* [dpdk-dev] [PATCH 00/10] net/i40e: base code update
@ 2017-05-27 3:47 Jingjing Wu
2017-05-27 3:47 ` [dpdk-dev] [PATCH 01/10] net/i40e/base: use new virtchnl header file Jingjing Wu
` (11 more replies)
0 siblings, 12 replies; 69+ messages in thread
From: Jingjing Wu @ 2017-05-27 3:47 UTC (permalink / raw)
To: helin.zhang; +Cc: dev, jingjing.wu
i40e base code upate. The main changes are:
- use virtchnl.h instead of i40e_virtchnl.h
- add support for Adaptive Virtual Function
- add new AQ commands for read/write PHY registers
- add new phy types for 25G
Jingjing Wu (10):
net/i40e/base: use new virtchnl header file
net/i40e/base: sync nvmupdate command and adminq subtask
net/i40e/base: add AQ command for read/write PHY registers
net/i40e/base: add support for Adaptive Virtual Function
net/i40e/base: store the requested FEC information
net/i40e/base: add new phy types for 25G AOC and ACC
net/i40e/base: report supported link modes
net/i40e/base: track id can be 0
net/i40e/base: update FW AQ API version to 1.7
net/i40e/base: update base code info
drivers/net/i40e/base/README | 2 +-
drivers/net/i40e/base/i40e_adminq_cmd.h | 42 +-
drivers/net/i40e/base/i40e_common.c | 126 +++++-
drivers/net/i40e/base/i40e_devids.h | 1 +
drivers/net/i40e/base/i40e_nvm.c | 6 +
drivers/net/i40e/base/i40e_prototype.h | 27 +-
drivers/net/i40e/base/i40e_type.h | 15 +-
drivers/net/i40e/base/i40e_virtchnl.h | 445 --------------------
drivers/net/i40e/base/virtchnl.h | 719 ++++++++++++++++++++++++++++++++
drivers/net/i40e/i40e_ethdev.h | 6 +-
drivers/net/i40e/i40e_ethdev_vf.c | 204 ++++-----
drivers/net/i40e/i40e_pf.c | 188 ++++-----
drivers/net/i40e/i40e_pf.h | 32 +-
drivers/net/i40e/rte_pmd_i40e.h | 2 +-
14 files changed, 1115 insertions(+), 700 deletions(-)
delete mode 100644 drivers/net/i40e/base/i40e_virtchnl.h
create mode 100644 drivers/net/i40e/base/virtchnl.h
--
2.4.11
^ permalink raw reply [flat|nested] 69+ messages in thread* [dpdk-dev] [PATCH 01/10] net/i40e/base: use new virtchnl header file 2017-05-27 3:47 [dpdk-dev] [PATCH 00/10] net/i40e: base code update Jingjing Wu @ 2017-05-27 3:47 ` Jingjing Wu 2017-05-27 3:47 ` [dpdk-dev] [PATCH 02/10] net/i40e/base: sync nvmupdate command and adminq subtask Jingjing Wu ` (10 subsequent siblings) 11 siblings, 0 replies; 69+ messages in thread From: Jingjing Wu @ 2017-05-27 3:47 UTC (permalink / raw) To: helin.zhang; +Cc: dev, jingjing.wu Modify the necessary files to be compatible with the new virtchnl.h file instead of relying on i40e_virtchnl.h variant. This mostly changes references to VIRTCHNL_ variables by removing prefix of I40E_. Signed-off-by: Jingjing Wu <jingjing.wu@intel.com> --- drivers/net/i40e/base/i40e_common.c | 24 +- drivers/net/i40e/base/i40e_prototype.h | 6 +- drivers/net/i40e/base/i40e_type.h | 12 +- drivers/net/i40e/base/i40e_virtchnl.h | 445 -------------------- drivers/net/i40e/base/virtchnl.h | 719 +++++++++++++++++++++++++++++++++ drivers/net/i40e/i40e_ethdev.h | 6 +- drivers/net/i40e/i40e_ethdev_vf.c | 204 +++++----- drivers/net/i40e/i40e_pf.c | 188 ++++----- drivers/net/i40e/i40e_pf.h | 32 +- drivers/net/i40e/rte_pmd_i40e.h | 2 +- 10 files changed, 956 insertions(+), 682 deletions(-) delete mode 100644 drivers/net/i40e/base/i40e_virtchnl.h create mode 100644 drivers/net/i40e/base/virtchnl.h diff --git a/drivers/net/i40e/base/i40e_common.c b/drivers/net/i40e/base/i40e_common.c index 03e94bc..6c09c27 100644 --- a/drivers/net/i40e/base/i40e_common.c +++ b/drivers/net/i40e/base/i40e_common.c @@ -34,7 +34,7 @@ POSSIBILITY OF SUCH DAMAGE. #include "i40e_type.h" #include "i40e_adminq.h" #include "i40e_prototype.h" -#include "i40e_virtchnl.h" +#include "virtchnl.h" /** @@ -5523,7 +5523,7 @@ enum i40e_status_code i40e_aq_add_rem_control_packet_filter(struct i40e_hw *hw, } if (mac_addr) - i40e_memcpy(cmd->mac, mac_addr, I40E_ETH_LENGTH_OF_ADDRESS, + i40e_memcpy(cmd->mac, mac_addr, ETH_ALEN, I40E_NONDMA_TO_NONDMA); cmd->etype = CPU_TO_LE16(ethtype); @@ -6879,7 +6879,7 @@ void i40e_write_rx_ctl(struct i40e_hw *hw, u32 reg_addr, u32 reg_val) * completion before returning. **/ enum i40e_status_code i40e_aq_send_msg_to_pf(struct i40e_hw *hw, - enum i40e_virtchnl_ops v_opcode, + enum virtchnl_ops v_opcode, enum i40e_status_code v_retval, u8 *msg, u16 msglen, struct i40e_asq_cmd_details *cmd_details) @@ -6918,9 +6918,9 @@ enum i40e_status_code i40e_aq_send_msg_to_pf(struct i40e_hw *hw, * with appropriate information. **/ void i40e_vf_parse_hw_config(struct i40e_hw *hw, - struct i40e_virtchnl_vf_resource *msg) + struct virtchnl_vf_resource *msg) { - struct i40e_virtchnl_vsi_resource *vsi_res; + struct virtchnl_vsi_resource *vsi_res; int i; vsi_res = &msg->vsi_res[0]; @@ -6930,19 +6930,17 @@ void i40e_vf_parse_hw_config(struct i40e_hw *hw, hw->dev_caps.num_tx_qp = msg->num_queue_pairs; hw->dev_caps.num_msix_vectors_vf = msg->max_vectors; hw->dev_caps.dcb = msg->vf_offload_flags & - I40E_VIRTCHNL_VF_OFFLOAD_L2; - hw->dev_caps.fcoe = (msg->vf_offload_flags & - I40E_VIRTCHNL_VF_OFFLOAD_FCOE) ? 1 : 0; + VIRTCHNL_VF_OFFLOAD_L2; hw->dev_caps.iwarp = (msg->vf_offload_flags & - I40E_VIRTCHNL_VF_OFFLOAD_IWARP) ? 1 : 0; + VIRTCHNL_VF_OFFLOAD_IWARP) ? 1 : 0; for (i = 0; i < msg->num_vsis; i++) { - if (vsi_res->vsi_type == I40E_VSI_SRIOV) { + if (vsi_res->vsi_type == VIRTCHNL_VSI_SRIOV) { i40e_memcpy(hw->mac.perm_addr, vsi_res->default_mac_addr, - I40E_ETH_LENGTH_OF_ADDRESS, + ETH_ALEN, I40E_NONDMA_TO_NONDMA); i40e_memcpy(hw->mac.addr, vsi_res->default_mac_addr, - I40E_ETH_LENGTH_OF_ADDRESS, + ETH_ALEN, I40E_NONDMA_TO_NONDMA); } vsi_res++; @@ -6959,7 +6957,7 @@ void i40e_vf_parse_hw_config(struct i40e_hw *hw, **/ enum i40e_status_code i40e_vf_reset(struct i40e_hw *hw) { - return i40e_aq_send_msg_to_pf(hw, I40E_VIRTCHNL_OP_RESET_VF, + return i40e_aq_send_msg_to_pf(hw, VIRTCHNL_OP_RESET_VF, I40E_SUCCESS, NULL, 0, NULL); } #endif /* VF_DRIVER */ diff --git a/drivers/net/i40e/base/i40e_prototype.h b/drivers/net/i40e/base/i40e_prototype.h index 4bd589e..6ec4304 100644 --- a/drivers/net/i40e/base/i40e_prototype.h +++ b/drivers/net/i40e/base/i40e_prototype.h @@ -36,7 +36,7 @@ POSSIBILITY OF SUCH DAMAGE. #include "i40e_type.h" #include "i40e_alloc.h" -#include "i40e_virtchnl.h" +#include "virtchnl.h" /* Prototypes for shared code functions that are not in * the standard function pointer structures. These are @@ -504,10 +504,10 @@ void i40e_destroy_spinlock(struct i40e_spinlock *sp); /* i40e_common for VF drivers*/ void i40e_vf_parse_hw_config(struct i40e_hw *hw, - struct i40e_virtchnl_vf_resource *msg); + struct virtchnl_vf_resource *msg); enum i40e_status_code i40e_vf_reset(struct i40e_hw *hw); enum i40e_status_code i40e_aq_send_msg_to_pf(struct i40e_hw *hw, - enum i40e_virtchnl_ops v_opcode, + enum virtchnl_ops v_opcode, enum i40e_status_code v_retval, u8 *msg, u16 msglen, struct i40e_asq_cmd_details *cmd_details); diff --git a/drivers/net/i40e/base/i40e_type.h b/drivers/net/i40e/base/i40e_type.h index 84d5757..52a114a 100644 --- a/drivers/net/i40e/base/i40e_type.h +++ b/drivers/net/i40e/base/i40e_type.h @@ -92,7 +92,9 @@ POSSIBILITY OF SUCH DAMAGE. struct i40e_hw; typedef void (*I40E_ADMINQ_CALLBACK)(struct i40e_hw *, struct i40e_aq_desc *); -#define I40E_ETH_LENGTH_OF_ADDRESS 6 +#ifndef ETH_ALEN +#define ETH_ALEN 6 +#endif /* Data type manipulation macros. */ #define I40E_HI_DWORD(x) ((u32)((((x) >> 16) >> 16) & 0xFFFFFFFF)) #define I40E_LO_DWORD(x) ((u32)((x) & 0xFFFFFFFF)) @@ -439,10 +441,10 @@ struct i40e_hw_capabilities { struct i40e_mac_info { enum i40e_mac_type type; - u8 addr[I40E_ETH_LENGTH_OF_ADDRESS]; - u8 perm_addr[I40E_ETH_LENGTH_OF_ADDRESS]; - u8 san_addr[I40E_ETH_LENGTH_OF_ADDRESS]; - u8 port_addr[I40E_ETH_LENGTH_OF_ADDRESS]; + u8 addr[ETH_ALEN]; + u8 perm_addr[ETH_ALEN]; + u8 san_addr[ETH_ALEN]; + u8 port_addr[ETH_ALEN]; u16 max_fcoeq; }; diff --git a/drivers/net/i40e/base/i40e_virtchnl.h b/drivers/net/i40e/base/i40e_virtchnl.h deleted file mode 100644 index 7a24c0f..0000000 --- a/drivers/net/i40e/base/i40e_virtchnl.h +++ /dev/null @@ -1,445 +0,0 @@ -/******************************************************************************* - -Copyright (c) 2013 - 2015, Intel Corporation -All rights reserved. - -Redistribution and use in source and binary forms, with or without -modification, are permitted provided that the following conditions are met: - - 1. Redistributions of source code must retain the above copyright notice, - this list of conditions and the following disclaimer. - - 2. Redistributions in binary form must reproduce the above copyright - notice, this list of conditions and the following disclaimer in the - documentation and/or other materials provided with the distribution. - - 3. Neither the name of the Intel Corporation nor the names of its - contributors may be used to endorse or promote products derived from - this software without specific prior written permission. - -THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS" -AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE -IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE -ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE -LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR -CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF -SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS -INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN -CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) -ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE -POSSIBILITY OF SUCH DAMAGE. - -***************************************************************************/ - -#ifndef _I40E_VIRTCHNL_H_ -#define _I40E_VIRTCHNL_H_ - -#include "i40e_type.h" - -/* Description: - * This header file describes the VF-PF communication protocol used - * by the various i40e drivers. - * - * Admin queue buffer usage: - * desc->opcode is always i40e_aqc_opc_send_msg_to_pf - * flags, retval, datalen, and data addr are all used normally. - * Firmware copies the cookie fields when sending messages between the PF and - * VF, but uses all other fields internally. Due to this limitation, we - * must send all messages as "indirect", i.e. using an external buffer. - * - * All the vsi indexes are relative to the VF. Each VF can have maximum of - * three VSIs. All the queue indexes are relative to the VSI. Each VF can - * have a maximum of sixteen queues for all of its VSIs. - * - * The PF is required to return a status code in v_retval for all messages - * except RESET_VF, which does not require any response. The return value is of - * i40e_status_code type, defined in the i40e_type.h. - * - * In general, VF driver initialization should roughly follow the order of these - * opcodes. The VF driver must first validate the API version of the PF driver, - * then request a reset, then get resources, then configure queues and - * interrupts. After these operations are complete, the VF driver may start - * its queues, optionally add MAC and VLAN filters, and process traffic. - */ - -/* Opcodes for VF-PF communication. These are placed in the v_opcode field - * of the virtchnl_msg structure. - */ -enum i40e_virtchnl_ops { -/* The PF sends status change events to VFs using - * the I40E_VIRTCHNL_OP_EVENT opcode. - * VFs send requests to the PF using the other ops. - */ - I40E_VIRTCHNL_OP_UNKNOWN = 0, - I40E_VIRTCHNL_OP_VERSION = 1, /* must ALWAYS be 1 */ - I40E_VIRTCHNL_OP_RESET_VF = 2, - I40E_VIRTCHNL_OP_GET_VF_RESOURCES = 3, - I40E_VIRTCHNL_OP_CONFIG_TX_QUEUE = 4, - I40E_VIRTCHNL_OP_CONFIG_RX_QUEUE = 5, - I40E_VIRTCHNL_OP_CONFIG_VSI_QUEUES = 6, - I40E_VIRTCHNL_OP_CONFIG_IRQ_MAP = 7, - I40E_VIRTCHNL_OP_ENABLE_QUEUES = 8, - I40E_VIRTCHNL_OP_DISABLE_QUEUES = 9, - I40E_VIRTCHNL_OP_ADD_ETHER_ADDRESS = 10, - I40E_VIRTCHNL_OP_DEL_ETHER_ADDRESS = 11, - I40E_VIRTCHNL_OP_ADD_VLAN = 12, - I40E_VIRTCHNL_OP_DEL_VLAN = 13, - I40E_VIRTCHNL_OP_CONFIG_PROMISCUOUS_MODE = 14, - I40E_VIRTCHNL_OP_GET_STATS = 15, - I40E_VIRTCHNL_OP_FCOE = 16, - I40E_VIRTCHNL_OP_EVENT = 17, /* must ALWAYS be 17 */ -#ifdef I40E_SOL_VF_SUPPORT - I40E_VIRTCHNL_OP_GET_ADDNL_SOL_CONFIG = 19, -#endif - I40E_VIRTCHNL_OP_CONFIG_RSS_KEY = 23, - I40E_VIRTCHNL_OP_CONFIG_RSS_LUT = 24, - I40E_VIRTCHNL_OP_GET_RSS_HENA_CAPS = 25, - I40E_VIRTCHNL_OP_SET_RSS_HENA = 26, - -}; - -/* Virtual channel message descriptor. This overlays the admin queue - * descriptor. All other data is passed in external buffers. - */ - -struct i40e_virtchnl_msg { - u8 pad[8]; /* AQ flags/opcode/len/retval fields */ - enum i40e_virtchnl_ops v_opcode; /* avoid confusion with desc->opcode */ - enum i40e_status_code v_retval; /* ditto for desc->retval */ - u32 vfid; /* used by PF when sending to VF */ -}; - -/* Message descriptions and data structures.*/ - -/* I40E_VIRTCHNL_OP_VERSION - * VF posts its version number to the PF. PF responds with its version number - * in the same format, along with a return code. - * Reply from PF has its major/minor versions also in param0 and param1. - * If there is a major version mismatch, then the VF cannot operate. - * If there is a minor version mismatch, then the VF can operate but should - * add a warning to the system log. - * - * This enum element MUST always be specified as == 1, regardless of other - * changes in the API. The PF must always respond to this message without - * error regardless of version mismatch. - */ -#define I40E_VIRTCHNL_VERSION_MAJOR 1 -#define I40E_VIRTCHNL_VERSION_MINOR 1 -#define I40E_VIRTCHNL_VERSION_MINOR_NO_VF_CAPS 0 - -struct i40e_virtchnl_version_info { - u32 major; - u32 minor; -}; - -/* I40E_VIRTCHNL_OP_RESET_VF - * VF sends this request to PF with no parameters - * PF does NOT respond! VF driver must delay then poll VFGEN_RSTAT register - * until reset completion is indicated. The admin queue must be reinitialized - * after this operation. - * - * When reset is complete, PF must ensure that all queues in all VSIs associated - * with the VF are stopped, all queue configurations in the HMC are set to 0, - * and all MAC and VLAN filters (except the default MAC address) on all VSIs - * are cleared. - */ - -/* I40E_VIRTCHNL_OP_GET_VF_RESOURCES - * Version 1.0 VF sends this request to PF with no parameters - * Version 1.1 VF sends this request to PF with u32 bitmap of its capabilities - * PF responds with an indirect message containing - * i40e_virtchnl_vf_resource and one or more - * i40e_virtchnl_vsi_resource structures. - */ - -struct i40e_virtchnl_vsi_resource { - u16 vsi_id; - u16 num_queue_pairs; - enum i40e_vsi_type vsi_type; - u16 qset_handle; - u8 default_mac_addr[I40E_ETH_LENGTH_OF_ADDRESS]; -}; -/* VF offload flags */ -#define I40E_VIRTCHNL_VF_OFFLOAD_L2 0x00000001 -#define I40E_VIRTCHNL_VF_OFFLOAD_IWARP 0x00000002 -#define I40E_VIRTCHNL_VF_OFFLOAD_FCOE 0x00000004 -#define I40E_VIRTCHNL_VF_OFFLOAD_RSS_AQ 0x00000008 -#define I40E_VIRTCHNL_VF_OFFLOAD_RSS_REG 0x00000010 -#define I40E_VIRTCHNL_VF_OFFLOAD_WB_ON_ITR 0x00000020 -#define I40E_VIRTCHNL_VF_OFFLOAD_VLAN 0x00010000 -#define I40E_VIRTCHNL_VF_OFFLOAD_RX_POLLING 0x00020000 -#define I40E_VIRTCHNL_VF_OFFLOAD_RSS_PCTYPE_V2 0x00040000 -#define I40E_VIRTCHNL_VF_OFFLOAD_RSS_PF 0X00080000 -#define I40E_VIRTCHNL_VF_OFFLOAD_ENCAP 0X00100000 -#define I40E_VIRTCHNL_VF_OFFLOAD_ENCAP_CSUM 0X00200000 -#define I40E_VIRTCHNL_VF_OFFLOAD_RX_ENCAP_CSUM 0X00400000 - -#define I40E_VF_BASE_MODE_OFFLOADS (I40E_VIRTCHNL_VF_OFFLOAD_L2 | \ - I40E_VIRTCHNL_VF_OFFLOAD_VLAN | \ - I40E_VIRTCHNL_VF_OFFLOAD_RSS_PF) - -struct i40e_virtchnl_vf_resource { - u16 num_vsis; - u16 num_queue_pairs; - u16 max_vectors; - u16 max_mtu; - - u32 vf_offload_flags; - u32 rss_key_size; - u32 rss_lut_size; - - struct i40e_virtchnl_vsi_resource vsi_res[1]; -}; - -/* I40E_VIRTCHNL_OP_CONFIG_TX_QUEUE - * VF sends this message to set up parameters for one TX queue. - * External data buffer contains one instance of i40e_virtchnl_txq_info. - * PF configures requested queue and returns a status code. - */ - -/* Tx queue config info */ -struct i40e_virtchnl_txq_info { - u16 vsi_id; - u16 queue_id; - u16 ring_len; /* number of descriptors, multiple of 8 */ - u16 headwb_enabled; - u64 dma_ring_addr; - u64 dma_headwb_addr; -}; - -/* I40E_VIRTCHNL_OP_CONFIG_RX_QUEUE - * VF sends this message to set up parameters for one RX queue. - * External data buffer contains one instance of i40e_virtchnl_rxq_info. - * PF configures requested queue and returns a status code. - */ - -/* Rx queue config info */ -struct i40e_virtchnl_rxq_info { - u16 vsi_id; - u16 queue_id; - u32 ring_len; /* number of descriptors, multiple of 32 */ - u16 hdr_size; - u16 splithdr_enabled; - u32 databuffer_size; - u32 max_pkt_size; - u64 dma_ring_addr; - enum i40e_hmc_obj_rx_hsplit_0 rx_split_pos; -}; - -/* I40E_VIRTCHNL_OP_CONFIG_VSI_QUEUES - * VF sends this message to set parameters for all active TX and RX queues - * associated with the specified VSI. - * PF configures queues and returns status. - * If the number of queues specified is greater than the number of queues - * associated with the VSI, an error is returned and no queues are configured. - */ -struct i40e_virtchnl_queue_pair_info { - /* NOTE: vsi_id and queue_id should be identical for both queues. */ - struct i40e_virtchnl_txq_info txq; - struct i40e_virtchnl_rxq_info rxq; -}; - -struct i40e_virtchnl_vsi_queue_config_info { - u16 vsi_id; - u16 num_queue_pairs; - struct i40e_virtchnl_queue_pair_info qpair[1]; -}; - -/* I40E_VIRTCHNL_OP_CONFIG_IRQ_MAP - * VF uses this message to map vectors to queues. - * The rxq_map and txq_map fields are bitmaps used to indicate which queues - * are to be associated with the specified vector. - * The "other" causes are always mapped to vector 0. - * PF configures interrupt mapping and returns status. - */ -struct i40e_virtchnl_vector_map { - u16 vsi_id; - u16 vector_id; - u16 rxq_map; - u16 txq_map; - u16 rxitr_idx; - u16 txitr_idx; -}; - -struct i40e_virtchnl_irq_map_info { - u16 num_vectors; - struct i40e_virtchnl_vector_map vecmap[1]; -}; - -/* I40E_VIRTCHNL_OP_ENABLE_QUEUES - * I40E_VIRTCHNL_OP_DISABLE_QUEUES - * VF sends these message to enable or disable TX/RX queue pairs. - * The queues fields are bitmaps indicating which queues to act upon. - * (Currently, we only support 16 queues per VF, but we make the field - * u32 to allow for expansion.) - * PF performs requested action and returns status. - */ -struct i40e_virtchnl_queue_select { - u16 vsi_id; - u16 pad; - u32 rx_queues; - u32 tx_queues; -}; - -/* I40E_VIRTCHNL_OP_ADD_ETHER_ADDRESS - * VF sends this message in order to add one or more unicast or multicast - * address filters for the specified VSI. - * PF adds the filters and returns status. - */ - -/* I40E_VIRTCHNL_OP_DEL_ETHER_ADDRESS - * VF sends this message in order to remove one or more unicast or multicast - * filters for the specified VSI. - * PF removes the filters and returns status. - */ - -struct i40e_virtchnl_ether_addr { - u8 addr[I40E_ETH_LENGTH_OF_ADDRESS]; - u8 pad[2]; -}; - -struct i40e_virtchnl_ether_addr_list { - u16 vsi_id; - u16 num_elements; - struct i40e_virtchnl_ether_addr list[1]; -}; - -#ifdef I40E_SOL_VF_SUPPORT -/* I40E_VIRTCHNL_OP_GET_ADDNL_SOL_CONFIG - * VF sends this message to get the default MTU and list of additional ethernet - * addresses it is allowed to use. - * PF responds with an indirect message containing - * i40e_virtchnl_addnl_solaris_config with zero or more - * i40e_virtchnl_ether_addr structures. - * - * It is expected that this operation will only ever be needed for Solaris VFs - * running under a Solaris PF. - */ -struct i40e_virtchnl_addnl_solaris_config { - u16 default_mtu; - struct i40e_virtchnl_ether_addr_list al; -}; - -#endif -/* I40E_VIRTCHNL_OP_ADD_VLAN - * VF sends this message to add one or more VLAN tag filters for receives. - * PF adds the filters and returns status. - * If a port VLAN is configured by the PF, this operation will return an - * error to the VF. - */ - -/* I40E_VIRTCHNL_OP_DEL_VLAN - * VF sends this message to remove one or more VLAN tag filters for receives. - * PF removes the filters and returns status. - * If a port VLAN is configured by the PF, this operation will return an - * error to the VF. - */ - -struct i40e_virtchnl_vlan_filter_list { - u16 vsi_id; - u16 num_elements; - u16 vlan_id[1]; -}; - -/* I40E_VIRTCHNL_OP_CONFIG_PROMISCUOUS_MODE - * VF sends VSI id and flags. - * PF returns status code in retval. - * Note: we assume that broadcast accept mode is always enabled. - */ -struct i40e_virtchnl_promisc_info { - u16 vsi_id; - u16 flags; -}; - -#define I40E_FLAG_VF_UNICAST_PROMISC 0x00000001 -#define I40E_FLAG_VF_MULTICAST_PROMISC 0x00000002 - -/* I40E_VIRTCHNL_OP_GET_STATS - * VF sends this message to request stats for the selected VSI. VF uses - * the i40e_virtchnl_queue_select struct to specify the VSI. The queue_id - * field is ignored by the PF. - * - * PF replies with struct i40e_eth_stats in an external buffer. - */ - -/* I40E_VIRTCHNL_OP_CONFIG_RSS_KEY - * I40E_VIRTCHNL_OP_CONFIG_RSS_LUT - * VF sends these messages to configure RSS. Only supported if both PF - * and VF drivers set the I40E_VIRTCHNL_VF_OFFLOAD_RSS_PF bit during - * configuration negotiation. If this is the case, then the rss fields in - * the vf resource struct are valid. - * Both the key and LUT are initialized to 0 by the PF, meaning that - * RSS is effectively disabled until set up by the VF. - */ -struct i40e_virtchnl_rss_key { - u16 vsi_id; - u16 key_len; - u8 key[1]; /* RSS hash key, packed bytes */ -}; - -struct i40e_virtchnl_rss_lut { - u16 vsi_id; - u16 lut_entries; - u8 lut[1]; /* RSS lookup table*/ -}; - -/* I40E_VIRTCHNL_OP_GET_RSS_HENA_CAPS - * I40E_VIRTCHNL_OP_SET_RSS_HENA - * VF sends these messages to get and set the hash filter enable bits for RSS. - * By default, the PF sets these to all possible traffic types that the - * hardware supports. The VF can query this value if it wants to change the - * traffic types that are hashed by the hardware. - * Traffic types are defined in the i40e_filter_pctype enum in i40e_type.h - */ -struct i40e_virtchnl_rss_hena { - u64 hena; -}; - -/* I40E_VIRTCHNL_OP_EVENT - * PF sends this message to inform the VF driver of events that may affect it. - * No direct response is expected from the VF, though it may generate other - * messages in response to this one. - */ -enum i40e_virtchnl_event_codes { - I40E_VIRTCHNL_EVENT_UNKNOWN = 0, - I40E_VIRTCHNL_EVENT_LINK_CHANGE, - I40E_VIRTCHNL_EVENT_RESET_IMPENDING, - I40E_VIRTCHNL_EVENT_PF_DRIVER_CLOSE, -}; -#define I40E_PF_EVENT_SEVERITY_INFO 0 -#define I40E_PF_EVENT_SEVERITY_ATTENTION 1 -#define I40E_PF_EVENT_SEVERITY_ACTION_REQUIRED 2 -#define I40E_PF_EVENT_SEVERITY_CERTAIN_DOOM 255 - -struct i40e_virtchnl_pf_event { - enum i40e_virtchnl_event_codes event; - union { - struct { - enum i40e_aq_link_speed link_speed; - bool link_status; - } link_event; - } event_data; - - int severity; -}; - -/* VF reset states - these are written into the RSTAT register: - * I40E_VFGEN_RSTAT1 on the PF - * I40E_VFGEN_RSTAT on the VF - * When the PF initiates a reset, it writes 0 - * When the reset is complete, it writes 1 - * When the PF detects that the VF has recovered, it writes 2 - * VF checks this register periodically to determine if a reset has occurred, - * then polls it to know when the reset is complete. - * If either the PF or VF reads the register while the hardware - * is in a reset state, it will return DEADBEEF, which, when masked - * will result in 3. - */ -enum i40e_vfr_states { - I40E_VFR_INPROGRESS = 0, - I40E_VFR_COMPLETED, - I40E_VFR_VFACTIVE, - I40E_VFR_UNKNOWN, -}; - -#endif /* _I40E_VIRTCHNL_H_ */ diff --git a/drivers/net/i40e/base/virtchnl.h b/drivers/net/i40e/base/virtchnl.h new file mode 100644 index 0000000..3fbc517 --- /dev/null +++ b/drivers/net/i40e/base/virtchnl.h @@ -0,0 +1,719 @@ +/******************************************************************************* + +Copyright (c) 2013 - 2015, Intel Corporation +All rights reserved. + +Redistribution and use in source and binary forms, with or without +modification, are permitted provided that the following conditions are met: + + 1. Redistributions of source code must retain the above copyright notice, + this list of conditions and the following disclaimer. + + 2. Redistributions in binary form must reproduce the above copyright + notice, this list of conditions and the following disclaimer in the + documentation and/or other materials provided with the distribution. + + 3. Neither the name of the Intel Corporation nor the names of its + contributors may be used to endorse or promote products derived from + this software without specific prior written permission. + +THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS" +AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE +IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE +ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE +LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR +CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF +SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS +INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN +CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) +ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE +POSSIBILITY OF SUCH DAMAGE. + +***************************************************************************/ + +#ifndef _VIRTCHNL_H_ +#define _VIRTCHNL_H_ + +/* Description: + * This header file describes the VF-PF communication protocol used + * by the drivers for all devices starting from our 40G product line + * + * Admin queue buffer usage: + * desc->opcode is always aqc_opc_send_msg_to_pf + * flags, retval, datalen, and data addr are all used normally. + * The Firmware copies the cookie fields when sending messages between the + * PF and VF, but uses all other fields internally. Due to this limitation, + * we must send all messages as "indirect", i.e. using an external buffer. + * + * All the vsi indexes are relative to the VF. Each VF can have maximum of + * three VSIs. All the queue indexes are relative to the VSI. Each VF can + * have a maximum of sixteen queues for all of its VSIs. + * + * The PF is required to return a status code in v_retval for all messages + * except RESET_VF, which does not require any response. The return value + * is of status_code type, defined in the shared type.h. + * + * In general, VF driver initialization should roughly follow the order of + * these opcodes. The VF driver must first validate the API version of the + * PF driver, then request a reset, then get resources, then configure + * queues and interrupts. After these operations are complete, the VF + * driver may start its queues, optionally add MAC and VLAN filters, and + * process traffic. + */ + +/* START GENERIC DEFINES + * Need to ensure the following enums and defines hold the same meaning and + * value in CPK and other projects + */ + +/* Error Codes */ +enum virtchnl_status_code { + VIRTCHNL_STATUS_SUCCESS = 0, + VIRTCHNL_ERR_PARAM = -5, + VIRTCHNL_STATUS_ERR_OPCODE_MISMATCH = -38, + VIRTCHNL_STATUS_ERR_CQP_COMPL_ERROR = -39, + VIRTCHNL_STATUS_ERR_INVALID_VF_ID = -40, + VIRTCHNL_STATUS_NOT_SUPPORTED = -64, +}; + +#define VIRTCHNL_LINK_SPEED_100MB_SHIFT 0x1 +#define VIRTCHNL_LINK_SPEED_1000MB_SHIFT 0x2 +#define VIRTCHNL_LINK_SPEED_10GB_SHIFT 0x3 +#define VIRTCHNL_LINK_SPEED_40GB_SHIFT 0x4 +#define VIRTCHNL_LINK_SPEED_20GB_SHIFT 0x5 +#define VIRTCHNL_LINK_SPEED_25GB_SHIFT 0x6 + +enum virtchnl_link_speed { + VIRTCHNL_LINK_SPEED_UNKNOWN = 0, + VIRTCHNL_LINK_SPEED_100MB = (1 << VIRTCHNL_LINK_SPEED_100MB_SHIFT), + VIRTCHNL_LINK_SPEED_1GB = (1 << VIRTCHNL_LINK_SPEED_1000MB_SHIFT), + VIRTCHNL_LINK_SPEED_10GB = (1 << VIRTCHNL_LINK_SPEED_10GB_SHIFT), + VIRTCHNL_LINK_SPEED_40GB = (1 << VIRTCHNL_LINK_SPEED_40GB_SHIFT), + VIRTCHNL_LINK_SPEED_20GB = (1 << VIRTCHNL_LINK_SPEED_20GB_SHIFT), + VIRTCHNL_LINK_SPEED_25GB = (1 << VIRTCHNL_LINK_SPEED_25GB_SHIFT), +}; + +/* for hsplit_0 field of Rx HMC context */ +/* deprecated with AVF 1.0 */ +enum virtchnl_rx_hsplit { + VIRTCHNL_RX_HSPLIT_NO_SPLIT = 0, + VIRTCHNL_RX_HSPLIT_SPLIT_L2 = 1, + VIRTCHNL_RX_HSPLIT_SPLIT_IP = 2, + VIRTCHNL_RX_HSPLIT_SPLIT_TCP_UDP = 4, + VIRTCHNL_RX_HSPLIT_SPLIT_SCTP = 8, +}; + +#define VIRTCHNL_ETH_LENGTH_OF_ADDRESS 6 + +/* END GENERIC DEFINES */ + +/* Opcodes for VF-PF communication. These are placed in the v_opcode field + * of the virtchnl_msg structure. + */ +enum virtchnl_ops { +/* The PF sends status change events to VFs using + * the VIRTCHNL_OP_EVENT opcode. + * VFs send requests to the PF using the other ops. + * Use of "advanced opcode" features must be negotiated as part of capabilities + * exchange and are not considered part of base mode feature set. + */ + VIRTCHNL_OP_UNKNOWN = 0, + VIRTCHNL_OP_VERSION = 1, /* must ALWAYS be 1 */ + VIRTCHNL_OP_RESET_VF = 2, + VIRTCHNL_OP_GET_VF_RESOURCES = 3, + VIRTCHNL_OP_CONFIG_TX_QUEUE = 4, + VIRTCHNL_OP_CONFIG_RX_QUEUE = 5, + VIRTCHNL_OP_CONFIG_VSI_QUEUES = 6, + VIRTCHNL_OP_CONFIG_IRQ_MAP = 7, + VIRTCHNL_OP_ENABLE_QUEUES = 8, + VIRTCHNL_OP_DISABLE_QUEUES = 9, + VIRTCHNL_OP_ADD_ETH_ADDR = 10, + VIRTCHNL_OP_DEL_ETH_ADDR = 11, + VIRTCHNL_OP_ADD_VLAN = 12, + VIRTCHNL_OP_DEL_VLAN = 13, + VIRTCHNL_OP_CONFIG_PROMISCUOUS_MODE = 14, /* advanced opcode */ + VIRTCHNL_OP_GET_STATS = 15, + VIRTCHNL_OP_RSVD = 16, + VIRTCHNL_OP_EVENT = 17, /* must ALWAYS be 17 */ +#ifdef VIRTCHNL_SOL_VF_SUPPORT + VIRTCHNL_OP_GET_ADDNL_SOL_CONFIG = 19, +#endif +#ifdef VIRTCHNL_IWARP + VIRTCHNL_OP_IWARP = 20, /* advanced opcode */ + VIRTCHNL_OP_CONFIG_IWARP_IRQ_MAP = 21, /* advanced opcode */ + VIRTCHNL_OP_RELEASE_IWARP_IRQ_MAP = 22, /* advanced opcode */ +#endif + VIRTCHNL_OP_CONFIG_RSS_KEY = 23, + VIRTCHNL_OP_CONFIG_RSS_LUT = 24, + VIRTCHNL_OP_GET_RSS_HENA_CAPS = 25, + VIRTCHNL_OP_SET_RSS_HENA = 26, + VIRTCHNL_OP_ENABLE_VLAN_STRIPPING = 27, + VIRTCHNL_OP_DISABLE_VLAN_STRIPPING = 28, + VIRTCHNL_OP_REQUEST_QUEUES = 29, + +}; + +/* Virtual channel message descriptor. This overlays the admin queue + * descriptor. All other data is passed in external buffers. + */ + +struct virtchnl_msg { + u8 pad[8]; /* AQ flags/opcode/len/retval fields */ + enum virtchnl_ops v_opcode; /* avoid confusion with desc->opcode */ + enum virtchnl_status_code v_retval; /* ditto for desc->retval */ + u32 vfid; /* used by PF when sending to VF */ +}; + +/* Message descriptions and data structures.*/ + +/* VIRTCHNL_OP_VERSION + * VF posts its version number to the PF. PF responds with its version number + * in the same format, along with a return code. + * Reply from PF has its major/minor versions also in param0 and param1. + * If there is a major version mismatch, then the VF cannot operate. + * If there is a minor version mismatch, then the VF can operate but should + * add a warning to the system log. + * + * This enum element MUST always be specified as == 1, regardless of other + * changes in the API. The PF must always respond to this message without + * error regardless of version mismatch. + */ +#define VIRTCHNL_VERSION_MAJOR 1 +#define VIRTCHNL_VERSION_MINOR 1 +#define VIRTCHNL_VERSION_MINOR_NO_VF_CAPS 0 + +struct virtchnl_version_info { + u32 major; + u32 minor; +}; + +#define VF_IS_V10(_v) (((_v)->major == 1) && ((_v)->minor == 0)) +#define VF_IS_V11(_ver) (((_ver)->major == 1) && ((_ver)->minor == 1)) + +/* VIRTCHNL_OP_RESET_VF + * VF sends this request to PF with no parameters + * PF does NOT respond! VF driver must delay then poll VFGEN_RSTAT register + * until reset completion is indicated. The admin queue must be reinitialized + * after this operation. + * + * When reset is complete, PF must ensure that all queues in all VSIs associated + * with the VF are stopped, all queue configurations in the HMC are set to 0, + * and all MAC and VLAN filters (except the default MAC address) on all VSIs + * are cleared. + */ + +/* VSI types that use VIRTCHNL interface for VF-PF communication. VSI_SRIOV + * vsi_type should always be 6 for backward compatibility. Add other fields + * as needed. + */ +enum virtchnl_vsi_type { + VIRTCHNL_VSI_TYPE_INVALID = 0, + VIRTCHNL_VSI_SRIOV = 6, +}; + +/* VIRTCHNL_OP_GET_VF_RESOURCES + * Version 1.0 VF sends this request to PF with no parameters + * Version 1.1 VF sends this request to PF with u32 bitmap of its capabilities + * PF responds with an indirect message containing + * virtchnl_vf_resource and one or more + * virtchnl_vsi_resource structures. + */ + +struct virtchnl_vsi_resource { + u16 vsi_id; + u16 num_queue_pairs; + enum virtchnl_vsi_type vsi_type; + u16 qset_handle; + u8 default_mac_addr[VIRTCHNL_ETH_LENGTH_OF_ADDRESS]; +}; + +/* VF offload flags + * VIRTCHNL_VF_OFFLOAD_L2 flag is inclusive of base mode L2 offloads including + * TX/RX Checksum offloading and TSO for non-tunnelled packets. + */ +#define VIRTCHNL_VF_OFFLOAD_L2 0x00000001 +#define VIRTCHNL_VF_OFFLOAD_IWARP 0x00000002 +#define VIRTCHNL_VF_OFFLOAD_RSVD 0x00000004 +#define VIRTCHNL_VF_OFFLOAD_RSS_AQ 0x00000008 +#define VIRTCHNL_VF_OFFLOAD_RSS_REG 0x00000010 +#define VIRTCHNL_VF_OFFLOAD_WB_ON_ITR 0x00000020 +#define VIRTCHNL_VF_OFFLOAD_REQ_QUEUES 0x00000040 +#define VIRTCHNL_VF_OFFLOAD_VLAN 0x00010000 +#define VIRTCHNL_VF_OFFLOAD_RX_POLLING 0x00020000 +#define VIRTCHNL_VF_OFFLOAD_RSS_PCTYPE_V2 0x00040000 +#define VIRTCHNL_VF_OFFLOAD_RSS_PF 0X00080000 +#define VIRTCHNL_VF_OFFLOAD_ENCAP 0X00100000 +#define VIRTCHNL_VF_OFFLOAD_ENCAP_CSUM 0X00200000 +#define VIRTCHNL_VF_OFFLOAD_RX_ENCAP_CSUM 0X00400000 + +#define VF_BASE_MODE_OFFLOADS (VIRTCHNL_VF_OFFLOAD_L2 | \ + VIRTCHNL_VF_OFFLOAD_VLAN | \ + VIRTCHNL_VF_OFFLOAD_RSS_PF) + +struct virtchnl_vf_resource { + u16 num_vsis; + u16 num_queue_pairs; + u16 max_vectors; + u16 max_mtu; + + u32 vf_offload_flags; + u32 rss_key_size; + u32 rss_lut_size; + + struct virtchnl_vsi_resource vsi_res[1]; +}; + +/* VIRTCHNL_OP_CONFIG_TX_QUEUE + * VF sends this message to set up parameters for one TX queue. + * External data buffer contains one instance of virtchnl_txq_info. + * PF configures requested queue and returns a status code. + */ + +/* Tx queue config info */ +struct virtchnl_txq_info { + u16 vsi_id; + u16 queue_id; + u16 ring_len; /* number of descriptors, multiple of 8 */ + u16 headwb_enabled; /* deprecated with AVF 1.0 */ + u64 dma_ring_addr; + u64 dma_headwb_addr; /* deprecated with AVF 1.0 */ +}; + +/* VIRTCHNL_OP_CONFIG_RX_QUEUE + * VF sends this message to set up parameters for one RX queue. + * External data buffer contains one instance of virtchnl_rxq_info. + * PF configures requested queue and returns a status code. + */ + +/* Rx queue config info */ +struct virtchnl_rxq_info { + u16 vsi_id; + u16 queue_id; + u32 ring_len; /* number of descriptors, multiple of 32 */ + u16 hdr_size; + u16 splithdr_enabled; /* deprecated with AVF 1.0 */ + u32 databuffer_size; + u32 max_pkt_size; + u64 dma_ring_addr; + enum virtchnl_rx_hsplit rx_split_pos; /* deprecated with AVF 1.0 */ +}; + +/* VIRTCHNL_OP_CONFIG_VSI_QUEUES + * VF sends this message to set parameters for all active TX and RX queues + * associated with the specified VSI. + * PF configures queues and returns status. + * If the number of queues specified is greater than the number of queues + * associated with the VSI, an error is returned and no queues are configured. + */ +struct virtchnl_queue_pair_info { + /* NOTE: vsi_id and queue_id should be identical for both queues. */ + struct virtchnl_txq_info txq; + struct virtchnl_rxq_info rxq; +}; + +struct virtchnl_vsi_queue_config_info { + u16 vsi_id; + u16 num_queue_pairs; + struct virtchnl_queue_pair_info qpair[1]; +}; + +/* VIRTCHNL_OP_REQUEST_QUEUES + * VF sends this message to request the PF to allocate additional queues to + * this VF. Each VF gets a guaranteed number of queues on init but asking for + * additional queues must be negotiated. This is a best effort request as it + * is possible the PF does not have enough queues left to support the request. + * If the PF cannot support the number requested it will respond with the + * maximum number it is able to support; otherwise it will respond with the + * number requested. + */ + +/* VF resource request */ +struct virtchnl_vf_res_request { + u16 num_queue_pairs; +}; + +/* VIRTCHNL_OP_CONFIG_IRQ_MAP + * VF uses this message to map vectors to queues. + * The rxq_map and txq_map fields are bitmaps used to indicate which queues + * are to be associated with the specified vector. + * The "other" causes are always mapped to vector 0. + * PF configures interrupt mapping and returns status. + */ +struct virtchnl_vector_map { + u16 vsi_id; + u16 vector_id; + u16 rxq_map; + u16 txq_map; + u16 rxitr_idx; + u16 txitr_idx; +}; + +struct virtchnl_irq_map_info { + u16 num_vectors; + struct virtchnl_vector_map vecmap[1]; +}; + +/* VIRTCHNL_OP_ENABLE_QUEUES + * VIRTCHNL_OP_DISABLE_QUEUES + * VF sends these message to enable or disable TX/RX queue pairs. + * The queues fields are bitmaps indicating which queues to act upon. + * (Currently, we only support 16 queues per VF, but we make the field + * u32 to allow for expansion.) + * PF performs requested action and returns status. + */ +struct virtchnl_queue_select { + u16 vsi_id; + u16 pad; + u32 rx_queues; + u32 tx_queues; +}; + +/* VIRTCHNL_OP_ADD_ETH_ADDR + * VF sends this message in order to add one or more unicast or multicast + * address filters for the specified VSI. + * PF adds the filters and returns status. + */ + +/* VIRTCHNL_OP_DEL_ETH_ADDR + * VF sends this message in order to remove one or more unicast or multicast + * filters for the specified VSI. + * PF removes the filters and returns status. + */ + +struct virtchnl_ether_addr { + u8 addr[VIRTCHNL_ETH_LENGTH_OF_ADDRESS]; + u8 pad[2]; +}; + +struct virtchnl_ether_addr_list { + u16 vsi_id; + u16 num_elements; + struct virtchnl_ether_addr list[1]; +}; + +#ifdef VIRTCHNL_SOL_VF_SUPPORT +/* VIRTCHNL_OP_GET_ADDNL_SOL_CONFIG + * VF sends this message to get the default MTU and list of additional ethernet + * addresses it is allowed to use. + * PF responds with an indirect message containing + * virtchnl_addnl_solaris_config with zero or more + * virtchnl_ether_addr structures. + * + * It is expected that this operation will only ever be needed for Solaris VFs + * running under a Solaris PF. + */ +struct virtchnl_addnl_solaris_config { + u16 default_mtu; + struct virtchnl_ether_addr_list al; +}; + +#endif +/* VIRTCHNL_OP_ADD_VLAN + * VF sends this message to add one or more VLAN tag filters for receives. + * PF adds the filters and returns status. + * If a port VLAN is configured by the PF, this operation will return an + * error to the VF. + */ + +/* VIRTCHNL_OP_DEL_VLAN + * VF sends this message to remove one or more VLAN tag filters for receives. + * PF removes the filters and returns status. + * If a port VLAN is configured by the PF, this operation will return an + * error to the VF. + */ + +struct virtchnl_vlan_filter_list { + u16 vsi_id; + u16 num_elements; + u16 vlan_id[1]; +}; + +/* VIRTCHNL_OP_CONFIG_PROMISCUOUS_MODE + * VF sends VSI id and flags. + * PF returns status code in retval. + * Note: we assume that broadcast accept mode is always enabled. + */ +struct virtchnl_promisc_info { + u16 vsi_id; + u16 flags; +}; + +#define FLAG_VF_UNICAST_PROMISC 0x00000001 +#define FLAG_VF_MULTICAST_PROMISC 0x00000002 + +/* VIRTCHNL_OP_GET_STATS + * VF sends this message to request stats for the selected VSI. VF uses + * the virtchnl_queue_select struct to specify the VSI. The queue_id + * field is ignored by the PF. + * + * PF replies with struct eth_stats in an external buffer. + */ + +/* VIRTCHNL_OP_CONFIG_RSS_KEY + * VIRTCHNL_OP_CONFIG_RSS_LUT + * VF sends these messages to configure RSS. Only supported if both PF + * and VF drivers set the VIRTCHNL_VF_OFFLOAD_RSS_PF bit during + * configuration negotiation. If this is the case, then the rss fields in + * the vf resource struct are valid. + * Both the key and LUT are initialized to 0 by the PF, meaning that + * RSS is effectively disabled until set up by the VF. + */ +struct virtchnl_rss_key { + u16 vsi_id; + u16 key_len; + u8 key[1]; /* RSS hash key, packed bytes */ +}; + +struct virtchnl_rss_lut { + u16 vsi_id; + u16 lut_entries; + u8 lut[1]; /* RSS lookup table*/ +}; + +/* VIRTCHNL_OP_GET_RSS_HENA_CAPS + * VIRTCHNL_OP_SET_RSS_HENA + * VF sends these messages to get and set the hash filter enable bits for RSS. + * By default, the PF sets these to all possible traffic types that the + * hardware supports. The VF can query this value if it wants to change the + * traffic types that are hashed by the hardware. + */ +struct virtchnl_rss_hena { + u64 hena; +}; + +/* VIRTCHNL_OP_EVENT + * PF sends this message to inform the VF driver of events that may affect it. + * No direct response is expected from the VF, though it may generate other + * messages in response to this one. + */ +enum virtchnl_event_codes { + VIRTCHNL_EVENT_UNKNOWN = 0, + VIRTCHNL_EVENT_LINK_CHANGE, + VIRTCHNL_EVENT_RESET_IMPENDING, + VIRTCHNL_EVENT_PF_DRIVER_CLOSE, +}; + +#define PF_EVENT_SEVERITY_INFO 0 +#define PF_EVENT_SEVERITY_ATTENTION 1 +#define PF_EVENT_SEVERITY_ACTION_REQUIRED 2 +#define PF_EVENT_SEVERITY_CERTAIN_DOOM 255 + +struct virtchnl_pf_event { + enum virtchnl_event_codes event; + union { + struct { + enum virtchnl_link_speed link_speed; + bool link_status; + } link_event; + } event_data; + + int severity; +}; +#ifdef VIRTCHNL_IWARP + +/* VIRTCHNL_OP_CONFIG_IWARP_IRQ_MAP + * VF uses this message to request PF to map IWARP vectors to IWARP queues. + * The request for this originates from the VF IWARP driver through + * a client interface between VF LAN and VF IWARP driver. + * A vector could have an AEQ and CEQ attached to it although + * there is a single AEQ per VF IWARP instance in which case + * most vectors will have an INVALID_IDX for aeq and valid idx for ceq. + * There will never be a case where there will be multiple CEQs attached + * to a single vector. + * PF configures interrupt mapping and returns status. + */ + +/* HW does not define a type value for AEQ; only for RX/TX and CEQ. + * In order for us to keep the interface simple, SW will define a + * unique type value for AEQ. +*/ +#define QUEUE_TYPE_PE_AEQ 0x80 +#define QUEUE_INVALID_IDX 0xFFFF + +struct virtchnl_iwarp_qv_info { + u32 v_idx; /* msix_vector */ + u16 ceq_idx; + u16 aeq_idx; + u8 itr_idx; +}; + +struct virtchnl_iwarp_qvlist_info { + u32 num_vectors; + struct virtchnl_iwarp_qv_info qv_info[1]; +}; +#endif + +/* VF reset states - these are written into the RSTAT register: + * VFGEN_RSTAT on the VF + * When the PF initiates a reset, it writes 0 + * When the reset is complete, it writes 1 + * When the PF detects that the VF has recovered, it writes 2 + * VF checks this register periodically to determine if a reset has occurred, + * then polls it to know when the reset is complete. + * If either the PF or VF reads the register while the hardware + * is in a reset state, it will return DEADBEEF, which, when masked + * will result in 3. + */ +enum virtchnl_vfr_states { + VIRTCHNL_VFR_INPROGRESS = 0, + VIRTCHNL_VFR_COMPLETED, + VIRTCHNL_VFR_VFACTIVE, +}; + +/** + * virtchnl_vc_validate_vf_msg + * @ver: Virtchnl version info + * @v_opcode: Opcode for the message + * @msg: pointer to the msg buffer + * @msglen: msg length + * + * validate msg format against struct for each opcode + */ +static inline int +virtchnl_vc_validate_vf_msg(struct virtchnl_version_info *ver, u32 v_opcode, + u8 *msg, u16 msglen) +{ + bool err_msg_format = false; + int valid_len = 0; + + /* Validate message length. */ + switch (v_opcode) { + case VIRTCHNL_OP_VERSION: + valid_len = sizeof(struct virtchnl_version_info); + break; + case VIRTCHNL_OP_RESET_VF: + break; + case VIRTCHNL_OP_GET_VF_RESOURCES: + if (VF_IS_V11(ver)) + valid_len = sizeof(u32); + break; + case VIRTCHNL_OP_CONFIG_TX_QUEUE: + valid_len = sizeof(struct virtchnl_txq_info); + break; + case VIRTCHNL_OP_CONFIG_RX_QUEUE: + valid_len = sizeof(struct virtchnl_rxq_info); + break; + case VIRTCHNL_OP_CONFIG_VSI_QUEUES: + valid_len = sizeof(struct virtchnl_vsi_queue_config_info); + if (msglen >= valid_len) { + struct virtchnl_vsi_queue_config_info *vqc = + (struct virtchnl_vsi_queue_config_info *)msg; + valid_len += (vqc->num_queue_pairs * + sizeof(struct + virtchnl_queue_pair_info)); + if (vqc->num_queue_pairs == 0) + err_msg_format = true; + } + break; + case VIRTCHNL_OP_CONFIG_IRQ_MAP: + valid_len = sizeof(struct virtchnl_irq_map_info); + if (msglen >= valid_len) { + struct virtchnl_irq_map_info *vimi = + (struct virtchnl_irq_map_info *)msg; + valid_len += (vimi->num_vectors * + sizeof(struct virtchnl_vector_map)); + if (vimi->num_vectors == 0) + err_msg_format = true; + } + break; + case VIRTCHNL_OP_ENABLE_QUEUES: + case VIRTCHNL_OP_DISABLE_QUEUES: + valid_len = sizeof(struct virtchnl_queue_select); + break; + case VIRTCHNL_OP_ADD_ETH_ADDR: + case VIRTCHNL_OP_DEL_ETH_ADDR: + valid_len = sizeof(struct virtchnl_ether_addr_list); + if (msglen >= valid_len) { + struct virtchnl_ether_addr_list *veal = + (struct virtchnl_ether_addr_list *)msg; + valid_len += veal->num_elements * + sizeof(struct virtchnl_ether_addr); + if (veal->num_elements == 0) + err_msg_format = true; + } + break; + case VIRTCHNL_OP_ADD_VLAN: + case VIRTCHNL_OP_DEL_VLAN: + valid_len = sizeof(struct virtchnl_vlan_filter_list); + if (msglen >= valid_len) { + struct virtchnl_vlan_filter_list *vfl = + (struct virtchnl_vlan_filter_list *)msg; + valid_len += vfl->num_elements * sizeof(u16); + if (vfl->num_elements == 0) + err_msg_format = true; + } + break; + case VIRTCHNL_OP_CONFIG_PROMISCUOUS_MODE: + valid_len = sizeof(struct virtchnl_promisc_info); + break; + case VIRTCHNL_OP_GET_STATS: + valid_len = sizeof(struct virtchnl_queue_select); + break; +#ifdef VIRTCHNL_IWARP + case VIRTCHNL_OP_IWARP: + /* These messages are opaque to us and will be validated in + * the RDMA client code. We just need to check for nonzero + * length. The firmware will enforce max length restrictions. + */ + if (msglen) + valid_len = msglen; + else + err_msg_format = true; + break; + case VIRTCHNL_OP_RELEASE_IWARP_IRQ_MAP: + break; + case VIRTCHNL_OP_CONFIG_IWARP_IRQ_MAP: + valid_len = sizeof(struct virtchnl_iwarp_qvlist_info); + if (msglen >= valid_len) { + struct virtchnl_iwarp_qvlist_info *qv = + (struct virtchnl_iwarp_qvlist_info *)msg; + if (qv->num_vectors == 0) { + err_msg_format = true; + break; + } + valid_len += ((qv->num_vectors - 1) * + sizeof(struct virtchnl_iwarp_qv_info)); + } + break; +#endif + case VIRTCHNL_OP_CONFIG_RSS_KEY: + valid_len = sizeof(struct virtchnl_rss_key); + if (msglen >= valid_len) { + struct virtchnl_rss_key *vrk = + (struct virtchnl_rss_key *)msg; + valid_len += vrk->key_len - 1; + } + break; + case VIRTCHNL_OP_CONFIG_RSS_LUT: + valid_len = sizeof(struct virtchnl_rss_lut); + if (msglen >= valid_len) { + struct virtchnl_rss_lut *vrl = + (struct virtchnl_rss_lut *)msg; + valid_len += vrl->lut_entries - 1; + } + break; + case VIRTCHNL_OP_GET_RSS_HENA_CAPS: + break; + case VIRTCHNL_OP_SET_RSS_HENA: + valid_len = sizeof(struct virtchnl_rss_hena); + break; + case VIRTCHNL_OP_ENABLE_VLAN_STRIPPING: + case VIRTCHNL_OP_DISABLE_VLAN_STRIPPING: + break; + case VIRTCHNL_OP_REQUEST_QUEUES: + valid_len = sizeof(struct virtchnl_vf_res_request); + break; + /* These are always errors coming from the VF. */ + case VIRTCHNL_OP_EVENT: + case VIRTCHNL_OP_UNKNOWN: + default: + return VIRTCHNL_ERR_PARAM; + } + + if ((valid_len != msglen) || (err_msg_format)) + return VIRTCHNL_STATUS_ERR_OPCODE_MISMATCH; + else + return 0; +} + +#endif /* _VIRTCHNL_H_ */ diff --git a/drivers/net/i40e/i40e_ethdev.h b/drivers/net/i40e/i40e_ethdev.h index 2ff8282..25abbd4 100644 --- a/drivers/net/i40e/i40e_ethdev.h +++ b/drivers/net/i40e/i40e_ethdev.h @@ -102,7 +102,7 @@ /* Linux PF host with virtchnl version 1.1 */ #define PF_IS_V11(vf) \ - (((vf)->version_major == I40E_VIRTCHNL_VERSION_MAJOR) && \ + (((vf)->version_major == VIRTCHNL_VERSION_MAJOR) && \ ((vf)->version_minor == 1)) /* index flex payload per layer */ @@ -745,8 +745,8 @@ struct i40e_vf { uint8_t *aq_resp; /* buffer to store the adminq response from PF */ /* VSI info */ - struct i40e_virtchnl_vf_resource *vf_res; /* All VSIs */ - struct i40e_virtchnl_vsi_resource *vsi_res; /* LAN VSI */ + struct virtchnl_vf_resource *vf_res; /* All VSIs */ + struct virtchnl_vsi_resource *vsi_res; /* LAN VSI */ struct i40e_vsi vsi; uint64_t flags; }; diff --git a/drivers/net/i40e/i40e_ethdev_vf.c b/drivers/net/i40e/i40e_ethdev_vf.c index 859b5e8..754051e 100644 --- a/drivers/net/i40e/i40e_ethdev_vf.c +++ b/drivers/net/i40e/i40e_ethdev_vf.c @@ -77,7 +77,7 @@ #define MAX_RESET_WAIT_CNT 20 struct i40evf_arq_msg_info { - enum i40e_virtchnl_ops ops; + enum virtchnl_ops ops; enum i40e_status_code result; uint16_t buf_len; uint16_t msg_len; @@ -85,7 +85,7 @@ struct i40evf_arq_msg_info { }; struct vf_cmd_info { - enum i40e_virtchnl_ops ops; + enum virtchnl_ops ops; uint8_t *in_args; uint32_t in_args_size; uint8_t *out_buffer; @@ -244,7 +244,7 @@ i40evf_read_pfmsg(struct rte_eth_dev *dev, struct i40evf_arq_msg_info *data) struct i40e_hw *hw = I40E_DEV_PRIVATE_TO_HW(dev->data->dev_private); struct i40e_vf *vf = I40EVF_DEV_PRIVATE_TO_VF(dev->data->dev_private); struct i40e_arq_event_info event; - enum i40e_virtchnl_ops opcode; + enum virtchnl_ops opcode; enum i40e_status_code retval; int ret; enum i40evf_aq_result result = I40EVF_MSG_NON; @@ -259,16 +259,16 @@ i40evf_read_pfmsg(struct rte_eth_dev *dev, struct i40evf_arq_msg_info *data) return result; } - opcode = (enum i40e_virtchnl_ops)rte_le_to_cpu_32(event.desc.cookie_high); + opcode = (enum virtchnl_ops)rte_le_to_cpu_32(event.desc.cookie_high); retval = (enum i40e_status_code)rte_le_to_cpu_32(event.desc.cookie_low); /* pf sys event */ - if (opcode == I40E_VIRTCHNL_OP_EVENT) { - struct i40e_virtchnl_pf_event *vpe = - (struct i40e_virtchnl_pf_event *)event.msg_buf; + if (opcode == VIRTCHNL_OP_EVENT) { + struct virtchnl_pf_event *vpe = + (struct virtchnl_pf_event *)event.msg_buf; result = I40EVF_MSG_SYS; switch (vpe->event) { - case I40E_VIRTCHNL_EVENT_LINK_CHANGE: + case VIRTCHNL_EVENT_LINK_CHANGE: vf->link_up = vpe->event_data.link_event.link_status; vf->link_speed = @@ -277,12 +277,12 @@ i40evf_read_pfmsg(struct rte_eth_dev *dev, struct i40evf_arq_msg_info *data) PMD_DRV_LOG(INFO, "Link status update:%s", vf->link_up ? "up" : "down"); break; - case I40E_VIRTCHNL_EVENT_RESET_IMPENDING: + case VIRTCHNL_EVENT_RESET_IMPENDING: vf->vf_reset = true; vf->pend_msg |= PFMSG_RESET_IMPENDING; PMD_DRV_LOG(INFO, "vf is reseting"); break; - case I40E_VIRTCHNL_EVENT_PF_DRIVER_CLOSE: + case VIRTCHNL_EVENT_PF_DRIVER_CLOSE: vf->dev_closed = true; vf->pend_msg |= PFMSG_DRIVER_CLOSE; PMD_DRV_LOG(INFO, "PF driver closed"); @@ -312,17 +312,17 @@ static inline void _clear_cmd(struct i40e_vf *vf) { rte_wmb(); - vf->pend_cmd = I40E_VIRTCHNL_OP_UNKNOWN; + vf->pend_cmd = VIRTCHNL_OP_UNKNOWN; } /* * Check there is pending cmd in execution. If none, set new command. */ static inline int -_atomic_set_cmd(struct i40e_vf *vf, enum i40e_virtchnl_ops ops) +_atomic_set_cmd(struct i40e_vf *vf, enum virtchnl_ops ops) { int ret = rte_atomic32_cmpset(&vf->pend_cmd, - I40E_VIRTCHNL_OP_UNKNOWN, ops); + VIRTCHNL_OP_UNKNOWN, ops); if (!ret) PMD_DRV_LOG(ERR, "There is incomplete cmd %d", vf->pend_cmd); @@ -347,7 +347,7 @@ i40evf_execute_vf_cmd(struct rte_eth_dev *dev, struct vf_cmd_info *args) info.msg = args->out_buffer; info.buf_len = args->out_size; - info.ops = I40E_VIRTCHNL_OP_UNKNOWN; + info.ops = VIRTCHNL_OP_UNKNOWN; info.result = I40E_SUCCESS; err = i40e_aq_send_msg_to_pf(hw, args->ops, I40E_SUCCESS, @@ -359,12 +359,12 @@ i40evf_execute_vf_cmd(struct rte_eth_dev *dev, struct vf_cmd_info *args) } switch (args->ops) { - case I40E_VIRTCHNL_OP_RESET_VF: + case VIRTCHNL_OP_RESET_VF: /*no need to process in this function */ err = 0; break; - case I40E_VIRTCHNL_OP_VERSION: - case I40E_VIRTCHNL_OP_GET_VF_RESOURCES: + case VIRTCHNL_OP_VERSION: + case VIRTCHNL_OP_GET_VF_RESOURCES: /* for init adminq commands, need to poll the response */ err = -1; do { @@ -385,7 +385,7 @@ i40evf_execute_vf_cmd(struct rte_eth_dev *dev, struct vf_cmd_info *args) /* for other adminq in running time, waiting the cmd done flag */ err = -1; do { - if (vf->pend_cmd == I40E_VIRTCHNL_OP_UNKNOWN) { + if (vf->pend_cmd == VIRTCHNL_OP_UNKNOWN) { err = 0; break; } @@ -404,15 +404,15 @@ i40evf_execute_vf_cmd(struct rte_eth_dev *dev, struct vf_cmd_info *args) static int i40evf_check_api_version(struct rte_eth_dev *dev) { - struct i40e_virtchnl_version_info version, *pver; + struct virtchnl_version_info version, *pver; int err; struct vf_cmd_info args; struct i40e_vf *vf = I40EVF_DEV_PRIVATE_TO_VF(dev->data->dev_private); - version.major = I40E_VIRTCHNL_VERSION_MAJOR; - version.minor = I40E_VIRTCHNL_VERSION_MINOR; + version.major = VIRTCHNL_VERSION_MAJOR; + version.minor = VIRTCHNL_VERSION_MINOR; - args.ops = I40E_VIRTCHNL_OP_VERSION; + args.ops = VIRTCHNL_OP_VERSION; args.in_args = (uint8_t *)&version; args.in_args_size = sizeof(version); args.out_buffer = vf->aq_resp; @@ -424,19 +424,19 @@ i40evf_check_api_version(struct rte_eth_dev *dev) return err; } - pver = (struct i40e_virtchnl_version_info *)args.out_buffer; + pver = (struct virtchnl_version_info *)args.out_buffer; vf->version_major = pver->major; vf->version_minor = pver->minor; if (vf->version_major == I40E_DPDK_VERSION_MAJOR) PMD_DRV_LOG(INFO, "Peer is DPDK PF host"); - else if ((vf->version_major == I40E_VIRTCHNL_VERSION_MAJOR) && - (vf->version_minor <= I40E_VIRTCHNL_VERSION_MINOR)) + else if ((vf->version_major == VIRTCHNL_VERSION_MAJOR) && + (vf->version_minor <= VIRTCHNL_VERSION_MINOR)) PMD_DRV_LOG(INFO, "Peer is Linux PF host"); else { PMD_INIT_LOG(ERR, "PF/VF API version mismatch:(%u.%u)-(%u.%u)", vf->version_major, vf->version_minor, - I40E_VIRTCHNL_VERSION_MAJOR, - I40E_VIRTCHNL_VERSION_MINOR); + VIRTCHNL_VERSION_MAJOR, + VIRTCHNL_VERSION_MINOR); return -1; } @@ -452,15 +452,15 @@ i40evf_get_vf_resource(struct rte_eth_dev *dev) struct vf_cmd_info args; uint32_t caps, len; - args.ops = I40E_VIRTCHNL_OP_GET_VF_RESOURCES; + args.ops = VIRTCHNL_OP_GET_VF_RESOURCES; args.out_buffer = vf->aq_resp; args.out_size = I40E_AQ_BUF_SZ; if (PF_IS_V11(vf)) { - caps = I40E_VIRTCHNL_VF_OFFLOAD_L2 | - I40E_VIRTCHNL_VF_OFFLOAD_RSS_AQ | - I40E_VIRTCHNL_VF_OFFLOAD_RSS_REG | - I40E_VIRTCHNL_VF_OFFLOAD_VLAN | - I40E_VIRTCHNL_VF_OFFLOAD_RX_POLLING; + caps = VIRTCHNL_VF_OFFLOAD_L2 | + VIRTCHNL_VF_OFFLOAD_RSS_AQ | + VIRTCHNL_VF_OFFLOAD_RSS_REG | + VIRTCHNL_VF_OFFLOAD_VLAN | + VIRTCHNL_VF_OFFLOAD_RX_POLLING; args.in_args = (uint8_t *)∩︀ args.in_args_size = sizeof(caps); } else { @@ -474,8 +474,8 @@ i40evf_get_vf_resource(struct rte_eth_dev *dev) return err; } - len = sizeof(struct i40e_virtchnl_vf_resource) + - I40E_MAX_VF_VSI * sizeof(struct i40e_virtchnl_vsi_resource); + len = sizeof(struct virtchnl_vf_resource) + + I40E_MAX_VF_VSI * sizeof(struct virtchnl_vsi_resource); (void)rte_memcpy(vf->vf_res, args.out_buffer, RTE_MIN(args.out_size, len)); @@ -492,18 +492,18 @@ i40evf_config_promisc(struct rte_eth_dev *dev, struct i40e_vf *vf = I40EVF_DEV_PRIVATE_TO_VF(dev->data->dev_private); int err; struct vf_cmd_info args; - struct i40e_virtchnl_promisc_info promisc; + struct virtchnl_promisc_info promisc; promisc.flags = 0; promisc.vsi_id = vf->vsi_res->vsi_id; if (enable_unicast) - promisc.flags |= I40E_FLAG_VF_UNICAST_PROMISC; + promisc.flags |= FLAG_VF_UNICAST_PROMISC; if (enable_multicast) - promisc.flags |= I40E_FLAG_VF_MULTICAST_PROMISC; + promisc.flags |= FLAG_VF_MULTICAST_PROMISC; - args.ops = I40E_VIRTCHNL_OP_CONFIG_PROMISCUOUS_MODE; + args.ops = VIRTCHNL_OP_CONFIG_PROMISCUOUS_MODE; args.in_args = (uint8_t *)&promisc; args.in_args_size = sizeof(promisc); args.out_buffer = vf->aq_resp; @@ -525,12 +525,12 @@ i40evf_config_vlan_offload(struct rte_eth_dev *dev, struct i40e_vf *vf = I40EVF_DEV_PRIVATE_TO_VF(dev->data->dev_private); int err; struct vf_cmd_info args; - struct i40e_virtchnl_vlan_offload_info offload; + struct virtchnl_vlan_offload_info offload; offload.vsi_id = vf->vsi_res->vsi_id; offload.enable_vlan_strip = enable_vlan_strip; - args.ops = (enum i40e_virtchnl_ops)I40E_VIRTCHNL_OP_CFG_VLAN_OFFLOAD; + args.ops = (enum virtchnl_ops)I40E_VIRTCHNL_OP_CFG_VLAN_OFFLOAD; args.in_args = (uint8_t *)&offload; args.in_args_size = sizeof(offload); args.out_buffer = vf->aq_resp; @@ -550,7 +550,7 @@ i40evf_config_vlan_pvid(struct rte_eth_dev *dev, struct i40e_vf *vf = I40EVF_DEV_PRIVATE_TO_VF(dev->data->dev_private); int err; struct vf_cmd_info args; - struct i40e_virtchnl_pvid_info tpid_info; + struct virtchnl_pvid_info tpid_info; if (info == NULL) { PMD_DRV_LOG(ERR, "invalid parameters"); @@ -561,7 +561,7 @@ i40evf_config_vlan_pvid(struct rte_eth_dev *dev, tpid_info.vsi_id = vf->vsi_res->vsi_id; (void)rte_memcpy(&tpid_info.info, info, sizeof(*info)); - args.ops = (enum i40e_virtchnl_ops)I40E_VIRTCHNL_OP_CFG_VLAN_PVID; + args.ops = (enum virtchnl_ops)I40E_VIRTCHNL_OP_CFG_VLAN_PVID; args.in_args = (uint8_t *)&tpid_info; args.in_args_size = sizeof(tpid_info); args.out_buffer = vf->aq_resp; @@ -575,7 +575,7 @@ i40evf_config_vlan_pvid(struct rte_eth_dev *dev, } static void -i40evf_fill_virtchnl_vsi_txq_info(struct i40e_virtchnl_txq_info *txq_info, +i40evf_fill_virtchnl_vsi_txq_info(struct virtchnl_txq_info *txq_info, uint16_t vsi_id, uint16_t queue_id, uint16_t nb_txq, @@ -590,7 +590,7 @@ i40evf_fill_virtchnl_vsi_txq_info(struct i40e_virtchnl_txq_info *txq_info, } static void -i40evf_fill_virtchnl_vsi_rxq_info(struct i40e_virtchnl_rxq_info *rxq_info, +i40evf_fill_virtchnl_vsi_rxq_info(struct virtchnl_rxq_info *rxq_info, uint16_t vsi_id, uint16_t queue_id, uint16_t nb_rxq, @@ -618,8 +618,8 @@ i40evf_configure_vsi_queues(struct rte_eth_dev *dev) (struct i40e_rx_queue **)dev->data->rx_queues; struct i40e_tx_queue **txq = (struct i40e_tx_queue **)dev->data->tx_queues; - struct i40e_virtchnl_vsi_queue_config_info *vc_vqci; - struct i40e_virtchnl_queue_pair_info *vc_qpi; + struct virtchnl_vsi_queue_config_info *vc_vqci; + struct virtchnl_queue_pair_info *vc_qpi; struct vf_cmd_info args; uint16_t i, nb_qp = vf->num_queue_pairs; const uint32_t size = @@ -628,7 +628,7 @@ i40evf_configure_vsi_queues(struct rte_eth_dev *dev) int ret; memset(buff, 0, sizeof(buff)); - vc_vqci = (struct i40e_virtchnl_vsi_queue_config_info *)buff; + vc_vqci = (struct virtchnl_vsi_queue_config_info *)buff; vc_vqci->vsi_id = vf->vsi_res->vsi_id; vc_vqci->num_queue_pairs = nb_qp; @@ -640,7 +640,7 @@ i40evf_configure_vsi_queues(struct rte_eth_dev *dev) vf->max_pkt_len, rxq[i]); } memset(&args, 0, sizeof(args)); - args.ops = I40E_VIRTCHNL_OP_CONFIG_VSI_QUEUES; + args.ops = VIRTCHNL_OP_CONFIG_VSI_QUEUES; args.in_args = (uint8_t *)vc_vqci; args.in_args_size = size; args.out_buffer = vf->aq_resp; @@ -648,7 +648,7 @@ i40evf_configure_vsi_queues(struct rte_eth_dev *dev) ret = i40evf_execute_vf_cmd(dev, &args); if (ret) PMD_DRV_LOG(ERR, "Failed to execute command of " - "I40E_VIRTCHNL_OP_CONFIG_VSI_QUEUES"); + "VIRTCHNL_OP_CONFIG_VSI_QUEUES"); return ret; } @@ -662,8 +662,8 @@ i40evf_configure_vsi_queues_ext(struct rte_eth_dev *dev) (struct i40e_rx_queue **)dev->data->rx_queues; struct i40e_tx_queue **txq = (struct i40e_tx_queue **)dev->data->tx_queues; - struct i40e_virtchnl_vsi_queue_config_ext_info *vc_vqcei; - struct i40e_virtchnl_queue_pair_ext_info *vc_qpei; + struct virtchnl_vsi_queue_config_ext_info *vc_vqcei; + struct virtchnl_queue_pair_ext_info *vc_qpei; struct vf_cmd_info args; uint16_t i, nb_qp = vf->num_queue_pairs; const uint32_t size = @@ -672,7 +672,7 @@ i40evf_configure_vsi_queues_ext(struct rte_eth_dev *dev) int ret; memset(buff, 0, sizeof(buff)); - vc_vqcei = (struct i40e_virtchnl_vsi_queue_config_ext_info *)buff; + vc_vqcei = (struct virtchnl_vsi_queue_config_ext_info *)buff; vc_vqcei->vsi_id = vf->vsi_res->vsi_id; vc_vqcei->num_queue_pairs = nb_qp; vc_qpei = vc_vqcei->qpair; @@ -693,7 +693,7 @@ i40evf_configure_vsi_queues_ext(struct rte_eth_dev *dev) } memset(&args, 0, sizeof(args)); args.ops = - (enum i40e_virtchnl_ops)I40E_VIRTCHNL_OP_CONFIG_VSI_QUEUES_EXT; + (enum virtchnl_ops)VIRTCHNL_OP_CONFIG_VSI_QUEUES_EXT; args.in_args = (uint8_t *)vc_vqcei; args.in_args_size = size; args.out_buffer = vf->aq_resp; @@ -701,7 +701,7 @@ i40evf_configure_vsi_queues_ext(struct rte_eth_dev *dev) ret = i40evf_execute_vf_cmd(dev, &args); if (ret) PMD_DRV_LOG(ERR, "Failed to execute command of " - "I40E_VIRTCHNL_OP_CONFIG_VSI_QUEUES_EXT"); + "VIRTCHNL_OP_CONFIG_VSI_QUEUES_EXT"); return ret; } @@ -724,9 +724,9 @@ i40evf_config_irq_map(struct rte_eth_dev *dev) { struct i40e_vf *vf = I40EVF_DEV_PRIVATE_TO_VF(dev->data->dev_private); struct vf_cmd_info args; - uint8_t cmd_buffer[sizeof(struct i40e_virtchnl_irq_map_info) + \ - sizeof(struct i40e_virtchnl_vector_map)]; - struct i40e_virtchnl_irq_map_info *map_info; + uint8_t cmd_buffer[sizeof(struct virtchnl_irq_map_info) + \ + sizeof(struct virtchnl_vector_map)]; + struct virtchnl_irq_map_info *map_info; struct rte_pci_device *pci_dev = I40E_DEV_TO_PCI(dev); struct rte_intr_handle *intr_handle = &pci_dev->intr_handle; uint32_t vector_id; @@ -741,7 +741,7 @@ i40evf_config_irq_map(struct rte_eth_dev *dev) vector_id = I40E_MISC_VEC_ID; } - map_info = (struct i40e_virtchnl_irq_map_info *)cmd_buffer; + map_info = (struct virtchnl_irq_map_info *)cmd_buffer; map_info->num_vectors = 1; map_info->vecmap[0].rxitr_idx = I40E_ITR_INDEX_DEFAULT; map_info->vecmap[0].vsi_id = vf->vsi_res->vsi_id; @@ -756,7 +756,7 @@ i40evf_config_irq_map(struct rte_eth_dev *dev) intr_handle->intr_vec[i] = vector_id; } - args.ops = I40E_VIRTCHNL_OP_CONFIG_IRQ_MAP; + args.ops = VIRTCHNL_OP_CONFIG_IRQ_MAP; args.in_args = (u8 *)cmd_buffer; args.in_args_size = sizeof(cmd_buffer); args.out_buffer = vf->aq_resp; @@ -773,7 +773,7 @@ i40evf_switch_queue(struct rte_eth_dev *dev, bool isrx, uint16_t qid, bool on) { struct i40e_vf *vf = I40EVF_DEV_PRIVATE_TO_VF(dev->data->dev_private); - struct i40e_virtchnl_queue_select queue_select; + struct virtchnl_queue_select queue_select; int err; struct vf_cmd_info args; memset(&queue_select, 0, sizeof(queue_select)); @@ -785,9 +785,9 @@ i40evf_switch_queue(struct rte_eth_dev *dev, bool isrx, uint16_t qid, queue_select.tx_queues |= 1 << qid; if (on) - args.ops = I40E_VIRTCHNL_OP_ENABLE_QUEUES; + args.ops = VIRTCHNL_OP_ENABLE_QUEUES; else - args.ops = I40E_VIRTCHNL_OP_DISABLE_QUEUES; + args.ops = VIRTCHNL_OP_DISABLE_QUEUES; args.in_args = (u8 *)&queue_select; args.in_args_size = sizeof(queue_select); args.out_buffer = vf->aq_resp; @@ -861,10 +861,10 @@ i40evf_add_mac_addr(struct rte_eth_dev *dev, __rte_unused uint32_t index, __rte_unused uint32_t pool) { - struct i40e_virtchnl_ether_addr_list *list; + struct virtchnl_ether_addr_list *list; struct i40e_vf *vf = I40EVF_DEV_PRIVATE_TO_VF(dev->data->dev_private); - uint8_t cmd_buffer[sizeof(struct i40e_virtchnl_ether_addr_list) + \ - sizeof(struct i40e_virtchnl_ether_addr)]; + uint8_t cmd_buffer[sizeof(struct virtchnl_ether_addr_list) + \ + sizeof(struct virtchnl_ether_addr)]; int err; struct vf_cmd_info args; @@ -876,13 +876,13 @@ i40evf_add_mac_addr(struct rte_eth_dev *dev, return I40E_ERR_INVALID_MAC_ADDR; } - list = (struct i40e_virtchnl_ether_addr_list *)cmd_buffer; + list = (struct virtchnl_ether_addr_list *)cmd_buffer; list->vsi_id = vf->vsi_res->vsi_id; list->num_elements = 1; (void)rte_memcpy(list->list[0].addr, addr->addr_bytes, sizeof(addr->addr_bytes)); - args.ops = I40E_VIRTCHNL_OP_ADD_ETHER_ADDRESS; + args.ops = VIRTCHNL_OP_ADD_ETH_ADDR; args.in_args = cmd_buffer; args.in_args_size = sizeof(cmd_buffer); args.out_buffer = vf->aq_resp; @@ -899,10 +899,10 @@ static void i40evf_del_mac_addr_by_addr(struct rte_eth_dev *dev, struct ether_addr *addr) { - struct i40e_virtchnl_ether_addr_list *list; + struct virtchnl_ether_addr_list *list; struct i40e_vf *vf = I40EVF_DEV_PRIVATE_TO_VF(dev->data->dev_private); - uint8_t cmd_buffer[sizeof(struct i40e_virtchnl_ether_addr_list) + \ - sizeof(struct i40e_virtchnl_ether_addr)]; + uint8_t cmd_buffer[sizeof(struct virtchnl_ether_addr_list) + \ + sizeof(struct virtchnl_ether_addr)]; int err; struct vf_cmd_info args; @@ -914,13 +914,13 @@ i40evf_del_mac_addr_by_addr(struct rte_eth_dev *dev, return; } - list = (struct i40e_virtchnl_ether_addr_list *)cmd_buffer; + list = (struct virtchnl_ether_addr_list *)cmd_buffer; list->vsi_id = vf->vsi_res->vsi_id; list->num_elements = 1; (void)rte_memcpy(list->list[0].addr, addr->addr_bytes, sizeof(addr->addr_bytes)); - args.ops = I40E_VIRTCHNL_OP_DEL_ETHER_ADDRESS; + args.ops = VIRTCHNL_OP_DEL_ETH_ADDR; args.in_args = cmd_buffer; args.in_args_size = sizeof(cmd_buffer); args.out_buffer = vf->aq_resp; @@ -947,13 +947,13 @@ static int i40evf_update_stats(struct rte_eth_dev *dev, struct i40e_eth_stats **pstats) { struct i40e_vf *vf = I40EVF_DEV_PRIVATE_TO_VF(dev->data->dev_private); - struct i40e_virtchnl_queue_select q_stats; + struct virtchnl_queue_select q_stats; int err; struct vf_cmd_info args; memset(&q_stats, 0, sizeof(q_stats)); q_stats.vsi_id = vf->vsi_res->vsi_id; - args.ops = I40E_VIRTCHNL_OP_GET_STATS; + args.ops = VIRTCHNL_OP_GET_STATS; args.in_args = (u8 *)&q_stats; args.in_args_size = sizeof(q_stats); args.out_buffer = vf->aq_resp; @@ -1050,18 +1050,18 @@ static int i40evf_add_vlan(struct rte_eth_dev *dev, uint16_t vlanid) { struct i40e_vf *vf = I40EVF_DEV_PRIVATE_TO_VF(dev->data->dev_private); - struct i40e_virtchnl_vlan_filter_list *vlan_list; - uint8_t cmd_buffer[sizeof(struct i40e_virtchnl_vlan_filter_list) + + struct virtchnl_vlan_filter_list *vlan_list; + uint8_t cmd_buffer[sizeof(struct virtchnl_vlan_filter_list) + sizeof(uint16_t)]; int err; struct vf_cmd_info args; - vlan_list = (struct i40e_virtchnl_vlan_filter_list *)cmd_buffer; + vlan_list = (struct virtchnl_vlan_filter_list *)cmd_buffer; vlan_list->vsi_id = vf->vsi_res->vsi_id; vlan_list->num_elements = 1; vlan_list->vlan_id[0] = vlanid; - args.ops = I40E_VIRTCHNL_OP_ADD_VLAN; + args.ops = VIRTCHNL_OP_ADD_VLAN; args.in_args = (u8 *)&cmd_buffer; args.in_args_size = sizeof(cmd_buffer); args.out_buffer = vf->aq_resp; @@ -1077,18 +1077,18 @@ static int i40evf_del_vlan(struct rte_eth_dev *dev, uint16_t vlanid) { struct i40e_vf *vf = I40EVF_DEV_PRIVATE_TO_VF(dev->data->dev_private); - struct i40e_virtchnl_vlan_filter_list *vlan_list; - uint8_t cmd_buffer[sizeof(struct i40e_virtchnl_vlan_filter_list) + + struct virtchnl_vlan_filter_list *vlan_list; + uint8_t cmd_buffer[sizeof(struct virtchnl_vlan_filter_list) + sizeof(uint16_t)]; int err; struct vf_cmd_info args; - vlan_list = (struct i40e_virtchnl_vlan_filter_list *)cmd_buffer; + vlan_list = (struct virtchnl_vlan_filter_list *)cmd_buffer; vlan_list->vsi_id = vf->vsi_res->vsi_id; vlan_list->num_elements = 1; vlan_list->vlan_id[0] = vlanid; - args.ops = I40E_VIRTCHNL_OP_DEL_VLAN; + args.ops = VIRTCHNL_OP_DEL_VLAN; args.in_args = (u8 *)&cmd_buffer; args.in_args_size = sizeof(cmd_buffer); args.out_buffer = vf->aq_resp; @@ -1178,7 +1178,7 @@ i40evf_reset_vf(struct i40e_hw *hw) reset = rd32(hw, I40E_VFGEN_RSTAT) & I40E_VFGEN_RSTAT_VFR_STATE_MASK; reset = reset >> I40E_VFGEN_RSTAT_VFR_STATE_SHIFT; - if (I40E_VFR_COMPLETED == reset || I40E_VFR_VFACTIVE == reset) + if (VIRTCHNL_VFR_COMPLETED == reset || VIRTCHNL_VFR_VFACTIVE == reset) break; else rte_delay_ms(50); @@ -1242,8 +1242,8 @@ i40evf_init_vf(struct rte_eth_dev *dev) PMD_INIT_LOG(ERR, "check_api version failed"); goto err_aq; } - bufsz = sizeof(struct i40e_virtchnl_vf_resource) + - (I40E_MAX_VF_VSI * sizeof(struct i40e_virtchnl_vsi_resource)); + bufsz = sizeof(struct virtchnl_vf_resource) + + (I40E_MAX_VF_VSI * sizeof(struct virtchnl_vsi_resource)); vf->vf_res = rte_zmalloc("vf_res", bufsz, 0); if (!vf->vf_res) { PMD_INIT_LOG(ERR, "unable to allocate vf_res memory"); @@ -1257,7 +1257,7 @@ i40evf_init_vf(struct rte_eth_dev *dev) /* got VF config message back from PF, now we can parse it */ for (i = 0; i < vf->vf_res->num_vsis; i++) { - if (vf->vf_res->vsi_res[i].vsi_type == I40E_VSI_SRIOV) + if (vf->vf_res->vsi_res[i].vsi_type == VIRTCHNL_VSI_SRIOV) vf->vsi_res = &vf->vf_res->vsi_res[i]; } @@ -1322,21 +1322,21 @@ i40evf_handle_pf_event(__rte_unused struct rte_eth_dev *dev, uint8_t *msg, __rte_unused uint16_t msglen) { - struct i40e_virtchnl_pf_event *pf_msg = - (struct i40e_virtchnl_pf_event *)msg; + struct virtchnl_pf_event *pf_msg = + (struct virtchnl_pf_event *)msg; struct i40e_vf *vf = I40EVF_DEV_PRIVATE_TO_VF(dev->data->dev_private); switch (pf_msg->event) { - case I40E_VIRTCHNL_EVENT_RESET_IMPENDING: + case VIRTCHNL_EVENT_RESET_IMPENDING: PMD_DRV_LOG(DEBUG, "VIRTCHNL_EVENT_RESET_IMPENDING event"); _rte_eth_dev_callback_process(dev, RTE_ETH_EVENT_INTR_RESET, NULL); break; - case I40E_VIRTCHNL_EVENT_LINK_CHANGE: + case VIRTCHNL_EVENT_LINK_CHANGE: PMD_DRV_LOG(DEBUG, "VIRTCHNL_EVENT_LINK_CHANGE event"); vf->link_up = pf_msg->event_data.link_event.link_status; vf->link_speed = pf_msg->event_data.link_event.link_speed; break; - case I40E_VIRTCHNL_EVENT_PF_DRIVER_CLOSE: + case VIRTCHNL_EVENT_PF_DRIVER_CLOSE: PMD_DRV_LOG(DEBUG, "VIRTCHNL_EVENT_PF_DRIVER_CLOSE event"); break; default: @@ -1352,7 +1352,7 @@ i40evf_handle_aq_msg(struct rte_eth_dev *dev) struct i40e_vf *vf = I40EVF_DEV_PRIVATE_TO_VF(dev->data->dev_private); struct i40e_arq_event_info info; uint16_t pending, aq_opc; - enum i40e_virtchnl_ops msg_opc; + enum virtchnl_ops msg_opc; enum i40e_status_code msg_ret; int ret; @@ -1377,13 +1377,13 @@ i40evf_handle_aq_msg(struct rte_eth_dev *dev) * cookie_high of struct i40e_aq_desc, while return error code * are stored in cookie_low, Which is done by * i40e_aq_send_msg_to_vf in PF driver.*/ - msg_opc = (enum i40e_virtchnl_ops)rte_le_to_cpu_32( + msg_opc = (enum virtchnl_ops)rte_le_to_cpu_32( info.desc.cookie_high); msg_ret = (enum i40e_status_code)rte_le_to_cpu_32( info.desc.cookie_low); switch (aq_opc) { case i40e_aqc_opc_send_msg_to_vf: - if (msg_opc == I40E_VIRTCHNL_OP_EVENT) + if (msg_opc == VIRTCHNL_OP_EVENT) /* process event*/ i40evf_handle_pf_event(dev, info.msg_buf, info.msg_len); @@ -1592,8 +1592,8 @@ i40evf_dev_configure(struct rte_eth_dev *dev) */ if (!conf->rxmode.hw_strip_crc) { vf = I40EVF_DEV_PRIVATE_TO_VF(dev->data->dev_private); - if ((vf->version_major == I40E_VIRTCHNL_VERSION_MAJOR) && - (vf->version_minor <= I40E_VIRTCHNL_VERSION_MINOR)) { + if ((vf->version_major == VIRTCHNL_VERSION_MAJOR) && + (vf->version_minor <= VIRTCHNL_VERSION_MINOR)) { /* Peer is running non-DPDK PF driver. */ PMD_INIT_LOG(ERR, "VF can't disable HW CRC Strip"); return -EINVAL; @@ -2001,7 +2001,7 @@ i40evf_dev_rx_queue_intr_disable(struct rte_eth_dev *dev, uint16_t queue_id) static void i40evf_add_del_all_mac_addr(struct rte_eth_dev *dev, bool add) { - struct i40e_virtchnl_ether_addr_list *list; + struct virtchnl_ether_addr_list *list; struct i40e_vf *vf = I40EVF_DEV_PRIVATE_TO_VF(dev->data->dev_private); int err, i, j; int next_begin = 0; @@ -2012,11 +2012,11 @@ i40evf_add_del_all_mac_addr(struct rte_eth_dev *dev, bool add) do { j = 0; - len = sizeof(struct i40e_virtchnl_ether_addr_list); + len = sizeof(struct virtchnl_ether_addr_list); for (i = begin; i < I40E_NUM_MACADDR_MAX; i++, next_begin++) { if (is_zero_ether_addr(&dev->data->mac_addrs[i])) continue; - len += sizeof(struct i40e_virtchnl_ether_addr); + len += sizeof(struct virtchnl_ether_addr); if (len >= I40E_AQ_BUF_SZ) { next_begin = i + 1; break; @@ -2043,8 +2043,8 @@ i40evf_add_del_all_mac_addr(struct rte_eth_dev *dev, bool add) } list->vsi_id = vf->vsi_res->vsi_id; list->num_elements = j; - args.ops = add ? I40E_VIRTCHNL_OP_ADD_ETHER_ADDRESS : - I40E_VIRTCHNL_OP_DEL_ETHER_ADDRESS; + args.ops = add ? VIRTCHNL_OP_ADD_ETH_ADDR : + VIRTCHNL_OP_DEL_ETH_ADDR; args.in_args = (uint8_t *)list; args.in_args_size = len; args.out_buffer = vf->aq_resp; diff --git a/drivers/net/i40e/i40e_pf.c b/drivers/net/i40e/i40e_pf.c index 0758503..2e57b66 100644 --- a/drivers/net/i40e/i40e_pf.c +++ b/drivers/net/i40e/i40e_pf.c @@ -61,7 +61,7 @@ static int i40e_pf_host_switch_queues(struct i40e_pf_vf *vf, - struct i40e_virtchnl_queue_select *qsel, + struct virtchnl_queue_select *qsel, bool on); /** @@ -128,7 +128,7 @@ i40e_pf_host_vf_reset(struct i40e_pf_vf *vf, bool do_hw_reset) struct i40e_pf *pf; uint16_t vf_id, abs_vf_id, vf_msix_num; int ret; - struct i40e_virtchnl_queue_select qsel; + struct virtchnl_queue_select qsel; if (vf == NULL) return -EINVAL; @@ -139,7 +139,7 @@ i40e_pf_host_vf_reset(struct i40e_pf_vf *vf, bool do_hw_reset) abs_vf_id = vf_id + hw->func_caps.vf_base_id; /* Notify VF that we are in VFR progress */ - I40E_WRITE_REG(hw, I40E_VFGEN_RSTAT1(vf_id), I40E_VFR_INPROGRESS); + I40E_WRITE_REG(hw, I40E_VFGEN_RSTAT1(vf_id), VIRTCHNL_VFR_INPROGRESS); /* * If require a SW VF reset, a VFLR interrupt will be generated, @@ -220,7 +220,7 @@ i40e_pf_host_vf_reset(struct i40e_pf_vf *vf, bool do_hw_reset) } /* Reset done, Set COMPLETE flag and clear reset bit */ - I40E_WRITE_REG(hw, I40E_VFGEN_RSTAT1(vf_id), I40E_VFR_COMPLETED); + I40E_WRITE_REG(hw, I40E_VFGEN_RSTAT1(vf_id), VIRTCHNL_VFR_COMPLETED); val = I40E_READ_REG(hw, I40E_VPGEN_VFRTRIG(vf_id)); val &= ~I40E_VPGEN_VFRTRIG_VFSWR_MASK; I40E_WRITE_REG(hw, I40E_VPGEN_VFRTRIG(vf_id), val); @@ -248,7 +248,7 @@ i40e_pf_host_vf_reset(struct i40e_pf_vf *vf, bool do_hw_reset) return -EFAULT; } - I40E_WRITE_REG(hw, I40E_VFGEN_RSTAT1(vf_id), I40E_VFR_VFACTIVE); + I40E_WRITE_REG(hw, I40E_VFGEN_RSTAT1(vf_id), VIRTCHNL_VFR_VFACTIVE); return ret; } @@ -277,7 +277,7 @@ i40e_pf_host_send_msg_to_vf(struct i40e_pf_vf *vf, static void i40e_pf_host_process_cmd_version(struct i40e_pf_vf *vf, bool b_op) { - struct i40e_virtchnl_version_info info; + struct virtchnl_version_info info; /* Respond like a Linux PF host in order to support both DPDK VF and * Linux VF driver. The expense is original DPDK host specific feature @@ -286,16 +286,16 @@ i40e_pf_host_process_cmd_version(struct i40e_pf_vf *vf, bool b_op) * DPDK VF also can't identify host driver by version number returned. * It always assume talking with Linux PF. */ - info.major = I40E_VIRTCHNL_VERSION_MAJOR; - info.minor = I40E_VIRTCHNL_VERSION_MINOR_NO_VF_CAPS; + info.major = VIRTCHNL_VERSION_MAJOR; + info.minor = VIRTCHNL_VERSION_MINOR_NO_VF_CAPS; if (b_op) - i40e_pf_host_send_msg_to_vf(vf, I40E_VIRTCHNL_OP_VERSION, + i40e_pf_host_send_msg_to_vf(vf, VIRTCHNL_OP_VERSION, I40E_SUCCESS, (uint8_t *)&info, sizeof(info)); else - i40e_pf_host_send_msg_to_vf(vf, I40E_VIRTCHNL_OP_VERSION, + i40e_pf_host_send_msg_to_vf(vf, VIRTCHNL_OP_VERSION, I40E_NOT_SUPPORTED, (uint8_t *)&info, sizeof(info)); @@ -313,22 +313,22 @@ i40e_pf_host_process_cmd_reset_vf(struct i40e_pf_vf *vf) static int i40e_pf_host_process_cmd_get_vf_resource(struct i40e_pf_vf *vf, bool b_op) { - struct i40e_virtchnl_vf_resource *vf_res = NULL; + struct virtchnl_vf_resource *vf_res = NULL; struct i40e_hw *hw = I40E_PF_TO_HW(vf->pf); uint32_t len = 0; int ret = I40E_SUCCESS; if (!b_op) { i40e_pf_host_send_msg_to_vf(vf, - I40E_VIRTCHNL_OP_GET_VF_RESOURCES, + VIRTCHNL_OP_GET_VF_RESOURCES, I40E_NOT_SUPPORTED, NULL, 0); return ret; } /* only have 1 VSI by default */ - len = sizeof(struct i40e_virtchnl_vf_resource) + + len = sizeof(struct virtchnl_vf_resource) + I40E_DEFAULT_VF_VSI_NUM * - sizeof(struct i40e_virtchnl_vsi_resource); + sizeof(struct virtchnl_vsi_resource); vf_res = rte_zmalloc("i40e_vf_res", len, 0); if (vf_res == NULL) { @@ -339,8 +339,8 @@ i40e_pf_host_process_cmd_get_vf_resource(struct i40e_pf_vf *vf, bool b_op) goto send_msg; } - vf_res->vf_offload_flags = I40E_VIRTCHNL_VF_OFFLOAD_L2 | - I40E_VIRTCHNL_VF_OFFLOAD_VLAN; + vf_res->vf_offload_flags = VIRTCHNL_VF_OFFLOAD_L2 | + VIRTCHNL_VF_OFFLOAD_VLAN; vf_res->max_vectors = hw->func_caps.num_msix_vectors_vf; vf_res->num_queue_pairs = vf->vsi->nb_qps; vf_res->num_vsis = I40E_DEFAULT_VF_VSI_NUM; @@ -353,7 +353,7 @@ i40e_pf_host_process_cmd_get_vf_resource(struct i40e_pf_vf *vf, bool b_op) (struct ether_addr *)vf_res->vsi_res[0].default_mac_addr); send_msg: - i40e_pf_host_send_msg_to_vf(vf, I40E_VIRTCHNL_OP_GET_VF_RESOURCES, + i40e_pf_host_send_msg_to_vf(vf, VIRTCHNL_OP_GET_VF_RESOURCES, ret, (uint8_t *)vf_res, len); rte_free(vf_res); @@ -363,7 +363,7 @@ i40e_pf_host_process_cmd_get_vf_resource(struct i40e_pf_vf *vf, bool b_op) static int i40e_pf_host_hmc_config_rxq(struct i40e_hw *hw, struct i40e_pf_vf *vf, - struct i40e_virtchnl_rxq_info *rxq, + struct virtchnl_rxq_info *rxq, uint8_t crcstrip) { int err = I40E_SUCCESS; @@ -431,7 +431,7 @@ i40e_vsi_get_tc_of_queue(struct i40e_vsi *vsi, static int i40e_pf_host_hmc_config_txq(struct i40e_hw *hw, struct i40e_pf_vf *vf, - struct i40e_virtchnl_txq_info *txq) + struct virtchnl_txq_info *txq) { int err = I40E_SUCCESS; struct i40e_hmc_obj_txq tx_ctx; @@ -480,14 +480,14 @@ i40e_pf_host_process_cmd_config_vsi_queues(struct i40e_pf_vf *vf, { struct i40e_hw *hw = I40E_PF_TO_HW(vf->pf); struct i40e_vsi *vsi = vf->vsi; - struct i40e_virtchnl_vsi_queue_config_info *vc_vqci = - (struct i40e_virtchnl_vsi_queue_config_info *)msg; - struct i40e_virtchnl_queue_pair_info *vc_qpi; + struct virtchnl_vsi_queue_config_info *vc_vqci = + (struct virtchnl_vsi_queue_config_info *)msg; + struct virtchnl_queue_pair_info *vc_qpi; int i, ret = I40E_SUCCESS; if (!b_op) { i40e_pf_host_send_msg_to_vf(vf, - I40E_VIRTCHNL_OP_CONFIG_VSI_QUEUES, + VIRTCHNL_OP_CONFIG_VSI_QUEUES, I40E_NOT_SUPPORTED, NULL, 0); return ret; } @@ -511,9 +511,9 @@ i40e_pf_host_process_cmd_config_vsi_queues(struct i40e_pf_vf *vf, /* * Apply VF RX queue setting to HMC. - * If the opcode is I40E_VIRTCHNL_OP_CONFIG_VSI_QUEUES_EXT, + * If the opcode is VIRTCHNL_OP_CONFIG_VSI_QUEUES_EXT, * then the extra information of - * 'struct i40e_virtchnl_queue_pair_extra_info' is needed, + * 'struct virtchnl_queue_pair_extra_info' is needed, * otherwise set the last parameter to NULL. */ if (i40e_pf_host_hmc_config_rxq(hw, vf, &vc_qpi[i].rxq, @@ -533,7 +533,7 @@ i40e_pf_host_process_cmd_config_vsi_queues(struct i40e_pf_vf *vf, } send_msg: - i40e_pf_host_send_msg_to_vf(vf, I40E_VIRTCHNL_OP_CONFIG_VSI_QUEUES, + i40e_pf_host_send_msg_to_vf(vf, VIRTCHNL_OP_CONFIG_VSI_QUEUES, ret, NULL, 0); return ret; @@ -547,15 +547,15 @@ i40e_pf_host_process_cmd_config_vsi_queues_ext(struct i40e_pf_vf *vf, { struct i40e_hw *hw = I40E_PF_TO_HW(vf->pf); struct i40e_vsi *vsi = vf->vsi; - struct i40e_virtchnl_vsi_queue_config_ext_info *vc_vqcei = - (struct i40e_virtchnl_vsi_queue_config_ext_info *)msg; - struct i40e_virtchnl_queue_pair_ext_info *vc_qpei; + struct virtchnl_vsi_queue_config_ext_info *vc_vqcei = + (struct virtchnl_vsi_queue_config_ext_info *)msg; + struct virtchnl_queue_pair_ext_info *vc_qpei; int i, ret = I40E_SUCCESS; if (!b_op) { i40e_pf_host_send_msg_to_vf( vf, - I40E_VIRTCHNL_OP_CONFIG_VSI_QUEUES_EXT, + VIRTCHNL_OP_CONFIG_VSI_QUEUES_EXT, I40E_NOT_SUPPORTED, NULL, 0); return ret; } @@ -578,9 +578,9 @@ i40e_pf_host_process_cmd_config_vsi_queues_ext(struct i40e_pf_vf *vf, } /* * Apply VF RX queue setting to HMC. - * If the opcode is I40E_VIRTCHNL_OP_CONFIG_VSI_QUEUES_EXT, + * If the opcode is VIRTCHNL_OP_CONFIG_VSI_QUEUES_EXT, * then the extra information of - * 'struct i40e_virtchnl_queue_pair_ext_info' is needed, + * 'struct virtchnl_queue_pair_ext_info' is needed, * otherwise set the last parameter to NULL. */ if (i40e_pf_host_hmc_config_rxq(hw, vf, &vc_qpei[i].rxq, @@ -600,7 +600,7 @@ i40e_pf_host_process_cmd_config_vsi_queues_ext(struct i40e_pf_vf *vf, } send_msg: - i40e_pf_host_send_msg_to_vf(vf, I40E_VIRTCHNL_OP_CONFIG_VSI_QUEUES_EXT, + i40e_pf_host_send_msg_to_vf(vf, VIRTCHNL_OP_CONFIG_VSI_QUEUES_EXT, ret, NULL, 0); return ret; @@ -608,7 +608,7 @@ i40e_pf_host_process_cmd_config_vsi_queues_ext(struct i40e_pf_vf *vf, static void i40e_pf_config_irq_link_list(struct i40e_pf_vf *vf, - struct i40e_virtchnl_vector_map *vvm) + struct virtchnl_vector_map *vvm) { #define BITS_PER_CHAR 8 uint64_t linklistmap = 0, tempmap; @@ -711,9 +711,9 @@ i40e_pf_host_process_cmd_config_irq_map(struct i40e_pf_vf *vf, int ret = I40E_SUCCESS; struct i40e_pf *pf = vf->pf; struct i40e_hw *hw = I40E_PF_TO_HW(vf->pf); - struct i40e_virtchnl_irq_map_info *irqmap = - (struct i40e_virtchnl_irq_map_info *)msg; - struct i40e_virtchnl_vector_map *map; + struct virtchnl_irq_map_info *irqmap = + (struct virtchnl_irq_map_info *)msg; + struct virtchnl_vector_map *map; int i; uint16_t vector_id; unsigned long qbit_max; @@ -721,12 +721,12 @@ i40e_pf_host_process_cmd_config_irq_map(struct i40e_pf_vf *vf, if (!b_op) { i40e_pf_host_send_msg_to_vf( vf, - I40E_VIRTCHNL_OP_CONFIG_IRQ_MAP, + VIRTCHNL_OP_CONFIG_IRQ_MAP, I40E_NOT_SUPPORTED, NULL, 0); return ret; } - if (msg == NULL || msglen < sizeof(struct i40e_virtchnl_irq_map_info)) { + if (msg == NULL || msglen < sizeof(struct virtchnl_irq_map_info)) { PMD_DRV_LOG(ERR, "buffer too short"); ret = I40E_ERR_PARAM; goto send_msg; @@ -773,7 +773,7 @@ i40e_pf_host_process_cmd_config_irq_map(struct i40e_pf_vf *vf, } send_msg: - i40e_pf_host_send_msg_to_vf(vf, I40E_VIRTCHNL_OP_CONFIG_IRQ_MAP, + i40e_pf_host_send_msg_to_vf(vf, VIRTCHNL_OP_CONFIG_IRQ_MAP, ret, NULL, 0); return ret; @@ -781,7 +781,7 @@ i40e_pf_host_process_cmd_config_irq_map(struct i40e_pf_vf *vf, static int i40e_pf_host_switch_queues(struct i40e_pf_vf *vf, - struct i40e_virtchnl_queue_select *qsel, + struct virtchnl_queue_select *qsel, bool on) { int ret = I40E_SUCCESS; @@ -831,8 +831,8 @@ i40e_pf_host_process_cmd_enable_queues(struct i40e_pf_vf *vf, uint16_t msglen) { int ret = I40E_SUCCESS; - struct i40e_virtchnl_queue_select *q_sel = - (struct i40e_virtchnl_queue_select *)msg; + struct virtchnl_queue_select *q_sel = + (struct virtchnl_queue_select *)msg; if (msg == NULL || msglen != sizeof(*q_sel)) { ret = I40E_ERR_PARAM; @@ -841,7 +841,7 @@ i40e_pf_host_process_cmd_enable_queues(struct i40e_pf_vf *vf, ret = i40e_pf_host_switch_queues(vf, q_sel, true); send_msg: - i40e_pf_host_send_msg_to_vf(vf, I40E_VIRTCHNL_OP_ENABLE_QUEUES, + i40e_pf_host_send_msg_to_vf(vf, VIRTCHNL_OP_ENABLE_QUEUES, ret, NULL, 0); return ret; @@ -854,13 +854,13 @@ i40e_pf_host_process_cmd_disable_queues(struct i40e_pf_vf *vf, bool b_op) { int ret = I40E_SUCCESS; - struct i40e_virtchnl_queue_select *q_sel = - (struct i40e_virtchnl_queue_select *)msg; + struct virtchnl_queue_select *q_sel = + (struct virtchnl_queue_select *)msg; if (!b_op) { i40e_pf_host_send_msg_to_vf( vf, - I40E_VIRTCHNL_OP_DISABLE_QUEUES, + VIRTCHNL_OP_DISABLE_QUEUES, I40E_NOT_SUPPORTED, NULL, 0); return ret; } @@ -872,7 +872,7 @@ i40e_pf_host_process_cmd_disable_queues(struct i40e_pf_vf *vf, ret = i40e_pf_host_switch_queues(vf, q_sel, false); send_msg: - i40e_pf_host_send_msg_to_vf(vf, I40E_VIRTCHNL_OP_DISABLE_QUEUES, + i40e_pf_host_send_msg_to_vf(vf, VIRTCHNL_OP_DISABLE_QUEUES, ret, NULL, 0); return ret; @@ -886,8 +886,8 @@ i40e_pf_host_process_cmd_add_ether_address(struct i40e_pf_vf *vf, bool b_op) { int ret = I40E_SUCCESS; - struct i40e_virtchnl_ether_addr_list *addr_list = - (struct i40e_virtchnl_ether_addr_list *)msg; + struct virtchnl_ether_addr_list *addr_list = + (struct virtchnl_ether_addr_list *)msg; struct i40e_mac_filter_info filter; int i; struct ether_addr *mac; @@ -895,7 +895,7 @@ i40e_pf_host_process_cmd_add_ether_address(struct i40e_pf_vf *vf, if (!b_op) { i40e_pf_host_send_msg_to_vf( vf, - I40E_VIRTCHNL_OP_ADD_ETHER_ADDRESS, + VIRTCHNL_OP_ADD_ETH_ADDR, I40E_NOT_SUPPORTED, NULL, 0); return ret; } @@ -920,7 +920,7 @@ i40e_pf_host_process_cmd_add_ether_address(struct i40e_pf_vf *vf, } send_msg: - i40e_pf_host_send_msg_to_vf(vf, I40E_VIRTCHNL_OP_ADD_ETHER_ADDRESS, + i40e_pf_host_send_msg_to_vf(vf, VIRTCHNL_OP_ADD_ETH_ADDR, ret, NULL, 0); return ret; @@ -933,15 +933,15 @@ i40e_pf_host_process_cmd_del_ether_address(struct i40e_pf_vf *vf, bool b_op) { int ret = I40E_SUCCESS; - struct i40e_virtchnl_ether_addr_list *addr_list = - (struct i40e_virtchnl_ether_addr_list *)msg; + struct virtchnl_ether_addr_list *addr_list = + (struct virtchnl_ether_addr_list *)msg; int i; struct ether_addr *mac; if (!b_op) { i40e_pf_host_send_msg_to_vf( vf, - I40E_VIRTCHNL_OP_DEL_ETHER_ADDRESS, + VIRTCHNL_OP_DEL_ETH_ADDR, I40E_NOT_SUPPORTED, NULL, 0); return ret; } @@ -962,7 +962,7 @@ i40e_pf_host_process_cmd_del_ether_address(struct i40e_pf_vf *vf, } send_msg: - i40e_pf_host_send_msg_to_vf(vf, I40E_VIRTCHNL_OP_DEL_ETHER_ADDRESS, + i40e_pf_host_send_msg_to_vf(vf, VIRTCHNL_OP_DEL_ETH_ADDR, ret, NULL, 0); return ret; @@ -974,15 +974,15 @@ i40e_pf_host_process_cmd_add_vlan(struct i40e_pf_vf *vf, bool b_op) { int ret = I40E_SUCCESS; - struct i40e_virtchnl_vlan_filter_list *vlan_filter_list = - (struct i40e_virtchnl_vlan_filter_list *)msg; + struct virtchnl_vlan_filter_list *vlan_filter_list = + (struct virtchnl_vlan_filter_list *)msg; int i; uint16_t *vid; if (!b_op) { i40e_pf_host_send_msg_to_vf( vf, - I40E_VIRTCHNL_OP_ADD_VLAN, + VIRTCHNL_OP_ADD_VLAN, I40E_NOT_SUPPORTED, NULL, 0); return ret; } @@ -1002,7 +1002,7 @@ i40e_pf_host_process_cmd_add_vlan(struct i40e_pf_vf *vf, } send_msg: - i40e_pf_host_send_msg_to_vf(vf, I40E_VIRTCHNL_OP_ADD_VLAN, + i40e_pf_host_send_msg_to_vf(vf, VIRTCHNL_OP_ADD_VLAN, ret, NULL, 0); return ret; @@ -1015,15 +1015,15 @@ i40e_pf_host_process_cmd_del_vlan(struct i40e_pf_vf *vf, bool b_op) { int ret = I40E_SUCCESS; - struct i40e_virtchnl_vlan_filter_list *vlan_filter_list = - (struct i40e_virtchnl_vlan_filter_list *)msg; + struct virtchnl_vlan_filter_list *vlan_filter_list = + (struct virtchnl_vlan_filter_list *)msg; int i; uint16_t *vid; if (!b_op) { i40e_pf_host_send_msg_to_vf( vf, - I40E_VIRTCHNL_OP_DEL_VLAN, + VIRTCHNL_OP_DEL_VLAN, I40E_NOT_SUPPORTED, NULL, 0); return ret; } @@ -1042,7 +1042,7 @@ i40e_pf_host_process_cmd_del_vlan(struct i40e_pf_vf *vf, } send_msg: - i40e_pf_host_send_msg_to_vf(vf, I40E_VIRTCHNL_OP_DEL_VLAN, + i40e_pf_host_send_msg_to_vf(vf, VIRTCHNL_OP_DEL_VLAN, ret, NULL, 0); return ret; @@ -1056,15 +1056,15 @@ i40e_pf_host_process_cmd_config_promisc_mode( bool b_op) { int ret = I40E_SUCCESS; - struct i40e_virtchnl_promisc_info *promisc = - (struct i40e_virtchnl_promisc_info *)msg; + struct virtchnl_promisc_info *promisc = + (struct virtchnl_promisc_info *)msg; struct i40e_hw *hw = I40E_PF_TO_HW(vf->pf); bool unicast = FALSE, multicast = FALSE; if (!b_op) { i40e_pf_host_send_msg_to_vf( vf, - I40E_VIRTCHNL_OP_CONFIG_PROMISCUOUS_MODE, + VIRTCHNL_OP_CONFIG_PROMISCUOUS_MODE, I40E_NOT_SUPPORTED, NULL, 0); return ret; } @@ -1074,21 +1074,21 @@ i40e_pf_host_process_cmd_config_promisc_mode( goto send_msg; } - if (promisc->flags & I40E_FLAG_VF_UNICAST_PROMISC) + if (promisc->flags & FLAG_VF_UNICAST_PROMISC) unicast = TRUE; ret = i40e_aq_set_vsi_unicast_promiscuous(hw, vf->vsi->seid, unicast, NULL, true); if (ret != I40E_SUCCESS) goto send_msg; - if (promisc->flags & I40E_FLAG_VF_MULTICAST_PROMISC) + if (promisc->flags & FLAG_VF_MULTICAST_PROMISC) multicast = TRUE; ret = i40e_aq_set_vsi_multicast_promiscuous(hw, vf->vsi->seid, multicast, NULL); send_msg: i40e_pf_host_send_msg_to_vf(vf, - I40E_VIRTCHNL_OP_CONFIG_PROMISCUOUS_MODE, ret, NULL, 0); + VIRTCHNL_OP_CONFIG_PROMISCUOUS_MODE, ret, NULL, 0); return ret; } @@ -1099,12 +1099,12 @@ i40e_pf_host_process_cmd_get_stats(struct i40e_pf_vf *vf, bool b_op) i40e_update_vsi_stats(vf->vsi); if (b_op) - i40e_pf_host_send_msg_to_vf(vf, I40E_VIRTCHNL_OP_GET_STATS, + i40e_pf_host_send_msg_to_vf(vf, VIRTCHNL_OP_GET_STATS, I40E_SUCCESS, (uint8_t *)&vf->vsi->eth_stats, sizeof(vf->vsi->eth_stats)); else - i40e_pf_host_send_msg_to_vf(vf, I40E_VIRTCHNL_OP_GET_STATS, + i40e_pf_host_send_msg_to_vf(vf, VIRTCHNL_OP_GET_STATS, I40E_NOT_SUPPORTED, (uint8_t *)&vf->vsi->eth_stats, sizeof(vf->vsi->eth_stats)); @@ -1120,8 +1120,8 @@ i40e_pf_host_process_cmd_cfg_vlan_offload( bool b_op) { int ret = I40E_SUCCESS; - struct i40e_virtchnl_vlan_offload_info *offload = - (struct i40e_virtchnl_vlan_offload_info *)msg; + struct virtchnl_vlan_offload_info *offload = + (struct virtchnl_vlan_offload_info *)msg; if (!b_op) { i40e_pf_host_send_msg_to_vf( @@ -1155,8 +1155,8 @@ i40e_pf_host_process_cmd_cfg_pvid(struct i40e_pf_vf *vf, bool b_op) { int ret = I40E_SUCCESS; - struct i40e_virtchnl_pvid_info *tpid_info = - (struct i40e_virtchnl_pvid_info *)msg; + struct virtchnl_pvid_info *tpid_info = + (struct virtchnl_pvid_info *)msg; if (!b_op) { i40e_pf_host_send_msg_to_vf( @@ -1183,9 +1183,9 @@ i40e_pf_host_process_cmd_cfg_pvid(struct i40e_pf_vf *vf, void i40e_notify_vf_link_status(struct rte_eth_dev *dev, struct i40e_pf_vf *vf) { - struct i40e_virtchnl_pf_event event; + struct virtchnl_pf_event event; - event.event = I40E_VIRTCHNL_EVENT_LINK_CHANGE; + event.event = VIRTCHNL_EVENT_LINK_CHANGE; event.event_data.link_event.link_status = dev->data->dev_link.link_status; @@ -1215,7 +1215,7 @@ i40e_notify_vf_link_status(struct rte_eth_dev *dev, struct i40e_pf_vf *vf) break; } - i40e_pf_host_send_msg_to_vf(vf, I40E_VIRTCHNL_OP_EVENT, + i40e_pf_host_send_msg_to_vf(vf, VIRTCHNL_OP_EVENT, I40E_SUCCESS, (uint8_t *)&event, sizeof(event)); } @@ -1273,71 +1273,71 @@ i40e_pf_host_handle_vf_msg(struct rte_eth_dev *dev, } switch (opcode) { - case I40E_VIRTCHNL_OP_VERSION : + case VIRTCHNL_OP_VERSION: PMD_DRV_LOG(INFO, "OP_VERSION received"); i40e_pf_host_process_cmd_version(vf, b_op); break; - case I40E_VIRTCHNL_OP_RESET_VF : + case VIRTCHNL_OP_RESET_VF: PMD_DRV_LOG(INFO, "OP_RESET_VF received"); i40e_pf_host_process_cmd_reset_vf(vf); break; - case I40E_VIRTCHNL_OP_GET_VF_RESOURCES: + case VIRTCHNL_OP_GET_VF_RESOURCES: PMD_DRV_LOG(INFO, "OP_GET_VF_RESOURCES received"); i40e_pf_host_process_cmd_get_vf_resource(vf, b_op); break; - case I40E_VIRTCHNL_OP_CONFIG_VSI_QUEUES: + case VIRTCHNL_OP_CONFIG_VSI_QUEUES: PMD_DRV_LOG(INFO, "OP_CONFIG_VSI_QUEUES received"); i40e_pf_host_process_cmd_config_vsi_queues(vf, msg, msglen, b_op); break; - case I40E_VIRTCHNL_OP_CONFIG_VSI_QUEUES_EXT: + case VIRTCHNL_OP_CONFIG_VSI_QUEUES_EXT: PMD_DRV_LOG(INFO, "OP_CONFIG_VSI_QUEUES_EXT received"); i40e_pf_host_process_cmd_config_vsi_queues_ext(vf, msg, msglen, b_op); break; - case I40E_VIRTCHNL_OP_CONFIG_IRQ_MAP: + case VIRTCHNL_OP_CONFIG_IRQ_MAP: PMD_DRV_LOG(INFO, "OP_CONFIG_IRQ_MAP received"); i40e_pf_host_process_cmd_config_irq_map(vf, msg, msglen, b_op); break; - case I40E_VIRTCHNL_OP_ENABLE_QUEUES: + case VIRTCHNL_OP_ENABLE_QUEUES: PMD_DRV_LOG(INFO, "OP_ENABLE_QUEUES received"); if (b_op) { i40e_pf_host_process_cmd_enable_queues(vf, msg, msglen); i40e_notify_vf_link_status(dev, vf); } else { i40e_pf_host_send_msg_to_vf( - vf, I40E_VIRTCHNL_OP_ENABLE_QUEUES, + vf, VIRTCHNL_OP_ENABLE_QUEUES, I40E_NOT_SUPPORTED, NULL, 0); } break; - case I40E_VIRTCHNL_OP_DISABLE_QUEUES: + case VIRTCHNL_OP_DISABLE_QUEUES: PMD_DRV_LOG(INFO, "OP_DISABLE_QUEUE received"); i40e_pf_host_process_cmd_disable_queues(vf, msg, msglen, b_op); break; - case I40E_VIRTCHNL_OP_ADD_ETHER_ADDRESS: + case VIRTCHNL_OP_ADD_ETH_ADDR: PMD_DRV_LOG(INFO, "OP_ADD_ETHER_ADDRESS received"); i40e_pf_host_process_cmd_add_ether_address(vf, msg, msglen, b_op); break; - case I40E_VIRTCHNL_OP_DEL_ETHER_ADDRESS: + case VIRTCHNL_OP_DEL_ETH_ADDR: PMD_DRV_LOG(INFO, "OP_DEL_ETHER_ADDRESS received"); i40e_pf_host_process_cmd_del_ether_address(vf, msg, msglen, b_op); break; - case I40E_VIRTCHNL_OP_ADD_VLAN: + case VIRTCHNL_OP_ADD_VLAN: PMD_DRV_LOG(INFO, "OP_ADD_VLAN received"); i40e_pf_host_process_cmd_add_vlan(vf, msg, msglen, b_op); break; - case I40E_VIRTCHNL_OP_DEL_VLAN: + case VIRTCHNL_OP_DEL_VLAN: PMD_DRV_LOG(INFO, "OP_DEL_VLAN received"); i40e_pf_host_process_cmd_del_vlan(vf, msg, msglen, b_op); break; - case I40E_VIRTCHNL_OP_CONFIG_PROMISCUOUS_MODE: + case VIRTCHNL_OP_CONFIG_PROMISCUOUS_MODE: PMD_DRV_LOG(INFO, "OP_CONFIG_PROMISCUOUS_MODE received"); i40e_pf_host_process_cmd_config_promisc_mode(vf, msg, msglen, b_op); break; - case I40E_VIRTCHNL_OP_GET_STATS: + case VIRTCHNL_OP_GET_STATS: PMD_DRV_LOG(INFO, "OP_GET_STATS received"); i40e_pf_host_process_cmd_get_stats(vf, b_op); break; diff --git a/drivers/net/i40e/i40e_pf.h b/drivers/net/i40e/i40e_pf.h index b4c2287..2fddaf7 100644 --- a/drivers/net/i40e/i40e_pf.h +++ b/drivers/net/i40e/i40e_pf.h @@ -35,7 +35,7 @@ #define _I40E_PF_H_ /* VERSION info to exchange between VF and PF host. In case VF works with - * ND kernel driver, it reads I40E_VIRTCHNL_VERSION_MAJOR/MINOR. In + * ND kernel driver, it reads VIRTCHNL_VERSION_MAJOR/MINOR. In * case works with DPDK host, it reads version below. Then VF realize who it * is talking to and use proper language to communicate. * */ @@ -49,45 +49,45 @@ #define I40E_DPDK_OFFSET 0x100 /* DPDK pf driver specific command to VF */ -enum i40e_virtchnl_ops_dpdk { +enum virtchnl_ops_dpdk { /* * Keep some gap between Linux PF commands and * DPDK PF extended commands. */ - I40E_VIRTCHNL_OP_CFG_VLAN_OFFLOAD = I40E_VIRTCHNL_OP_VERSION + + I40E_VIRTCHNL_OP_CFG_VLAN_OFFLOAD = VIRTCHNL_OP_VERSION + I40E_DPDK_OFFSET, I40E_VIRTCHNL_OP_CFG_VLAN_PVID, - I40E_VIRTCHNL_OP_CONFIG_VSI_QUEUES_EXT, + VIRTCHNL_OP_CONFIG_VSI_QUEUES_EXT, }; /* A structure to support extended info of a receive queue. */ -struct i40e_virtchnl_rxq_ext_info { +struct virtchnl_rxq_ext_info { uint8_t crcstrip; }; /* * A structure to support extended info of queue pairs, an additional field - * is added, comparing to original 'struct i40e_virtchnl_queue_pair_info'. + * is added, comparing to original 'struct virtchnl_queue_pair_info'. */ -struct i40e_virtchnl_queue_pair_ext_info { +struct virtchnl_queue_pair_ext_info { /* vsi_id and queue_id should be identical for both rx and tx queues.*/ - struct i40e_virtchnl_txq_info txq; - struct i40e_virtchnl_rxq_info rxq; - struct i40e_virtchnl_rxq_ext_info rxq_ext; + struct virtchnl_txq_info txq; + struct virtchnl_rxq_info rxq; + struct virtchnl_rxq_ext_info rxq_ext; }; /* * A structure to support extended info of VSI queue pairs, - * 'struct i40e_virtchnl_queue_pair_ext_info' is used, see its original - * of 'struct i40e_virtchnl_queue_pair_info'. + * 'struct virtchnl_queue_pair_ext_info' is used, see its original + * of 'struct virtchnl_queue_pair_info'. */ -struct i40e_virtchnl_vsi_queue_config_ext_info { +struct virtchnl_vsi_queue_config_ext_info { uint16_t vsi_id; uint16_t num_queue_pairs; - struct i40e_virtchnl_queue_pair_ext_info qpair[0]; + struct virtchnl_queue_pair_ext_info qpair[0]; }; -struct i40e_virtchnl_vlan_offload_info { +struct virtchnl_vlan_offload_info { uint16_t vsi_id; uint8_t enable_vlan_strip; uint8_t reserved; @@ -106,7 +106,7 @@ struct i40e_virtchnl_vlan_offload_info { * enable op, needs to specify the pvid. PF returns status * code in retval. */ -struct i40e_virtchnl_pvid_info { +struct virtchnl_pvid_info { uint16_t vsi_id; struct i40e_vsi_vlan_pvid_info info; }; diff --git a/drivers/net/i40e/rte_pmd_i40e.h b/drivers/net/i40e/rte_pmd_i40e.h index 1efb2c4..cbe742e 100644 --- a/drivers/net/i40e/rte_pmd_i40e.h +++ b/drivers/net/i40e/rte_pmd_i40e.h @@ -59,7 +59,7 @@ enum rte_pmd_i40e_mb_event_rsp { */ struct rte_pmd_i40e_mb_event_param { uint16_t vfid; /**< Virtual Function number */ - uint16_t msg_type; /**< VF to PF message type, see i40e_virtchnl_ops */ + uint16_t msg_type; /**< VF to PF message type, see virtchnl_ops */ uint16_t retval; /**< return value */ void *msg; /**< pointer to message */ uint16_t msglen; /**< length of the message */ -- 2.4.11 ^ permalink raw reply [flat|nested] 69+ messages in thread
* [dpdk-dev] [PATCH 02/10] net/i40e/base: sync nvmupdate command and adminq subtask 2017-05-27 3:47 [dpdk-dev] [PATCH 00/10] net/i40e: base code update Jingjing Wu 2017-05-27 3:47 ` [dpdk-dev] [PATCH 01/10] net/i40e/base: use new virtchnl header file Jingjing Wu @ 2017-05-27 3:47 ` Jingjing Wu 2017-05-27 3:47 ` [dpdk-dev] [PATCH 03/10] net/i40e/base: add AQ command for read/write PHY registers Jingjing Wu ` (9 subsequent siblings) 11 siblings, 0 replies; 69+ messages in thread From: Jingjing Wu @ 2017-05-27 3:47 UTC (permalink / raw) To: helin.zhang; +Cc: dev, jingjing.wu During NVMupdate, state machine gets into unrecoverable state because i40e_clean_adminq_subtask can get scheduled after the admin queue command but before other state variables are updated. This patch adds locking around admin queue command and update of state variables so that adminq_subtask will have accurate information whenever it gets scheduled. Signed-off-by: Jingjing Wu <jingjing.wu@intel.com> --- drivers/net/i40e/base/i40e_nvm.c | 6 ++++++ 1 file changed, 6 insertions(+) diff --git a/drivers/net/i40e/base/i40e_nvm.c b/drivers/net/i40e/base/i40e_nvm.c index e896502..90521fa 100644 --- a/drivers/net/i40e/base/i40e_nvm.c +++ b/drivers/net/i40e/base/i40e_nvm.c @@ -899,6 +899,11 @@ enum i40e_status_code i40e_nvmupd_command(struct i40e_hw *hw, hw->nvmupd_state = I40E_NVMUPD_STATE_INIT; } + /* Acquire lock to prevent race condition where adminq_task + * can execute after i40e_nvmupd_nvm_read/write but before state + * variables (nvm_wait_opcode, nvm_release_on_done) are updated + */ + i40e_acquire_spinlock(&hw->aq.arq_spinlock); switch (hw->nvmupd_state) { case I40E_NVMUPD_STATE_INIT: status = i40e_nvmupd_state_init(hw, cmd, bytes, perrno); @@ -934,6 +939,7 @@ enum i40e_status_code i40e_nvmupd_command(struct i40e_hw *hw, *perrno = -ESRCH; break; } + i40e_release_spinlock(&hw->aq.arq_spinlock); return status; } -- 2.4.11 ^ permalink raw reply [flat|nested] 69+ messages in thread
* [dpdk-dev] [PATCH 03/10] net/i40e/base: add AQ command for read/write PHY registers 2017-05-27 3:47 [dpdk-dev] [PATCH 00/10] net/i40e: base code update Jingjing Wu 2017-05-27 3:47 ` [dpdk-dev] [PATCH 01/10] net/i40e/base: use new virtchnl header file Jingjing Wu 2017-05-27 3:47 ` [dpdk-dev] [PATCH 02/10] net/i40e/base: sync nvmupdate command and adminq subtask Jingjing Wu @ 2017-05-27 3:47 ` Jingjing Wu 2017-05-27 3:47 ` [dpdk-dev] [PATCH 04/10] net/i40e/base: add support for Adaptive Virtual Function Jingjing Wu ` (8 subsequent siblings) 11 siblings, 0 replies; 69+ messages in thread From: Jingjing Wu @ 2017-05-27 3:47 UTC (permalink / raw) To: helin.zhang; +Cc: dev, jingjing.wu This patch adds new additional command for accessing to PHY registers. Signed-off-by: Jingjing Wu <jingjing.wu@intel.com> --- drivers/net/i40e/base/i40e_adminq_cmd.h | 18 +++++++++ drivers/net/i40e/base/i40e_common.c | 70 +++++++++++++++++++++++++++++++++ drivers/net/i40e/base/i40e_prototype.h | 9 +++++ 3 files changed, 97 insertions(+) diff --git a/drivers/net/i40e/base/i40e_adminq_cmd.h b/drivers/net/i40e/base/i40e_adminq_cmd.h index 09f5bf5..83b28f8 100644 --- a/drivers/net/i40e/base/i40e_adminq_cmd.h +++ b/drivers/net/i40e/base/i40e_adminq_cmd.h @@ -245,6 +245,8 @@ enum i40e_admin_queue_opc { i40e_aqc_opc_set_phy_debug = 0x0622, i40e_aqc_opc_upload_ext_phy_fm = 0x0625, i40e_aqc_opc_run_phy_activity = 0x0626, + i40e_aqc_opc_set_phy_register = 0x0628, + i40e_aqc_opc_get_phy_register = 0x0629, /* NVM commands */ i40e_aqc_opc_nvm_read = 0x0701, @@ -2128,6 +2130,22 @@ struct i40e_aqc_run_phy_activity { I40E_CHECK_CMD_LENGTH(i40e_aqc_run_phy_activity); +/* Set PHY Register command (0x0628) */ +/* Get PHY Register command (0x0629) */ +struct i40e_aqc_phy_register_access { + u8 phy_interface; +#define I40E_AQ_PHY_REG_ACCESS_INTERNAL 0 +#define I40E_AQ_PHY_REG_ACCESS_EXTERNAL 1 +#define I40E_AQ_PHY_REG_ACCESS_EXTERNAL_MODULE 2 + u8 dev_addres; + u8 reserved1[2]; + u32 reg_address; + u32 reg_value; + u8 reserved2[4]; +}; + +I40E_CHECK_CMD_LENGTH(i40e_aqc_phy_register_access); + /* NVM Read command (indirect 0x0701) * NVM Erase commands (direct 0x0702) * NVM Update commands (indirect 0x0703) diff --git a/drivers/net/i40e/base/i40e_common.c b/drivers/net/i40e/base/i40e_common.c index 6c09c27..280d9da 100644 --- a/drivers/net/i40e/base/i40e_common.c +++ b/drivers/net/i40e/base/i40e_common.c @@ -6863,6 +6863,76 @@ void i40e_write_rx_ctl(struct i40e_hw *hw, u32 reg_addr, u32 reg_val) if (status || use_register) wr32(hw, reg_addr, reg_val); } + +/** + * i40e_aq_set_phy_register + * @hw: pointer to the hw struct + * @phy_select: select which phy should be accessed + * @dev_addr: PHY device address + * @reg_addr: PHY register address + * @reg_val: new register value + * @cmd_details: pointer to command details structure or NULL + * + * Write the external PHY register. + **/ +enum i40e_status_code i40e_aq_set_phy_register(struct i40e_hw *hw, + u8 phy_select, u8 dev_addr, + u32 reg_addr, u32 reg_val, + struct i40e_asq_cmd_details *cmd_details) +{ + struct i40e_aq_desc desc; + struct i40e_aqc_phy_register_access *cmd = + (struct i40e_aqc_phy_register_access *)&desc.params.raw; + enum i40e_status_code status; + + i40e_fill_default_direct_cmd_desc(&desc, + i40e_aqc_opc_set_phy_register); + + cmd->phy_interface = phy_select; + cmd->dev_addres = dev_addr; + cmd->reg_address = reg_addr; + cmd->reg_value = reg_val; + + status = i40e_asq_send_command(hw, &desc, NULL, 0, cmd_details); + + return status; +} + +/** + * i40e_aq_get_phy_register + * @hw: pointer to the hw struct + * @phy_select: select which phy should be accessed + * @dev_addr: PHY device address + * @reg_addr: PHY register address + * @reg_val: read register value + * @cmd_details: pointer to command details structure or NULL + * + * Read the external PHY register. + **/ +enum i40e_status_code i40e_aq_get_phy_register(struct i40e_hw *hw, + u8 phy_select, u8 dev_addr, + u32 reg_addr, u32 *reg_val, + struct i40e_asq_cmd_details *cmd_details) +{ + struct i40e_aq_desc desc; + struct i40e_aqc_phy_register_access *cmd = + (struct i40e_aqc_phy_register_access *)&desc.params.raw; + enum i40e_status_code status; + + i40e_fill_default_direct_cmd_desc(&desc, + i40e_aqc_opc_get_phy_register); + + cmd->phy_interface = phy_select; + cmd->dev_addres = dev_addr; + cmd->reg_address = reg_addr; + + status = i40e_asq_send_command(hw, &desc, NULL, 0, cmd_details); + if (!status) + *reg_val = cmd->reg_value; + + return status; +} + #ifdef VF_DRIVER /** diff --git a/drivers/net/i40e/base/i40e_prototype.h b/drivers/net/i40e/base/i40e_prototype.h index 6ec4304..54b6750 100644 --- a/drivers/net/i40e/base/i40e_prototype.h +++ b/drivers/net/i40e/base/i40e_prototype.h @@ -533,6 +533,15 @@ enum i40e_status_code i40e_aq_rx_ctl_write_register(struct i40e_hw *hw, u32 reg_addr, u32 reg_val, struct i40e_asq_cmd_details *cmd_details); void i40e_write_rx_ctl(struct i40e_hw *hw, u32 reg_addr, u32 reg_val); +enum i40e_status_code i40e_aq_set_phy_register(struct i40e_hw *hw, + u8 phy_select, u8 dev_addr, + u32 reg_addr, u32 reg_val, + struct i40e_asq_cmd_details *cmd_details); +enum i40e_status_code i40e_aq_get_phy_register(struct i40e_hw *hw, + u8 phy_select, u8 dev_addr, + u32 reg_addr, u32 *reg_val, + struct i40e_asq_cmd_details *cmd_details); + enum i40e_status_code i40e_aq_set_arp_proxy_config(struct i40e_hw *hw, struct i40e_aqc_arp_proxy_data *proxy_config, struct i40e_asq_cmd_details *cmd_details); -- 2.4.11 ^ permalink raw reply [flat|nested] 69+ messages in thread
* [dpdk-dev] [PATCH 04/10] net/i40e/base: add support for Adaptive Virtual Function 2017-05-27 3:47 [dpdk-dev] [PATCH 00/10] net/i40e: base code update Jingjing Wu ` (2 preceding siblings ...) 2017-05-27 3:47 ` [dpdk-dev] [PATCH 03/10] net/i40e/base: add AQ command for read/write PHY registers Jingjing Wu @ 2017-05-27 3:47 ` Jingjing Wu 2017-05-27 3:47 ` [dpdk-dev] [PATCH 05/10] net/i40e/base: store the requested FEC information Jingjing Wu ` (7 subsequent siblings) 11 siblings, 0 replies; 69+ messages in thread From: Jingjing Wu @ 2017-05-27 3:47 UTC (permalink / raw) To: helin.zhang; +Cc: dev, jingjing.wu Add device id define and mac_type assignment needed for Adaptive Virtual Function. Signed-off-by: Jingjing Wu <jingjing.wu@intel.com> --- drivers/net/i40e/base/i40e_common.c | 1 + drivers/net/i40e/base/i40e_devids.h | 1 + 2 files changed, 2 insertions(+) diff --git a/drivers/net/i40e/base/i40e_common.c b/drivers/net/i40e/base/i40e_common.c index 280d9da..5f6dd7b 100644 --- a/drivers/net/i40e/base/i40e_common.c +++ b/drivers/net/i40e/base/i40e_common.c @@ -93,6 +93,7 @@ STATIC enum i40e_status_code i40e_set_mac_type(struct i40e_hw *hw) #if defined(INTEGRATED_VF) || defined(VF_DRIVER) case I40E_DEV_ID_VF: case I40E_DEV_ID_VF_HV: + case I40E_DEV_ID_ADAPTIVE_VF: hw->mac.type = I40E_MAC_VF; break; #endif diff --git a/drivers/net/i40e/base/i40e_devids.h b/drivers/net/i40e/base/i40e_devids.h index 4546689..f4a8784 100644 --- a/drivers/net/i40e/base/i40e_devids.h +++ b/drivers/net/i40e/base/i40e_devids.h @@ -54,6 +54,7 @@ POSSIBILITY OF SUCH DAMAGE. #if defined(INTEGRATED_VF) || defined(VF_DRIVER) || defined(I40E_NDIS_SUPPORT) #define I40E_DEV_ID_VF 0x154C #define I40E_DEV_ID_VF_HV 0x1571 +#define I40E_DEV_ID_ADAPTIVE_VF 0x1889 #endif /* VF_DRIVER */ #ifdef X722_A0_SUPPORT #define I40E_DEV_ID_X722_A0 0x374C -- 2.4.11 ^ permalink raw reply [flat|nested] 69+ messages in thread
* [dpdk-dev] [PATCH 05/10] net/i40e/base: store the requested FEC information 2017-05-27 3:47 [dpdk-dev] [PATCH 00/10] net/i40e: base code update Jingjing Wu ` (3 preceding siblings ...) 2017-05-27 3:47 ` [dpdk-dev] [PATCH 04/10] net/i40e/base: add support for Adaptive Virtual Function Jingjing Wu @ 2017-05-27 3:47 ` Jingjing Wu 2017-05-27 3:47 ` [dpdk-dev] [PATCH 06/10] net/i40e/base: add new phy types for 25G AOC and ACC Jingjing Wu ` (6 subsequent siblings) 11 siblings, 0 replies; 69+ messages in thread From: Jingjing Wu @ 2017-05-27 3:47 UTC (permalink / raw) To: helin.zhang; +Cc: dev, jingjing.wu Store information about FEC modes, that were requested. It will be used in printing link status information function and this way there is no need to call admin queue there. Signed-off-by: Jingjing Wu <jingjing.wu@intel.com> --- drivers/net/i40e/base/i40e_common.c | 4 ++++ drivers/net/i40e/base/i40e_type.h | 1 + 2 files changed, 5 insertions(+) diff --git a/drivers/net/i40e/base/i40e_common.c b/drivers/net/i40e/base/i40e_common.c index 5f6dd7b..7d59b59 100644 --- a/drivers/net/i40e/base/i40e_common.c +++ b/drivers/net/i40e/base/i40e_common.c @@ -2826,6 +2826,10 @@ enum i40e_status_code i40e_update_link_info(struct i40e_hw *hw) if (status) return status; + hw->phy.link_info.req_fec_info = + abilities.fec_cfg_curr_mod_ext_info & + (I40E_AQ_REQUEST_FEC_KR | I40E_AQ_REQUEST_FEC_RS); + i40e_memcpy(hw->phy.link_info.module_type, &abilities.module_type, sizeof(hw->phy.link_info.module_type), I40E_NONDMA_TO_NONDMA); } diff --git a/drivers/net/i40e/base/i40e_type.h b/drivers/net/i40e/base/i40e_type.h index 52a114a..af5347b 100644 --- a/drivers/net/i40e/base/i40e_type.h +++ b/drivers/net/i40e/base/i40e_type.h @@ -271,6 +271,7 @@ struct i40e_link_status { enum i40e_aq_link_speed link_speed; u8 link_info; u8 an_info; + u8 req_fec_info; u8 fec_info; u8 ext_info; u8 loopback; -- 2.4.11 ^ permalink raw reply [flat|nested] 69+ messages in thread
* [dpdk-dev] [PATCH 06/10] net/i40e/base: add new phy types for 25G AOC and ACC 2017-05-27 3:47 [dpdk-dev] [PATCH 00/10] net/i40e: base code update Jingjing Wu ` (4 preceding siblings ...) 2017-05-27 3:47 ` [dpdk-dev] [PATCH 05/10] net/i40e/base: store the requested FEC information Jingjing Wu @ 2017-05-27 3:47 ` Jingjing Wu 2017-05-27 3:47 ` [dpdk-dev] [PATCH 07/10] net/i40e/base: report supported link modes Jingjing Wu ` (5 subsequent siblings) 11 siblings, 0 replies; 69+ messages in thread From: Jingjing Wu @ 2017-05-27 3:47 UTC (permalink / raw) To: helin.zhang; +Cc: dev, jingjing.wu This patch adds new phy types for 25G Active Optical Cables (AOC) and Active Copper Cables (ACC) support. Signed-off-by: Jingjing Wu <jingjing.wu@intel.com> --- drivers/net/i40e/base/i40e_adminq_cmd.h | 4 ++++ drivers/net/i40e/base/i40e_common.c | 2 ++ 2 files changed, 6 insertions(+) diff --git a/drivers/net/i40e/base/i40e_adminq_cmd.h b/drivers/net/i40e/base/i40e_adminq_cmd.h index 83b28f8..70079a0 100644 --- a/drivers/net/i40e/base/i40e_adminq_cmd.h +++ b/drivers/net/i40e/base/i40e_adminq_cmd.h @@ -1849,6 +1849,8 @@ enum i40e_aq_phy_type { I40E_PHY_TYPE_25GBASE_CR = 0x20, I40E_PHY_TYPE_25GBASE_SR = 0x21, I40E_PHY_TYPE_25GBASE_LR = 0x22, + I40E_PHY_TYPE_25GBASE_AOC = 0x23, + I40E_PHY_TYPE_25GBASE_ACC = 0x24, I40E_PHY_TYPE_MAX }; @@ -1906,6 +1908,8 @@ struct i40e_aq_get_phy_abilities_resp { #define I40E_AQ_PHY_TYPE_EXT_25G_CR 0x02 #define I40E_AQ_PHY_TYPE_EXT_25G_SR 0x04 #define I40E_AQ_PHY_TYPE_EXT_25G_LR 0x08 +#define I40E_AQ_PHY_TYPE_EXT_25G_AOC 0x10 +#define I40E_AQ_PHY_TYPE_EXT_25G_ACC 0x20 u8 fec_cfg_curr_mod_ext_info; #define I40E_AQ_ENABLE_FEC_KR 0x01 #define I40E_AQ_ENABLE_FEC_RS 0x02 diff --git a/drivers/net/i40e/base/i40e_common.c b/drivers/net/i40e/base/i40e_common.c index 7d59b59..004c062 100644 --- a/drivers/net/i40e/base/i40e_common.c +++ b/drivers/net/i40e/base/i40e_common.c @@ -1296,6 +1296,8 @@ STATIC enum i40e_media_type i40e_get_media_type(struct i40e_hw *hw) case I40E_PHY_TYPE_40GBASE_AOC: case I40E_PHY_TYPE_10GBASE_AOC: case I40E_PHY_TYPE_25GBASE_CR: + case I40E_PHY_TYPE_25GBASE_AOC: + case I40E_PHY_TYPE_25GBASE_ACC: media = I40E_MEDIA_TYPE_DA; break; case I40E_PHY_TYPE_1000BASE_KX: -- 2.4.11 ^ permalink raw reply [flat|nested] 69+ messages in thread
* [dpdk-dev] [PATCH 07/10] net/i40e/base: report supported link modes 2017-05-27 3:47 [dpdk-dev] [PATCH 00/10] net/i40e: base code update Jingjing Wu ` (5 preceding siblings ...) 2017-05-27 3:47 ` [dpdk-dev] [PATCH 06/10] net/i40e/base: add new phy types for 25G AOC and ACC Jingjing Wu @ 2017-05-27 3:47 ` Jingjing Wu 2017-05-27 3:47 ` [dpdk-dev] [PATCH 08/10] net/i40e/base: track id can be 0 Jingjing Wu ` (4 subsequent siblings) 11 siblings, 0 replies; 69+ messages in thread From: Jingjing Wu @ 2017-05-27 3:47 UTC (permalink / raw) To: helin.zhang; +Cc: dev, jingjing.wu Signed-off-by: Jingjing Wu <jingjing.wu@intel.com> --- drivers/net/i40e/base/i40e_adminq_cmd.h | 20 ++++++++++++++++++-- drivers/net/i40e/base/i40e_common.c | 18 +++++++++++++++--- 2 files changed, 33 insertions(+), 5 deletions(-) diff --git a/drivers/net/i40e/base/i40e_adminq_cmd.h b/drivers/net/i40e/base/i40e_adminq_cmd.h index 70079a0..03a7b28 100644 --- a/drivers/net/i40e/base/i40e_adminq_cmd.h +++ b/drivers/net/i40e/base/i40e_adminq_cmd.h @@ -1831,6 +1831,8 @@ enum i40e_aq_phy_type { I40E_PHY_TYPE_10GBASE_CR1_CU = 0xB, I40E_PHY_TYPE_10GBASE_AOC = 0xC, I40E_PHY_TYPE_40GBASE_AOC = 0xD, + I40E_PHY_TYPE_UNRECOGNIZED = 0xE, + I40E_PHY_TYPE_UNSUPPORTED = 0xF, I40E_PHY_TYPE_100BASE_TX = 0x11, I40E_PHY_TYPE_1000BASE_T = 0x12, I40E_PHY_TYPE_10GBASE_T = 0x13, @@ -1851,6 +1853,8 @@ enum i40e_aq_phy_type { I40E_PHY_TYPE_25GBASE_LR = 0x22, I40E_PHY_TYPE_25GBASE_AOC = 0x23, I40E_PHY_TYPE_25GBASE_ACC = 0x24, + I40E_PHY_TYPE_EMPTY = 0xFE, + I40E_PHY_TYPE_DEFAULT = 0xFF, I40E_PHY_TYPE_MAX }; @@ -2039,19 +2043,31 @@ struct i40e_aqc_get_link_status { #define I40E_AQ_25G_SERDES_UCODE_ERR 0X04 #define I40E_AQ_25G_NIMB_UCODE_ERR 0X05 u8 loopback; /* use defines from i40e_aqc_set_lb_mode */ +/* Since firmware API 1.7 loopback field keeps power class info as well */ +#define I40E_AQ_LOOPBACK_MASK 0x07 +#define I40E_AQ_PWR_CLASS_SHIFT_LB 6 +#define I40E_AQ_PWR_CLASS_MASK_LB (0x03 << I40E_AQ_PWR_CLASS_SHIFT_LB) __le16 max_frame_size; u8 config; #define I40E_AQ_CONFIG_FEC_KR_ENA 0x01 #define I40E_AQ_CONFIG_FEC_RS_ENA 0x02 #define I40E_AQ_CONFIG_CRC_ENA 0x04 #define I40E_AQ_CONFIG_PACING_MASK 0x78 - u8 power_desc; + union { + struct { + u8 power_desc; #define I40E_AQ_LINK_POWER_CLASS_1 0x00 #define I40E_AQ_LINK_POWER_CLASS_2 0x01 #define I40E_AQ_LINK_POWER_CLASS_3 0x02 #define I40E_AQ_LINK_POWER_CLASS_4 0x03 #define I40E_AQ_PWR_CLASS_MASK 0x03 - u8 reserved[4]; + u8 reserved[4]; + }; + struct { + u8 link_type[4]; + u8 link_type_ext; + }; + }; }; I40E_CHECK_CMD_LENGTH(i40e_aqc_get_link_status); diff --git a/drivers/net/i40e/base/i40e_common.c b/drivers/net/i40e/base/i40e_common.c index 004c062..bb4b7eb 100644 --- a/drivers/net/i40e/base/i40e_common.c +++ b/drivers/net/i40e/base/i40e_common.c @@ -1692,8 +1692,14 @@ enum i40e_status_code i40e_aq_get_phy_capabilities(struct i40e_hw *hw, status = I40E_ERR_UNKNOWN_PHY; if (report_init) { - hw->phy.phy_types = LE32_TO_CPU(abilities->phy_type); - hw->phy.phy_types |= ((u64)abilities->phy_type_ext << 32); + if (hw->aq.api_maj_ver == I40E_FW_API_VERSION_MAJOR && + hw->aq.api_min_ver >= 7) { + status = i40e_aq_get_link_info(hw, true, NULL, NULL); + } else { + hw->phy.phy_types = LE32_TO_CPU(abilities->phy_type); + hw->phy.phy_types |= + ((u64)abilities->phy_type_ext << 32); + } } return status; @@ -1955,7 +1961,7 @@ enum i40e_status_code i40e_aq_get_link_info(struct i40e_hw *hw, hw_link_info->fec_info = resp->config & (I40E_AQ_CONFIG_FEC_KR_ENA | I40E_AQ_CONFIG_FEC_RS_ENA); hw_link_info->ext_info = resp->ext_info; - hw_link_info->loopback = resp->loopback; + hw_link_info->loopback = resp->loopback & I40E_AQ_LOOPBACK_MASK; hw_link_info->max_frame_size = LE16_TO_CPU(resp->max_frame_size); hw_link_info->pacing = resp->config & I40E_AQ_CONFIG_PACING_MASK; @@ -1986,6 +1992,12 @@ enum i40e_status_code i40e_aq_get_link_info(struct i40e_hw *hw, hw->aq.fw_min_ver < 40)) && hw_link_info->phy_type == 0xE) hw_link_info->phy_type = I40E_PHY_TYPE_10GBASE_SFPP_CU; + if (hw->aq.api_maj_ver == I40E_FW_API_VERSION_MAJOR && + hw->aq.api_min_ver >= 7) { + hw->phy.phy_types = LE32_TO_CPU(*(__le32 *)resp->link_type); + hw->phy.phy_types |= ((u64)resp->link_type_ext << 32); + } + /* save link status information */ if (link) i40e_memcpy(link, hw_link_info, sizeof(*hw_link_info), -- 2.4.11 ^ permalink raw reply [flat|nested] 69+ messages in thread
* [dpdk-dev] [PATCH 08/10] net/i40e/base: track id can be 0 2017-05-27 3:47 [dpdk-dev] [PATCH 00/10] net/i40e: base code update Jingjing Wu ` (6 preceding siblings ...) 2017-05-27 3:47 ` [dpdk-dev] [PATCH 07/10] net/i40e/base: report supported link modes Jingjing Wu @ 2017-05-27 3:47 ` Jingjing Wu 2017-05-27 3:47 ` [dpdk-dev] [PATCH 09/10] net/i40e/base: update FW AQ API version to 1.7 Jingjing Wu ` (3 subsequent siblings) 11 siblings, 0 replies; 69+ messages in thread From: Jingjing Wu @ 2017-05-27 3:47 UTC (permalink / raw) To: helin.zhang; +Cc: dev, jingjing.wu track_id == 0 is valid for “read only” profiles when profile does not have any “write” commands. Signed-off-by: Jingjing Wu <jingjing.wu@intel.com> --- drivers/net/i40e/base/i40e_common.c | 7 +------ drivers/net/i40e/base/i40e_prototype.h | 12 +++++++----- 2 files changed, 8 insertions(+), 11 deletions(-) diff --git a/drivers/net/i40e/base/i40e_common.c b/drivers/net/i40e/base/i40e_common.c index bb4b7eb..38c30a1 100644 --- a/drivers/net/i40e/base/i40e_common.c +++ b/drivers/net/i40e/base/i40e_common.c @@ -7370,11 +7370,6 @@ i40e_write_profile(struct i40e_hw *hw, struct i40e_profile_segment *profile, u32 offset = 0, info = 0; u32 i; - if (!track_id) { - i40e_debug(hw, I40E_DEBUG_PACKAGE, "Track_id can't be 0."); - return I40E_NOT_SUPPORTED; - } - dev_cnt = profile->device_table_count; for (i = 0; i < dev_cnt; i++) { @@ -7449,6 +7444,6 @@ i40e_add_pinfo_to_list(struct i40e_hw *hw, memcpy(pinfo->name, profile->name, I40E_DDP_NAME_SIZE); status = i40e_aq_write_ddp(hw, (void *)sec, sec->data_end, - track_id, &offset, &info, NULL); + track_id, &offset, &info, NULL); return status; } diff --git a/drivers/net/i40e/base/i40e_prototype.h b/drivers/net/i40e/base/i40e_prototype.h index 54b6750..9171e97 100644 --- a/drivers/net/i40e/base/i40e_prototype.h +++ b/drivers/net/i40e/base/i40e_prototype.h @@ -575,12 +575,14 @@ u8 i40e_get_phy_address(struct i40e_hw *hw, u8 dev_num); enum i40e_status_code i40e_blink_phy_link_led(struct i40e_hw *hw, u32 time, u32 interval); enum i40e_status_code i40e_aq_write_ddp(struct i40e_hw *hw, void *buff, - u16 buff_size, u32 track_id, - u32 *error_offset, u32 *error_info, - struct i40e_asq_cmd_details *cmd_details); + u16 buff_size, u32 track_id, + u32 *error_offset, u32 *error_info, + struct i40e_asq_cmd_details * + cmd_details); enum i40e_status_code i40e_aq_get_ddp_list(struct i40e_hw *hw, void *buff, - u16 buff_size, u8 flags, - struct i40e_asq_cmd_details *cmd_details); + u16 buff_size, u8 flags, + struct i40e_asq_cmd_details * + cmd_details); struct i40e_generic_seg_header * i40e_find_segment_in_package(u32 segment_type, struct i40e_package_header *pkg_header); -- 2.4.11 ^ permalink raw reply [flat|nested] 69+ messages in thread
* [dpdk-dev] [PATCH 09/10] net/i40e/base: update FW AQ API version to 1.7 2017-05-27 3:47 [dpdk-dev] [PATCH 00/10] net/i40e: base code update Jingjing Wu ` (7 preceding siblings ...) 2017-05-27 3:47 ` [dpdk-dev] [PATCH 08/10] net/i40e/base: track id can be 0 Jingjing Wu @ 2017-05-27 3:47 ` Jingjing Wu 2017-05-27 3:47 ` [dpdk-dev] [PATCH 10/10] net/i40e/base: update base code info Jingjing Wu ` (2 subsequent siblings) 11 siblings, 0 replies; 69+ messages in thread From: Jingjing Wu @ 2017-05-27 3:47 UTC (permalink / raw) To: helin.zhang; +Cc: dev, jingjing.wu Signed-off-by: Jingjing Wu <jingjing.wu@intel.com> --- drivers/net/i40e/base/i40e_type.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/net/i40e/base/i40e_type.h b/drivers/net/i40e/base/i40e_type.h index af5347b..3892f52 100644 --- a/drivers/net/i40e/base/i40e_type.h +++ b/drivers/net/i40e/base/i40e_type.h @@ -200,7 +200,7 @@ enum i40e_memcpy_type { }; #define I40E_FW_API_VERSION_MINOR_X722 0x0005 -#define I40E_FW_API_VERSION_MINOR_X710 0x0005 +#define I40E_FW_API_VERSION_MINOR_X710 0x0007 /* These are structs for managing the hardware information and the operations. -- 2.4.11 ^ permalink raw reply [flat|nested] 69+ messages in thread
* [dpdk-dev] [PATCH 10/10] net/i40e/base: update base code info 2017-05-27 3:47 [dpdk-dev] [PATCH 00/10] net/i40e: base code update Jingjing Wu ` (8 preceding siblings ...) 2017-05-27 3:47 ` [dpdk-dev] [PATCH 09/10] net/i40e/base: update FW AQ API version to 1.7 Jingjing Wu @ 2017-05-27 3:47 ` Jingjing Wu 2017-05-30 11:10 ` [dpdk-dev] [PATCH 00/10] net/i40e: base code update Ferruh Yigit 2017-06-26 9:48 ` [dpdk-dev] [PATCH v2 00/16] " Jingjing Wu 11 siblings, 0 replies; 69+ messages in thread From: Jingjing Wu @ 2017-05-27 3:47 UTC (permalink / raw) To: helin.zhang; +Cc: dev, jingjing.wu Signed-off-by: Jingjing Wu <jingjing.wu@intel.com> --- drivers/net/i40e/base/README | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/net/i40e/base/README b/drivers/net/i40e/base/README index 0da9f67..270d7f1 100644 --- a/drivers/net/i40e/base/README +++ b/drivers/net/i40e/base/README @@ -34,7 +34,7 @@ Intel® I40E driver ================== This directory contains source code of FreeBSD i40e driver of version -cid-i40e.2017.03.21.tar.gz released by the team which develops +cid-i40e.2017.05.24.tar.gz released by the team which develops basic drivers for any i40e NIC. The directory of base/ contains the original source package. This driver is valid for the product(s) listed below -- 2.4.11 ^ permalink raw reply [flat|nested] 69+ messages in thread
* Re: [dpdk-dev] [PATCH 00/10] net/i40e: base code update 2017-05-27 3:47 [dpdk-dev] [PATCH 00/10] net/i40e: base code update Jingjing Wu ` (9 preceding siblings ...) 2017-05-27 3:47 ` [dpdk-dev] [PATCH 10/10] net/i40e/base: update base code info Jingjing Wu @ 2017-05-30 11:10 ` Ferruh Yigit 2017-06-02 2:24 ` Wu, Jingjing 2017-06-26 9:48 ` [dpdk-dev] [PATCH v2 00/16] " Jingjing Wu 11 siblings, 1 reply; 69+ messages in thread From: Ferruh Yigit @ 2017-05-30 11:10 UTC (permalink / raw) To: Jingjing Wu, helin.zhang; +Cc: dev On 5/27/2017 4:47 AM, Jingjing Wu wrote: > i40e base code upate. The main changes are: > - use virtchnl.h instead of i40e_virtchnl.h > - add support for Adaptive Virtual Function > - add new AQ commands for read/write PHY registers > - add new phy types for 25G Hi Jingjing, Patchset giving build error with clang [1] because of assignment between different enum types, can you please check the build errors? [1] .../drivers/net/i40e/i40e_pf.c:349:32: error: implicit conversion from enumeration type 'enum i40e_vsi_type' to different enumeration type 'enum virtchnl_vsi_type' [-Werror,-Wenum-conversion] vf_res->vsi_res[0].vsi_type = I40E_VSI_SRIOV; ~ ^~~~~~~~~~~~~~ .../drivers/net/i40e/i40e_pf.c:1195:44: error: implicit conversion from enumeration type 'enum i40e_aq_link_speed' to different enumeration type 'enum virtchnl_link_speed' [-Werror,-Wenum-conversion] event.event_data.link_event.link_speed = I40E_LINK_SPEED_100MB; ~ ^~~~~~~~~~~~~~~~~~~~~ .../drivers/net/i40e/i40e_pf.c:1198:44: error: implicit conversion from enumeration type 'enum i40e_aq_link_speed' to different enumeration type 'enum virtchnl_link_speed' [-Werror,-Wenum-conversion] event.event_data.link_event.link_speed = I40E_LINK_SPEED_1GB; ~ ^~~~~~~~~~~~~~~~~~~ .../drivers/net/i40e/i40e_pf.c:1201:44: error: implicit conversion from enumeration type 'enum i40e_aq_link_speed' to different enumeration type 'enum virtchnl_link_speed' [-Werror,-Wenum-conversion] event.event_data.link_event.link_speed = I40E_LINK_SPEED_10GB; ~ ^~~~~~~~~~~~~~~~~~~~ .../drivers/net/i40e/i40e_pf.c:1204:44: error: implicit conversion from enumeration type 'enum i40e_aq_link_speed' to different enumeration type 'enum virtchnl_link_speed' [-Werror,-Wenum-conversion] event.event_data.link_event.link_speed = I40E_LINK_SPEED_20GB; ~ ^~~~~~~~~~~~~~~~~~~~ .../drivers/net/i40e/i40e_pf.c:1207:44: error: implicit conversion from enumeration type 'enum i40e_aq_link_speed' to different enumeration type 'enum virtchnl_link_speed' [-Werror,-Wenum-conversion] event.event_data.link_event.link_speed = I40E_LINK_SPEED_25GB; ~ ^~~~~~~~~~~~~~~~~~~~ .../drivers/net/i40e/i40e_pf.c:1210:44: error: implicit conversion from enumeration type 'enum i40e_aq_link_speed' to different enumeration type 'enum virtchnl_link_speed' [-Werror,-Wenum-conversion] event.event_data.link_event.link_speed = I40E_LINK_SPEED_40GB; ~ ^~~~~~~~~~~~~~~~~~~~ .../drivers/net/i40e/i40e_pf.c:1214:4: error: implicit conversion from enumeration type 'enum i40e_aq_link_speed' to different enumeration type 'enum virtchnl_link_speed' [-Werror,-Wenum-conversion] I40E_LINK_SPEED_UNKNOWN; ^~~~~~~~~~~~~~~~~~~~~~~ 8 errors generated. .../mk/internal/rte.compile-pre.mk:138: recipe for target 'i40e_pf.o' failed make[4]: *** [i40e_pf.o] Error 1 make[4]: *** Waiting for unfinished jobs.... .../drivers/net/i40e/i40e_ethdev_vf.c:275:32: error: implicit conversion from enumeration type 'enum virtchnl_link_speed' to different enumeration type 'enum i40e_aq_link_speed' [-Werror,-Wenum-conversion] vpe->event_data.link_event.link_speed; ~~~~~~~~~~~~~~~~~~~~~~~~~~~^~~~~~~~~~ .../drivers/net/i40e/i40e_ethdev_vf.c:1274:30: error: implicit conversion from enumeration type 'enum virtchnl_vsi_type' to different enumeration type 'enum i40e_vsi_type' [-Werror,-Wenum-conversion] vf->vsi.type = vf->vsi_res->vsi_type; ~ ~~~~~~~~~~~~~^~~~~~~~ .../drivers/net/i40e/i40e_ethdev_vf.c:1339:50: error: implicit conversion from enumeration type 'enum virtchnl_link_speed' to different enumeration type 'enum i40e_aq_link_speed' [-Werror,-Wenum-conversion] vf->link_speed = pf_msg->event_data.link_event.link_speed; ~ ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~^~~~~~~~~~ <...> ^ permalink raw reply [flat|nested] 69+ messages in thread
* Re: [dpdk-dev] [PATCH 00/10] net/i40e: base code update 2017-05-30 11:10 ` [dpdk-dev] [PATCH 00/10] net/i40e: base code update Ferruh Yigit @ 2017-06-02 2:24 ` Wu, Jingjing 0 siblings, 0 replies; 69+ messages in thread From: Wu, Jingjing @ 2017-06-02 2:24 UTC (permalink / raw) To: Yigit, Ferruh, Zhang, Helin; +Cc: dev > -----Original Message----- > From: Yigit, Ferruh > Sent: Tuesday, May 30, 2017 7:10 PM > To: Wu, Jingjing <jingjing.wu@intel.com>; Zhang, Helin > <helin.zhang@intel.com> > Cc: dev@dpdk.org > Subject: Re: [dpdk-dev] [PATCH 00/10] net/i40e: base code update > > On 5/27/2017 4:47 AM, Jingjing Wu wrote: > > i40e base code upate. The main changes are: > > - use virtchnl.h instead of i40e_virtchnl.h > > - add support for Adaptive Virtual Function > > - add new AQ commands for read/write PHY registers > > - add new phy types for 25G > > Hi Jingjing, > > Patchset giving build error with clang [1] because of assignment between > different enum types, can you please check the build errors? OK, will fix it in v2 patch. Thanks ^ permalink raw reply [flat|nested] 69+ messages in thread
* [dpdk-dev] [PATCH v2 00/16] net/i40e: base code update 2017-05-27 3:47 [dpdk-dev] [PATCH 00/10] net/i40e: base code update Jingjing Wu ` (10 preceding siblings ...) 2017-05-30 11:10 ` [dpdk-dev] [PATCH 00/10] net/i40e: base code update Ferruh Yigit @ 2017-06-26 9:48 ` Jingjing Wu 2017-06-26 9:48 ` [dpdk-dev] [PATCH v2 01/16] net/i40e/base: use new virtchnl header file Jingjing Wu ` (16 more replies) 11 siblings, 17 replies; 69+ messages in thread From: Jingjing Wu @ 2017-06-26 9:48 UTC (permalink / raw) To: helin.zhang; +Cc: dev, jingjing.wu v2 changes: - Fix compile issue - update from package from 2017.05.24 to 2017.06.23 i40e base code upate. The main changes are: - use virtchnl.h instead of i40e_virtchnl.h - add support for Adaptive Virtual Function - add new AQ commands for read/write PHY registers - add new phy types for 25G - extend processing of DDP Jingjing Wu (16): net/i40e/base: use new virtchnl header file net/i40e/base: sync nvmupdate command and adminq subtask net/i40e/base: add AQ command for read/write PHY registers net/i40e/base: add support for Adaptive Virtual Function net/i40e/base: store the requested FEC information net/i40e/base: add new phy types for 25G AOC and ACC net/i40e/base: report supported link modes net/i40e/base: track id can be 0 net/i40e/base: update FW AQ API version to 1.7 net/i40e/base: add support for switch parameters net/i40e/base: use admin queue for setting LEDs behavior net/i40e/base: avoid potential null pointer dereference net/i40e/base: avoid reset timeout issue net/i40e/base: add EEPROM checksum verification net/i40e/base: extend processing of DDP net/i40e: use set switch aq instead of register setting drivers/net/i40e/base/README | 2 +- drivers/net/i40e/base/i40e_adminq.c | 12 + drivers/net/i40e/base/i40e_adminq_cmd.h | 71 ++- drivers/net/i40e/base/i40e_common.c | 549 +++++++++++++++++++---- drivers/net/i40e/base/i40e_devids.h | 1 + drivers/net/i40e/base/i40e_nvm.c | 20 +- drivers/net/i40e/base/i40e_prototype.h | 33 +- drivers/net/i40e/base/i40e_register.h | 2 +- drivers/net/i40e/base/i40e_type.h | 49 +- drivers/net/i40e/base/i40e_virtchnl.h | 445 ------------------ drivers/net/i40e/base/virtchnl.h | 772 ++++++++++++++++++++++++++++++++ drivers/net/i40e/i40e_ethdev.c | 104 +++-- drivers/net/i40e/i40e_ethdev.h | 8 +- drivers/net/i40e/i40e_ethdev_vf.c | 206 ++++----- drivers/net/i40e/i40e_pf.c | 206 ++++----- drivers/net/i40e/i40e_pf.h | 32 +- drivers/net/i40e/rte_pmd_i40e.h | 2 +- 17 files changed, 1694 insertions(+), 820 deletions(-) delete mode 100644 drivers/net/i40e/base/i40e_virtchnl.h create mode 100644 drivers/net/i40e/base/virtchnl.h -- 2.4.11 ^ permalink raw reply [flat|nested] 69+ messages in thread
* [dpdk-dev] [PATCH v2 01/16] net/i40e/base: use new virtchnl header file 2017-06-26 9:48 ` [dpdk-dev] [PATCH v2 00/16] " Jingjing Wu @ 2017-06-26 9:48 ` Jingjing Wu 2017-06-26 9:48 ` [dpdk-dev] [PATCH v2 02/16] net/i40e/base: sync nvmupdate command and adminq subtask Jingjing Wu ` (15 subsequent siblings) 16 siblings, 0 replies; 69+ messages in thread From: Jingjing Wu @ 2017-06-26 9:48 UTC (permalink / raw) To: helin.zhang; +Cc: dev, jingjing.wu Modify the necessary files to be compatible with the new virtchnl.h file instead of relying on i40e_virtchnl.h variant. This mostly changes references to VIRTCHNL_ variables by removing prefix of I40E_. Signed-off-by: Jingjing Wu <jingjing.wu@intel.com> --- drivers/net/i40e/base/i40e_common.c | 24 +- drivers/net/i40e/base/i40e_prototype.h | 6 +- drivers/net/i40e/base/i40e_type.h | 12 +- drivers/net/i40e/base/i40e_virtchnl.h | 445 ------------------- drivers/net/i40e/base/virtchnl.h | 772 +++++++++++++++++++++++++++++++++ drivers/net/i40e/i40e_ethdev.h | 8 +- drivers/net/i40e/i40e_ethdev_vf.c | 206 ++++----- drivers/net/i40e/i40e_pf.c | 206 ++++----- drivers/net/i40e/i40e_pf.h | 32 +- drivers/net/i40e/rte_pmd_i40e.h | 2 +- 10 files changed, 1020 insertions(+), 693 deletions(-) delete mode 100644 drivers/net/i40e/base/i40e_virtchnl.h create mode 100644 drivers/net/i40e/base/virtchnl.h diff --git a/drivers/net/i40e/base/i40e_common.c b/drivers/net/i40e/base/i40e_common.c index 03e94bc..6c09c27 100644 --- a/drivers/net/i40e/base/i40e_common.c +++ b/drivers/net/i40e/base/i40e_common.c @@ -34,7 +34,7 @@ POSSIBILITY OF SUCH DAMAGE. #include "i40e_type.h" #include "i40e_adminq.h" #include "i40e_prototype.h" -#include "i40e_virtchnl.h" +#include "virtchnl.h" /** @@ -5523,7 +5523,7 @@ enum i40e_status_code i40e_aq_add_rem_control_packet_filter(struct i40e_hw *hw, } if (mac_addr) - i40e_memcpy(cmd->mac, mac_addr, I40E_ETH_LENGTH_OF_ADDRESS, + i40e_memcpy(cmd->mac, mac_addr, ETH_ALEN, I40E_NONDMA_TO_NONDMA); cmd->etype = CPU_TO_LE16(ethtype); @@ -6879,7 +6879,7 @@ void i40e_write_rx_ctl(struct i40e_hw *hw, u32 reg_addr, u32 reg_val) * completion before returning. **/ enum i40e_status_code i40e_aq_send_msg_to_pf(struct i40e_hw *hw, - enum i40e_virtchnl_ops v_opcode, + enum virtchnl_ops v_opcode, enum i40e_status_code v_retval, u8 *msg, u16 msglen, struct i40e_asq_cmd_details *cmd_details) @@ -6918,9 +6918,9 @@ enum i40e_status_code i40e_aq_send_msg_to_pf(struct i40e_hw *hw, * with appropriate information. **/ void i40e_vf_parse_hw_config(struct i40e_hw *hw, - struct i40e_virtchnl_vf_resource *msg) + struct virtchnl_vf_resource *msg) { - struct i40e_virtchnl_vsi_resource *vsi_res; + struct virtchnl_vsi_resource *vsi_res; int i; vsi_res = &msg->vsi_res[0]; @@ -6930,19 +6930,17 @@ void i40e_vf_parse_hw_config(struct i40e_hw *hw, hw->dev_caps.num_tx_qp = msg->num_queue_pairs; hw->dev_caps.num_msix_vectors_vf = msg->max_vectors; hw->dev_caps.dcb = msg->vf_offload_flags & - I40E_VIRTCHNL_VF_OFFLOAD_L2; - hw->dev_caps.fcoe = (msg->vf_offload_flags & - I40E_VIRTCHNL_VF_OFFLOAD_FCOE) ? 1 : 0; + VIRTCHNL_VF_OFFLOAD_L2; hw->dev_caps.iwarp = (msg->vf_offload_flags & - I40E_VIRTCHNL_VF_OFFLOAD_IWARP) ? 1 : 0; + VIRTCHNL_VF_OFFLOAD_IWARP) ? 1 : 0; for (i = 0; i < msg->num_vsis; i++) { - if (vsi_res->vsi_type == I40E_VSI_SRIOV) { + if (vsi_res->vsi_type == VIRTCHNL_VSI_SRIOV) { i40e_memcpy(hw->mac.perm_addr, vsi_res->default_mac_addr, - I40E_ETH_LENGTH_OF_ADDRESS, + ETH_ALEN, I40E_NONDMA_TO_NONDMA); i40e_memcpy(hw->mac.addr, vsi_res->default_mac_addr, - I40E_ETH_LENGTH_OF_ADDRESS, + ETH_ALEN, I40E_NONDMA_TO_NONDMA); } vsi_res++; @@ -6959,7 +6957,7 @@ void i40e_vf_parse_hw_config(struct i40e_hw *hw, **/ enum i40e_status_code i40e_vf_reset(struct i40e_hw *hw) { - return i40e_aq_send_msg_to_pf(hw, I40E_VIRTCHNL_OP_RESET_VF, + return i40e_aq_send_msg_to_pf(hw, VIRTCHNL_OP_RESET_VF, I40E_SUCCESS, NULL, 0, NULL); } #endif /* VF_DRIVER */ diff --git a/drivers/net/i40e/base/i40e_prototype.h b/drivers/net/i40e/base/i40e_prototype.h index 4bd589e..6ec4304 100644 --- a/drivers/net/i40e/base/i40e_prototype.h +++ b/drivers/net/i40e/base/i40e_prototype.h @@ -36,7 +36,7 @@ POSSIBILITY OF SUCH DAMAGE. #include "i40e_type.h" #include "i40e_alloc.h" -#include "i40e_virtchnl.h" +#include "virtchnl.h" /* Prototypes for shared code functions that are not in * the standard function pointer structures. These are @@ -504,10 +504,10 @@ void i40e_destroy_spinlock(struct i40e_spinlock *sp); /* i40e_common for VF drivers*/ void i40e_vf_parse_hw_config(struct i40e_hw *hw, - struct i40e_virtchnl_vf_resource *msg); + struct virtchnl_vf_resource *msg); enum i40e_status_code i40e_vf_reset(struct i40e_hw *hw); enum i40e_status_code i40e_aq_send_msg_to_pf(struct i40e_hw *hw, - enum i40e_virtchnl_ops v_opcode, + enum virtchnl_ops v_opcode, enum i40e_status_code v_retval, u8 *msg, u16 msglen, struct i40e_asq_cmd_details *cmd_details); diff --git a/drivers/net/i40e/base/i40e_type.h b/drivers/net/i40e/base/i40e_type.h index 84d5757..52a114a 100644 --- a/drivers/net/i40e/base/i40e_type.h +++ b/drivers/net/i40e/base/i40e_type.h @@ -92,7 +92,9 @@ POSSIBILITY OF SUCH DAMAGE. struct i40e_hw; typedef void (*I40E_ADMINQ_CALLBACK)(struct i40e_hw *, struct i40e_aq_desc *); -#define I40E_ETH_LENGTH_OF_ADDRESS 6 +#ifndef ETH_ALEN +#define ETH_ALEN 6 +#endif /* Data type manipulation macros. */ #define I40E_HI_DWORD(x) ((u32)((((x) >> 16) >> 16) & 0xFFFFFFFF)) #define I40E_LO_DWORD(x) ((u32)((x) & 0xFFFFFFFF)) @@ -439,10 +441,10 @@ struct i40e_hw_capabilities { struct i40e_mac_info { enum i40e_mac_type type; - u8 addr[I40E_ETH_LENGTH_OF_ADDRESS]; - u8 perm_addr[I40E_ETH_LENGTH_OF_ADDRESS]; - u8 san_addr[I40E_ETH_LENGTH_OF_ADDRESS]; - u8 port_addr[I40E_ETH_LENGTH_OF_ADDRESS]; + u8 addr[ETH_ALEN]; + u8 perm_addr[ETH_ALEN]; + u8 san_addr[ETH_ALEN]; + u8 port_addr[ETH_ALEN]; u16 max_fcoeq; }; diff --git a/drivers/net/i40e/base/i40e_virtchnl.h b/drivers/net/i40e/base/i40e_virtchnl.h deleted file mode 100644 index 7a24c0f..0000000 --- a/drivers/net/i40e/base/i40e_virtchnl.h +++ /dev/null @@ -1,445 +0,0 @@ -/******************************************************************************* - -Copyright (c) 2013 - 2015, Intel Corporation -All rights reserved. - -Redistribution and use in source and binary forms, with or without -modification, are permitted provided that the following conditions are met: - - 1. Redistributions of source code must retain the above copyright notice, - this list of conditions and the following disclaimer. - - 2. Redistributions in binary form must reproduce the above copyright - notice, this list of conditions and the following disclaimer in the - documentation and/or other materials provided with the distribution. - - 3. Neither the name of the Intel Corporation nor the names of its - contributors may be used to endorse or promote products derived from - this software without specific prior written permission. - -THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS" -AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE -IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE -ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE -LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR -CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF -SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS -INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN -CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) -ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE -POSSIBILITY OF SUCH DAMAGE. - -***************************************************************************/ - -#ifndef _I40E_VIRTCHNL_H_ -#define _I40E_VIRTCHNL_H_ - -#include "i40e_type.h" - -/* Description: - * This header file describes the VF-PF communication protocol used - * by the various i40e drivers. - * - * Admin queue buffer usage: - * desc->opcode is always i40e_aqc_opc_send_msg_to_pf - * flags, retval, datalen, and data addr are all used normally. - * Firmware copies the cookie fields when sending messages between the PF and - * VF, but uses all other fields internally. Due to this limitation, we - * must send all messages as "indirect", i.e. using an external buffer. - * - * All the vsi indexes are relative to the VF. Each VF can have maximum of - * three VSIs. All the queue indexes are relative to the VSI. Each VF can - * have a maximum of sixteen queues for all of its VSIs. - * - * The PF is required to return a status code in v_retval for all messages - * except RESET_VF, which does not require any response. The return value is of - * i40e_status_code type, defined in the i40e_type.h. - * - * In general, VF driver initialization should roughly follow the order of these - * opcodes. The VF driver must first validate the API version of the PF driver, - * then request a reset, then get resources, then configure queues and - * interrupts. After these operations are complete, the VF driver may start - * its queues, optionally add MAC and VLAN filters, and process traffic. - */ - -/* Opcodes for VF-PF communication. These are placed in the v_opcode field - * of the virtchnl_msg structure. - */ -enum i40e_virtchnl_ops { -/* The PF sends status change events to VFs using - * the I40E_VIRTCHNL_OP_EVENT opcode. - * VFs send requests to the PF using the other ops. - */ - I40E_VIRTCHNL_OP_UNKNOWN = 0, - I40E_VIRTCHNL_OP_VERSION = 1, /* must ALWAYS be 1 */ - I40E_VIRTCHNL_OP_RESET_VF = 2, - I40E_VIRTCHNL_OP_GET_VF_RESOURCES = 3, - I40E_VIRTCHNL_OP_CONFIG_TX_QUEUE = 4, - I40E_VIRTCHNL_OP_CONFIG_RX_QUEUE = 5, - I40E_VIRTCHNL_OP_CONFIG_VSI_QUEUES = 6, - I40E_VIRTCHNL_OP_CONFIG_IRQ_MAP = 7, - I40E_VIRTCHNL_OP_ENABLE_QUEUES = 8, - I40E_VIRTCHNL_OP_DISABLE_QUEUES = 9, - I40E_VIRTCHNL_OP_ADD_ETHER_ADDRESS = 10, - I40E_VIRTCHNL_OP_DEL_ETHER_ADDRESS = 11, - I40E_VIRTCHNL_OP_ADD_VLAN = 12, - I40E_VIRTCHNL_OP_DEL_VLAN = 13, - I40E_VIRTCHNL_OP_CONFIG_PROMISCUOUS_MODE = 14, - I40E_VIRTCHNL_OP_GET_STATS = 15, - I40E_VIRTCHNL_OP_FCOE = 16, - I40E_VIRTCHNL_OP_EVENT = 17, /* must ALWAYS be 17 */ -#ifdef I40E_SOL_VF_SUPPORT - I40E_VIRTCHNL_OP_GET_ADDNL_SOL_CONFIG = 19, -#endif - I40E_VIRTCHNL_OP_CONFIG_RSS_KEY = 23, - I40E_VIRTCHNL_OP_CONFIG_RSS_LUT = 24, - I40E_VIRTCHNL_OP_GET_RSS_HENA_CAPS = 25, - I40E_VIRTCHNL_OP_SET_RSS_HENA = 26, - -}; - -/* Virtual channel message descriptor. This overlays the admin queue - * descriptor. All other data is passed in external buffers. - */ - -struct i40e_virtchnl_msg { - u8 pad[8]; /* AQ flags/opcode/len/retval fields */ - enum i40e_virtchnl_ops v_opcode; /* avoid confusion with desc->opcode */ - enum i40e_status_code v_retval; /* ditto for desc->retval */ - u32 vfid; /* used by PF when sending to VF */ -}; - -/* Message descriptions and data structures.*/ - -/* I40E_VIRTCHNL_OP_VERSION - * VF posts its version number to the PF. PF responds with its version number - * in the same format, along with a return code. - * Reply from PF has its major/minor versions also in param0 and param1. - * If there is a major version mismatch, then the VF cannot operate. - * If there is a minor version mismatch, then the VF can operate but should - * add a warning to the system log. - * - * This enum element MUST always be specified as == 1, regardless of other - * changes in the API. The PF must always respond to this message without - * error regardless of version mismatch. - */ -#define I40E_VIRTCHNL_VERSION_MAJOR 1 -#define I40E_VIRTCHNL_VERSION_MINOR 1 -#define I40E_VIRTCHNL_VERSION_MINOR_NO_VF_CAPS 0 - -struct i40e_virtchnl_version_info { - u32 major; - u32 minor; -}; - -/* I40E_VIRTCHNL_OP_RESET_VF - * VF sends this request to PF with no parameters - * PF does NOT respond! VF driver must delay then poll VFGEN_RSTAT register - * until reset completion is indicated. The admin queue must be reinitialized - * after this operation. - * - * When reset is complete, PF must ensure that all queues in all VSIs associated - * with the VF are stopped, all queue configurations in the HMC are set to 0, - * and all MAC and VLAN filters (except the default MAC address) on all VSIs - * are cleared. - */ - -/* I40E_VIRTCHNL_OP_GET_VF_RESOURCES - * Version 1.0 VF sends this request to PF with no parameters - * Version 1.1 VF sends this request to PF with u32 bitmap of its capabilities - * PF responds with an indirect message containing - * i40e_virtchnl_vf_resource and one or more - * i40e_virtchnl_vsi_resource structures. - */ - -struct i40e_virtchnl_vsi_resource { - u16 vsi_id; - u16 num_queue_pairs; - enum i40e_vsi_type vsi_type; - u16 qset_handle; - u8 default_mac_addr[I40E_ETH_LENGTH_OF_ADDRESS]; -}; -/* VF offload flags */ -#define I40E_VIRTCHNL_VF_OFFLOAD_L2 0x00000001 -#define I40E_VIRTCHNL_VF_OFFLOAD_IWARP 0x00000002 -#define I40E_VIRTCHNL_VF_OFFLOAD_FCOE 0x00000004 -#define I40E_VIRTCHNL_VF_OFFLOAD_RSS_AQ 0x00000008 -#define I40E_VIRTCHNL_VF_OFFLOAD_RSS_REG 0x00000010 -#define I40E_VIRTCHNL_VF_OFFLOAD_WB_ON_ITR 0x00000020 -#define I40E_VIRTCHNL_VF_OFFLOAD_VLAN 0x00010000 -#define I40E_VIRTCHNL_VF_OFFLOAD_RX_POLLING 0x00020000 -#define I40E_VIRTCHNL_VF_OFFLOAD_RSS_PCTYPE_V2 0x00040000 -#define I40E_VIRTCHNL_VF_OFFLOAD_RSS_PF 0X00080000 -#define I40E_VIRTCHNL_VF_OFFLOAD_ENCAP 0X00100000 -#define I40E_VIRTCHNL_VF_OFFLOAD_ENCAP_CSUM 0X00200000 -#define I40E_VIRTCHNL_VF_OFFLOAD_RX_ENCAP_CSUM 0X00400000 - -#define I40E_VF_BASE_MODE_OFFLOADS (I40E_VIRTCHNL_VF_OFFLOAD_L2 | \ - I40E_VIRTCHNL_VF_OFFLOAD_VLAN | \ - I40E_VIRTCHNL_VF_OFFLOAD_RSS_PF) - -struct i40e_virtchnl_vf_resource { - u16 num_vsis; - u16 num_queue_pairs; - u16 max_vectors; - u16 max_mtu; - - u32 vf_offload_flags; - u32 rss_key_size; - u32 rss_lut_size; - - struct i40e_virtchnl_vsi_resource vsi_res[1]; -}; - -/* I40E_VIRTCHNL_OP_CONFIG_TX_QUEUE - * VF sends this message to set up parameters for one TX queue. - * External data buffer contains one instance of i40e_virtchnl_txq_info. - * PF configures requested queue and returns a status code. - */ - -/* Tx queue config info */ -struct i40e_virtchnl_txq_info { - u16 vsi_id; - u16 queue_id; - u16 ring_len; /* number of descriptors, multiple of 8 */ - u16 headwb_enabled; - u64 dma_ring_addr; - u64 dma_headwb_addr; -}; - -/* I40E_VIRTCHNL_OP_CONFIG_RX_QUEUE - * VF sends this message to set up parameters for one RX queue. - * External data buffer contains one instance of i40e_virtchnl_rxq_info. - * PF configures requested queue and returns a status code. - */ - -/* Rx queue config info */ -struct i40e_virtchnl_rxq_info { - u16 vsi_id; - u16 queue_id; - u32 ring_len; /* number of descriptors, multiple of 32 */ - u16 hdr_size; - u16 splithdr_enabled; - u32 databuffer_size; - u32 max_pkt_size; - u64 dma_ring_addr; - enum i40e_hmc_obj_rx_hsplit_0 rx_split_pos; -}; - -/* I40E_VIRTCHNL_OP_CONFIG_VSI_QUEUES - * VF sends this message to set parameters for all active TX and RX queues - * associated with the specified VSI. - * PF configures queues and returns status. - * If the number of queues specified is greater than the number of queues - * associated with the VSI, an error is returned and no queues are configured. - */ -struct i40e_virtchnl_queue_pair_info { - /* NOTE: vsi_id and queue_id should be identical for both queues. */ - struct i40e_virtchnl_txq_info txq; - struct i40e_virtchnl_rxq_info rxq; -}; - -struct i40e_virtchnl_vsi_queue_config_info { - u16 vsi_id; - u16 num_queue_pairs; - struct i40e_virtchnl_queue_pair_info qpair[1]; -}; - -/* I40E_VIRTCHNL_OP_CONFIG_IRQ_MAP - * VF uses this message to map vectors to queues. - * The rxq_map and txq_map fields are bitmaps used to indicate which queues - * are to be associated with the specified vector. - * The "other" causes are always mapped to vector 0. - * PF configures interrupt mapping and returns status. - */ -struct i40e_virtchnl_vector_map { - u16 vsi_id; - u16 vector_id; - u16 rxq_map; - u16 txq_map; - u16 rxitr_idx; - u16 txitr_idx; -}; - -struct i40e_virtchnl_irq_map_info { - u16 num_vectors; - struct i40e_virtchnl_vector_map vecmap[1]; -}; - -/* I40E_VIRTCHNL_OP_ENABLE_QUEUES - * I40E_VIRTCHNL_OP_DISABLE_QUEUES - * VF sends these message to enable or disable TX/RX queue pairs. - * The queues fields are bitmaps indicating which queues to act upon. - * (Currently, we only support 16 queues per VF, but we make the field - * u32 to allow for expansion.) - * PF performs requested action and returns status. - */ -struct i40e_virtchnl_queue_select { - u16 vsi_id; - u16 pad; - u32 rx_queues; - u32 tx_queues; -}; - -/* I40E_VIRTCHNL_OP_ADD_ETHER_ADDRESS - * VF sends this message in order to add one or more unicast or multicast - * address filters for the specified VSI. - * PF adds the filters and returns status. - */ - -/* I40E_VIRTCHNL_OP_DEL_ETHER_ADDRESS - * VF sends this message in order to remove one or more unicast or multicast - * filters for the specified VSI. - * PF removes the filters and returns status. - */ - -struct i40e_virtchnl_ether_addr { - u8 addr[I40E_ETH_LENGTH_OF_ADDRESS]; - u8 pad[2]; -}; - -struct i40e_virtchnl_ether_addr_list { - u16 vsi_id; - u16 num_elements; - struct i40e_virtchnl_ether_addr list[1]; -}; - -#ifdef I40E_SOL_VF_SUPPORT -/* I40E_VIRTCHNL_OP_GET_ADDNL_SOL_CONFIG - * VF sends this message to get the default MTU and list of additional ethernet - * addresses it is allowed to use. - * PF responds with an indirect message containing - * i40e_virtchnl_addnl_solaris_config with zero or more - * i40e_virtchnl_ether_addr structures. - * - * It is expected that this operation will only ever be needed for Solaris VFs - * running under a Solaris PF. - */ -struct i40e_virtchnl_addnl_solaris_config { - u16 default_mtu; - struct i40e_virtchnl_ether_addr_list al; -}; - -#endif -/* I40E_VIRTCHNL_OP_ADD_VLAN - * VF sends this message to add one or more VLAN tag filters for receives. - * PF adds the filters and returns status. - * If a port VLAN is configured by the PF, this operation will return an - * error to the VF. - */ - -/* I40E_VIRTCHNL_OP_DEL_VLAN - * VF sends this message to remove one or more VLAN tag filters for receives. - * PF removes the filters and returns status. - * If a port VLAN is configured by the PF, this operation will return an - * error to the VF. - */ - -struct i40e_virtchnl_vlan_filter_list { - u16 vsi_id; - u16 num_elements; - u16 vlan_id[1]; -}; - -/* I40E_VIRTCHNL_OP_CONFIG_PROMISCUOUS_MODE - * VF sends VSI id and flags. - * PF returns status code in retval. - * Note: we assume that broadcast accept mode is always enabled. - */ -struct i40e_virtchnl_promisc_info { - u16 vsi_id; - u16 flags; -}; - -#define I40E_FLAG_VF_UNICAST_PROMISC 0x00000001 -#define I40E_FLAG_VF_MULTICAST_PROMISC 0x00000002 - -/* I40E_VIRTCHNL_OP_GET_STATS - * VF sends this message to request stats for the selected VSI. VF uses - * the i40e_virtchnl_queue_select struct to specify the VSI. The queue_id - * field is ignored by the PF. - * - * PF replies with struct i40e_eth_stats in an external buffer. - */ - -/* I40E_VIRTCHNL_OP_CONFIG_RSS_KEY - * I40E_VIRTCHNL_OP_CONFIG_RSS_LUT - * VF sends these messages to configure RSS. Only supported if both PF - * and VF drivers set the I40E_VIRTCHNL_VF_OFFLOAD_RSS_PF bit during - * configuration negotiation. If this is the case, then the rss fields in - * the vf resource struct are valid. - * Both the key and LUT are initialized to 0 by the PF, meaning that - * RSS is effectively disabled until set up by the VF. - */ -struct i40e_virtchnl_rss_key { - u16 vsi_id; - u16 key_len; - u8 key[1]; /* RSS hash key, packed bytes */ -}; - -struct i40e_virtchnl_rss_lut { - u16 vsi_id; - u16 lut_entries; - u8 lut[1]; /* RSS lookup table*/ -}; - -/* I40E_VIRTCHNL_OP_GET_RSS_HENA_CAPS - * I40E_VIRTCHNL_OP_SET_RSS_HENA - * VF sends these messages to get and set the hash filter enable bits for RSS. - * By default, the PF sets these to all possible traffic types that the - * hardware supports. The VF can query this value if it wants to change the - * traffic types that are hashed by the hardware. - * Traffic types are defined in the i40e_filter_pctype enum in i40e_type.h - */ -struct i40e_virtchnl_rss_hena { - u64 hena; -}; - -/* I40E_VIRTCHNL_OP_EVENT - * PF sends this message to inform the VF driver of events that may affect it. - * No direct response is expected from the VF, though it may generate other - * messages in response to this one. - */ -enum i40e_virtchnl_event_codes { - I40E_VIRTCHNL_EVENT_UNKNOWN = 0, - I40E_VIRTCHNL_EVENT_LINK_CHANGE, - I40E_VIRTCHNL_EVENT_RESET_IMPENDING, - I40E_VIRTCHNL_EVENT_PF_DRIVER_CLOSE, -}; -#define I40E_PF_EVENT_SEVERITY_INFO 0 -#define I40E_PF_EVENT_SEVERITY_ATTENTION 1 -#define I40E_PF_EVENT_SEVERITY_ACTION_REQUIRED 2 -#define I40E_PF_EVENT_SEVERITY_CERTAIN_DOOM 255 - -struct i40e_virtchnl_pf_event { - enum i40e_virtchnl_event_codes event; - union { - struct { - enum i40e_aq_link_speed link_speed; - bool link_status; - } link_event; - } event_data; - - int severity; -}; - -/* VF reset states - these are written into the RSTAT register: - * I40E_VFGEN_RSTAT1 on the PF - * I40E_VFGEN_RSTAT on the VF - * When the PF initiates a reset, it writes 0 - * When the reset is complete, it writes 1 - * When the PF detects that the VF has recovered, it writes 2 - * VF checks this register periodically to determine if a reset has occurred, - * then polls it to know when the reset is complete. - * If either the PF or VF reads the register while the hardware - * is in a reset state, it will return DEADBEEF, which, when masked - * will result in 3. - */ -enum i40e_vfr_states { - I40E_VFR_INPROGRESS = 0, - I40E_VFR_COMPLETED, - I40E_VFR_VFACTIVE, - I40E_VFR_UNKNOWN, -}; - -#endif /* _I40E_VIRTCHNL_H_ */ diff --git a/drivers/net/i40e/base/virtchnl.h b/drivers/net/i40e/base/virtchnl.h new file mode 100644 index 0000000..f00dd36 --- /dev/null +++ b/drivers/net/i40e/base/virtchnl.h @@ -0,0 +1,772 @@ +/******************************************************************************* + +Copyright (c) 2013 - 2015, Intel Corporation +All rights reserved. + +Redistribution and use in source and binary forms, with or without +modification, are permitted provided that the following conditions are met: + + 1. Redistributions of source code must retain the above copyright notice, + this list of conditions and the following disclaimer. + + 2. Redistributions in binary form must reproduce the above copyright + notice, this list of conditions and the following disclaimer in the + documentation and/or other materials provided with the distribution. + + 3. Neither the name of the Intel Corporation nor the names of its + contributors may be used to endorse or promote products derived from + this software without specific prior written permission. + +THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS" +AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE +IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE +ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE +LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR +CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF +SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS +INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN +CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) +ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE +POSSIBILITY OF SUCH DAMAGE. + +***************************************************************************/ + +#ifndef _VIRTCHNL_H_ +#define _VIRTCHNL_H_ + +/* Description: + * This header file describes the VF-PF communication protocol used + * by the drivers for all devices starting from our 40G product line + * + * Admin queue buffer usage: + * desc->opcode is always aqc_opc_send_msg_to_pf + * flags, retval, datalen, and data addr are all used normally. + * The Firmware copies the cookie fields when sending messages between the + * PF and VF, but uses all other fields internally. Due to this limitation, + * we must send all messages as "indirect", i.e. using an external buffer. + * + * All the VSI indexes are relative to the VF. Each VF can have maximum of + * three VSIs. All the queue indexes are relative to the VSI. Each VF can + * have a maximum of sixteen queues for all of its VSIs. + * + * The PF is required to return a status code in v_retval for all messages + * except RESET_VF, which does not require any response. The return value + * is of status_code type, defined in the shared type.h. + * + * In general, VF driver initialization should roughly follow the order of + * these opcodes. The VF driver must first validate the API version of the + * PF driver, then request a reset, then get resources, then configure + * queues and interrupts. After these operations are complete, the VF + * driver may start its queues, optionally add MAC and VLAN filters, and + * process traffic. + */ + +/* START GENERIC DEFINES + * Need to ensure the following enums and defines hold the same meaning and + * value in current and future projects + */ + +/* Error Codes */ +enum virtchnl_status_code { + VIRTCHNL_STATUS_SUCCESS = 0, + VIRTCHNL_ERR_PARAM = -5, + VIRTCHNL_STATUS_ERR_OPCODE_MISMATCH = -38, + VIRTCHNL_STATUS_ERR_CQP_COMPL_ERROR = -39, + VIRTCHNL_STATUS_ERR_INVALID_VF_ID = -40, + VIRTCHNL_STATUS_NOT_SUPPORTED = -64, +}; + +#define VIRTCHNL_LINK_SPEED_100MB_SHIFT 0x1 +#define VIRTCHNL_LINK_SPEED_1000MB_SHIFT 0x2 +#define VIRTCHNL_LINK_SPEED_10GB_SHIFT 0x3 +#define VIRTCHNL_LINK_SPEED_40GB_SHIFT 0x4 +#define VIRTCHNL_LINK_SPEED_20GB_SHIFT 0x5 +#define VIRTCHNL_LINK_SPEED_25GB_SHIFT 0x6 + +enum virtchnl_link_speed { + VIRTCHNL_LINK_SPEED_UNKNOWN = 0, + VIRTCHNL_LINK_SPEED_100MB = BIT(VIRTCHNL_LINK_SPEED_100MB_SHIFT), + VIRTCHNL_LINK_SPEED_1GB = BIT(VIRTCHNL_LINK_SPEED_1000MB_SHIFT), + VIRTCHNL_LINK_SPEED_10GB = BIT(VIRTCHNL_LINK_SPEED_10GB_SHIFT), + VIRTCHNL_LINK_SPEED_40GB = BIT(VIRTCHNL_LINK_SPEED_40GB_SHIFT), + VIRTCHNL_LINK_SPEED_20GB = BIT(VIRTCHNL_LINK_SPEED_20GB_SHIFT), + VIRTCHNL_LINK_SPEED_25GB = BIT(VIRTCHNL_LINK_SPEED_25GB_SHIFT), +}; + +/* for hsplit_0 field of Rx HMC context */ +/* deprecated with AVF 1.0 */ +enum virtchnl_rx_hsplit { + VIRTCHNL_RX_HSPLIT_NO_SPLIT = 0, + VIRTCHNL_RX_HSPLIT_SPLIT_L2 = 1, + VIRTCHNL_RX_HSPLIT_SPLIT_IP = 2, + VIRTCHNL_RX_HSPLIT_SPLIT_TCP_UDP = 4, + VIRTCHNL_RX_HSPLIT_SPLIT_SCTP = 8, +}; + +#define VIRTCHNL_ETH_LENGTH_OF_ADDRESS 6 +/* END GENERIC DEFINES */ + +/* Opcodes for VF-PF communication. These are placed in the v_opcode field + * of the virtchnl_msg structure. + */ +enum virtchnl_ops { +/* The PF sends status change events to VFs using + * the VIRTCHNL_OP_EVENT opcode. + * VFs send requests to the PF using the other ops. + * Use of "advanced opcode" features must be negotiated as part of capabilities + * exchange and are not considered part of base mode feature set. + */ + VIRTCHNL_OP_UNKNOWN = 0, + VIRTCHNL_OP_VERSION = 1, /* must ALWAYS be 1 */ + VIRTCHNL_OP_RESET_VF = 2, + VIRTCHNL_OP_GET_VF_RESOURCES = 3, + VIRTCHNL_OP_CONFIG_TX_QUEUE = 4, + VIRTCHNL_OP_CONFIG_RX_QUEUE = 5, + VIRTCHNL_OP_CONFIG_VSI_QUEUES = 6, + VIRTCHNL_OP_CONFIG_IRQ_MAP = 7, + VIRTCHNL_OP_ENABLE_QUEUES = 8, + VIRTCHNL_OP_DISABLE_QUEUES = 9, + VIRTCHNL_OP_ADD_ETH_ADDR = 10, + VIRTCHNL_OP_DEL_ETH_ADDR = 11, + VIRTCHNL_OP_ADD_VLAN = 12, + VIRTCHNL_OP_DEL_VLAN = 13, + VIRTCHNL_OP_CONFIG_PROMISCUOUS_MODE = 14, + VIRTCHNL_OP_GET_STATS = 15, + VIRTCHNL_OP_RSVD = 16, + VIRTCHNL_OP_EVENT = 17, /* must ALWAYS be 17 */ +#ifdef VIRTCHNL_SOL_VF_SUPPORT + VIRTCHNL_OP_GET_ADDNL_SOL_CONFIG = 19, +#endif +#ifdef VIRTCHNL_IWARP + VIRTCHNL_OP_IWARP = 20, /* advanced opcode */ + VIRTCHNL_OP_CONFIG_IWARP_IRQ_MAP = 21, /* advanced opcode */ + VIRTCHNL_OP_RELEASE_IWARP_IRQ_MAP = 22, /* advanced opcode */ +#endif + VIRTCHNL_OP_CONFIG_RSS_KEY = 23, + VIRTCHNL_OP_CONFIG_RSS_LUT = 24, + VIRTCHNL_OP_GET_RSS_HENA_CAPS = 25, + VIRTCHNL_OP_SET_RSS_HENA = 26, + VIRTCHNL_OP_ENABLE_VLAN_STRIPPING = 27, + VIRTCHNL_OP_DISABLE_VLAN_STRIPPING = 28, + VIRTCHNL_OP_REQUEST_QUEUES = 29, + +}; + +/* This macro is used to generate a compilation error if a structure + * is not exactly the correct length. It gives a divide by zero error if the + * structure is not of the correct size, otherwise it creates an enum that is + * never used. + */ +#define VIRTCHNL_CHECK_STRUCT_LEN(n, X) enum virtchnl_static_assert_enum_##X \ + {virtchnl_static_assert_##X = (n) / ((sizeof(struct X) == (n)) ? 1 : 0)} + +/* Virtual channel message descriptor. This overlays the admin queue + * descriptor. All other data is passed in external buffers. + */ + +struct virtchnl_msg { + u8 pad[8]; /* AQ flags/opcode/len/retval fields */ + enum virtchnl_ops v_opcode; /* avoid confusion with desc->opcode */ + enum virtchnl_status_code v_retval; /* ditto for desc->retval */ + u32 vfid; /* used by PF when sending to VF */ +}; + +VIRTCHNL_CHECK_STRUCT_LEN(20, virtchnl_msg); + +/* Message descriptions and data structures.*/ + +/* VIRTCHNL_OP_VERSION + * VF posts its version number to the PF. PF responds with its version number + * in the same format, along with a return code. + * Reply from PF has its major/minor versions also in param0 and param1. + * If there is a major version mismatch, then the VF cannot operate. + * If there is a minor version mismatch, then the VF can operate but should + * add a warning to the system log. + * + * This enum element MUST always be specified as == 1, regardless of other + * changes in the API. The PF must always respond to this message without + * error regardless of version mismatch. + */ +#define VIRTCHNL_VERSION_MAJOR 1 +#define VIRTCHNL_VERSION_MINOR 1 +#define VIRTCHNL_VERSION_MINOR_NO_VF_CAPS 0 + +struct virtchnl_version_info { + u32 major; + u32 minor; +}; + +VIRTCHNL_CHECK_STRUCT_LEN(8, virtchnl_version_info); + +#define VF_IS_V10(_v) (((_v)->major == 1) && ((_v)->minor == 0)) +#define VF_IS_V11(_ver) (((_ver)->major == 1) && ((_ver)->minor == 1)) + +/* VIRTCHNL_OP_RESET_VF + * VF sends this request to PF with no parameters + * PF does NOT respond! VF driver must delay then poll VFGEN_RSTAT register + * until reset completion is indicated. The admin queue must be reinitialized + * after this operation. + * + * When reset is complete, PF must ensure that all queues in all VSIs associated + * with the VF are stopped, all queue configurations in the HMC are set to 0, + * and all MAC and VLAN filters (except the default MAC address) on all VSIs + * are cleared. + */ + +/* VSI types that use VIRTCHNL interface for VF-PF communication. VSI_SRIOV + * vsi_type should always be 6 for backward compatibility. Add other fields + * as needed. + */ +enum virtchnl_vsi_type { + VIRTCHNL_VSI_TYPE_INVALID = 0, + VIRTCHNL_VSI_SRIOV = 6, +}; + +/* VIRTCHNL_OP_GET_VF_RESOURCES + * Version 1.0 VF sends this request to PF with no parameters + * Version 1.1 VF sends this request to PF with u32 bitmap of its capabilities + * PF responds with an indirect message containing + * virtchnl_vf_resource and one or more + * virtchnl_vsi_resource structures. + */ + +struct virtchnl_vsi_resource { + u16 vsi_id; + u16 num_queue_pairs; + enum virtchnl_vsi_type vsi_type; + u16 qset_handle; + u8 default_mac_addr[VIRTCHNL_ETH_LENGTH_OF_ADDRESS]; +}; + +VIRTCHNL_CHECK_STRUCT_LEN(16, virtchnl_vsi_resource); + +/* VF offload flags + * VIRTCHNL_VF_OFFLOAD_L2 flag is inclusive of base mode L2 offloads including + * TX/RX Checksum offloading and TSO for non-tunnelled packets. + */ +#define VIRTCHNL_VF_OFFLOAD_L2 0x00000001 +#define VIRTCHNL_VF_OFFLOAD_IWARP 0x00000002 +#define VIRTCHNL_VF_OFFLOAD_RSVD 0x00000004 +#define VIRTCHNL_VF_OFFLOAD_RSS_AQ 0x00000008 +#define VIRTCHNL_VF_OFFLOAD_RSS_REG 0x00000010 +#define VIRTCHNL_VF_OFFLOAD_WB_ON_ITR 0x00000020 +#define VIRTCHNL_VF_OFFLOAD_REQ_QUEUES 0x00000040 +#define VIRTCHNL_VF_OFFLOAD_VLAN 0x00010000 +#define VIRTCHNL_VF_OFFLOAD_RX_POLLING 0x00020000 +#define VIRTCHNL_VF_OFFLOAD_RSS_PCTYPE_V2 0x00040000 +#define VIRTCHNL_VF_OFFLOAD_RSS_PF 0X00080000 +#define VIRTCHNL_VF_OFFLOAD_ENCAP 0X00100000 +#define VIRTCHNL_VF_OFFLOAD_ENCAP_CSUM 0X00200000 +#define VIRTCHNL_VF_OFFLOAD_RX_ENCAP_CSUM 0X00400000 + +#define VF_BASE_MODE_OFFLOADS (VIRTCHNL_VF_OFFLOAD_L2 | \ + VIRTCHNL_VF_OFFLOAD_VLAN | \ + VIRTCHNL_VF_OFFLOAD_RSS_PF) + +struct virtchnl_vf_resource { + u16 num_vsis; + u16 num_queue_pairs; + u16 max_vectors; + u16 max_mtu; + + u32 vf_offload_flags; + u32 rss_key_size; + u32 rss_lut_size; + + struct virtchnl_vsi_resource vsi_res[1]; +}; + +VIRTCHNL_CHECK_STRUCT_LEN(36, virtchnl_vf_resource); + +/* VIRTCHNL_OP_CONFIG_TX_QUEUE + * VF sends this message to set up parameters for one TX queue. + * External data buffer contains one instance of virtchnl_txq_info. + * PF configures requested queue and returns a status code. + */ + +/* Tx queue config info */ +struct virtchnl_txq_info { + u16 vsi_id; + u16 queue_id; + u16 ring_len; /* number of descriptors, multiple of 8 */ + u16 headwb_enabled; /* deprecated with AVF 1.0 */ + u64 dma_ring_addr; + u64 dma_headwb_addr; /* deprecated with AVF 1.0 */ +}; + +VIRTCHNL_CHECK_STRUCT_LEN(24, virtchnl_txq_info); + +/* VIRTCHNL_OP_CONFIG_RX_QUEUE + * VF sends this message to set up parameters for one RX queue. + * External data buffer contains one instance of virtchnl_rxq_info. + * PF configures requested queue and returns a status code. + */ + +/* Rx queue config info */ +struct virtchnl_rxq_info { + u16 vsi_id; + u16 queue_id; + u32 ring_len; /* number of descriptors, multiple of 32 */ + u16 hdr_size; + u16 splithdr_enabled; /* deprecated with AVF 1.0 */ + u32 databuffer_size; + u32 max_pkt_size; + u32 pad1; + u64 dma_ring_addr; + enum virtchnl_rx_hsplit rx_split_pos; /* deprecated with AVF 1.0 */ + u32 pad2; +}; + +VIRTCHNL_CHECK_STRUCT_LEN(40, virtchnl_rxq_info); + +/* VIRTCHNL_OP_CONFIG_VSI_QUEUES + * VF sends this message to set parameters for all active TX and RX queues + * associated with the specified VSI. + * PF configures queues and returns status. + * If the number of queues specified is greater than the number of queues + * associated with the VSI, an error is returned and no queues are configured. + */ +struct virtchnl_queue_pair_info { + /* NOTE: vsi_id and queue_id should be identical for both queues. */ + struct virtchnl_txq_info txq; + struct virtchnl_rxq_info rxq; +}; + +VIRTCHNL_CHECK_STRUCT_LEN(64, virtchnl_queue_pair_info); + +struct virtchnl_vsi_queue_config_info { + u16 vsi_id; + u16 num_queue_pairs; + u32 pad; + struct virtchnl_queue_pair_info qpair[1]; +}; + +VIRTCHNL_CHECK_STRUCT_LEN(72, virtchnl_vsi_queue_config_info); + +/* VIRTCHNL_OP_REQUEST_QUEUES + * VF sends this message to request the PF to allocate additional queues to + * this VF. Each VF gets a guaranteed number of queues on init but asking for + * additional queues must be negotiated. This is a best effort request as it + * is possible the PF does not have enough queues left to support the request. + * If the PF cannot support the number requested it will respond with the + * maximum number it is able to support; otherwise it will respond with the + * number requested. + */ + +/* VF resource request */ +struct virtchnl_vf_res_request { + u16 num_queue_pairs; +}; + +/* VIRTCHNL_OP_CONFIG_IRQ_MAP + * VF uses this message to map vectors to queues. + * The rxq_map and txq_map fields are bitmaps used to indicate which queues + * are to be associated with the specified vector. + * The "other" causes are always mapped to vector 0. + * PF configures interrupt mapping and returns status. + */ +struct virtchnl_vector_map { + u16 vsi_id; + u16 vector_id; + u16 rxq_map; + u16 txq_map; + u16 rxitr_idx; + u16 txitr_idx; +}; + +VIRTCHNL_CHECK_STRUCT_LEN(12, virtchnl_vector_map); + +struct virtchnl_irq_map_info { + u16 num_vectors; + struct virtchnl_vector_map vecmap[1]; +}; + +VIRTCHNL_CHECK_STRUCT_LEN(14, virtchnl_irq_map_info); + +/* VIRTCHNL_OP_ENABLE_QUEUES + * VIRTCHNL_OP_DISABLE_QUEUES + * VF sends these message to enable or disable TX/RX queue pairs. + * The queues fields are bitmaps indicating which queues to act upon. + * (Currently, we only support 16 queues per VF, but we make the field + * u32 to allow for expansion.) + * PF performs requested action and returns status. + */ +struct virtchnl_queue_select { + u16 vsi_id; + u16 pad; + u32 rx_queues; + u32 tx_queues; +}; + +VIRTCHNL_CHECK_STRUCT_LEN(12, virtchnl_queue_select); + +/* VIRTCHNL_OP_ADD_ETH_ADDR + * VF sends this message in order to add one or more unicast or multicast + * address filters for the specified VSI. + * PF adds the filters and returns status. + */ + +/* VIRTCHNL_OP_DEL_ETH_ADDR + * VF sends this message in order to remove one or more unicast or multicast + * filters for the specified VSI. + * PF removes the filters and returns status. + */ + +struct virtchnl_ether_addr { + u8 addr[VIRTCHNL_ETH_LENGTH_OF_ADDRESS]; + u8 pad[2]; +}; + +VIRTCHNL_CHECK_STRUCT_LEN(8, virtchnl_ether_addr); + +struct virtchnl_ether_addr_list { + u16 vsi_id; + u16 num_elements; + struct virtchnl_ether_addr list[1]; +}; + +VIRTCHNL_CHECK_STRUCT_LEN(12, virtchnl_ether_addr_list); + +#ifdef VIRTCHNL_SOL_VF_SUPPORT +/* VIRTCHNL_OP_GET_ADDNL_SOL_CONFIG + * VF sends this message to get the default MTU and list of additional ethernet + * addresses it is allowed to use. + * PF responds with an indirect message containing + * virtchnl_addnl_solaris_config with zero or more + * virtchnl_ether_addr structures. + * + * It is expected that this operation will only ever be needed for Solaris VFs + * running under a Solaris PF. + */ +struct virtchnl_addnl_solaris_config { + u16 default_mtu; + struct virtchnl_ether_addr_list al; +}; + +#endif +/* VIRTCHNL_OP_ADD_VLAN + * VF sends this message to add one or more VLAN tag filters for receives. + * PF adds the filters and returns status. + * If a port VLAN is configured by the PF, this operation will return an + * error to the VF. + */ + +/* VIRTCHNL_OP_DEL_VLAN + * VF sends this message to remove one or more VLAN tag filters for receives. + * PF removes the filters and returns status. + * If a port VLAN is configured by the PF, this operation will return an + * error to the VF. + */ + +struct virtchnl_vlan_filter_list { + u16 vsi_id; + u16 num_elements; + u16 vlan_id[1]; +}; + +VIRTCHNL_CHECK_STRUCT_LEN(6, virtchnl_vlan_filter_list); + +/* VIRTCHNL_OP_CONFIG_PROMISCUOUS_MODE + * VF sends VSI id and flags. + * PF returns status code in retval. + * Note: we assume that broadcast accept mode is always enabled. + */ +struct virtchnl_promisc_info { + u16 vsi_id; + u16 flags; +}; + +VIRTCHNL_CHECK_STRUCT_LEN(4, virtchnl_promisc_info); + +#define FLAG_VF_UNICAST_PROMISC 0x00000001 +#define FLAG_VF_MULTICAST_PROMISC 0x00000002 + +/* VIRTCHNL_OP_GET_STATS + * VF sends this message to request stats for the selected VSI. VF uses + * the virtchnl_queue_select struct to specify the VSI. The queue_id + * field is ignored by the PF. + * + * PF replies with struct eth_stats in an external buffer. + */ + +/* VIRTCHNL_OP_CONFIG_RSS_KEY + * VIRTCHNL_OP_CONFIG_RSS_LUT + * VF sends these messages to configure RSS. Only supported if both PF + * and VF drivers set the VIRTCHNL_VF_OFFLOAD_RSS_PF bit during + * configuration negotiation. If this is the case, then the RSS fields in + * the VF resource struct are valid. + * Both the key and LUT are initialized to 0 by the PF, meaning that + * RSS is effectively disabled until set up by the VF. + */ +struct virtchnl_rss_key { + u16 vsi_id; + u16 key_len; + u8 key[1]; /* RSS hash key, packed bytes */ +}; + +VIRTCHNL_CHECK_STRUCT_LEN(6, virtchnl_rss_key); + +struct virtchnl_rss_lut { + u16 vsi_id; + u16 lut_entries; + u8 lut[1]; /* RSS lookup table*/ +}; + +VIRTCHNL_CHECK_STRUCT_LEN(6, virtchnl_rss_lut); + +/* VIRTCHNL_OP_GET_RSS_HENA_CAPS + * VIRTCHNL_OP_SET_RSS_HENA + * VF sends these messages to get and set the hash filter enable bits for RSS. + * By default, the PF sets these to all possible traffic types that the + * hardware supports. The VF can query this value if it wants to change the + * traffic types that are hashed by the hardware. + */ +struct virtchnl_rss_hena { + u64 hena; +}; + +VIRTCHNL_CHECK_STRUCT_LEN(8, virtchnl_rss_hena); + +/* VIRTCHNL_OP_EVENT + * PF sends this message to inform the VF driver of events that may affect it. + * No direct response is expected from the VF, though it may generate other + * messages in response to this one. + */ +enum virtchnl_event_codes { + VIRTCHNL_EVENT_UNKNOWN = 0, + VIRTCHNL_EVENT_LINK_CHANGE, + VIRTCHNL_EVENT_RESET_IMPENDING, + VIRTCHNL_EVENT_PF_DRIVER_CLOSE, +}; + +#define PF_EVENT_SEVERITY_INFO 0 +#define PF_EVENT_SEVERITY_ATTENTION 1 +#define PF_EVENT_SEVERITY_ACTION_REQUIRED 2 +#define PF_EVENT_SEVERITY_CERTAIN_DOOM 255 + +struct virtchnl_pf_event { + enum virtchnl_event_codes event; + union { + struct { + enum virtchnl_link_speed link_speed; + bool link_status; + } link_event; + } event_data; + + int severity; +}; + +VIRTCHNL_CHECK_STRUCT_LEN(16, virtchnl_pf_event); + +#ifdef VIRTCHNL_IWARP + +/* VIRTCHNL_OP_CONFIG_IWARP_IRQ_MAP + * VF uses this message to request PF to map IWARP vectors to IWARP queues. + * The request for this originates from the VF IWARP driver through + * a client interface between VF LAN and VF IWARP driver. + * A vector could have an AEQ and CEQ attached to it although + * there is a single AEQ per VF IWARP instance in which case + * most vectors will have an INVALID_IDX for aeq and valid idx for ceq. + * There will never be a case where there will be multiple CEQs attached + * to a single vector. + * PF configures interrupt mapping and returns status. + */ + +/* HW does not define a type value for AEQ; only for RX/TX and CEQ. + * In order for us to keep the interface simple, SW will define a + * unique type value for AEQ. + */ +#define QUEUE_TYPE_PE_AEQ 0x80 +#define QUEUE_INVALID_IDX 0xFFFF + +struct virtchnl_iwarp_qv_info { + u32 v_idx; /* msix_vector */ + u16 ceq_idx; + u16 aeq_idx; + u8 itr_idx; +}; + +VIRTCHNL_CHECK_STRUCT_LEN(12, virtchnl_iwarp_qv_info); + +struct virtchnl_iwarp_qvlist_info { + u32 num_vectors; + struct virtchnl_iwarp_qv_info qv_info[1]; +}; + +VIRTCHNL_CHECK_STRUCT_LEN(16, virtchnl_iwarp_qvlist_info); + +#endif + +/* VF reset states - these are written into the RSTAT register: + * VFGEN_RSTAT on the VF + * When the PF initiates a reset, it writes 0 + * When the reset is complete, it writes 1 + * When the PF detects that the VF has recovered, it writes 2 + * VF checks this register periodically to determine if a reset has occurred, + * then polls it to know when the reset is complete. + * If either the PF or VF reads the register while the hardware + * is in a reset state, it will return DEADBEEF, which, when masked + * will result in 3. + */ +enum virtchnl_vfr_states { + VIRTCHNL_VFR_INPROGRESS = 0, + VIRTCHNL_VFR_COMPLETED, + VIRTCHNL_VFR_VFACTIVE, +}; + +/** + * virtchnl_vc_validate_vf_msg + * @ver: Virtchnl version info + * @v_opcode: Opcode for the message + * @msg: pointer to the msg buffer + * @msglen: msg length + * + * validate msg format against struct for each opcode + */ +static inline int +virtchnl_vc_validate_vf_msg(struct virtchnl_version_info *ver, u32 v_opcode, + u8 *msg, u16 msglen) +{ + bool err_msg_format = false; + int valid_len = 0; + + /* Validate message length. */ + switch (v_opcode) { + case VIRTCHNL_OP_VERSION: + valid_len = sizeof(struct virtchnl_version_info); + break; + case VIRTCHNL_OP_RESET_VF: + break; + case VIRTCHNL_OP_GET_VF_RESOURCES: + if (VF_IS_V11(ver)) + valid_len = sizeof(u32); + break; + case VIRTCHNL_OP_CONFIG_TX_QUEUE: + valid_len = sizeof(struct virtchnl_txq_info); + break; + case VIRTCHNL_OP_CONFIG_RX_QUEUE: + valid_len = sizeof(struct virtchnl_rxq_info); + break; + case VIRTCHNL_OP_CONFIG_VSI_QUEUES: + valid_len = sizeof(struct virtchnl_vsi_queue_config_info); + if (msglen >= valid_len) { + struct virtchnl_vsi_queue_config_info *vqc = + (struct virtchnl_vsi_queue_config_info *)msg; + valid_len += (vqc->num_queue_pairs * + sizeof(struct + virtchnl_queue_pair_info)); + if (vqc->num_queue_pairs == 0) + err_msg_format = true; + } + break; + case VIRTCHNL_OP_CONFIG_IRQ_MAP: + valid_len = sizeof(struct virtchnl_irq_map_info); + if (msglen >= valid_len) { + struct virtchnl_irq_map_info *vimi = + (struct virtchnl_irq_map_info *)msg; + valid_len += (vimi->num_vectors * + sizeof(struct virtchnl_vector_map)); + if (vimi->num_vectors == 0) + err_msg_format = true; + } + break; + case VIRTCHNL_OP_ENABLE_QUEUES: + case VIRTCHNL_OP_DISABLE_QUEUES: + valid_len = sizeof(struct virtchnl_queue_select); + break; + case VIRTCHNL_OP_ADD_ETH_ADDR: + case VIRTCHNL_OP_DEL_ETH_ADDR: + valid_len = sizeof(struct virtchnl_ether_addr_list); + if (msglen >= valid_len) { + struct virtchnl_ether_addr_list *veal = + (struct virtchnl_ether_addr_list *)msg; + valid_len += veal->num_elements * + sizeof(struct virtchnl_ether_addr); + if (veal->num_elements == 0) + err_msg_format = true; + } + break; + case VIRTCHNL_OP_ADD_VLAN: + case VIRTCHNL_OP_DEL_VLAN: + valid_len = sizeof(struct virtchnl_vlan_filter_list); + if (msglen >= valid_len) { + struct virtchnl_vlan_filter_list *vfl = + (struct virtchnl_vlan_filter_list *)msg; + valid_len += vfl->num_elements * sizeof(u16); + if (vfl->num_elements == 0) + err_msg_format = true; + } + break; + case VIRTCHNL_OP_CONFIG_PROMISCUOUS_MODE: + valid_len = sizeof(struct virtchnl_promisc_info); + break; + case VIRTCHNL_OP_GET_STATS: + valid_len = sizeof(struct virtchnl_queue_select); + break; +#ifdef VIRTCHNL_IWARP + case VIRTCHNL_OP_IWARP: + /* These messages are opaque to us and will be validated in + * the RDMA client code. We just need to check for nonzero + * length. The firmware will enforce max length restrictions. + */ + if (msglen) + valid_len = msglen; + else + err_msg_format = true; + break; + case VIRTCHNL_OP_RELEASE_IWARP_IRQ_MAP: + break; + case VIRTCHNL_OP_CONFIG_IWARP_IRQ_MAP: + valid_len = sizeof(struct virtchnl_iwarp_qvlist_info); + if (msglen >= valid_len) { + struct virtchnl_iwarp_qvlist_info *qv = + (struct virtchnl_iwarp_qvlist_info *)msg; + if (qv->num_vectors == 0) { + err_msg_format = true; + break; + } + valid_len += ((qv->num_vectors - 1) * + sizeof(struct virtchnl_iwarp_qv_info)); + } + break; +#endif + case VIRTCHNL_OP_CONFIG_RSS_KEY: + valid_len = sizeof(struct virtchnl_rss_key); + if (msglen >= valid_len) { + struct virtchnl_rss_key *vrk = + (struct virtchnl_rss_key *)msg; + valid_len += vrk->key_len - 1; + } + break; + case VIRTCHNL_OP_CONFIG_RSS_LUT: + valid_len = sizeof(struct virtchnl_rss_lut); + if (msglen >= valid_len) { + struct virtchnl_rss_lut *vrl = + (struct virtchnl_rss_lut *)msg; + valid_len += vrl->lut_entries - 1; + } + break; + case VIRTCHNL_OP_GET_RSS_HENA_CAPS: + break; + case VIRTCHNL_OP_SET_RSS_HENA: + valid_len = sizeof(struct virtchnl_rss_hena); + break; + case VIRTCHNL_OP_ENABLE_VLAN_STRIPPING: + case VIRTCHNL_OP_DISABLE_VLAN_STRIPPING: + break; + case VIRTCHNL_OP_REQUEST_QUEUES: + valid_len = sizeof(struct virtchnl_vf_res_request); + break; + /* These are always errors coming from the VF. */ + case VIRTCHNL_OP_EVENT: + case VIRTCHNL_OP_UNKNOWN: + default: + return VIRTCHNL_ERR_PARAM; + } + /* few more checks */ + if ((valid_len != msglen) || (err_msg_format)) + return VIRTCHNL_STATUS_ERR_OPCODE_MISMATCH; + + return 0; +} +#endif /* _VIRTCHNL_H_ */ diff --git a/drivers/net/i40e/i40e_ethdev.h b/drivers/net/i40e/i40e_ethdev.h index 183dc17..bda7450 100644 --- a/drivers/net/i40e/i40e_ethdev.h +++ b/drivers/net/i40e/i40e_ethdev.h @@ -102,7 +102,7 @@ /* Linux PF host with virtchnl version 1.1 */ #define PF_IS_V11(vf) \ - (((vf)->version_major == I40E_VIRTCHNL_VERSION_MAJOR) && \ + (((vf)->version_major == VIRTCHNL_VERSION_MAJOR) && \ ((vf)->version_minor == 1)) /* index flex payload per layer */ @@ -767,7 +767,7 @@ struct i40e_vf { /* Event from pf */ bool dev_closed; bool link_up; - enum i40e_aq_link_speed link_speed; + enum virtchnl_link_speed link_speed; bool vf_reset; volatile uint32_t pend_cmd; /* pending command not finished yet */ int32_t cmd_retval; /* return value of the cmd response from PF */ @@ -775,8 +775,8 @@ struct i40e_vf { uint8_t *aq_resp; /* buffer to store the adminq response from PF */ /* VSI info */ - struct i40e_virtchnl_vf_resource *vf_res; /* All VSIs */ - struct i40e_virtchnl_vsi_resource *vsi_res; /* LAN VSI */ + struct virtchnl_vf_resource *vf_res; /* All VSIs */ + struct virtchnl_vsi_resource *vsi_res; /* LAN VSI */ struct i40e_vsi vsi; uint64_t flags; }; diff --git a/drivers/net/i40e/i40e_ethdev_vf.c b/drivers/net/i40e/i40e_ethdev_vf.c index 2d5a9b5..456dd11 100644 --- a/drivers/net/i40e/i40e_ethdev_vf.c +++ b/drivers/net/i40e/i40e_ethdev_vf.c @@ -77,7 +77,7 @@ #define MAX_RESET_WAIT_CNT 20 struct i40evf_arq_msg_info { - enum i40e_virtchnl_ops ops; + enum virtchnl_ops ops; enum i40e_status_code result; uint16_t buf_len; uint16_t msg_len; @@ -85,7 +85,7 @@ struct i40evf_arq_msg_info { }; struct vf_cmd_info { - enum i40e_virtchnl_ops ops; + enum virtchnl_ops ops; uint8_t *in_args; uint32_t in_args_size; uint8_t *out_buffer; @@ -244,7 +244,7 @@ i40evf_read_pfmsg(struct rte_eth_dev *dev, struct i40evf_arq_msg_info *data) struct i40e_hw *hw = I40E_DEV_PRIVATE_TO_HW(dev->data->dev_private); struct i40e_vf *vf = I40EVF_DEV_PRIVATE_TO_VF(dev->data->dev_private); struct i40e_arq_event_info event; - enum i40e_virtchnl_ops opcode; + enum virtchnl_ops opcode; enum i40e_status_code retval; int ret; enum i40evf_aq_result result = I40EVF_MSG_NON; @@ -259,16 +259,16 @@ i40evf_read_pfmsg(struct rte_eth_dev *dev, struct i40evf_arq_msg_info *data) return result; } - opcode = (enum i40e_virtchnl_ops)rte_le_to_cpu_32(event.desc.cookie_high); + opcode = (enum virtchnl_ops)rte_le_to_cpu_32(event.desc.cookie_high); retval = (enum i40e_status_code)rte_le_to_cpu_32(event.desc.cookie_low); /* pf sys event */ - if (opcode == I40E_VIRTCHNL_OP_EVENT) { - struct i40e_virtchnl_pf_event *vpe = - (struct i40e_virtchnl_pf_event *)event.msg_buf; + if (opcode == VIRTCHNL_OP_EVENT) { + struct virtchnl_pf_event *vpe = + (struct virtchnl_pf_event *)event.msg_buf; result = I40EVF_MSG_SYS; switch (vpe->event) { - case I40E_VIRTCHNL_EVENT_LINK_CHANGE: + case VIRTCHNL_EVENT_LINK_CHANGE: vf->link_up = vpe->event_data.link_event.link_status; vf->link_speed = @@ -277,12 +277,12 @@ i40evf_read_pfmsg(struct rte_eth_dev *dev, struct i40evf_arq_msg_info *data) PMD_DRV_LOG(INFO, "Link status update:%s", vf->link_up ? "up" : "down"); break; - case I40E_VIRTCHNL_EVENT_RESET_IMPENDING: + case VIRTCHNL_EVENT_RESET_IMPENDING: vf->vf_reset = true; vf->pend_msg |= PFMSG_RESET_IMPENDING; PMD_DRV_LOG(INFO, "vf is reseting"); break; - case I40E_VIRTCHNL_EVENT_PF_DRIVER_CLOSE: + case VIRTCHNL_EVENT_PF_DRIVER_CLOSE: vf->dev_closed = true; vf->pend_msg |= PFMSG_DRIVER_CLOSE; PMD_DRV_LOG(INFO, "PF driver closed"); @@ -312,17 +312,17 @@ static inline void _clear_cmd(struct i40e_vf *vf) { rte_wmb(); - vf->pend_cmd = I40E_VIRTCHNL_OP_UNKNOWN; + vf->pend_cmd = VIRTCHNL_OP_UNKNOWN; } /* * Check there is pending cmd in execution. If none, set new command. */ static inline int -_atomic_set_cmd(struct i40e_vf *vf, enum i40e_virtchnl_ops ops) +_atomic_set_cmd(struct i40e_vf *vf, enum virtchnl_ops ops) { int ret = rte_atomic32_cmpset(&vf->pend_cmd, - I40E_VIRTCHNL_OP_UNKNOWN, ops); + VIRTCHNL_OP_UNKNOWN, ops); if (!ret) PMD_DRV_LOG(ERR, "There is incomplete cmd %d", vf->pend_cmd); @@ -347,7 +347,7 @@ i40evf_execute_vf_cmd(struct rte_eth_dev *dev, struct vf_cmd_info *args) info.msg = args->out_buffer; info.buf_len = args->out_size; - info.ops = I40E_VIRTCHNL_OP_UNKNOWN; + info.ops = VIRTCHNL_OP_UNKNOWN; info.result = I40E_SUCCESS; err = i40e_aq_send_msg_to_pf(hw, args->ops, I40E_SUCCESS, @@ -359,12 +359,12 @@ i40evf_execute_vf_cmd(struct rte_eth_dev *dev, struct vf_cmd_info *args) } switch (args->ops) { - case I40E_VIRTCHNL_OP_RESET_VF: + case VIRTCHNL_OP_RESET_VF: /*no need to process in this function */ err = 0; break; - case I40E_VIRTCHNL_OP_VERSION: - case I40E_VIRTCHNL_OP_GET_VF_RESOURCES: + case VIRTCHNL_OP_VERSION: + case VIRTCHNL_OP_GET_VF_RESOURCES: /* for init adminq commands, need to poll the response */ err = -1; do { @@ -385,7 +385,7 @@ i40evf_execute_vf_cmd(struct rte_eth_dev *dev, struct vf_cmd_info *args) /* for other adminq in running time, waiting the cmd done flag */ err = -1; do { - if (vf->pend_cmd == I40E_VIRTCHNL_OP_UNKNOWN) { + if (vf->pend_cmd == VIRTCHNL_OP_UNKNOWN) { err = 0; break; } @@ -404,15 +404,15 @@ i40evf_execute_vf_cmd(struct rte_eth_dev *dev, struct vf_cmd_info *args) static int i40evf_check_api_version(struct rte_eth_dev *dev) { - struct i40e_virtchnl_version_info version, *pver; + struct virtchnl_version_info version, *pver; int err; struct vf_cmd_info args; struct i40e_vf *vf = I40EVF_DEV_PRIVATE_TO_VF(dev->data->dev_private); - version.major = I40E_VIRTCHNL_VERSION_MAJOR; - version.minor = I40E_VIRTCHNL_VERSION_MINOR; + version.major = VIRTCHNL_VERSION_MAJOR; + version.minor = VIRTCHNL_VERSION_MINOR; - args.ops = I40E_VIRTCHNL_OP_VERSION; + args.ops = VIRTCHNL_OP_VERSION; args.in_args = (uint8_t *)&version; args.in_args_size = sizeof(version); args.out_buffer = vf->aq_resp; @@ -424,19 +424,19 @@ i40evf_check_api_version(struct rte_eth_dev *dev) return err; } - pver = (struct i40e_virtchnl_version_info *)args.out_buffer; + pver = (struct virtchnl_version_info *)args.out_buffer; vf->version_major = pver->major; vf->version_minor = pver->minor; if (vf->version_major == I40E_DPDK_VERSION_MAJOR) PMD_DRV_LOG(INFO, "Peer is DPDK PF host"); - else if ((vf->version_major == I40E_VIRTCHNL_VERSION_MAJOR) && - (vf->version_minor <= I40E_VIRTCHNL_VERSION_MINOR)) + else if ((vf->version_major == VIRTCHNL_VERSION_MAJOR) && + (vf->version_minor <= VIRTCHNL_VERSION_MINOR)) PMD_DRV_LOG(INFO, "Peer is Linux PF host"); else { PMD_INIT_LOG(ERR, "PF/VF API version mismatch:(%u.%u)-(%u.%u)", vf->version_major, vf->version_minor, - I40E_VIRTCHNL_VERSION_MAJOR, - I40E_VIRTCHNL_VERSION_MINOR); + VIRTCHNL_VERSION_MAJOR, + VIRTCHNL_VERSION_MINOR); return -1; } @@ -452,15 +452,15 @@ i40evf_get_vf_resource(struct rte_eth_dev *dev) struct vf_cmd_info args; uint32_t caps, len; - args.ops = I40E_VIRTCHNL_OP_GET_VF_RESOURCES; + args.ops = VIRTCHNL_OP_GET_VF_RESOURCES; args.out_buffer = vf->aq_resp; args.out_size = I40E_AQ_BUF_SZ; if (PF_IS_V11(vf)) { - caps = I40E_VIRTCHNL_VF_OFFLOAD_L2 | - I40E_VIRTCHNL_VF_OFFLOAD_RSS_AQ | - I40E_VIRTCHNL_VF_OFFLOAD_RSS_REG | - I40E_VIRTCHNL_VF_OFFLOAD_VLAN | - I40E_VIRTCHNL_VF_OFFLOAD_RX_POLLING; + caps = VIRTCHNL_VF_OFFLOAD_L2 | + VIRTCHNL_VF_OFFLOAD_RSS_AQ | + VIRTCHNL_VF_OFFLOAD_RSS_REG | + VIRTCHNL_VF_OFFLOAD_VLAN | + VIRTCHNL_VF_OFFLOAD_RX_POLLING; args.in_args = (uint8_t *)∩︀ args.in_args_size = sizeof(caps); } else { @@ -474,8 +474,8 @@ i40evf_get_vf_resource(struct rte_eth_dev *dev) return err; } - len = sizeof(struct i40e_virtchnl_vf_resource) + - I40E_MAX_VF_VSI * sizeof(struct i40e_virtchnl_vsi_resource); + len = sizeof(struct virtchnl_vf_resource) + + I40E_MAX_VF_VSI * sizeof(struct virtchnl_vsi_resource); (void)rte_memcpy(vf->vf_res, args.out_buffer, RTE_MIN(args.out_size, len)); @@ -492,18 +492,18 @@ i40evf_config_promisc(struct rte_eth_dev *dev, struct i40e_vf *vf = I40EVF_DEV_PRIVATE_TO_VF(dev->data->dev_private); int err; struct vf_cmd_info args; - struct i40e_virtchnl_promisc_info promisc; + struct virtchnl_promisc_info promisc; promisc.flags = 0; promisc.vsi_id = vf->vsi_res->vsi_id; if (enable_unicast) - promisc.flags |= I40E_FLAG_VF_UNICAST_PROMISC; + promisc.flags |= FLAG_VF_UNICAST_PROMISC; if (enable_multicast) - promisc.flags |= I40E_FLAG_VF_MULTICAST_PROMISC; + promisc.flags |= FLAG_VF_MULTICAST_PROMISC; - args.ops = I40E_VIRTCHNL_OP_CONFIG_PROMISCUOUS_MODE; + args.ops = VIRTCHNL_OP_CONFIG_PROMISCUOUS_MODE; args.in_args = (uint8_t *)&promisc; args.in_args_size = sizeof(promisc); args.out_buffer = vf->aq_resp; @@ -525,12 +525,12 @@ i40evf_config_vlan_offload(struct rte_eth_dev *dev, struct i40e_vf *vf = I40EVF_DEV_PRIVATE_TO_VF(dev->data->dev_private); int err; struct vf_cmd_info args; - struct i40e_virtchnl_vlan_offload_info offload; + struct virtchnl_vlan_offload_info offload; offload.vsi_id = vf->vsi_res->vsi_id; offload.enable_vlan_strip = enable_vlan_strip; - args.ops = (enum i40e_virtchnl_ops)I40E_VIRTCHNL_OP_CFG_VLAN_OFFLOAD; + args.ops = (enum virtchnl_ops)I40E_VIRTCHNL_OP_CFG_VLAN_OFFLOAD; args.in_args = (uint8_t *)&offload; args.in_args_size = sizeof(offload); args.out_buffer = vf->aq_resp; @@ -550,7 +550,7 @@ i40evf_config_vlan_pvid(struct rte_eth_dev *dev, struct i40e_vf *vf = I40EVF_DEV_PRIVATE_TO_VF(dev->data->dev_private); int err; struct vf_cmd_info args; - struct i40e_virtchnl_pvid_info tpid_info; + struct virtchnl_pvid_info tpid_info; if (info == NULL) { PMD_DRV_LOG(ERR, "invalid parameters"); @@ -561,7 +561,7 @@ i40evf_config_vlan_pvid(struct rte_eth_dev *dev, tpid_info.vsi_id = vf->vsi_res->vsi_id; (void)rte_memcpy(&tpid_info.info, info, sizeof(*info)); - args.ops = (enum i40e_virtchnl_ops)I40E_VIRTCHNL_OP_CFG_VLAN_PVID; + args.ops = (enum virtchnl_ops)I40E_VIRTCHNL_OP_CFG_VLAN_PVID; args.in_args = (uint8_t *)&tpid_info; args.in_args_size = sizeof(tpid_info); args.out_buffer = vf->aq_resp; @@ -575,7 +575,7 @@ i40evf_config_vlan_pvid(struct rte_eth_dev *dev, } static void -i40evf_fill_virtchnl_vsi_txq_info(struct i40e_virtchnl_txq_info *txq_info, +i40evf_fill_virtchnl_vsi_txq_info(struct virtchnl_txq_info *txq_info, uint16_t vsi_id, uint16_t queue_id, uint16_t nb_txq, @@ -590,7 +590,7 @@ i40evf_fill_virtchnl_vsi_txq_info(struct i40e_virtchnl_txq_info *txq_info, } static void -i40evf_fill_virtchnl_vsi_rxq_info(struct i40e_virtchnl_rxq_info *rxq_info, +i40evf_fill_virtchnl_vsi_rxq_info(struct virtchnl_rxq_info *rxq_info, uint16_t vsi_id, uint16_t queue_id, uint16_t nb_rxq, @@ -618,8 +618,8 @@ i40evf_configure_vsi_queues(struct rte_eth_dev *dev) (struct i40e_rx_queue **)dev->data->rx_queues; struct i40e_tx_queue **txq = (struct i40e_tx_queue **)dev->data->tx_queues; - struct i40e_virtchnl_vsi_queue_config_info *vc_vqci; - struct i40e_virtchnl_queue_pair_info *vc_qpi; + struct virtchnl_vsi_queue_config_info *vc_vqci; + struct virtchnl_queue_pair_info *vc_qpi; struct vf_cmd_info args; uint16_t i, nb_qp = vf->num_queue_pairs; const uint32_t size = @@ -628,7 +628,7 @@ i40evf_configure_vsi_queues(struct rte_eth_dev *dev) int ret; memset(buff, 0, sizeof(buff)); - vc_vqci = (struct i40e_virtchnl_vsi_queue_config_info *)buff; + vc_vqci = (struct virtchnl_vsi_queue_config_info *)buff; vc_vqci->vsi_id = vf->vsi_res->vsi_id; vc_vqci->num_queue_pairs = nb_qp; @@ -640,7 +640,7 @@ i40evf_configure_vsi_queues(struct rte_eth_dev *dev) vf->max_pkt_len, rxq[i]); } memset(&args, 0, sizeof(args)); - args.ops = I40E_VIRTCHNL_OP_CONFIG_VSI_QUEUES; + args.ops = VIRTCHNL_OP_CONFIG_VSI_QUEUES; args.in_args = (uint8_t *)vc_vqci; args.in_args_size = size; args.out_buffer = vf->aq_resp; @@ -648,7 +648,7 @@ i40evf_configure_vsi_queues(struct rte_eth_dev *dev) ret = i40evf_execute_vf_cmd(dev, &args); if (ret) PMD_DRV_LOG(ERR, "Failed to execute command of " - "I40E_VIRTCHNL_OP_CONFIG_VSI_QUEUES"); + "VIRTCHNL_OP_CONFIG_VSI_QUEUES"); return ret; } @@ -662,8 +662,8 @@ i40evf_configure_vsi_queues_ext(struct rte_eth_dev *dev) (struct i40e_rx_queue **)dev->data->rx_queues; struct i40e_tx_queue **txq = (struct i40e_tx_queue **)dev->data->tx_queues; - struct i40e_virtchnl_vsi_queue_config_ext_info *vc_vqcei; - struct i40e_virtchnl_queue_pair_ext_info *vc_qpei; + struct virtchnl_vsi_queue_config_ext_info *vc_vqcei; + struct virtchnl_queue_pair_ext_info *vc_qpei; struct vf_cmd_info args; uint16_t i, nb_qp = vf->num_queue_pairs; const uint32_t size = @@ -672,7 +672,7 @@ i40evf_configure_vsi_queues_ext(struct rte_eth_dev *dev) int ret; memset(buff, 0, sizeof(buff)); - vc_vqcei = (struct i40e_virtchnl_vsi_queue_config_ext_info *)buff; + vc_vqcei = (struct virtchnl_vsi_queue_config_ext_info *)buff; vc_vqcei->vsi_id = vf->vsi_res->vsi_id; vc_vqcei->num_queue_pairs = nb_qp; vc_qpei = vc_vqcei->qpair; @@ -693,7 +693,7 @@ i40evf_configure_vsi_queues_ext(struct rte_eth_dev *dev) } memset(&args, 0, sizeof(args)); args.ops = - (enum i40e_virtchnl_ops)I40E_VIRTCHNL_OP_CONFIG_VSI_QUEUES_EXT; + (enum virtchnl_ops)VIRTCHNL_OP_CONFIG_VSI_QUEUES_EXT; args.in_args = (uint8_t *)vc_vqcei; args.in_args_size = size; args.out_buffer = vf->aq_resp; @@ -701,7 +701,7 @@ i40evf_configure_vsi_queues_ext(struct rte_eth_dev *dev) ret = i40evf_execute_vf_cmd(dev, &args); if (ret) PMD_DRV_LOG(ERR, "Failed to execute command of " - "I40E_VIRTCHNL_OP_CONFIG_VSI_QUEUES_EXT"); + "VIRTCHNL_OP_CONFIG_VSI_QUEUES_EXT"); return ret; } @@ -724,9 +724,9 @@ i40evf_config_irq_map(struct rte_eth_dev *dev) { struct i40e_vf *vf = I40EVF_DEV_PRIVATE_TO_VF(dev->data->dev_private); struct vf_cmd_info args; - uint8_t cmd_buffer[sizeof(struct i40e_virtchnl_irq_map_info) + \ - sizeof(struct i40e_virtchnl_vector_map)]; - struct i40e_virtchnl_irq_map_info *map_info; + uint8_t cmd_buffer[sizeof(struct virtchnl_irq_map_info) + \ + sizeof(struct virtchnl_vector_map)]; + struct virtchnl_irq_map_info *map_info; struct rte_pci_device *pci_dev = RTE_ETH_DEV_TO_PCI(dev); struct rte_intr_handle *intr_handle = &pci_dev->intr_handle; uint32_t vector_id; @@ -741,7 +741,7 @@ i40evf_config_irq_map(struct rte_eth_dev *dev) vector_id = I40E_MISC_VEC_ID; } - map_info = (struct i40e_virtchnl_irq_map_info *)cmd_buffer; + map_info = (struct virtchnl_irq_map_info *)cmd_buffer; map_info->num_vectors = 1; map_info->vecmap[0].rxitr_idx = I40E_ITR_INDEX_DEFAULT; map_info->vecmap[0].vsi_id = vf->vsi_res->vsi_id; @@ -756,7 +756,7 @@ i40evf_config_irq_map(struct rte_eth_dev *dev) intr_handle->intr_vec[i] = vector_id; } - args.ops = I40E_VIRTCHNL_OP_CONFIG_IRQ_MAP; + args.ops = VIRTCHNL_OP_CONFIG_IRQ_MAP; args.in_args = (u8 *)cmd_buffer; args.in_args_size = sizeof(cmd_buffer); args.out_buffer = vf->aq_resp; @@ -773,7 +773,7 @@ i40evf_switch_queue(struct rte_eth_dev *dev, bool isrx, uint16_t qid, bool on) { struct i40e_vf *vf = I40EVF_DEV_PRIVATE_TO_VF(dev->data->dev_private); - struct i40e_virtchnl_queue_select queue_select; + struct virtchnl_queue_select queue_select; int err; struct vf_cmd_info args; memset(&queue_select, 0, sizeof(queue_select)); @@ -785,9 +785,9 @@ i40evf_switch_queue(struct rte_eth_dev *dev, bool isrx, uint16_t qid, queue_select.tx_queues |= 1 << qid; if (on) - args.ops = I40E_VIRTCHNL_OP_ENABLE_QUEUES; + args.ops = VIRTCHNL_OP_ENABLE_QUEUES; else - args.ops = I40E_VIRTCHNL_OP_DISABLE_QUEUES; + args.ops = VIRTCHNL_OP_DISABLE_QUEUES; args.in_args = (u8 *)&queue_select; args.in_args_size = sizeof(queue_select); args.out_buffer = vf->aq_resp; @@ -861,10 +861,10 @@ i40evf_add_mac_addr(struct rte_eth_dev *dev, __rte_unused uint32_t index, __rte_unused uint32_t pool) { - struct i40e_virtchnl_ether_addr_list *list; + struct virtchnl_ether_addr_list *list; struct i40e_vf *vf = I40EVF_DEV_PRIVATE_TO_VF(dev->data->dev_private); - uint8_t cmd_buffer[sizeof(struct i40e_virtchnl_ether_addr_list) + \ - sizeof(struct i40e_virtchnl_ether_addr)]; + uint8_t cmd_buffer[sizeof(struct virtchnl_ether_addr_list) + \ + sizeof(struct virtchnl_ether_addr)]; int err; struct vf_cmd_info args; @@ -876,13 +876,13 @@ i40evf_add_mac_addr(struct rte_eth_dev *dev, return I40E_ERR_INVALID_MAC_ADDR; } - list = (struct i40e_virtchnl_ether_addr_list *)cmd_buffer; + list = (struct virtchnl_ether_addr_list *)cmd_buffer; list->vsi_id = vf->vsi_res->vsi_id; list->num_elements = 1; (void)rte_memcpy(list->list[0].addr, addr->addr_bytes, sizeof(addr->addr_bytes)); - args.ops = I40E_VIRTCHNL_OP_ADD_ETHER_ADDRESS; + args.ops = VIRTCHNL_OP_ADD_ETH_ADDR; args.in_args = cmd_buffer; args.in_args_size = sizeof(cmd_buffer); args.out_buffer = vf->aq_resp; @@ -899,10 +899,10 @@ static void i40evf_del_mac_addr_by_addr(struct rte_eth_dev *dev, struct ether_addr *addr) { - struct i40e_virtchnl_ether_addr_list *list; + struct virtchnl_ether_addr_list *list; struct i40e_vf *vf = I40EVF_DEV_PRIVATE_TO_VF(dev->data->dev_private); - uint8_t cmd_buffer[sizeof(struct i40e_virtchnl_ether_addr_list) + \ - sizeof(struct i40e_virtchnl_ether_addr)]; + uint8_t cmd_buffer[sizeof(struct virtchnl_ether_addr_list) + \ + sizeof(struct virtchnl_ether_addr)]; int err; struct vf_cmd_info args; @@ -914,13 +914,13 @@ i40evf_del_mac_addr_by_addr(struct rte_eth_dev *dev, return; } - list = (struct i40e_virtchnl_ether_addr_list *)cmd_buffer; + list = (struct virtchnl_ether_addr_list *)cmd_buffer; list->vsi_id = vf->vsi_res->vsi_id; list->num_elements = 1; (void)rte_memcpy(list->list[0].addr, addr->addr_bytes, sizeof(addr->addr_bytes)); - args.ops = I40E_VIRTCHNL_OP_DEL_ETHER_ADDRESS; + args.ops = VIRTCHNL_OP_DEL_ETH_ADDR; args.in_args = cmd_buffer; args.in_args_size = sizeof(cmd_buffer); args.out_buffer = vf->aq_resp; @@ -947,13 +947,13 @@ static int i40evf_update_stats(struct rte_eth_dev *dev, struct i40e_eth_stats **pstats) { struct i40e_vf *vf = I40EVF_DEV_PRIVATE_TO_VF(dev->data->dev_private); - struct i40e_virtchnl_queue_select q_stats; + struct virtchnl_queue_select q_stats; int err; struct vf_cmd_info args; memset(&q_stats, 0, sizeof(q_stats)); q_stats.vsi_id = vf->vsi_res->vsi_id; - args.ops = I40E_VIRTCHNL_OP_GET_STATS; + args.ops = VIRTCHNL_OP_GET_STATS; args.in_args = (u8 *)&q_stats; args.in_args_size = sizeof(q_stats); args.out_buffer = vf->aq_resp; @@ -1050,18 +1050,18 @@ static int i40evf_add_vlan(struct rte_eth_dev *dev, uint16_t vlanid) { struct i40e_vf *vf = I40EVF_DEV_PRIVATE_TO_VF(dev->data->dev_private); - struct i40e_virtchnl_vlan_filter_list *vlan_list; - uint8_t cmd_buffer[sizeof(struct i40e_virtchnl_vlan_filter_list) + + struct virtchnl_vlan_filter_list *vlan_list; + uint8_t cmd_buffer[sizeof(struct virtchnl_vlan_filter_list) + sizeof(uint16_t)]; int err; struct vf_cmd_info args; - vlan_list = (struct i40e_virtchnl_vlan_filter_list *)cmd_buffer; + vlan_list = (struct virtchnl_vlan_filter_list *)cmd_buffer; vlan_list->vsi_id = vf->vsi_res->vsi_id; vlan_list->num_elements = 1; vlan_list->vlan_id[0] = vlanid; - args.ops = I40E_VIRTCHNL_OP_ADD_VLAN; + args.ops = VIRTCHNL_OP_ADD_VLAN; args.in_args = (u8 *)&cmd_buffer; args.in_args_size = sizeof(cmd_buffer); args.out_buffer = vf->aq_resp; @@ -1077,18 +1077,18 @@ static int i40evf_del_vlan(struct rte_eth_dev *dev, uint16_t vlanid) { struct i40e_vf *vf = I40EVF_DEV_PRIVATE_TO_VF(dev->data->dev_private); - struct i40e_virtchnl_vlan_filter_list *vlan_list; - uint8_t cmd_buffer[sizeof(struct i40e_virtchnl_vlan_filter_list) + + struct virtchnl_vlan_filter_list *vlan_list; + uint8_t cmd_buffer[sizeof(struct virtchnl_vlan_filter_list) + sizeof(uint16_t)]; int err; struct vf_cmd_info args; - vlan_list = (struct i40e_virtchnl_vlan_filter_list *)cmd_buffer; + vlan_list = (struct virtchnl_vlan_filter_list *)cmd_buffer; vlan_list->vsi_id = vf->vsi_res->vsi_id; vlan_list->num_elements = 1; vlan_list->vlan_id[0] = vlanid; - args.ops = I40E_VIRTCHNL_OP_DEL_VLAN; + args.ops = VIRTCHNL_OP_DEL_VLAN; args.in_args = (u8 *)&cmd_buffer; args.in_args_size = sizeof(cmd_buffer); args.out_buffer = vf->aq_resp; @@ -1178,7 +1178,7 @@ i40evf_reset_vf(struct i40e_hw *hw) reset = rd32(hw, I40E_VFGEN_RSTAT) & I40E_VFGEN_RSTAT_VFR_STATE_MASK; reset = reset >> I40E_VFGEN_RSTAT_VFR_STATE_SHIFT; - if (I40E_VFR_COMPLETED == reset || I40E_VFR_VFACTIVE == reset) + if (VIRTCHNL_VFR_COMPLETED == reset || VIRTCHNL_VFR_VFACTIVE == reset) break; else rte_delay_ms(50); @@ -1242,8 +1242,8 @@ i40evf_init_vf(struct rte_eth_dev *dev) PMD_INIT_LOG(ERR, "check_api version failed"); goto err_aq; } - bufsz = sizeof(struct i40e_virtchnl_vf_resource) + - (I40E_MAX_VF_VSI * sizeof(struct i40e_virtchnl_vsi_resource)); + bufsz = sizeof(struct virtchnl_vf_resource) + + (I40E_MAX_VF_VSI * sizeof(struct virtchnl_vsi_resource)); vf->vf_res = rte_zmalloc("vf_res", bufsz, 0); if (!vf->vf_res) { PMD_INIT_LOG(ERR, "unable to allocate vf_res memory"); @@ -1257,7 +1257,7 @@ i40evf_init_vf(struct rte_eth_dev *dev) /* got VF config message back from PF, now we can parse it */ for (i = 0; i < vf->vf_res->num_vsis; i++) { - if (vf->vf_res->vsi_res[i].vsi_type == I40E_VSI_SRIOV) + if (vf->vf_res->vsi_res[i].vsi_type == VIRTCHNL_VSI_SRIOV) vf->vsi_res = &vf->vf_res->vsi_res[i]; } @@ -1269,7 +1269,7 @@ i40evf_init_vf(struct rte_eth_dev *dev) if (hw->mac.type == I40E_MAC_X722_VF) vf->flags = I40E_FLAG_RSS_AQ_CAPABLE; vf->vsi.vsi_id = vf->vsi_res->vsi_id; - vf->vsi.type = vf->vsi_res->vsi_type; + vf->vsi.type = (enum i40e_vsi_type)vf->vsi_res->vsi_type; vf->vsi.nb_qps = vf->vsi_res->num_queue_pairs; vf->vsi.adapter = I40E_DEV_PRIVATE_TO_ADAPTER(dev->data->dev_private); @@ -1321,21 +1321,21 @@ static void i40evf_handle_pf_event(struct rte_eth_dev *dev, uint8_t *msg, __rte_unused uint16_t msglen) { - struct i40e_virtchnl_pf_event *pf_msg = - (struct i40e_virtchnl_pf_event *)msg; + struct virtchnl_pf_event *pf_msg = + (struct virtchnl_pf_event *)msg; struct i40e_vf *vf = I40EVF_DEV_PRIVATE_TO_VF(dev->data->dev_private); switch (pf_msg->event) { - case I40E_VIRTCHNL_EVENT_RESET_IMPENDING: + case VIRTCHNL_EVENT_RESET_IMPENDING: PMD_DRV_LOG(DEBUG, "VIRTCHNL_EVENT_RESET_IMPENDING event"); _rte_eth_dev_callback_process(dev, RTE_ETH_EVENT_INTR_RESET, NULL); break; - case I40E_VIRTCHNL_EVENT_LINK_CHANGE: + case VIRTCHNL_EVENT_LINK_CHANGE: PMD_DRV_LOG(DEBUG, "VIRTCHNL_EVENT_LINK_CHANGE event"); vf->link_up = pf_msg->event_data.link_event.link_status; vf->link_speed = pf_msg->event_data.link_event.link_speed; break; - case I40E_VIRTCHNL_EVENT_PF_DRIVER_CLOSE: + case VIRTCHNL_EVENT_PF_DRIVER_CLOSE: PMD_DRV_LOG(DEBUG, "VIRTCHNL_EVENT_PF_DRIVER_CLOSE event"); break; default: @@ -1351,7 +1351,7 @@ i40evf_handle_aq_msg(struct rte_eth_dev *dev) struct i40e_vf *vf = I40EVF_DEV_PRIVATE_TO_VF(dev->data->dev_private); struct i40e_arq_event_info info; uint16_t pending, aq_opc; - enum i40e_virtchnl_ops msg_opc; + enum virtchnl_ops msg_opc; enum i40e_status_code msg_ret; int ret; @@ -1376,13 +1376,13 @@ i40evf_handle_aq_msg(struct rte_eth_dev *dev) * cookie_high of struct i40e_aq_desc, while return error code * are stored in cookie_low, Which is done by * i40e_aq_send_msg_to_vf in PF driver.*/ - msg_opc = (enum i40e_virtchnl_ops)rte_le_to_cpu_32( + msg_opc = (enum virtchnl_ops)rte_le_to_cpu_32( info.desc.cookie_high); msg_ret = (enum i40e_status_code)rte_le_to_cpu_32( info.desc.cookie_low); switch (aq_opc) { case i40e_aqc_opc_send_msg_to_vf: - if (msg_opc == I40E_VIRTCHNL_OP_EVENT) + if (msg_opc == VIRTCHNL_OP_EVENT) /* process event*/ i40evf_handle_pf_event(dev, info.msg_buf, info.msg_len); @@ -1591,8 +1591,8 @@ i40evf_dev_configure(struct rte_eth_dev *dev) */ if (!conf->rxmode.hw_strip_crc) { vf = I40EVF_DEV_PRIVATE_TO_VF(dev->data->dev_private); - if ((vf->version_major == I40E_VIRTCHNL_VERSION_MAJOR) && - (vf->version_minor <= I40E_VIRTCHNL_VERSION_MINOR)) { + if ((vf->version_major == VIRTCHNL_VERSION_MAJOR) && + (vf->version_minor <= VIRTCHNL_VERSION_MINOR)) { /* Peer is running non-DPDK PF driver. */ PMD_INIT_LOG(ERR, "VF can't disable HW CRC Strip"); return -EINVAL; @@ -2000,7 +2000,7 @@ i40evf_dev_rx_queue_intr_disable(struct rte_eth_dev *dev, uint16_t queue_id) static void i40evf_add_del_all_mac_addr(struct rte_eth_dev *dev, bool add) { - struct i40e_virtchnl_ether_addr_list *list; + struct virtchnl_ether_addr_list *list; struct i40e_vf *vf = I40EVF_DEV_PRIVATE_TO_VF(dev->data->dev_private); int err, i, j; int next_begin = 0; @@ -2011,11 +2011,11 @@ i40evf_add_del_all_mac_addr(struct rte_eth_dev *dev, bool add) do { j = 0; - len = sizeof(struct i40e_virtchnl_ether_addr_list); + len = sizeof(struct virtchnl_ether_addr_list); for (i = begin; i < I40E_NUM_MACADDR_MAX; i++, next_begin++) { if (is_zero_ether_addr(&dev->data->mac_addrs[i])) continue; - len += sizeof(struct i40e_virtchnl_ether_addr); + len += sizeof(struct virtchnl_ether_addr); if (len >= I40E_AQ_BUF_SZ) { next_begin = i + 1; break; @@ -2042,8 +2042,8 @@ i40evf_add_del_all_mac_addr(struct rte_eth_dev *dev, bool add) } list->vsi_id = vf->vsi_res->vsi_id; list->num_elements = j; - args.ops = add ? I40E_VIRTCHNL_OP_ADD_ETHER_ADDRESS : - I40E_VIRTCHNL_OP_DEL_ETHER_ADDRESS; + args.ops = add ? VIRTCHNL_OP_ADD_ETH_ADDR : + VIRTCHNL_OP_DEL_ETH_ADDR; args.in_args = (uint8_t *)list; args.in_args_size = len; args.out_buffer = vf->aq_resp; diff --git a/drivers/net/i40e/i40e_pf.c b/drivers/net/i40e/i40e_pf.c index 0758503..f9b4be5 100644 --- a/drivers/net/i40e/i40e_pf.c +++ b/drivers/net/i40e/i40e_pf.c @@ -61,7 +61,7 @@ static int i40e_pf_host_switch_queues(struct i40e_pf_vf *vf, - struct i40e_virtchnl_queue_select *qsel, + struct virtchnl_queue_select *qsel, bool on); /** @@ -128,7 +128,7 @@ i40e_pf_host_vf_reset(struct i40e_pf_vf *vf, bool do_hw_reset) struct i40e_pf *pf; uint16_t vf_id, abs_vf_id, vf_msix_num; int ret; - struct i40e_virtchnl_queue_select qsel; + struct virtchnl_queue_select qsel; if (vf == NULL) return -EINVAL; @@ -139,7 +139,7 @@ i40e_pf_host_vf_reset(struct i40e_pf_vf *vf, bool do_hw_reset) abs_vf_id = vf_id + hw->func_caps.vf_base_id; /* Notify VF that we are in VFR progress */ - I40E_WRITE_REG(hw, I40E_VFGEN_RSTAT1(vf_id), I40E_VFR_INPROGRESS); + I40E_WRITE_REG(hw, I40E_VFGEN_RSTAT1(vf_id), VIRTCHNL_VFR_INPROGRESS); /* * If require a SW VF reset, a VFLR interrupt will be generated, @@ -220,7 +220,7 @@ i40e_pf_host_vf_reset(struct i40e_pf_vf *vf, bool do_hw_reset) } /* Reset done, Set COMPLETE flag and clear reset bit */ - I40E_WRITE_REG(hw, I40E_VFGEN_RSTAT1(vf_id), I40E_VFR_COMPLETED); + I40E_WRITE_REG(hw, I40E_VFGEN_RSTAT1(vf_id), VIRTCHNL_VFR_COMPLETED); val = I40E_READ_REG(hw, I40E_VPGEN_VFRTRIG(vf_id)); val &= ~I40E_VPGEN_VFRTRIG_VFSWR_MASK; I40E_WRITE_REG(hw, I40E_VPGEN_VFRTRIG(vf_id), val); @@ -248,7 +248,7 @@ i40e_pf_host_vf_reset(struct i40e_pf_vf *vf, bool do_hw_reset) return -EFAULT; } - I40E_WRITE_REG(hw, I40E_VFGEN_RSTAT1(vf_id), I40E_VFR_VFACTIVE); + I40E_WRITE_REG(hw, I40E_VFGEN_RSTAT1(vf_id), VIRTCHNL_VFR_VFACTIVE); return ret; } @@ -277,7 +277,7 @@ i40e_pf_host_send_msg_to_vf(struct i40e_pf_vf *vf, static void i40e_pf_host_process_cmd_version(struct i40e_pf_vf *vf, bool b_op) { - struct i40e_virtchnl_version_info info; + struct virtchnl_version_info info; /* Respond like a Linux PF host in order to support both DPDK VF and * Linux VF driver. The expense is original DPDK host specific feature @@ -286,16 +286,16 @@ i40e_pf_host_process_cmd_version(struct i40e_pf_vf *vf, bool b_op) * DPDK VF also can't identify host driver by version number returned. * It always assume talking with Linux PF. */ - info.major = I40E_VIRTCHNL_VERSION_MAJOR; - info.minor = I40E_VIRTCHNL_VERSION_MINOR_NO_VF_CAPS; + info.major = VIRTCHNL_VERSION_MAJOR; + info.minor = VIRTCHNL_VERSION_MINOR_NO_VF_CAPS; if (b_op) - i40e_pf_host_send_msg_to_vf(vf, I40E_VIRTCHNL_OP_VERSION, + i40e_pf_host_send_msg_to_vf(vf, VIRTCHNL_OP_VERSION, I40E_SUCCESS, (uint8_t *)&info, sizeof(info)); else - i40e_pf_host_send_msg_to_vf(vf, I40E_VIRTCHNL_OP_VERSION, + i40e_pf_host_send_msg_to_vf(vf, VIRTCHNL_OP_VERSION, I40E_NOT_SUPPORTED, (uint8_t *)&info, sizeof(info)); @@ -313,22 +313,22 @@ i40e_pf_host_process_cmd_reset_vf(struct i40e_pf_vf *vf) static int i40e_pf_host_process_cmd_get_vf_resource(struct i40e_pf_vf *vf, bool b_op) { - struct i40e_virtchnl_vf_resource *vf_res = NULL; + struct virtchnl_vf_resource *vf_res = NULL; struct i40e_hw *hw = I40E_PF_TO_HW(vf->pf); uint32_t len = 0; int ret = I40E_SUCCESS; if (!b_op) { i40e_pf_host_send_msg_to_vf(vf, - I40E_VIRTCHNL_OP_GET_VF_RESOURCES, + VIRTCHNL_OP_GET_VF_RESOURCES, I40E_NOT_SUPPORTED, NULL, 0); return ret; } /* only have 1 VSI by default */ - len = sizeof(struct i40e_virtchnl_vf_resource) + + len = sizeof(struct virtchnl_vf_resource) + I40E_DEFAULT_VF_VSI_NUM * - sizeof(struct i40e_virtchnl_vsi_resource); + sizeof(struct virtchnl_vsi_resource); vf_res = rte_zmalloc("i40e_vf_res", len, 0); if (vf_res == NULL) { @@ -339,21 +339,21 @@ i40e_pf_host_process_cmd_get_vf_resource(struct i40e_pf_vf *vf, bool b_op) goto send_msg; } - vf_res->vf_offload_flags = I40E_VIRTCHNL_VF_OFFLOAD_L2 | - I40E_VIRTCHNL_VF_OFFLOAD_VLAN; + vf_res->vf_offload_flags = VIRTCHNL_VF_OFFLOAD_L2 | + VIRTCHNL_VF_OFFLOAD_VLAN; vf_res->max_vectors = hw->func_caps.num_msix_vectors_vf; vf_res->num_queue_pairs = vf->vsi->nb_qps; vf_res->num_vsis = I40E_DEFAULT_VF_VSI_NUM; /* Change below setting if PF host can support more VSIs for VF */ - vf_res->vsi_res[0].vsi_type = I40E_VSI_SRIOV; + vf_res->vsi_res[0].vsi_type = VIRTCHNL_VSI_SRIOV; vf_res->vsi_res[0].vsi_id = vf->vsi->vsi_id; vf_res->vsi_res[0].num_queue_pairs = vf->vsi->nb_qps; ether_addr_copy(&vf->mac_addr, (struct ether_addr *)vf_res->vsi_res[0].default_mac_addr); send_msg: - i40e_pf_host_send_msg_to_vf(vf, I40E_VIRTCHNL_OP_GET_VF_RESOURCES, + i40e_pf_host_send_msg_to_vf(vf, VIRTCHNL_OP_GET_VF_RESOURCES, ret, (uint8_t *)vf_res, len); rte_free(vf_res); @@ -363,7 +363,7 @@ i40e_pf_host_process_cmd_get_vf_resource(struct i40e_pf_vf *vf, bool b_op) static int i40e_pf_host_hmc_config_rxq(struct i40e_hw *hw, struct i40e_pf_vf *vf, - struct i40e_virtchnl_rxq_info *rxq, + struct virtchnl_rxq_info *rxq, uint8_t crcstrip) { int err = I40E_SUCCESS; @@ -431,7 +431,7 @@ i40e_vsi_get_tc_of_queue(struct i40e_vsi *vsi, static int i40e_pf_host_hmc_config_txq(struct i40e_hw *hw, struct i40e_pf_vf *vf, - struct i40e_virtchnl_txq_info *txq) + struct virtchnl_txq_info *txq) { int err = I40E_SUCCESS; struct i40e_hmc_obj_txq tx_ctx; @@ -480,14 +480,14 @@ i40e_pf_host_process_cmd_config_vsi_queues(struct i40e_pf_vf *vf, { struct i40e_hw *hw = I40E_PF_TO_HW(vf->pf); struct i40e_vsi *vsi = vf->vsi; - struct i40e_virtchnl_vsi_queue_config_info *vc_vqci = - (struct i40e_virtchnl_vsi_queue_config_info *)msg; - struct i40e_virtchnl_queue_pair_info *vc_qpi; + struct virtchnl_vsi_queue_config_info *vc_vqci = + (struct virtchnl_vsi_queue_config_info *)msg; + struct virtchnl_queue_pair_info *vc_qpi; int i, ret = I40E_SUCCESS; if (!b_op) { i40e_pf_host_send_msg_to_vf(vf, - I40E_VIRTCHNL_OP_CONFIG_VSI_QUEUES, + VIRTCHNL_OP_CONFIG_VSI_QUEUES, I40E_NOT_SUPPORTED, NULL, 0); return ret; } @@ -511,9 +511,9 @@ i40e_pf_host_process_cmd_config_vsi_queues(struct i40e_pf_vf *vf, /* * Apply VF RX queue setting to HMC. - * If the opcode is I40E_VIRTCHNL_OP_CONFIG_VSI_QUEUES_EXT, + * If the opcode is VIRTCHNL_OP_CONFIG_VSI_QUEUES_EXT, * then the extra information of - * 'struct i40e_virtchnl_queue_pair_extra_info' is needed, + * 'struct virtchnl_queue_pair_extra_info' is needed, * otherwise set the last parameter to NULL. */ if (i40e_pf_host_hmc_config_rxq(hw, vf, &vc_qpi[i].rxq, @@ -533,7 +533,7 @@ i40e_pf_host_process_cmd_config_vsi_queues(struct i40e_pf_vf *vf, } send_msg: - i40e_pf_host_send_msg_to_vf(vf, I40E_VIRTCHNL_OP_CONFIG_VSI_QUEUES, + i40e_pf_host_send_msg_to_vf(vf, VIRTCHNL_OP_CONFIG_VSI_QUEUES, ret, NULL, 0); return ret; @@ -547,15 +547,15 @@ i40e_pf_host_process_cmd_config_vsi_queues_ext(struct i40e_pf_vf *vf, { struct i40e_hw *hw = I40E_PF_TO_HW(vf->pf); struct i40e_vsi *vsi = vf->vsi; - struct i40e_virtchnl_vsi_queue_config_ext_info *vc_vqcei = - (struct i40e_virtchnl_vsi_queue_config_ext_info *)msg; - struct i40e_virtchnl_queue_pair_ext_info *vc_qpei; + struct virtchnl_vsi_queue_config_ext_info *vc_vqcei = + (struct virtchnl_vsi_queue_config_ext_info *)msg; + struct virtchnl_queue_pair_ext_info *vc_qpei; int i, ret = I40E_SUCCESS; if (!b_op) { i40e_pf_host_send_msg_to_vf( vf, - I40E_VIRTCHNL_OP_CONFIG_VSI_QUEUES_EXT, + VIRTCHNL_OP_CONFIG_VSI_QUEUES_EXT, I40E_NOT_SUPPORTED, NULL, 0); return ret; } @@ -578,9 +578,9 @@ i40e_pf_host_process_cmd_config_vsi_queues_ext(struct i40e_pf_vf *vf, } /* * Apply VF RX queue setting to HMC. - * If the opcode is I40E_VIRTCHNL_OP_CONFIG_VSI_QUEUES_EXT, + * If the opcode is VIRTCHNL_OP_CONFIG_VSI_QUEUES_EXT, * then the extra information of - * 'struct i40e_virtchnl_queue_pair_ext_info' is needed, + * 'struct virtchnl_queue_pair_ext_info' is needed, * otherwise set the last parameter to NULL. */ if (i40e_pf_host_hmc_config_rxq(hw, vf, &vc_qpei[i].rxq, @@ -600,7 +600,7 @@ i40e_pf_host_process_cmd_config_vsi_queues_ext(struct i40e_pf_vf *vf, } send_msg: - i40e_pf_host_send_msg_to_vf(vf, I40E_VIRTCHNL_OP_CONFIG_VSI_QUEUES_EXT, + i40e_pf_host_send_msg_to_vf(vf, VIRTCHNL_OP_CONFIG_VSI_QUEUES_EXT, ret, NULL, 0); return ret; @@ -608,7 +608,7 @@ i40e_pf_host_process_cmd_config_vsi_queues_ext(struct i40e_pf_vf *vf, static void i40e_pf_config_irq_link_list(struct i40e_pf_vf *vf, - struct i40e_virtchnl_vector_map *vvm) + struct virtchnl_vector_map *vvm) { #define BITS_PER_CHAR 8 uint64_t linklistmap = 0, tempmap; @@ -711,9 +711,9 @@ i40e_pf_host_process_cmd_config_irq_map(struct i40e_pf_vf *vf, int ret = I40E_SUCCESS; struct i40e_pf *pf = vf->pf; struct i40e_hw *hw = I40E_PF_TO_HW(vf->pf); - struct i40e_virtchnl_irq_map_info *irqmap = - (struct i40e_virtchnl_irq_map_info *)msg; - struct i40e_virtchnl_vector_map *map; + struct virtchnl_irq_map_info *irqmap = + (struct virtchnl_irq_map_info *)msg; + struct virtchnl_vector_map *map; int i; uint16_t vector_id; unsigned long qbit_max; @@ -721,12 +721,12 @@ i40e_pf_host_process_cmd_config_irq_map(struct i40e_pf_vf *vf, if (!b_op) { i40e_pf_host_send_msg_to_vf( vf, - I40E_VIRTCHNL_OP_CONFIG_IRQ_MAP, + VIRTCHNL_OP_CONFIG_IRQ_MAP, I40E_NOT_SUPPORTED, NULL, 0); return ret; } - if (msg == NULL || msglen < sizeof(struct i40e_virtchnl_irq_map_info)) { + if (msg == NULL || msglen < sizeof(struct virtchnl_irq_map_info)) { PMD_DRV_LOG(ERR, "buffer too short"); ret = I40E_ERR_PARAM; goto send_msg; @@ -773,7 +773,7 @@ i40e_pf_host_process_cmd_config_irq_map(struct i40e_pf_vf *vf, } send_msg: - i40e_pf_host_send_msg_to_vf(vf, I40E_VIRTCHNL_OP_CONFIG_IRQ_MAP, + i40e_pf_host_send_msg_to_vf(vf, VIRTCHNL_OP_CONFIG_IRQ_MAP, ret, NULL, 0); return ret; @@ -781,7 +781,7 @@ i40e_pf_host_process_cmd_config_irq_map(struct i40e_pf_vf *vf, static int i40e_pf_host_switch_queues(struct i40e_pf_vf *vf, - struct i40e_virtchnl_queue_select *qsel, + struct virtchnl_queue_select *qsel, bool on) { int ret = I40E_SUCCESS; @@ -831,8 +831,8 @@ i40e_pf_host_process_cmd_enable_queues(struct i40e_pf_vf *vf, uint16_t msglen) { int ret = I40E_SUCCESS; - struct i40e_virtchnl_queue_select *q_sel = - (struct i40e_virtchnl_queue_select *)msg; + struct virtchnl_queue_select *q_sel = + (struct virtchnl_queue_select *)msg; if (msg == NULL || msglen != sizeof(*q_sel)) { ret = I40E_ERR_PARAM; @@ -841,7 +841,7 @@ i40e_pf_host_process_cmd_enable_queues(struct i40e_pf_vf *vf, ret = i40e_pf_host_switch_queues(vf, q_sel, true); send_msg: - i40e_pf_host_send_msg_to_vf(vf, I40E_VIRTCHNL_OP_ENABLE_QUEUES, + i40e_pf_host_send_msg_to_vf(vf, VIRTCHNL_OP_ENABLE_QUEUES, ret, NULL, 0); return ret; @@ -854,13 +854,13 @@ i40e_pf_host_process_cmd_disable_queues(struct i40e_pf_vf *vf, bool b_op) { int ret = I40E_SUCCESS; - struct i40e_virtchnl_queue_select *q_sel = - (struct i40e_virtchnl_queue_select *)msg; + struct virtchnl_queue_select *q_sel = + (struct virtchnl_queue_select *)msg; if (!b_op) { i40e_pf_host_send_msg_to_vf( vf, - I40E_VIRTCHNL_OP_DISABLE_QUEUES, + VIRTCHNL_OP_DISABLE_QUEUES, I40E_NOT_SUPPORTED, NULL, 0); return ret; } @@ -872,7 +872,7 @@ i40e_pf_host_process_cmd_disable_queues(struct i40e_pf_vf *vf, ret = i40e_pf_host_switch_queues(vf, q_sel, false); send_msg: - i40e_pf_host_send_msg_to_vf(vf, I40E_VIRTCHNL_OP_DISABLE_QUEUES, + i40e_pf_host_send_msg_to_vf(vf, VIRTCHNL_OP_DISABLE_QUEUES, ret, NULL, 0); return ret; @@ -886,8 +886,8 @@ i40e_pf_host_process_cmd_add_ether_address(struct i40e_pf_vf *vf, bool b_op) { int ret = I40E_SUCCESS; - struct i40e_virtchnl_ether_addr_list *addr_list = - (struct i40e_virtchnl_ether_addr_list *)msg; + struct virtchnl_ether_addr_list *addr_list = + (struct virtchnl_ether_addr_list *)msg; struct i40e_mac_filter_info filter; int i; struct ether_addr *mac; @@ -895,7 +895,7 @@ i40e_pf_host_process_cmd_add_ether_address(struct i40e_pf_vf *vf, if (!b_op) { i40e_pf_host_send_msg_to_vf( vf, - I40E_VIRTCHNL_OP_ADD_ETHER_ADDRESS, + VIRTCHNL_OP_ADD_ETH_ADDR, I40E_NOT_SUPPORTED, NULL, 0); return ret; } @@ -920,7 +920,7 @@ i40e_pf_host_process_cmd_add_ether_address(struct i40e_pf_vf *vf, } send_msg: - i40e_pf_host_send_msg_to_vf(vf, I40E_VIRTCHNL_OP_ADD_ETHER_ADDRESS, + i40e_pf_host_send_msg_to_vf(vf, VIRTCHNL_OP_ADD_ETH_ADDR, ret, NULL, 0); return ret; @@ -933,15 +933,15 @@ i40e_pf_host_process_cmd_del_ether_address(struct i40e_pf_vf *vf, bool b_op) { int ret = I40E_SUCCESS; - struct i40e_virtchnl_ether_addr_list *addr_list = - (struct i40e_virtchnl_ether_addr_list *)msg; + struct virtchnl_ether_addr_list *addr_list = + (struct virtchnl_ether_addr_list *)msg; int i; struct ether_addr *mac; if (!b_op) { i40e_pf_host_send_msg_to_vf( vf, - I40E_VIRTCHNL_OP_DEL_ETHER_ADDRESS, + VIRTCHNL_OP_DEL_ETH_ADDR, I40E_NOT_SUPPORTED, NULL, 0); return ret; } @@ -962,7 +962,7 @@ i40e_pf_host_process_cmd_del_ether_address(struct i40e_pf_vf *vf, } send_msg: - i40e_pf_host_send_msg_to_vf(vf, I40E_VIRTCHNL_OP_DEL_ETHER_ADDRESS, + i40e_pf_host_send_msg_to_vf(vf, VIRTCHNL_OP_DEL_ETH_ADDR, ret, NULL, 0); return ret; @@ -974,15 +974,15 @@ i40e_pf_host_process_cmd_add_vlan(struct i40e_pf_vf *vf, bool b_op) { int ret = I40E_SUCCESS; - struct i40e_virtchnl_vlan_filter_list *vlan_filter_list = - (struct i40e_virtchnl_vlan_filter_list *)msg; + struct virtchnl_vlan_filter_list *vlan_filter_list = + (struct virtchnl_vlan_filter_list *)msg; int i; uint16_t *vid; if (!b_op) { i40e_pf_host_send_msg_to_vf( vf, - I40E_VIRTCHNL_OP_ADD_VLAN, + VIRTCHNL_OP_ADD_VLAN, I40E_NOT_SUPPORTED, NULL, 0); return ret; } @@ -1002,7 +1002,7 @@ i40e_pf_host_process_cmd_add_vlan(struct i40e_pf_vf *vf, } send_msg: - i40e_pf_host_send_msg_to_vf(vf, I40E_VIRTCHNL_OP_ADD_VLAN, + i40e_pf_host_send_msg_to_vf(vf, VIRTCHNL_OP_ADD_VLAN, ret, NULL, 0); return ret; @@ -1015,15 +1015,15 @@ i40e_pf_host_process_cmd_del_vlan(struct i40e_pf_vf *vf, bool b_op) { int ret = I40E_SUCCESS; - struct i40e_virtchnl_vlan_filter_list *vlan_filter_list = - (struct i40e_virtchnl_vlan_filter_list *)msg; + struct virtchnl_vlan_filter_list *vlan_filter_list = + (struct virtchnl_vlan_filter_list *)msg; int i; uint16_t *vid; if (!b_op) { i40e_pf_host_send_msg_to_vf( vf, - I40E_VIRTCHNL_OP_DEL_VLAN, + VIRTCHNL_OP_DEL_VLAN, I40E_NOT_SUPPORTED, NULL, 0); return ret; } @@ -1042,7 +1042,7 @@ i40e_pf_host_process_cmd_del_vlan(struct i40e_pf_vf *vf, } send_msg: - i40e_pf_host_send_msg_to_vf(vf, I40E_VIRTCHNL_OP_DEL_VLAN, + i40e_pf_host_send_msg_to_vf(vf, VIRTCHNL_OP_DEL_VLAN, ret, NULL, 0); return ret; @@ -1056,15 +1056,15 @@ i40e_pf_host_process_cmd_config_promisc_mode( bool b_op) { int ret = I40E_SUCCESS; - struct i40e_virtchnl_promisc_info *promisc = - (struct i40e_virtchnl_promisc_info *)msg; + struct virtchnl_promisc_info *promisc = + (struct virtchnl_promisc_info *)msg; struct i40e_hw *hw = I40E_PF_TO_HW(vf->pf); bool unicast = FALSE, multicast = FALSE; if (!b_op) { i40e_pf_host_send_msg_to_vf( vf, - I40E_VIRTCHNL_OP_CONFIG_PROMISCUOUS_MODE, + VIRTCHNL_OP_CONFIG_PROMISCUOUS_MODE, I40E_NOT_SUPPORTED, NULL, 0); return ret; } @@ -1074,21 +1074,21 @@ i40e_pf_host_process_cmd_config_promisc_mode( goto send_msg; } - if (promisc->flags & I40E_FLAG_VF_UNICAST_PROMISC) + if (promisc->flags & FLAG_VF_UNICAST_PROMISC) unicast = TRUE; ret = i40e_aq_set_vsi_unicast_promiscuous(hw, vf->vsi->seid, unicast, NULL, true); if (ret != I40E_SUCCESS) goto send_msg; - if (promisc->flags & I40E_FLAG_VF_MULTICAST_PROMISC) + if (promisc->flags & FLAG_VF_MULTICAST_PROMISC) multicast = TRUE; ret = i40e_aq_set_vsi_multicast_promiscuous(hw, vf->vsi->seid, multicast, NULL); send_msg: i40e_pf_host_send_msg_to_vf(vf, - I40E_VIRTCHNL_OP_CONFIG_PROMISCUOUS_MODE, ret, NULL, 0); + VIRTCHNL_OP_CONFIG_PROMISCUOUS_MODE, ret, NULL, 0); return ret; } @@ -1099,12 +1099,12 @@ i40e_pf_host_process_cmd_get_stats(struct i40e_pf_vf *vf, bool b_op) i40e_update_vsi_stats(vf->vsi); if (b_op) - i40e_pf_host_send_msg_to_vf(vf, I40E_VIRTCHNL_OP_GET_STATS, + i40e_pf_host_send_msg_to_vf(vf, VIRTCHNL_OP_GET_STATS, I40E_SUCCESS, (uint8_t *)&vf->vsi->eth_stats, sizeof(vf->vsi->eth_stats)); else - i40e_pf_host_send_msg_to_vf(vf, I40E_VIRTCHNL_OP_GET_STATS, + i40e_pf_host_send_msg_to_vf(vf, VIRTCHNL_OP_GET_STATS, I40E_NOT_SUPPORTED, (uint8_t *)&vf->vsi->eth_stats, sizeof(vf->vsi->eth_stats)); @@ -1120,8 +1120,8 @@ i40e_pf_host_process_cmd_cfg_vlan_offload( bool b_op) { int ret = I40E_SUCCESS; - struct i40e_virtchnl_vlan_offload_info *offload = - (struct i40e_virtchnl_vlan_offload_info *)msg; + struct virtchnl_vlan_offload_info *offload = + (struct virtchnl_vlan_offload_info *)msg; if (!b_op) { i40e_pf_host_send_msg_to_vf( @@ -1155,8 +1155,8 @@ i40e_pf_host_process_cmd_cfg_pvid(struct i40e_pf_vf *vf, bool b_op) { int ret = I40E_SUCCESS; - struct i40e_virtchnl_pvid_info *tpid_info = - (struct i40e_virtchnl_pvid_info *)msg; + struct virtchnl_pvid_info *tpid_info = + (struct virtchnl_pvid_info *)msg; if (!b_op) { i40e_pf_host_send_msg_to_vf( @@ -1183,39 +1183,39 @@ i40e_pf_host_process_cmd_cfg_pvid(struct i40e_pf_vf *vf, void i40e_notify_vf_link_status(struct rte_eth_dev *dev, struct i40e_pf_vf *vf) { - struct i40e_virtchnl_pf_event event; + struct virtchnl_pf_event event; - event.event = I40E_VIRTCHNL_EVENT_LINK_CHANGE; + event.event = VIRTCHNL_EVENT_LINK_CHANGE; event.event_data.link_event.link_status = dev->data->dev_link.link_status; - /* need to convert the ETH_SPEED_xxx into I40E_LINK_SPEED_xxx */ + /* need to convert the ETH_SPEED_xxx into VIRTCHNL_LINK_SPEED_xxx */ switch (dev->data->dev_link.link_speed) { case ETH_SPEED_NUM_100M: - event.event_data.link_event.link_speed = I40E_LINK_SPEED_100MB; + event.event_data.link_event.link_speed = VIRTCHNL_LINK_SPEED_100MB; break; case ETH_SPEED_NUM_1G: - event.event_data.link_event.link_speed = I40E_LINK_SPEED_1GB; + event.event_data.link_event.link_speed = VIRTCHNL_LINK_SPEED_1GB; break; case ETH_SPEED_NUM_10G: - event.event_data.link_event.link_speed = I40E_LINK_SPEED_10GB; + event.event_data.link_event.link_speed = VIRTCHNL_LINK_SPEED_10GB; break; case ETH_SPEED_NUM_20G: - event.event_data.link_event.link_speed = I40E_LINK_SPEED_20GB; + event.event_data.link_event.link_speed = VIRTCHNL_LINK_SPEED_20GB; break; case ETH_SPEED_NUM_25G: - event.event_data.link_event.link_speed = I40E_LINK_SPEED_25GB; + event.event_data.link_event.link_speed = VIRTCHNL_LINK_SPEED_25GB; break; case ETH_SPEED_NUM_40G: - event.event_data.link_event.link_speed = I40E_LINK_SPEED_40GB; + event.event_data.link_event.link_speed = VIRTCHNL_LINK_SPEED_40GB; break; default: event.event_data.link_event.link_speed = - I40E_LINK_SPEED_UNKNOWN; + VIRTCHNL_LINK_SPEED_UNKNOWN; break; } - i40e_pf_host_send_msg_to_vf(vf, I40E_VIRTCHNL_OP_EVENT, + i40e_pf_host_send_msg_to_vf(vf, VIRTCHNL_OP_EVENT, I40E_SUCCESS, (uint8_t *)&event, sizeof(event)); } @@ -1273,71 +1273,71 @@ i40e_pf_host_handle_vf_msg(struct rte_eth_dev *dev, } switch (opcode) { - case I40E_VIRTCHNL_OP_VERSION : + case VIRTCHNL_OP_VERSION: PMD_DRV_LOG(INFO, "OP_VERSION received"); i40e_pf_host_process_cmd_version(vf, b_op); break; - case I40E_VIRTCHNL_OP_RESET_VF : + case VIRTCHNL_OP_RESET_VF: PMD_DRV_LOG(INFO, "OP_RESET_VF received"); i40e_pf_host_process_cmd_reset_vf(vf); break; - case I40E_VIRTCHNL_OP_GET_VF_RESOURCES: + case VIRTCHNL_OP_GET_VF_RESOURCES: PMD_DRV_LOG(INFO, "OP_GET_VF_RESOURCES received"); i40e_pf_host_process_cmd_get_vf_resource(vf, b_op); break; - case I40E_VIRTCHNL_OP_CONFIG_VSI_QUEUES: + case VIRTCHNL_OP_CONFIG_VSI_QUEUES: PMD_DRV_LOG(INFO, "OP_CONFIG_VSI_QUEUES received"); i40e_pf_host_process_cmd_config_vsi_queues(vf, msg, msglen, b_op); break; - case I40E_VIRTCHNL_OP_CONFIG_VSI_QUEUES_EXT: + case VIRTCHNL_OP_CONFIG_VSI_QUEUES_EXT: PMD_DRV_LOG(INFO, "OP_CONFIG_VSI_QUEUES_EXT received"); i40e_pf_host_process_cmd_config_vsi_queues_ext(vf, msg, msglen, b_op); break; - case I40E_VIRTCHNL_OP_CONFIG_IRQ_MAP: + case VIRTCHNL_OP_CONFIG_IRQ_MAP: PMD_DRV_LOG(INFO, "OP_CONFIG_IRQ_MAP received"); i40e_pf_host_process_cmd_config_irq_map(vf, msg, msglen, b_op); break; - case I40E_VIRTCHNL_OP_ENABLE_QUEUES: + case VIRTCHNL_OP_ENABLE_QUEUES: PMD_DRV_LOG(INFO, "OP_ENABLE_QUEUES received"); if (b_op) { i40e_pf_host_process_cmd_enable_queues(vf, msg, msglen); i40e_notify_vf_link_status(dev, vf); } else { i40e_pf_host_send_msg_to_vf( - vf, I40E_VIRTCHNL_OP_ENABLE_QUEUES, + vf, VIRTCHNL_OP_ENABLE_QUEUES, I40E_NOT_SUPPORTED, NULL, 0); } break; - case I40E_VIRTCHNL_OP_DISABLE_QUEUES: + case VIRTCHNL_OP_DISABLE_QUEUES: PMD_DRV_LOG(INFO, "OP_DISABLE_QUEUE received"); i40e_pf_host_process_cmd_disable_queues(vf, msg, msglen, b_op); break; - case I40E_VIRTCHNL_OP_ADD_ETHER_ADDRESS: + case VIRTCHNL_OP_ADD_ETH_ADDR: PMD_DRV_LOG(INFO, "OP_ADD_ETHER_ADDRESS received"); i40e_pf_host_process_cmd_add_ether_address(vf, msg, msglen, b_op); break; - case I40E_VIRTCHNL_OP_DEL_ETHER_ADDRESS: + case VIRTCHNL_OP_DEL_ETH_ADDR: PMD_DRV_LOG(INFO, "OP_DEL_ETHER_ADDRESS received"); i40e_pf_host_process_cmd_del_ether_address(vf, msg, msglen, b_op); break; - case I40E_VIRTCHNL_OP_ADD_VLAN: + case VIRTCHNL_OP_ADD_VLAN: PMD_DRV_LOG(INFO, "OP_ADD_VLAN received"); i40e_pf_host_process_cmd_add_vlan(vf, msg, msglen, b_op); break; - case I40E_VIRTCHNL_OP_DEL_VLAN: + case VIRTCHNL_OP_DEL_VLAN: PMD_DRV_LOG(INFO, "OP_DEL_VLAN received"); i40e_pf_host_process_cmd_del_vlan(vf, msg, msglen, b_op); break; - case I40E_VIRTCHNL_OP_CONFIG_PROMISCUOUS_MODE: + case VIRTCHNL_OP_CONFIG_PROMISCUOUS_MODE: PMD_DRV_LOG(INFO, "OP_CONFIG_PROMISCUOUS_MODE received"); i40e_pf_host_process_cmd_config_promisc_mode(vf, msg, msglen, b_op); break; - case I40E_VIRTCHNL_OP_GET_STATS: + case VIRTCHNL_OP_GET_STATS: PMD_DRV_LOG(INFO, "OP_GET_STATS received"); i40e_pf_host_process_cmd_get_stats(vf, b_op); break; diff --git a/drivers/net/i40e/i40e_pf.h b/drivers/net/i40e/i40e_pf.h index 0961f06..4f93a4d 100644 --- a/drivers/net/i40e/i40e_pf.h +++ b/drivers/net/i40e/i40e_pf.h @@ -35,7 +35,7 @@ #define _I40E_PF_H_ /* VERSION info to exchange between VF and PF host. In case VF works with - * ND kernel driver, it reads I40E_VIRTCHNL_VERSION_MAJOR/MINOR. In + * ND kernel driver, it reads VIRTCHNL_VERSION_MAJOR/MINOR. In * case works with DPDK host, it reads version below. Then VF realize who it * is talking to and use proper language to communicate. * */ @@ -49,45 +49,45 @@ #define I40E_DPDK_OFFSET 0x100 /* DPDK pf driver specific command to VF */ -enum i40e_virtchnl_ops_dpdk { +enum virtchnl_ops_dpdk { /* * Keep some gap between Linux PF commands and * DPDK PF extended commands. */ - I40E_VIRTCHNL_OP_CFG_VLAN_OFFLOAD = I40E_VIRTCHNL_OP_VERSION + + I40E_VIRTCHNL_OP_CFG_VLAN_OFFLOAD = VIRTCHNL_OP_VERSION + I40E_DPDK_OFFSET, I40E_VIRTCHNL_OP_CFG_VLAN_PVID, - I40E_VIRTCHNL_OP_CONFIG_VSI_QUEUES_EXT, + VIRTCHNL_OP_CONFIG_VSI_QUEUES_EXT, }; /* A structure to support extended info of a receive queue. */ -struct i40e_virtchnl_rxq_ext_info { +struct virtchnl_rxq_ext_info { uint8_t crcstrip; }; /* * A structure to support extended info of queue pairs, an additional field - * is added, comparing to original 'struct i40e_virtchnl_queue_pair_info'. + * is added, comparing to original 'struct virtchnl_queue_pair_info'. */ -struct i40e_virtchnl_queue_pair_ext_info { +struct virtchnl_queue_pair_ext_info { /* vsi_id and queue_id should be identical for both rx and tx queues.*/ - struct i40e_virtchnl_txq_info txq; - struct i40e_virtchnl_rxq_info rxq; - struct i40e_virtchnl_rxq_ext_info rxq_ext; + struct virtchnl_txq_info txq; + struct virtchnl_rxq_info rxq; + struct virtchnl_rxq_ext_info rxq_ext; }; /* * A structure to support extended info of VSI queue pairs, - * 'struct i40e_virtchnl_queue_pair_ext_info' is used, see its original - * of 'struct i40e_virtchnl_queue_pair_info'. + * 'struct virtchnl_queue_pair_ext_info' is used, see its original + * of 'struct virtchnl_queue_pair_info'. */ -struct i40e_virtchnl_vsi_queue_config_ext_info { +struct virtchnl_vsi_queue_config_ext_info { uint16_t vsi_id; uint16_t num_queue_pairs; - struct i40e_virtchnl_queue_pair_ext_info qpair[0]; + struct virtchnl_queue_pair_ext_info qpair[0]; }; -struct i40e_virtchnl_vlan_offload_info { +struct virtchnl_vlan_offload_info { uint16_t vsi_id; uint8_t enable_vlan_strip; uint8_t reserved; @@ -106,7 +106,7 @@ struct i40e_virtchnl_vlan_offload_info { * enable op, needs to specify the pvid. PF returns status * code in retval. */ -struct i40e_virtchnl_pvid_info { +struct virtchnl_pvid_info { uint16_t vsi_id; struct i40e_vsi_vlan_pvid_info info; }; diff --git a/drivers/net/i40e/rte_pmd_i40e.h b/drivers/net/i40e/rte_pmd_i40e.h index 1efb2c4..cbe742e 100644 --- a/drivers/net/i40e/rte_pmd_i40e.h +++ b/drivers/net/i40e/rte_pmd_i40e.h @@ -59,7 +59,7 @@ enum rte_pmd_i40e_mb_event_rsp { */ struct rte_pmd_i40e_mb_event_param { uint16_t vfid; /**< Virtual Function number */ - uint16_t msg_type; /**< VF to PF message type, see i40e_virtchnl_ops */ + uint16_t msg_type; /**< VF to PF message type, see virtchnl_ops */ uint16_t retval; /**< return value */ void *msg; /**< pointer to message */ uint16_t msglen; /**< length of the message */ -- 2.4.11 ^ permalink raw reply [flat|nested] 69+ messages in thread
* [dpdk-dev] [PATCH v2 02/16] net/i40e/base: sync nvmupdate command and adminq subtask 2017-06-26 9:48 ` [dpdk-dev] [PATCH v2 00/16] " Jingjing Wu 2017-06-26 9:48 ` [dpdk-dev] [PATCH v2 01/16] net/i40e/base: use new virtchnl header file Jingjing Wu @ 2017-06-26 9:48 ` Jingjing Wu 2017-06-26 9:48 ` [dpdk-dev] [PATCH v2 03/16] net/i40e/base: add AQ command for read/write PHY registers Jingjing Wu ` (14 subsequent siblings) 16 siblings, 0 replies; 69+ messages in thread From: Jingjing Wu @ 2017-06-26 9:48 UTC (permalink / raw) To: helin.zhang; +Cc: dev, jingjing.wu During NVMupdate, state machine gets into unrecoverable state because i40e_clean_adminq_subtask can get scheduled after the admin queue command but before other state variables are updated. This patch adds locking around admin queue command and update of state variables so that adminq_subtask will have accurate information whenever it gets scheduled. Signed-off-by: Jingjing Wu <jingjing.wu@intel.com> --- drivers/net/i40e/base/i40e_nvm.c | 6 ++++++ 1 file changed, 6 insertions(+) diff --git a/drivers/net/i40e/base/i40e_nvm.c b/drivers/net/i40e/base/i40e_nvm.c index e896502..90521fa 100644 --- a/drivers/net/i40e/base/i40e_nvm.c +++ b/drivers/net/i40e/base/i40e_nvm.c @@ -899,6 +899,11 @@ enum i40e_status_code i40e_nvmupd_command(struct i40e_hw *hw, hw->nvmupd_state = I40E_NVMUPD_STATE_INIT; } + /* Acquire lock to prevent race condition where adminq_task + * can execute after i40e_nvmupd_nvm_read/write but before state + * variables (nvm_wait_opcode, nvm_release_on_done) are updated + */ + i40e_acquire_spinlock(&hw->aq.arq_spinlock); switch (hw->nvmupd_state) { case I40E_NVMUPD_STATE_INIT: status = i40e_nvmupd_state_init(hw, cmd, bytes, perrno); @@ -934,6 +939,7 @@ enum i40e_status_code i40e_nvmupd_command(struct i40e_hw *hw, *perrno = -ESRCH; break; } + i40e_release_spinlock(&hw->aq.arq_spinlock); return status; } -- 2.4.11 ^ permalink raw reply [flat|nested] 69+ messages in thread
* [dpdk-dev] [PATCH v2 03/16] net/i40e/base: add AQ command for read/write PHY registers 2017-06-26 9:48 ` [dpdk-dev] [PATCH v2 00/16] " Jingjing Wu 2017-06-26 9:48 ` [dpdk-dev] [PATCH v2 01/16] net/i40e/base: use new virtchnl header file Jingjing Wu 2017-06-26 9:48 ` [dpdk-dev] [PATCH v2 02/16] net/i40e/base: sync nvmupdate command and adminq subtask Jingjing Wu @ 2017-06-26 9:48 ` Jingjing Wu 2017-06-26 9:48 ` [dpdk-dev] [PATCH v2 04/16] net/i40e/base: add support for Adaptive Virtual Function Jingjing Wu ` (13 subsequent siblings) 16 siblings, 0 replies; 69+ messages in thread From: Jingjing Wu @ 2017-06-26 9:48 UTC (permalink / raw) To: helin.zhang; +Cc: dev, jingjing.wu This patch adds new additional command for accessing to PHY registers. Signed-off-by: Jingjing Wu <jingjing.wu@intel.com> --- drivers/net/i40e/base/i40e_adminq_cmd.h | 18 +++++++++ drivers/net/i40e/base/i40e_common.c | 70 +++++++++++++++++++++++++++++++++ drivers/net/i40e/base/i40e_prototype.h | 9 +++++ 3 files changed, 97 insertions(+) diff --git a/drivers/net/i40e/base/i40e_adminq_cmd.h b/drivers/net/i40e/base/i40e_adminq_cmd.h index 09f5bf5..83b28f8 100644 --- a/drivers/net/i40e/base/i40e_adminq_cmd.h +++ b/drivers/net/i40e/base/i40e_adminq_cmd.h @@ -245,6 +245,8 @@ enum i40e_admin_queue_opc { i40e_aqc_opc_set_phy_debug = 0x0622, i40e_aqc_opc_upload_ext_phy_fm = 0x0625, i40e_aqc_opc_run_phy_activity = 0x0626, + i40e_aqc_opc_set_phy_register = 0x0628, + i40e_aqc_opc_get_phy_register = 0x0629, /* NVM commands */ i40e_aqc_opc_nvm_read = 0x0701, @@ -2128,6 +2130,22 @@ struct i40e_aqc_run_phy_activity { I40E_CHECK_CMD_LENGTH(i40e_aqc_run_phy_activity); +/* Set PHY Register command (0x0628) */ +/* Get PHY Register command (0x0629) */ +struct i40e_aqc_phy_register_access { + u8 phy_interface; +#define I40E_AQ_PHY_REG_ACCESS_INTERNAL 0 +#define I40E_AQ_PHY_REG_ACCESS_EXTERNAL 1 +#define I40E_AQ_PHY_REG_ACCESS_EXTERNAL_MODULE 2 + u8 dev_addres; + u8 reserved1[2]; + u32 reg_address; + u32 reg_value; + u8 reserved2[4]; +}; + +I40E_CHECK_CMD_LENGTH(i40e_aqc_phy_register_access); + /* NVM Read command (indirect 0x0701) * NVM Erase commands (direct 0x0702) * NVM Update commands (indirect 0x0703) diff --git a/drivers/net/i40e/base/i40e_common.c b/drivers/net/i40e/base/i40e_common.c index 6c09c27..280d9da 100644 --- a/drivers/net/i40e/base/i40e_common.c +++ b/drivers/net/i40e/base/i40e_common.c @@ -6863,6 +6863,76 @@ void i40e_write_rx_ctl(struct i40e_hw *hw, u32 reg_addr, u32 reg_val) if (status || use_register) wr32(hw, reg_addr, reg_val); } + +/** + * i40e_aq_set_phy_register + * @hw: pointer to the hw struct + * @phy_select: select which phy should be accessed + * @dev_addr: PHY device address + * @reg_addr: PHY register address + * @reg_val: new register value + * @cmd_details: pointer to command details structure or NULL + * + * Write the external PHY register. + **/ +enum i40e_status_code i40e_aq_set_phy_register(struct i40e_hw *hw, + u8 phy_select, u8 dev_addr, + u32 reg_addr, u32 reg_val, + struct i40e_asq_cmd_details *cmd_details) +{ + struct i40e_aq_desc desc; + struct i40e_aqc_phy_register_access *cmd = + (struct i40e_aqc_phy_register_access *)&desc.params.raw; + enum i40e_status_code status; + + i40e_fill_default_direct_cmd_desc(&desc, + i40e_aqc_opc_set_phy_register); + + cmd->phy_interface = phy_select; + cmd->dev_addres = dev_addr; + cmd->reg_address = reg_addr; + cmd->reg_value = reg_val; + + status = i40e_asq_send_command(hw, &desc, NULL, 0, cmd_details); + + return status; +} + +/** + * i40e_aq_get_phy_register + * @hw: pointer to the hw struct + * @phy_select: select which phy should be accessed + * @dev_addr: PHY device address + * @reg_addr: PHY register address + * @reg_val: read register value + * @cmd_details: pointer to command details structure or NULL + * + * Read the external PHY register. + **/ +enum i40e_status_code i40e_aq_get_phy_register(struct i40e_hw *hw, + u8 phy_select, u8 dev_addr, + u32 reg_addr, u32 *reg_val, + struct i40e_asq_cmd_details *cmd_details) +{ + struct i40e_aq_desc desc; + struct i40e_aqc_phy_register_access *cmd = + (struct i40e_aqc_phy_register_access *)&desc.params.raw; + enum i40e_status_code status; + + i40e_fill_default_direct_cmd_desc(&desc, + i40e_aqc_opc_get_phy_register); + + cmd->phy_interface = phy_select; + cmd->dev_addres = dev_addr; + cmd->reg_address = reg_addr; + + status = i40e_asq_send_command(hw, &desc, NULL, 0, cmd_details); + if (!status) + *reg_val = cmd->reg_value; + + return status; +} + #ifdef VF_DRIVER /** diff --git a/drivers/net/i40e/base/i40e_prototype.h b/drivers/net/i40e/base/i40e_prototype.h index 6ec4304..54b6750 100644 --- a/drivers/net/i40e/base/i40e_prototype.h +++ b/drivers/net/i40e/base/i40e_prototype.h @@ -533,6 +533,15 @@ enum i40e_status_code i40e_aq_rx_ctl_write_register(struct i40e_hw *hw, u32 reg_addr, u32 reg_val, struct i40e_asq_cmd_details *cmd_details); void i40e_write_rx_ctl(struct i40e_hw *hw, u32 reg_addr, u32 reg_val); +enum i40e_status_code i40e_aq_set_phy_register(struct i40e_hw *hw, + u8 phy_select, u8 dev_addr, + u32 reg_addr, u32 reg_val, + struct i40e_asq_cmd_details *cmd_details); +enum i40e_status_code i40e_aq_get_phy_register(struct i40e_hw *hw, + u8 phy_select, u8 dev_addr, + u32 reg_addr, u32 *reg_val, + struct i40e_asq_cmd_details *cmd_details); + enum i40e_status_code i40e_aq_set_arp_proxy_config(struct i40e_hw *hw, struct i40e_aqc_arp_proxy_data *proxy_config, struct i40e_asq_cmd_details *cmd_details); -- 2.4.11 ^ permalink raw reply [flat|nested] 69+ messages in thread
* [dpdk-dev] [PATCH v2 04/16] net/i40e/base: add support for Adaptive Virtual Function 2017-06-26 9:48 ` [dpdk-dev] [PATCH v2 00/16] " Jingjing Wu ` (2 preceding siblings ...) 2017-06-26 9:48 ` [dpdk-dev] [PATCH v2 03/16] net/i40e/base: add AQ command for read/write PHY registers Jingjing Wu @ 2017-06-26 9:48 ` Jingjing Wu 2017-06-26 9:48 ` [dpdk-dev] [PATCH v2 05/16] net/i40e/base: store the requested FEC information Jingjing Wu ` (12 subsequent siblings) 16 siblings, 0 replies; 69+ messages in thread From: Jingjing Wu @ 2017-06-26 9:48 UTC (permalink / raw) To: helin.zhang; +Cc: dev, jingjing.wu Add device id define and mac_type assignment needed for Adaptive Virtual Function. Signed-off-by: Jingjing Wu <jingjing.wu@intel.com> --- drivers/net/i40e/base/i40e_common.c | 1 + drivers/net/i40e/base/i40e_devids.h | 1 + 2 files changed, 2 insertions(+) diff --git a/drivers/net/i40e/base/i40e_common.c b/drivers/net/i40e/base/i40e_common.c index 280d9da..5f6dd7b 100644 --- a/drivers/net/i40e/base/i40e_common.c +++ b/drivers/net/i40e/base/i40e_common.c @@ -93,6 +93,7 @@ STATIC enum i40e_status_code i40e_set_mac_type(struct i40e_hw *hw) #if defined(INTEGRATED_VF) || defined(VF_DRIVER) case I40E_DEV_ID_VF: case I40E_DEV_ID_VF_HV: + case I40E_DEV_ID_ADAPTIVE_VF: hw->mac.type = I40E_MAC_VF; break; #endif diff --git a/drivers/net/i40e/base/i40e_devids.h b/drivers/net/i40e/base/i40e_devids.h index 4546689..f4a8784 100644 --- a/drivers/net/i40e/base/i40e_devids.h +++ b/drivers/net/i40e/base/i40e_devids.h @@ -54,6 +54,7 @@ POSSIBILITY OF SUCH DAMAGE. #if defined(INTEGRATED_VF) || defined(VF_DRIVER) || defined(I40E_NDIS_SUPPORT) #define I40E_DEV_ID_VF 0x154C #define I40E_DEV_ID_VF_HV 0x1571 +#define I40E_DEV_ID_ADAPTIVE_VF 0x1889 #endif /* VF_DRIVER */ #ifdef X722_A0_SUPPORT #define I40E_DEV_ID_X722_A0 0x374C -- 2.4.11 ^ permalink raw reply [flat|nested] 69+ messages in thread
* [dpdk-dev] [PATCH v2 05/16] net/i40e/base: store the requested FEC information 2017-06-26 9:48 ` [dpdk-dev] [PATCH v2 00/16] " Jingjing Wu ` (3 preceding siblings ...) 2017-06-26 9:48 ` [dpdk-dev] [PATCH v2 04/16] net/i40e/base: add support for Adaptive Virtual Function Jingjing Wu @ 2017-06-26 9:48 ` Jingjing Wu 2017-06-26 9:48 ` [dpdk-dev] [PATCH v2 06/16] net/i40e/base: add new phy types for 25G AOC and ACC Jingjing Wu ` (11 subsequent siblings) 16 siblings, 0 replies; 69+ messages in thread From: Jingjing Wu @ 2017-06-26 9:48 UTC (permalink / raw) To: helin.zhang; +Cc: dev, jingjing.wu Store information about FEC modes, that were requested. It will be used in printing link status information function and this way there is no need to call admin queue there. Signed-off-by: Jingjing Wu <jingjing.wu@intel.com> --- drivers/net/i40e/base/i40e_common.c | 4 ++++ drivers/net/i40e/base/i40e_type.h | 1 + 2 files changed, 5 insertions(+) diff --git a/drivers/net/i40e/base/i40e_common.c b/drivers/net/i40e/base/i40e_common.c index 5f6dd7b..7d59b59 100644 --- a/drivers/net/i40e/base/i40e_common.c +++ b/drivers/net/i40e/base/i40e_common.c @@ -2826,6 +2826,10 @@ enum i40e_status_code i40e_update_link_info(struct i40e_hw *hw) if (status) return status; + hw->phy.link_info.req_fec_info = + abilities.fec_cfg_curr_mod_ext_info & + (I40E_AQ_REQUEST_FEC_KR | I40E_AQ_REQUEST_FEC_RS); + i40e_memcpy(hw->phy.link_info.module_type, &abilities.module_type, sizeof(hw->phy.link_info.module_type), I40E_NONDMA_TO_NONDMA); } diff --git a/drivers/net/i40e/base/i40e_type.h b/drivers/net/i40e/base/i40e_type.h index 52a114a..af5347b 100644 --- a/drivers/net/i40e/base/i40e_type.h +++ b/drivers/net/i40e/base/i40e_type.h @@ -271,6 +271,7 @@ struct i40e_link_status { enum i40e_aq_link_speed link_speed; u8 link_info; u8 an_info; + u8 req_fec_info; u8 fec_info; u8 ext_info; u8 loopback; -- 2.4.11 ^ permalink raw reply [flat|nested] 69+ messages in thread
* [dpdk-dev] [PATCH v2 06/16] net/i40e/base: add new phy types for 25G AOC and ACC 2017-06-26 9:48 ` [dpdk-dev] [PATCH v2 00/16] " Jingjing Wu ` (4 preceding siblings ...) 2017-06-26 9:48 ` [dpdk-dev] [PATCH v2 05/16] net/i40e/base: store the requested FEC information Jingjing Wu @ 2017-06-26 9:48 ` Jingjing Wu 2017-06-26 9:48 ` [dpdk-dev] [PATCH v2 07/16] net/i40e/base: report supported link modes Jingjing Wu ` (10 subsequent siblings) 16 siblings, 0 replies; 69+ messages in thread From: Jingjing Wu @ 2017-06-26 9:48 UTC (permalink / raw) To: helin.zhang; +Cc: dev, jingjing.wu This patch adds new phy types for 25G Active Optical Cables (AOC) and Active Copper Cables (ACC) support. Signed-off-by: Jingjing Wu <jingjing.wu@intel.com> --- drivers/net/i40e/base/i40e_adminq_cmd.h | 4 ++++ drivers/net/i40e/base/i40e_common.c | 2 ++ 2 files changed, 6 insertions(+) diff --git a/drivers/net/i40e/base/i40e_adminq_cmd.h b/drivers/net/i40e/base/i40e_adminq_cmd.h index 83b28f8..70079a0 100644 --- a/drivers/net/i40e/base/i40e_adminq_cmd.h +++ b/drivers/net/i40e/base/i40e_adminq_cmd.h @@ -1849,6 +1849,8 @@ enum i40e_aq_phy_type { I40E_PHY_TYPE_25GBASE_CR = 0x20, I40E_PHY_TYPE_25GBASE_SR = 0x21, I40E_PHY_TYPE_25GBASE_LR = 0x22, + I40E_PHY_TYPE_25GBASE_AOC = 0x23, + I40E_PHY_TYPE_25GBASE_ACC = 0x24, I40E_PHY_TYPE_MAX }; @@ -1906,6 +1908,8 @@ struct i40e_aq_get_phy_abilities_resp { #define I40E_AQ_PHY_TYPE_EXT_25G_CR 0x02 #define I40E_AQ_PHY_TYPE_EXT_25G_SR 0x04 #define I40E_AQ_PHY_TYPE_EXT_25G_LR 0x08 +#define I40E_AQ_PHY_TYPE_EXT_25G_AOC 0x10 +#define I40E_AQ_PHY_TYPE_EXT_25G_ACC 0x20 u8 fec_cfg_curr_mod_ext_info; #define I40E_AQ_ENABLE_FEC_KR 0x01 #define I40E_AQ_ENABLE_FEC_RS 0x02 diff --git a/drivers/net/i40e/base/i40e_common.c b/drivers/net/i40e/base/i40e_common.c index 7d59b59..004c062 100644 --- a/drivers/net/i40e/base/i40e_common.c +++ b/drivers/net/i40e/base/i40e_common.c @@ -1296,6 +1296,8 @@ STATIC enum i40e_media_type i40e_get_media_type(struct i40e_hw *hw) case I40E_PHY_TYPE_40GBASE_AOC: case I40E_PHY_TYPE_10GBASE_AOC: case I40E_PHY_TYPE_25GBASE_CR: + case I40E_PHY_TYPE_25GBASE_AOC: + case I40E_PHY_TYPE_25GBASE_ACC: media = I40E_MEDIA_TYPE_DA; break; case I40E_PHY_TYPE_1000BASE_KX: -- 2.4.11 ^ permalink raw reply [flat|nested] 69+ messages in thread
* [dpdk-dev] [PATCH v2 07/16] net/i40e/base: report supported link modes 2017-06-26 9:48 ` [dpdk-dev] [PATCH v2 00/16] " Jingjing Wu ` (5 preceding siblings ...) 2017-06-26 9:48 ` [dpdk-dev] [PATCH v2 06/16] net/i40e/base: add new phy types for 25G AOC and ACC Jingjing Wu @ 2017-06-26 9:48 ` Jingjing Wu 2017-06-26 9:48 ` [dpdk-dev] [PATCH v2 08/16] net/i40e/base: track id can be 0 Jingjing Wu ` (9 subsequent siblings) 16 siblings, 0 replies; 69+ messages in thread From: Jingjing Wu @ 2017-06-26 9:48 UTC (permalink / raw) To: helin.zhang; +Cc: dev, jingjing.wu Signed-off-by: Jingjing Wu <jingjing.wu@intel.com> --- drivers/net/i40e/base/i40e_adminq_cmd.h | 22 +++++++++++++++++++--- drivers/net/i40e/base/i40e_common.c | 18 +++++++++++++++--- 2 files changed, 34 insertions(+), 6 deletions(-) diff --git a/drivers/net/i40e/base/i40e_adminq_cmd.h b/drivers/net/i40e/base/i40e_adminq_cmd.h index 70079a0..c1ca290 100644 --- a/drivers/net/i40e/base/i40e_adminq_cmd.h +++ b/drivers/net/i40e/base/i40e_adminq_cmd.h @@ -1831,6 +1831,8 @@ enum i40e_aq_phy_type { I40E_PHY_TYPE_10GBASE_CR1_CU = 0xB, I40E_PHY_TYPE_10GBASE_AOC = 0xC, I40E_PHY_TYPE_40GBASE_AOC = 0xD, + I40E_PHY_TYPE_UNRECOGNIZED = 0xE, + I40E_PHY_TYPE_UNSUPPORTED = 0xF, I40E_PHY_TYPE_100BASE_TX = 0x11, I40E_PHY_TYPE_1000BASE_T = 0x12, I40E_PHY_TYPE_10GBASE_T = 0x13, @@ -1851,7 +1853,9 @@ enum i40e_aq_phy_type { I40E_PHY_TYPE_25GBASE_LR = 0x22, I40E_PHY_TYPE_25GBASE_AOC = 0x23, I40E_PHY_TYPE_25GBASE_ACC = 0x24, - I40E_PHY_TYPE_MAX + I40E_PHY_TYPE_MAX, + I40E_PHY_TYPE_EMPTY = 0xFE, + I40E_PHY_TYPE_DEFAULT = 0xFF, }; #define I40E_LINK_SPEED_100MB_SHIFT 0x1 @@ -2039,19 +2043,31 @@ struct i40e_aqc_get_link_status { #define I40E_AQ_25G_SERDES_UCODE_ERR 0X04 #define I40E_AQ_25G_NIMB_UCODE_ERR 0X05 u8 loopback; /* use defines from i40e_aqc_set_lb_mode */ +/* Since firmware API 1.7 loopback field keeps power class info as well */ +#define I40E_AQ_LOOPBACK_MASK 0x07 +#define I40E_AQ_PWR_CLASS_SHIFT_LB 6 +#define I40E_AQ_PWR_CLASS_MASK_LB (0x03 << I40E_AQ_PWR_CLASS_SHIFT_LB) __le16 max_frame_size; u8 config; #define I40E_AQ_CONFIG_FEC_KR_ENA 0x01 #define I40E_AQ_CONFIG_FEC_RS_ENA 0x02 #define I40E_AQ_CONFIG_CRC_ENA 0x04 #define I40E_AQ_CONFIG_PACING_MASK 0x78 - u8 power_desc; + union { + struct { + u8 power_desc; #define I40E_AQ_LINK_POWER_CLASS_1 0x00 #define I40E_AQ_LINK_POWER_CLASS_2 0x01 #define I40E_AQ_LINK_POWER_CLASS_3 0x02 #define I40E_AQ_LINK_POWER_CLASS_4 0x03 #define I40E_AQ_PWR_CLASS_MASK 0x03 - u8 reserved[4]; + u8 reserved[4]; + }; + struct { + u8 link_type[4]; + u8 link_type_ext; + }; + }; }; I40E_CHECK_CMD_LENGTH(i40e_aqc_get_link_status); diff --git a/drivers/net/i40e/base/i40e_common.c b/drivers/net/i40e/base/i40e_common.c index 004c062..bb4b7eb 100644 --- a/drivers/net/i40e/base/i40e_common.c +++ b/drivers/net/i40e/base/i40e_common.c @@ -1692,8 +1692,14 @@ enum i40e_status_code i40e_aq_get_phy_capabilities(struct i40e_hw *hw, status = I40E_ERR_UNKNOWN_PHY; if (report_init) { - hw->phy.phy_types = LE32_TO_CPU(abilities->phy_type); - hw->phy.phy_types |= ((u64)abilities->phy_type_ext << 32); + if (hw->aq.api_maj_ver == I40E_FW_API_VERSION_MAJOR && + hw->aq.api_min_ver >= 7) { + status = i40e_aq_get_link_info(hw, true, NULL, NULL); + } else { + hw->phy.phy_types = LE32_TO_CPU(abilities->phy_type); + hw->phy.phy_types |= + ((u64)abilities->phy_type_ext << 32); + } } return status; @@ -1955,7 +1961,7 @@ enum i40e_status_code i40e_aq_get_link_info(struct i40e_hw *hw, hw_link_info->fec_info = resp->config & (I40E_AQ_CONFIG_FEC_KR_ENA | I40E_AQ_CONFIG_FEC_RS_ENA); hw_link_info->ext_info = resp->ext_info; - hw_link_info->loopback = resp->loopback; + hw_link_info->loopback = resp->loopback & I40E_AQ_LOOPBACK_MASK; hw_link_info->max_frame_size = LE16_TO_CPU(resp->max_frame_size); hw_link_info->pacing = resp->config & I40E_AQ_CONFIG_PACING_MASK; @@ -1986,6 +1992,12 @@ enum i40e_status_code i40e_aq_get_link_info(struct i40e_hw *hw, hw->aq.fw_min_ver < 40)) && hw_link_info->phy_type == 0xE) hw_link_info->phy_type = I40E_PHY_TYPE_10GBASE_SFPP_CU; + if (hw->aq.api_maj_ver == I40E_FW_API_VERSION_MAJOR && + hw->aq.api_min_ver >= 7) { + hw->phy.phy_types = LE32_TO_CPU(*(__le32 *)resp->link_type); + hw->phy.phy_types |= ((u64)resp->link_type_ext << 32); + } + /* save link status information */ if (link) i40e_memcpy(link, hw_link_info, sizeof(*hw_link_info), -- 2.4.11 ^ permalink raw reply [flat|nested] 69+ messages in thread
* [dpdk-dev] [PATCH v2 08/16] net/i40e/base: track id can be 0 2017-06-26 9:48 ` [dpdk-dev] [PATCH v2 00/16] " Jingjing Wu ` (6 preceding siblings ...) 2017-06-26 9:48 ` [dpdk-dev] [PATCH v2 07/16] net/i40e/base: report supported link modes Jingjing Wu @ 2017-06-26 9:48 ` Jingjing Wu 2017-06-26 9:48 ` [dpdk-dev] [PATCH v2 09/16] net/i40e/base: update FW AQ API version to 1.7 Jingjing Wu ` (8 subsequent siblings) 16 siblings, 0 replies; 69+ messages in thread From: Jingjing Wu @ 2017-06-26 9:48 UTC (permalink / raw) To: helin.zhang; +Cc: dev, jingjing.wu track_id == 0 is valid for “read only” profiles when profile does not have any “write” commands. Signed-off-by: Jingjing Wu <jingjing.wu@intel.com> --- drivers/net/i40e/base/i40e_common.c | 7 +------ drivers/net/i40e/base/i40e_prototype.h | 12 +++++++----- 2 files changed, 8 insertions(+), 11 deletions(-) diff --git a/drivers/net/i40e/base/i40e_common.c b/drivers/net/i40e/base/i40e_common.c index bb4b7eb..38c30a1 100644 --- a/drivers/net/i40e/base/i40e_common.c +++ b/drivers/net/i40e/base/i40e_common.c @@ -7370,11 +7370,6 @@ i40e_write_profile(struct i40e_hw *hw, struct i40e_profile_segment *profile, u32 offset = 0, info = 0; u32 i; - if (!track_id) { - i40e_debug(hw, I40E_DEBUG_PACKAGE, "Track_id can't be 0."); - return I40E_NOT_SUPPORTED; - } - dev_cnt = profile->device_table_count; for (i = 0; i < dev_cnt; i++) { @@ -7449,6 +7444,6 @@ i40e_add_pinfo_to_list(struct i40e_hw *hw, memcpy(pinfo->name, profile->name, I40E_DDP_NAME_SIZE); status = i40e_aq_write_ddp(hw, (void *)sec, sec->data_end, - track_id, &offset, &info, NULL); + track_id, &offset, &info, NULL); return status; } diff --git a/drivers/net/i40e/base/i40e_prototype.h b/drivers/net/i40e/base/i40e_prototype.h index 54b6750..9171e97 100644 --- a/drivers/net/i40e/base/i40e_prototype.h +++ b/drivers/net/i40e/base/i40e_prototype.h @@ -575,12 +575,14 @@ u8 i40e_get_phy_address(struct i40e_hw *hw, u8 dev_num); enum i40e_status_code i40e_blink_phy_link_led(struct i40e_hw *hw, u32 time, u32 interval); enum i40e_status_code i40e_aq_write_ddp(struct i40e_hw *hw, void *buff, - u16 buff_size, u32 track_id, - u32 *error_offset, u32 *error_info, - struct i40e_asq_cmd_details *cmd_details); + u16 buff_size, u32 track_id, + u32 *error_offset, u32 *error_info, + struct i40e_asq_cmd_details * + cmd_details); enum i40e_status_code i40e_aq_get_ddp_list(struct i40e_hw *hw, void *buff, - u16 buff_size, u8 flags, - struct i40e_asq_cmd_details *cmd_details); + u16 buff_size, u8 flags, + struct i40e_asq_cmd_details * + cmd_details); struct i40e_generic_seg_header * i40e_find_segment_in_package(u32 segment_type, struct i40e_package_header *pkg_header); -- 2.4.11 ^ permalink raw reply [flat|nested] 69+ messages in thread
* [dpdk-dev] [PATCH v2 09/16] net/i40e/base: update FW AQ API version to 1.7 2017-06-26 9:48 ` [dpdk-dev] [PATCH v2 00/16] " Jingjing Wu ` (7 preceding siblings ...) 2017-06-26 9:48 ` [dpdk-dev] [PATCH v2 08/16] net/i40e/base: track id can be 0 Jingjing Wu @ 2017-06-26 9:48 ` Jingjing Wu 2017-06-26 9:48 ` [dpdk-dev] [PATCH v2 10/16] net/i40e/base: add support for switch parameters Jingjing Wu ` (7 subsequent siblings) 16 siblings, 0 replies; 69+ messages in thread From: Jingjing Wu @ 2017-06-26 9:48 UTC (permalink / raw) To: helin.zhang; +Cc: dev, jingjing.wu Signed-off-by: Jingjing Wu <jingjing.wu@intel.com> --- drivers/net/i40e/base/i40e_adminq_cmd.h | 10 +++++++++- drivers/net/i40e/base/i40e_common.c | 5 +++-- drivers/net/i40e/base/i40e_type.h | 4 ---- 3 files changed, 12 insertions(+), 7 deletions(-) diff --git a/drivers/net/i40e/base/i40e_adminq_cmd.h b/drivers/net/i40e/base/i40e_adminq_cmd.h index c1ca290..dd9fb55 100644 --- a/drivers/net/i40e/base/i40e_adminq_cmd.h +++ b/drivers/net/i40e/base/i40e_adminq_cmd.h @@ -41,7 +41,15 @@ POSSIBILITY OF SUCH DAMAGE. */ #define I40E_FW_API_VERSION_MAJOR 0x0001 -#define I40E_FW_API_VERSION_MINOR 0x0005 +#define I40E_FW_API_VERSION_MINOR_X722 0x0005 +#define I40E_FW_API_VERSION_MINOR_X710 0x0007 + +#define I40E_FW_MINOR_VERSION(_h) ((_h)->mac.type == I40E_MAC_XL710 ? \ + I40E_FW_API_VERSION_MINOR_X710 : \ + I40E_FW_API_VERSION_MINOR_X722) + +/* API version 1.7 implements additional link and PHY-specific APIs */ +#define I40E_MINOR_VER_GET_LINK_INFO_XL710 0x0007 struct i40e_aq_desc { __le16 flags; diff --git a/drivers/net/i40e/base/i40e_common.c b/drivers/net/i40e/base/i40e_common.c index 38c30a1..9895a77 100644 --- a/drivers/net/i40e/base/i40e_common.c +++ b/drivers/net/i40e/base/i40e_common.c @@ -1692,8 +1692,9 @@ enum i40e_status_code i40e_aq_get_phy_capabilities(struct i40e_hw *hw, status = I40E_ERR_UNKNOWN_PHY; if (report_init) { - if (hw->aq.api_maj_ver == I40E_FW_API_VERSION_MAJOR && - hw->aq.api_min_ver >= 7) { + if (hw->mac.type == I40E_MAC_XL710 && + hw->aq.api_maj_ver == I40E_FW_API_VERSION_MAJOR && + hw->aq.api_min_ver >= I40E_MINOR_VER_GET_LINK_INFO_XL710) { status = i40e_aq_get_link_info(hw, true, NULL, NULL); } else { hw->phy.phy_types = LE32_TO_CPU(abilities->phy_type); diff --git a/drivers/net/i40e/base/i40e_type.h b/drivers/net/i40e/base/i40e_type.h index af5347b..6ca1a0f 100644 --- a/drivers/net/i40e/base/i40e_type.h +++ b/drivers/net/i40e/base/i40e_type.h @@ -199,10 +199,6 @@ enum i40e_memcpy_type { I40E_DMA_TO_NONDMA }; -#define I40E_FW_API_VERSION_MINOR_X722 0x0005 -#define I40E_FW_API_VERSION_MINOR_X710 0x0005 - - /* These are structs for managing the hardware information and the operations. * The structures of function pointers are filled out at init time when we * know for sure exactly which hardware we're working with. This gives us the -- 2.4.11 ^ permalink raw reply [flat|nested] 69+ messages in thread
* [dpdk-dev] [PATCH v2 10/16] net/i40e/base: add support for switch parameters 2017-06-26 9:48 ` [dpdk-dev] [PATCH v2 00/16] " Jingjing Wu ` (8 preceding siblings ...) 2017-06-26 9:48 ` [dpdk-dev] [PATCH v2 09/16] net/i40e/base: update FW AQ API version to 1.7 Jingjing Wu @ 2017-06-26 9:48 ` Jingjing Wu 2017-06-26 9:48 ` [dpdk-dev] [PATCH v2 11/16] net/i40e/base: use admin queue for setting LEDs behavior Jingjing Wu ` (6 subsequent siblings) 16 siblings, 0 replies; 69+ messages in thread From: Jingjing Wu @ 2017-06-26 9:48 UTC (permalink / raw) To: helin.zhang; +Cc: dev, jingjing.wu Adds double VLAN tagging ethertype fields to Set Switch Parameters AQ command. These were added in firmware API 1.7. Callers of i40e_aq_set_switch_config() can specify the ethertypes to use by filling out the corresponding fields in struct i40e_hw. Signed-off-by: Jingjing Wu <jingjing.wu@intel.com> --- drivers/net/i40e/base/i40e_adminq.c | 6 ++++++ drivers/net/i40e/base/i40e_adminq_cmd.h | 17 ++++++++++++++++- drivers/net/i40e/base/i40e_common.c | 6 +++++- drivers/net/i40e/base/i40e_type.h | 6 ++++++ 4 files changed, 33 insertions(+), 2 deletions(-) diff --git a/drivers/net/i40e/base/i40e_adminq.c b/drivers/net/i40e/base/i40e_adminq.c index a60292a..19871c7 100644 --- a/drivers/net/i40e/base/i40e_adminq.c +++ b/drivers/net/i40e/base/i40e_adminq.c @@ -682,6 +682,12 @@ enum i40e_status_code i40e_init_adminq(struct i40e_hw *hw) &oem_lo); hw->nvm.oem_ver = ((u32)oem_hi << 16) | oem_lo; + /* The ability to RX (not drop) 802.1ad frames was added in API 1.7 */ + if ((hw->aq.api_maj_ver > 1) || + ((hw->aq.api_maj_ver == 1) && + (hw->aq.api_min_ver >= 7))) + hw->flags |= I40E_HW_FLAG_802_1AD_CAPABLE; + if (hw->aq.api_maj_ver > I40E_FW_API_VERSION_MAJOR) { ret_code = I40E_ERR_FIRMWARE_API_VERSION; goto init_adminq_free_arq; diff --git a/drivers/net/i40e/base/i40e_adminq_cmd.h b/drivers/net/i40e/base/i40e_adminq_cmd.h index dd9fb55..c36da2a 100644 --- a/drivers/net/i40e/base/i40e_adminq_cmd.h +++ b/drivers/net/i40e/base/i40e_adminq_cmd.h @@ -787,7 +787,22 @@ struct i40e_aqc_set_switch_config { #define I40E_AQ_SET_SWITCH_CFG_L2_FILTER 0x0002 #define I40E_AQ_SET_SWITCH_CFG_HW_ATR_EVICT 0x0004 __le16 valid_flags; - u8 reserved[12]; + /* The ethertype in switch_tag is dropped on ingress and used + * internally by the switch. Set this to zero for the default + * of 0x88a8 (802.1ad). Should be zero for firmware API + * versions lower than 1.7. + */ + __le16 switch_tag; + /* The ethertypes in first_tag and second_tag are used to + * match the outer and inner VLAN tags (respectively) when HW + * double VLAN tagging is enabled via the set port parameters + * AQ command. Otherwise these are both ignored. Set them to + * zero for their defaults of 0x8100 (802.1Q). Should be zero + * for firmware API versions lower than 1.7. + */ + __le16 first_tag; + __le16 second_tag; + u8 reserved[6]; }; I40E_CHECK_CMD_LENGTH(i40e_aqc_set_switch_config); diff --git a/drivers/net/i40e/base/i40e_common.c b/drivers/net/i40e/base/i40e_common.c index 9895a77..b4901ef 100644 --- a/drivers/net/i40e/base/i40e_common.c +++ b/drivers/net/i40e/base/i40e_common.c @@ -2695,7 +2695,11 @@ enum i40e_status_code i40e_aq_set_switch_config(struct i40e_hw *hw, i40e_aqc_opc_set_switch_config); scfg->flags = CPU_TO_LE16(flags); scfg->valid_flags = CPU_TO_LE16(valid_flags); - + if (hw->flags & I40E_HW_FLAG_802_1AD_CAPABLE) { + scfg->switch_tag = CPU_TO_LE16(hw->switch_tag); + scfg->first_tag = CPU_TO_LE16(hw->first_tag); + scfg->second_tag = CPU_TO_LE16(hw->second_tag); + } status = i40e_asq_send_command(hw, &desc, NULL, 0, cmd_details); return status; diff --git a/drivers/net/i40e/base/i40e_type.h b/drivers/net/i40e/base/i40e_type.h index 6ca1a0f..152b4a7 100644 --- a/drivers/net/i40e/base/i40e_type.h +++ b/drivers/net/i40e/base/i40e_type.h @@ -700,8 +700,14 @@ struct i40e_hw { u16 wol_proxy_vsi_seid; #define I40E_HW_FLAG_AQ_SRCTL_ACCESS_ENABLE BIT_ULL(0) +#define I40E_HW_FLAG_802_1AD_CAPABLE BIT_ULL(1) u64 flags; + /* Used in set switch config AQ command */ + u16 switch_tag; + u16 first_tag; + u16 second_tag; + /* debug mask */ u32 debug_mask; char err_str[16]; -- 2.4.11 ^ permalink raw reply [flat|nested] 69+ messages in thread
* [dpdk-dev] [PATCH v2 11/16] net/i40e/base: use admin queue for setting LEDs behavior 2017-06-26 9:48 ` [dpdk-dev] [PATCH v2 00/16] " Jingjing Wu ` (9 preceding siblings ...) 2017-06-26 9:48 ` [dpdk-dev] [PATCH v2 10/16] net/i40e/base: add support for switch parameters Jingjing Wu @ 2017-06-26 9:48 ` Jingjing Wu 2017-06-26 9:48 ` [dpdk-dev] [PATCH v2 12/16] net/i40e/base: avoid potential null pointer dereference Jingjing Wu ` (5 subsequent siblings) 16 siblings, 0 replies; 69+ messages in thread From: Jingjing Wu @ 2017-06-26 9:48 UTC (permalink / raw) To: helin.zhang; +Cc: dev, jingjing.wu Instead of accessing register directly, use newly added AQC in order to blink LEDs. Introduce and utilize a new flag to prevent excessive API version checking. Signed-off-by: Jingjing Wu <jingjing.wu@intel.com> --- drivers/net/i40e/base/i40e_adminq.c | 6 ++ drivers/net/i40e/base/i40e_common.c | 148 ++++++++++++++++++++++++++-------- drivers/net/i40e/base/i40e_register.h | 2 +- drivers/net/i40e/base/i40e_type.h | 1 + 4 files changed, 121 insertions(+), 36 deletions(-) diff --git a/drivers/net/i40e/base/i40e_adminq.c b/drivers/net/i40e/base/i40e_adminq.c index 19871c7..8cc8c5e 100644 --- a/drivers/net/i40e/base/i40e_adminq.c +++ b/drivers/net/i40e/base/i40e_adminq.c @@ -688,6 +688,12 @@ enum i40e_status_code i40e_init_adminq(struct i40e_hw *hw) (hw->aq.api_min_ver >= 7))) hw->flags |= I40E_HW_FLAG_802_1AD_CAPABLE; + if (hw->mac.type == I40E_MAC_XL710 && + hw->aq.api_maj_ver == I40E_FW_API_VERSION_MAJOR && + hw->aq.api_min_ver >= I40E_MINOR_VER_GET_LINK_INFO_XL710) { + hw->flags |= I40E_HW_FLAG_AQ_PHY_ACCESS_CAPABLE; + } + if (hw->aq.api_maj_ver > I40E_FW_API_VERSION_MAJOR) { ret_code = I40E_ERR_FIRMWARE_API_VERSION; goto init_adminq_free_arq; diff --git a/drivers/net/i40e/base/i40e_common.c b/drivers/net/i40e/base/i40e_common.c index b4901ef..3f3b275 100644 --- a/drivers/net/i40e/base/i40e_common.c +++ b/drivers/net/i40e/base/i40e_common.c @@ -6678,24 +6678,38 @@ enum i40e_status_code i40e_led_get_phy(struct i40e_hw *hw, u16 *led_addr, u16 temp_addr; u8 port_num; u32 i; - - temp_addr = I40E_PHY_LED_PROV_REG_1; - i = rd32(hw, I40E_PFGEN_PORTNUM); - port_num = (u8)(i & I40E_PFGEN_PORTNUM_PORT_NUM_MASK); - phy_addr = i40e_get_phy_address(hw, port_num); - - for (gpio_led_port = 0; gpio_led_port < 3; gpio_led_port++, - temp_addr++) { - status = i40e_read_phy_register_clause45(hw, + u32 reg_val_aq; + + if (hw->flags & I40E_HW_FLAG_AQ_PHY_ACCESS_CAPABLE) { + status = + i40e_aq_get_phy_register(hw, + I40E_AQ_PHY_REG_ACCESS_EXTERNAL, + I40E_PHY_COM_REG_PAGE, + I40E_PHY_LED_PROV_REG_1, + ®_val_aq, NULL); + if (status) + return status; + *val = (u16)reg_val_aq; + } else { + temp_addr = I40E_PHY_LED_PROV_REG_1; + i = rd32(hw, I40E_PFGEN_PORTNUM); + port_num = (u8)(i & I40E_PFGEN_PORTNUM_PORT_NUM_MASK); + phy_addr = i40e_get_phy_address(hw, port_num); + + for (gpio_led_port = 0; gpio_led_port < 3; gpio_led_port++, + temp_addr++) { + status = + i40e_read_phy_register_clause45(hw, I40E_PHY_COM_REG_PAGE, temp_addr, phy_addr, ®_val); - if (status) - return status; - *val = reg_val; - if (reg_val & I40E_PHY_LED_LINK_MODE_MASK) { - *led_addr = temp_addr; - break; + if (status) + return status; + *val = reg_val; + if (reg_val & I40E_PHY_LED_LINK_MODE_MASK) { + *led_addr = temp_addr; + break; + } } } return status; @@ -6713,51 +6727,115 @@ enum i40e_status_code i40e_led_set_phy(struct i40e_hw *hw, bool on, u16 led_addr, u32 mode) { enum i40e_status_code status = I40E_SUCCESS; - u16 led_ctl = 0; - u16 led_reg = 0; + u32 led_ctl = 0; + u32 led_reg = 0; u8 phy_addr = 0; u8 port_num; u32 i; - i = rd32(hw, I40E_PFGEN_PORTNUM); - port_num = (u8)(i & I40E_PFGEN_PORTNUM_PORT_NUM_MASK); - phy_addr = i40e_get_phy_address(hw, port_num); - status = i40e_read_phy_register_clause45(hw, I40E_PHY_COM_REG_PAGE, - led_addr, phy_addr, &led_reg); + if (hw->flags & I40E_HW_FLAG_AQ_PHY_ACCESS_CAPABLE) { + status = + i40e_aq_get_phy_register(hw, + I40E_AQ_PHY_REG_ACCESS_EXTERNAL, + I40E_PHY_COM_REG_PAGE, + I40E_PHY_LED_PROV_REG_1, + &led_reg, NULL); + } else { + i = rd32(hw, I40E_PFGEN_PORTNUM); + port_num = (u8)(i & I40E_PFGEN_PORTNUM_PORT_NUM_MASK); + phy_addr = i40e_get_phy_address(hw, port_num); + status = i40e_read_phy_register_clause45(hw, + I40E_PHY_COM_REG_PAGE, + led_addr, phy_addr, + (u16 *)&led_reg); + } if (status) return status; led_ctl = led_reg; if (led_reg & I40E_PHY_LED_LINK_MODE_MASK) { led_reg = 0; - status = i40e_write_phy_register_clause45(hw, - I40E_PHY_COM_REG_PAGE, - led_addr, phy_addr, - led_reg); + if (hw->aq.api_maj_ver == I40E_FW_API_VERSION_MAJOR && + hw->aq.api_min_ver >= I40E_MINOR_VER_GET_LINK_INFO_XL710) { + status = i40e_aq_set_phy_register(hw, + I40E_AQ_PHY_REG_ACCESS_EXTERNAL, + I40E_PHY_COM_REG_PAGE, + I40E_PHY_LED_PROV_REG_1, + led_reg, NULL); + } else { + status = i40e_write_phy_register_clause45(hw, + I40E_PHY_COM_REG_PAGE, + led_addr, phy_addr, + (u16)led_reg); + } if (status) return status; } - status = i40e_read_phy_register_clause45(hw, I40E_PHY_COM_REG_PAGE, - led_addr, phy_addr, &led_reg); + if (hw->flags & I40E_HW_FLAG_AQ_PHY_ACCESS_CAPABLE) { + status = + i40e_aq_get_phy_register(hw, + I40E_AQ_PHY_REG_ACCESS_EXTERNAL, + I40E_PHY_COM_REG_PAGE, + I40E_PHY_LED_PROV_REG_1, + &led_reg, NULL); + } else { + status = i40e_read_phy_register_clause45(hw, + I40E_PHY_COM_REG_PAGE, + led_addr, phy_addr, + (u16 *)&led_reg); + } if (status) goto restore_config; if (on) led_reg = I40E_PHY_LED_MANUAL_ON; else led_reg = 0; - status = i40e_write_phy_register_clause45(hw, I40E_PHY_COM_REG_PAGE, - led_addr, phy_addr, led_reg); + + if (hw->flags & I40E_HW_FLAG_AQ_PHY_ACCESS_CAPABLE) { + status = + i40e_aq_set_phy_register(hw, + I40E_AQ_PHY_REG_ACCESS_EXTERNAL, + I40E_PHY_COM_REG_PAGE, + I40E_PHY_LED_PROV_REG_1, + led_reg, NULL); + } else { + status = + i40e_write_phy_register_clause45(hw, I40E_PHY_COM_REG_PAGE, + led_addr, phy_addr, + (u16)led_reg); + } if (status) goto restore_config; if (mode & I40E_PHY_LED_MODE_ORIG) { led_ctl = (mode & I40E_PHY_LED_MODE_MASK); - status = i40e_write_phy_register_clause45(hw, - I40E_PHY_COM_REG_PAGE, - led_addr, phy_addr, led_ctl); + if (hw->flags & I40E_HW_FLAG_AQ_PHY_ACCESS_CAPABLE) { + status = i40e_aq_set_phy_register(hw, + I40E_AQ_PHY_REG_ACCESS_EXTERNAL, + I40E_PHY_COM_REG_PAGE, + I40E_PHY_LED_PROV_REG_1, + led_ctl, NULL); + } else { + status = i40e_write_phy_register_clause45(hw, + I40E_PHY_COM_REG_PAGE, + led_addr, phy_addr, + (u16)led_ctl); + } } return status; restore_config: - status = i40e_write_phy_register_clause45(hw, I40E_PHY_COM_REG_PAGE, - led_addr, phy_addr, led_ctl); + if (hw->flags & I40E_HW_FLAG_AQ_PHY_ACCESS_CAPABLE) { + status = + i40e_aq_set_phy_register(hw, + I40E_AQ_PHY_REG_ACCESS_EXTERNAL, + I40E_PHY_COM_REG_PAGE, + I40E_PHY_LED_PROV_REG_1, + led_ctl, NULL); + } else { + status = + i40e_write_phy_register_clause45(hw, + I40E_PHY_COM_REG_PAGE, + led_addr, phy_addr, + (u16)led_ctl); + } return status; } #endif /* PF_DRIVER */ diff --git a/drivers/net/i40e/base/i40e_register.h b/drivers/net/i40e/base/i40e_register.h index b150fbd..a482ab9 100644 --- a/drivers/net/i40e/base/i40e_register.h +++ b/drivers/net/i40e/base/i40e_register.h @@ -2805,7 +2805,7 @@ POSSIBILITY OF SUCH DAMAGE. #define I40E_GLV_RUPP_MAX_INDEX 383 #define I40E_GLV_RUPP_RUPP_SHIFT 0 #define I40E_GLV_RUPP_RUPP_MASK I40E_MASK(0xFFFFFFFF, I40E_GLV_RUPP_RUPP_SHIFT) -#define I40E_GLV_TEPC(_VSI) (0x00344000 + ((_VSI) * 8)) /* _i=0...383 */ /* Reset: CORER */ +#define I40E_GLV_TEPC(_i) (0x00344000 + ((_i) * 8)) /* _i=0...383 */ /* Reset: CORER */ #define I40E_GLV_TEPC_MAX_INDEX 383 #define I40E_GLV_TEPC_TEPC_SHIFT 0 #define I40E_GLV_TEPC_TEPC_MASK I40E_MASK(0xFFFFFFFF, I40E_GLV_TEPC_TEPC_SHIFT) diff --git a/drivers/net/i40e/base/i40e_type.h b/drivers/net/i40e/base/i40e_type.h index 152b4a7..9d7b1a2 100644 --- a/drivers/net/i40e/base/i40e_type.h +++ b/drivers/net/i40e/base/i40e_type.h @@ -701,6 +701,7 @@ struct i40e_hw { #define I40E_HW_FLAG_AQ_SRCTL_ACCESS_ENABLE BIT_ULL(0) #define I40E_HW_FLAG_802_1AD_CAPABLE BIT_ULL(1) +#define I40E_HW_FLAG_AQ_PHY_ACCESS_CAPABLE BIT_ULL(2) u64 flags; /* Used in set switch config AQ command */ -- 2.4.11 ^ permalink raw reply [flat|nested] 69+ messages in thread
* [dpdk-dev] [PATCH v2 12/16] net/i40e/base: avoid potential null pointer dereference 2017-06-26 9:48 ` [dpdk-dev] [PATCH v2 00/16] " Jingjing Wu ` (10 preceding siblings ...) 2017-06-26 9:48 ` [dpdk-dev] [PATCH v2 11/16] net/i40e/base: use admin queue for setting LEDs behavior Jingjing Wu @ 2017-06-26 9:48 ` Jingjing Wu 2017-06-26 9:48 ` [dpdk-dev] [PATCH v2 13/16] net/i40e/base: avoid reset timeout issue Jingjing Wu ` (4 subsequent siblings) 16 siblings, 0 replies; 69+ messages in thread From: Jingjing Wu @ 2017-06-26 9:48 UTC (permalink / raw) To: helin.zhang; +Cc: dev, jingjing.wu Signed-off-by: Jingjing Wu <jingjing.wu@intel.com> --- drivers/net/i40e/base/i40e_common.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) diff --git a/drivers/net/i40e/base/i40e_common.c b/drivers/net/i40e/base/i40e_common.c index 3f3b275..78ed2a8 100644 --- a/drivers/net/i40e/base/i40e_common.c +++ b/drivers/net/i40e/base/i40e_common.c @@ -330,13 +330,15 @@ void i40e_debug_aq(struct i40e_hw *hw, enum i40e_debug_mask mask, void *desc, void *buffer, u16 buf_len) { struct i40e_aq_desc *aq_desc = (struct i40e_aq_desc *)desc; - u16 len = LE16_TO_CPU(aq_desc->datalen); u8 *buf = (u8 *)buffer; + u16 len; u16 i = 0; if ((!(mask & hw->debug_mask)) || (desc == NULL)) return; + len = LE16_TO_CPU(aq_desc->datalen); + i40e_debug(hw, mask, "AQ CMD: opcode 0x%04X, flags 0x%04X, datalen 0x%04X, retval 0x%04X\n", LE16_TO_CPU(aq_desc->opcode), -- 2.4.11 ^ permalink raw reply [flat|nested] 69+ messages in thread
* [dpdk-dev] [PATCH v2 13/16] net/i40e/base: avoid reset timeout issue 2017-06-26 9:48 ` [dpdk-dev] [PATCH v2 00/16] " Jingjing Wu ` (11 preceding siblings ...) 2017-06-26 9:48 ` [dpdk-dev] [PATCH v2 12/16] net/i40e/base: avoid potential null pointer dereference Jingjing Wu @ 2017-06-26 9:48 ` Jingjing Wu 2017-06-26 9:48 ` [dpdk-dev] [PATCH v2 14/16] net/i40e/base: add EEPROM checksum verification Jingjing Wu ` (3 subsequent siblings) 16 siblings, 0 replies; 69+ messages in thread From: Jingjing Wu @ 2017-06-26 9:48 UTC (permalink / raw) To: helin.zhang; +Cc: dev, jingjing.wu This patch allows detection of upcoming core reset in case NIC gets stuck while performing FLR reset. The i40e_pf_reset() function returns I40E_ERR_NOT_READY when global reset was detected. Signed-off-by: Jingjing Wu <jingjing.wu@intel.com> --- drivers/net/i40e/base/i40e_common.c | 8 ++++++++ 1 file changed, 8 insertions(+) diff --git a/drivers/net/i40e/base/i40e_common.c b/drivers/net/i40e/base/i40e_common.c index 78ed2a8..ed2e01a 100644 --- a/drivers/net/i40e/base/i40e_common.c +++ b/drivers/net/i40e/base/i40e_common.c @@ -1382,6 +1382,8 @@ enum i40e_status_code i40e_pf_reset(struct i40e_hw *hw) * we don't need to do the PF Reset */ if (!cnt) { + u32 reg2 = 0; + reg = rd32(hw, I40E_PFGEN_CTRL); wr32(hw, I40E_PFGEN_CTRL, (reg | I40E_PFGEN_CTRL_PFSWR_MASK)); @@ -1389,6 +1391,12 @@ enum i40e_status_code i40e_pf_reset(struct i40e_hw *hw) reg = rd32(hw, I40E_PFGEN_CTRL); if (!(reg & I40E_PFGEN_CTRL_PFSWR_MASK)) break; + reg2 = rd32(hw, I40E_GLGEN_RSTAT); + if (reg2 & I40E_GLGEN_RSTAT_DEVSTATE_MASK) { + DEBUGOUT("Core reset upcoming.\n"); + DEBUGOUT1("I40E_GLGEN_RSTAT = 0x%x\n", reg2); + return I40E_ERR_NOT_READY; + } i40e_msec_delay(1); } if (reg & I40E_PFGEN_CTRL_PFSWR_MASK) { -- 2.4.11 ^ permalink raw reply [flat|nested] 69+ messages in thread
* [dpdk-dev] [PATCH v2 14/16] net/i40e/base: add EEPROM checksum verification 2017-06-26 9:48 ` [dpdk-dev] [PATCH v2 00/16] " Jingjing Wu ` (12 preceding siblings ...) 2017-06-26 9:48 ` [dpdk-dev] [PATCH v2 13/16] net/i40e/base: avoid reset timeout issue Jingjing Wu @ 2017-06-26 9:48 ` Jingjing Wu 2017-06-26 9:48 ` [dpdk-dev] [PATCH v2 15/16] net/i40e/base: extend processing of DDP Jingjing Wu ` (2 subsequent siblings) 16 siblings, 0 replies; 69+ messages in thread From: Jingjing Wu @ 2017-06-26 9:48 UTC (permalink / raw) To: helin.zhang; +Cc: dev, jingjing.wu This patch ensures PFs mutually exclusive access to NVM. Signed-off-by: Jingjing Wu <jingjing.wu@intel.com> --- drivers/net/i40e/base/i40e_nvm.c | 14 ++++++++++---- 1 file changed, 10 insertions(+), 4 deletions(-) diff --git a/drivers/net/i40e/base/i40e_nvm.c b/drivers/net/i40e/base/i40e_nvm.c index 90521fa..a1e7830 100644 --- a/drivers/net/i40e/base/i40e_nvm.c +++ b/drivers/net/i40e/base/i40e_nvm.c @@ -749,12 +749,18 @@ enum i40e_status_code i40e_validate_nvm_checksum(struct i40e_hw *hw, DEBUGFUNC("i40e_validate_nvm_checksum"); - if (hw->flags & I40E_HW_FLAG_AQ_SRCTL_ACCESS_ENABLE) - ret_code = i40e_acquire_nvm(hw, I40E_RESOURCE_READ); + /* acquire_nvm provides exclusive NVM lock to synchronize access across + * PFs. X710 uses i40e_read_nvm_word_srctl which polls for done bit + * twice (first time to be able to write address to I40E_GLNVM_SRCTL + * register, second to read data from I40E_GLNVM_SRDATA. One PF can see + * done bit and try to write address, while another one will interpret + * it as a good time to read data. It will cause invalid data to be + * read. + */ + ret_code = i40e_acquire_nvm(hw, I40E_RESOURCE_READ); if (!ret_code) { ret_code = i40e_calc_nvm_checksum(hw, &checksum_local); - if (hw->flags & I40E_HW_FLAG_AQ_SRCTL_ACCESS_ENABLE) - i40e_release_nvm(hw); + i40e_release_nvm(hw); if (ret_code != I40E_SUCCESS) goto i40e_validate_nvm_checksum_exit; } else { -- 2.4.11 ^ permalink raw reply [flat|nested] 69+ messages in thread
* [dpdk-dev] [PATCH v2 15/16] net/i40e/base: extend processing of DDP 2017-06-26 9:48 ` [dpdk-dev] [PATCH v2 00/16] " Jingjing Wu ` (13 preceding siblings ...) 2017-06-26 9:48 ` [dpdk-dev] [PATCH v2 14/16] net/i40e/base: add EEPROM checksum verification Jingjing Wu @ 2017-06-26 9:48 ` Jingjing Wu 2017-06-26 13:18 ` Ferruh Yigit 2017-06-26 9:48 ` [dpdk-dev] [PATCH v2 16/16] net/i40e: use set switch aq instead of register setting Jingjing Wu 2017-06-27 13:29 ` [dpdk-dev] [PATCH v3 00/17] net/i40e: base code update Jingjing Wu 16 siblings, 1 reply; 69+ messages in thread From: Jingjing Wu @ 2017-06-26 9:48 UTC (permalink / raw) To: helin.zhang; +Cc: dev, jingjing.wu, Andrey Chilikin, Beilei Xing This patch adds extended processing of DDP packages: - Execution of adminq command sections to support AQ-depended profiles, for example, for programming cloud filters types. - Ability to write a profile without registering it in the list of applied profiles, to be used for AQ-depended profiles. - Profile rollback is implemented to support restoration of original parser/analyzer configuration without the need of core reset, for example, for deploying new profile without resetting device. - Search for a specific section in a profile, to be used by driver to access metadata sections with description of PCTYPE/PTYPEs defined in the profile. Signed-off-by: Andrey Chilikin <andrey.chilikin@intel.com> Signed-off-by: Beilei Xing <beilei.xing@intel.com> Signed-off-by: Jingjing Wu <jingjing.wu@intel.com> --- drivers/net/i40e/base/README | 2 +- drivers/net/i40e/base/i40e_common.c | 258 ++++++++++++++++++++++++++++++--- drivers/net/i40e/base/i40e_prototype.h | 6 + drivers/net/i40e/base/i40e_type.h | 25 +++- 4 files changed, 266 insertions(+), 25 deletions(-) diff --git a/drivers/net/i40e/base/README b/drivers/net/i40e/base/README index 0da9f67..59e76c2 100644 --- a/drivers/net/i40e/base/README +++ b/drivers/net/i40e/base/README @@ -34,7 +34,7 @@ Intel® I40E driver ================== This directory contains source code of FreeBSD i40e driver of version -cid-i40e.2017.03.21.tar.gz released by the team which develops +cid-i40e.2017.06.23.tar.gz released by the team which develops basic drivers for any i40e NIC. The directory of base/ contains the original source package. This driver is valid for the product(s) listed below diff --git a/drivers/net/i40e/base/i40e_common.c b/drivers/net/i40e/base/i40e_common.c index ed2e01a..900d379 100644 --- a/drivers/net/i40e/base/i40e_common.c +++ b/drivers/net/i40e/base/i40e_common.c @@ -7441,6 +7441,165 @@ i40e_find_segment_in_package(u32 segment_type, return NULL; } +/* Get section table in profile */ +#define I40E_SECTION_TABLE(profile, sec_tbl) \ + do { \ + struct i40e_profile_segment *p = (profile); \ + u32 count; \ + u32 *nvm; \ + count = p->device_table_count; \ + nvm = (u32 *)&p->device_table[count]; \ + sec_tbl = (struct i40e_section_table *)&nvm[nvm[0] + 1]; \ + } while (0) + +/* Get section header in profile */ +#define I40E_SECTION_HEADER(profile, offset) \ + (struct i40e_profile_section_header *)((u8 *)(profile) + (offset)) + +/** + * i40e_find_section_in_profile + * @section_type: the section type to search for (i.e., SECTION_TYPE_NOTE) + * @profile: pointer to the i40e segment header to be searched + * + * This function searches i40e segment for a particular section type. On + * success it returns a pointer to the section header, otherwise it will + * return NULL. + **/ +struct i40e_profile_section_header * +i40e_find_section_in_profile(u32 section_type, + struct i40e_profile_segment *profile) +{ + struct i40e_profile_section_header *sec; + struct i40e_section_table *sec_tbl; + u32 sec_off; + u32 i; + + if (profile->header.type != SEGMENT_TYPE_I40E) + return NULL; + + I40E_SECTION_TABLE(profile, sec_tbl); + + for (i = 0; i < sec_tbl->section_count; i++) { + sec_off = sec_tbl->section_offset[i]; + sec = I40E_SECTION_HEADER(profile, sec_off); + if (sec->section.type == section_type) + return sec; + } + + return NULL; +} + +/** + * i40e_ddp_exec_aq_section - Execute generic AQ for DDP + * @hw: pointer to the hw struct + * @aq: command buffer containing all data to execute AQ + **/ +STATIC enum +i40e_status_code i40e_ddp_exec_aq_section(struct i40e_hw *hw, + struct i40e_profile_aq_section *aq) +{ + enum i40e_status_code status; + struct i40e_aq_desc desc; + u8 *msg = NULL; + u16 msglen; + + i40e_fill_default_direct_cmd_desc(&desc, aq->opcode); + desc.flags |= CPU_TO_LE16(aq->flags); + i40e_memcpy(desc.params.raw, aq->param, sizeof(desc.params.raw), + I40E_NONDMA_TO_NONDMA); + + msglen = aq->datalen; + if (msglen) { + desc.flags |= CPU_TO_LE16((u16)(I40E_AQ_FLAG_BUF | + I40E_AQ_FLAG_RD)); + if (msglen > I40E_AQ_LARGE_BUF) + desc.flags |= CPU_TO_LE16((u16)I40E_AQ_FLAG_LB); + desc.datalen = CPU_TO_LE16(msglen); + msg = &aq->data[0]; + } + + status = i40e_asq_send_command(hw, &desc, msg, msglen, NULL); + + if (status != I40E_SUCCESS) { + i40e_debug(hw, I40E_DEBUG_PACKAGE, + "unable to exec DDP AQ opcode %u, error %d\n", + aq->opcode, status); + return status; + } + + /* copy returned desc to aq_buf */ + i40e_memcpy(aq->param, desc.params.raw, sizeof(desc.params.raw), + I40E_NONDMA_TO_NONDMA); + + return I40E_SUCCESS; +} + +/** + * i40e_validate_profile + * @hw: pointer to the hardware structure + * @profile: pointer to the profile segment of the package to be validated + * @track_id: package tracking id + * @rollback: flag if the profile is for rollback. + * + * Validates supported devices and profile's sections. + */ +STATIC enum i40e_status_code +i40e_validate_profile(struct i40e_hw *hw, struct i40e_profile_segment *profile, + u32 track_id, bool rollback) +{ + struct i40e_profile_section_header *sec = NULL; + enum i40e_status_code status = I40E_SUCCESS; + struct i40e_section_table *sec_tbl; + u32 vendor_dev_id; + u32 dev_cnt; + u32 sec_off; + u32 i; + + if (track_id == I40E_DDP_TRACKID_INVALID) { + i40e_debug(hw, I40E_DEBUG_PACKAGE, "Invalid track_id\n"); + return I40E_NOT_SUPPORTED; + } + + dev_cnt = profile->device_table_count; + for (i = 0; i < dev_cnt; i++) { + vendor_dev_id = profile->device_table[i].vendor_dev_id; + if ((vendor_dev_id >> 16) == I40E_INTEL_VENDOR_ID && + hw->device_id == (vendor_dev_id & 0xFFFF)) + break; + } + if (dev_cnt && (i == dev_cnt)) { + i40e_debug(hw, I40E_DEBUG_PACKAGE, + "Device doesn't support DDP\n"); + return I40E_ERR_DEVICE_NOT_SUPPORTED; + } + + I40E_SECTION_TABLE(profile, sec_tbl); + + /* Validate sections types */ + for (i = 0; i < sec_tbl->section_count; i++) { + sec_off = sec_tbl->section_offset[i]; + sec = I40E_SECTION_HEADER(profile, sec_off); + if (rollback) { + if (sec->section.type == SECTION_TYPE_MMIO || + sec->section.type == SECTION_TYPE_AQ || + sec->section.type == SECTION_TYPE_RB_AQ) { + i40e_debug(hw, I40E_DEBUG_PACKAGE, + "Not a roll-back package\n"); + return I40E_NOT_SUPPORTED; + } + } else { + if (sec->section.type == SECTION_TYPE_RB_AQ || + sec->section.type == SECTION_TYPE_RB_MMIO) { + i40e_debug(hw, I40E_DEBUG_PACKAGE, + "Not an original package\n"); + return I40E_NOT_SUPPORTED; + } + } + } + + return status; +} + /** * i40e_write_profile * @hw: pointer to the hardware structure @@ -7456,47 +7615,99 @@ i40e_write_profile(struct i40e_hw *hw, struct i40e_profile_segment *profile, enum i40e_status_code status = I40E_SUCCESS; struct i40e_section_table *sec_tbl; struct i40e_profile_section_header *sec = NULL; - u32 dev_cnt; - u32 vendor_dev_id; - u32 *nvm; + struct i40e_profile_aq_section *ddp_aq; u32 section_size = 0; u32 offset = 0, info = 0; + u32 sec_off; u32 i; - dev_cnt = profile->device_table_count; + status = i40e_validate_profile(hw, profile, track_id, false); + if (status) + return status; - for (i = 0; i < dev_cnt; i++) { - vendor_dev_id = profile->device_table[i].vendor_dev_id; - if ((vendor_dev_id >> 16) == I40E_INTEL_VENDOR_ID) - if (hw->device_id == (vendor_dev_id & 0xFFFF)) + I40E_SECTION_TABLE(profile, sec_tbl); + + for (i = 0; i < sec_tbl->section_count; i++) { + sec_off = sec_tbl->section_offset[i]; + sec = I40E_SECTION_HEADER(profile, sec_off); + /* Process generic admin command */ + if (sec->section.type == SECTION_TYPE_AQ) { + ddp_aq = (struct i40e_profile_aq_section *)&sec[1]; + status = i40e_ddp_exec_aq_section(hw, ddp_aq); + if (status) { + i40e_debug(hw, I40E_DEBUG_PACKAGE, + "Failed to execute aq: section %d, opcode %u\n", + i, ddp_aq->opcode); break; + } + sec->section.type = SECTION_TYPE_RB_AQ; + } + + /* Skip any non-mmio sections */ + if (sec->section.type != SECTION_TYPE_MMIO) + continue; + + section_size = sec->section.size + + sizeof(struct i40e_profile_section_header); + + /* Write MMIO section */ + status = i40e_aq_write_ddp(hw, (void *)sec, (u16)section_size, + track_id, &offset, &info, NULL); + if (status) { + i40e_debug(hw, I40E_DEBUG_PACKAGE, + "Failed to write profile: section %d, offset %d, info %d\n", + i, offset, info); + break; + } } - if (i == dev_cnt) { - i40e_debug(hw, I40E_DEBUG_PACKAGE, "Device doesn't support DDP"); - return I40E_ERR_DEVICE_NOT_SUPPORTED; - } + return status; +} + +/** + * i40e_rollback_profile + * @hw: pointer to the hardware structure + * @profile: pointer to the profile segment of the package to be removed + * @track_id: package tracking id + * + * Rolls back previously loaded package. + */ +enum i40e_status_code +i40e_rollback_profile(struct i40e_hw *hw, struct i40e_profile_segment *profile, + u32 track_id) +{ + struct i40e_profile_section_header *sec = NULL; + enum i40e_status_code status = I40E_SUCCESS; + struct i40e_section_table *sec_tbl; + u32 offset = 0, info = 0; + u32 section_size = 0; + u32 sec_off; + int i; - nvm = (u32 *)&profile->device_table[dev_cnt]; - sec_tbl = (struct i40e_section_table *)&nvm[nvm[0] + 1]; + status = i40e_validate_profile(hw, profile, track_id, true); + if (status) + return status; - for (i = 0; i < sec_tbl->section_count; i++) { - sec = (struct i40e_profile_section_header *)((u8 *)profile + - sec_tbl->section_offset[i]); + I40E_SECTION_TABLE(profile, sec_tbl); - /* Skip 'AQ', 'note' and 'name' sections */ - if (sec->section.type != SECTION_TYPE_MMIO) + /* For rollback write sections in reverse */ + for (i = sec_tbl->section_count - 1; i >= 0; i--) { + sec_off = sec_tbl->section_offset[i]; + sec = I40E_SECTION_HEADER(profile, sec_off); + + /* Skip any non-rollback sections */ + if (sec->section.type != SECTION_TYPE_RB_MMIO) continue; section_size = sec->section.size + sizeof(struct i40e_profile_section_header); - /* Write profile */ + /* Write roll-back MMIO section */ status = i40e_aq_write_ddp(hw, (void *)sec, (u16)section_size, track_id, &offset, &info, NULL); if (status) { i40e_debug(hw, I40E_DEBUG_PACKAGE, - "Failed to write profile: offset %d, info %d", - offset, info); + "Failed to write profile: section %d, offset %d, info %d\n", + i, offset, info); break; } } @@ -7534,7 +7745,8 @@ i40e_add_pinfo_to_list(struct i40e_hw *hw, pinfo->track_id = track_id; pinfo->version = profile->version; pinfo->op = I40E_DDP_ADD_TRACKID; - memcpy(pinfo->name, profile->name, I40E_DDP_NAME_SIZE); + i40e_memcpy(pinfo->name, profile->name, I40E_DDP_NAME_SIZE, + I40E_NONDMA_TO_NONDMA); status = i40e_aq_write_ddp(hw, (void *)sec, sec->data_end, track_id, &offset, &info, NULL); diff --git a/drivers/net/i40e/base/i40e_prototype.h b/drivers/net/i40e/base/i40e_prototype.h index 9171e97..acb2023 100644 --- a/drivers/net/i40e/base/i40e_prototype.h +++ b/drivers/net/i40e/base/i40e_prototype.h @@ -586,10 +586,16 @@ enum i40e_status_code i40e_aq_get_ddp_list(struct i40e_hw *hw, void *buff, struct i40e_generic_seg_header * i40e_find_segment_in_package(u32 segment_type, struct i40e_package_header *pkg_header); +struct i40e_profile_section_header * +i40e_find_section_in_profile(u32 section_type, + struct i40e_profile_segment *profile); enum i40e_status_code i40e_write_profile(struct i40e_hw *hw, struct i40e_profile_segment *i40e_seg, u32 track_id); enum i40e_status_code +i40e_rollback_profile(struct i40e_hw *hw, struct i40e_profile_segment *i40e_seg, + u32 track_id); +enum i40e_status_code i40e_add_pinfo_to_list(struct i40e_hw *hw, struct i40e_profile_segment *profile, u8 *profile_info_sec, u32 track_id); diff --git a/drivers/net/i40e/base/i40e_type.h b/drivers/net/i40e/base/i40e_type.h index 9d7b1a2..dca725a 100644 --- a/drivers/net/i40e/base/i40e_type.h +++ b/drivers/net/i40e/base/i40e_type.h @@ -1918,8 +1918,10 @@ struct i40e_generic_seg_header { struct i40e_metadata_segment { struct i40e_generic_seg_header header; struct i40e_ddp_version version; +#define I40E_DDP_TRACKID_RDONLY 0 +#define I40E_DDP_TRACKID_INVALID 0xFFFFFFFF u32 track_id; - char name[I40E_DDP_NAME_SIZE]; + char name[I40E_DDP_NAME_SIZE]; }; struct i40e_device_id_entry { @@ -1946,15 +1948,36 @@ struct i40e_profile_section_header { struct { #define SECTION_TYPE_INFO 0x00000010 #define SECTION_TYPE_MMIO 0x00000800 +#define SECTION_TYPE_RB_MMIO 0x00001800 #define SECTION_TYPE_AQ 0x00000801 +#define SECTION_TYPE_RB_AQ 0x00001801 #define SECTION_TYPE_NOTE 0x80000000 #define SECTION_TYPE_NAME 0x80000001 +#define SECTION_TYPE_PROTO 0x80000002 +#define SECTION_TYPE_PCTYPE 0x80000003 +#define SECTION_TYPE_PTYPE 0x80000004 u32 type; u32 offset; u32 size; } section; }; +struct i40e_profile_tlv_section_record { + u8 rtype; + u8 type; + u16 len; + u8 data[12]; +}; + +/* Generic AQ section in proflie */ +struct i40e_profile_aq_section { + u16 opcode; + u16 flags; + u8 param[16]; + u16 datalen; + u8 data[1]; +}; + struct i40e_profile_info { u32 track_id; struct i40e_ddp_version version; -- 2.4.11 ^ permalink raw reply [flat|nested] 69+ messages in thread
* Re: [dpdk-dev] [PATCH v2 15/16] net/i40e/base: extend processing of DDP 2017-06-26 9:48 ` [dpdk-dev] [PATCH v2 15/16] net/i40e/base: extend processing of DDP Jingjing Wu @ 2017-06-26 13:18 ` Ferruh Yigit 0 siblings, 0 replies; 69+ messages in thread From: Ferruh Yigit @ 2017-06-26 13:18 UTC (permalink / raw) To: Jingjing Wu, helin.zhang; +Cc: dev, Andrey Chilikin, Beilei Xing On 6/26/2017 10:48 AM, Jingjing Wu wrote: > This patch adds extended processing of DDP packages: > - Execution of adminq command sections to support AQ-depended profiles, > for example, for programming cloud filters types. > - Ability to write a profile without registering it in the list of > applied profiles, to be used for AQ-depended profiles. > - Profile rollback is implemented to support restoration of original > parser/analyzer configuration without the need of core reset, > for example, for deploying new profile without resetting device. > - Search for a specific section in a profile, to be used by driver > to access metadata sections with description of PCTYPE/PTYPEs > defined in the profile. > > Signed-off-by: Andrey Chilikin <andrey.chilikin@intel.com> > Signed-off-by: Beilei Xing <beilei.xing@intel.com> > Signed-off-by: Jingjing Wu <jingjing.wu@intel.com> > --- > drivers/net/i40e/base/README | 2 +- > drivers/net/i40e/base/i40e_common.c | 258 ++++++++++++++++++++++++++++++--- > drivers/net/i40e/base/i40e_prototype.h | 6 + > drivers/net/i40e/base/i40e_type.h | 25 +++- > 4 files changed, 266 insertions(+), 25 deletions(-) > > diff --git a/drivers/net/i40e/base/README b/drivers/net/i40e/base/README > index 0da9f67..59e76c2 100644 > --- a/drivers/net/i40e/base/README > +++ b/drivers/net/i40e/base/README > @@ -34,7 +34,7 @@ Intel® I40E driver > ================== > > This directory contains source code of FreeBSD i40e driver of version > -cid-i40e.2017.03.21.tar.gz released by the team which develops > +cid-i40e.2017.06.23.tar.gz released by the team which develops > basic drivers for any i40e NIC. The directory of base/ contains the > original source package. > This driver is valid for the product(s) listed below Hi Jingjing, Can you please extract this part into different patch? Thanks, ferruh ^ permalink raw reply [flat|nested] 69+ messages in thread
* [dpdk-dev] [PATCH v2 16/16] net/i40e: use set switch aq instead of register setting 2017-06-26 9:48 ` [dpdk-dev] [PATCH v2 00/16] " Jingjing Wu ` (14 preceding siblings ...) 2017-06-26 9:48 ` [dpdk-dev] [PATCH v2 15/16] net/i40e/base: extend processing of DDP Jingjing Wu @ 2017-06-26 9:48 ` Jingjing Wu 2017-06-26 13:20 ` Ferruh Yigit 2017-06-27 13:29 ` [dpdk-dev] [PATCH v3 00/17] net/i40e: base code update Jingjing Wu 16 siblings, 1 reply; 69+ messages in thread From: Jingjing Wu @ 2017-06-26 9:48 UTC (permalink / raw) To: helin.zhang; +Cc: dev, jingjing.wu TPID can be set by set_switch_config aq, change the TPID setting by set_switch_config on new FW release. Signed-off-by: Jingjing Wu <jingjing.wu@intel.com> --- drivers/net/i40e/i40e_ethdev.c | 104 ++++++++++++++++++++++++++--------------- 1 file changed, 66 insertions(+), 38 deletions(-) diff --git a/drivers/net/i40e/i40e_ethdev.c b/drivers/net/i40e/i40e_ethdev.c index 4ee1113..ae93424 100644 --- a/drivers/net/i40e/i40e_ethdev.c +++ b/drivers/net/i40e/i40e_ethdev.c @@ -2973,71 +2973,99 @@ i40e_vlan_filter_set(struct rte_eth_dev *dev, uint16_t vlan_id, int on) } static int -i40e_vlan_tpid_set(struct rte_eth_dev *dev, - enum rte_vlan_type vlan_type, - uint16_t tpid) +i40e_vlan_tpid_set_by_registers(struct rte_eth_dev *dev, + enum rte_vlan_type vlan_type, + uint16_t tpid, int qinq) { struct i40e_hw *hw = I40E_DEV_PRIVATE_TO_HW(dev->data->dev_private); - uint64_t reg_r = 0, reg_w = 0; - uint16_t reg_id = 0; - int ret = 0; - int qinq = dev->data->dev_conf.rxmode.hw_vlan_extend; + uint64_t reg_r = 0; + uint64_t reg_w = 0; + uint16_t reg_id = 3; + int ret; - switch (vlan_type) { - case ETH_VLAN_TYPE_OUTER: - if (qinq) + if (qinq) { + if (vlan_type == ETH_VLAN_TYPE_OUTER) reg_id = 2; - else + else if (vlan_type == ETH_VLAN_TYPE_INNER) reg_id = 3; - break; - case ETH_VLAN_TYPE_INNER: - if (qinq) + } else { + if (vlan_type == ETH_VLAN_TYPE_OUTER) reg_id = 3; - else { - ret = -EINVAL; - PMD_DRV_LOG(ERR, - "Unsupported vlan type in single vlan."); - return ret; - } - break; - default: - ret = -EINVAL; - PMD_DRV_LOG(ERR, "Unsupported vlan type %d", vlan_type); - return ret; } + ret = i40e_aq_debug_read_register(hw, I40E_GL_SWT_L2TAGCTRL(reg_id), ®_r, NULL); if (ret != I40E_SUCCESS) { PMD_DRV_LOG(ERR, "Fail to debug read from I40E_GL_SWT_L2TAGCTRL[%d]", reg_id); - ret = -EIO; - return ret; + return -EIO; } PMD_DRV_LOG(DEBUG, - "Debug read from I40E_GL_SWT_L2TAGCTRL[%d]: 0x%08"PRIx64, - reg_id, reg_r); + "Debug read from I40E_GL_SWT_L2TAGCTRL[%d]: 0x%08"PRIx64, + reg_id, reg_r); reg_w = reg_r & (~(I40E_GL_SWT_L2TAGCTRL_ETHERTYPE_MASK)); reg_w |= ((uint64_t)tpid << I40E_GL_SWT_L2TAGCTRL_ETHERTYPE_SHIFT); if (reg_r == reg_w) { - ret = 0; PMD_DRV_LOG(DEBUG, "No need to write"); - return ret; + return 0; } ret = i40e_aq_debug_write_register(hw, I40E_GL_SWT_L2TAGCTRL(reg_id), reg_w, NULL); if (ret != I40E_SUCCESS) { - ret = -EIO; PMD_DRV_LOG(ERR, - "Fail to debug write to I40E_GL_SWT_L2TAGCTRL[%d]", - reg_id); - return ret; + "Fail to debug write to I40E_GL_SWT_L2TAGCTRL[%d]", + reg_id); + return -EIO; } PMD_DRV_LOG(DEBUG, - "Debug write 0x%08"PRIx64" to I40E_GL_SWT_L2TAGCTRL[%d]", - reg_w, reg_id); + "Debug write 0x%08"PRIx64" to I40E_GL_SWT_L2TAGCTRL[%d]", + reg_w, reg_id); + + return 0; +} + +static int +i40e_vlan_tpid_set(struct rte_eth_dev *dev, + enum rte_vlan_type vlan_type, + uint16_t tpid) +{ + struct i40e_hw *hw = I40E_DEV_PRIVATE_TO_HW(dev->data->dev_private); + int qinq = dev->data->dev_conf.rxmode.hw_vlan_extend; + int ret = 0; + + if ((vlan_type != ETH_VLAN_TYPE_INNER && + vlan_type != ETH_VLAN_TYPE_OUTER) || + (!qinq && vlan_type == ETH_VLAN_TYPE_INNER)) { + PMD_DRV_LOG(ERR, + "Unsupported vlan type."); + return -EINVAL; + + } + /* 802.1ad frames ability is added in NVM API 1.7*/ + if (hw->flags & I40E_HW_FLAG_802_1AD_CAPABLE) { + if (qinq) { + if (vlan_type == ETH_VLAN_TYPE_OUTER) + hw->first_tag = rte_cpu_to_le_16(tpid); + else if (vlan_type == ETH_VLAN_TYPE_INNER) + hw->second_tag = rte_cpu_to_le_16(tpid); + } else { + if (vlan_type == ETH_VLAN_TYPE_OUTER) + hw->second_tag = rte_cpu_to_le_16(tpid); + } + ret = i40e_aq_set_switch_config(hw, 0, 0, NULL); + if (ret != I40E_SUCCESS) { + PMD_DRV_LOG(ERR, + "Set switch config failed aq_err: %d", + hw->aq.asq_last_status); + ret = -EIO; + } + } else + /* If NVM API < 1.7, keep the register setting */ + ret = i40e_vlan_tpid_set_by_registers(dev, vlan_type, + tpid, qinq); return ret; } @@ -3066,7 +3094,7 @@ i40e_vlan_offload_set(struct rte_eth_dev *dev, int mask) if (mask & ETH_VLAN_EXTEND_MASK) { if (dev->data->dev_conf.rxmode.hw_vlan_extend) { i40e_vsi_config_double_vlan(vsi, TRUE); - /* Set global registers with default ether type value */ + /* Set global registers with default ether type value. */ i40e_vlan_tpid_set(dev, ETH_VLAN_TYPE_OUTER, ETHER_TYPE_VLAN); i40e_vlan_tpid_set(dev, ETH_VLAN_TYPE_INNER, -- 2.4.11 ^ permalink raw reply [flat|nested] 69+ messages in thread
* Re: [dpdk-dev] [PATCH v2 16/16] net/i40e: use set switch aq instead of register setting 2017-06-26 9:48 ` [dpdk-dev] [PATCH v2 16/16] net/i40e: use set switch aq instead of register setting Jingjing Wu @ 2017-06-26 13:20 ` Ferruh Yigit 0 siblings, 0 replies; 69+ messages in thread From: Ferruh Yigit @ 2017-06-26 13:20 UTC (permalink / raw) To: Jingjing Wu, helin.zhang; +Cc: dev On 6/26/2017 10:48 AM, Jingjing Wu wrote: > TPID can be set by set_switch_config aq, change the TPID setting > by set_switch_config on new FW release. Hi Jingjing, Since you will already send a new version of the patchset, can you please address the checkpatch warnings in this patch too? Thanks, ferruh > > Signed-off-by: Jingjing Wu <jingjing.wu@intel.com> <...> ^ permalink raw reply [flat|nested] 69+ messages in thread
* [dpdk-dev] [PATCH v3 00/17] net/i40e: base code update 2017-06-26 9:48 ` [dpdk-dev] [PATCH v2 00/16] " Jingjing Wu ` (15 preceding siblings ...) 2017-06-26 9:48 ` [dpdk-dev] [PATCH v2 16/16] net/i40e: use set switch aq instead of register setting Jingjing Wu @ 2017-06-27 13:29 ` Jingjing Wu 2017-06-27 13:29 ` [dpdk-dev] [PATCH v3 01/17] net/i40e/base: use new virtchnl header file Jingjing Wu ` (17 more replies) 16 siblings, 18 replies; 69+ messages in thread From: Jingjing Wu @ 2017-06-27 13:29 UTC (permalink / raw) To: dev; +Cc: jingjing.wu, helin.zhang v3 chanes: - extra patch to update README - remove unnecessary code - fix checkpatch warnings v2 changes: - fix compile issue - update from package from 2017.05.24 to 2017.06.23 i40e base code upate. The main changes are: - use virtchnl.h instead of i40e_virtchnl.h - add support for Adaptive Virtual Function - add new AQ commands for read/write PHY registers - add new phy types for 25G - extend processing of DDP Jingjing Wu (17): net/i40e/base: use new virtchnl header file net/i40e/base: sync nvmupdate command and adminq subtask net/i40e/base: add AQ command for read/write PHY registers net/i40e/base: add support for Adaptive Virtual Function net/i40e/base: store the requested FEC information net/i40e/base: add new phy types for 25G AOC and ACC net/i40e/base: report supported link modes net/i40e/base: track id can be 0 net/i40e/base: update FW AQ API version to 1.7 net/i40e/base: add support for switch parameters net/i40e/base: use admin queue for setting LEDs behavior net/i40e/base: avoid potential null pointer dereference net/i40e/base: avoid reset timeout issue net/i40e/base: add EEPROM checksum verification net/i40e/base: extend processing of DDP net/i40e: use set switch aq instead of register setting net/i40e/base: update base code info drivers/net/i40e/base/README | 2 +- drivers/net/i40e/base/i40e_adminq.c | 12 + drivers/net/i40e/base/i40e_adminq_cmd.h | 71 ++- drivers/net/i40e/base/i40e_common.c | 549 +++++++++++++++++++---- drivers/net/i40e/base/i40e_devids.h | 1 + drivers/net/i40e/base/i40e_nvm.c | 20 +- drivers/net/i40e/base/i40e_prototype.h | 33 +- drivers/net/i40e/base/i40e_register.h | 2 +- drivers/net/i40e/base/i40e_type.h | 49 +- drivers/net/i40e/base/i40e_virtchnl.h | 445 ------------------ drivers/net/i40e/base/virtchnl.h | 772 ++++++++++++++++++++++++++++++++ drivers/net/i40e/i40e_ethdev.c | 102 +++-- drivers/net/i40e/i40e_ethdev.h | 8 +- drivers/net/i40e/i40e_ethdev_vf.c | 206 ++++----- drivers/net/i40e/i40e_pf.c | 206 ++++----- drivers/net/i40e/i40e_pf.h | 32 +- drivers/net/i40e/rte_pmd_i40e.h | 2 +- 17 files changed, 1690 insertions(+), 822 deletions(-) delete mode 100644 drivers/net/i40e/base/i40e_virtchnl.h create mode 100644 drivers/net/i40e/base/virtchnl.h -- 2.4.11 ^ permalink raw reply [flat|nested] 69+ messages in thread
* [dpdk-dev] [PATCH v3 01/17] net/i40e/base: use new virtchnl header file 2017-06-27 13:29 ` [dpdk-dev] [PATCH v3 00/17] net/i40e: base code update Jingjing Wu @ 2017-06-27 13:29 ` Jingjing Wu 2017-06-27 13:29 ` [dpdk-dev] [PATCH v3 02/17] net/i40e/base: sync nvmupdate command and adminq subtask Jingjing Wu ` (16 subsequent siblings) 17 siblings, 0 replies; 69+ messages in thread From: Jingjing Wu @ 2017-06-27 13:29 UTC (permalink / raw) To: dev; +Cc: jingjing.wu, helin.zhang Modify the necessary files to be compatible with the new virtchnl.h file instead of relying on i40e_virtchnl.h variant. This mostly changes references to VIRTCHNL_ variables by removing prefix of I40E_. Signed-off-by: Jingjing Wu <jingjing.wu@intel.com> --- drivers/net/i40e/base/i40e_common.c | 24 +- drivers/net/i40e/base/i40e_prototype.h | 6 +- drivers/net/i40e/base/i40e_type.h | 12 +- drivers/net/i40e/base/i40e_virtchnl.h | 445 ------------------- drivers/net/i40e/base/virtchnl.h | 772 +++++++++++++++++++++++++++++++++ drivers/net/i40e/i40e_ethdev.h | 8 +- drivers/net/i40e/i40e_ethdev_vf.c | 206 ++++----- drivers/net/i40e/i40e_pf.c | 206 ++++----- drivers/net/i40e/i40e_pf.h | 32 +- drivers/net/i40e/rte_pmd_i40e.h | 2 +- 10 files changed, 1020 insertions(+), 693 deletions(-) delete mode 100644 drivers/net/i40e/base/i40e_virtchnl.h create mode 100644 drivers/net/i40e/base/virtchnl.h diff --git a/drivers/net/i40e/base/i40e_common.c b/drivers/net/i40e/base/i40e_common.c index 03e94bc..6c09c27 100644 --- a/drivers/net/i40e/base/i40e_common.c +++ b/drivers/net/i40e/base/i40e_common.c @@ -34,7 +34,7 @@ POSSIBILITY OF SUCH DAMAGE. #include "i40e_type.h" #include "i40e_adminq.h" #include "i40e_prototype.h" -#include "i40e_virtchnl.h" +#include "virtchnl.h" /** @@ -5523,7 +5523,7 @@ enum i40e_status_code i40e_aq_add_rem_control_packet_filter(struct i40e_hw *hw, } if (mac_addr) - i40e_memcpy(cmd->mac, mac_addr, I40E_ETH_LENGTH_OF_ADDRESS, + i40e_memcpy(cmd->mac, mac_addr, ETH_ALEN, I40E_NONDMA_TO_NONDMA); cmd->etype = CPU_TO_LE16(ethtype); @@ -6879,7 +6879,7 @@ void i40e_write_rx_ctl(struct i40e_hw *hw, u32 reg_addr, u32 reg_val) * completion before returning. **/ enum i40e_status_code i40e_aq_send_msg_to_pf(struct i40e_hw *hw, - enum i40e_virtchnl_ops v_opcode, + enum virtchnl_ops v_opcode, enum i40e_status_code v_retval, u8 *msg, u16 msglen, struct i40e_asq_cmd_details *cmd_details) @@ -6918,9 +6918,9 @@ enum i40e_status_code i40e_aq_send_msg_to_pf(struct i40e_hw *hw, * with appropriate information. **/ void i40e_vf_parse_hw_config(struct i40e_hw *hw, - struct i40e_virtchnl_vf_resource *msg) + struct virtchnl_vf_resource *msg) { - struct i40e_virtchnl_vsi_resource *vsi_res; + struct virtchnl_vsi_resource *vsi_res; int i; vsi_res = &msg->vsi_res[0]; @@ -6930,19 +6930,17 @@ void i40e_vf_parse_hw_config(struct i40e_hw *hw, hw->dev_caps.num_tx_qp = msg->num_queue_pairs; hw->dev_caps.num_msix_vectors_vf = msg->max_vectors; hw->dev_caps.dcb = msg->vf_offload_flags & - I40E_VIRTCHNL_VF_OFFLOAD_L2; - hw->dev_caps.fcoe = (msg->vf_offload_flags & - I40E_VIRTCHNL_VF_OFFLOAD_FCOE) ? 1 : 0; + VIRTCHNL_VF_OFFLOAD_L2; hw->dev_caps.iwarp = (msg->vf_offload_flags & - I40E_VIRTCHNL_VF_OFFLOAD_IWARP) ? 1 : 0; + VIRTCHNL_VF_OFFLOAD_IWARP) ? 1 : 0; for (i = 0; i < msg->num_vsis; i++) { - if (vsi_res->vsi_type == I40E_VSI_SRIOV) { + if (vsi_res->vsi_type == VIRTCHNL_VSI_SRIOV) { i40e_memcpy(hw->mac.perm_addr, vsi_res->default_mac_addr, - I40E_ETH_LENGTH_OF_ADDRESS, + ETH_ALEN, I40E_NONDMA_TO_NONDMA); i40e_memcpy(hw->mac.addr, vsi_res->default_mac_addr, - I40E_ETH_LENGTH_OF_ADDRESS, + ETH_ALEN, I40E_NONDMA_TO_NONDMA); } vsi_res++; @@ -6959,7 +6957,7 @@ void i40e_vf_parse_hw_config(struct i40e_hw *hw, **/ enum i40e_status_code i40e_vf_reset(struct i40e_hw *hw) { - return i40e_aq_send_msg_to_pf(hw, I40E_VIRTCHNL_OP_RESET_VF, + return i40e_aq_send_msg_to_pf(hw, VIRTCHNL_OP_RESET_VF, I40E_SUCCESS, NULL, 0, NULL); } #endif /* VF_DRIVER */ diff --git a/drivers/net/i40e/base/i40e_prototype.h b/drivers/net/i40e/base/i40e_prototype.h index 4bd589e..6ec4304 100644 --- a/drivers/net/i40e/base/i40e_prototype.h +++ b/drivers/net/i40e/base/i40e_prototype.h @@ -36,7 +36,7 @@ POSSIBILITY OF SUCH DAMAGE. #include "i40e_type.h" #include "i40e_alloc.h" -#include "i40e_virtchnl.h" +#include "virtchnl.h" /* Prototypes for shared code functions that are not in * the standard function pointer structures. These are @@ -504,10 +504,10 @@ void i40e_destroy_spinlock(struct i40e_spinlock *sp); /* i40e_common for VF drivers*/ void i40e_vf_parse_hw_config(struct i40e_hw *hw, - struct i40e_virtchnl_vf_resource *msg); + struct virtchnl_vf_resource *msg); enum i40e_status_code i40e_vf_reset(struct i40e_hw *hw); enum i40e_status_code i40e_aq_send_msg_to_pf(struct i40e_hw *hw, - enum i40e_virtchnl_ops v_opcode, + enum virtchnl_ops v_opcode, enum i40e_status_code v_retval, u8 *msg, u16 msglen, struct i40e_asq_cmd_details *cmd_details); diff --git a/drivers/net/i40e/base/i40e_type.h b/drivers/net/i40e/base/i40e_type.h index 84d5757..52a114a 100644 --- a/drivers/net/i40e/base/i40e_type.h +++ b/drivers/net/i40e/base/i40e_type.h @@ -92,7 +92,9 @@ POSSIBILITY OF SUCH DAMAGE. struct i40e_hw; typedef void (*I40E_ADMINQ_CALLBACK)(struct i40e_hw *, struct i40e_aq_desc *); -#define I40E_ETH_LENGTH_OF_ADDRESS 6 +#ifndef ETH_ALEN +#define ETH_ALEN 6 +#endif /* Data type manipulation macros. */ #define I40E_HI_DWORD(x) ((u32)((((x) >> 16) >> 16) & 0xFFFFFFFF)) #define I40E_LO_DWORD(x) ((u32)((x) & 0xFFFFFFFF)) @@ -439,10 +441,10 @@ struct i40e_hw_capabilities { struct i40e_mac_info { enum i40e_mac_type type; - u8 addr[I40E_ETH_LENGTH_OF_ADDRESS]; - u8 perm_addr[I40E_ETH_LENGTH_OF_ADDRESS]; - u8 san_addr[I40E_ETH_LENGTH_OF_ADDRESS]; - u8 port_addr[I40E_ETH_LENGTH_OF_ADDRESS]; + u8 addr[ETH_ALEN]; + u8 perm_addr[ETH_ALEN]; + u8 san_addr[ETH_ALEN]; + u8 port_addr[ETH_ALEN]; u16 max_fcoeq; }; diff --git a/drivers/net/i40e/base/i40e_virtchnl.h b/drivers/net/i40e/base/i40e_virtchnl.h deleted file mode 100644 index 7a24c0f..0000000 --- a/drivers/net/i40e/base/i40e_virtchnl.h +++ /dev/null @@ -1,445 +0,0 @@ -/******************************************************************************* - -Copyright (c) 2013 - 2015, Intel Corporation -All rights reserved. - -Redistribution and use in source and binary forms, with or without -modification, are permitted provided that the following conditions are met: - - 1. Redistributions of source code must retain the above copyright notice, - this list of conditions and the following disclaimer. - - 2. Redistributions in binary form must reproduce the above copyright - notice, this list of conditions and the following disclaimer in the - documentation and/or other materials provided with the distribution. - - 3. Neither the name of the Intel Corporation nor the names of its - contributors may be used to endorse or promote products derived from - this software without specific prior written permission. - -THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS" -AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE -IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE -ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE -LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR -CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF -SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS -INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN -CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) -ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE -POSSIBILITY OF SUCH DAMAGE. - -***************************************************************************/ - -#ifndef _I40E_VIRTCHNL_H_ -#define _I40E_VIRTCHNL_H_ - -#include "i40e_type.h" - -/* Description: - * This header file describes the VF-PF communication protocol used - * by the various i40e drivers. - * - * Admin queue buffer usage: - * desc->opcode is always i40e_aqc_opc_send_msg_to_pf - * flags, retval, datalen, and data addr are all used normally. - * Firmware copies the cookie fields when sending messages between the PF and - * VF, but uses all other fields internally. Due to this limitation, we - * must send all messages as "indirect", i.e. using an external buffer. - * - * All the vsi indexes are relative to the VF. Each VF can have maximum of - * three VSIs. All the queue indexes are relative to the VSI. Each VF can - * have a maximum of sixteen queues for all of its VSIs. - * - * The PF is required to return a status code in v_retval for all messages - * except RESET_VF, which does not require any response. The return value is of - * i40e_status_code type, defined in the i40e_type.h. - * - * In general, VF driver initialization should roughly follow the order of these - * opcodes. The VF driver must first validate the API version of the PF driver, - * then request a reset, then get resources, then configure queues and - * interrupts. After these operations are complete, the VF driver may start - * its queues, optionally add MAC and VLAN filters, and process traffic. - */ - -/* Opcodes for VF-PF communication. These are placed in the v_opcode field - * of the virtchnl_msg structure. - */ -enum i40e_virtchnl_ops { -/* The PF sends status change events to VFs using - * the I40E_VIRTCHNL_OP_EVENT opcode. - * VFs send requests to the PF using the other ops. - */ - I40E_VIRTCHNL_OP_UNKNOWN = 0, - I40E_VIRTCHNL_OP_VERSION = 1, /* must ALWAYS be 1 */ - I40E_VIRTCHNL_OP_RESET_VF = 2, - I40E_VIRTCHNL_OP_GET_VF_RESOURCES = 3, - I40E_VIRTCHNL_OP_CONFIG_TX_QUEUE = 4, - I40E_VIRTCHNL_OP_CONFIG_RX_QUEUE = 5, - I40E_VIRTCHNL_OP_CONFIG_VSI_QUEUES = 6, - I40E_VIRTCHNL_OP_CONFIG_IRQ_MAP = 7, - I40E_VIRTCHNL_OP_ENABLE_QUEUES = 8, - I40E_VIRTCHNL_OP_DISABLE_QUEUES = 9, - I40E_VIRTCHNL_OP_ADD_ETHER_ADDRESS = 10, - I40E_VIRTCHNL_OP_DEL_ETHER_ADDRESS = 11, - I40E_VIRTCHNL_OP_ADD_VLAN = 12, - I40E_VIRTCHNL_OP_DEL_VLAN = 13, - I40E_VIRTCHNL_OP_CONFIG_PROMISCUOUS_MODE = 14, - I40E_VIRTCHNL_OP_GET_STATS = 15, - I40E_VIRTCHNL_OP_FCOE = 16, - I40E_VIRTCHNL_OP_EVENT = 17, /* must ALWAYS be 17 */ -#ifdef I40E_SOL_VF_SUPPORT - I40E_VIRTCHNL_OP_GET_ADDNL_SOL_CONFIG = 19, -#endif - I40E_VIRTCHNL_OP_CONFIG_RSS_KEY = 23, - I40E_VIRTCHNL_OP_CONFIG_RSS_LUT = 24, - I40E_VIRTCHNL_OP_GET_RSS_HENA_CAPS = 25, - I40E_VIRTCHNL_OP_SET_RSS_HENA = 26, - -}; - -/* Virtual channel message descriptor. This overlays the admin queue - * descriptor. All other data is passed in external buffers. - */ - -struct i40e_virtchnl_msg { - u8 pad[8]; /* AQ flags/opcode/len/retval fields */ - enum i40e_virtchnl_ops v_opcode; /* avoid confusion with desc->opcode */ - enum i40e_status_code v_retval; /* ditto for desc->retval */ - u32 vfid; /* used by PF when sending to VF */ -}; - -/* Message descriptions and data structures.*/ - -/* I40E_VIRTCHNL_OP_VERSION - * VF posts its version number to the PF. PF responds with its version number - * in the same format, along with a return code. - * Reply from PF has its major/minor versions also in param0 and param1. - * If there is a major version mismatch, then the VF cannot operate. - * If there is a minor version mismatch, then the VF can operate but should - * add a warning to the system log. - * - * This enum element MUST always be specified as == 1, regardless of other - * changes in the API. The PF must always respond to this message without - * error regardless of version mismatch. - */ -#define I40E_VIRTCHNL_VERSION_MAJOR 1 -#define I40E_VIRTCHNL_VERSION_MINOR 1 -#define I40E_VIRTCHNL_VERSION_MINOR_NO_VF_CAPS 0 - -struct i40e_virtchnl_version_info { - u32 major; - u32 minor; -}; - -/* I40E_VIRTCHNL_OP_RESET_VF - * VF sends this request to PF with no parameters - * PF does NOT respond! VF driver must delay then poll VFGEN_RSTAT register - * until reset completion is indicated. The admin queue must be reinitialized - * after this operation. - * - * When reset is complete, PF must ensure that all queues in all VSIs associated - * with the VF are stopped, all queue configurations in the HMC are set to 0, - * and all MAC and VLAN filters (except the default MAC address) on all VSIs - * are cleared. - */ - -/* I40E_VIRTCHNL_OP_GET_VF_RESOURCES - * Version 1.0 VF sends this request to PF with no parameters - * Version 1.1 VF sends this request to PF with u32 bitmap of its capabilities - * PF responds with an indirect message containing - * i40e_virtchnl_vf_resource and one or more - * i40e_virtchnl_vsi_resource structures. - */ - -struct i40e_virtchnl_vsi_resource { - u16 vsi_id; - u16 num_queue_pairs; - enum i40e_vsi_type vsi_type; - u16 qset_handle; - u8 default_mac_addr[I40E_ETH_LENGTH_OF_ADDRESS]; -}; -/* VF offload flags */ -#define I40E_VIRTCHNL_VF_OFFLOAD_L2 0x00000001 -#define I40E_VIRTCHNL_VF_OFFLOAD_IWARP 0x00000002 -#define I40E_VIRTCHNL_VF_OFFLOAD_FCOE 0x00000004 -#define I40E_VIRTCHNL_VF_OFFLOAD_RSS_AQ 0x00000008 -#define I40E_VIRTCHNL_VF_OFFLOAD_RSS_REG 0x00000010 -#define I40E_VIRTCHNL_VF_OFFLOAD_WB_ON_ITR 0x00000020 -#define I40E_VIRTCHNL_VF_OFFLOAD_VLAN 0x00010000 -#define I40E_VIRTCHNL_VF_OFFLOAD_RX_POLLING 0x00020000 -#define I40E_VIRTCHNL_VF_OFFLOAD_RSS_PCTYPE_V2 0x00040000 -#define I40E_VIRTCHNL_VF_OFFLOAD_RSS_PF 0X00080000 -#define I40E_VIRTCHNL_VF_OFFLOAD_ENCAP 0X00100000 -#define I40E_VIRTCHNL_VF_OFFLOAD_ENCAP_CSUM 0X00200000 -#define I40E_VIRTCHNL_VF_OFFLOAD_RX_ENCAP_CSUM 0X00400000 - -#define I40E_VF_BASE_MODE_OFFLOADS (I40E_VIRTCHNL_VF_OFFLOAD_L2 | \ - I40E_VIRTCHNL_VF_OFFLOAD_VLAN | \ - I40E_VIRTCHNL_VF_OFFLOAD_RSS_PF) - -struct i40e_virtchnl_vf_resource { - u16 num_vsis; - u16 num_queue_pairs; - u16 max_vectors; - u16 max_mtu; - - u32 vf_offload_flags; - u32 rss_key_size; - u32 rss_lut_size; - - struct i40e_virtchnl_vsi_resource vsi_res[1]; -}; - -/* I40E_VIRTCHNL_OP_CONFIG_TX_QUEUE - * VF sends this message to set up parameters for one TX queue. - * External data buffer contains one instance of i40e_virtchnl_txq_info. - * PF configures requested queue and returns a status code. - */ - -/* Tx queue config info */ -struct i40e_virtchnl_txq_info { - u16 vsi_id; - u16 queue_id; - u16 ring_len; /* number of descriptors, multiple of 8 */ - u16 headwb_enabled; - u64 dma_ring_addr; - u64 dma_headwb_addr; -}; - -/* I40E_VIRTCHNL_OP_CONFIG_RX_QUEUE - * VF sends this message to set up parameters for one RX queue. - * External data buffer contains one instance of i40e_virtchnl_rxq_info. - * PF configures requested queue and returns a status code. - */ - -/* Rx queue config info */ -struct i40e_virtchnl_rxq_info { - u16 vsi_id; - u16 queue_id; - u32 ring_len; /* number of descriptors, multiple of 32 */ - u16 hdr_size; - u16 splithdr_enabled; - u32 databuffer_size; - u32 max_pkt_size; - u64 dma_ring_addr; - enum i40e_hmc_obj_rx_hsplit_0 rx_split_pos; -}; - -/* I40E_VIRTCHNL_OP_CONFIG_VSI_QUEUES - * VF sends this message to set parameters for all active TX and RX queues - * associated with the specified VSI. - * PF configures queues and returns status. - * If the number of queues specified is greater than the number of queues - * associated with the VSI, an error is returned and no queues are configured. - */ -struct i40e_virtchnl_queue_pair_info { - /* NOTE: vsi_id and queue_id should be identical for both queues. */ - struct i40e_virtchnl_txq_info txq; - struct i40e_virtchnl_rxq_info rxq; -}; - -struct i40e_virtchnl_vsi_queue_config_info { - u16 vsi_id; - u16 num_queue_pairs; - struct i40e_virtchnl_queue_pair_info qpair[1]; -}; - -/* I40E_VIRTCHNL_OP_CONFIG_IRQ_MAP - * VF uses this message to map vectors to queues. - * The rxq_map and txq_map fields are bitmaps used to indicate which queues - * are to be associated with the specified vector. - * The "other" causes are always mapped to vector 0. - * PF configures interrupt mapping and returns status. - */ -struct i40e_virtchnl_vector_map { - u16 vsi_id; - u16 vector_id; - u16 rxq_map; - u16 txq_map; - u16 rxitr_idx; - u16 txitr_idx; -}; - -struct i40e_virtchnl_irq_map_info { - u16 num_vectors; - struct i40e_virtchnl_vector_map vecmap[1]; -}; - -/* I40E_VIRTCHNL_OP_ENABLE_QUEUES - * I40E_VIRTCHNL_OP_DISABLE_QUEUES - * VF sends these message to enable or disable TX/RX queue pairs. - * The queues fields are bitmaps indicating which queues to act upon. - * (Currently, we only support 16 queues per VF, but we make the field - * u32 to allow for expansion.) - * PF performs requested action and returns status. - */ -struct i40e_virtchnl_queue_select { - u16 vsi_id; - u16 pad; - u32 rx_queues; - u32 tx_queues; -}; - -/* I40E_VIRTCHNL_OP_ADD_ETHER_ADDRESS - * VF sends this message in order to add one or more unicast or multicast - * address filters for the specified VSI. - * PF adds the filters and returns status. - */ - -/* I40E_VIRTCHNL_OP_DEL_ETHER_ADDRESS - * VF sends this message in order to remove one or more unicast or multicast - * filters for the specified VSI. - * PF removes the filters and returns status. - */ - -struct i40e_virtchnl_ether_addr { - u8 addr[I40E_ETH_LENGTH_OF_ADDRESS]; - u8 pad[2]; -}; - -struct i40e_virtchnl_ether_addr_list { - u16 vsi_id; - u16 num_elements; - struct i40e_virtchnl_ether_addr list[1]; -}; - -#ifdef I40E_SOL_VF_SUPPORT -/* I40E_VIRTCHNL_OP_GET_ADDNL_SOL_CONFIG - * VF sends this message to get the default MTU and list of additional ethernet - * addresses it is allowed to use. - * PF responds with an indirect message containing - * i40e_virtchnl_addnl_solaris_config with zero or more - * i40e_virtchnl_ether_addr structures. - * - * It is expected that this operation will only ever be needed for Solaris VFs - * running under a Solaris PF. - */ -struct i40e_virtchnl_addnl_solaris_config { - u16 default_mtu; - struct i40e_virtchnl_ether_addr_list al; -}; - -#endif -/* I40E_VIRTCHNL_OP_ADD_VLAN - * VF sends this message to add one or more VLAN tag filters for receives. - * PF adds the filters and returns status. - * If a port VLAN is configured by the PF, this operation will return an - * error to the VF. - */ - -/* I40E_VIRTCHNL_OP_DEL_VLAN - * VF sends this message to remove one or more VLAN tag filters for receives. - * PF removes the filters and returns status. - * If a port VLAN is configured by the PF, this operation will return an - * error to the VF. - */ - -struct i40e_virtchnl_vlan_filter_list { - u16 vsi_id; - u16 num_elements; - u16 vlan_id[1]; -}; - -/* I40E_VIRTCHNL_OP_CONFIG_PROMISCUOUS_MODE - * VF sends VSI id and flags. - * PF returns status code in retval. - * Note: we assume that broadcast accept mode is always enabled. - */ -struct i40e_virtchnl_promisc_info { - u16 vsi_id; - u16 flags; -}; - -#define I40E_FLAG_VF_UNICAST_PROMISC 0x00000001 -#define I40E_FLAG_VF_MULTICAST_PROMISC 0x00000002 - -/* I40E_VIRTCHNL_OP_GET_STATS - * VF sends this message to request stats for the selected VSI. VF uses - * the i40e_virtchnl_queue_select struct to specify the VSI. The queue_id - * field is ignored by the PF. - * - * PF replies with struct i40e_eth_stats in an external buffer. - */ - -/* I40E_VIRTCHNL_OP_CONFIG_RSS_KEY - * I40E_VIRTCHNL_OP_CONFIG_RSS_LUT - * VF sends these messages to configure RSS. Only supported if both PF - * and VF drivers set the I40E_VIRTCHNL_VF_OFFLOAD_RSS_PF bit during - * configuration negotiation. If this is the case, then the rss fields in - * the vf resource struct are valid. - * Both the key and LUT are initialized to 0 by the PF, meaning that - * RSS is effectively disabled until set up by the VF. - */ -struct i40e_virtchnl_rss_key { - u16 vsi_id; - u16 key_len; - u8 key[1]; /* RSS hash key, packed bytes */ -}; - -struct i40e_virtchnl_rss_lut { - u16 vsi_id; - u16 lut_entries; - u8 lut[1]; /* RSS lookup table*/ -}; - -/* I40E_VIRTCHNL_OP_GET_RSS_HENA_CAPS - * I40E_VIRTCHNL_OP_SET_RSS_HENA - * VF sends these messages to get and set the hash filter enable bits for RSS. - * By default, the PF sets these to all possible traffic types that the - * hardware supports. The VF can query this value if it wants to change the - * traffic types that are hashed by the hardware. - * Traffic types are defined in the i40e_filter_pctype enum in i40e_type.h - */ -struct i40e_virtchnl_rss_hena { - u64 hena; -}; - -/* I40E_VIRTCHNL_OP_EVENT - * PF sends this message to inform the VF driver of events that may affect it. - * No direct response is expected from the VF, though it may generate other - * messages in response to this one. - */ -enum i40e_virtchnl_event_codes { - I40E_VIRTCHNL_EVENT_UNKNOWN = 0, - I40E_VIRTCHNL_EVENT_LINK_CHANGE, - I40E_VIRTCHNL_EVENT_RESET_IMPENDING, - I40E_VIRTCHNL_EVENT_PF_DRIVER_CLOSE, -}; -#define I40E_PF_EVENT_SEVERITY_INFO 0 -#define I40E_PF_EVENT_SEVERITY_ATTENTION 1 -#define I40E_PF_EVENT_SEVERITY_ACTION_REQUIRED 2 -#define I40E_PF_EVENT_SEVERITY_CERTAIN_DOOM 255 - -struct i40e_virtchnl_pf_event { - enum i40e_virtchnl_event_codes event; - union { - struct { - enum i40e_aq_link_speed link_speed; - bool link_status; - } link_event; - } event_data; - - int severity; -}; - -/* VF reset states - these are written into the RSTAT register: - * I40E_VFGEN_RSTAT1 on the PF - * I40E_VFGEN_RSTAT on the VF - * When the PF initiates a reset, it writes 0 - * When the reset is complete, it writes 1 - * When the PF detects that the VF has recovered, it writes 2 - * VF checks this register periodically to determine if a reset has occurred, - * then polls it to know when the reset is complete. - * If either the PF or VF reads the register while the hardware - * is in a reset state, it will return DEADBEEF, which, when masked - * will result in 3. - */ -enum i40e_vfr_states { - I40E_VFR_INPROGRESS = 0, - I40E_VFR_COMPLETED, - I40E_VFR_VFACTIVE, - I40E_VFR_UNKNOWN, -}; - -#endif /* _I40E_VIRTCHNL_H_ */ diff --git a/drivers/net/i40e/base/virtchnl.h b/drivers/net/i40e/base/virtchnl.h new file mode 100644 index 0000000..f00dd36 --- /dev/null +++ b/drivers/net/i40e/base/virtchnl.h @@ -0,0 +1,772 @@ +/******************************************************************************* + +Copyright (c) 2013 - 2015, Intel Corporation +All rights reserved. + +Redistribution and use in source and binary forms, with or without +modification, are permitted provided that the following conditions are met: + + 1. Redistributions of source code must retain the above copyright notice, + this list of conditions and the following disclaimer. + + 2. Redistributions in binary form must reproduce the above copyright + notice, this list of conditions and the following disclaimer in the + documentation and/or other materials provided with the distribution. + + 3. Neither the name of the Intel Corporation nor the names of its + contributors may be used to endorse or promote products derived from + this software without specific prior written permission. + +THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS" +AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE +IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE +ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE +LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR +CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF +SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS +INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN +CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) +ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE +POSSIBILITY OF SUCH DAMAGE. + +***************************************************************************/ + +#ifndef _VIRTCHNL_H_ +#define _VIRTCHNL_H_ + +/* Description: + * This header file describes the VF-PF communication protocol used + * by the drivers for all devices starting from our 40G product line + * + * Admin queue buffer usage: + * desc->opcode is always aqc_opc_send_msg_to_pf + * flags, retval, datalen, and data addr are all used normally. + * The Firmware copies the cookie fields when sending messages between the + * PF and VF, but uses all other fields internally. Due to this limitation, + * we must send all messages as "indirect", i.e. using an external buffer. + * + * All the VSI indexes are relative to the VF. Each VF can have maximum of + * three VSIs. All the queue indexes are relative to the VSI. Each VF can + * have a maximum of sixteen queues for all of its VSIs. + * + * The PF is required to return a status code in v_retval for all messages + * except RESET_VF, which does not require any response. The return value + * is of status_code type, defined in the shared type.h. + * + * In general, VF driver initialization should roughly follow the order of + * these opcodes. The VF driver must first validate the API version of the + * PF driver, then request a reset, then get resources, then configure + * queues and interrupts. After these operations are complete, the VF + * driver may start its queues, optionally add MAC and VLAN filters, and + * process traffic. + */ + +/* START GENERIC DEFINES + * Need to ensure the following enums and defines hold the same meaning and + * value in current and future projects + */ + +/* Error Codes */ +enum virtchnl_status_code { + VIRTCHNL_STATUS_SUCCESS = 0, + VIRTCHNL_ERR_PARAM = -5, + VIRTCHNL_STATUS_ERR_OPCODE_MISMATCH = -38, + VIRTCHNL_STATUS_ERR_CQP_COMPL_ERROR = -39, + VIRTCHNL_STATUS_ERR_INVALID_VF_ID = -40, + VIRTCHNL_STATUS_NOT_SUPPORTED = -64, +}; + +#define VIRTCHNL_LINK_SPEED_100MB_SHIFT 0x1 +#define VIRTCHNL_LINK_SPEED_1000MB_SHIFT 0x2 +#define VIRTCHNL_LINK_SPEED_10GB_SHIFT 0x3 +#define VIRTCHNL_LINK_SPEED_40GB_SHIFT 0x4 +#define VIRTCHNL_LINK_SPEED_20GB_SHIFT 0x5 +#define VIRTCHNL_LINK_SPEED_25GB_SHIFT 0x6 + +enum virtchnl_link_speed { + VIRTCHNL_LINK_SPEED_UNKNOWN = 0, + VIRTCHNL_LINK_SPEED_100MB = BIT(VIRTCHNL_LINK_SPEED_100MB_SHIFT), + VIRTCHNL_LINK_SPEED_1GB = BIT(VIRTCHNL_LINK_SPEED_1000MB_SHIFT), + VIRTCHNL_LINK_SPEED_10GB = BIT(VIRTCHNL_LINK_SPEED_10GB_SHIFT), + VIRTCHNL_LINK_SPEED_40GB = BIT(VIRTCHNL_LINK_SPEED_40GB_SHIFT), + VIRTCHNL_LINK_SPEED_20GB = BIT(VIRTCHNL_LINK_SPEED_20GB_SHIFT), + VIRTCHNL_LINK_SPEED_25GB = BIT(VIRTCHNL_LINK_SPEED_25GB_SHIFT), +}; + +/* for hsplit_0 field of Rx HMC context */ +/* deprecated with AVF 1.0 */ +enum virtchnl_rx_hsplit { + VIRTCHNL_RX_HSPLIT_NO_SPLIT = 0, + VIRTCHNL_RX_HSPLIT_SPLIT_L2 = 1, + VIRTCHNL_RX_HSPLIT_SPLIT_IP = 2, + VIRTCHNL_RX_HSPLIT_SPLIT_TCP_UDP = 4, + VIRTCHNL_RX_HSPLIT_SPLIT_SCTP = 8, +}; + +#define VIRTCHNL_ETH_LENGTH_OF_ADDRESS 6 +/* END GENERIC DEFINES */ + +/* Opcodes for VF-PF communication. These are placed in the v_opcode field + * of the virtchnl_msg structure. + */ +enum virtchnl_ops { +/* The PF sends status change events to VFs using + * the VIRTCHNL_OP_EVENT opcode. + * VFs send requests to the PF using the other ops. + * Use of "advanced opcode" features must be negotiated as part of capabilities + * exchange and are not considered part of base mode feature set. + */ + VIRTCHNL_OP_UNKNOWN = 0, + VIRTCHNL_OP_VERSION = 1, /* must ALWAYS be 1 */ + VIRTCHNL_OP_RESET_VF = 2, + VIRTCHNL_OP_GET_VF_RESOURCES = 3, + VIRTCHNL_OP_CONFIG_TX_QUEUE = 4, + VIRTCHNL_OP_CONFIG_RX_QUEUE = 5, + VIRTCHNL_OP_CONFIG_VSI_QUEUES = 6, + VIRTCHNL_OP_CONFIG_IRQ_MAP = 7, + VIRTCHNL_OP_ENABLE_QUEUES = 8, + VIRTCHNL_OP_DISABLE_QUEUES = 9, + VIRTCHNL_OP_ADD_ETH_ADDR = 10, + VIRTCHNL_OP_DEL_ETH_ADDR = 11, + VIRTCHNL_OP_ADD_VLAN = 12, + VIRTCHNL_OP_DEL_VLAN = 13, + VIRTCHNL_OP_CONFIG_PROMISCUOUS_MODE = 14, + VIRTCHNL_OP_GET_STATS = 15, + VIRTCHNL_OP_RSVD = 16, + VIRTCHNL_OP_EVENT = 17, /* must ALWAYS be 17 */ +#ifdef VIRTCHNL_SOL_VF_SUPPORT + VIRTCHNL_OP_GET_ADDNL_SOL_CONFIG = 19, +#endif +#ifdef VIRTCHNL_IWARP + VIRTCHNL_OP_IWARP = 20, /* advanced opcode */ + VIRTCHNL_OP_CONFIG_IWARP_IRQ_MAP = 21, /* advanced opcode */ + VIRTCHNL_OP_RELEASE_IWARP_IRQ_MAP = 22, /* advanced opcode */ +#endif + VIRTCHNL_OP_CONFIG_RSS_KEY = 23, + VIRTCHNL_OP_CONFIG_RSS_LUT = 24, + VIRTCHNL_OP_GET_RSS_HENA_CAPS = 25, + VIRTCHNL_OP_SET_RSS_HENA = 26, + VIRTCHNL_OP_ENABLE_VLAN_STRIPPING = 27, + VIRTCHNL_OP_DISABLE_VLAN_STRIPPING = 28, + VIRTCHNL_OP_REQUEST_QUEUES = 29, + +}; + +/* This macro is used to generate a compilation error if a structure + * is not exactly the correct length. It gives a divide by zero error if the + * structure is not of the correct size, otherwise it creates an enum that is + * never used. + */ +#define VIRTCHNL_CHECK_STRUCT_LEN(n, X) enum virtchnl_static_assert_enum_##X \ + {virtchnl_static_assert_##X = (n) / ((sizeof(struct X) == (n)) ? 1 : 0)} + +/* Virtual channel message descriptor. This overlays the admin queue + * descriptor. All other data is passed in external buffers. + */ + +struct virtchnl_msg { + u8 pad[8]; /* AQ flags/opcode/len/retval fields */ + enum virtchnl_ops v_opcode; /* avoid confusion with desc->opcode */ + enum virtchnl_status_code v_retval; /* ditto for desc->retval */ + u32 vfid; /* used by PF when sending to VF */ +}; + +VIRTCHNL_CHECK_STRUCT_LEN(20, virtchnl_msg); + +/* Message descriptions and data structures.*/ + +/* VIRTCHNL_OP_VERSION + * VF posts its version number to the PF. PF responds with its version number + * in the same format, along with a return code. + * Reply from PF has its major/minor versions also in param0 and param1. + * If there is a major version mismatch, then the VF cannot operate. + * If there is a minor version mismatch, then the VF can operate but should + * add a warning to the system log. + * + * This enum element MUST always be specified as == 1, regardless of other + * changes in the API. The PF must always respond to this message without + * error regardless of version mismatch. + */ +#define VIRTCHNL_VERSION_MAJOR 1 +#define VIRTCHNL_VERSION_MINOR 1 +#define VIRTCHNL_VERSION_MINOR_NO_VF_CAPS 0 + +struct virtchnl_version_info { + u32 major; + u32 minor; +}; + +VIRTCHNL_CHECK_STRUCT_LEN(8, virtchnl_version_info); + +#define VF_IS_V10(_v) (((_v)->major == 1) && ((_v)->minor == 0)) +#define VF_IS_V11(_ver) (((_ver)->major == 1) && ((_ver)->minor == 1)) + +/* VIRTCHNL_OP_RESET_VF + * VF sends this request to PF with no parameters + * PF does NOT respond! VF driver must delay then poll VFGEN_RSTAT register + * until reset completion is indicated. The admin queue must be reinitialized + * after this operation. + * + * When reset is complete, PF must ensure that all queues in all VSIs associated + * with the VF are stopped, all queue configurations in the HMC are set to 0, + * and all MAC and VLAN filters (except the default MAC address) on all VSIs + * are cleared. + */ + +/* VSI types that use VIRTCHNL interface for VF-PF communication. VSI_SRIOV + * vsi_type should always be 6 for backward compatibility. Add other fields + * as needed. + */ +enum virtchnl_vsi_type { + VIRTCHNL_VSI_TYPE_INVALID = 0, + VIRTCHNL_VSI_SRIOV = 6, +}; + +/* VIRTCHNL_OP_GET_VF_RESOURCES + * Version 1.0 VF sends this request to PF with no parameters + * Version 1.1 VF sends this request to PF with u32 bitmap of its capabilities + * PF responds with an indirect message containing + * virtchnl_vf_resource and one or more + * virtchnl_vsi_resource structures. + */ + +struct virtchnl_vsi_resource { + u16 vsi_id; + u16 num_queue_pairs; + enum virtchnl_vsi_type vsi_type; + u16 qset_handle; + u8 default_mac_addr[VIRTCHNL_ETH_LENGTH_OF_ADDRESS]; +}; + +VIRTCHNL_CHECK_STRUCT_LEN(16, virtchnl_vsi_resource); + +/* VF offload flags + * VIRTCHNL_VF_OFFLOAD_L2 flag is inclusive of base mode L2 offloads including + * TX/RX Checksum offloading and TSO for non-tunnelled packets. + */ +#define VIRTCHNL_VF_OFFLOAD_L2 0x00000001 +#define VIRTCHNL_VF_OFFLOAD_IWARP 0x00000002 +#define VIRTCHNL_VF_OFFLOAD_RSVD 0x00000004 +#define VIRTCHNL_VF_OFFLOAD_RSS_AQ 0x00000008 +#define VIRTCHNL_VF_OFFLOAD_RSS_REG 0x00000010 +#define VIRTCHNL_VF_OFFLOAD_WB_ON_ITR 0x00000020 +#define VIRTCHNL_VF_OFFLOAD_REQ_QUEUES 0x00000040 +#define VIRTCHNL_VF_OFFLOAD_VLAN 0x00010000 +#define VIRTCHNL_VF_OFFLOAD_RX_POLLING 0x00020000 +#define VIRTCHNL_VF_OFFLOAD_RSS_PCTYPE_V2 0x00040000 +#define VIRTCHNL_VF_OFFLOAD_RSS_PF 0X00080000 +#define VIRTCHNL_VF_OFFLOAD_ENCAP 0X00100000 +#define VIRTCHNL_VF_OFFLOAD_ENCAP_CSUM 0X00200000 +#define VIRTCHNL_VF_OFFLOAD_RX_ENCAP_CSUM 0X00400000 + +#define VF_BASE_MODE_OFFLOADS (VIRTCHNL_VF_OFFLOAD_L2 | \ + VIRTCHNL_VF_OFFLOAD_VLAN | \ + VIRTCHNL_VF_OFFLOAD_RSS_PF) + +struct virtchnl_vf_resource { + u16 num_vsis; + u16 num_queue_pairs; + u16 max_vectors; + u16 max_mtu; + + u32 vf_offload_flags; + u32 rss_key_size; + u32 rss_lut_size; + + struct virtchnl_vsi_resource vsi_res[1]; +}; + +VIRTCHNL_CHECK_STRUCT_LEN(36, virtchnl_vf_resource); + +/* VIRTCHNL_OP_CONFIG_TX_QUEUE + * VF sends this message to set up parameters for one TX queue. + * External data buffer contains one instance of virtchnl_txq_info. + * PF configures requested queue and returns a status code. + */ + +/* Tx queue config info */ +struct virtchnl_txq_info { + u16 vsi_id; + u16 queue_id; + u16 ring_len; /* number of descriptors, multiple of 8 */ + u16 headwb_enabled; /* deprecated with AVF 1.0 */ + u64 dma_ring_addr; + u64 dma_headwb_addr; /* deprecated with AVF 1.0 */ +}; + +VIRTCHNL_CHECK_STRUCT_LEN(24, virtchnl_txq_info); + +/* VIRTCHNL_OP_CONFIG_RX_QUEUE + * VF sends this message to set up parameters for one RX queue. + * External data buffer contains one instance of virtchnl_rxq_info. + * PF configures requested queue and returns a status code. + */ + +/* Rx queue config info */ +struct virtchnl_rxq_info { + u16 vsi_id; + u16 queue_id; + u32 ring_len; /* number of descriptors, multiple of 32 */ + u16 hdr_size; + u16 splithdr_enabled; /* deprecated with AVF 1.0 */ + u32 databuffer_size; + u32 max_pkt_size; + u32 pad1; + u64 dma_ring_addr; + enum virtchnl_rx_hsplit rx_split_pos; /* deprecated with AVF 1.0 */ + u32 pad2; +}; + +VIRTCHNL_CHECK_STRUCT_LEN(40, virtchnl_rxq_info); + +/* VIRTCHNL_OP_CONFIG_VSI_QUEUES + * VF sends this message to set parameters for all active TX and RX queues + * associated with the specified VSI. + * PF configures queues and returns status. + * If the number of queues specified is greater than the number of queues + * associated with the VSI, an error is returned and no queues are configured. + */ +struct virtchnl_queue_pair_info { + /* NOTE: vsi_id and queue_id should be identical for both queues. */ + struct virtchnl_txq_info txq; + struct virtchnl_rxq_info rxq; +}; + +VIRTCHNL_CHECK_STRUCT_LEN(64, virtchnl_queue_pair_info); + +struct virtchnl_vsi_queue_config_info { + u16 vsi_id; + u16 num_queue_pairs; + u32 pad; + struct virtchnl_queue_pair_info qpair[1]; +}; + +VIRTCHNL_CHECK_STRUCT_LEN(72, virtchnl_vsi_queue_config_info); + +/* VIRTCHNL_OP_REQUEST_QUEUES + * VF sends this message to request the PF to allocate additional queues to + * this VF. Each VF gets a guaranteed number of queues on init but asking for + * additional queues must be negotiated. This is a best effort request as it + * is possible the PF does not have enough queues left to support the request. + * If the PF cannot support the number requested it will respond with the + * maximum number it is able to support; otherwise it will respond with the + * number requested. + */ + +/* VF resource request */ +struct virtchnl_vf_res_request { + u16 num_queue_pairs; +}; + +/* VIRTCHNL_OP_CONFIG_IRQ_MAP + * VF uses this message to map vectors to queues. + * The rxq_map and txq_map fields are bitmaps used to indicate which queues + * are to be associated with the specified vector. + * The "other" causes are always mapped to vector 0. + * PF configures interrupt mapping and returns status. + */ +struct virtchnl_vector_map { + u16 vsi_id; + u16 vector_id; + u16 rxq_map; + u16 txq_map; + u16 rxitr_idx; + u16 txitr_idx; +}; + +VIRTCHNL_CHECK_STRUCT_LEN(12, virtchnl_vector_map); + +struct virtchnl_irq_map_info { + u16 num_vectors; + struct virtchnl_vector_map vecmap[1]; +}; + +VIRTCHNL_CHECK_STRUCT_LEN(14, virtchnl_irq_map_info); + +/* VIRTCHNL_OP_ENABLE_QUEUES + * VIRTCHNL_OP_DISABLE_QUEUES + * VF sends these message to enable or disable TX/RX queue pairs. + * The queues fields are bitmaps indicating which queues to act upon. + * (Currently, we only support 16 queues per VF, but we make the field + * u32 to allow for expansion.) + * PF performs requested action and returns status. + */ +struct virtchnl_queue_select { + u16 vsi_id; + u16 pad; + u32 rx_queues; + u32 tx_queues; +}; + +VIRTCHNL_CHECK_STRUCT_LEN(12, virtchnl_queue_select); + +/* VIRTCHNL_OP_ADD_ETH_ADDR + * VF sends this message in order to add one or more unicast or multicast + * address filters for the specified VSI. + * PF adds the filters and returns status. + */ + +/* VIRTCHNL_OP_DEL_ETH_ADDR + * VF sends this message in order to remove one or more unicast or multicast + * filters for the specified VSI. + * PF removes the filters and returns status. + */ + +struct virtchnl_ether_addr { + u8 addr[VIRTCHNL_ETH_LENGTH_OF_ADDRESS]; + u8 pad[2]; +}; + +VIRTCHNL_CHECK_STRUCT_LEN(8, virtchnl_ether_addr); + +struct virtchnl_ether_addr_list { + u16 vsi_id; + u16 num_elements; + struct virtchnl_ether_addr list[1]; +}; + +VIRTCHNL_CHECK_STRUCT_LEN(12, virtchnl_ether_addr_list); + +#ifdef VIRTCHNL_SOL_VF_SUPPORT +/* VIRTCHNL_OP_GET_ADDNL_SOL_CONFIG + * VF sends this message to get the default MTU and list of additional ethernet + * addresses it is allowed to use. + * PF responds with an indirect message containing + * virtchnl_addnl_solaris_config with zero or more + * virtchnl_ether_addr structures. + * + * It is expected that this operation will only ever be needed for Solaris VFs + * running under a Solaris PF. + */ +struct virtchnl_addnl_solaris_config { + u16 default_mtu; + struct virtchnl_ether_addr_list al; +}; + +#endif +/* VIRTCHNL_OP_ADD_VLAN + * VF sends this message to add one or more VLAN tag filters for receives. + * PF adds the filters and returns status. + * If a port VLAN is configured by the PF, this operation will return an + * error to the VF. + */ + +/* VIRTCHNL_OP_DEL_VLAN + * VF sends this message to remove one or more VLAN tag filters for receives. + * PF removes the filters and returns status. + * If a port VLAN is configured by the PF, this operation will return an + * error to the VF. + */ + +struct virtchnl_vlan_filter_list { + u16 vsi_id; + u16 num_elements; + u16 vlan_id[1]; +}; + +VIRTCHNL_CHECK_STRUCT_LEN(6, virtchnl_vlan_filter_list); + +/* VIRTCHNL_OP_CONFIG_PROMISCUOUS_MODE + * VF sends VSI id and flags. + * PF returns status code in retval. + * Note: we assume that broadcast accept mode is always enabled. + */ +struct virtchnl_promisc_info { + u16 vsi_id; + u16 flags; +}; + +VIRTCHNL_CHECK_STRUCT_LEN(4, virtchnl_promisc_info); + +#define FLAG_VF_UNICAST_PROMISC 0x00000001 +#define FLAG_VF_MULTICAST_PROMISC 0x00000002 + +/* VIRTCHNL_OP_GET_STATS + * VF sends this message to request stats for the selected VSI. VF uses + * the virtchnl_queue_select struct to specify the VSI. The queue_id + * field is ignored by the PF. + * + * PF replies with struct eth_stats in an external buffer. + */ + +/* VIRTCHNL_OP_CONFIG_RSS_KEY + * VIRTCHNL_OP_CONFIG_RSS_LUT + * VF sends these messages to configure RSS. Only supported if both PF + * and VF drivers set the VIRTCHNL_VF_OFFLOAD_RSS_PF bit during + * configuration negotiation. If this is the case, then the RSS fields in + * the VF resource struct are valid. + * Both the key and LUT are initialized to 0 by the PF, meaning that + * RSS is effectively disabled until set up by the VF. + */ +struct virtchnl_rss_key { + u16 vsi_id; + u16 key_len; + u8 key[1]; /* RSS hash key, packed bytes */ +}; + +VIRTCHNL_CHECK_STRUCT_LEN(6, virtchnl_rss_key); + +struct virtchnl_rss_lut { + u16 vsi_id; + u16 lut_entries; + u8 lut[1]; /* RSS lookup table*/ +}; + +VIRTCHNL_CHECK_STRUCT_LEN(6, virtchnl_rss_lut); + +/* VIRTCHNL_OP_GET_RSS_HENA_CAPS + * VIRTCHNL_OP_SET_RSS_HENA + * VF sends these messages to get and set the hash filter enable bits for RSS. + * By default, the PF sets these to all possible traffic types that the + * hardware supports. The VF can query this value if it wants to change the + * traffic types that are hashed by the hardware. + */ +struct virtchnl_rss_hena { + u64 hena; +}; + +VIRTCHNL_CHECK_STRUCT_LEN(8, virtchnl_rss_hena); + +/* VIRTCHNL_OP_EVENT + * PF sends this message to inform the VF driver of events that may affect it. + * No direct response is expected from the VF, though it may generate other + * messages in response to this one. + */ +enum virtchnl_event_codes { + VIRTCHNL_EVENT_UNKNOWN = 0, + VIRTCHNL_EVENT_LINK_CHANGE, + VIRTCHNL_EVENT_RESET_IMPENDING, + VIRTCHNL_EVENT_PF_DRIVER_CLOSE, +}; + +#define PF_EVENT_SEVERITY_INFO 0 +#define PF_EVENT_SEVERITY_ATTENTION 1 +#define PF_EVENT_SEVERITY_ACTION_REQUIRED 2 +#define PF_EVENT_SEVERITY_CERTAIN_DOOM 255 + +struct virtchnl_pf_event { + enum virtchnl_event_codes event; + union { + struct { + enum virtchnl_link_speed link_speed; + bool link_status; + } link_event; + } event_data; + + int severity; +}; + +VIRTCHNL_CHECK_STRUCT_LEN(16, virtchnl_pf_event); + +#ifdef VIRTCHNL_IWARP + +/* VIRTCHNL_OP_CONFIG_IWARP_IRQ_MAP + * VF uses this message to request PF to map IWARP vectors to IWARP queues. + * The request for this originates from the VF IWARP driver through + * a client interface between VF LAN and VF IWARP driver. + * A vector could have an AEQ and CEQ attached to it although + * there is a single AEQ per VF IWARP instance in which case + * most vectors will have an INVALID_IDX for aeq and valid idx for ceq. + * There will never be a case where there will be multiple CEQs attached + * to a single vector. + * PF configures interrupt mapping and returns status. + */ + +/* HW does not define a type value for AEQ; only for RX/TX and CEQ. + * In order for us to keep the interface simple, SW will define a + * unique type value for AEQ. + */ +#define QUEUE_TYPE_PE_AEQ 0x80 +#define QUEUE_INVALID_IDX 0xFFFF + +struct virtchnl_iwarp_qv_info { + u32 v_idx; /* msix_vector */ + u16 ceq_idx; + u16 aeq_idx; + u8 itr_idx; +}; + +VIRTCHNL_CHECK_STRUCT_LEN(12, virtchnl_iwarp_qv_info); + +struct virtchnl_iwarp_qvlist_info { + u32 num_vectors; + struct virtchnl_iwarp_qv_info qv_info[1]; +}; + +VIRTCHNL_CHECK_STRUCT_LEN(16, virtchnl_iwarp_qvlist_info); + +#endif + +/* VF reset states - these are written into the RSTAT register: + * VFGEN_RSTAT on the VF + * When the PF initiates a reset, it writes 0 + * When the reset is complete, it writes 1 + * When the PF detects that the VF has recovered, it writes 2 + * VF checks this register periodically to determine if a reset has occurred, + * then polls it to know when the reset is complete. + * If either the PF or VF reads the register while the hardware + * is in a reset state, it will return DEADBEEF, which, when masked + * will result in 3. + */ +enum virtchnl_vfr_states { + VIRTCHNL_VFR_INPROGRESS = 0, + VIRTCHNL_VFR_COMPLETED, + VIRTCHNL_VFR_VFACTIVE, +}; + +/** + * virtchnl_vc_validate_vf_msg + * @ver: Virtchnl version info + * @v_opcode: Opcode for the message + * @msg: pointer to the msg buffer + * @msglen: msg length + * + * validate msg format against struct for each opcode + */ +static inline int +virtchnl_vc_validate_vf_msg(struct virtchnl_version_info *ver, u32 v_opcode, + u8 *msg, u16 msglen) +{ + bool err_msg_format = false; + int valid_len = 0; + + /* Validate message length. */ + switch (v_opcode) { + case VIRTCHNL_OP_VERSION: + valid_len = sizeof(struct virtchnl_version_info); + break; + case VIRTCHNL_OP_RESET_VF: + break; + case VIRTCHNL_OP_GET_VF_RESOURCES: + if (VF_IS_V11(ver)) + valid_len = sizeof(u32); + break; + case VIRTCHNL_OP_CONFIG_TX_QUEUE: + valid_len = sizeof(struct virtchnl_txq_info); + break; + case VIRTCHNL_OP_CONFIG_RX_QUEUE: + valid_len = sizeof(struct virtchnl_rxq_info); + break; + case VIRTCHNL_OP_CONFIG_VSI_QUEUES: + valid_len = sizeof(struct virtchnl_vsi_queue_config_info); + if (msglen >= valid_len) { + struct virtchnl_vsi_queue_config_info *vqc = + (struct virtchnl_vsi_queue_config_info *)msg; + valid_len += (vqc->num_queue_pairs * + sizeof(struct + virtchnl_queue_pair_info)); + if (vqc->num_queue_pairs == 0) + err_msg_format = true; + } + break; + case VIRTCHNL_OP_CONFIG_IRQ_MAP: + valid_len = sizeof(struct virtchnl_irq_map_info); + if (msglen >= valid_len) { + struct virtchnl_irq_map_info *vimi = + (struct virtchnl_irq_map_info *)msg; + valid_len += (vimi->num_vectors * + sizeof(struct virtchnl_vector_map)); + if (vimi->num_vectors == 0) + err_msg_format = true; + } + break; + case VIRTCHNL_OP_ENABLE_QUEUES: + case VIRTCHNL_OP_DISABLE_QUEUES: + valid_len = sizeof(struct virtchnl_queue_select); + break; + case VIRTCHNL_OP_ADD_ETH_ADDR: + case VIRTCHNL_OP_DEL_ETH_ADDR: + valid_len = sizeof(struct virtchnl_ether_addr_list); + if (msglen >= valid_len) { + struct virtchnl_ether_addr_list *veal = + (struct virtchnl_ether_addr_list *)msg; + valid_len += veal->num_elements * + sizeof(struct virtchnl_ether_addr); + if (veal->num_elements == 0) + err_msg_format = true; + } + break; + case VIRTCHNL_OP_ADD_VLAN: + case VIRTCHNL_OP_DEL_VLAN: + valid_len = sizeof(struct virtchnl_vlan_filter_list); + if (msglen >= valid_len) { + struct virtchnl_vlan_filter_list *vfl = + (struct virtchnl_vlan_filter_list *)msg; + valid_len += vfl->num_elements * sizeof(u16); + if (vfl->num_elements == 0) + err_msg_format = true; + } + break; + case VIRTCHNL_OP_CONFIG_PROMISCUOUS_MODE: + valid_len = sizeof(struct virtchnl_promisc_info); + break; + case VIRTCHNL_OP_GET_STATS: + valid_len = sizeof(struct virtchnl_queue_select); + break; +#ifdef VIRTCHNL_IWARP + case VIRTCHNL_OP_IWARP: + /* These messages are opaque to us and will be validated in + * the RDMA client code. We just need to check for nonzero + * length. The firmware will enforce max length restrictions. + */ + if (msglen) + valid_len = msglen; + else + err_msg_format = true; + break; + case VIRTCHNL_OP_RELEASE_IWARP_IRQ_MAP: + break; + case VIRTCHNL_OP_CONFIG_IWARP_IRQ_MAP: + valid_len = sizeof(struct virtchnl_iwarp_qvlist_info); + if (msglen >= valid_len) { + struct virtchnl_iwarp_qvlist_info *qv = + (struct virtchnl_iwarp_qvlist_info *)msg; + if (qv->num_vectors == 0) { + err_msg_format = true; + break; + } + valid_len += ((qv->num_vectors - 1) * + sizeof(struct virtchnl_iwarp_qv_info)); + } + break; +#endif + case VIRTCHNL_OP_CONFIG_RSS_KEY: + valid_len = sizeof(struct virtchnl_rss_key); + if (msglen >= valid_len) { + struct virtchnl_rss_key *vrk = + (struct virtchnl_rss_key *)msg; + valid_len += vrk->key_len - 1; + } + break; + case VIRTCHNL_OP_CONFIG_RSS_LUT: + valid_len = sizeof(struct virtchnl_rss_lut); + if (msglen >= valid_len) { + struct virtchnl_rss_lut *vrl = + (struct virtchnl_rss_lut *)msg; + valid_len += vrl->lut_entries - 1; + } + break; + case VIRTCHNL_OP_GET_RSS_HENA_CAPS: + break; + case VIRTCHNL_OP_SET_RSS_HENA: + valid_len = sizeof(struct virtchnl_rss_hena); + break; + case VIRTCHNL_OP_ENABLE_VLAN_STRIPPING: + case VIRTCHNL_OP_DISABLE_VLAN_STRIPPING: + break; + case VIRTCHNL_OP_REQUEST_QUEUES: + valid_len = sizeof(struct virtchnl_vf_res_request); + break; + /* These are always errors coming from the VF. */ + case VIRTCHNL_OP_EVENT: + case VIRTCHNL_OP_UNKNOWN: + default: + return VIRTCHNL_ERR_PARAM; + } + /* few more checks */ + if ((valid_len != msglen) || (err_msg_format)) + return VIRTCHNL_STATUS_ERR_OPCODE_MISMATCH; + + return 0; +} +#endif /* _VIRTCHNL_H_ */ diff --git a/drivers/net/i40e/i40e_ethdev.h b/drivers/net/i40e/i40e_ethdev.h index 183dc17..bda7450 100644 --- a/drivers/net/i40e/i40e_ethdev.h +++ b/drivers/net/i40e/i40e_ethdev.h @@ -102,7 +102,7 @@ /* Linux PF host with virtchnl version 1.1 */ #define PF_IS_V11(vf) \ - (((vf)->version_major == I40E_VIRTCHNL_VERSION_MAJOR) && \ + (((vf)->version_major == VIRTCHNL_VERSION_MAJOR) && \ ((vf)->version_minor == 1)) /* index flex payload per layer */ @@ -767,7 +767,7 @@ struct i40e_vf { /* Event from pf */ bool dev_closed; bool link_up; - enum i40e_aq_link_speed link_speed; + enum virtchnl_link_speed link_speed; bool vf_reset; volatile uint32_t pend_cmd; /* pending command not finished yet */ int32_t cmd_retval; /* return value of the cmd response from PF */ @@ -775,8 +775,8 @@ struct i40e_vf { uint8_t *aq_resp; /* buffer to store the adminq response from PF */ /* VSI info */ - struct i40e_virtchnl_vf_resource *vf_res; /* All VSIs */ - struct i40e_virtchnl_vsi_resource *vsi_res; /* LAN VSI */ + struct virtchnl_vf_resource *vf_res; /* All VSIs */ + struct virtchnl_vsi_resource *vsi_res; /* LAN VSI */ struct i40e_vsi vsi; uint64_t flags; }; diff --git a/drivers/net/i40e/i40e_ethdev_vf.c b/drivers/net/i40e/i40e_ethdev_vf.c index 2d5a9b5..456dd11 100644 --- a/drivers/net/i40e/i40e_ethdev_vf.c +++ b/drivers/net/i40e/i40e_ethdev_vf.c @@ -77,7 +77,7 @@ #define MAX_RESET_WAIT_CNT 20 struct i40evf_arq_msg_info { - enum i40e_virtchnl_ops ops; + enum virtchnl_ops ops; enum i40e_status_code result; uint16_t buf_len; uint16_t msg_len; @@ -85,7 +85,7 @@ struct i40evf_arq_msg_info { }; struct vf_cmd_info { - enum i40e_virtchnl_ops ops; + enum virtchnl_ops ops; uint8_t *in_args; uint32_t in_args_size; uint8_t *out_buffer; @@ -244,7 +244,7 @@ i40evf_read_pfmsg(struct rte_eth_dev *dev, struct i40evf_arq_msg_info *data) struct i40e_hw *hw = I40E_DEV_PRIVATE_TO_HW(dev->data->dev_private); struct i40e_vf *vf = I40EVF_DEV_PRIVATE_TO_VF(dev->data->dev_private); struct i40e_arq_event_info event; - enum i40e_virtchnl_ops opcode; + enum virtchnl_ops opcode; enum i40e_status_code retval; int ret; enum i40evf_aq_result result = I40EVF_MSG_NON; @@ -259,16 +259,16 @@ i40evf_read_pfmsg(struct rte_eth_dev *dev, struct i40evf_arq_msg_info *data) return result; } - opcode = (enum i40e_virtchnl_ops)rte_le_to_cpu_32(event.desc.cookie_high); + opcode = (enum virtchnl_ops)rte_le_to_cpu_32(event.desc.cookie_high); retval = (enum i40e_status_code)rte_le_to_cpu_32(event.desc.cookie_low); /* pf sys event */ - if (opcode == I40E_VIRTCHNL_OP_EVENT) { - struct i40e_virtchnl_pf_event *vpe = - (struct i40e_virtchnl_pf_event *)event.msg_buf; + if (opcode == VIRTCHNL_OP_EVENT) { + struct virtchnl_pf_event *vpe = + (struct virtchnl_pf_event *)event.msg_buf; result = I40EVF_MSG_SYS; switch (vpe->event) { - case I40E_VIRTCHNL_EVENT_LINK_CHANGE: + case VIRTCHNL_EVENT_LINK_CHANGE: vf->link_up = vpe->event_data.link_event.link_status; vf->link_speed = @@ -277,12 +277,12 @@ i40evf_read_pfmsg(struct rte_eth_dev *dev, struct i40evf_arq_msg_info *data) PMD_DRV_LOG(INFO, "Link status update:%s", vf->link_up ? "up" : "down"); break; - case I40E_VIRTCHNL_EVENT_RESET_IMPENDING: + case VIRTCHNL_EVENT_RESET_IMPENDING: vf->vf_reset = true; vf->pend_msg |= PFMSG_RESET_IMPENDING; PMD_DRV_LOG(INFO, "vf is reseting"); break; - case I40E_VIRTCHNL_EVENT_PF_DRIVER_CLOSE: + case VIRTCHNL_EVENT_PF_DRIVER_CLOSE: vf->dev_closed = true; vf->pend_msg |= PFMSG_DRIVER_CLOSE; PMD_DRV_LOG(INFO, "PF driver closed"); @@ -312,17 +312,17 @@ static inline void _clear_cmd(struct i40e_vf *vf) { rte_wmb(); - vf->pend_cmd = I40E_VIRTCHNL_OP_UNKNOWN; + vf->pend_cmd = VIRTCHNL_OP_UNKNOWN; } /* * Check there is pending cmd in execution. If none, set new command. */ static inline int -_atomic_set_cmd(struct i40e_vf *vf, enum i40e_virtchnl_ops ops) +_atomic_set_cmd(struct i40e_vf *vf, enum virtchnl_ops ops) { int ret = rte_atomic32_cmpset(&vf->pend_cmd, - I40E_VIRTCHNL_OP_UNKNOWN, ops); + VIRTCHNL_OP_UNKNOWN, ops); if (!ret) PMD_DRV_LOG(ERR, "There is incomplete cmd %d", vf->pend_cmd); @@ -347,7 +347,7 @@ i40evf_execute_vf_cmd(struct rte_eth_dev *dev, struct vf_cmd_info *args) info.msg = args->out_buffer; info.buf_len = args->out_size; - info.ops = I40E_VIRTCHNL_OP_UNKNOWN; + info.ops = VIRTCHNL_OP_UNKNOWN; info.result = I40E_SUCCESS; err = i40e_aq_send_msg_to_pf(hw, args->ops, I40E_SUCCESS, @@ -359,12 +359,12 @@ i40evf_execute_vf_cmd(struct rte_eth_dev *dev, struct vf_cmd_info *args) } switch (args->ops) { - case I40E_VIRTCHNL_OP_RESET_VF: + case VIRTCHNL_OP_RESET_VF: /*no need to process in this function */ err = 0; break; - case I40E_VIRTCHNL_OP_VERSION: - case I40E_VIRTCHNL_OP_GET_VF_RESOURCES: + case VIRTCHNL_OP_VERSION: + case VIRTCHNL_OP_GET_VF_RESOURCES: /* for init adminq commands, need to poll the response */ err = -1; do { @@ -385,7 +385,7 @@ i40evf_execute_vf_cmd(struct rte_eth_dev *dev, struct vf_cmd_info *args) /* for other adminq in running time, waiting the cmd done flag */ err = -1; do { - if (vf->pend_cmd == I40E_VIRTCHNL_OP_UNKNOWN) { + if (vf->pend_cmd == VIRTCHNL_OP_UNKNOWN) { err = 0; break; } @@ -404,15 +404,15 @@ i40evf_execute_vf_cmd(struct rte_eth_dev *dev, struct vf_cmd_info *args) static int i40evf_check_api_version(struct rte_eth_dev *dev) { - struct i40e_virtchnl_version_info version, *pver; + struct virtchnl_version_info version, *pver; int err; struct vf_cmd_info args; struct i40e_vf *vf = I40EVF_DEV_PRIVATE_TO_VF(dev->data->dev_private); - version.major = I40E_VIRTCHNL_VERSION_MAJOR; - version.minor = I40E_VIRTCHNL_VERSION_MINOR; + version.major = VIRTCHNL_VERSION_MAJOR; + version.minor = VIRTCHNL_VERSION_MINOR; - args.ops = I40E_VIRTCHNL_OP_VERSION; + args.ops = VIRTCHNL_OP_VERSION; args.in_args = (uint8_t *)&version; args.in_args_size = sizeof(version); args.out_buffer = vf->aq_resp; @@ -424,19 +424,19 @@ i40evf_check_api_version(struct rte_eth_dev *dev) return err; } - pver = (struct i40e_virtchnl_version_info *)args.out_buffer; + pver = (struct virtchnl_version_info *)args.out_buffer; vf->version_major = pver->major; vf->version_minor = pver->minor; if (vf->version_major == I40E_DPDK_VERSION_MAJOR) PMD_DRV_LOG(INFO, "Peer is DPDK PF host"); - else if ((vf->version_major == I40E_VIRTCHNL_VERSION_MAJOR) && - (vf->version_minor <= I40E_VIRTCHNL_VERSION_MINOR)) + else if ((vf->version_major == VIRTCHNL_VERSION_MAJOR) && + (vf->version_minor <= VIRTCHNL_VERSION_MINOR)) PMD_DRV_LOG(INFO, "Peer is Linux PF host"); else { PMD_INIT_LOG(ERR, "PF/VF API version mismatch:(%u.%u)-(%u.%u)", vf->version_major, vf->version_minor, - I40E_VIRTCHNL_VERSION_MAJOR, - I40E_VIRTCHNL_VERSION_MINOR); + VIRTCHNL_VERSION_MAJOR, + VIRTCHNL_VERSION_MINOR); return -1; } @@ -452,15 +452,15 @@ i40evf_get_vf_resource(struct rte_eth_dev *dev) struct vf_cmd_info args; uint32_t caps, len; - args.ops = I40E_VIRTCHNL_OP_GET_VF_RESOURCES; + args.ops = VIRTCHNL_OP_GET_VF_RESOURCES; args.out_buffer = vf->aq_resp; args.out_size = I40E_AQ_BUF_SZ; if (PF_IS_V11(vf)) { - caps = I40E_VIRTCHNL_VF_OFFLOAD_L2 | - I40E_VIRTCHNL_VF_OFFLOAD_RSS_AQ | - I40E_VIRTCHNL_VF_OFFLOAD_RSS_REG | - I40E_VIRTCHNL_VF_OFFLOAD_VLAN | - I40E_VIRTCHNL_VF_OFFLOAD_RX_POLLING; + caps = VIRTCHNL_VF_OFFLOAD_L2 | + VIRTCHNL_VF_OFFLOAD_RSS_AQ | + VIRTCHNL_VF_OFFLOAD_RSS_REG | + VIRTCHNL_VF_OFFLOAD_VLAN | + VIRTCHNL_VF_OFFLOAD_RX_POLLING; args.in_args = (uint8_t *)∩︀ args.in_args_size = sizeof(caps); } else { @@ -474,8 +474,8 @@ i40evf_get_vf_resource(struct rte_eth_dev *dev) return err; } - len = sizeof(struct i40e_virtchnl_vf_resource) + - I40E_MAX_VF_VSI * sizeof(struct i40e_virtchnl_vsi_resource); + len = sizeof(struct virtchnl_vf_resource) + + I40E_MAX_VF_VSI * sizeof(struct virtchnl_vsi_resource); (void)rte_memcpy(vf->vf_res, args.out_buffer, RTE_MIN(args.out_size, len)); @@ -492,18 +492,18 @@ i40evf_config_promisc(struct rte_eth_dev *dev, struct i40e_vf *vf = I40EVF_DEV_PRIVATE_TO_VF(dev->data->dev_private); int err; struct vf_cmd_info args; - struct i40e_virtchnl_promisc_info promisc; + struct virtchnl_promisc_info promisc; promisc.flags = 0; promisc.vsi_id = vf->vsi_res->vsi_id; if (enable_unicast) - promisc.flags |= I40E_FLAG_VF_UNICAST_PROMISC; + promisc.flags |= FLAG_VF_UNICAST_PROMISC; if (enable_multicast) - promisc.flags |= I40E_FLAG_VF_MULTICAST_PROMISC; + promisc.flags |= FLAG_VF_MULTICAST_PROMISC; - args.ops = I40E_VIRTCHNL_OP_CONFIG_PROMISCUOUS_MODE; + args.ops = VIRTCHNL_OP_CONFIG_PROMISCUOUS_MODE; args.in_args = (uint8_t *)&promisc; args.in_args_size = sizeof(promisc); args.out_buffer = vf->aq_resp; @@ -525,12 +525,12 @@ i40evf_config_vlan_offload(struct rte_eth_dev *dev, struct i40e_vf *vf = I40EVF_DEV_PRIVATE_TO_VF(dev->data->dev_private); int err; struct vf_cmd_info args; - struct i40e_virtchnl_vlan_offload_info offload; + struct virtchnl_vlan_offload_info offload; offload.vsi_id = vf->vsi_res->vsi_id; offload.enable_vlan_strip = enable_vlan_strip; - args.ops = (enum i40e_virtchnl_ops)I40E_VIRTCHNL_OP_CFG_VLAN_OFFLOAD; + args.ops = (enum virtchnl_ops)I40E_VIRTCHNL_OP_CFG_VLAN_OFFLOAD; args.in_args = (uint8_t *)&offload; args.in_args_size = sizeof(offload); args.out_buffer = vf->aq_resp; @@ -550,7 +550,7 @@ i40evf_config_vlan_pvid(struct rte_eth_dev *dev, struct i40e_vf *vf = I40EVF_DEV_PRIVATE_TO_VF(dev->data->dev_private); int err; struct vf_cmd_info args; - struct i40e_virtchnl_pvid_info tpid_info; + struct virtchnl_pvid_info tpid_info; if (info == NULL) { PMD_DRV_LOG(ERR, "invalid parameters"); @@ -561,7 +561,7 @@ i40evf_config_vlan_pvid(struct rte_eth_dev *dev, tpid_info.vsi_id = vf->vsi_res->vsi_id; (void)rte_memcpy(&tpid_info.info, info, sizeof(*info)); - args.ops = (enum i40e_virtchnl_ops)I40E_VIRTCHNL_OP_CFG_VLAN_PVID; + args.ops = (enum virtchnl_ops)I40E_VIRTCHNL_OP_CFG_VLAN_PVID; args.in_args = (uint8_t *)&tpid_info; args.in_args_size = sizeof(tpid_info); args.out_buffer = vf->aq_resp; @@ -575,7 +575,7 @@ i40evf_config_vlan_pvid(struct rte_eth_dev *dev, } static void -i40evf_fill_virtchnl_vsi_txq_info(struct i40e_virtchnl_txq_info *txq_info, +i40evf_fill_virtchnl_vsi_txq_info(struct virtchnl_txq_info *txq_info, uint16_t vsi_id, uint16_t queue_id, uint16_t nb_txq, @@ -590,7 +590,7 @@ i40evf_fill_virtchnl_vsi_txq_info(struct i40e_virtchnl_txq_info *txq_info, } static void -i40evf_fill_virtchnl_vsi_rxq_info(struct i40e_virtchnl_rxq_info *rxq_info, +i40evf_fill_virtchnl_vsi_rxq_info(struct virtchnl_rxq_info *rxq_info, uint16_t vsi_id, uint16_t queue_id, uint16_t nb_rxq, @@ -618,8 +618,8 @@ i40evf_configure_vsi_queues(struct rte_eth_dev *dev) (struct i40e_rx_queue **)dev->data->rx_queues; struct i40e_tx_queue **txq = (struct i40e_tx_queue **)dev->data->tx_queues; - struct i40e_virtchnl_vsi_queue_config_info *vc_vqci; - struct i40e_virtchnl_queue_pair_info *vc_qpi; + struct virtchnl_vsi_queue_config_info *vc_vqci; + struct virtchnl_queue_pair_info *vc_qpi; struct vf_cmd_info args; uint16_t i, nb_qp = vf->num_queue_pairs; const uint32_t size = @@ -628,7 +628,7 @@ i40evf_configure_vsi_queues(struct rte_eth_dev *dev) int ret; memset(buff, 0, sizeof(buff)); - vc_vqci = (struct i40e_virtchnl_vsi_queue_config_info *)buff; + vc_vqci = (struct virtchnl_vsi_queue_config_info *)buff; vc_vqci->vsi_id = vf->vsi_res->vsi_id; vc_vqci->num_queue_pairs = nb_qp; @@ -640,7 +640,7 @@ i40evf_configure_vsi_queues(struct rte_eth_dev *dev) vf->max_pkt_len, rxq[i]); } memset(&args, 0, sizeof(args)); - args.ops = I40E_VIRTCHNL_OP_CONFIG_VSI_QUEUES; + args.ops = VIRTCHNL_OP_CONFIG_VSI_QUEUES; args.in_args = (uint8_t *)vc_vqci; args.in_args_size = size; args.out_buffer = vf->aq_resp; @@ -648,7 +648,7 @@ i40evf_configure_vsi_queues(struct rte_eth_dev *dev) ret = i40evf_execute_vf_cmd(dev, &args); if (ret) PMD_DRV_LOG(ERR, "Failed to execute command of " - "I40E_VIRTCHNL_OP_CONFIG_VSI_QUEUES"); + "VIRTCHNL_OP_CONFIG_VSI_QUEUES"); return ret; } @@ -662,8 +662,8 @@ i40evf_configure_vsi_queues_ext(struct rte_eth_dev *dev) (struct i40e_rx_queue **)dev->data->rx_queues; struct i40e_tx_queue **txq = (struct i40e_tx_queue **)dev->data->tx_queues; - struct i40e_virtchnl_vsi_queue_config_ext_info *vc_vqcei; - struct i40e_virtchnl_queue_pair_ext_info *vc_qpei; + struct virtchnl_vsi_queue_config_ext_info *vc_vqcei; + struct virtchnl_queue_pair_ext_info *vc_qpei; struct vf_cmd_info args; uint16_t i, nb_qp = vf->num_queue_pairs; const uint32_t size = @@ -672,7 +672,7 @@ i40evf_configure_vsi_queues_ext(struct rte_eth_dev *dev) int ret; memset(buff, 0, sizeof(buff)); - vc_vqcei = (struct i40e_virtchnl_vsi_queue_config_ext_info *)buff; + vc_vqcei = (struct virtchnl_vsi_queue_config_ext_info *)buff; vc_vqcei->vsi_id = vf->vsi_res->vsi_id; vc_vqcei->num_queue_pairs = nb_qp; vc_qpei = vc_vqcei->qpair; @@ -693,7 +693,7 @@ i40evf_configure_vsi_queues_ext(struct rte_eth_dev *dev) } memset(&args, 0, sizeof(args)); args.ops = - (enum i40e_virtchnl_ops)I40E_VIRTCHNL_OP_CONFIG_VSI_QUEUES_EXT; + (enum virtchnl_ops)VIRTCHNL_OP_CONFIG_VSI_QUEUES_EXT; args.in_args = (uint8_t *)vc_vqcei; args.in_args_size = size; args.out_buffer = vf->aq_resp; @@ -701,7 +701,7 @@ i40evf_configure_vsi_queues_ext(struct rte_eth_dev *dev) ret = i40evf_execute_vf_cmd(dev, &args); if (ret) PMD_DRV_LOG(ERR, "Failed to execute command of " - "I40E_VIRTCHNL_OP_CONFIG_VSI_QUEUES_EXT"); + "VIRTCHNL_OP_CONFIG_VSI_QUEUES_EXT"); return ret; } @@ -724,9 +724,9 @@ i40evf_config_irq_map(struct rte_eth_dev *dev) { struct i40e_vf *vf = I40EVF_DEV_PRIVATE_TO_VF(dev->data->dev_private); struct vf_cmd_info args; - uint8_t cmd_buffer[sizeof(struct i40e_virtchnl_irq_map_info) + \ - sizeof(struct i40e_virtchnl_vector_map)]; - struct i40e_virtchnl_irq_map_info *map_info; + uint8_t cmd_buffer[sizeof(struct virtchnl_irq_map_info) + \ + sizeof(struct virtchnl_vector_map)]; + struct virtchnl_irq_map_info *map_info; struct rte_pci_device *pci_dev = RTE_ETH_DEV_TO_PCI(dev); struct rte_intr_handle *intr_handle = &pci_dev->intr_handle; uint32_t vector_id; @@ -741,7 +741,7 @@ i40evf_config_irq_map(struct rte_eth_dev *dev) vector_id = I40E_MISC_VEC_ID; } - map_info = (struct i40e_virtchnl_irq_map_info *)cmd_buffer; + map_info = (struct virtchnl_irq_map_info *)cmd_buffer; map_info->num_vectors = 1; map_info->vecmap[0].rxitr_idx = I40E_ITR_INDEX_DEFAULT; map_info->vecmap[0].vsi_id = vf->vsi_res->vsi_id; @@ -756,7 +756,7 @@ i40evf_config_irq_map(struct rte_eth_dev *dev) intr_handle->intr_vec[i] = vector_id; } - args.ops = I40E_VIRTCHNL_OP_CONFIG_IRQ_MAP; + args.ops = VIRTCHNL_OP_CONFIG_IRQ_MAP; args.in_args = (u8 *)cmd_buffer; args.in_args_size = sizeof(cmd_buffer); args.out_buffer = vf->aq_resp; @@ -773,7 +773,7 @@ i40evf_switch_queue(struct rte_eth_dev *dev, bool isrx, uint16_t qid, bool on) { struct i40e_vf *vf = I40EVF_DEV_PRIVATE_TO_VF(dev->data->dev_private); - struct i40e_virtchnl_queue_select queue_select; + struct virtchnl_queue_select queue_select; int err; struct vf_cmd_info args; memset(&queue_select, 0, sizeof(queue_select)); @@ -785,9 +785,9 @@ i40evf_switch_queue(struct rte_eth_dev *dev, bool isrx, uint16_t qid, queue_select.tx_queues |= 1 << qid; if (on) - args.ops = I40E_VIRTCHNL_OP_ENABLE_QUEUES; + args.ops = VIRTCHNL_OP_ENABLE_QUEUES; else - args.ops = I40E_VIRTCHNL_OP_DISABLE_QUEUES; + args.ops = VIRTCHNL_OP_DISABLE_QUEUES; args.in_args = (u8 *)&queue_select; args.in_args_size = sizeof(queue_select); args.out_buffer = vf->aq_resp; @@ -861,10 +861,10 @@ i40evf_add_mac_addr(struct rte_eth_dev *dev, __rte_unused uint32_t index, __rte_unused uint32_t pool) { - struct i40e_virtchnl_ether_addr_list *list; + struct virtchnl_ether_addr_list *list; struct i40e_vf *vf = I40EVF_DEV_PRIVATE_TO_VF(dev->data->dev_private); - uint8_t cmd_buffer[sizeof(struct i40e_virtchnl_ether_addr_list) + \ - sizeof(struct i40e_virtchnl_ether_addr)]; + uint8_t cmd_buffer[sizeof(struct virtchnl_ether_addr_list) + \ + sizeof(struct virtchnl_ether_addr)]; int err; struct vf_cmd_info args; @@ -876,13 +876,13 @@ i40evf_add_mac_addr(struct rte_eth_dev *dev, return I40E_ERR_INVALID_MAC_ADDR; } - list = (struct i40e_virtchnl_ether_addr_list *)cmd_buffer; + list = (struct virtchnl_ether_addr_list *)cmd_buffer; list->vsi_id = vf->vsi_res->vsi_id; list->num_elements = 1; (void)rte_memcpy(list->list[0].addr, addr->addr_bytes, sizeof(addr->addr_bytes)); - args.ops = I40E_VIRTCHNL_OP_ADD_ETHER_ADDRESS; + args.ops = VIRTCHNL_OP_ADD_ETH_ADDR; args.in_args = cmd_buffer; args.in_args_size = sizeof(cmd_buffer); args.out_buffer = vf->aq_resp; @@ -899,10 +899,10 @@ static void i40evf_del_mac_addr_by_addr(struct rte_eth_dev *dev, struct ether_addr *addr) { - struct i40e_virtchnl_ether_addr_list *list; + struct virtchnl_ether_addr_list *list; struct i40e_vf *vf = I40EVF_DEV_PRIVATE_TO_VF(dev->data->dev_private); - uint8_t cmd_buffer[sizeof(struct i40e_virtchnl_ether_addr_list) + \ - sizeof(struct i40e_virtchnl_ether_addr)]; + uint8_t cmd_buffer[sizeof(struct virtchnl_ether_addr_list) + \ + sizeof(struct virtchnl_ether_addr)]; int err; struct vf_cmd_info args; @@ -914,13 +914,13 @@ i40evf_del_mac_addr_by_addr(struct rte_eth_dev *dev, return; } - list = (struct i40e_virtchnl_ether_addr_list *)cmd_buffer; + list = (struct virtchnl_ether_addr_list *)cmd_buffer; list->vsi_id = vf->vsi_res->vsi_id; list->num_elements = 1; (void)rte_memcpy(list->list[0].addr, addr->addr_bytes, sizeof(addr->addr_bytes)); - args.ops = I40E_VIRTCHNL_OP_DEL_ETHER_ADDRESS; + args.ops = VIRTCHNL_OP_DEL_ETH_ADDR; args.in_args = cmd_buffer; args.in_args_size = sizeof(cmd_buffer); args.out_buffer = vf->aq_resp; @@ -947,13 +947,13 @@ static int i40evf_update_stats(struct rte_eth_dev *dev, struct i40e_eth_stats **pstats) { struct i40e_vf *vf = I40EVF_DEV_PRIVATE_TO_VF(dev->data->dev_private); - struct i40e_virtchnl_queue_select q_stats; + struct virtchnl_queue_select q_stats; int err; struct vf_cmd_info args; memset(&q_stats, 0, sizeof(q_stats)); q_stats.vsi_id = vf->vsi_res->vsi_id; - args.ops = I40E_VIRTCHNL_OP_GET_STATS; + args.ops = VIRTCHNL_OP_GET_STATS; args.in_args = (u8 *)&q_stats; args.in_args_size = sizeof(q_stats); args.out_buffer = vf->aq_resp; @@ -1050,18 +1050,18 @@ static int i40evf_add_vlan(struct rte_eth_dev *dev, uint16_t vlanid) { struct i40e_vf *vf = I40EVF_DEV_PRIVATE_TO_VF(dev->data->dev_private); - struct i40e_virtchnl_vlan_filter_list *vlan_list; - uint8_t cmd_buffer[sizeof(struct i40e_virtchnl_vlan_filter_list) + + struct virtchnl_vlan_filter_list *vlan_list; + uint8_t cmd_buffer[sizeof(struct virtchnl_vlan_filter_list) + sizeof(uint16_t)]; int err; struct vf_cmd_info args; - vlan_list = (struct i40e_virtchnl_vlan_filter_list *)cmd_buffer; + vlan_list = (struct virtchnl_vlan_filter_list *)cmd_buffer; vlan_list->vsi_id = vf->vsi_res->vsi_id; vlan_list->num_elements = 1; vlan_list->vlan_id[0] = vlanid; - args.ops = I40E_VIRTCHNL_OP_ADD_VLAN; + args.ops = VIRTCHNL_OP_ADD_VLAN; args.in_args = (u8 *)&cmd_buffer; args.in_args_size = sizeof(cmd_buffer); args.out_buffer = vf->aq_resp; @@ -1077,18 +1077,18 @@ static int i40evf_del_vlan(struct rte_eth_dev *dev, uint16_t vlanid) { struct i40e_vf *vf = I40EVF_DEV_PRIVATE_TO_VF(dev->data->dev_private); - struct i40e_virtchnl_vlan_filter_list *vlan_list; - uint8_t cmd_buffer[sizeof(struct i40e_virtchnl_vlan_filter_list) + + struct virtchnl_vlan_filter_list *vlan_list; + uint8_t cmd_buffer[sizeof(struct virtchnl_vlan_filter_list) + sizeof(uint16_t)]; int err; struct vf_cmd_info args; - vlan_list = (struct i40e_virtchnl_vlan_filter_list *)cmd_buffer; + vlan_list = (struct virtchnl_vlan_filter_list *)cmd_buffer; vlan_list->vsi_id = vf->vsi_res->vsi_id; vlan_list->num_elements = 1; vlan_list->vlan_id[0] = vlanid; - args.ops = I40E_VIRTCHNL_OP_DEL_VLAN; + args.ops = VIRTCHNL_OP_DEL_VLAN; args.in_args = (u8 *)&cmd_buffer; args.in_args_size = sizeof(cmd_buffer); args.out_buffer = vf->aq_resp; @@ -1178,7 +1178,7 @@ i40evf_reset_vf(struct i40e_hw *hw) reset = rd32(hw, I40E_VFGEN_RSTAT) & I40E_VFGEN_RSTAT_VFR_STATE_MASK; reset = reset >> I40E_VFGEN_RSTAT_VFR_STATE_SHIFT; - if (I40E_VFR_COMPLETED == reset || I40E_VFR_VFACTIVE == reset) + if (VIRTCHNL_VFR_COMPLETED == reset || VIRTCHNL_VFR_VFACTIVE == reset) break; else rte_delay_ms(50); @@ -1242,8 +1242,8 @@ i40evf_init_vf(struct rte_eth_dev *dev) PMD_INIT_LOG(ERR, "check_api version failed"); goto err_aq; } - bufsz = sizeof(struct i40e_virtchnl_vf_resource) + - (I40E_MAX_VF_VSI * sizeof(struct i40e_virtchnl_vsi_resource)); + bufsz = sizeof(struct virtchnl_vf_resource) + + (I40E_MAX_VF_VSI * sizeof(struct virtchnl_vsi_resource)); vf->vf_res = rte_zmalloc("vf_res", bufsz, 0); if (!vf->vf_res) { PMD_INIT_LOG(ERR, "unable to allocate vf_res memory"); @@ -1257,7 +1257,7 @@ i40evf_init_vf(struct rte_eth_dev *dev) /* got VF config message back from PF, now we can parse it */ for (i = 0; i < vf->vf_res->num_vsis; i++) { - if (vf->vf_res->vsi_res[i].vsi_type == I40E_VSI_SRIOV) + if (vf->vf_res->vsi_res[i].vsi_type == VIRTCHNL_VSI_SRIOV) vf->vsi_res = &vf->vf_res->vsi_res[i]; } @@ -1269,7 +1269,7 @@ i40evf_init_vf(struct rte_eth_dev *dev) if (hw->mac.type == I40E_MAC_X722_VF) vf->flags = I40E_FLAG_RSS_AQ_CAPABLE; vf->vsi.vsi_id = vf->vsi_res->vsi_id; - vf->vsi.type = vf->vsi_res->vsi_type; + vf->vsi.type = (enum i40e_vsi_type)vf->vsi_res->vsi_type; vf->vsi.nb_qps = vf->vsi_res->num_queue_pairs; vf->vsi.adapter = I40E_DEV_PRIVATE_TO_ADAPTER(dev->data->dev_private); @@ -1321,21 +1321,21 @@ static void i40evf_handle_pf_event(struct rte_eth_dev *dev, uint8_t *msg, __rte_unused uint16_t msglen) { - struct i40e_virtchnl_pf_event *pf_msg = - (struct i40e_virtchnl_pf_event *)msg; + struct virtchnl_pf_event *pf_msg = + (struct virtchnl_pf_event *)msg; struct i40e_vf *vf = I40EVF_DEV_PRIVATE_TO_VF(dev->data->dev_private); switch (pf_msg->event) { - case I40E_VIRTCHNL_EVENT_RESET_IMPENDING: + case VIRTCHNL_EVENT_RESET_IMPENDING: PMD_DRV_LOG(DEBUG, "VIRTCHNL_EVENT_RESET_IMPENDING event"); _rte_eth_dev_callback_process(dev, RTE_ETH_EVENT_INTR_RESET, NULL); break; - case I40E_VIRTCHNL_EVENT_LINK_CHANGE: + case VIRTCHNL_EVENT_LINK_CHANGE: PMD_DRV_LOG(DEBUG, "VIRTCHNL_EVENT_LINK_CHANGE event"); vf->link_up = pf_msg->event_data.link_event.link_status; vf->link_speed = pf_msg->event_data.link_event.link_speed; break; - case I40E_VIRTCHNL_EVENT_PF_DRIVER_CLOSE: + case VIRTCHNL_EVENT_PF_DRIVER_CLOSE: PMD_DRV_LOG(DEBUG, "VIRTCHNL_EVENT_PF_DRIVER_CLOSE event"); break; default: @@ -1351,7 +1351,7 @@ i40evf_handle_aq_msg(struct rte_eth_dev *dev) struct i40e_vf *vf = I40EVF_DEV_PRIVATE_TO_VF(dev->data->dev_private); struct i40e_arq_event_info info; uint16_t pending, aq_opc; - enum i40e_virtchnl_ops msg_opc; + enum virtchnl_ops msg_opc; enum i40e_status_code msg_ret; int ret; @@ -1376,13 +1376,13 @@ i40evf_handle_aq_msg(struct rte_eth_dev *dev) * cookie_high of struct i40e_aq_desc, while return error code * are stored in cookie_low, Which is done by * i40e_aq_send_msg_to_vf in PF driver.*/ - msg_opc = (enum i40e_virtchnl_ops)rte_le_to_cpu_32( + msg_opc = (enum virtchnl_ops)rte_le_to_cpu_32( info.desc.cookie_high); msg_ret = (enum i40e_status_code)rte_le_to_cpu_32( info.desc.cookie_low); switch (aq_opc) { case i40e_aqc_opc_send_msg_to_vf: - if (msg_opc == I40E_VIRTCHNL_OP_EVENT) + if (msg_opc == VIRTCHNL_OP_EVENT) /* process event*/ i40evf_handle_pf_event(dev, info.msg_buf, info.msg_len); @@ -1591,8 +1591,8 @@ i40evf_dev_configure(struct rte_eth_dev *dev) */ if (!conf->rxmode.hw_strip_crc) { vf = I40EVF_DEV_PRIVATE_TO_VF(dev->data->dev_private); - if ((vf->version_major == I40E_VIRTCHNL_VERSION_MAJOR) && - (vf->version_minor <= I40E_VIRTCHNL_VERSION_MINOR)) { + if ((vf->version_major == VIRTCHNL_VERSION_MAJOR) && + (vf->version_minor <= VIRTCHNL_VERSION_MINOR)) { /* Peer is running non-DPDK PF driver. */ PMD_INIT_LOG(ERR, "VF can't disable HW CRC Strip"); return -EINVAL; @@ -2000,7 +2000,7 @@ i40evf_dev_rx_queue_intr_disable(struct rte_eth_dev *dev, uint16_t queue_id) static void i40evf_add_del_all_mac_addr(struct rte_eth_dev *dev, bool add) { - struct i40e_virtchnl_ether_addr_list *list; + struct virtchnl_ether_addr_list *list; struct i40e_vf *vf = I40EVF_DEV_PRIVATE_TO_VF(dev->data->dev_private); int err, i, j; int next_begin = 0; @@ -2011,11 +2011,11 @@ i40evf_add_del_all_mac_addr(struct rte_eth_dev *dev, bool add) do { j = 0; - len = sizeof(struct i40e_virtchnl_ether_addr_list); + len = sizeof(struct virtchnl_ether_addr_list); for (i = begin; i < I40E_NUM_MACADDR_MAX; i++, next_begin++) { if (is_zero_ether_addr(&dev->data->mac_addrs[i])) continue; - len += sizeof(struct i40e_virtchnl_ether_addr); + len += sizeof(struct virtchnl_ether_addr); if (len >= I40E_AQ_BUF_SZ) { next_begin = i + 1; break; @@ -2042,8 +2042,8 @@ i40evf_add_del_all_mac_addr(struct rte_eth_dev *dev, bool add) } list->vsi_id = vf->vsi_res->vsi_id; list->num_elements = j; - args.ops = add ? I40E_VIRTCHNL_OP_ADD_ETHER_ADDRESS : - I40E_VIRTCHNL_OP_DEL_ETHER_ADDRESS; + args.ops = add ? VIRTCHNL_OP_ADD_ETH_ADDR : + VIRTCHNL_OP_DEL_ETH_ADDR; args.in_args = (uint8_t *)list; args.in_args_size = len; args.out_buffer = vf->aq_resp; diff --git a/drivers/net/i40e/i40e_pf.c b/drivers/net/i40e/i40e_pf.c index 0758503..f9b4be5 100644 --- a/drivers/net/i40e/i40e_pf.c +++ b/drivers/net/i40e/i40e_pf.c @@ -61,7 +61,7 @@ static int i40e_pf_host_switch_queues(struct i40e_pf_vf *vf, - struct i40e_virtchnl_queue_select *qsel, + struct virtchnl_queue_select *qsel, bool on); /** @@ -128,7 +128,7 @@ i40e_pf_host_vf_reset(struct i40e_pf_vf *vf, bool do_hw_reset) struct i40e_pf *pf; uint16_t vf_id, abs_vf_id, vf_msix_num; int ret; - struct i40e_virtchnl_queue_select qsel; + struct virtchnl_queue_select qsel; if (vf == NULL) return -EINVAL; @@ -139,7 +139,7 @@ i40e_pf_host_vf_reset(struct i40e_pf_vf *vf, bool do_hw_reset) abs_vf_id = vf_id + hw->func_caps.vf_base_id; /* Notify VF that we are in VFR progress */ - I40E_WRITE_REG(hw, I40E_VFGEN_RSTAT1(vf_id), I40E_VFR_INPROGRESS); + I40E_WRITE_REG(hw, I40E_VFGEN_RSTAT1(vf_id), VIRTCHNL_VFR_INPROGRESS); /* * If require a SW VF reset, a VFLR interrupt will be generated, @@ -220,7 +220,7 @@ i40e_pf_host_vf_reset(struct i40e_pf_vf *vf, bool do_hw_reset) } /* Reset done, Set COMPLETE flag and clear reset bit */ - I40E_WRITE_REG(hw, I40E_VFGEN_RSTAT1(vf_id), I40E_VFR_COMPLETED); + I40E_WRITE_REG(hw, I40E_VFGEN_RSTAT1(vf_id), VIRTCHNL_VFR_COMPLETED); val = I40E_READ_REG(hw, I40E_VPGEN_VFRTRIG(vf_id)); val &= ~I40E_VPGEN_VFRTRIG_VFSWR_MASK; I40E_WRITE_REG(hw, I40E_VPGEN_VFRTRIG(vf_id), val); @@ -248,7 +248,7 @@ i40e_pf_host_vf_reset(struct i40e_pf_vf *vf, bool do_hw_reset) return -EFAULT; } - I40E_WRITE_REG(hw, I40E_VFGEN_RSTAT1(vf_id), I40E_VFR_VFACTIVE); + I40E_WRITE_REG(hw, I40E_VFGEN_RSTAT1(vf_id), VIRTCHNL_VFR_VFACTIVE); return ret; } @@ -277,7 +277,7 @@ i40e_pf_host_send_msg_to_vf(struct i40e_pf_vf *vf, static void i40e_pf_host_process_cmd_version(struct i40e_pf_vf *vf, bool b_op) { - struct i40e_virtchnl_version_info info; + struct virtchnl_version_info info; /* Respond like a Linux PF host in order to support both DPDK VF and * Linux VF driver. The expense is original DPDK host specific feature @@ -286,16 +286,16 @@ i40e_pf_host_process_cmd_version(struct i40e_pf_vf *vf, bool b_op) * DPDK VF also can't identify host driver by version number returned. * It always assume talking with Linux PF. */ - info.major = I40E_VIRTCHNL_VERSION_MAJOR; - info.minor = I40E_VIRTCHNL_VERSION_MINOR_NO_VF_CAPS; + info.major = VIRTCHNL_VERSION_MAJOR; + info.minor = VIRTCHNL_VERSION_MINOR_NO_VF_CAPS; if (b_op) - i40e_pf_host_send_msg_to_vf(vf, I40E_VIRTCHNL_OP_VERSION, + i40e_pf_host_send_msg_to_vf(vf, VIRTCHNL_OP_VERSION, I40E_SUCCESS, (uint8_t *)&info, sizeof(info)); else - i40e_pf_host_send_msg_to_vf(vf, I40E_VIRTCHNL_OP_VERSION, + i40e_pf_host_send_msg_to_vf(vf, VIRTCHNL_OP_VERSION, I40E_NOT_SUPPORTED, (uint8_t *)&info, sizeof(info)); @@ -313,22 +313,22 @@ i40e_pf_host_process_cmd_reset_vf(struct i40e_pf_vf *vf) static int i40e_pf_host_process_cmd_get_vf_resource(struct i40e_pf_vf *vf, bool b_op) { - struct i40e_virtchnl_vf_resource *vf_res = NULL; + struct virtchnl_vf_resource *vf_res = NULL; struct i40e_hw *hw = I40E_PF_TO_HW(vf->pf); uint32_t len = 0; int ret = I40E_SUCCESS; if (!b_op) { i40e_pf_host_send_msg_to_vf(vf, - I40E_VIRTCHNL_OP_GET_VF_RESOURCES, + VIRTCHNL_OP_GET_VF_RESOURCES, I40E_NOT_SUPPORTED, NULL, 0); return ret; } /* only have 1 VSI by default */ - len = sizeof(struct i40e_virtchnl_vf_resource) + + len = sizeof(struct virtchnl_vf_resource) + I40E_DEFAULT_VF_VSI_NUM * - sizeof(struct i40e_virtchnl_vsi_resource); + sizeof(struct virtchnl_vsi_resource); vf_res = rte_zmalloc("i40e_vf_res", len, 0); if (vf_res == NULL) { @@ -339,21 +339,21 @@ i40e_pf_host_process_cmd_get_vf_resource(struct i40e_pf_vf *vf, bool b_op) goto send_msg; } - vf_res->vf_offload_flags = I40E_VIRTCHNL_VF_OFFLOAD_L2 | - I40E_VIRTCHNL_VF_OFFLOAD_VLAN; + vf_res->vf_offload_flags = VIRTCHNL_VF_OFFLOAD_L2 | + VIRTCHNL_VF_OFFLOAD_VLAN; vf_res->max_vectors = hw->func_caps.num_msix_vectors_vf; vf_res->num_queue_pairs = vf->vsi->nb_qps; vf_res->num_vsis = I40E_DEFAULT_VF_VSI_NUM; /* Change below setting if PF host can support more VSIs for VF */ - vf_res->vsi_res[0].vsi_type = I40E_VSI_SRIOV; + vf_res->vsi_res[0].vsi_type = VIRTCHNL_VSI_SRIOV; vf_res->vsi_res[0].vsi_id = vf->vsi->vsi_id; vf_res->vsi_res[0].num_queue_pairs = vf->vsi->nb_qps; ether_addr_copy(&vf->mac_addr, (struct ether_addr *)vf_res->vsi_res[0].default_mac_addr); send_msg: - i40e_pf_host_send_msg_to_vf(vf, I40E_VIRTCHNL_OP_GET_VF_RESOURCES, + i40e_pf_host_send_msg_to_vf(vf, VIRTCHNL_OP_GET_VF_RESOURCES, ret, (uint8_t *)vf_res, len); rte_free(vf_res); @@ -363,7 +363,7 @@ i40e_pf_host_process_cmd_get_vf_resource(struct i40e_pf_vf *vf, bool b_op) static int i40e_pf_host_hmc_config_rxq(struct i40e_hw *hw, struct i40e_pf_vf *vf, - struct i40e_virtchnl_rxq_info *rxq, + struct virtchnl_rxq_info *rxq, uint8_t crcstrip) { int err = I40E_SUCCESS; @@ -431,7 +431,7 @@ i40e_vsi_get_tc_of_queue(struct i40e_vsi *vsi, static int i40e_pf_host_hmc_config_txq(struct i40e_hw *hw, struct i40e_pf_vf *vf, - struct i40e_virtchnl_txq_info *txq) + struct virtchnl_txq_info *txq) { int err = I40E_SUCCESS; struct i40e_hmc_obj_txq tx_ctx; @@ -480,14 +480,14 @@ i40e_pf_host_process_cmd_config_vsi_queues(struct i40e_pf_vf *vf, { struct i40e_hw *hw = I40E_PF_TO_HW(vf->pf); struct i40e_vsi *vsi = vf->vsi; - struct i40e_virtchnl_vsi_queue_config_info *vc_vqci = - (struct i40e_virtchnl_vsi_queue_config_info *)msg; - struct i40e_virtchnl_queue_pair_info *vc_qpi; + struct virtchnl_vsi_queue_config_info *vc_vqci = + (struct virtchnl_vsi_queue_config_info *)msg; + struct virtchnl_queue_pair_info *vc_qpi; int i, ret = I40E_SUCCESS; if (!b_op) { i40e_pf_host_send_msg_to_vf(vf, - I40E_VIRTCHNL_OP_CONFIG_VSI_QUEUES, + VIRTCHNL_OP_CONFIG_VSI_QUEUES, I40E_NOT_SUPPORTED, NULL, 0); return ret; } @@ -511,9 +511,9 @@ i40e_pf_host_process_cmd_config_vsi_queues(struct i40e_pf_vf *vf, /* * Apply VF RX queue setting to HMC. - * If the opcode is I40E_VIRTCHNL_OP_CONFIG_VSI_QUEUES_EXT, + * If the opcode is VIRTCHNL_OP_CONFIG_VSI_QUEUES_EXT, * then the extra information of - * 'struct i40e_virtchnl_queue_pair_extra_info' is needed, + * 'struct virtchnl_queue_pair_extra_info' is needed, * otherwise set the last parameter to NULL. */ if (i40e_pf_host_hmc_config_rxq(hw, vf, &vc_qpi[i].rxq, @@ -533,7 +533,7 @@ i40e_pf_host_process_cmd_config_vsi_queues(struct i40e_pf_vf *vf, } send_msg: - i40e_pf_host_send_msg_to_vf(vf, I40E_VIRTCHNL_OP_CONFIG_VSI_QUEUES, + i40e_pf_host_send_msg_to_vf(vf, VIRTCHNL_OP_CONFIG_VSI_QUEUES, ret, NULL, 0); return ret; @@ -547,15 +547,15 @@ i40e_pf_host_process_cmd_config_vsi_queues_ext(struct i40e_pf_vf *vf, { struct i40e_hw *hw = I40E_PF_TO_HW(vf->pf); struct i40e_vsi *vsi = vf->vsi; - struct i40e_virtchnl_vsi_queue_config_ext_info *vc_vqcei = - (struct i40e_virtchnl_vsi_queue_config_ext_info *)msg; - struct i40e_virtchnl_queue_pair_ext_info *vc_qpei; + struct virtchnl_vsi_queue_config_ext_info *vc_vqcei = + (struct virtchnl_vsi_queue_config_ext_info *)msg; + struct virtchnl_queue_pair_ext_info *vc_qpei; int i, ret = I40E_SUCCESS; if (!b_op) { i40e_pf_host_send_msg_to_vf( vf, - I40E_VIRTCHNL_OP_CONFIG_VSI_QUEUES_EXT, + VIRTCHNL_OP_CONFIG_VSI_QUEUES_EXT, I40E_NOT_SUPPORTED, NULL, 0); return ret; } @@ -578,9 +578,9 @@ i40e_pf_host_process_cmd_config_vsi_queues_ext(struct i40e_pf_vf *vf, } /* * Apply VF RX queue setting to HMC. - * If the opcode is I40E_VIRTCHNL_OP_CONFIG_VSI_QUEUES_EXT, + * If the opcode is VIRTCHNL_OP_CONFIG_VSI_QUEUES_EXT, * then the extra information of - * 'struct i40e_virtchnl_queue_pair_ext_info' is needed, + * 'struct virtchnl_queue_pair_ext_info' is needed, * otherwise set the last parameter to NULL. */ if (i40e_pf_host_hmc_config_rxq(hw, vf, &vc_qpei[i].rxq, @@ -600,7 +600,7 @@ i40e_pf_host_process_cmd_config_vsi_queues_ext(struct i40e_pf_vf *vf, } send_msg: - i40e_pf_host_send_msg_to_vf(vf, I40E_VIRTCHNL_OP_CONFIG_VSI_QUEUES_EXT, + i40e_pf_host_send_msg_to_vf(vf, VIRTCHNL_OP_CONFIG_VSI_QUEUES_EXT, ret, NULL, 0); return ret; @@ -608,7 +608,7 @@ i40e_pf_host_process_cmd_config_vsi_queues_ext(struct i40e_pf_vf *vf, static void i40e_pf_config_irq_link_list(struct i40e_pf_vf *vf, - struct i40e_virtchnl_vector_map *vvm) + struct virtchnl_vector_map *vvm) { #define BITS_PER_CHAR 8 uint64_t linklistmap = 0, tempmap; @@ -711,9 +711,9 @@ i40e_pf_host_process_cmd_config_irq_map(struct i40e_pf_vf *vf, int ret = I40E_SUCCESS; struct i40e_pf *pf = vf->pf; struct i40e_hw *hw = I40E_PF_TO_HW(vf->pf); - struct i40e_virtchnl_irq_map_info *irqmap = - (struct i40e_virtchnl_irq_map_info *)msg; - struct i40e_virtchnl_vector_map *map; + struct virtchnl_irq_map_info *irqmap = + (struct virtchnl_irq_map_info *)msg; + struct virtchnl_vector_map *map; int i; uint16_t vector_id; unsigned long qbit_max; @@ -721,12 +721,12 @@ i40e_pf_host_process_cmd_config_irq_map(struct i40e_pf_vf *vf, if (!b_op) { i40e_pf_host_send_msg_to_vf( vf, - I40E_VIRTCHNL_OP_CONFIG_IRQ_MAP, + VIRTCHNL_OP_CONFIG_IRQ_MAP, I40E_NOT_SUPPORTED, NULL, 0); return ret; } - if (msg == NULL || msglen < sizeof(struct i40e_virtchnl_irq_map_info)) { + if (msg == NULL || msglen < sizeof(struct virtchnl_irq_map_info)) { PMD_DRV_LOG(ERR, "buffer too short"); ret = I40E_ERR_PARAM; goto send_msg; @@ -773,7 +773,7 @@ i40e_pf_host_process_cmd_config_irq_map(struct i40e_pf_vf *vf, } send_msg: - i40e_pf_host_send_msg_to_vf(vf, I40E_VIRTCHNL_OP_CONFIG_IRQ_MAP, + i40e_pf_host_send_msg_to_vf(vf, VIRTCHNL_OP_CONFIG_IRQ_MAP, ret, NULL, 0); return ret; @@ -781,7 +781,7 @@ i40e_pf_host_process_cmd_config_irq_map(struct i40e_pf_vf *vf, static int i40e_pf_host_switch_queues(struct i40e_pf_vf *vf, - struct i40e_virtchnl_queue_select *qsel, + struct virtchnl_queue_select *qsel, bool on) { int ret = I40E_SUCCESS; @@ -831,8 +831,8 @@ i40e_pf_host_process_cmd_enable_queues(struct i40e_pf_vf *vf, uint16_t msglen) { int ret = I40E_SUCCESS; - struct i40e_virtchnl_queue_select *q_sel = - (struct i40e_virtchnl_queue_select *)msg; + struct virtchnl_queue_select *q_sel = + (struct virtchnl_queue_select *)msg; if (msg == NULL || msglen != sizeof(*q_sel)) { ret = I40E_ERR_PARAM; @@ -841,7 +841,7 @@ i40e_pf_host_process_cmd_enable_queues(struct i40e_pf_vf *vf, ret = i40e_pf_host_switch_queues(vf, q_sel, true); send_msg: - i40e_pf_host_send_msg_to_vf(vf, I40E_VIRTCHNL_OP_ENABLE_QUEUES, + i40e_pf_host_send_msg_to_vf(vf, VIRTCHNL_OP_ENABLE_QUEUES, ret, NULL, 0); return ret; @@ -854,13 +854,13 @@ i40e_pf_host_process_cmd_disable_queues(struct i40e_pf_vf *vf, bool b_op) { int ret = I40E_SUCCESS; - struct i40e_virtchnl_queue_select *q_sel = - (struct i40e_virtchnl_queue_select *)msg; + struct virtchnl_queue_select *q_sel = + (struct virtchnl_queue_select *)msg; if (!b_op) { i40e_pf_host_send_msg_to_vf( vf, - I40E_VIRTCHNL_OP_DISABLE_QUEUES, + VIRTCHNL_OP_DISABLE_QUEUES, I40E_NOT_SUPPORTED, NULL, 0); return ret; } @@ -872,7 +872,7 @@ i40e_pf_host_process_cmd_disable_queues(struct i40e_pf_vf *vf, ret = i40e_pf_host_switch_queues(vf, q_sel, false); send_msg: - i40e_pf_host_send_msg_to_vf(vf, I40E_VIRTCHNL_OP_DISABLE_QUEUES, + i40e_pf_host_send_msg_to_vf(vf, VIRTCHNL_OP_DISABLE_QUEUES, ret, NULL, 0); return ret; @@ -886,8 +886,8 @@ i40e_pf_host_process_cmd_add_ether_address(struct i40e_pf_vf *vf, bool b_op) { int ret = I40E_SUCCESS; - struct i40e_virtchnl_ether_addr_list *addr_list = - (struct i40e_virtchnl_ether_addr_list *)msg; + struct virtchnl_ether_addr_list *addr_list = + (struct virtchnl_ether_addr_list *)msg; struct i40e_mac_filter_info filter; int i; struct ether_addr *mac; @@ -895,7 +895,7 @@ i40e_pf_host_process_cmd_add_ether_address(struct i40e_pf_vf *vf, if (!b_op) { i40e_pf_host_send_msg_to_vf( vf, - I40E_VIRTCHNL_OP_ADD_ETHER_ADDRESS, + VIRTCHNL_OP_ADD_ETH_ADDR, I40E_NOT_SUPPORTED, NULL, 0); return ret; } @@ -920,7 +920,7 @@ i40e_pf_host_process_cmd_add_ether_address(struct i40e_pf_vf *vf, } send_msg: - i40e_pf_host_send_msg_to_vf(vf, I40E_VIRTCHNL_OP_ADD_ETHER_ADDRESS, + i40e_pf_host_send_msg_to_vf(vf, VIRTCHNL_OP_ADD_ETH_ADDR, ret, NULL, 0); return ret; @@ -933,15 +933,15 @@ i40e_pf_host_process_cmd_del_ether_address(struct i40e_pf_vf *vf, bool b_op) { int ret = I40E_SUCCESS; - struct i40e_virtchnl_ether_addr_list *addr_list = - (struct i40e_virtchnl_ether_addr_list *)msg; + struct virtchnl_ether_addr_list *addr_list = + (struct virtchnl_ether_addr_list *)msg; int i; struct ether_addr *mac; if (!b_op) { i40e_pf_host_send_msg_to_vf( vf, - I40E_VIRTCHNL_OP_DEL_ETHER_ADDRESS, + VIRTCHNL_OP_DEL_ETH_ADDR, I40E_NOT_SUPPORTED, NULL, 0); return ret; } @@ -962,7 +962,7 @@ i40e_pf_host_process_cmd_del_ether_address(struct i40e_pf_vf *vf, } send_msg: - i40e_pf_host_send_msg_to_vf(vf, I40E_VIRTCHNL_OP_DEL_ETHER_ADDRESS, + i40e_pf_host_send_msg_to_vf(vf, VIRTCHNL_OP_DEL_ETH_ADDR, ret, NULL, 0); return ret; @@ -974,15 +974,15 @@ i40e_pf_host_process_cmd_add_vlan(struct i40e_pf_vf *vf, bool b_op) { int ret = I40E_SUCCESS; - struct i40e_virtchnl_vlan_filter_list *vlan_filter_list = - (struct i40e_virtchnl_vlan_filter_list *)msg; + struct virtchnl_vlan_filter_list *vlan_filter_list = + (struct virtchnl_vlan_filter_list *)msg; int i; uint16_t *vid; if (!b_op) { i40e_pf_host_send_msg_to_vf( vf, - I40E_VIRTCHNL_OP_ADD_VLAN, + VIRTCHNL_OP_ADD_VLAN, I40E_NOT_SUPPORTED, NULL, 0); return ret; } @@ -1002,7 +1002,7 @@ i40e_pf_host_process_cmd_add_vlan(struct i40e_pf_vf *vf, } send_msg: - i40e_pf_host_send_msg_to_vf(vf, I40E_VIRTCHNL_OP_ADD_VLAN, + i40e_pf_host_send_msg_to_vf(vf, VIRTCHNL_OP_ADD_VLAN, ret, NULL, 0); return ret; @@ -1015,15 +1015,15 @@ i40e_pf_host_process_cmd_del_vlan(struct i40e_pf_vf *vf, bool b_op) { int ret = I40E_SUCCESS; - struct i40e_virtchnl_vlan_filter_list *vlan_filter_list = - (struct i40e_virtchnl_vlan_filter_list *)msg; + struct virtchnl_vlan_filter_list *vlan_filter_list = + (struct virtchnl_vlan_filter_list *)msg; int i; uint16_t *vid; if (!b_op) { i40e_pf_host_send_msg_to_vf( vf, - I40E_VIRTCHNL_OP_DEL_VLAN, + VIRTCHNL_OP_DEL_VLAN, I40E_NOT_SUPPORTED, NULL, 0); return ret; } @@ -1042,7 +1042,7 @@ i40e_pf_host_process_cmd_del_vlan(struct i40e_pf_vf *vf, } send_msg: - i40e_pf_host_send_msg_to_vf(vf, I40E_VIRTCHNL_OP_DEL_VLAN, + i40e_pf_host_send_msg_to_vf(vf, VIRTCHNL_OP_DEL_VLAN, ret, NULL, 0); return ret; @@ -1056,15 +1056,15 @@ i40e_pf_host_process_cmd_config_promisc_mode( bool b_op) { int ret = I40E_SUCCESS; - struct i40e_virtchnl_promisc_info *promisc = - (struct i40e_virtchnl_promisc_info *)msg; + struct virtchnl_promisc_info *promisc = + (struct virtchnl_promisc_info *)msg; struct i40e_hw *hw = I40E_PF_TO_HW(vf->pf); bool unicast = FALSE, multicast = FALSE; if (!b_op) { i40e_pf_host_send_msg_to_vf( vf, - I40E_VIRTCHNL_OP_CONFIG_PROMISCUOUS_MODE, + VIRTCHNL_OP_CONFIG_PROMISCUOUS_MODE, I40E_NOT_SUPPORTED, NULL, 0); return ret; } @@ -1074,21 +1074,21 @@ i40e_pf_host_process_cmd_config_promisc_mode( goto send_msg; } - if (promisc->flags & I40E_FLAG_VF_UNICAST_PROMISC) + if (promisc->flags & FLAG_VF_UNICAST_PROMISC) unicast = TRUE; ret = i40e_aq_set_vsi_unicast_promiscuous(hw, vf->vsi->seid, unicast, NULL, true); if (ret != I40E_SUCCESS) goto send_msg; - if (promisc->flags & I40E_FLAG_VF_MULTICAST_PROMISC) + if (promisc->flags & FLAG_VF_MULTICAST_PROMISC) multicast = TRUE; ret = i40e_aq_set_vsi_multicast_promiscuous(hw, vf->vsi->seid, multicast, NULL); send_msg: i40e_pf_host_send_msg_to_vf(vf, - I40E_VIRTCHNL_OP_CONFIG_PROMISCUOUS_MODE, ret, NULL, 0); + VIRTCHNL_OP_CONFIG_PROMISCUOUS_MODE, ret, NULL, 0); return ret; } @@ -1099,12 +1099,12 @@ i40e_pf_host_process_cmd_get_stats(struct i40e_pf_vf *vf, bool b_op) i40e_update_vsi_stats(vf->vsi); if (b_op) - i40e_pf_host_send_msg_to_vf(vf, I40E_VIRTCHNL_OP_GET_STATS, + i40e_pf_host_send_msg_to_vf(vf, VIRTCHNL_OP_GET_STATS, I40E_SUCCESS, (uint8_t *)&vf->vsi->eth_stats, sizeof(vf->vsi->eth_stats)); else - i40e_pf_host_send_msg_to_vf(vf, I40E_VIRTCHNL_OP_GET_STATS, + i40e_pf_host_send_msg_to_vf(vf, VIRTCHNL_OP_GET_STATS, I40E_NOT_SUPPORTED, (uint8_t *)&vf->vsi->eth_stats, sizeof(vf->vsi->eth_stats)); @@ -1120,8 +1120,8 @@ i40e_pf_host_process_cmd_cfg_vlan_offload( bool b_op) { int ret = I40E_SUCCESS; - struct i40e_virtchnl_vlan_offload_info *offload = - (struct i40e_virtchnl_vlan_offload_info *)msg; + struct virtchnl_vlan_offload_info *offload = + (struct virtchnl_vlan_offload_info *)msg; if (!b_op) { i40e_pf_host_send_msg_to_vf( @@ -1155,8 +1155,8 @@ i40e_pf_host_process_cmd_cfg_pvid(struct i40e_pf_vf *vf, bool b_op) { int ret = I40E_SUCCESS; - struct i40e_virtchnl_pvid_info *tpid_info = - (struct i40e_virtchnl_pvid_info *)msg; + struct virtchnl_pvid_info *tpid_info = + (struct virtchnl_pvid_info *)msg; if (!b_op) { i40e_pf_host_send_msg_to_vf( @@ -1183,39 +1183,39 @@ i40e_pf_host_process_cmd_cfg_pvid(struct i40e_pf_vf *vf, void i40e_notify_vf_link_status(struct rte_eth_dev *dev, struct i40e_pf_vf *vf) { - struct i40e_virtchnl_pf_event event; + struct virtchnl_pf_event event; - event.event = I40E_VIRTCHNL_EVENT_LINK_CHANGE; + event.event = VIRTCHNL_EVENT_LINK_CHANGE; event.event_data.link_event.link_status = dev->data->dev_link.link_status; - /* need to convert the ETH_SPEED_xxx into I40E_LINK_SPEED_xxx */ + /* need to convert the ETH_SPEED_xxx into VIRTCHNL_LINK_SPEED_xxx */ switch (dev->data->dev_link.link_speed) { case ETH_SPEED_NUM_100M: - event.event_data.link_event.link_speed = I40E_LINK_SPEED_100MB; + event.event_data.link_event.link_speed = VIRTCHNL_LINK_SPEED_100MB; break; case ETH_SPEED_NUM_1G: - event.event_data.link_event.link_speed = I40E_LINK_SPEED_1GB; + event.event_data.link_event.link_speed = VIRTCHNL_LINK_SPEED_1GB; break; case ETH_SPEED_NUM_10G: - event.event_data.link_event.link_speed = I40E_LINK_SPEED_10GB; + event.event_data.link_event.link_speed = VIRTCHNL_LINK_SPEED_10GB; break; case ETH_SPEED_NUM_20G: - event.event_data.link_event.link_speed = I40E_LINK_SPEED_20GB; + event.event_data.link_event.link_speed = VIRTCHNL_LINK_SPEED_20GB; break; case ETH_SPEED_NUM_25G: - event.event_data.link_event.link_speed = I40E_LINK_SPEED_25GB; + event.event_data.link_event.link_speed = VIRTCHNL_LINK_SPEED_25GB; break; case ETH_SPEED_NUM_40G: - event.event_data.link_event.link_speed = I40E_LINK_SPEED_40GB; + event.event_data.link_event.link_speed = VIRTCHNL_LINK_SPEED_40GB; break; default: event.event_data.link_event.link_speed = - I40E_LINK_SPEED_UNKNOWN; + VIRTCHNL_LINK_SPEED_UNKNOWN; break; } - i40e_pf_host_send_msg_to_vf(vf, I40E_VIRTCHNL_OP_EVENT, + i40e_pf_host_send_msg_to_vf(vf, VIRTCHNL_OP_EVENT, I40E_SUCCESS, (uint8_t *)&event, sizeof(event)); } @@ -1273,71 +1273,71 @@ i40e_pf_host_handle_vf_msg(struct rte_eth_dev *dev, } switch (opcode) { - case I40E_VIRTCHNL_OP_VERSION : + case VIRTCHNL_OP_VERSION: PMD_DRV_LOG(INFO, "OP_VERSION received"); i40e_pf_host_process_cmd_version(vf, b_op); break; - case I40E_VIRTCHNL_OP_RESET_VF : + case VIRTCHNL_OP_RESET_VF: PMD_DRV_LOG(INFO, "OP_RESET_VF received"); i40e_pf_host_process_cmd_reset_vf(vf); break; - case I40E_VIRTCHNL_OP_GET_VF_RESOURCES: + case VIRTCHNL_OP_GET_VF_RESOURCES: PMD_DRV_LOG(INFO, "OP_GET_VF_RESOURCES received"); i40e_pf_host_process_cmd_get_vf_resource(vf, b_op); break; - case I40E_VIRTCHNL_OP_CONFIG_VSI_QUEUES: + case VIRTCHNL_OP_CONFIG_VSI_QUEUES: PMD_DRV_LOG(INFO, "OP_CONFIG_VSI_QUEUES received"); i40e_pf_host_process_cmd_config_vsi_queues(vf, msg, msglen, b_op); break; - case I40E_VIRTCHNL_OP_CONFIG_VSI_QUEUES_EXT: + case VIRTCHNL_OP_CONFIG_VSI_QUEUES_EXT: PMD_DRV_LOG(INFO, "OP_CONFIG_VSI_QUEUES_EXT received"); i40e_pf_host_process_cmd_config_vsi_queues_ext(vf, msg, msglen, b_op); break; - case I40E_VIRTCHNL_OP_CONFIG_IRQ_MAP: + case VIRTCHNL_OP_CONFIG_IRQ_MAP: PMD_DRV_LOG(INFO, "OP_CONFIG_IRQ_MAP received"); i40e_pf_host_process_cmd_config_irq_map(vf, msg, msglen, b_op); break; - case I40E_VIRTCHNL_OP_ENABLE_QUEUES: + case VIRTCHNL_OP_ENABLE_QUEUES: PMD_DRV_LOG(INFO, "OP_ENABLE_QUEUES received"); if (b_op) { i40e_pf_host_process_cmd_enable_queues(vf, msg, msglen); i40e_notify_vf_link_status(dev, vf); } else { i40e_pf_host_send_msg_to_vf( - vf, I40E_VIRTCHNL_OP_ENABLE_QUEUES, + vf, VIRTCHNL_OP_ENABLE_QUEUES, I40E_NOT_SUPPORTED, NULL, 0); } break; - case I40E_VIRTCHNL_OP_DISABLE_QUEUES: + case VIRTCHNL_OP_DISABLE_QUEUES: PMD_DRV_LOG(INFO, "OP_DISABLE_QUEUE received"); i40e_pf_host_process_cmd_disable_queues(vf, msg, msglen, b_op); break; - case I40E_VIRTCHNL_OP_ADD_ETHER_ADDRESS: + case VIRTCHNL_OP_ADD_ETH_ADDR: PMD_DRV_LOG(INFO, "OP_ADD_ETHER_ADDRESS received"); i40e_pf_host_process_cmd_add_ether_address(vf, msg, msglen, b_op); break; - case I40E_VIRTCHNL_OP_DEL_ETHER_ADDRESS: + case VIRTCHNL_OP_DEL_ETH_ADDR: PMD_DRV_LOG(INFO, "OP_DEL_ETHER_ADDRESS received"); i40e_pf_host_process_cmd_del_ether_address(vf, msg, msglen, b_op); break; - case I40E_VIRTCHNL_OP_ADD_VLAN: + case VIRTCHNL_OP_ADD_VLAN: PMD_DRV_LOG(INFO, "OP_ADD_VLAN received"); i40e_pf_host_process_cmd_add_vlan(vf, msg, msglen, b_op); break; - case I40E_VIRTCHNL_OP_DEL_VLAN: + case VIRTCHNL_OP_DEL_VLAN: PMD_DRV_LOG(INFO, "OP_DEL_VLAN received"); i40e_pf_host_process_cmd_del_vlan(vf, msg, msglen, b_op); break; - case I40E_VIRTCHNL_OP_CONFIG_PROMISCUOUS_MODE: + case VIRTCHNL_OP_CONFIG_PROMISCUOUS_MODE: PMD_DRV_LOG(INFO, "OP_CONFIG_PROMISCUOUS_MODE received"); i40e_pf_host_process_cmd_config_promisc_mode(vf, msg, msglen, b_op); break; - case I40E_VIRTCHNL_OP_GET_STATS: + case VIRTCHNL_OP_GET_STATS: PMD_DRV_LOG(INFO, "OP_GET_STATS received"); i40e_pf_host_process_cmd_get_stats(vf, b_op); break; diff --git a/drivers/net/i40e/i40e_pf.h b/drivers/net/i40e/i40e_pf.h index 0961f06..4f93a4d 100644 --- a/drivers/net/i40e/i40e_pf.h +++ b/drivers/net/i40e/i40e_pf.h @@ -35,7 +35,7 @@ #define _I40E_PF_H_ /* VERSION info to exchange between VF and PF host. In case VF works with - * ND kernel driver, it reads I40E_VIRTCHNL_VERSION_MAJOR/MINOR. In + * ND kernel driver, it reads VIRTCHNL_VERSION_MAJOR/MINOR. In * case works with DPDK host, it reads version below. Then VF realize who it * is talking to and use proper language to communicate. * */ @@ -49,45 +49,45 @@ #define I40E_DPDK_OFFSET 0x100 /* DPDK pf driver specific command to VF */ -enum i40e_virtchnl_ops_dpdk { +enum virtchnl_ops_dpdk { /* * Keep some gap between Linux PF commands and * DPDK PF extended commands. */ - I40E_VIRTCHNL_OP_CFG_VLAN_OFFLOAD = I40E_VIRTCHNL_OP_VERSION + + I40E_VIRTCHNL_OP_CFG_VLAN_OFFLOAD = VIRTCHNL_OP_VERSION + I40E_DPDK_OFFSET, I40E_VIRTCHNL_OP_CFG_VLAN_PVID, - I40E_VIRTCHNL_OP_CONFIG_VSI_QUEUES_EXT, + VIRTCHNL_OP_CONFIG_VSI_QUEUES_EXT, }; /* A structure to support extended info of a receive queue. */ -struct i40e_virtchnl_rxq_ext_info { +struct virtchnl_rxq_ext_info { uint8_t crcstrip; }; /* * A structure to support extended info of queue pairs, an additional field - * is added, comparing to original 'struct i40e_virtchnl_queue_pair_info'. + * is added, comparing to original 'struct virtchnl_queue_pair_info'. */ -struct i40e_virtchnl_queue_pair_ext_info { +struct virtchnl_queue_pair_ext_info { /* vsi_id and queue_id should be identical for both rx and tx queues.*/ - struct i40e_virtchnl_txq_info txq; - struct i40e_virtchnl_rxq_info rxq; - struct i40e_virtchnl_rxq_ext_info rxq_ext; + struct virtchnl_txq_info txq; + struct virtchnl_rxq_info rxq; + struct virtchnl_rxq_ext_info rxq_ext; }; /* * A structure to support extended info of VSI queue pairs, - * 'struct i40e_virtchnl_queue_pair_ext_info' is used, see its original - * of 'struct i40e_virtchnl_queue_pair_info'. + * 'struct virtchnl_queue_pair_ext_info' is used, see its original + * of 'struct virtchnl_queue_pair_info'. */ -struct i40e_virtchnl_vsi_queue_config_ext_info { +struct virtchnl_vsi_queue_config_ext_info { uint16_t vsi_id; uint16_t num_queue_pairs; - struct i40e_virtchnl_queue_pair_ext_info qpair[0]; + struct virtchnl_queue_pair_ext_info qpair[0]; }; -struct i40e_virtchnl_vlan_offload_info { +struct virtchnl_vlan_offload_info { uint16_t vsi_id; uint8_t enable_vlan_strip; uint8_t reserved; @@ -106,7 +106,7 @@ struct i40e_virtchnl_vlan_offload_info { * enable op, needs to specify the pvid. PF returns status * code in retval. */ -struct i40e_virtchnl_pvid_info { +struct virtchnl_pvid_info { uint16_t vsi_id; struct i40e_vsi_vlan_pvid_info info; }; diff --git a/drivers/net/i40e/rte_pmd_i40e.h b/drivers/net/i40e/rte_pmd_i40e.h index 1efb2c4..cbe742e 100644 --- a/drivers/net/i40e/rte_pmd_i40e.h +++ b/drivers/net/i40e/rte_pmd_i40e.h @@ -59,7 +59,7 @@ enum rte_pmd_i40e_mb_event_rsp { */ struct rte_pmd_i40e_mb_event_param { uint16_t vfid; /**< Virtual Function number */ - uint16_t msg_type; /**< VF to PF message type, see i40e_virtchnl_ops */ + uint16_t msg_type; /**< VF to PF message type, see virtchnl_ops */ uint16_t retval; /**< return value */ void *msg; /**< pointer to message */ uint16_t msglen; /**< length of the message */ -- 2.4.11 ^ permalink raw reply [flat|nested] 69+ messages in thread
* [dpdk-dev] [PATCH v3 02/17] net/i40e/base: sync nvmupdate command and adminq subtask 2017-06-27 13:29 ` [dpdk-dev] [PATCH v3 00/17] net/i40e: base code update Jingjing Wu 2017-06-27 13:29 ` [dpdk-dev] [PATCH v3 01/17] net/i40e/base: use new virtchnl header file Jingjing Wu @ 2017-06-27 13:29 ` Jingjing Wu 2017-06-27 13:29 ` [dpdk-dev] [PATCH v3 03/17] net/i40e/base: add AQ command for read/write PHY registers Jingjing Wu ` (15 subsequent siblings) 17 siblings, 0 replies; 69+ messages in thread From: Jingjing Wu @ 2017-06-27 13:29 UTC (permalink / raw) To: dev; +Cc: jingjing.wu, helin.zhang During NVMupdate, state machine gets into unrecoverable state because i40e_clean_adminq_subtask can get scheduled after the admin queue command but before other state variables are updated. This patch adds locking around admin queue command and update of state variables so that adminq_subtask will have accurate information whenever it gets scheduled. Signed-off-by: Jingjing Wu <jingjing.wu@intel.com> --- drivers/net/i40e/base/i40e_nvm.c | 6 ++++++ 1 file changed, 6 insertions(+) diff --git a/drivers/net/i40e/base/i40e_nvm.c b/drivers/net/i40e/base/i40e_nvm.c index e896502..90521fa 100644 --- a/drivers/net/i40e/base/i40e_nvm.c +++ b/drivers/net/i40e/base/i40e_nvm.c @@ -899,6 +899,11 @@ enum i40e_status_code i40e_nvmupd_command(struct i40e_hw *hw, hw->nvmupd_state = I40E_NVMUPD_STATE_INIT; } + /* Acquire lock to prevent race condition where adminq_task + * can execute after i40e_nvmupd_nvm_read/write but before state + * variables (nvm_wait_opcode, nvm_release_on_done) are updated + */ + i40e_acquire_spinlock(&hw->aq.arq_spinlock); switch (hw->nvmupd_state) { case I40E_NVMUPD_STATE_INIT: status = i40e_nvmupd_state_init(hw, cmd, bytes, perrno); @@ -934,6 +939,7 @@ enum i40e_status_code i40e_nvmupd_command(struct i40e_hw *hw, *perrno = -ESRCH; break; } + i40e_release_spinlock(&hw->aq.arq_spinlock); return status; } -- 2.4.11 ^ permalink raw reply [flat|nested] 69+ messages in thread
* [dpdk-dev] [PATCH v3 03/17] net/i40e/base: add AQ command for read/write PHY registers 2017-06-27 13:29 ` [dpdk-dev] [PATCH v3 00/17] net/i40e: base code update Jingjing Wu 2017-06-27 13:29 ` [dpdk-dev] [PATCH v3 01/17] net/i40e/base: use new virtchnl header file Jingjing Wu 2017-06-27 13:29 ` [dpdk-dev] [PATCH v3 02/17] net/i40e/base: sync nvmupdate command and adminq subtask Jingjing Wu @ 2017-06-27 13:29 ` Jingjing Wu 2017-06-27 13:29 ` [dpdk-dev] [PATCH v3 04/17] net/i40e/base: add support for Adaptive Virtual Function Jingjing Wu ` (14 subsequent siblings) 17 siblings, 0 replies; 69+ messages in thread From: Jingjing Wu @ 2017-06-27 13:29 UTC (permalink / raw) To: dev; +Cc: jingjing.wu, helin.zhang This patch adds new additional command for accessing to PHY registers. Signed-off-by: Jingjing Wu <jingjing.wu@intel.com> --- drivers/net/i40e/base/i40e_adminq_cmd.h | 18 +++++++++ drivers/net/i40e/base/i40e_common.c | 70 +++++++++++++++++++++++++++++++++ drivers/net/i40e/base/i40e_prototype.h | 9 +++++ 3 files changed, 97 insertions(+) diff --git a/drivers/net/i40e/base/i40e_adminq_cmd.h b/drivers/net/i40e/base/i40e_adminq_cmd.h index 09f5bf5..83b28f8 100644 --- a/drivers/net/i40e/base/i40e_adminq_cmd.h +++ b/drivers/net/i40e/base/i40e_adminq_cmd.h @@ -245,6 +245,8 @@ enum i40e_admin_queue_opc { i40e_aqc_opc_set_phy_debug = 0x0622, i40e_aqc_opc_upload_ext_phy_fm = 0x0625, i40e_aqc_opc_run_phy_activity = 0x0626, + i40e_aqc_opc_set_phy_register = 0x0628, + i40e_aqc_opc_get_phy_register = 0x0629, /* NVM commands */ i40e_aqc_opc_nvm_read = 0x0701, @@ -2128,6 +2130,22 @@ struct i40e_aqc_run_phy_activity { I40E_CHECK_CMD_LENGTH(i40e_aqc_run_phy_activity); +/* Set PHY Register command (0x0628) */ +/* Get PHY Register command (0x0629) */ +struct i40e_aqc_phy_register_access { + u8 phy_interface; +#define I40E_AQ_PHY_REG_ACCESS_INTERNAL 0 +#define I40E_AQ_PHY_REG_ACCESS_EXTERNAL 1 +#define I40E_AQ_PHY_REG_ACCESS_EXTERNAL_MODULE 2 + u8 dev_addres; + u8 reserved1[2]; + u32 reg_address; + u32 reg_value; + u8 reserved2[4]; +}; + +I40E_CHECK_CMD_LENGTH(i40e_aqc_phy_register_access); + /* NVM Read command (indirect 0x0701) * NVM Erase commands (direct 0x0702) * NVM Update commands (indirect 0x0703) diff --git a/drivers/net/i40e/base/i40e_common.c b/drivers/net/i40e/base/i40e_common.c index 6c09c27..280d9da 100644 --- a/drivers/net/i40e/base/i40e_common.c +++ b/drivers/net/i40e/base/i40e_common.c @@ -6863,6 +6863,76 @@ void i40e_write_rx_ctl(struct i40e_hw *hw, u32 reg_addr, u32 reg_val) if (status || use_register) wr32(hw, reg_addr, reg_val); } + +/** + * i40e_aq_set_phy_register + * @hw: pointer to the hw struct + * @phy_select: select which phy should be accessed + * @dev_addr: PHY device address + * @reg_addr: PHY register address + * @reg_val: new register value + * @cmd_details: pointer to command details structure or NULL + * + * Write the external PHY register. + **/ +enum i40e_status_code i40e_aq_set_phy_register(struct i40e_hw *hw, + u8 phy_select, u8 dev_addr, + u32 reg_addr, u32 reg_val, + struct i40e_asq_cmd_details *cmd_details) +{ + struct i40e_aq_desc desc; + struct i40e_aqc_phy_register_access *cmd = + (struct i40e_aqc_phy_register_access *)&desc.params.raw; + enum i40e_status_code status; + + i40e_fill_default_direct_cmd_desc(&desc, + i40e_aqc_opc_set_phy_register); + + cmd->phy_interface = phy_select; + cmd->dev_addres = dev_addr; + cmd->reg_address = reg_addr; + cmd->reg_value = reg_val; + + status = i40e_asq_send_command(hw, &desc, NULL, 0, cmd_details); + + return status; +} + +/** + * i40e_aq_get_phy_register + * @hw: pointer to the hw struct + * @phy_select: select which phy should be accessed + * @dev_addr: PHY device address + * @reg_addr: PHY register address + * @reg_val: read register value + * @cmd_details: pointer to command details structure or NULL + * + * Read the external PHY register. + **/ +enum i40e_status_code i40e_aq_get_phy_register(struct i40e_hw *hw, + u8 phy_select, u8 dev_addr, + u32 reg_addr, u32 *reg_val, + struct i40e_asq_cmd_details *cmd_details) +{ + struct i40e_aq_desc desc; + struct i40e_aqc_phy_register_access *cmd = + (struct i40e_aqc_phy_register_access *)&desc.params.raw; + enum i40e_status_code status; + + i40e_fill_default_direct_cmd_desc(&desc, + i40e_aqc_opc_get_phy_register); + + cmd->phy_interface = phy_select; + cmd->dev_addres = dev_addr; + cmd->reg_address = reg_addr; + + status = i40e_asq_send_command(hw, &desc, NULL, 0, cmd_details); + if (!status) + *reg_val = cmd->reg_value; + + return status; +} + #ifdef VF_DRIVER /** diff --git a/drivers/net/i40e/base/i40e_prototype.h b/drivers/net/i40e/base/i40e_prototype.h index 6ec4304..54b6750 100644 --- a/drivers/net/i40e/base/i40e_prototype.h +++ b/drivers/net/i40e/base/i40e_prototype.h @@ -533,6 +533,15 @@ enum i40e_status_code i40e_aq_rx_ctl_write_register(struct i40e_hw *hw, u32 reg_addr, u32 reg_val, struct i40e_asq_cmd_details *cmd_details); void i40e_write_rx_ctl(struct i40e_hw *hw, u32 reg_addr, u32 reg_val); +enum i40e_status_code i40e_aq_set_phy_register(struct i40e_hw *hw, + u8 phy_select, u8 dev_addr, + u32 reg_addr, u32 reg_val, + struct i40e_asq_cmd_details *cmd_details); +enum i40e_status_code i40e_aq_get_phy_register(struct i40e_hw *hw, + u8 phy_select, u8 dev_addr, + u32 reg_addr, u32 *reg_val, + struct i40e_asq_cmd_details *cmd_details); + enum i40e_status_code i40e_aq_set_arp_proxy_config(struct i40e_hw *hw, struct i40e_aqc_arp_proxy_data *proxy_config, struct i40e_asq_cmd_details *cmd_details); -- 2.4.11 ^ permalink raw reply [flat|nested] 69+ messages in thread
* [dpdk-dev] [PATCH v3 04/17] net/i40e/base: add support for Adaptive Virtual Function 2017-06-27 13:29 ` [dpdk-dev] [PATCH v3 00/17] net/i40e: base code update Jingjing Wu ` (2 preceding siblings ...) 2017-06-27 13:29 ` [dpdk-dev] [PATCH v3 03/17] net/i40e/base: add AQ command for read/write PHY registers Jingjing Wu @ 2017-06-27 13:29 ` Jingjing Wu 2017-06-27 13:29 ` [dpdk-dev] [PATCH v3 05/17] net/i40e/base: store the requested FEC information Jingjing Wu ` (13 subsequent siblings) 17 siblings, 0 replies; 69+ messages in thread From: Jingjing Wu @ 2017-06-27 13:29 UTC (permalink / raw) To: dev; +Cc: jingjing.wu, helin.zhang Add device id define and mac_type assignment needed for Adaptive Virtual Function. Signed-off-by: Jingjing Wu <jingjing.wu@intel.com> --- drivers/net/i40e/base/i40e_common.c | 1 + drivers/net/i40e/base/i40e_devids.h | 1 + 2 files changed, 2 insertions(+) diff --git a/drivers/net/i40e/base/i40e_common.c b/drivers/net/i40e/base/i40e_common.c index 280d9da..5f6dd7b 100644 --- a/drivers/net/i40e/base/i40e_common.c +++ b/drivers/net/i40e/base/i40e_common.c @@ -93,6 +93,7 @@ STATIC enum i40e_status_code i40e_set_mac_type(struct i40e_hw *hw) #if defined(INTEGRATED_VF) || defined(VF_DRIVER) case I40E_DEV_ID_VF: case I40E_DEV_ID_VF_HV: + case I40E_DEV_ID_ADAPTIVE_VF: hw->mac.type = I40E_MAC_VF; break; #endif diff --git a/drivers/net/i40e/base/i40e_devids.h b/drivers/net/i40e/base/i40e_devids.h index 4546689..f4a8784 100644 --- a/drivers/net/i40e/base/i40e_devids.h +++ b/drivers/net/i40e/base/i40e_devids.h @@ -54,6 +54,7 @@ POSSIBILITY OF SUCH DAMAGE. #if defined(INTEGRATED_VF) || defined(VF_DRIVER) || defined(I40E_NDIS_SUPPORT) #define I40E_DEV_ID_VF 0x154C #define I40E_DEV_ID_VF_HV 0x1571 +#define I40E_DEV_ID_ADAPTIVE_VF 0x1889 #endif /* VF_DRIVER */ #ifdef X722_A0_SUPPORT #define I40E_DEV_ID_X722_A0 0x374C -- 2.4.11 ^ permalink raw reply [flat|nested] 69+ messages in thread
* [dpdk-dev] [PATCH v3 05/17] net/i40e/base: store the requested FEC information 2017-06-27 13:29 ` [dpdk-dev] [PATCH v3 00/17] net/i40e: base code update Jingjing Wu ` (3 preceding siblings ...) 2017-06-27 13:29 ` [dpdk-dev] [PATCH v3 04/17] net/i40e/base: add support for Adaptive Virtual Function Jingjing Wu @ 2017-06-27 13:29 ` Jingjing Wu 2017-06-27 13:29 ` [dpdk-dev] [PATCH v3 06/17] net/i40e/base: add new phy types for 25G AOC and ACC Jingjing Wu ` (12 subsequent siblings) 17 siblings, 0 replies; 69+ messages in thread From: Jingjing Wu @ 2017-06-27 13:29 UTC (permalink / raw) To: dev; +Cc: jingjing.wu, helin.zhang Store information about FEC modes, that were requested. It will be used in printing link status information function and this way there is no need to call admin queue there. Signed-off-by: Jingjing Wu <jingjing.wu@intel.com> --- drivers/net/i40e/base/i40e_common.c | 4 ++++ drivers/net/i40e/base/i40e_type.h | 1 + 2 files changed, 5 insertions(+) diff --git a/drivers/net/i40e/base/i40e_common.c b/drivers/net/i40e/base/i40e_common.c index 5f6dd7b..7d59b59 100644 --- a/drivers/net/i40e/base/i40e_common.c +++ b/drivers/net/i40e/base/i40e_common.c @@ -2826,6 +2826,10 @@ enum i40e_status_code i40e_update_link_info(struct i40e_hw *hw) if (status) return status; + hw->phy.link_info.req_fec_info = + abilities.fec_cfg_curr_mod_ext_info & + (I40E_AQ_REQUEST_FEC_KR | I40E_AQ_REQUEST_FEC_RS); + i40e_memcpy(hw->phy.link_info.module_type, &abilities.module_type, sizeof(hw->phy.link_info.module_type), I40E_NONDMA_TO_NONDMA); } diff --git a/drivers/net/i40e/base/i40e_type.h b/drivers/net/i40e/base/i40e_type.h index 52a114a..af5347b 100644 --- a/drivers/net/i40e/base/i40e_type.h +++ b/drivers/net/i40e/base/i40e_type.h @@ -271,6 +271,7 @@ struct i40e_link_status { enum i40e_aq_link_speed link_speed; u8 link_info; u8 an_info; + u8 req_fec_info; u8 fec_info; u8 ext_info; u8 loopback; -- 2.4.11 ^ permalink raw reply [flat|nested] 69+ messages in thread
* [dpdk-dev] [PATCH v3 06/17] net/i40e/base: add new phy types for 25G AOC and ACC 2017-06-27 13:29 ` [dpdk-dev] [PATCH v3 00/17] net/i40e: base code update Jingjing Wu ` (4 preceding siblings ...) 2017-06-27 13:29 ` [dpdk-dev] [PATCH v3 05/17] net/i40e/base: store the requested FEC information Jingjing Wu @ 2017-06-27 13:29 ` Jingjing Wu 2017-06-27 13:29 ` [dpdk-dev] [PATCH v3 07/17] net/i40e/base: report supported link modes Jingjing Wu ` (11 subsequent siblings) 17 siblings, 0 replies; 69+ messages in thread From: Jingjing Wu @ 2017-06-27 13:29 UTC (permalink / raw) To: dev; +Cc: jingjing.wu, helin.zhang This patch adds new phy types for 25G Active Optical Cables (AOC) and Active Copper Cables (ACC) support. Signed-off-by: Jingjing Wu <jingjing.wu@intel.com> --- drivers/net/i40e/base/i40e_adminq_cmd.h | 4 ++++ drivers/net/i40e/base/i40e_common.c | 2 ++ 2 files changed, 6 insertions(+) diff --git a/drivers/net/i40e/base/i40e_adminq_cmd.h b/drivers/net/i40e/base/i40e_adminq_cmd.h index 83b28f8..70079a0 100644 --- a/drivers/net/i40e/base/i40e_adminq_cmd.h +++ b/drivers/net/i40e/base/i40e_adminq_cmd.h @@ -1849,6 +1849,8 @@ enum i40e_aq_phy_type { I40E_PHY_TYPE_25GBASE_CR = 0x20, I40E_PHY_TYPE_25GBASE_SR = 0x21, I40E_PHY_TYPE_25GBASE_LR = 0x22, + I40E_PHY_TYPE_25GBASE_AOC = 0x23, + I40E_PHY_TYPE_25GBASE_ACC = 0x24, I40E_PHY_TYPE_MAX }; @@ -1906,6 +1908,8 @@ struct i40e_aq_get_phy_abilities_resp { #define I40E_AQ_PHY_TYPE_EXT_25G_CR 0x02 #define I40E_AQ_PHY_TYPE_EXT_25G_SR 0x04 #define I40E_AQ_PHY_TYPE_EXT_25G_LR 0x08 +#define I40E_AQ_PHY_TYPE_EXT_25G_AOC 0x10 +#define I40E_AQ_PHY_TYPE_EXT_25G_ACC 0x20 u8 fec_cfg_curr_mod_ext_info; #define I40E_AQ_ENABLE_FEC_KR 0x01 #define I40E_AQ_ENABLE_FEC_RS 0x02 diff --git a/drivers/net/i40e/base/i40e_common.c b/drivers/net/i40e/base/i40e_common.c index 7d59b59..004c062 100644 --- a/drivers/net/i40e/base/i40e_common.c +++ b/drivers/net/i40e/base/i40e_common.c @@ -1296,6 +1296,8 @@ STATIC enum i40e_media_type i40e_get_media_type(struct i40e_hw *hw) case I40E_PHY_TYPE_40GBASE_AOC: case I40E_PHY_TYPE_10GBASE_AOC: case I40E_PHY_TYPE_25GBASE_CR: + case I40E_PHY_TYPE_25GBASE_AOC: + case I40E_PHY_TYPE_25GBASE_ACC: media = I40E_MEDIA_TYPE_DA; break; case I40E_PHY_TYPE_1000BASE_KX: -- 2.4.11 ^ permalink raw reply [flat|nested] 69+ messages in thread
* [dpdk-dev] [PATCH v3 07/17] net/i40e/base: report supported link modes 2017-06-27 13:29 ` [dpdk-dev] [PATCH v3 00/17] net/i40e: base code update Jingjing Wu ` (5 preceding siblings ...) 2017-06-27 13:29 ` [dpdk-dev] [PATCH v3 06/17] net/i40e/base: add new phy types for 25G AOC and ACC Jingjing Wu @ 2017-06-27 13:29 ` Jingjing Wu 2017-06-27 13:29 ` [dpdk-dev] [PATCH v3 08/17] net/i40e/base: track id can be 0 Jingjing Wu ` (10 subsequent siblings) 17 siblings, 0 replies; 69+ messages in thread From: Jingjing Wu @ 2017-06-27 13:29 UTC (permalink / raw) To: dev; +Cc: jingjing.wu, helin.zhang Signed-off-by: Jingjing Wu <jingjing.wu@intel.com> --- drivers/net/i40e/base/i40e_adminq_cmd.h | 22 +++++++++++++++++++--- drivers/net/i40e/base/i40e_common.c | 18 +++++++++++++++--- 2 files changed, 34 insertions(+), 6 deletions(-) diff --git a/drivers/net/i40e/base/i40e_adminq_cmd.h b/drivers/net/i40e/base/i40e_adminq_cmd.h index 70079a0..c1ca290 100644 --- a/drivers/net/i40e/base/i40e_adminq_cmd.h +++ b/drivers/net/i40e/base/i40e_adminq_cmd.h @@ -1831,6 +1831,8 @@ enum i40e_aq_phy_type { I40E_PHY_TYPE_10GBASE_CR1_CU = 0xB, I40E_PHY_TYPE_10GBASE_AOC = 0xC, I40E_PHY_TYPE_40GBASE_AOC = 0xD, + I40E_PHY_TYPE_UNRECOGNIZED = 0xE, + I40E_PHY_TYPE_UNSUPPORTED = 0xF, I40E_PHY_TYPE_100BASE_TX = 0x11, I40E_PHY_TYPE_1000BASE_T = 0x12, I40E_PHY_TYPE_10GBASE_T = 0x13, @@ -1851,7 +1853,9 @@ enum i40e_aq_phy_type { I40E_PHY_TYPE_25GBASE_LR = 0x22, I40E_PHY_TYPE_25GBASE_AOC = 0x23, I40E_PHY_TYPE_25GBASE_ACC = 0x24, - I40E_PHY_TYPE_MAX + I40E_PHY_TYPE_MAX, + I40E_PHY_TYPE_EMPTY = 0xFE, + I40E_PHY_TYPE_DEFAULT = 0xFF, }; #define I40E_LINK_SPEED_100MB_SHIFT 0x1 @@ -2039,19 +2043,31 @@ struct i40e_aqc_get_link_status { #define I40E_AQ_25G_SERDES_UCODE_ERR 0X04 #define I40E_AQ_25G_NIMB_UCODE_ERR 0X05 u8 loopback; /* use defines from i40e_aqc_set_lb_mode */ +/* Since firmware API 1.7 loopback field keeps power class info as well */ +#define I40E_AQ_LOOPBACK_MASK 0x07 +#define I40E_AQ_PWR_CLASS_SHIFT_LB 6 +#define I40E_AQ_PWR_CLASS_MASK_LB (0x03 << I40E_AQ_PWR_CLASS_SHIFT_LB) __le16 max_frame_size; u8 config; #define I40E_AQ_CONFIG_FEC_KR_ENA 0x01 #define I40E_AQ_CONFIG_FEC_RS_ENA 0x02 #define I40E_AQ_CONFIG_CRC_ENA 0x04 #define I40E_AQ_CONFIG_PACING_MASK 0x78 - u8 power_desc; + union { + struct { + u8 power_desc; #define I40E_AQ_LINK_POWER_CLASS_1 0x00 #define I40E_AQ_LINK_POWER_CLASS_2 0x01 #define I40E_AQ_LINK_POWER_CLASS_3 0x02 #define I40E_AQ_LINK_POWER_CLASS_4 0x03 #define I40E_AQ_PWR_CLASS_MASK 0x03 - u8 reserved[4]; + u8 reserved[4]; + }; + struct { + u8 link_type[4]; + u8 link_type_ext; + }; + }; }; I40E_CHECK_CMD_LENGTH(i40e_aqc_get_link_status); diff --git a/drivers/net/i40e/base/i40e_common.c b/drivers/net/i40e/base/i40e_common.c index 004c062..bb4b7eb 100644 --- a/drivers/net/i40e/base/i40e_common.c +++ b/drivers/net/i40e/base/i40e_common.c @@ -1692,8 +1692,14 @@ enum i40e_status_code i40e_aq_get_phy_capabilities(struct i40e_hw *hw, status = I40E_ERR_UNKNOWN_PHY; if (report_init) { - hw->phy.phy_types = LE32_TO_CPU(abilities->phy_type); - hw->phy.phy_types |= ((u64)abilities->phy_type_ext << 32); + if (hw->aq.api_maj_ver == I40E_FW_API_VERSION_MAJOR && + hw->aq.api_min_ver >= 7) { + status = i40e_aq_get_link_info(hw, true, NULL, NULL); + } else { + hw->phy.phy_types = LE32_TO_CPU(abilities->phy_type); + hw->phy.phy_types |= + ((u64)abilities->phy_type_ext << 32); + } } return status; @@ -1955,7 +1961,7 @@ enum i40e_status_code i40e_aq_get_link_info(struct i40e_hw *hw, hw_link_info->fec_info = resp->config & (I40E_AQ_CONFIG_FEC_KR_ENA | I40E_AQ_CONFIG_FEC_RS_ENA); hw_link_info->ext_info = resp->ext_info; - hw_link_info->loopback = resp->loopback; + hw_link_info->loopback = resp->loopback & I40E_AQ_LOOPBACK_MASK; hw_link_info->max_frame_size = LE16_TO_CPU(resp->max_frame_size); hw_link_info->pacing = resp->config & I40E_AQ_CONFIG_PACING_MASK; @@ -1986,6 +1992,12 @@ enum i40e_status_code i40e_aq_get_link_info(struct i40e_hw *hw, hw->aq.fw_min_ver < 40)) && hw_link_info->phy_type == 0xE) hw_link_info->phy_type = I40E_PHY_TYPE_10GBASE_SFPP_CU; + if (hw->aq.api_maj_ver == I40E_FW_API_VERSION_MAJOR && + hw->aq.api_min_ver >= 7) { + hw->phy.phy_types = LE32_TO_CPU(*(__le32 *)resp->link_type); + hw->phy.phy_types |= ((u64)resp->link_type_ext << 32); + } + /* save link status information */ if (link) i40e_memcpy(link, hw_link_info, sizeof(*hw_link_info), -- 2.4.11 ^ permalink raw reply [flat|nested] 69+ messages in thread
* [dpdk-dev] [PATCH v3 08/17] net/i40e/base: track id can be 0 2017-06-27 13:29 ` [dpdk-dev] [PATCH v3 00/17] net/i40e: base code update Jingjing Wu ` (6 preceding siblings ...) 2017-06-27 13:29 ` [dpdk-dev] [PATCH v3 07/17] net/i40e/base: report supported link modes Jingjing Wu @ 2017-06-27 13:29 ` Jingjing Wu 2017-06-27 13:29 ` [dpdk-dev] [PATCH v3 09/17] net/i40e/base: update FW AQ API version to 1.7 Jingjing Wu ` (9 subsequent siblings) 17 siblings, 0 replies; 69+ messages in thread From: Jingjing Wu @ 2017-06-27 13:29 UTC (permalink / raw) To: dev; +Cc: jingjing.wu, helin.zhang track_id == 0 is valid for “read only” profiles when profile does not have any “write” commands. Signed-off-by: Jingjing Wu <jingjing.wu@intel.com> --- drivers/net/i40e/base/i40e_common.c | 7 +------ drivers/net/i40e/base/i40e_prototype.h | 12 +++++++----- 2 files changed, 8 insertions(+), 11 deletions(-) diff --git a/drivers/net/i40e/base/i40e_common.c b/drivers/net/i40e/base/i40e_common.c index bb4b7eb..38c30a1 100644 --- a/drivers/net/i40e/base/i40e_common.c +++ b/drivers/net/i40e/base/i40e_common.c @@ -7370,11 +7370,6 @@ i40e_write_profile(struct i40e_hw *hw, struct i40e_profile_segment *profile, u32 offset = 0, info = 0; u32 i; - if (!track_id) { - i40e_debug(hw, I40E_DEBUG_PACKAGE, "Track_id can't be 0."); - return I40E_NOT_SUPPORTED; - } - dev_cnt = profile->device_table_count; for (i = 0; i < dev_cnt; i++) { @@ -7449,6 +7444,6 @@ i40e_add_pinfo_to_list(struct i40e_hw *hw, memcpy(pinfo->name, profile->name, I40E_DDP_NAME_SIZE); status = i40e_aq_write_ddp(hw, (void *)sec, sec->data_end, - track_id, &offset, &info, NULL); + track_id, &offset, &info, NULL); return status; } diff --git a/drivers/net/i40e/base/i40e_prototype.h b/drivers/net/i40e/base/i40e_prototype.h index 54b6750..9171e97 100644 --- a/drivers/net/i40e/base/i40e_prototype.h +++ b/drivers/net/i40e/base/i40e_prototype.h @@ -575,12 +575,14 @@ u8 i40e_get_phy_address(struct i40e_hw *hw, u8 dev_num); enum i40e_status_code i40e_blink_phy_link_led(struct i40e_hw *hw, u32 time, u32 interval); enum i40e_status_code i40e_aq_write_ddp(struct i40e_hw *hw, void *buff, - u16 buff_size, u32 track_id, - u32 *error_offset, u32 *error_info, - struct i40e_asq_cmd_details *cmd_details); + u16 buff_size, u32 track_id, + u32 *error_offset, u32 *error_info, + struct i40e_asq_cmd_details * + cmd_details); enum i40e_status_code i40e_aq_get_ddp_list(struct i40e_hw *hw, void *buff, - u16 buff_size, u8 flags, - struct i40e_asq_cmd_details *cmd_details); + u16 buff_size, u8 flags, + struct i40e_asq_cmd_details * + cmd_details); struct i40e_generic_seg_header * i40e_find_segment_in_package(u32 segment_type, struct i40e_package_header *pkg_header); -- 2.4.11 ^ permalink raw reply [flat|nested] 69+ messages in thread
* [dpdk-dev] [PATCH v3 09/17] net/i40e/base: update FW AQ API version to 1.7 2017-06-27 13:29 ` [dpdk-dev] [PATCH v3 00/17] net/i40e: base code update Jingjing Wu ` (7 preceding siblings ...) 2017-06-27 13:29 ` [dpdk-dev] [PATCH v3 08/17] net/i40e/base: track id can be 0 Jingjing Wu @ 2017-06-27 13:29 ` Jingjing Wu 2017-06-27 13:29 ` [dpdk-dev] [PATCH v3 10/17] net/i40e/base: add support for switch parameters Jingjing Wu ` (8 subsequent siblings) 17 siblings, 0 replies; 69+ messages in thread From: Jingjing Wu @ 2017-06-27 13:29 UTC (permalink / raw) To: dev; +Cc: jingjing.wu, helin.zhang Signed-off-by: Jingjing Wu <jingjing.wu@intel.com> --- drivers/net/i40e/base/i40e_adminq_cmd.h | 10 +++++++++- drivers/net/i40e/base/i40e_common.c | 5 +++-- drivers/net/i40e/base/i40e_type.h | 4 ---- 3 files changed, 12 insertions(+), 7 deletions(-) diff --git a/drivers/net/i40e/base/i40e_adminq_cmd.h b/drivers/net/i40e/base/i40e_adminq_cmd.h index c1ca290..dd9fb55 100644 --- a/drivers/net/i40e/base/i40e_adminq_cmd.h +++ b/drivers/net/i40e/base/i40e_adminq_cmd.h @@ -41,7 +41,15 @@ POSSIBILITY OF SUCH DAMAGE. */ #define I40E_FW_API_VERSION_MAJOR 0x0001 -#define I40E_FW_API_VERSION_MINOR 0x0005 +#define I40E_FW_API_VERSION_MINOR_X722 0x0005 +#define I40E_FW_API_VERSION_MINOR_X710 0x0007 + +#define I40E_FW_MINOR_VERSION(_h) ((_h)->mac.type == I40E_MAC_XL710 ? \ + I40E_FW_API_VERSION_MINOR_X710 : \ + I40E_FW_API_VERSION_MINOR_X722) + +/* API version 1.7 implements additional link and PHY-specific APIs */ +#define I40E_MINOR_VER_GET_LINK_INFO_XL710 0x0007 struct i40e_aq_desc { __le16 flags; diff --git a/drivers/net/i40e/base/i40e_common.c b/drivers/net/i40e/base/i40e_common.c index 38c30a1..9895a77 100644 --- a/drivers/net/i40e/base/i40e_common.c +++ b/drivers/net/i40e/base/i40e_common.c @@ -1692,8 +1692,9 @@ enum i40e_status_code i40e_aq_get_phy_capabilities(struct i40e_hw *hw, status = I40E_ERR_UNKNOWN_PHY; if (report_init) { - if (hw->aq.api_maj_ver == I40E_FW_API_VERSION_MAJOR && - hw->aq.api_min_ver >= 7) { + if (hw->mac.type == I40E_MAC_XL710 && + hw->aq.api_maj_ver == I40E_FW_API_VERSION_MAJOR && + hw->aq.api_min_ver >= I40E_MINOR_VER_GET_LINK_INFO_XL710) { status = i40e_aq_get_link_info(hw, true, NULL, NULL); } else { hw->phy.phy_types = LE32_TO_CPU(abilities->phy_type); diff --git a/drivers/net/i40e/base/i40e_type.h b/drivers/net/i40e/base/i40e_type.h index af5347b..6ca1a0f 100644 --- a/drivers/net/i40e/base/i40e_type.h +++ b/drivers/net/i40e/base/i40e_type.h @@ -199,10 +199,6 @@ enum i40e_memcpy_type { I40E_DMA_TO_NONDMA }; -#define I40E_FW_API_VERSION_MINOR_X722 0x0005 -#define I40E_FW_API_VERSION_MINOR_X710 0x0005 - - /* These are structs for managing the hardware information and the operations. * The structures of function pointers are filled out at init time when we * know for sure exactly which hardware we're working with. This gives us the -- 2.4.11 ^ permalink raw reply [flat|nested] 69+ messages in thread
* [dpdk-dev] [PATCH v3 10/17] net/i40e/base: add support for switch parameters 2017-06-27 13:29 ` [dpdk-dev] [PATCH v3 00/17] net/i40e: base code update Jingjing Wu ` (8 preceding siblings ...) 2017-06-27 13:29 ` [dpdk-dev] [PATCH v3 09/17] net/i40e/base: update FW AQ API version to 1.7 Jingjing Wu @ 2017-06-27 13:29 ` Jingjing Wu 2017-06-27 13:29 ` [dpdk-dev] [PATCH v3 11/17] net/i40e/base: use admin queue for setting LEDs behavior Jingjing Wu ` (7 subsequent siblings) 17 siblings, 0 replies; 69+ messages in thread From: Jingjing Wu @ 2017-06-27 13:29 UTC (permalink / raw) To: dev; +Cc: jingjing.wu, helin.zhang Adds double VLAN tagging ethertype fields to Set Switch Parameters AQ command. These were added in firmware API 1.7. Callers of i40e_aq_set_switch_config() can specify the ethertypes to use by filling out the corresponding fields in struct i40e_hw. Signed-off-by: Jingjing Wu <jingjing.wu@intel.com> --- drivers/net/i40e/base/i40e_adminq.c | 6 ++++++ drivers/net/i40e/base/i40e_adminq_cmd.h | 17 ++++++++++++++++- drivers/net/i40e/base/i40e_common.c | 6 +++++- drivers/net/i40e/base/i40e_type.h | 6 ++++++ 4 files changed, 33 insertions(+), 2 deletions(-) diff --git a/drivers/net/i40e/base/i40e_adminq.c b/drivers/net/i40e/base/i40e_adminq.c index a60292a..19871c7 100644 --- a/drivers/net/i40e/base/i40e_adminq.c +++ b/drivers/net/i40e/base/i40e_adminq.c @@ -682,6 +682,12 @@ enum i40e_status_code i40e_init_adminq(struct i40e_hw *hw) &oem_lo); hw->nvm.oem_ver = ((u32)oem_hi << 16) | oem_lo; + /* The ability to RX (not drop) 802.1ad frames was added in API 1.7 */ + if ((hw->aq.api_maj_ver > 1) || + ((hw->aq.api_maj_ver == 1) && + (hw->aq.api_min_ver >= 7))) + hw->flags |= I40E_HW_FLAG_802_1AD_CAPABLE; + if (hw->aq.api_maj_ver > I40E_FW_API_VERSION_MAJOR) { ret_code = I40E_ERR_FIRMWARE_API_VERSION; goto init_adminq_free_arq; diff --git a/drivers/net/i40e/base/i40e_adminq_cmd.h b/drivers/net/i40e/base/i40e_adminq_cmd.h index dd9fb55..c36da2a 100644 --- a/drivers/net/i40e/base/i40e_adminq_cmd.h +++ b/drivers/net/i40e/base/i40e_adminq_cmd.h @@ -787,7 +787,22 @@ struct i40e_aqc_set_switch_config { #define I40E_AQ_SET_SWITCH_CFG_L2_FILTER 0x0002 #define I40E_AQ_SET_SWITCH_CFG_HW_ATR_EVICT 0x0004 __le16 valid_flags; - u8 reserved[12]; + /* The ethertype in switch_tag is dropped on ingress and used + * internally by the switch. Set this to zero for the default + * of 0x88a8 (802.1ad). Should be zero for firmware API + * versions lower than 1.7. + */ + __le16 switch_tag; + /* The ethertypes in first_tag and second_tag are used to + * match the outer and inner VLAN tags (respectively) when HW + * double VLAN tagging is enabled via the set port parameters + * AQ command. Otherwise these are both ignored. Set them to + * zero for their defaults of 0x8100 (802.1Q). Should be zero + * for firmware API versions lower than 1.7. + */ + __le16 first_tag; + __le16 second_tag; + u8 reserved[6]; }; I40E_CHECK_CMD_LENGTH(i40e_aqc_set_switch_config); diff --git a/drivers/net/i40e/base/i40e_common.c b/drivers/net/i40e/base/i40e_common.c index 9895a77..b4901ef 100644 --- a/drivers/net/i40e/base/i40e_common.c +++ b/drivers/net/i40e/base/i40e_common.c @@ -2695,7 +2695,11 @@ enum i40e_status_code i40e_aq_set_switch_config(struct i40e_hw *hw, i40e_aqc_opc_set_switch_config); scfg->flags = CPU_TO_LE16(flags); scfg->valid_flags = CPU_TO_LE16(valid_flags); - + if (hw->flags & I40E_HW_FLAG_802_1AD_CAPABLE) { + scfg->switch_tag = CPU_TO_LE16(hw->switch_tag); + scfg->first_tag = CPU_TO_LE16(hw->first_tag); + scfg->second_tag = CPU_TO_LE16(hw->second_tag); + } status = i40e_asq_send_command(hw, &desc, NULL, 0, cmd_details); return status; diff --git a/drivers/net/i40e/base/i40e_type.h b/drivers/net/i40e/base/i40e_type.h index 6ca1a0f..152b4a7 100644 --- a/drivers/net/i40e/base/i40e_type.h +++ b/drivers/net/i40e/base/i40e_type.h @@ -700,8 +700,14 @@ struct i40e_hw { u16 wol_proxy_vsi_seid; #define I40E_HW_FLAG_AQ_SRCTL_ACCESS_ENABLE BIT_ULL(0) +#define I40E_HW_FLAG_802_1AD_CAPABLE BIT_ULL(1) u64 flags; + /* Used in set switch config AQ command */ + u16 switch_tag; + u16 first_tag; + u16 second_tag; + /* debug mask */ u32 debug_mask; char err_str[16]; -- 2.4.11 ^ permalink raw reply [flat|nested] 69+ messages in thread
* [dpdk-dev] [PATCH v3 11/17] net/i40e/base: use admin queue for setting LEDs behavior 2017-06-27 13:29 ` [dpdk-dev] [PATCH v3 00/17] net/i40e: base code update Jingjing Wu ` (9 preceding siblings ...) 2017-06-27 13:29 ` [dpdk-dev] [PATCH v3 10/17] net/i40e/base: add support for switch parameters Jingjing Wu @ 2017-06-27 13:29 ` Jingjing Wu 2017-06-27 13:29 ` [dpdk-dev] [PATCH v3 12/17] net/i40e/base: avoid potential null pointer dereference Jingjing Wu ` (6 subsequent siblings) 17 siblings, 0 replies; 69+ messages in thread From: Jingjing Wu @ 2017-06-27 13:29 UTC (permalink / raw) To: dev; +Cc: jingjing.wu, helin.zhang Instead of accessing register directly, use newly added AQC in order to blink LEDs. Introduce and utilize a new flag to prevent excessive API version checking. Signed-off-by: Jingjing Wu <jingjing.wu@intel.com> --- drivers/net/i40e/base/i40e_adminq.c | 6 ++ drivers/net/i40e/base/i40e_common.c | 148 ++++++++++++++++++++++++++-------- drivers/net/i40e/base/i40e_register.h | 2 +- drivers/net/i40e/base/i40e_type.h | 1 + 4 files changed, 121 insertions(+), 36 deletions(-) diff --git a/drivers/net/i40e/base/i40e_adminq.c b/drivers/net/i40e/base/i40e_adminq.c index 19871c7..8cc8c5e 100644 --- a/drivers/net/i40e/base/i40e_adminq.c +++ b/drivers/net/i40e/base/i40e_adminq.c @@ -688,6 +688,12 @@ enum i40e_status_code i40e_init_adminq(struct i40e_hw *hw) (hw->aq.api_min_ver >= 7))) hw->flags |= I40E_HW_FLAG_802_1AD_CAPABLE; + if (hw->mac.type == I40E_MAC_XL710 && + hw->aq.api_maj_ver == I40E_FW_API_VERSION_MAJOR && + hw->aq.api_min_ver >= I40E_MINOR_VER_GET_LINK_INFO_XL710) { + hw->flags |= I40E_HW_FLAG_AQ_PHY_ACCESS_CAPABLE; + } + if (hw->aq.api_maj_ver > I40E_FW_API_VERSION_MAJOR) { ret_code = I40E_ERR_FIRMWARE_API_VERSION; goto init_adminq_free_arq; diff --git a/drivers/net/i40e/base/i40e_common.c b/drivers/net/i40e/base/i40e_common.c index b4901ef..3f3b275 100644 --- a/drivers/net/i40e/base/i40e_common.c +++ b/drivers/net/i40e/base/i40e_common.c @@ -6678,24 +6678,38 @@ enum i40e_status_code i40e_led_get_phy(struct i40e_hw *hw, u16 *led_addr, u16 temp_addr; u8 port_num; u32 i; - - temp_addr = I40E_PHY_LED_PROV_REG_1; - i = rd32(hw, I40E_PFGEN_PORTNUM); - port_num = (u8)(i & I40E_PFGEN_PORTNUM_PORT_NUM_MASK); - phy_addr = i40e_get_phy_address(hw, port_num); - - for (gpio_led_port = 0; gpio_led_port < 3; gpio_led_port++, - temp_addr++) { - status = i40e_read_phy_register_clause45(hw, + u32 reg_val_aq; + + if (hw->flags & I40E_HW_FLAG_AQ_PHY_ACCESS_CAPABLE) { + status = + i40e_aq_get_phy_register(hw, + I40E_AQ_PHY_REG_ACCESS_EXTERNAL, + I40E_PHY_COM_REG_PAGE, + I40E_PHY_LED_PROV_REG_1, + ®_val_aq, NULL); + if (status) + return status; + *val = (u16)reg_val_aq; + } else { + temp_addr = I40E_PHY_LED_PROV_REG_1; + i = rd32(hw, I40E_PFGEN_PORTNUM); + port_num = (u8)(i & I40E_PFGEN_PORTNUM_PORT_NUM_MASK); + phy_addr = i40e_get_phy_address(hw, port_num); + + for (gpio_led_port = 0; gpio_led_port < 3; gpio_led_port++, + temp_addr++) { + status = + i40e_read_phy_register_clause45(hw, I40E_PHY_COM_REG_PAGE, temp_addr, phy_addr, ®_val); - if (status) - return status; - *val = reg_val; - if (reg_val & I40E_PHY_LED_LINK_MODE_MASK) { - *led_addr = temp_addr; - break; + if (status) + return status; + *val = reg_val; + if (reg_val & I40E_PHY_LED_LINK_MODE_MASK) { + *led_addr = temp_addr; + break; + } } } return status; @@ -6713,51 +6727,115 @@ enum i40e_status_code i40e_led_set_phy(struct i40e_hw *hw, bool on, u16 led_addr, u32 mode) { enum i40e_status_code status = I40E_SUCCESS; - u16 led_ctl = 0; - u16 led_reg = 0; + u32 led_ctl = 0; + u32 led_reg = 0; u8 phy_addr = 0; u8 port_num; u32 i; - i = rd32(hw, I40E_PFGEN_PORTNUM); - port_num = (u8)(i & I40E_PFGEN_PORTNUM_PORT_NUM_MASK); - phy_addr = i40e_get_phy_address(hw, port_num); - status = i40e_read_phy_register_clause45(hw, I40E_PHY_COM_REG_PAGE, - led_addr, phy_addr, &led_reg); + if (hw->flags & I40E_HW_FLAG_AQ_PHY_ACCESS_CAPABLE) { + status = + i40e_aq_get_phy_register(hw, + I40E_AQ_PHY_REG_ACCESS_EXTERNAL, + I40E_PHY_COM_REG_PAGE, + I40E_PHY_LED_PROV_REG_1, + &led_reg, NULL); + } else { + i = rd32(hw, I40E_PFGEN_PORTNUM); + port_num = (u8)(i & I40E_PFGEN_PORTNUM_PORT_NUM_MASK); + phy_addr = i40e_get_phy_address(hw, port_num); + status = i40e_read_phy_register_clause45(hw, + I40E_PHY_COM_REG_PAGE, + led_addr, phy_addr, + (u16 *)&led_reg); + } if (status) return status; led_ctl = led_reg; if (led_reg & I40E_PHY_LED_LINK_MODE_MASK) { led_reg = 0; - status = i40e_write_phy_register_clause45(hw, - I40E_PHY_COM_REG_PAGE, - led_addr, phy_addr, - led_reg); + if (hw->aq.api_maj_ver == I40E_FW_API_VERSION_MAJOR && + hw->aq.api_min_ver >= I40E_MINOR_VER_GET_LINK_INFO_XL710) { + status = i40e_aq_set_phy_register(hw, + I40E_AQ_PHY_REG_ACCESS_EXTERNAL, + I40E_PHY_COM_REG_PAGE, + I40E_PHY_LED_PROV_REG_1, + led_reg, NULL); + } else { + status = i40e_write_phy_register_clause45(hw, + I40E_PHY_COM_REG_PAGE, + led_addr, phy_addr, + (u16)led_reg); + } if (status) return status; } - status = i40e_read_phy_register_clause45(hw, I40E_PHY_COM_REG_PAGE, - led_addr, phy_addr, &led_reg); + if (hw->flags & I40E_HW_FLAG_AQ_PHY_ACCESS_CAPABLE) { + status = + i40e_aq_get_phy_register(hw, + I40E_AQ_PHY_REG_ACCESS_EXTERNAL, + I40E_PHY_COM_REG_PAGE, + I40E_PHY_LED_PROV_REG_1, + &led_reg, NULL); + } else { + status = i40e_read_phy_register_clause45(hw, + I40E_PHY_COM_REG_PAGE, + led_addr, phy_addr, + (u16 *)&led_reg); + } if (status) goto restore_config; if (on) led_reg = I40E_PHY_LED_MANUAL_ON; else led_reg = 0; - status = i40e_write_phy_register_clause45(hw, I40E_PHY_COM_REG_PAGE, - led_addr, phy_addr, led_reg); + + if (hw->flags & I40E_HW_FLAG_AQ_PHY_ACCESS_CAPABLE) { + status = + i40e_aq_set_phy_register(hw, + I40E_AQ_PHY_REG_ACCESS_EXTERNAL, + I40E_PHY_COM_REG_PAGE, + I40E_PHY_LED_PROV_REG_1, + led_reg, NULL); + } else { + status = + i40e_write_phy_register_clause45(hw, I40E_PHY_COM_REG_PAGE, + led_addr, phy_addr, + (u16)led_reg); + } if (status) goto restore_config; if (mode & I40E_PHY_LED_MODE_ORIG) { led_ctl = (mode & I40E_PHY_LED_MODE_MASK); - status = i40e_write_phy_register_clause45(hw, - I40E_PHY_COM_REG_PAGE, - led_addr, phy_addr, led_ctl); + if (hw->flags & I40E_HW_FLAG_AQ_PHY_ACCESS_CAPABLE) { + status = i40e_aq_set_phy_register(hw, + I40E_AQ_PHY_REG_ACCESS_EXTERNAL, + I40E_PHY_COM_REG_PAGE, + I40E_PHY_LED_PROV_REG_1, + led_ctl, NULL); + } else { + status = i40e_write_phy_register_clause45(hw, + I40E_PHY_COM_REG_PAGE, + led_addr, phy_addr, + (u16)led_ctl); + } } return status; restore_config: - status = i40e_write_phy_register_clause45(hw, I40E_PHY_COM_REG_PAGE, - led_addr, phy_addr, led_ctl); + if (hw->flags & I40E_HW_FLAG_AQ_PHY_ACCESS_CAPABLE) { + status = + i40e_aq_set_phy_register(hw, + I40E_AQ_PHY_REG_ACCESS_EXTERNAL, + I40E_PHY_COM_REG_PAGE, + I40E_PHY_LED_PROV_REG_1, + led_ctl, NULL); + } else { + status = + i40e_write_phy_register_clause45(hw, + I40E_PHY_COM_REG_PAGE, + led_addr, phy_addr, + (u16)led_ctl); + } return status; } #endif /* PF_DRIVER */ diff --git a/drivers/net/i40e/base/i40e_register.h b/drivers/net/i40e/base/i40e_register.h index b150fbd..a482ab9 100644 --- a/drivers/net/i40e/base/i40e_register.h +++ b/drivers/net/i40e/base/i40e_register.h @@ -2805,7 +2805,7 @@ POSSIBILITY OF SUCH DAMAGE. #define I40E_GLV_RUPP_MAX_INDEX 383 #define I40E_GLV_RUPP_RUPP_SHIFT 0 #define I40E_GLV_RUPP_RUPP_MASK I40E_MASK(0xFFFFFFFF, I40E_GLV_RUPP_RUPP_SHIFT) -#define I40E_GLV_TEPC(_VSI) (0x00344000 + ((_VSI) * 8)) /* _i=0...383 */ /* Reset: CORER */ +#define I40E_GLV_TEPC(_i) (0x00344000 + ((_i) * 8)) /* _i=0...383 */ /* Reset: CORER */ #define I40E_GLV_TEPC_MAX_INDEX 383 #define I40E_GLV_TEPC_TEPC_SHIFT 0 #define I40E_GLV_TEPC_TEPC_MASK I40E_MASK(0xFFFFFFFF, I40E_GLV_TEPC_TEPC_SHIFT) diff --git a/drivers/net/i40e/base/i40e_type.h b/drivers/net/i40e/base/i40e_type.h index 152b4a7..9d7b1a2 100644 --- a/drivers/net/i40e/base/i40e_type.h +++ b/drivers/net/i40e/base/i40e_type.h @@ -701,6 +701,7 @@ struct i40e_hw { #define I40E_HW_FLAG_AQ_SRCTL_ACCESS_ENABLE BIT_ULL(0) #define I40E_HW_FLAG_802_1AD_CAPABLE BIT_ULL(1) +#define I40E_HW_FLAG_AQ_PHY_ACCESS_CAPABLE BIT_ULL(2) u64 flags; /* Used in set switch config AQ command */ -- 2.4.11 ^ permalink raw reply [flat|nested] 69+ messages in thread
* [dpdk-dev] [PATCH v3 12/17] net/i40e/base: avoid potential null pointer dereference 2017-06-27 13:29 ` [dpdk-dev] [PATCH v3 00/17] net/i40e: base code update Jingjing Wu ` (10 preceding siblings ...) 2017-06-27 13:29 ` [dpdk-dev] [PATCH v3 11/17] net/i40e/base: use admin queue for setting LEDs behavior Jingjing Wu @ 2017-06-27 13:29 ` Jingjing Wu 2017-06-27 13:29 ` [dpdk-dev] [PATCH v3 13/17] net/i40e/base: avoid reset timeout issue Jingjing Wu ` (5 subsequent siblings) 17 siblings, 0 replies; 69+ messages in thread From: Jingjing Wu @ 2017-06-27 13:29 UTC (permalink / raw) To: dev; +Cc: jingjing.wu, helin.zhang Signed-off-by: Jingjing Wu <jingjing.wu@intel.com> --- drivers/net/i40e/base/i40e_common.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) diff --git a/drivers/net/i40e/base/i40e_common.c b/drivers/net/i40e/base/i40e_common.c index 3f3b275..78ed2a8 100644 --- a/drivers/net/i40e/base/i40e_common.c +++ b/drivers/net/i40e/base/i40e_common.c @@ -330,13 +330,15 @@ void i40e_debug_aq(struct i40e_hw *hw, enum i40e_debug_mask mask, void *desc, void *buffer, u16 buf_len) { struct i40e_aq_desc *aq_desc = (struct i40e_aq_desc *)desc; - u16 len = LE16_TO_CPU(aq_desc->datalen); u8 *buf = (u8 *)buffer; + u16 len; u16 i = 0; if ((!(mask & hw->debug_mask)) || (desc == NULL)) return; + len = LE16_TO_CPU(aq_desc->datalen); + i40e_debug(hw, mask, "AQ CMD: opcode 0x%04X, flags 0x%04X, datalen 0x%04X, retval 0x%04X\n", LE16_TO_CPU(aq_desc->opcode), -- 2.4.11 ^ permalink raw reply [flat|nested] 69+ messages in thread
* [dpdk-dev] [PATCH v3 13/17] net/i40e/base: avoid reset timeout issue 2017-06-27 13:29 ` [dpdk-dev] [PATCH v3 00/17] net/i40e: base code update Jingjing Wu ` (11 preceding siblings ...) 2017-06-27 13:29 ` [dpdk-dev] [PATCH v3 12/17] net/i40e/base: avoid potential null pointer dereference Jingjing Wu @ 2017-06-27 13:29 ` Jingjing Wu 2017-06-27 13:29 ` [dpdk-dev] [PATCH v3 14/17] net/i40e/base: add EEPROM checksum verification Jingjing Wu ` (4 subsequent siblings) 17 siblings, 0 replies; 69+ messages in thread From: Jingjing Wu @ 2017-06-27 13:29 UTC (permalink / raw) To: dev; +Cc: jingjing.wu, helin.zhang This patch allows detection of upcoming core reset in case NIC gets stuck while performing FLR reset. The i40e_pf_reset() function returns I40E_ERR_NOT_READY when global reset was detected. Signed-off-by: Jingjing Wu <jingjing.wu@intel.com> --- drivers/net/i40e/base/i40e_common.c | 8 ++++++++ 1 file changed, 8 insertions(+) diff --git a/drivers/net/i40e/base/i40e_common.c b/drivers/net/i40e/base/i40e_common.c index 78ed2a8..ed2e01a 100644 --- a/drivers/net/i40e/base/i40e_common.c +++ b/drivers/net/i40e/base/i40e_common.c @@ -1382,6 +1382,8 @@ enum i40e_status_code i40e_pf_reset(struct i40e_hw *hw) * we don't need to do the PF Reset */ if (!cnt) { + u32 reg2 = 0; + reg = rd32(hw, I40E_PFGEN_CTRL); wr32(hw, I40E_PFGEN_CTRL, (reg | I40E_PFGEN_CTRL_PFSWR_MASK)); @@ -1389,6 +1391,12 @@ enum i40e_status_code i40e_pf_reset(struct i40e_hw *hw) reg = rd32(hw, I40E_PFGEN_CTRL); if (!(reg & I40E_PFGEN_CTRL_PFSWR_MASK)) break; + reg2 = rd32(hw, I40E_GLGEN_RSTAT); + if (reg2 & I40E_GLGEN_RSTAT_DEVSTATE_MASK) { + DEBUGOUT("Core reset upcoming.\n"); + DEBUGOUT1("I40E_GLGEN_RSTAT = 0x%x\n", reg2); + return I40E_ERR_NOT_READY; + } i40e_msec_delay(1); } if (reg & I40E_PFGEN_CTRL_PFSWR_MASK) { -- 2.4.11 ^ permalink raw reply [flat|nested] 69+ messages in thread
* [dpdk-dev] [PATCH v3 14/17] net/i40e/base: add EEPROM checksum verification 2017-06-27 13:29 ` [dpdk-dev] [PATCH v3 00/17] net/i40e: base code update Jingjing Wu ` (12 preceding siblings ...) 2017-06-27 13:29 ` [dpdk-dev] [PATCH v3 13/17] net/i40e/base: avoid reset timeout issue Jingjing Wu @ 2017-06-27 13:29 ` Jingjing Wu 2017-06-27 13:29 ` [dpdk-dev] [PATCH v3 15/17] net/i40e/base: extend processing of DDP Jingjing Wu ` (3 subsequent siblings) 17 siblings, 0 replies; 69+ messages in thread From: Jingjing Wu @ 2017-06-27 13:29 UTC (permalink / raw) To: dev; +Cc: jingjing.wu, helin.zhang This patch ensures PFs mutually exclusive access to NVM. Signed-off-by: Jingjing Wu <jingjing.wu@intel.com> --- drivers/net/i40e/base/i40e_nvm.c | 14 ++++++++++---- 1 file changed, 10 insertions(+), 4 deletions(-) diff --git a/drivers/net/i40e/base/i40e_nvm.c b/drivers/net/i40e/base/i40e_nvm.c index 90521fa..a1e7830 100644 --- a/drivers/net/i40e/base/i40e_nvm.c +++ b/drivers/net/i40e/base/i40e_nvm.c @@ -749,12 +749,18 @@ enum i40e_status_code i40e_validate_nvm_checksum(struct i40e_hw *hw, DEBUGFUNC("i40e_validate_nvm_checksum"); - if (hw->flags & I40E_HW_FLAG_AQ_SRCTL_ACCESS_ENABLE) - ret_code = i40e_acquire_nvm(hw, I40E_RESOURCE_READ); + /* acquire_nvm provides exclusive NVM lock to synchronize access across + * PFs. X710 uses i40e_read_nvm_word_srctl which polls for done bit + * twice (first time to be able to write address to I40E_GLNVM_SRCTL + * register, second to read data from I40E_GLNVM_SRDATA. One PF can see + * done bit and try to write address, while another one will interpret + * it as a good time to read data. It will cause invalid data to be + * read. + */ + ret_code = i40e_acquire_nvm(hw, I40E_RESOURCE_READ); if (!ret_code) { ret_code = i40e_calc_nvm_checksum(hw, &checksum_local); - if (hw->flags & I40E_HW_FLAG_AQ_SRCTL_ACCESS_ENABLE) - i40e_release_nvm(hw); + i40e_release_nvm(hw); if (ret_code != I40E_SUCCESS) goto i40e_validate_nvm_checksum_exit; } else { -- 2.4.11 ^ permalink raw reply [flat|nested] 69+ messages in thread
* [dpdk-dev] [PATCH v3 15/17] net/i40e/base: extend processing of DDP 2017-06-27 13:29 ` [dpdk-dev] [PATCH v3 00/17] net/i40e: base code update Jingjing Wu ` (13 preceding siblings ...) 2017-06-27 13:29 ` [dpdk-dev] [PATCH v3 14/17] net/i40e/base: add EEPROM checksum verification Jingjing Wu @ 2017-06-27 13:29 ` Jingjing Wu 2017-06-27 13:29 ` [dpdk-dev] [PATCH v3 16/17] net/i40e: use set switch aq instead of register setting Jingjing Wu ` (2 subsequent siblings) 17 siblings, 0 replies; 69+ messages in thread From: Jingjing Wu @ 2017-06-27 13:29 UTC (permalink / raw) To: dev; +Cc: jingjing.wu, helin.zhang, Andrey Chilikin, Beilei Xing This patch adds extended processing of DDP packages: - Execution of adminq command sections to support AQ-depended profiles, for example, for programming cloud filters types. - Ability to write a profile without registering it in the list of applied profiles, to be used for AQ-depended profiles. - Profile rollback is implemented to support restoration of original parser/analyzer configuration without the need of core reset, for example, for deploying new profile without resetting device. - Search for a specific section in a profile, to be used by driver to access metadata sections with description of PCTYPE/PTYPEs defined in the profile. Signed-off-by: Andrey Chilikin <andrey.chilikin@intel.com> Signed-off-by: Beilei Xing <beilei.xing@intel.com> Signed-off-by: Jingjing Wu <jingjing.wu@intel.com> --- drivers/net/i40e/base/i40e_common.c | 258 ++++++++++++++++++++++++++++++--- drivers/net/i40e/base/i40e_prototype.h | 6 + drivers/net/i40e/base/i40e_type.h | 25 +++- 3 files changed, 265 insertions(+), 24 deletions(-) diff --git a/drivers/net/i40e/base/i40e_common.c b/drivers/net/i40e/base/i40e_common.c index ed2e01a..900d379 100644 --- a/drivers/net/i40e/base/i40e_common.c +++ b/drivers/net/i40e/base/i40e_common.c @@ -7441,6 +7441,165 @@ i40e_find_segment_in_package(u32 segment_type, return NULL; } +/* Get section table in profile */ +#define I40E_SECTION_TABLE(profile, sec_tbl) \ + do { \ + struct i40e_profile_segment *p = (profile); \ + u32 count; \ + u32 *nvm; \ + count = p->device_table_count; \ + nvm = (u32 *)&p->device_table[count]; \ + sec_tbl = (struct i40e_section_table *)&nvm[nvm[0] + 1]; \ + } while (0) + +/* Get section header in profile */ +#define I40E_SECTION_HEADER(profile, offset) \ + (struct i40e_profile_section_header *)((u8 *)(profile) + (offset)) + +/** + * i40e_find_section_in_profile + * @section_type: the section type to search for (i.e., SECTION_TYPE_NOTE) + * @profile: pointer to the i40e segment header to be searched + * + * This function searches i40e segment for a particular section type. On + * success it returns a pointer to the section header, otherwise it will + * return NULL. + **/ +struct i40e_profile_section_header * +i40e_find_section_in_profile(u32 section_type, + struct i40e_profile_segment *profile) +{ + struct i40e_profile_section_header *sec; + struct i40e_section_table *sec_tbl; + u32 sec_off; + u32 i; + + if (profile->header.type != SEGMENT_TYPE_I40E) + return NULL; + + I40E_SECTION_TABLE(profile, sec_tbl); + + for (i = 0; i < sec_tbl->section_count; i++) { + sec_off = sec_tbl->section_offset[i]; + sec = I40E_SECTION_HEADER(profile, sec_off); + if (sec->section.type == section_type) + return sec; + } + + return NULL; +} + +/** + * i40e_ddp_exec_aq_section - Execute generic AQ for DDP + * @hw: pointer to the hw struct + * @aq: command buffer containing all data to execute AQ + **/ +STATIC enum +i40e_status_code i40e_ddp_exec_aq_section(struct i40e_hw *hw, + struct i40e_profile_aq_section *aq) +{ + enum i40e_status_code status; + struct i40e_aq_desc desc; + u8 *msg = NULL; + u16 msglen; + + i40e_fill_default_direct_cmd_desc(&desc, aq->opcode); + desc.flags |= CPU_TO_LE16(aq->flags); + i40e_memcpy(desc.params.raw, aq->param, sizeof(desc.params.raw), + I40E_NONDMA_TO_NONDMA); + + msglen = aq->datalen; + if (msglen) { + desc.flags |= CPU_TO_LE16((u16)(I40E_AQ_FLAG_BUF | + I40E_AQ_FLAG_RD)); + if (msglen > I40E_AQ_LARGE_BUF) + desc.flags |= CPU_TO_LE16((u16)I40E_AQ_FLAG_LB); + desc.datalen = CPU_TO_LE16(msglen); + msg = &aq->data[0]; + } + + status = i40e_asq_send_command(hw, &desc, msg, msglen, NULL); + + if (status != I40E_SUCCESS) { + i40e_debug(hw, I40E_DEBUG_PACKAGE, + "unable to exec DDP AQ opcode %u, error %d\n", + aq->opcode, status); + return status; + } + + /* copy returned desc to aq_buf */ + i40e_memcpy(aq->param, desc.params.raw, sizeof(desc.params.raw), + I40E_NONDMA_TO_NONDMA); + + return I40E_SUCCESS; +} + +/** + * i40e_validate_profile + * @hw: pointer to the hardware structure + * @profile: pointer to the profile segment of the package to be validated + * @track_id: package tracking id + * @rollback: flag if the profile is for rollback. + * + * Validates supported devices and profile's sections. + */ +STATIC enum i40e_status_code +i40e_validate_profile(struct i40e_hw *hw, struct i40e_profile_segment *profile, + u32 track_id, bool rollback) +{ + struct i40e_profile_section_header *sec = NULL; + enum i40e_status_code status = I40E_SUCCESS; + struct i40e_section_table *sec_tbl; + u32 vendor_dev_id; + u32 dev_cnt; + u32 sec_off; + u32 i; + + if (track_id == I40E_DDP_TRACKID_INVALID) { + i40e_debug(hw, I40E_DEBUG_PACKAGE, "Invalid track_id\n"); + return I40E_NOT_SUPPORTED; + } + + dev_cnt = profile->device_table_count; + for (i = 0; i < dev_cnt; i++) { + vendor_dev_id = profile->device_table[i].vendor_dev_id; + if ((vendor_dev_id >> 16) == I40E_INTEL_VENDOR_ID && + hw->device_id == (vendor_dev_id & 0xFFFF)) + break; + } + if (dev_cnt && (i == dev_cnt)) { + i40e_debug(hw, I40E_DEBUG_PACKAGE, + "Device doesn't support DDP\n"); + return I40E_ERR_DEVICE_NOT_SUPPORTED; + } + + I40E_SECTION_TABLE(profile, sec_tbl); + + /* Validate sections types */ + for (i = 0; i < sec_tbl->section_count; i++) { + sec_off = sec_tbl->section_offset[i]; + sec = I40E_SECTION_HEADER(profile, sec_off); + if (rollback) { + if (sec->section.type == SECTION_TYPE_MMIO || + sec->section.type == SECTION_TYPE_AQ || + sec->section.type == SECTION_TYPE_RB_AQ) { + i40e_debug(hw, I40E_DEBUG_PACKAGE, + "Not a roll-back package\n"); + return I40E_NOT_SUPPORTED; + } + } else { + if (sec->section.type == SECTION_TYPE_RB_AQ || + sec->section.type == SECTION_TYPE_RB_MMIO) { + i40e_debug(hw, I40E_DEBUG_PACKAGE, + "Not an original package\n"); + return I40E_NOT_SUPPORTED; + } + } + } + + return status; +} + /** * i40e_write_profile * @hw: pointer to the hardware structure @@ -7456,47 +7615,99 @@ i40e_write_profile(struct i40e_hw *hw, struct i40e_profile_segment *profile, enum i40e_status_code status = I40E_SUCCESS; struct i40e_section_table *sec_tbl; struct i40e_profile_section_header *sec = NULL; - u32 dev_cnt; - u32 vendor_dev_id; - u32 *nvm; + struct i40e_profile_aq_section *ddp_aq; u32 section_size = 0; u32 offset = 0, info = 0; + u32 sec_off; u32 i; - dev_cnt = profile->device_table_count; + status = i40e_validate_profile(hw, profile, track_id, false); + if (status) + return status; - for (i = 0; i < dev_cnt; i++) { - vendor_dev_id = profile->device_table[i].vendor_dev_id; - if ((vendor_dev_id >> 16) == I40E_INTEL_VENDOR_ID) - if (hw->device_id == (vendor_dev_id & 0xFFFF)) + I40E_SECTION_TABLE(profile, sec_tbl); + + for (i = 0; i < sec_tbl->section_count; i++) { + sec_off = sec_tbl->section_offset[i]; + sec = I40E_SECTION_HEADER(profile, sec_off); + /* Process generic admin command */ + if (sec->section.type == SECTION_TYPE_AQ) { + ddp_aq = (struct i40e_profile_aq_section *)&sec[1]; + status = i40e_ddp_exec_aq_section(hw, ddp_aq); + if (status) { + i40e_debug(hw, I40E_DEBUG_PACKAGE, + "Failed to execute aq: section %d, opcode %u\n", + i, ddp_aq->opcode); break; + } + sec->section.type = SECTION_TYPE_RB_AQ; + } + + /* Skip any non-mmio sections */ + if (sec->section.type != SECTION_TYPE_MMIO) + continue; + + section_size = sec->section.size + + sizeof(struct i40e_profile_section_header); + + /* Write MMIO section */ + status = i40e_aq_write_ddp(hw, (void *)sec, (u16)section_size, + track_id, &offset, &info, NULL); + if (status) { + i40e_debug(hw, I40E_DEBUG_PACKAGE, + "Failed to write profile: section %d, offset %d, info %d\n", + i, offset, info); + break; + } } - if (i == dev_cnt) { - i40e_debug(hw, I40E_DEBUG_PACKAGE, "Device doesn't support DDP"); - return I40E_ERR_DEVICE_NOT_SUPPORTED; - } + return status; +} + +/** + * i40e_rollback_profile + * @hw: pointer to the hardware structure + * @profile: pointer to the profile segment of the package to be removed + * @track_id: package tracking id + * + * Rolls back previously loaded package. + */ +enum i40e_status_code +i40e_rollback_profile(struct i40e_hw *hw, struct i40e_profile_segment *profile, + u32 track_id) +{ + struct i40e_profile_section_header *sec = NULL; + enum i40e_status_code status = I40E_SUCCESS; + struct i40e_section_table *sec_tbl; + u32 offset = 0, info = 0; + u32 section_size = 0; + u32 sec_off; + int i; - nvm = (u32 *)&profile->device_table[dev_cnt]; - sec_tbl = (struct i40e_section_table *)&nvm[nvm[0] + 1]; + status = i40e_validate_profile(hw, profile, track_id, true); + if (status) + return status; - for (i = 0; i < sec_tbl->section_count; i++) { - sec = (struct i40e_profile_section_header *)((u8 *)profile + - sec_tbl->section_offset[i]); + I40E_SECTION_TABLE(profile, sec_tbl); - /* Skip 'AQ', 'note' and 'name' sections */ - if (sec->section.type != SECTION_TYPE_MMIO) + /* For rollback write sections in reverse */ + for (i = sec_tbl->section_count - 1; i >= 0; i--) { + sec_off = sec_tbl->section_offset[i]; + sec = I40E_SECTION_HEADER(profile, sec_off); + + /* Skip any non-rollback sections */ + if (sec->section.type != SECTION_TYPE_RB_MMIO) continue; section_size = sec->section.size + sizeof(struct i40e_profile_section_header); - /* Write profile */ + /* Write roll-back MMIO section */ status = i40e_aq_write_ddp(hw, (void *)sec, (u16)section_size, track_id, &offset, &info, NULL); if (status) { i40e_debug(hw, I40E_DEBUG_PACKAGE, - "Failed to write profile: offset %d, info %d", - offset, info); + "Failed to write profile: section %d, offset %d, info %d\n", + i, offset, info); break; } } @@ -7534,7 +7745,8 @@ i40e_add_pinfo_to_list(struct i40e_hw *hw, pinfo->track_id = track_id; pinfo->version = profile->version; pinfo->op = I40E_DDP_ADD_TRACKID; - memcpy(pinfo->name, profile->name, I40E_DDP_NAME_SIZE); + i40e_memcpy(pinfo->name, profile->name, I40E_DDP_NAME_SIZE, + I40E_NONDMA_TO_NONDMA); status = i40e_aq_write_ddp(hw, (void *)sec, sec->data_end, track_id, &offset, &info, NULL); diff --git a/drivers/net/i40e/base/i40e_prototype.h b/drivers/net/i40e/base/i40e_prototype.h index 9171e97..acb2023 100644 --- a/drivers/net/i40e/base/i40e_prototype.h +++ b/drivers/net/i40e/base/i40e_prototype.h @@ -586,10 +586,16 @@ enum i40e_status_code i40e_aq_get_ddp_list(struct i40e_hw *hw, void *buff, struct i40e_generic_seg_header * i40e_find_segment_in_package(u32 segment_type, struct i40e_package_header *pkg_header); +struct i40e_profile_section_header * +i40e_find_section_in_profile(u32 section_type, + struct i40e_profile_segment *profile); enum i40e_status_code i40e_write_profile(struct i40e_hw *hw, struct i40e_profile_segment *i40e_seg, u32 track_id); enum i40e_status_code +i40e_rollback_profile(struct i40e_hw *hw, struct i40e_profile_segment *i40e_seg, + u32 track_id); +enum i40e_status_code i40e_add_pinfo_to_list(struct i40e_hw *hw, struct i40e_profile_segment *profile, u8 *profile_info_sec, u32 track_id); diff --git a/drivers/net/i40e/base/i40e_type.h b/drivers/net/i40e/base/i40e_type.h index 9d7b1a2..dca725a 100644 --- a/drivers/net/i40e/base/i40e_type.h +++ b/drivers/net/i40e/base/i40e_type.h @@ -1918,8 +1918,10 @@ struct i40e_generic_seg_header { struct i40e_metadata_segment { struct i40e_generic_seg_header header; struct i40e_ddp_version version; +#define I40E_DDP_TRACKID_RDONLY 0 +#define I40E_DDP_TRACKID_INVALID 0xFFFFFFFF u32 track_id; - char name[I40E_DDP_NAME_SIZE]; + char name[I40E_DDP_NAME_SIZE]; }; struct i40e_device_id_entry { @@ -1946,15 +1948,36 @@ struct i40e_profile_section_header { struct { #define SECTION_TYPE_INFO 0x00000010 #define SECTION_TYPE_MMIO 0x00000800 +#define SECTION_TYPE_RB_MMIO 0x00001800 #define SECTION_TYPE_AQ 0x00000801 +#define SECTION_TYPE_RB_AQ 0x00001801 #define SECTION_TYPE_NOTE 0x80000000 #define SECTION_TYPE_NAME 0x80000001 +#define SECTION_TYPE_PROTO 0x80000002 +#define SECTION_TYPE_PCTYPE 0x80000003 +#define SECTION_TYPE_PTYPE 0x80000004 u32 type; u32 offset; u32 size; } section; }; +struct i40e_profile_tlv_section_record { + u8 rtype; + u8 type; + u16 len; + u8 data[12]; +}; + +/* Generic AQ section in proflie */ +struct i40e_profile_aq_section { + u16 opcode; + u16 flags; + u8 param[16]; + u16 datalen; + u8 data[1]; +}; + struct i40e_profile_info { u32 track_id; struct i40e_ddp_version version; -- 2.4.11 ^ permalink raw reply [flat|nested] 69+ messages in thread
* [dpdk-dev] [PATCH v3 16/17] net/i40e: use set switch aq instead of register setting 2017-06-27 13:29 ` [dpdk-dev] [PATCH v3 00/17] net/i40e: base code update Jingjing Wu ` (14 preceding siblings ...) 2017-06-27 13:29 ` [dpdk-dev] [PATCH v3 15/17] net/i40e/base: extend processing of DDP Jingjing Wu @ 2017-06-27 13:29 ` Jingjing Wu 2017-06-27 13:29 ` [dpdk-dev] [PATCH v3 17/17] net/i40e/base: update base code info Jingjing Wu 2017-06-27 14:00 ` [dpdk-dev] [PATCH v4 00/17] net/i40e: base code update Jingjing Wu 17 siblings, 0 replies; 69+ messages in thread From: Jingjing Wu @ 2017-06-27 13:29 UTC (permalink / raw) To: dev; +Cc: jingjing.wu, helin.zhang TPID can be set by set_switch_config aq, change the TPID setting by set_switch_config on new FW release. Signed-off-by: Jingjing Wu <jingjing.wu@intel.com> --- drivers/net/i40e/i40e_ethdev.c | 102 +++++++++++++++++++++++++---------------- 1 file changed, 62 insertions(+), 40 deletions(-) diff --git a/drivers/net/i40e/i40e_ethdev.c b/drivers/net/i40e/i40e_ethdev.c index 4ee1113..b1706a0 100644 --- a/drivers/net/i40e/i40e_ethdev.c +++ b/drivers/net/i40e/i40e_ethdev.c @@ -2973,71 +2973,93 @@ i40e_vlan_filter_set(struct rte_eth_dev *dev, uint16_t vlan_id, int on) } static int -i40e_vlan_tpid_set(struct rte_eth_dev *dev, - enum rte_vlan_type vlan_type, - uint16_t tpid) +i40e_vlan_tpid_set_by_registers(struct rte_eth_dev *dev, + enum rte_vlan_type vlan_type, + uint16_t tpid, int qinq) { struct i40e_hw *hw = I40E_DEV_PRIVATE_TO_HW(dev->data->dev_private); - uint64_t reg_r = 0, reg_w = 0; - uint16_t reg_id = 0; - int ret = 0; - int qinq = dev->data->dev_conf.rxmode.hw_vlan_extend; + uint64_t reg_r = 0; + uint64_t reg_w = 0; + uint16_t reg_id = 3; + int ret; - switch (vlan_type) { - case ETH_VLAN_TYPE_OUTER: - if (qinq) + if (qinq) { + if (vlan_type == ETH_VLAN_TYPE_OUTER) reg_id = 2; - else - reg_id = 3; - break; - case ETH_VLAN_TYPE_INNER: - if (qinq) - reg_id = 3; - else { - ret = -EINVAL; - PMD_DRV_LOG(ERR, - "Unsupported vlan type in single vlan."); - return ret; - } - break; - default: - ret = -EINVAL; - PMD_DRV_LOG(ERR, "Unsupported vlan type %d", vlan_type); - return ret; } + ret = i40e_aq_debug_read_register(hw, I40E_GL_SWT_L2TAGCTRL(reg_id), ®_r, NULL); if (ret != I40E_SUCCESS) { PMD_DRV_LOG(ERR, "Fail to debug read from I40E_GL_SWT_L2TAGCTRL[%d]", reg_id); - ret = -EIO; - return ret; + return -EIO; } PMD_DRV_LOG(DEBUG, - "Debug read from I40E_GL_SWT_L2TAGCTRL[%d]: 0x%08"PRIx64, - reg_id, reg_r); + "Debug read from I40E_GL_SWT_L2TAGCTRL[%d]: 0x%08"PRIx64, + reg_id, reg_r); reg_w = reg_r & (~(I40E_GL_SWT_L2TAGCTRL_ETHERTYPE_MASK)); reg_w |= ((uint64_t)tpid << I40E_GL_SWT_L2TAGCTRL_ETHERTYPE_SHIFT); if (reg_r == reg_w) { - ret = 0; PMD_DRV_LOG(DEBUG, "No need to write"); - return ret; + return 0; } ret = i40e_aq_debug_write_register(hw, I40E_GL_SWT_L2TAGCTRL(reg_id), reg_w, NULL); if (ret != I40E_SUCCESS) { - ret = -EIO; PMD_DRV_LOG(ERR, - "Fail to debug write to I40E_GL_SWT_L2TAGCTRL[%d]", - reg_id); - return ret; + "Fail to debug write to I40E_GL_SWT_L2TAGCTRL[%d]", + reg_id); + return -EIO; } PMD_DRV_LOG(DEBUG, - "Debug write 0x%08"PRIx64" to I40E_GL_SWT_L2TAGCTRL[%d]", - reg_w, reg_id); + "Debug write 0x%08"PRIx64" to I40E_GL_SWT_L2TAGCTRL[%d]", + reg_w, reg_id); + + return 0; +} + +static int +i40e_vlan_tpid_set(struct rte_eth_dev *dev, + enum rte_vlan_type vlan_type, + uint16_t tpid) +{ + struct i40e_hw *hw = I40E_DEV_PRIVATE_TO_HW(dev->data->dev_private); + int qinq = dev->data->dev_conf.rxmode.hw_vlan_extend; + int ret = 0; + + if ((vlan_type != ETH_VLAN_TYPE_INNER && + vlan_type != ETH_VLAN_TYPE_OUTER) || + (!qinq && vlan_type == ETH_VLAN_TYPE_INNER)) { + PMD_DRV_LOG(ERR, + "Unsupported vlan type."); + return -EINVAL; + } + /* 802.1ad frames ability is added in NVM API 1.7*/ + if (hw->flags & I40E_HW_FLAG_802_1AD_CAPABLE) { + if (qinq) { + if (vlan_type == ETH_VLAN_TYPE_OUTER) + hw->first_tag = rte_cpu_to_le_16(tpid); + else if (vlan_type == ETH_VLAN_TYPE_INNER) + hw->second_tag = rte_cpu_to_le_16(tpid); + } else { + if (vlan_type == ETH_VLAN_TYPE_OUTER) + hw->second_tag = rte_cpu_to_le_16(tpid); + } + ret = i40e_aq_set_switch_config(hw, 0, 0, NULL); + if (ret != I40E_SUCCESS) { + PMD_DRV_LOG(ERR, + "Set switch config failed aq_err: %d", + hw->aq.asq_last_status); + ret = -EIO; + } + } else + /* If NVM API < 1.7, keep the register setting */ + ret = i40e_vlan_tpid_set_by_registers(dev, vlan_type, + tpid, qinq); return ret; } @@ -3066,7 +3088,7 @@ i40e_vlan_offload_set(struct rte_eth_dev *dev, int mask) if (mask & ETH_VLAN_EXTEND_MASK) { if (dev->data->dev_conf.rxmode.hw_vlan_extend) { i40e_vsi_config_double_vlan(vsi, TRUE); - /* Set global registers with default ether type value */ + /* Set global registers with default ethertype. */ i40e_vlan_tpid_set(dev, ETH_VLAN_TYPE_OUTER, ETHER_TYPE_VLAN); i40e_vlan_tpid_set(dev, ETH_VLAN_TYPE_INNER, -- 2.4.11 ^ permalink raw reply [flat|nested] 69+ messages in thread
* [dpdk-dev] [PATCH v3 17/17] net/i40e/base: update base code info 2017-06-27 13:29 ` [dpdk-dev] [PATCH v3 00/17] net/i40e: base code update Jingjing Wu ` (15 preceding siblings ...) 2017-06-27 13:29 ` [dpdk-dev] [PATCH v3 16/17] net/i40e: use set switch aq instead of register setting Jingjing Wu @ 2017-06-27 13:29 ` Jingjing Wu 2017-06-27 14:00 ` [dpdk-dev] [PATCH v4 00/17] net/i40e: base code update Jingjing Wu 17 siblings, 0 replies; 69+ messages in thread From: Jingjing Wu @ 2017-06-27 13:29 UTC (permalink / raw) To: dev; +Cc: jingjing.wu, helin.zhang Signed-off-by: Jingjing Wu <jingjing.wu@intel.com> --- drivers/net/i40e/base/README | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/net/i40e/base/README b/drivers/net/i40e/base/README index 0da9f67..59e76c2 100644 --- a/drivers/net/i40e/base/README +++ b/drivers/net/i40e/base/README @@ -34,7 +34,7 @@ Intel® I40E driver ================== This directory contains source code of FreeBSD i40e driver of version -cid-i40e.2017.03.21.tar.gz released by the team which develops +cid-i40e.2017.06.23.tar.gz released by the team which develops basic drivers for any i40e NIC. The directory of base/ contains the original source package. This driver is valid for the product(s) listed below -- 2.4.11 ^ permalink raw reply [flat|nested] 69+ messages in thread
* [dpdk-dev] [PATCH v4 00/17] net/i40e: base code update 2017-06-27 13:29 ` [dpdk-dev] [PATCH v3 00/17] net/i40e: base code update Jingjing Wu ` (16 preceding siblings ...) 2017-06-27 13:29 ` [dpdk-dev] [PATCH v3 17/17] net/i40e/base: update base code info Jingjing Wu @ 2017-06-27 14:00 ` Jingjing Wu 2017-06-27 14:00 ` [dpdk-dev] [PATCH v4 01/17] net/i40e/base: use new virtchnl header file Jingjing Wu ` (17 more replies) 17 siblings, 18 replies; 69+ messages in thread From: Jingjing Wu @ 2017-06-27 14:00 UTC (permalink / raw) To: dev; +Cc: jingjing.wu, helin.zhang v4 changes: - fix more checkpatch warnings v3 changes: - extra patch to update README - remove unnecessary code - fix checkpatch warning v2 changes: - fix compile issue - update from package from 2017.05.24 to 2017.06.23 i40e base code upate. The main changes are: - use virtchnl.h instead of i40e_virtchnl.h - add support for Adaptive Virtual Function - add new AQ commands for read/write PHY registers - add new phy types for 25G - extend processing of DDP Jingjing Wu (17): net/i40e/base: use new virtchnl header file net/i40e/base: sync nvmupdate command and adminq subtask net/i40e/base: add AQ command for read/write PHY registers net/i40e/base: add support for Adaptive Virtual Function net/i40e/base: store the requested FEC information net/i40e/base: add new phy types for 25G AOC and ACC net/i40e/base: report supported link modes net/i40e/base: track id can be 0 net/i40e/base: update FW AQ API version to 1.7 net/i40e/base: add support for switch parameters net/i40e/base: use admin queue for setting LEDs behavior net/i40e/base: avoid potential null pointer dereference net/i40e/base: avoid reset timeout issue net/i40e/base: add EEPROM checksum verification net/i40e/base: extend processing of DDP net/i40e: use set switch aq instead of register setting net/i40e/base: update base code info drivers/net/i40e/base/README | 2 +- drivers/net/i40e/base/i40e_adminq.c | 12 + drivers/net/i40e/base/i40e_adminq_cmd.h | 71 ++- drivers/net/i40e/base/i40e_common.c | 549 +++++++++++++++++++---- drivers/net/i40e/base/i40e_devids.h | 1 + drivers/net/i40e/base/i40e_nvm.c | 20 +- drivers/net/i40e/base/i40e_prototype.h | 33 +- drivers/net/i40e/base/i40e_register.h | 2 +- drivers/net/i40e/base/i40e_type.h | 49 +- drivers/net/i40e/base/i40e_virtchnl.h | 445 ------------------ drivers/net/i40e/base/virtchnl.h | 772 ++++++++++++++++++++++++++++++++ drivers/net/i40e/i40e_ethdev.c | 102 +++-- drivers/net/i40e/i40e_ethdev.h | 8 +- drivers/net/i40e/i40e_ethdev_vf.c | 209 +++++---- drivers/net/i40e/i40e_pf.c | 235 +++++----- drivers/net/i40e/i40e_pf.h | 32 +- drivers/net/i40e/rte_pmd_i40e.h | 2 +- 17 files changed, 1709 insertions(+), 835 deletions(-) delete mode 100644 drivers/net/i40e/base/i40e_virtchnl.h create mode 100644 drivers/net/i40e/base/virtchnl.h -- 2.4.11 ^ permalink raw reply [flat|nested] 69+ messages in thread
* [dpdk-dev] [PATCH v4 01/17] net/i40e/base: use new virtchnl header file 2017-06-27 14:00 ` [dpdk-dev] [PATCH v4 00/17] net/i40e: base code update Jingjing Wu @ 2017-06-27 14:00 ` Jingjing Wu 2017-06-27 14:00 ` [dpdk-dev] [PATCH v4 02/17] net/i40e/base: sync nvmupdate command and adminq subtask Jingjing Wu ` (16 subsequent siblings) 17 siblings, 0 replies; 69+ messages in thread From: Jingjing Wu @ 2017-06-27 14:00 UTC (permalink / raw) To: dev; +Cc: jingjing.wu, helin.zhang Modify the necessary files to be compatible with the new virtchnl.h file instead of relying on i40e_virtchnl.h variant. This mostly changes references to VIRTCHNL_ variables by removing prefix of I40E_. Signed-off-by: Jingjing Wu <jingjing.wu@intel.com> --- drivers/net/i40e/base/i40e_common.c | 24 +- drivers/net/i40e/base/i40e_prototype.h | 6 +- drivers/net/i40e/base/i40e_type.h | 12 +- drivers/net/i40e/base/i40e_virtchnl.h | 445 ------------------- drivers/net/i40e/base/virtchnl.h | 772 +++++++++++++++++++++++++++++++++ drivers/net/i40e/i40e_ethdev.h | 8 +- drivers/net/i40e/i40e_ethdev_vf.c | 209 +++++---- drivers/net/i40e/i40e_pf.c | 235 +++++----- drivers/net/i40e/i40e_pf.h | 32 +- drivers/net/i40e/rte_pmd_i40e.h | 2 +- 10 files changed, 1039 insertions(+), 706 deletions(-) delete mode 100644 drivers/net/i40e/base/i40e_virtchnl.h create mode 100644 drivers/net/i40e/base/virtchnl.h diff --git a/drivers/net/i40e/base/i40e_common.c b/drivers/net/i40e/base/i40e_common.c index 03e94bc..6c09c27 100644 --- a/drivers/net/i40e/base/i40e_common.c +++ b/drivers/net/i40e/base/i40e_common.c @@ -34,7 +34,7 @@ POSSIBILITY OF SUCH DAMAGE. #include "i40e_type.h" #include "i40e_adminq.h" #include "i40e_prototype.h" -#include "i40e_virtchnl.h" +#include "virtchnl.h" /** @@ -5523,7 +5523,7 @@ enum i40e_status_code i40e_aq_add_rem_control_packet_filter(struct i40e_hw *hw, } if (mac_addr) - i40e_memcpy(cmd->mac, mac_addr, I40E_ETH_LENGTH_OF_ADDRESS, + i40e_memcpy(cmd->mac, mac_addr, ETH_ALEN, I40E_NONDMA_TO_NONDMA); cmd->etype = CPU_TO_LE16(ethtype); @@ -6879,7 +6879,7 @@ void i40e_write_rx_ctl(struct i40e_hw *hw, u32 reg_addr, u32 reg_val) * completion before returning. **/ enum i40e_status_code i40e_aq_send_msg_to_pf(struct i40e_hw *hw, - enum i40e_virtchnl_ops v_opcode, + enum virtchnl_ops v_opcode, enum i40e_status_code v_retval, u8 *msg, u16 msglen, struct i40e_asq_cmd_details *cmd_details) @@ -6918,9 +6918,9 @@ enum i40e_status_code i40e_aq_send_msg_to_pf(struct i40e_hw *hw, * with appropriate information. **/ void i40e_vf_parse_hw_config(struct i40e_hw *hw, - struct i40e_virtchnl_vf_resource *msg) + struct virtchnl_vf_resource *msg) { - struct i40e_virtchnl_vsi_resource *vsi_res; + struct virtchnl_vsi_resource *vsi_res; int i; vsi_res = &msg->vsi_res[0]; @@ -6930,19 +6930,17 @@ void i40e_vf_parse_hw_config(struct i40e_hw *hw, hw->dev_caps.num_tx_qp = msg->num_queue_pairs; hw->dev_caps.num_msix_vectors_vf = msg->max_vectors; hw->dev_caps.dcb = msg->vf_offload_flags & - I40E_VIRTCHNL_VF_OFFLOAD_L2; - hw->dev_caps.fcoe = (msg->vf_offload_flags & - I40E_VIRTCHNL_VF_OFFLOAD_FCOE) ? 1 : 0; + VIRTCHNL_VF_OFFLOAD_L2; hw->dev_caps.iwarp = (msg->vf_offload_flags & - I40E_VIRTCHNL_VF_OFFLOAD_IWARP) ? 1 : 0; + VIRTCHNL_VF_OFFLOAD_IWARP) ? 1 : 0; for (i = 0; i < msg->num_vsis; i++) { - if (vsi_res->vsi_type == I40E_VSI_SRIOV) { + if (vsi_res->vsi_type == VIRTCHNL_VSI_SRIOV) { i40e_memcpy(hw->mac.perm_addr, vsi_res->default_mac_addr, - I40E_ETH_LENGTH_OF_ADDRESS, + ETH_ALEN, I40E_NONDMA_TO_NONDMA); i40e_memcpy(hw->mac.addr, vsi_res->default_mac_addr, - I40E_ETH_LENGTH_OF_ADDRESS, + ETH_ALEN, I40E_NONDMA_TO_NONDMA); } vsi_res++; @@ -6959,7 +6957,7 @@ void i40e_vf_parse_hw_config(struct i40e_hw *hw, **/ enum i40e_status_code i40e_vf_reset(struct i40e_hw *hw) { - return i40e_aq_send_msg_to_pf(hw, I40E_VIRTCHNL_OP_RESET_VF, + return i40e_aq_send_msg_to_pf(hw, VIRTCHNL_OP_RESET_VF, I40E_SUCCESS, NULL, 0, NULL); } #endif /* VF_DRIVER */ diff --git a/drivers/net/i40e/base/i40e_prototype.h b/drivers/net/i40e/base/i40e_prototype.h index 4bd589e..6ec4304 100644 --- a/drivers/net/i40e/base/i40e_prototype.h +++ b/drivers/net/i40e/base/i40e_prototype.h @@ -36,7 +36,7 @@ POSSIBILITY OF SUCH DAMAGE. #include "i40e_type.h" #include "i40e_alloc.h" -#include "i40e_virtchnl.h" +#include "virtchnl.h" /* Prototypes for shared code functions that are not in * the standard function pointer structures. These are @@ -504,10 +504,10 @@ void i40e_destroy_spinlock(struct i40e_spinlock *sp); /* i40e_common for VF drivers*/ void i40e_vf_parse_hw_config(struct i40e_hw *hw, - struct i40e_virtchnl_vf_resource *msg); + struct virtchnl_vf_resource *msg); enum i40e_status_code i40e_vf_reset(struct i40e_hw *hw); enum i40e_status_code i40e_aq_send_msg_to_pf(struct i40e_hw *hw, - enum i40e_virtchnl_ops v_opcode, + enum virtchnl_ops v_opcode, enum i40e_status_code v_retval, u8 *msg, u16 msglen, struct i40e_asq_cmd_details *cmd_details); diff --git a/drivers/net/i40e/base/i40e_type.h b/drivers/net/i40e/base/i40e_type.h index 84d5757..52a114a 100644 --- a/drivers/net/i40e/base/i40e_type.h +++ b/drivers/net/i40e/base/i40e_type.h @@ -92,7 +92,9 @@ POSSIBILITY OF SUCH DAMAGE. struct i40e_hw; typedef void (*I40E_ADMINQ_CALLBACK)(struct i40e_hw *, struct i40e_aq_desc *); -#define I40E_ETH_LENGTH_OF_ADDRESS 6 +#ifndef ETH_ALEN +#define ETH_ALEN 6 +#endif /* Data type manipulation macros. */ #define I40E_HI_DWORD(x) ((u32)((((x) >> 16) >> 16) & 0xFFFFFFFF)) #define I40E_LO_DWORD(x) ((u32)((x) & 0xFFFFFFFF)) @@ -439,10 +441,10 @@ struct i40e_hw_capabilities { struct i40e_mac_info { enum i40e_mac_type type; - u8 addr[I40E_ETH_LENGTH_OF_ADDRESS]; - u8 perm_addr[I40E_ETH_LENGTH_OF_ADDRESS]; - u8 san_addr[I40E_ETH_LENGTH_OF_ADDRESS]; - u8 port_addr[I40E_ETH_LENGTH_OF_ADDRESS]; + u8 addr[ETH_ALEN]; + u8 perm_addr[ETH_ALEN]; + u8 san_addr[ETH_ALEN]; + u8 port_addr[ETH_ALEN]; u16 max_fcoeq; }; diff --git a/drivers/net/i40e/base/i40e_virtchnl.h b/drivers/net/i40e/base/i40e_virtchnl.h deleted file mode 100644 index 7a24c0f..0000000 --- a/drivers/net/i40e/base/i40e_virtchnl.h +++ /dev/null @@ -1,445 +0,0 @@ -/******************************************************************************* - -Copyright (c) 2013 - 2015, Intel Corporation -All rights reserved. - -Redistribution and use in source and binary forms, with or without -modification, are permitted provided that the following conditions are met: - - 1. Redistributions of source code must retain the above copyright notice, - this list of conditions and the following disclaimer. - - 2. Redistributions in binary form must reproduce the above copyright - notice, this list of conditions and the following disclaimer in the - documentation and/or other materials provided with the distribution. - - 3. Neither the name of the Intel Corporation nor the names of its - contributors may be used to endorse or promote products derived from - this software without specific prior written permission. - -THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS" -AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE -IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE -ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE -LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR -CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF -SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS -INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN -CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) -ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE -POSSIBILITY OF SUCH DAMAGE. - -***************************************************************************/ - -#ifndef _I40E_VIRTCHNL_H_ -#define _I40E_VIRTCHNL_H_ - -#include "i40e_type.h" - -/* Description: - * This header file describes the VF-PF communication protocol used - * by the various i40e drivers. - * - * Admin queue buffer usage: - * desc->opcode is always i40e_aqc_opc_send_msg_to_pf - * flags, retval, datalen, and data addr are all used normally. - * Firmware copies the cookie fields when sending messages between the PF and - * VF, but uses all other fields internally. Due to this limitation, we - * must send all messages as "indirect", i.e. using an external buffer. - * - * All the vsi indexes are relative to the VF. Each VF can have maximum of - * three VSIs. All the queue indexes are relative to the VSI. Each VF can - * have a maximum of sixteen queues for all of its VSIs. - * - * The PF is required to return a status code in v_retval for all messages - * except RESET_VF, which does not require any response. The return value is of - * i40e_status_code type, defined in the i40e_type.h. - * - * In general, VF driver initialization should roughly follow the order of these - * opcodes. The VF driver must first validate the API version of the PF driver, - * then request a reset, then get resources, then configure queues and - * interrupts. After these operations are complete, the VF driver may start - * its queues, optionally add MAC and VLAN filters, and process traffic. - */ - -/* Opcodes for VF-PF communication. These are placed in the v_opcode field - * of the virtchnl_msg structure. - */ -enum i40e_virtchnl_ops { -/* The PF sends status change events to VFs using - * the I40E_VIRTCHNL_OP_EVENT opcode. - * VFs send requests to the PF using the other ops. - */ - I40E_VIRTCHNL_OP_UNKNOWN = 0, - I40E_VIRTCHNL_OP_VERSION = 1, /* must ALWAYS be 1 */ - I40E_VIRTCHNL_OP_RESET_VF = 2, - I40E_VIRTCHNL_OP_GET_VF_RESOURCES = 3, - I40E_VIRTCHNL_OP_CONFIG_TX_QUEUE = 4, - I40E_VIRTCHNL_OP_CONFIG_RX_QUEUE = 5, - I40E_VIRTCHNL_OP_CONFIG_VSI_QUEUES = 6, - I40E_VIRTCHNL_OP_CONFIG_IRQ_MAP = 7, - I40E_VIRTCHNL_OP_ENABLE_QUEUES = 8, - I40E_VIRTCHNL_OP_DISABLE_QUEUES = 9, - I40E_VIRTCHNL_OP_ADD_ETHER_ADDRESS = 10, - I40E_VIRTCHNL_OP_DEL_ETHER_ADDRESS = 11, - I40E_VIRTCHNL_OP_ADD_VLAN = 12, - I40E_VIRTCHNL_OP_DEL_VLAN = 13, - I40E_VIRTCHNL_OP_CONFIG_PROMISCUOUS_MODE = 14, - I40E_VIRTCHNL_OP_GET_STATS = 15, - I40E_VIRTCHNL_OP_FCOE = 16, - I40E_VIRTCHNL_OP_EVENT = 17, /* must ALWAYS be 17 */ -#ifdef I40E_SOL_VF_SUPPORT - I40E_VIRTCHNL_OP_GET_ADDNL_SOL_CONFIG = 19, -#endif - I40E_VIRTCHNL_OP_CONFIG_RSS_KEY = 23, - I40E_VIRTCHNL_OP_CONFIG_RSS_LUT = 24, - I40E_VIRTCHNL_OP_GET_RSS_HENA_CAPS = 25, - I40E_VIRTCHNL_OP_SET_RSS_HENA = 26, - -}; - -/* Virtual channel message descriptor. This overlays the admin queue - * descriptor. All other data is passed in external buffers. - */ - -struct i40e_virtchnl_msg { - u8 pad[8]; /* AQ flags/opcode/len/retval fields */ - enum i40e_virtchnl_ops v_opcode; /* avoid confusion with desc->opcode */ - enum i40e_status_code v_retval; /* ditto for desc->retval */ - u32 vfid; /* used by PF when sending to VF */ -}; - -/* Message descriptions and data structures.*/ - -/* I40E_VIRTCHNL_OP_VERSION - * VF posts its version number to the PF. PF responds with its version number - * in the same format, along with a return code. - * Reply from PF has its major/minor versions also in param0 and param1. - * If there is a major version mismatch, then the VF cannot operate. - * If there is a minor version mismatch, then the VF can operate but should - * add a warning to the system log. - * - * This enum element MUST always be specified as == 1, regardless of other - * changes in the API. The PF must always respond to this message without - * error regardless of version mismatch. - */ -#define I40E_VIRTCHNL_VERSION_MAJOR 1 -#define I40E_VIRTCHNL_VERSION_MINOR 1 -#define I40E_VIRTCHNL_VERSION_MINOR_NO_VF_CAPS 0 - -struct i40e_virtchnl_version_info { - u32 major; - u32 minor; -}; - -/* I40E_VIRTCHNL_OP_RESET_VF - * VF sends this request to PF with no parameters - * PF does NOT respond! VF driver must delay then poll VFGEN_RSTAT register - * until reset completion is indicated. The admin queue must be reinitialized - * after this operation. - * - * When reset is complete, PF must ensure that all queues in all VSIs associated - * with the VF are stopped, all queue configurations in the HMC are set to 0, - * and all MAC and VLAN filters (except the default MAC address) on all VSIs - * are cleared. - */ - -/* I40E_VIRTCHNL_OP_GET_VF_RESOURCES - * Version 1.0 VF sends this request to PF with no parameters - * Version 1.1 VF sends this request to PF with u32 bitmap of its capabilities - * PF responds with an indirect message containing - * i40e_virtchnl_vf_resource and one or more - * i40e_virtchnl_vsi_resource structures. - */ - -struct i40e_virtchnl_vsi_resource { - u16 vsi_id; - u16 num_queue_pairs; - enum i40e_vsi_type vsi_type; - u16 qset_handle; - u8 default_mac_addr[I40E_ETH_LENGTH_OF_ADDRESS]; -}; -/* VF offload flags */ -#define I40E_VIRTCHNL_VF_OFFLOAD_L2 0x00000001 -#define I40E_VIRTCHNL_VF_OFFLOAD_IWARP 0x00000002 -#define I40E_VIRTCHNL_VF_OFFLOAD_FCOE 0x00000004 -#define I40E_VIRTCHNL_VF_OFFLOAD_RSS_AQ 0x00000008 -#define I40E_VIRTCHNL_VF_OFFLOAD_RSS_REG 0x00000010 -#define I40E_VIRTCHNL_VF_OFFLOAD_WB_ON_ITR 0x00000020 -#define I40E_VIRTCHNL_VF_OFFLOAD_VLAN 0x00010000 -#define I40E_VIRTCHNL_VF_OFFLOAD_RX_POLLING 0x00020000 -#define I40E_VIRTCHNL_VF_OFFLOAD_RSS_PCTYPE_V2 0x00040000 -#define I40E_VIRTCHNL_VF_OFFLOAD_RSS_PF 0X00080000 -#define I40E_VIRTCHNL_VF_OFFLOAD_ENCAP 0X00100000 -#define I40E_VIRTCHNL_VF_OFFLOAD_ENCAP_CSUM 0X00200000 -#define I40E_VIRTCHNL_VF_OFFLOAD_RX_ENCAP_CSUM 0X00400000 - -#define I40E_VF_BASE_MODE_OFFLOADS (I40E_VIRTCHNL_VF_OFFLOAD_L2 | \ - I40E_VIRTCHNL_VF_OFFLOAD_VLAN | \ - I40E_VIRTCHNL_VF_OFFLOAD_RSS_PF) - -struct i40e_virtchnl_vf_resource { - u16 num_vsis; - u16 num_queue_pairs; - u16 max_vectors; - u16 max_mtu; - - u32 vf_offload_flags; - u32 rss_key_size; - u32 rss_lut_size; - - struct i40e_virtchnl_vsi_resource vsi_res[1]; -}; - -/* I40E_VIRTCHNL_OP_CONFIG_TX_QUEUE - * VF sends this message to set up parameters for one TX queue. - * External data buffer contains one instance of i40e_virtchnl_txq_info. - * PF configures requested queue and returns a status code. - */ - -/* Tx queue config info */ -struct i40e_virtchnl_txq_info { - u16 vsi_id; - u16 queue_id; - u16 ring_len; /* number of descriptors, multiple of 8 */ - u16 headwb_enabled; - u64 dma_ring_addr; - u64 dma_headwb_addr; -}; - -/* I40E_VIRTCHNL_OP_CONFIG_RX_QUEUE - * VF sends this message to set up parameters for one RX queue. - * External data buffer contains one instance of i40e_virtchnl_rxq_info. - * PF configures requested queue and returns a status code. - */ - -/* Rx queue config info */ -struct i40e_virtchnl_rxq_info { - u16 vsi_id; - u16 queue_id; - u32 ring_len; /* number of descriptors, multiple of 32 */ - u16 hdr_size; - u16 splithdr_enabled; - u32 databuffer_size; - u32 max_pkt_size; - u64 dma_ring_addr; - enum i40e_hmc_obj_rx_hsplit_0 rx_split_pos; -}; - -/* I40E_VIRTCHNL_OP_CONFIG_VSI_QUEUES - * VF sends this message to set parameters for all active TX and RX queues - * associated with the specified VSI. - * PF configures queues and returns status. - * If the number of queues specified is greater than the number of queues - * associated with the VSI, an error is returned and no queues are configured. - */ -struct i40e_virtchnl_queue_pair_info { - /* NOTE: vsi_id and queue_id should be identical for both queues. */ - struct i40e_virtchnl_txq_info txq; - struct i40e_virtchnl_rxq_info rxq; -}; - -struct i40e_virtchnl_vsi_queue_config_info { - u16 vsi_id; - u16 num_queue_pairs; - struct i40e_virtchnl_queue_pair_info qpair[1]; -}; - -/* I40E_VIRTCHNL_OP_CONFIG_IRQ_MAP - * VF uses this message to map vectors to queues. - * The rxq_map and txq_map fields are bitmaps used to indicate which queues - * are to be associated with the specified vector. - * The "other" causes are always mapped to vector 0. - * PF configures interrupt mapping and returns status. - */ -struct i40e_virtchnl_vector_map { - u16 vsi_id; - u16 vector_id; - u16 rxq_map; - u16 txq_map; - u16 rxitr_idx; - u16 txitr_idx; -}; - -struct i40e_virtchnl_irq_map_info { - u16 num_vectors; - struct i40e_virtchnl_vector_map vecmap[1]; -}; - -/* I40E_VIRTCHNL_OP_ENABLE_QUEUES - * I40E_VIRTCHNL_OP_DISABLE_QUEUES - * VF sends these message to enable or disable TX/RX queue pairs. - * The queues fields are bitmaps indicating which queues to act upon. - * (Currently, we only support 16 queues per VF, but we make the field - * u32 to allow for expansion.) - * PF performs requested action and returns status. - */ -struct i40e_virtchnl_queue_select { - u16 vsi_id; - u16 pad; - u32 rx_queues; - u32 tx_queues; -}; - -/* I40E_VIRTCHNL_OP_ADD_ETHER_ADDRESS - * VF sends this message in order to add one or more unicast or multicast - * address filters for the specified VSI. - * PF adds the filters and returns status. - */ - -/* I40E_VIRTCHNL_OP_DEL_ETHER_ADDRESS - * VF sends this message in order to remove one or more unicast or multicast - * filters for the specified VSI. - * PF removes the filters and returns status. - */ - -struct i40e_virtchnl_ether_addr { - u8 addr[I40E_ETH_LENGTH_OF_ADDRESS]; - u8 pad[2]; -}; - -struct i40e_virtchnl_ether_addr_list { - u16 vsi_id; - u16 num_elements; - struct i40e_virtchnl_ether_addr list[1]; -}; - -#ifdef I40E_SOL_VF_SUPPORT -/* I40E_VIRTCHNL_OP_GET_ADDNL_SOL_CONFIG - * VF sends this message to get the default MTU and list of additional ethernet - * addresses it is allowed to use. - * PF responds with an indirect message containing - * i40e_virtchnl_addnl_solaris_config with zero or more - * i40e_virtchnl_ether_addr structures. - * - * It is expected that this operation will only ever be needed for Solaris VFs - * running under a Solaris PF. - */ -struct i40e_virtchnl_addnl_solaris_config { - u16 default_mtu; - struct i40e_virtchnl_ether_addr_list al; -}; - -#endif -/* I40E_VIRTCHNL_OP_ADD_VLAN - * VF sends this message to add one or more VLAN tag filters for receives. - * PF adds the filters and returns status. - * If a port VLAN is configured by the PF, this operation will return an - * error to the VF. - */ - -/* I40E_VIRTCHNL_OP_DEL_VLAN - * VF sends this message to remove one or more VLAN tag filters for receives. - * PF removes the filters and returns status. - * If a port VLAN is configured by the PF, this operation will return an - * error to the VF. - */ - -struct i40e_virtchnl_vlan_filter_list { - u16 vsi_id; - u16 num_elements; - u16 vlan_id[1]; -}; - -/* I40E_VIRTCHNL_OP_CONFIG_PROMISCUOUS_MODE - * VF sends VSI id and flags. - * PF returns status code in retval. - * Note: we assume that broadcast accept mode is always enabled. - */ -struct i40e_virtchnl_promisc_info { - u16 vsi_id; - u16 flags; -}; - -#define I40E_FLAG_VF_UNICAST_PROMISC 0x00000001 -#define I40E_FLAG_VF_MULTICAST_PROMISC 0x00000002 - -/* I40E_VIRTCHNL_OP_GET_STATS - * VF sends this message to request stats for the selected VSI. VF uses - * the i40e_virtchnl_queue_select struct to specify the VSI. The queue_id - * field is ignored by the PF. - * - * PF replies with struct i40e_eth_stats in an external buffer. - */ - -/* I40E_VIRTCHNL_OP_CONFIG_RSS_KEY - * I40E_VIRTCHNL_OP_CONFIG_RSS_LUT - * VF sends these messages to configure RSS. Only supported if both PF - * and VF drivers set the I40E_VIRTCHNL_VF_OFFLOAD_RSS_PF bit during - * configuration negotiation. If this is the case, then the rss fields in - * the vf resource struct are valid. - * Both the key and LUT are initialized to 0 by the PF, meaning that - * RSS is effectively disabled until set up by the VF. - */ -struct i40e_virtchnl_rss_key { - u16 vsi_id; - u16 key_len; - u8 key[1]; /* RSS hash key, packed bytes */ -}; - -struct i40e_virtchnl_rss_lut { - u16 vsi_id; - u16 lut_entries; - u8 lut[1]; /* RSS lookup table*/ -}; - -/* I40E_VIRTCHNL_OP_GET_RSS_HENA_CAPS - * I40E_VIRTCHNL_OP_SET_RSS_HENA - * VF sends these messages to get and set the hash filter enable bits for RSS. - * By default, the PF sets these to all possible traffic types that the - * hardware supports. The VF can query this value if it wants to change the - * traffic types that are hashed by the hardware. - * Traffic types are defined in the i40e_filter_pctype enum in i40e_type.h - */ -struct i40e_virtchnl_rss_hena { - u64 hena; -}; - -/* I40E_VIRTCHNL_OP_EVENT - * PF sends this message to inform the VF driver of events that may affect it. - * No direct response is expected from the VF, though it may generate other - * messages in response to this one. - */ -enum i40e_virtchnl_event_codes { - I40E_VIRTCHNL_EVENT_UNKNOWN = 0, - I40E_VIRTCHNL_EVENT_LINK_CHANGE, - I40E_VIRTCHNL_EVENT_RESET_IMPENDING, - I40E_VIRTCHNL_EVENT_PF_DRIVER_CLOSE, -}; -#define I40E_PF_EVENT_SEVERITY_INFO 0 -#define I40E_PF_EVENT_SEVERITY_ATTENTION 1 -#define I40E_PF_EVENT_SEVERITY_ACTION_REQUIRED 2 -#define I40E_PF_EVENT_SEVERITY_CERTAIN_DOOM 255 - -struct i40e_virtchnl_pf_event { - enum i40e_virtchnl_event_codes event; - union { - struct { - enum i40e_aq_link_speed link_speed; - bool link_status; - } link_event; - } event_data; - - int severity; -}; - -/* VF reset states - these are written into the RSTAT register: - * I40E_VFGEN_RSTAT1 on the PF - * I40E_VFGEN_RSTAT on the VF - * When the PF initiates a reset, it writes 0 - * When the reset is complete, it writes 1 - * When the PF detects that the VF has recovered, it writes 2 - * VF checks this register periodically to determine if a reset has occurred, - * then polls it to know when the reset is complete. - * If either the PF or VF reads the register while the hardware - * is in a reset state, it will return DEADBEEF, which, when masked - * will result in 3. - */ -enum i40e_vfr_states { - I40E_VFR_INPROGRESS = 0, - I40E_VFR_COMPLETED, - I40E_VFR_VFACTIVE, - I40E_VFR_UNKNOWN, -}; - -#endif /* _I40E_VIRTCHNL_H_ */ diff --git a/drivers/net/i40e/base/virtchnl.h b/drivers/net/i40e/base/virtchnl.h new file mode 100644 index 0000000..f00dd36 --- /dev/null +++ b/drivers/net/i40e/base/virtchnl.h @@ -0,0 +1,772 @@ +/******************************************************************************* + +Copyright (c) 2013 - 2015, Intel Corporation +All rights reserved. + +Redistribution and use in source and binary forms, with or without +modification, are permitted provided that the following conditions are met: + + 1. Redistributions of source code must retain the above copyright notice, + this list of conditions and the following disclaimer. + + 2. Redistributions in binary form must reproduce the above copyright + notice, this list of conditions and the following disclaimer in the + documentation and/or other materials provided with the distribution. + + 3. Neither the name of the Intel Corporation nor the names of its + contributors may be used to endorse or promote products derived from + this software without specific prior written permission. + +THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS" +AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE +IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE +ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE +LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR +CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF +SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS +INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN +CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) +ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE +POSSIBILITY OF SUCH DAMAGE. + +***************************************************************************/ + +#ifndef _VIRTCHNL_H_ +#define _VIRTCHNL_H_ + +/* Description: + * This header file describes the VF-PF communication protocol used + * by the drivers for all devices starting from our 40G product line + * + * Admin queue buffer usage: + * desc->opcode is always aqc_opc_send_msg_to_pf + * flags, retval, datalen, and data addr are all used normally. + * The Firmware copies the cookie fields when sending messages between the + * PF and VF, but uses all other fields internally. Due to this limitation, + * we must send all messages as "indirect", i.e. using an external buffer. + * + * All the VSI indexes are relative to the VF. Each VF can have maximum of + * three VSIs. All the queue indexes are relative to the VSI. Each VF can + * have a maximum of sixteen queues for all of its VSIs. + * + * The PF is required to return a status code in v_retval for all messages + * except RESET_VF, which does not require any response. The return value + * is of status_code type, defined in the shared type.h. + * + * In general, VF driver initialization should roughly follow the order of + * these opcodes. The VF driver must first validate the API version of the + * PF driver, then request a reset, then get resources, then configure + * queues and interrupts. After these operations are complete, the VF + * driver may start its queues, optionally add MAC and VLAN filters, and + * process traffic. + */ + +/* START GENERIC DEFINES + * Need to ensure the following enums and defines hold the same meaning and + * value in current and future projects + */ + +/* Error Codes */ +enum virtchnl_status_code { + VIRTCHNL_STATUS_SUCCESS = 0, + VIRTCHNL_ERR_PARAM = -5, + VIRTCHNL_STATUS_ERR_OPCODE_MISMATCH = -38, + VIRTCHNL_STATUS_ERR_CQP_COMPL_ERROR = -39, + VIRTCHNL_STATUS_ERR_INVALID_VF_ID = -40, + VIRTCHNL_STATUS_NOT_SUPPORTED = -64, +}; + +#define VIRTCHNL_LINK_SPEED_100MB_SHIFT 0x1 +#define VIRTCHNL_LINK_SPEED_1000MB_SHIFT 0x2 +#define VIRTCHNL_LINK_SPEED_10GB_SHIFT 0x3 +#define VIRTCHNL_LINK_SPEED_40GB_SHIFT 0x4 +#define VIRTCHNL_LINK_SPEED_20GB_SHIFT 0x5 +#define VIRTCHNL_LINK_SPEED_25GB_SHIFT 0x6 + +enum virtchnl_link_speed { + VIRTCHNL_LINK_SPEED_UNKNOWN = 0, + VIRTCHNL_LINK_SPEED_100MB = BIT(VIRTCHNL_LINK_SPEED_100MB_SHIFT), + VIRTCHNL_LINK_SPEED_1GB = BIT(VIRTCHNL_LINK_SPEED_1000MB_SHIFT), + VIRTCHNL_LINK_SPEED_10GB = BIT(VIRTCHNL_LINK_SPEED_10GB_SHIFT), + VIRTCHNL_LINK_SPEED_40GB = BIT(VIRTCHNL_LINK_SPEED_40GB_SHIFT), + VIRTCHNL_LINK_SPEED_20GB = BIT(VIRTCHNL_LINK_SPEED_20GB_SHIFT), + VIRTCHNL_LINK_SPEED_25GB = BIT(VIRTCHNL_LINK_SPEED_25GB_SHIFT), +}; + +/* for hsplit_0 field of Rx HMC context */ +/* deprecated with AVF 1.0 */ +enum virtchnl_rx_hsplit { + VIRTCHNL_RX_HSPLIT_NO_SPLIT = 0, + VIRTCHNL_RX_HSPLIT_SPLIT_L2 = 1, + VIRTCHNL_RX_HSPLIT_SPLIT_IP = 2, + VIRTCHNL_RX_HSPLIT_SPLIT_TCP_UDP = 4, + VIRTCHNL_RX_HSPLIT_SPLIT_SCTP = 8, +}; + +#define VIRTCHNL_ETH_LENGTH_OF_ADDRESS 6 +/* END GENERIC DEFINES */ + +/* Opcodes for VF-PF communication. These are placed in the v_opcode field + * of the virtchnl_msg structure. + */ +enum virtchnl_ops { +/* The PF sends status change events to VFs using + * the VIRTCHNL_OP_EVENT opcode. + * VFs send requests to the PF using the other ops. + * Use of "advanced opcode" features must be negotiated as part of capabilities + * exchange and are not considered part of base mode feature set. + */ + VIRTCHNL_OP_UNKNOWN = 0, + VIRTCHNL_OP_VERSION = 1, /* must ALWAYS be 1 */ + VIRTCHNL_OP_RESET_VF = 2, + VIRTCHNL_OP_GET_VF_RESOURCES = 3, + VIRTCHNL_OP_CONFIG_TX_QUEUE = 4, + VIRTCHNL_OP_CONFIG_RX_QUEUE = 5, + VIRTCHNL_OP_CONFIG_VSI_QUEUES = 6, + VIRTCHNL_OP_CONFIG_IRQ_MAP = 7, + VIRTCHNL_OP_ENABLE_QUEUES = 8, + VIRTCHNL_OP_DISABLE_QUEUES = 9, + VIRTCHNL_OP_ADD_ETH_ADDR = 10, + VIRTCHNL_OP_DEL_ETH_ADDR = 11, + VIRTCHNL_OP_ADD_VLAN = 12, + VIRTCHNL_OP_DEL_VLAN = 13, + VIRTCHNL_OP_CONFIG_PROMISCUOUS_MODE = 14, + VIRTCHNL_OP_GET_STATS = 15, + VIRTCHNL_OP_RSVD = 16, + VIRTCHNL_OP_EVENT = 17, /* must ALWAYS be 17 */ +#ifdef VIRTCHNL_SOL_VF_SUPPORT + VIRTCHNL_OP_GET_ADDNL_SOL_CONFIG = 19, +#endif +#ifdef VIRTCHNL_IWARP + VIRTCHNL_OP_IWARP = 20, /* advanced opcode */ + VIRTCHNL_OP_CONFIG_IWARP_IRQ_MAP = 21, /* advanced opcode */ + VIRTCHNL_OP_RELEASE_IWARP_IRQ_MAP = 22, /* advanced opcode */ +#endif + VIRTCHNL_OP_CONFIG_RSS_KEY = 23, + VIRTCHNL_OP_CONFIG_RSS_LUT = 24, + VIRTCHNL_OP_GET_RSS_HENA_CAPS = 25, + VIRTCHNL_OP_SET_RSS_HENA = 26, + VIRTCHNL_OP_ENABLE_VLAN_STRIPPING = 27, + VIRTCHNL_OP_DISABLE_VLAN_STRIPPING = 28, + VIRTCHNL_OP_REQUEST_QUEUES = 29, + +}; + +/* This macro is used to generate a compilation error if a structure + * is not exactly the correct length. It gives a divide by zero error if the + * structure is not of the correct size, otherwise it creates an enum that is + * never used. + */ +#define VIRTCHNL_CHECK_STRUCT_LEN(n, X) enum virtchnl_static_assert_enum_##X \ + {virtchnl_static_assert_##X = (n) / ((sizeof(struct X) == (n)) ? 1 : 0)} + +/* Virtual channel message descriptor. This overlays the admin queue + * descriptor. All other data is passed in external buffers. + */ + +struct virtchnl_msg { + u8 pad[8]; /* AQ flags/opcode/len/retval fields */ + enum virtchnl_ops v_opcode; /* avoid confusion with desc->opcode */ + enum virtchnl_status_code v_retval; /* ditto for desc->retval */ + u32 vfid; /* used by PF when sending to VF */ +}; + +VIRTCHNL_CHECK_STRUCT_LEN(20, virtchnl_msg); + +/* Message descriptions and data structures.*/ + +/* VIRTCHNL_OP_VERSION + * VF posts its version number to the PF. PF responds with its version number + * in the same format, along with a return code. + * Reply from PF has its major/minor versions also in param0 and param1. + * If there is a major version mismatch, then the VF cannot operate. + * If there is a minor version mismatch, then the VF can operate but should + * add a warning to the system log. + * + * This enum element MUST always be specified as == 1, regardless of other + * changes in the API. The PF must always respond to this message without + * error regardless of version mismatch. + */ +#define VIRTCHNL_VERSION_MAJOR 1 +#define VIRTCHNL_VERSION_MINOR 1 +#define VIRTCHNL_VERSION_MINOR_NO_VF_CAPS 0 + +struct virtchnl_version_info { + u32 major; + u32 minor; +}; + +VIRTCHNL_CHECK_STRUCT_LEN(8, virtchnl_version_info); + +#define VF_IS_V10(_v) (((_v)->major == 1) && ((_v)->minor == 0)) +#define VF_IS_V11(_ver) (((_ver)->major == 1) && ((_ver)->minor == 1)) + +/* VIRTCHNL_OP_RESET_VF + * VF sends this request to PF with no parameters + * PF does NOT respond! VF driver must delay then poll VFGEN_RSTAT register + * until reset completion is indicated. The admin queue must be reinitialized + * after this operation. + * + * When reset is complete, PF must ensure that all queues in all VSIs associated + * with the VF are stopped, all queue configurations in the HMC are set to 0, + * and all MAC and VLAN filters (except the default MAC address) on all VSIs + * are cleared. + */ + +/* VSI types that use VIRTCHNL interface for VF-PF communication. VSI_SRIOV + * vsi_type should always be 6 for backward compatibility. Add other fields + * as needed. + */ +enum virtchnl_vsi_type { + VIRTCHNL_VSI_TYPE_INVALID = 0, + VIRTCHNL_VSI_SRIOV = 6, +}; + +/* VIRTCHNL_OP_GET_VF_RESOURCES + * Version 1.0 VF sends this request to PF with no parameters + * Version 1.1 VF sends this request to PF with u32 bitmap of its capabilities + * PF responds with an indirect message containing + * virtchnl_vf_resource and one or more + * virtchnl_vsi_resource structures. + */ + +struct virtchnl_vsi_resource { + u16 vsi_id; + u16 num_queue_pairs; + enum virtchnl_vsi_type vsi_type; + u16 qset_handle; + u8 default_mac_addr[VIRTCHNL_ETH_LENGTH_OF_ADDRESS]; +}; + +VIRTCHNL_CHECK_STRUCT_LEN(16, virtchnl_vsi_resource); + +/* VF offload flags + * VIRTCHNL_VF_OFFLOAD_L2 flag is inclusive of base mode L2 offloads including + * TX/RX Checksum offloading and TSO for non-tunnelled packets. + */ +#define VIRTCHNL_VF_OFFLOAD_L2 0x00000001 +#define VIRTCHNL_VF_OFFLOAD_IWARP 0x00000002 +#define VIRTCHNL_VF_OFFLOAD_RSVD 0x00000004 +#define VIRTCHNL_VF_OFFLOAD_RSS_AQ 0x00000008 +#define VIRTCHNL_VF_OFFLOAD_RSS_REG 0x00000010 +#define VIRTCHNL_VF_OFFLOAD_WB_ON_ITR 0x00000020 +#define VIRTCHNL_VF_OFFLOAD_REQ_QUEUES 0x00000040 +#define VIRTCHNL_VF_OFFLOAD_VLAN 0x00010000 +#define VIRTCHNL_VF_OFFLOAD_RX_POLLING 0x00020000 +#define VIRTCHNL_VF_OFFLOAD_RSS_PCTYPE_V2 0x00040000 +#define VIRTCHNL_VF_OFFLOAD_RSS_PF 0X00080000 +#define VIRTCHNL_VF_OFFLOAD_ENCAP 0X00100000 +#define VIRTCHNL_VF_OFFLOAD_ENCAP_CSUM 0X00200000 +#define VIRTCHNL_VF_OFFLOAD_RX_ENCAP_CSUM 0X00400000 + +#define VF_BASE_MODE_OFFLOADS (VIRTCHNL_VF_OFFLOAD_L2 | \ + VIRTCHNL_VF_OFFLOAD_VLAN | \ + VIRTCHNL_VF_OFFLOAD_RSS_PF) + +struct virtchnl_vf_resource { + u16 num_vsis; + u16 num_queue_pairs; + u16 max_vectors; + u16 max_mtu; + + u32 vf_offload_flags; + u32 rss_key_size; + u32 rss_lut_size; + + struct virtchnl_vsi_resource vsi_res[1]; +}; + +VIRTCHNL_CHECK_STRUCT_LEN(36, virtchnl_vf_resource); + +/* VIRTCHNL_OP_CONFIG_TX_QUEUE + * VF sends this message to set up parameters for one TX queue. + * External data buffer contains one instance of virtchnl_txq_info. + * PF configures requested queue and returns a status code. + */ + +/* Tx queue config info */ +struct virtchnl_txq_info { + u16 vsi_id; + u16 queue_id; + u16 ring_len; /* number of descriptors, multiple of 8 */ + u16 headwb_enabled; /* deprecated with AVF 1.0 */ + u64 dma_ring_addr; + u64 dma_headwb_addr; /* deprecated with AVF 1.0 */ +}; + +VIRTCHNL_CHECK_STRUCT_LEN(24, virtchnl_txq_info); + +/* VIRTCHNL_OP_CONFIG_RX_QUEUE + * VF sends this message to set up parameters for one RX queue. + * External data buffer contains one instance of virtchnl_rxq_info. + * PF configures requested queue and returns a status code. + */ + +/* Rx queue config info */ +struct virtchnl_rxq_info { + u16 vsi_id; + u16 queue_id; + u32 ring_len; /* number of descriptors, multiple of 32 */ + u16 hdr_size; + u16 splithdr_enabled; /* deprecated with AVF 1.0 */ + u32 databuffer_size; + u32 max_pkt_size; + u32 pad1; + u64 dma_ring_addr; + enum virtchnl_rx_hsplit rx_split_pos; /* deprecated with AVF 1.0 */ + u32 pad2; +}; + +VIRTCHNL_CHECK_STRUCT_LEN(40, virtchnl_rxq_info); + +/* VIRTCHNL_OP_CONFIG_VSI_QUEUES + * VF sends this message to set parameters for all active TX and RX queues + * associated with the specified VSI. + * PF configures queues and returns status. + * If the number of queues specified is greater than the number of queues + * associated with the VSI, an error is returned and no queues are configured. + */ +struct virtchnl_queue_pair_info { + /* NOTE: vsi_id and queue_id should be identical for both queues. */ + struct virtchnl_txq_info txq; + struct virtchnl_rxq_info rxq; +}; + +VIRTCHNL_CHECK_STRUCT_LEN(64, virtchnl_queue_pair_info); + +struct virtchnl_vsi_queue_config_info { + u16 vsi_id; + u16 num_queue_pairs; + u32 pad; + struct virtchnl_queue_pair_info qpair[1]; +}; + +VIRTCHNL_CHECK_STRUCT_LEN(72, virtchnl_vsi_queue_config_info); + +/* VIRTCHNL_OP_REQUEST_QUEUES + * VF sends this message to request the PF to allocate additional queues to + * this VF. Each VF gets a guaranteed number of queues on init but asking for + * additional queues must be negotiated. This is a best effort request as it + * is possible the PF does not have enough queues left to support the request. + * If the PF cannot support the number requested it will respond with the + * maximum number it is able to support; otherwise it will respond with the + * number requested. + */ + +/* VF resource request */ +struct virtchnl_vf_res_request { + u16 num_queue_pairs; +}; + +/* VIRTCHNL_OP_CONFIG_IRQ_MAP + * VF uses this message to map vectors to queues. + * The rxq_map and txq_map fields are bitmaps used to indicate which queues + * are to be associated with the specified vector. + * The "other" causes are always mapped to vector 0. + * PF configures interrupt mapping and returns status. + */ +struct virtchnl_vector_map { + u16 vsi_id; + u16 vector_id; + u16 rxq_map; + u16 txq_map; + u16 rxitr_idx; + u16 txitr_idx; +}; + +VIRTCHNL_CHECK_STRUCT_LEN(12, virtchnl_vector_map); + +struct virtchnl_irq_map_info { + u16 num_vectors; + struct virtchnl_vector_map vecmap[1]; +}; + +VIRTCHNL_CHECK_STRUCT_LEN(14, virtchnl_irq_map_info); + +/* VIRTCHNL_OP_ENABLE_QUEUES + * VIRTCHNL_OP_DISABLE_QUEUES + * VF sends these message to enable or disable TX/RX queue pairs. + * The queues fields are bitmaps indicating which queues to act upon. + * (Currently, we only support 16 queues per VF, but we make the field + * u32 to allow for expansion.) + * PF performs requested action and returns status. + */ +struct virtchnl_queue_select { + u16 vsi_id; + u16 pad; + u32 rx_queues; + u32 tx_queues; +}; + +VIRTCHNL_CHECK_STRUCT_LEN(12, virtchnl_queue_select); + +/* VIRTCHNL_OP_ADD_ETH_ADDR + * VF sends this message in order to add one or more unicast or multicast + * address filters for the specified VSI. + * PF adds the filters and returns status. + */ + +/* VIRTCHNL_OP_DEL_ETH_ADDR + * VF sends this message in order to remove one or more unicast or multicast + * filters for the specified VSI. + * PF removes the filters and returns status. + */ + +struct virtchnl_ether_addr { + u8 addr[VIRTCHNL_ETH_LENGTH_OF_ADDRESS]; + u8 pad[2]; +}; + +VIRTCHNL_CHECK_STRUCT_LEN(8, virtchnl_ether_addr); + +struct virtchnl_ether_addr_list { + u16 vsi_id; + u16 num_elements; + struct virtchnl_ether_addr list[1]; +}; + +VIRTCHNL_CHECK_STRUCT_LEN(12, virtchnl_ether_addr_list); + +#ifdef VIRTCHNL_SOL_VF_SUPPORT +/* VIRTCHNL_OP_GET_ADDNL_SOL_CONFIG + * VF sends this message to get the default MTU and list of additional ethernet + * addresses it is allowed to use. + * PF responds with an indirect message containing + * virtchnl_addnl_solaris_config with zero or more + * virtchnl_ether_addr structures. + * + * It is expected that this operation will only ever be needed for Solaris VFs + * running under a Solaris PF. + */ +struct virtchnl_addnl_solaris_config { + u16 default_mtu; + struct virtchnl_ether_addr_list al; +}; + +#endif +/* VIRTCHNL_OP_ADD_VLAN + * VF sends this message to add one or more VLAN tag filters for receives. + * PF adds the filters and returns status. + * If a port VLAN is configured by the PF, this operation will return an + * error to the VF. + */ + +/* VIRTCHNL_OP_DEL_VLAN + * VF sends this message to remove one or more VLAN tag filters for receives. + * PF removes the filters and returns status. + * If a port VLAN is configured by the PF, this operation will return an + * error to the VF. + */ + +struct virtchnl_vlan_filter_list { + u16 vsi_id; + u16 num_elements; + u16 vlan_id[1]; +}; + +VIRTCHNL_CHECK_STRUCT_LEN(6, virtchnl_vlan_filter_list); + +/* VIRTCHNL_OP_CONFIG_PROMISCUOUS_MODE + * VF sends VSI id and flags. + * PF returns status code in retval. + * Note: we assume that broadcast accept mode is always enabled. + */ +struct virtchnl_promisc_info { + u16 vsi_id; + u16 flags; +}; + +VIRTCHNL_CHECK_STRUCT_LEN(4, virtchnl_promisc_info); + +#define FLAG_VF_UNICAST_PROMISC 0x00000001 +#define FLAG_VF_MULTICAST_PROMISC 0x00000002 + +/* VIRTCHNL_OP_GET_STATS + * VF sends this message to request stats for the selected VSI. VF uses + * the virtchnl_queue_select struct to specify the VSI. The queue_id + * field is ignored by the PF. + * + * PF replies with struct eth_stats in an external buffer. + */ + +/* VIRTCHNL_OP_CONFIG_RSS_KEY + * VIRTCHNL_OP_CONFIG_RSS_LUT + * VF sends these messages to configure RSS. Only supported if both PF + * and VF drivers set the VIRTCHNL_VF_OFFLOAD_RSS_PF bit during + * configuration negotiation. If this is the case, then the RSS fields in + * the VF resource struct are valid. + * Both the key and LUT are initialized to 0 by the PF, meaning that + * RSS is effectively disabled until set up by the VF. + */ +struct virtchnl_rss_key { + u16 vsi_id; + u16 key_len; + u8 key[1]; /* RSS hash key, packed bytes */ +}; + +VIRTCHNL_CHECK_STRUCT_LEN(6, virtchnl_rss_key); + +struct virtchnl_rss_lut { + u16 vsi_id; + u16 lut_entries; + u8 lut[1]; /* RSS lookup table*/ +}; + +VIRTCHNL_CHECK_STRUCT_LEN(6, virtchnl_rss_lut); + +/* VIRTCHNL_OP_GET_RSS_HENA_CAPS + * VIRTCHNL_OP_SET_RSS_HENA + * VF sends these messages to get and set the hash filter enable bits for RSS. + * By default, the PF sets these to all possible traffic types that the + * hardware supports. The VF can query this value if it wants to change the + * traffic types that are hashed by the hardware. + */ +struct virtchnl_rss_hena { + u64 hena; +}; + +VIRTCHNL_CHECK_STRUCT_LEN(8, virtchnl_rss_hena); + +/* VIRTCHNL_OP_EVENT + * PF sends this message to inform the VF driver of events that may affect it. + * No direct response is expected from the VF, though it may generate other + * messages in response to this one. + */ +enum virtchnl_event_codes { + VIRTCHNL_EVENT_UNKNOWN = 0, + VIRTCHNL_EVENT_LINK_CHANGE, + VIRTCHNL_EVENT_RESET_IMPENDING, + VIRTCHNL_EVENT_PF_DRIVER_CLOSE, +}; + +#define PF_EVENT_SEVERITY_INFO 0 +#define PF_EVENT_SEVERITY_ATTENTION 1 +#define PF_EVENT_SEVERITY_ACTION_REQUIRED 2 +#define PF_EVENT_SEVERITY_CERTAIN_DOOM 255 + +struct virtchnl_pf_event { + enum virtchnl_event_codes event; + union { + struct { + enum virtchnl_link_speed link_speed; + bool link_status; + } link_event; + } event_data; + + int severity; +}; + +VIRTCHNL_CHECK_STRUCT_LEN(16, virtchnl_pf_event); + +#ifdef VIRTCHNL_IWARP + +/* VIRTCHNL_OP_CONFIG_IWARP_IRQ_MAP + * VF uses this message to request PF to map IWARP vectors to IWARP queues. + * The request for this originates from the VF IWARP driver through + * a client interface between VF LAN and VF IWARP driver. + * A vector could have an AEQ and CEQ attached to it although + * there is a single AEQ per VF IWARP instance in which case + * most vectors will have an INVALID_IDX for aeq and valid idx for ceq. + * There will never be a case where there will be multiple CEQs attached + * to a single vector. + * PF configures interrupt mapping and returns status. + */ + +/* HW does not define a type value for AEQ; only for RX/TX and CEQ. + * In order for us to keep the interface simple, SW will define a + * unique type value for AEQ. + */ +#define QUEUE_TYPE_PE_AEQ 0x80 +#define QUEUE_INVALID_IDX 0xFFFF + +struct virtchnl_iwarp_qv_info { + u32 v_idx; /* msix_vector */ + u16 ceq_idx; + u16 aeq_idx; + u8 itr_idx; +}; + +VIRTCHNL_CHECK_STRUCT_LEN(12, virtchnl_iwarp_qv_info); + +struct virtchnl_iwarp_qvlist_info { + u32 num_vectors; + struct virtchnl_iwarp_qv_info qv_info[1]; +}; + +VIRTCHNL_CHECK_STRUCT_LEN(16, virtchnl_iwarp_qvlist_info); + +#endif + +/* VF reset states - these are written into the RSTAT register: + * VFGEN_RSTAT on the VF + * When the PF initiates a reset, it writes 0 + * When the reset is complete, it writes 1 + * When the PF detects that the VF has recovered, it writes 2 + * VF checks this register periodically to determine if a reset has occurred, + * then polls it to know when the reset is complete. + * If either the PF or VF reads the register while the hardware + * is in a reset state, it will return DEADBEEF, which, when masked + * will result in 3. + */ +enum virtchnl_vfr_states { + VIRTCHNL_VFR_INPROGRESS = 0, + VIRTCHNL_VFR_COMPLETED, + VIRTCHNL_VFR_VFACTIVE, +}; + +/** + * virtchnl_vc_validate_vf_msg + * @ver: Virtchnl version info + * @v_opcode: Opcode for the message + * @msg: pointer to the msg buffer + * @msglen: msg length + * + * validate msg format against struct for each opcode + */ +static inline int +virtchnl_vc_validate_vf_msg(struct virtchnl_version_info *ver, u32 v_opcode, + u8 *msg, u16 msglen) +{ + bool err_msg_format = false; + int valid_len = 0; + + /* Validate message length. */ + switch (v_opcode) { + case VIRTCHNL_OP_VERSION: + valid_len = sizeof(struct virtchnl_version_info); + break; + case VIRTCHNL_OP_RESET_VF: + break; + case VIRTCHNL_OP_GET_VF_RESOURCES: + if (VF_IS_V11(ver)) + valid_len = sizeof(u32); + break; + case VIRTCHNL_OP_CONFIG_TX_QUEUE: + valid_len = sizeof(struct virtchnl_txq_info); + break; + case VIRTCHNL_OP_CONFIG_RX_QUEUE: + valid_len = sizeof(struct virtchnl_rxq_info); + break; + case VIRTCHNL_OP_CONFIG_VSI_QUEUES: + valid_len = sizeof(struct virtchnl_vsi_queue_config_info); + if (msglen >= valid_len) { + struct virtchnl_vsi_queue_config_info *vqc = + (struct virtchnl_vsi_queue_config_info *)msg; + valid_len += (vqc->num_queue_pairs * + sizeof(struct + virtchnl_queue_pair_info)); + if (vqc->num_queue_pairs == 0) + err_msg_format = true; + } + break; + case VIRTCHNL_OP_CONFIG_IRQ_MAP: + valid_len = sizeof(struct virtchnl_irq_map_info); + if (msglen >= valid_len) { + struct virtchnl_irq_map_info *vimi = + (struct virtchnl_irq_map_info *)msg; + valid_len += (vimi->num_vectors * + sizeof(struct virtchnl_vector_map)); + if (vimi->num_vectors == 0) + err_msg_format = true; + } + break; + case VIRTCHNL_OP_ENABLE_QUEUES: + case VIRTCHNL_OP_DISABLE_QUEUES: + valid_len = sizeof(struct virtchnl_queue_select); + break; + case VIRTCHNL_OP_ADD_ETH_ADDR: + case VIRTCHNL_OP_DEL_ETH_ADDR: + valid_len = sizeof(struct virtchnl_ether_addr_list); + if (msglen >= valid_len) { + struct virtchnl_ether_addr_list *veal = + (struct virtchnl_ether_addr_list *)msg; + valid_len += veal->num_elements * + sizeof(struct virtchnl_ether_addr); + if (veal->num_elements == 0) + err_msg_format = true; + } + break; + case VIRTCHNL_OP_ADD_VLAN: + case VIRTCHNL_OP_DEL_VLAN: + valid_len = sizeof(struct virtchnl_vlan_filter_list); + if (msglen >= valid_len) { + struct virtchnl_vlan_filter_list *vfl = + (struct virtchnl_vlan_filter_list *)msg; + valid_len += vfl->num_elements * sizeof(u16); + if (vfl->num_elements == 0) + err_msg_format = true; + } + break; + case VIRTCHNL_OP_CONFIG_PROMISCUOUS_MODE: + valid_len = sizeof(struct virtchnl_promisc_info); + break; + case VIRTCHNL_OP_GET_STATS: + valid_len = sizeof(struct virtchnl_queue_select); + break; +#ifdef VIRTCHNL_IWARP + case VIRTCHNL_OP_IWARP: + /* These messages are opaque to us and will be validated in + * the RDMA client code. We just need to check for nonzero + * length. The firmware will enforce max length restrictions. + */ + if (msglen) + valid_len = msglen; + else + err_msg_format = true; + break; + case VIRTCHNL_OP_RELEASE_IWARP_IRQ_MAP: + break; + case VIRTCHNL_OP_CONFIG_IWARP_IRQ_MAP: + valid_len = sizeof(struct virtchnl_iwarp_qvlist_info); + if (msglen >= valid_len) { + struct virtchnl_iwarp_qvlist_info *qv = + (struct virtchnl_iwarp_qvlist_info *)msg; + if (qv->num_vectors == 0) { + err_msg_format = true; + break; + } + valid_len += ((qv->num_vectors - 1) * + sizeof(struct virtchnl_iwarp_qv_info)); + } + break; +#endif + case VIRTCHNL_OP_CONFIG_RSS_KEY: + valid_len = sizeof(struct virtchnl_rss_key); + if (msglen >= valid_len) { + struct virtchnl_rss_key *vrk = + (struct virtchnl_rss_key *)msg; + valid_len += vrk->key_len - 1; + } + break; + case VIRTCHNL_OP_CONFIG_RSS_LUT: + valid_len = sizeof(struct virtchnl_rss_lut); + if (msglen >= valid_len) { + struct virtchnl_rss_lut *vrl = + (struct virtchnl_rss_lut *)msg; + valid_len += vrl->lut_entries - 1; + } + break; + case VIRTCHNL_OP_GET_RSS_HENA_CAPS: + break; + case VIRTCHNL_OP_SET_RSS_HENA: + valid_len = sizeof(struct virtchnl_rss_hena); + break; + case VIRTCHNL_OP_ENABLE_VLAN_STRIPPING: + case VIRTCHNL_OP_DISABLE_VLAN_STRIPPING: + break; + case VIRTCHNL_OP_REQUEST_QUEUES: + valid_len = sizeof(struct virtchnl_vf_res_request); + break; + /* These are always errors coming from the VF. */ + case VIRTCHNL_OP_EVENT: + case VIRTCHNL_OP_UNKNOWN: + default: + return VIRTCHNL_ERR_PARAM; + } + /* few more checks */ + if ((valid_len != msglen) || (err_msg_format)) + return VIRTCHNL_STATUS_ERR_OPCODE_MISMATCH; + + return 0; +} +#endif /* _VIRTCHNL_H_ */ diff --git a/drivers/net/i40e/i40e_ethdev.h b/drivers/net/i40e/i40e_ethdev.h index 183dc17..bda7450 100644 --- a/drivers/net/i40e/i40e_ethdev.h +++ b/drivers/net/i40e/i40e_ethdev.h @@ -102,7 +102,7 @@ /* Linux PF host with virtchnl version 1.1 */ #define PF_IS_V11(vf) \ - (((vf)->version_major == I40E_VIRTCHNL_VERSION_MAJOR) && \ + (((vf)->version_major == VIRTCHNL_VERSION_MAJOR) && \ ((vf)->version_minor == 1)) /* index flex payload per layer */ @@ -767,7 +767,7 @@ struct i40e_vf { /* Event from pf */ bool dev_closed; bool link_up; - enum i40e_aq_link_speed link_speed; + enum virtchnl_link_speed link_speed; bool vf_reset; volatile uint32_t pend_cmd; /* pending command not finished yet */ int32_t cmd_retval; /* return value of the cmd response from PF */ @@ -775,8 +775,8 @@ struct i40e_vf { uint8_t *aq_resp; /* buffer to store the adminq response from PF */ /* VSI info */ - struct i40e_virtchnl_vf_resource *vf_res; /* All VSIs */ - struct i40e_virtchnl_vsi_resource *vsi_res; /* LAN VSI */ + struct virtchnl_vf_resource *vf_res; /* All VSIs */ + struct virtchnl_vsi_resource *vsi_res; /* LAN VSI */ struct i40e_vsi vsi; uint64_t flags; }; diff --git a/drivers/net/i40e/i40e_ethdev_vf.c b/drivers/net/i40e/i40e_ethdev_vf.c index 2d5a9b5..6718eee 100644 --- a/drivers/net/i40e/i40e_ethdev_vf.c +++ b/drivers/net/i40e/i40e_ethdev_vf.c @@ -77,7 +77,7 @@ #define MAX_RESET_WAIT_CNT 20 struct i40evf_arq_msg_info { - enum i40e_virtchnl_ops ops; + enum virtchnl_ops ops; enum i40e_status_code result; uint16_t buf_len; uint16_t msg_len; @@ -85,7 +85,7 @@ struct i40evf_arq_msg_info { }; struct vf_cmd_info { - enum i40e_virtchnl_ops ops; + enum virtchnl_ops ops; uint8_t *in_args; uint32_t in_args_size; uint8_t *out_buffer; @@ -244,7 +244,7 @@ i40evf_read_pfmsg(struct rte_eth_dev *dev, struct i40evf_arq_msg_info *data) struct i40e_hw *hw = I40E_DEV_PRIVATE_TO_HW(dev->data->dev_private); struct i40e_vf *vf = I40EVF_DEV_PRIVATE_TO_VF(dev->data->dev_private); struct i40e_arq_event_info event; - enum i40e_virtchnl_ops opcode; + enum virtchnl_ops opcode; enum i40e_status_code retval; int ret; enum i40evf_aq_result result = I40EVF_MSG_NON; @@ -259,16 +259,16 @@ i40evf_read_pfmsg(struct rte_eth_dev *dev, struct i40evf_arq_msg_info *data) return result; } - opcode = (enum i40e_virtchnl_ops)rte_le_to_cpu_32(event.desc.cookie_high); + opcode = (enum virtchnl_ops)rte_le_to_cpu_32(event.desc.cookie_high); retval = (enum i40e_status_code)rte_le_to_cpu_32(event.desc.cookie_low); /* pf sys event */ - if (opcode == I40E_VIRTCHNL_OP_EVENT) { - struct i40e_virtchnl_pf_event *vpe = - (struct i40e_virtchnl_pf_event *)event.msg_buf; + if (opcode == VIRTCHNL_OP_EVENT) { + struct virtchnl_pf_event *vpe = + (struct virtchnl_pf_event *)event.msg_buf; result = I40EVF_MSG_SYS; switch (vpe->event) { - case I40E_VIRTCHNL_EVENT_LINK_CHANGE: + case VIRTCHNL_EVENT_LINK_CHANGE: vf->link_up = vpe->event_data.link_event.link_status; vf->link_speed = @@ -277,12 +277,12 @@ i40evf_read_pfmsg(struct rte_eth_dev *dev, struct i40evf_arq_msg_info *data) PMD_DRV_LOG(INFO, "Link status update:%s", vf->link_up ? "up" : "down"); break; - case I40E_VIRTCHNL_EVENT_RESET_IMPENDING: + case VIRTCHNL_EVENT_RESET_IMPENDING: vf->vf_reset = true; vf->pend_msg |= PFMSG_RESET_IMPENDING; PMD_DRV_LOG(INFO, "vf is reseting"); break; - case I40E_VIRTCHNL_EVENT_PF_DRIVER_CLOSE: + case VIRTCHNL_EVENT_PF_DRIVER_CLOSE: vf->dev_closed = true; vf->pend_msg |= PFMSG_DRIVER_CLOSE; PMD_DRV_LOG(INFO, "PF driver closed"); @@ -312,17 +312,17 @@ static inline void _clear_cmd(struct i40e_vf *vf) { rte_wmb(); - vf->pend_cmd = I40E_VIRTCHNL_OP_UNKNOWN; + vf->pend_cmd = VIRTCHNL_OP_UNKNOWN; } /* * Check there is pending cmd in execution. If none, set new command. */ static inline int -_atomic_set_cmd(struct i40e_vf *vf, enum i40e_virtchnl_ops ops) +_atomic_set_cmd(struct i40e_vf *vf, enum virtchnl_ops ops) { int ret = rte_atomic32_cmpset(&vf->pend_cmd, - I40E_VIRTCHNL_OP_UNKNOWN, ops); + VIRTCHNL_OP_UNKNOWN, ops); if (!ret) PMD_DRV_LOG(ERR, "There is incomplete cmd %d", vf->pend_cmd); @@ -347,7 +347,7 @@ i40evf_execute_vf_cmd(struct rte_eth_dev *dev, struct vf_cmd_info *args) info.msg = args->out_buffer; info.buf_len = args->out_size; - info.ops = I40E_VIRTCHNL_OP_UNKNOWN; + info.ops = VIRTCHNL_OP_UNKNOWN; info.result = I40E_SUCCESS; err = i40e_aq_send_msg_to_pf(hw, args->ops, I40E_SUCCESS, @@ -359,12 +359,12 @@ i40evf_execute_vf_cmd(struct rte_eth_dev *dev, struct vf_cmd_info *args) } switch (args->ops) { - case I40E_VIRTCHNL_OP_RESET_VF: + case VIRTCHNL_OP_RESET_VF: /*no need to process in this function */ err = 0; break; - case I40E_VIRTCHNL_OP_VERSION: - case I40E_VIRTCHNL_OP_GET_VF_RESOURCES: + case VIRTCHNL_OP_VERSION: + case VIRTCHNL_OP_GET_VF_RESOURCES: /* for init adminq commands, need to poll the response */ err = -1; do { @@ -385,7 +385,7 @@ i40evf_execute_vf_cmd(struct rte_eth_dev *dev, struct vf_cmd_info *args) /* for other adminq in running time, waiting the cmd done flag */ err = -1; do { - if (vf->pend_cmd == I40E_VIRTCHNL_OP_UNKNOWN) { + if (vf->pend_cmd == VIRTCHNL_OP_UNKNOWN) { err = 0; break; } @@ -404,15 +404,15 @@ i40evf_execute_vf_cmd(struct rte_eth_dev *dev, struct vf_cmd_info *args) static int i40evf_check_api_version(struct rte_eth_dev *dev) { - struct i40e_virtchnl_version_info version, *pver; + struct virtchnl_version_info version, *pver; int err; struct vf_cmd_info args; struct i40e_vf *vf = I40EVF_DEV_PRIVATE_TO_VF(dev->data->dev_private); - version.major = I40E_VIRTCHNL_VERSION_MAJOR; - version.minor = I40E_VIRTCHNL_VERSION_MINOR; + version.major = VIRTCHNL_VERSION_MAJOR; + version.minor = VIRTCHNL_VERSION_MINOR; - args.ops = I40E_VIRTCHNL_OP_VERSION; + args.ops = VIRTCHNL_OP_VERSION; args.in_args = (uint8_t *)&version; args.in_args_size = sizeof(version); args.out_buffer = vf->aq_resp; @@ -424,19 +424,19 @@ i40evf_check_api_version(struct rte_eth_dev *dev) return err; } - pver = (struct i40e_virtchnl_version_info *)args.out_buffer; + pver = (struct virtchnl_version_info *)args.out_buffer; vf->version_major = pver->major; vf->version_minor = pver->minor; if (vf->version_major == I40E_DPDK_VERSION_MAJOR) PMD_DRV_LOG(INFO, "Peer is DPDK PF host"); - else if ((vf->version_major == I40E_VIRTCHNL_VERSION_MAJOR) && - (vf->version_minor <= I40E_VIRTCHNL_VERSION_MINOR)) + else if ((vf->version_major == VIRTCHNL_VERSION_MAJOR) && + (vf->version_minor <= VIRTCHNL_VERSION_MINOR)) PMD_DRV_LOG(INFO, "Peer is Linux PF host"); else { PMD_INIT_LOG(ERR, "PF/VF API version mismatch:(%u.%u)-(%u.%u)", vf->version_major, vf->version_minor, - I40E_VIRTCHNL_VERSION_MAJOR, - I40E_VIRTCHNL_VERSION_MINOR); + VIRTCHNL_VERSION_MAJOR, + VIRTCHNL_VERSION_MINOR); return -1; } @@ -452,15 +452,15 @@ i40evf_get_vf_resource(struct rte_eth_dev *dev) struct vf_cmd_info args; uint32_t caps, len; - args.ops = I40E_VIRTCHNL_OP_GET_VF_RESOURCES; + args.ops = VIRTCHNL_OP_GET_VF_RESOURCES; args.out_buffer = vf->aq_resp; args.out_size = I40E_AQ_BUF_SZ; if (PF_IS_V11(vf)) { - caps = I40E_VIRTCHNL_VF_OFFLOAD_L2 | - I40E_VIRTCHNL_VF_OFFLOAD_RSS_AQ | - I40E_VIRTCHNL_VF_OFFLOAD_RSS_REG | - I40E_VIRTCHNL_VF_OFFLOAD_VLAN | - I40E_VIRTCHNL_VF_OFFLOAD_RX_POLLING; + caps = VIRTCHNL_VF_OFFLOAD_L2 | + VIRTCHNL_VF_OFFLOAD_RSS_AQ | + VIRTCHNL_VF_OFFLOAD_RSS_REG | + VIRTCHNL_VF_OFFLOAD_VLAN | + VIRTCHNL_VF_OFFLOAD_RX_POLLING; args.in_args = (uint8_t *)∩︀ args.in_args_size = sizeof(caps); } else { @@ -474,8 +474,8 @@ i40evf_get_vf_resource(struct rte_eth_dev *dev) return err; } - len = sizeof(struct i40e_virtchnl_vf_resource) + - I40E_MAX_VF_VSI * sizeof(struct i40e_virtchnl_vsi_resource); + len = sizeof(struct virtchnl_vf_resource) + + I40E_MAX_VF_VSI * sizeof(struct virtchnl_vsi_resource); (void)rte_memcpy(vf->vf_res, args.out_buffer, RTE_MIN(args.out_size, len)); @@ -492,18 +492,18 @@ i40evf_config_promisc(struct rte_eth_dev *dev, struct i40e_vf *vf = I40EVF_DEV_PRIVATE_TO_VF(dev->data->dev_private); int err; struct vf_cmd_info args; - struct i40e_virtchnl_promisc_info promisc; + struct virtchnl_promisc_info promisc; promisc.flags = 0; promisc.vsi_id = vf->vsi_res->vsi_id; if (enable_unicast) - promisc.flags |= I40E_FLAG_VF_UNICAST_PROMISC; + promisc.flags |= FLAG_VF_UNICAST_PROMISC; if (enable_multicast) - promisc.flags |= I40E_FLAG_VF_MULTICAST_PROMISC; + promisc.flags |= FLAG_VF_MULTICAST_PROMISC; - args.ops = I40E_VIRTCHNL_OP_CONFIG_PROMISCUOUS_MODE; + args.ops = VIRTCHNL_OP_CONFIG_PROMISCUOUS_MODE; args.in_args = (uint8_t *)&promisc; args.in_args_size = sizeof(promisc); args.out_buffer = vf->aq_resp; @@ -525,12 +525,12 @@ i40evf_config_vlan_offload(struct rte_eth_dev *dev, struct i40e_vf *vf = I40EVF_DEV_PRIVATE_TO_VF(dev->data->dev_private); int err; struct vf_cmd_info args; - struct i40e_virtchnl_vlan_offload_info offload; + struct virtchnl_vlan_offload_info offload; offload.vsi_id = vf->vsi_res->vsi_id; offload.enable_vlan_strip = enable_vlan_strip; - args.ops = (enum i40e_virtchnl_ops)I40E_VIRTCHNL_OP_CFG_VLAN_OFFLOAD; + args.ops = (enum virtchnl_ops)I40E_VIRTCHNL_OP_CFG_VLAN_OFFLOAD; args.in_args = (uint8_t *)&offload; args.in_args_size = sizeof(offload); args.out_buffer = vf->aq_resp; @@ -550,7 +550,7 @@ i40evf_config_vlan_pvid(struct rte_eth_dev *dev, struct i40e_vf *vf = I40EVF_DEV_PRIVATE_TO_VF(dev->data->dev_private); int err; struct vf_cmd_info args; - struct i40e_virtchnl_pvid_info tpid_info; + struct virtchnl_pvid_info tpid_info; if (info == NULL) { PMD_DRV_LOG(ERR, "invalid parameters"); @@ -561,7 +561,7 @@ i40evf_config_vlan_pvid(struct rte_eth_dev *dev, tpid_info.vsi_id = vf->vsi_res->vsi_id; (void)rte_memcpy(&tpid_info.info, info, sizeof(*info)); - args.ops = (enum i40e_virtchnl_ops)I40E_VIRTCHNL_OP_CFG_VLAN_PVID; + args.ops = (enum virtchnl_ops)I40E_VIRTCHNL_OP_CFG_VLAN_PVID; args.in_args = (uint8_t *)&tpid_info; args.in_args_size = sizeof(tpid_info); args.out_buffer = vf->aq_resp; @@ -575,7 +575,7 @@ i40evf_config_vlan_pvid(struct rte_eth_dev *dev, } static void -i40evf_fill_virtchnl_vsi_txq_info(struct i40e_virtchnl_txq_info *txq_info, +i40evf_fill_virtchnl_vsi_txq_info(struct virtchnl_txq_info *txq_info, uint16_t vsi_id, uint16_t queue_id, uint16_t nb_txq, @@ -590,7 +590,7 @@ i40evf_fill_virtchnl_vsi_txq_info(struct i40e_virtchnl_txq_info *txq_info, } static void -i40evf_fill_virtchnl_vsi_rxq_info(struct i40e_virtchnl_rxq_info *rxq_info, +i40evf_fill_virtchnl_vsi_rxq_info(struct virtchnl_rxq_info *rxq_info, uint16_t vsi_id, uint16_t queue_id, uint16_t nb_rxq, @@ -618,8 +618,8 @@ i40evf_configure_vsi_queues(struct rte_eth_dev *dev) (struct i40e_rx_queue **)dev->data->rx_queues; struct i40e_tx_queue **txq = (struct i40e_tx_queue **)dev->data->tx_queues; - struct i40e_virtchnl_vsi_queue_config_info *vc_vqci; - struct i40e_virtchnl_queue_pair_info *vc_qpi; + struct virtchnl_vsi_queue_config_info *vc_vqci; + struct virtchnl_queue_pair_info *vc_qpi; struct vf_cmd_info args; uint16_t i, nb_qp = vf->num_queue_pairs; const uint32_t size = @@ -628,7 +628,7 @@ i40evf_configure_vsi_queues(struct rte_eth_dev *dev) int ret; memset(buff, 0, sizeof(buff)); - vc_vqci = (struct i40e_virtchnl_vsi_queue_config_info *)buff; + vc_vqci = (struct virtchnl_vsi_queue_config_info *)buff; vc_vqci->vsi_id = vf->vsi_res->vsi_id; vc_vqci->num_queue_pairs = nb_qp; @@ -640,15 +640,14 @@ i40evf_configure_vsi_queues(struct rte_eth_dev *dev) vf->max_pkt_len, rxq[i]); } memset(&args, 0, sizeof(args)); - args.ops = I40E_VIRTCHNL_OP_CONFIG_VSI_QUEUES; + args.ops = VIRTCHNL_OP_CONFIG_VSI_QUEUES; args.in_args = (uint8_t *)vc_vqci; args.in_args_size = size; args.out_buffer = vf->aq_resp; args.out_size = I40E_AQ_BUF_SZ; ret = i40evf_execute_vf_cmd(dev, &args); if (ret) - PMD_DRV_LOG(ERR, "Failed to execute command of " - "I40E_VIRTCHNL_OP_CONFIG_VSI_QUEUES"); + PMD_DRV_LOG(ERR, "Failed to execute command of VIRTCHNL_OP_CONFIG_VSI_QUEUES"); return ret; } @@ -662,8 +661,8 @@ i40evf_configure_vsi_queues_ext(struct rte_eth_dev *dev) (struct i40e_rx_queue **)dev->data->rx_queues; struct i40e_tx_queue **txq = (struct i40e_tx_queue **)dev->data->tx_queues; - struct i40e_virtchnl_vsi_queue_config_ext_info *vc_vqcei; - struct i40e_virtchnl_queue_pair_ext_info *vc_qpei; + struct virtchnl_vsi_queue_config_ext_info *vc_vqcei; + struct virtchnl_queue_pair_ext_info *vc_qpei; struct vf_cmd_info args; uint16_t i, nb_qp = vf->num_queue_pairs; const uint32_t size = @@ -672,7 +671,7 @@ i40evf_configure_vsi_queues_ext(struct rte_eth_dev *dev) int ret; memset(buff, 0, sizeof(buff)); - vc_vqcei = (struct i40e_virtchnl_vsi_queue_config_ext_info *)buff; + vc_vqcei = (struct virtchnl_vsi_queue_config_ext_info *)buff; vc_vqcei->vsi_id = vf->vsi_res->vsi_id; vc_vqcei->num_queue_pairs = nb_qp; vc_qpei = vc_vqcei->qpair; @@ -693,15 +692,14 @@ i40evf_configure_vsi_queues_ext(struct rte_eth_dev *dev) } memset(&args, 0, sizeof(args)); args.ops = - (enum i40e_virtchnl_ops)I40E_VIRTCHNL_OP_CONFIG_VSI_QUEUES_EXT; + (enum virtchnl_ops)VIRTCHNL_OP_CONFIG_VSI_QUEUES_EXT; args.in_args = (uint8_t *)vc_vqcei; args.in_args_size = size; args.out_buffer = vf->aq_resp; args.out_size = I40E_AQ_BUF_SZ; ret = i40evf_execute_vf_cmd(dev, &args); if (ret) - PMD_DRV_LOG(ERR, "Failed to execute command of " - "I40E_VIRTCHNL_OP_CONFIG_VSI_QUEUES_EXT"); + PMD_DRV_LOG(ERR, "Failed to execute command of VIRTCHNL_OP_CONFIG_VSI_QUEUES_EXT"); return ret; } @@ -724,9 +722,9 @@ i40evf_config_irq_map(struct rte_eth_dev *dev) { struct i40e_vf *vf = I40EVF_DEV_PRIVATE_TO_VF(dev->data->dev_private); struct vf_cmd_info args; - uint8_t cmd_buffer[sizeof(struct i40e_virtchnl_irq_map_info) + \ - sizeof(struct i40e_virtchnl_vector_map)]; - struct i40e_virtchnl_irq_map_info *map_info; + uint8_t cmd_buffer[sizeof(struct virtchnl_irq_map_info) + + sizeof(struct virtchnl_vector_map)]; + struct virtchnl_irq_map_info *map_info; struct rte_pci_device *pci_dev = RTE_ETH_DEV_TO_PCI(dev); struct rte_intr_handle *intr_handle = &pci_dev->intr_handle; uint32_t vector_id; @@ -741,7 +739,7 @@ i40evf_config_irq_map(struct rte_eth_dev *dev) vector_id = I40E_MISC_VEC_ID; } - map_info = (struct i40e_virtchnl_irq_map_info *)cmd_buffer; + map_info = (struct virtchnl_irq_map_info *)cmd_buffer; map_info->num_vectors = 1; map_info->vecmap[0].rxitr_idx = I40E_ITR_INDEX_DEFAULT; map_info->vecmap[0].vsi_id = vf->vsi_res->vsi_id; @@ -756,7 +754,7 @@ i40evf_config_irq_map(struct rte_eth_dev *dev) intr_handle->intr_vec[i] = vector_id; } - args.ops = I40E_VIRTCHNL_OP_CONFIG_IRQ_MAP; + args.ops = VIRTCHNL_OP_CONFIG_IRQ_MAP; args.in_args = (u8 *)cmd_buffer; args.in_args_size = sizeof(cmd_buffer); args.out_buffer = vf->aq_resp; @@ -773,7 +771,7 @@ i40evf_switch_queue(struct rte_eth_dev *dev, bool isrx, uint16_t qid, bool on) { struct i40e_vf *vf = I40EVF_DEV_PRIVATE_TO_VF(dev->data->dev_private); - struct i40e_virtchnl_queue_select queue_select; + struct virtchnl_queue_select queue_select; int err; struct vf_cmd_info args; memset(&queue_select, 0, sizeof(queue_select)); @@ -785,9 +783,9 @@ i40evf_switch_queue(struct rte_eth_dev *dev, bool isrx, uint16_t qid, queue_select.tx_queues |= 1 << qid; if (on) - args.ops = I40E_VIRTCHNL_OP_ENABLE_QUEUES; + args.ops = VIRTCHNL_OP_ENABLE_QUEUES; else - args.ops = I40E_VIRTCHNL_OP_DISABLE_QUEUES; + args.ops = VIRTCHNL_OP_DISABLE_QUEUES; args.in_args = (u8 *)&queue_select; args.in_args_size = sizeof(queue_select); args.out_buffer = vf->aq_resp; @@ -861,10 +859,10 @@ i40evf_add_mac_addr(struct rte_eth_dev *dev, __rte_unused uint32_t index, __rte_unused uint32_t pool) { - struct i40e_virtchnl_ether_addr_list *list; + struct virtchnl_ether_addr_list *list; struct i40e_vf *vf = I40EVF_DEV_PRIVATE_TO_VF(dev->data->dev_private); - uint8_t cmd_buffer[sizeof(struct i40e_virtchnl_ether_addr_list) + \ - sizeof(struct i40e_virtchnl_ether_addr)]; + uint8_t cmd_buffer[sizeof(struct virtchnl_ether_addr_list) + + sizeof(struct virtchnl_ether_addr)]; int err; struct vf_cmd_info args; @@ -876,13 +874,13 @@ i40evf_add_mac_addr(struct rte_eth_dev *dev, return I40E_ERR_INVALID_MAC_ADDR; } - list = (struct i40e_virtchnl_ether_addr_list *)cmd_buffer; + list = (struct virtchnl_ether_addr_list *)cmd_buffer; list->vsi_id = vf->vsi_res->vsi_id; list->num_elements = 1; (void)rte_memcpy(list->list[0].addr, addr->addr_bytes, sizeof(addr->addr_bytes)); - args.ops = I40E_VIRTCHNL_OP_ADD_ETHER_ADDRESS; + args.ops = VIRTCHNL_OP_ADD_ETH_ADDR; args.in_args = cmd_buffer; args.in_args_size = sizeof(cmd_buffer); args.out_buffer = vf->aq_resp; @@ -899,10 +897,10 @@ static void i40evf_del_mac_addr_by_addr(struct rte_eth_dev *dev, struct ether_addr *addr) { - struct i40e_virtchnl_ether_addr_list *list; + struct virtchnl_ether_addr_list *list; struct i40e_vf *vf = I40EVF_DEV_PRIVATE_TO_VF(dev->data->dev_private); - uint8_t cmd_buffer[sizeof(struct i40e_virtchnl_ether_addr_list) + \ - sizeof(struct i40e_virtchnl_ether_addr)]; + uint8_t cmd_buffer[sizeof(struct virtchnl_ether_addr_list) + + sizeof(struct virtchnl_ether_addr)]; int err; struct vf_cmd_info args; @@ -914,13 +912,13 @@ i40evf_del_mac_addr_by_addr(struct rte_eth_dev *dev, return; } - list = (struct i40e_virtchnl_ether_addr_list *)cmd_buffer; + list = (struct virtchnl_ether_addr_list *)cmd_buffer; list->vsi_id = vf->vsi_res->vsi_id; list->num_elements = 1; (void)rte_memcpy(list->list[0].addr, addr->addr_bytes, sizeof(addr->addr_bytes)); - args.ops = I40E_VIRTCHNL_OP_DEL_ETHER_ADDRESS; + args.ops = VIRTCHNL_OP_DEL_ETH_ADDR; args.in_args = cmd_buffer; args.in_args_size = sizeof(cmd_buffer); args.out_buffer = vf->aq_resp; @@ -947,13 +945,13 @@ static int i40evf_update_stats(struct rte_eth_dev *dev, struct i40e_eth_stats **pstats) { struct i40e_vf *vf = I40EVF_DEV_PRIVATE_TO_VF(dev->data->dev_private); - struct i40e_virtchnl_queue_select q_stats; + struct virtchnl_queue_select q_stats; int err; struct vf_cmd_info args; memset(&q_stats, 0, sizeof(q_stats)); q_stats.vsi_id = vf->vsi_res->vsi_id; - args.ops = I40E_VIRTCHNL_OP_GET_STATS; + args.ops = VIRTCHNL_OP_GET_STATS; args.in_args = (u8 *)&q_stats; args.in_args_size = sizeof(q_stats); args.out_buffer = vf->aq_resp; @@ -1050,18 +1048,18 @@ static int i40evf_add_vlan(struct rte_eth_dev *dev, uint16_t vlanid) { struct i40e_vf *vf = I40EVF_DEV_PRIVATE_TO_VF(dev->data->dev_private); - struct i40e_virtchnl_vlan_filter_list *vlan_list; - uint8_t cmd_buffer[sizeof(struct i40e_virtchnl_vlan_filter_list) + + struct virtchnl_vlan_filter_list *vlan_list; + uint8_t cmd_buffer[sizeof(struct virtchnl_vlan_filter_list) + sizeof(uint16_t)]; int err; struct vf_cmd_info args; - vlan_list = (struct i40e_virtchnl_vlan_filter_list *)cmd_buffer; + vlan_list = (struct virtchnl_vlan_filter_list *)cmd_buffer; vlan_list->vsi_id = vf->vsi_res->vsi_id; vlan_list->num_elements = 1; vlan_list->vlan_id[0] = vlanid; - args.ops = I40E_VIRTCHNL_OP_ADD_VLAN; + args.ops = VIRTCHNL_OP_ADD_VLAN; args.in_args = (u8 *)&cmd_buffer; args.in_args_size = sizeof(cmd_buffer); args.out_buffer = vf->aq_resp; @@ -1077,18 +1075,18 @@ static int i40evf_del_vlan(struct rte_eth_dev *dev, uint16_t vlanid) { struct i40e_vf *vf = I40EVF_DEV_PRIVATE_TO_VF(dev->data->dev_private); - struct i40e_virtchnl_vlan_filter_list *vlan_list; - uint8_t cmd_buffer[sizeof(struct i40e_virtchnl_vlan_filter_list) + + struct virtchnl_vlan_filter_list *vlan_list; + uint8_t cmd_buffer[sizeof(struct virtchnl_vlan_filter_list) + sizeof(uint16_t)]; int err; struct vf_cmd_info args; - vlan_list = (struct i40e_virtchnl_vlan_filter_list *)cmd_buffer; + vlan_list = (struct virtchnl_vlan_filter_list *)cmd_buffer; vlan_list->vsi_id = vf->vsi_res->vsi_id; vlan_list->num_elements = 1; vlan_list->vlan_id[0] = vlanid; - args.ops = I40E_VIRTCHNL_OP_DEL_VLAN; + args.ops = VIRTCHNL_OP_DEL_VLAN; args.in_args = (u8 *)&cmd_buffer; args.in_args_size = sizeof(cmd_buffer); args.out_buffer = vf->aq_resp; @@ -1178,7 +1176,8 @@ i40evf_reset_vf(struct i40e_hw *hw) reset = rd32(hw, I40E_VFGEN_RSTAT) & I40E_VFGEN_RSTAT_VFR_STATE_MASK; reset = reset >> I40E_VFGEN_RSTAT_VFR_STATE_SHIFT; - if (I40E_VFR_COMPLETED == reset || I40E_VFR_VFACTIVE == reset) + if (VIRTCHNL_VFR_COMPLETED == reset || + VIRTCHNL_VFR_VFACTIVE == reset) break; else rte_delay_ms(50); @@ -1242,8 +1241,8 @@ i40evf_init_vf(struct rte_eth_dev *dev) PMD_INIT_LOG(ERR, "check_api version failed"); goto err_aq; } - bufsz = sizeof(struct i40e_virtchnl_vf_resource) + - (I40E_MAX_VF_VSI * sizeof(struct i40e_virtchnl_vsi_resource)); + bufsz = sizeof(struct virtchnl_vf_resource) + + (I40E_MAX_VF_VSI * sizeof(struct virtchnl_vsi_resource)); vf->vf_res = rte_zmalloc("vf_res", bufsz, 0); if (!vf->vf_res) { PMD_INIT_LOG(ERR, "unable to allocate vf_res memory"); @@ -1257,7 +1256,7 @@ i40evf_init_vf(struct rte_eth_dev *dev) /* got VF config message back from PF, now we can parse it */ for (i = 0; i < vf->vf_res->num_vsis; i++) { - if (vf->vf_res->vsi_res[i].vsi_type == I40E_VSI_SRIOV) + if (vf->vf_res->vsi_res[i].vsi_type == VIRTCHNL_VSI_SRIOV) vf->vsi_res = &vf->vf_res->vsi_res[i]; } @@ -1269,7 +1268,7 @@ i40evf_init_vf(struct rte_eth_dev *dev) if (hw->mac.type == I40E_MAC_X722_VF) vf->flags = I40E_FLAG_RSS_AQ_CAPABLE; vf->vsi.vsi_id = vf->vsi_res->vsi_id; - vf->vsi.type = vf->vsi_res->vsi_type; + vf->vsi.type = (enum i40e_vsi_type)vf->vsi_res->vsi_type; vf->vsi.nb_qps = vf->vsi_res->num_queue_pairs; vf->vsi.adapter = I40E_DEV_PRIVATE_TO_ADAPTER(dev->data->dev_private); @@ -1321,21 +1320,21 @@ static void i40evf_handle_pf_event(struct rte_eth_dev *dev, uint8_t *msg, __rte_unused uint16_t msglen) { - struct i40e_virtchnl_pf_event *pf_msg = - (struct i40e_virtchnl_pf_event *)msg; + struct virtchnl_pf_event *pf_msg = + (struct virtchnl_pf_event *)msg; struct i40e_vf *vf = I40EVF_DEV_PRIVATE_TO_VF(dev->data->dev_private); switch (pf_msg->event) { - case I40E_VIRTCHNL_EVENT_RESET_IMPENDING: + case VIRTCHNL_EVENT_RESET_IMPENDING: PMD_DRV_LOG(DEBUG, "VIRTCHNL_EVENT_RESET_IMPENDING event"); _rte_eth_dev_callback_process(dev, RTE_ETH_EVENT_INTR_RESET, NULL); break; - case I40E_VIRTCHNL_EVENT_LINK_CHANGE: + case VIRTCHNL_EVENT_LINK_CHANGE: PMD_DRV_LOG(DEBUG, "VIRTCHNL_EVENT_LINK_CHANGE event"); vf->link_up = pf_msg->event_data.link_event.link_status; vf->link_speed = pf_msg->event_data.link_event.link_speed; break; - case I40E_VIRTCHNL_EVENT_PF_DRIVER_CLOSE: + case VIRTCHNL_EVENT_PF_DRIVER_CLOSE: PMD_DRV_LOG(DEBUG, "VIRTCHNL_EVENT_PF_DRIVER_CLOSE event"); break; default: @@ -1351,7 +1350,7 @@ i40evf_handle_aq_msg(struct rte_eth_dev *dev) struct i40e_vf *vf = I40EVF_DEV_PRIVATE_TO_VF(dev->data->dev_private); struct i40e_arq_event_info info; uint16_t pending, aq_opc; - enum i40e_virtchnl_ops msg_opc; + enum virtchnl_ops msg_opc; enum i40e_status_code msg_ret; int ret; @@ -1376,13 +1375,13 @@ i40evf_handle_aq_msg(struct rte_eth_dev *dev) * cookie_high of struct i40e_aq_desc, while return error code * are stored in cookie_low, Which is done by * i40e_aq_send_msg_to_vf in PF driver.*/ - msg_opc = (enum i40e_virtchnl_ops)rte_le_to_cpu_32( + msg_opc = (enum virtchnl_ops)rte_le_to_cpu_32( info.desc.cookie_high); msg_ret = (enum i40e_status_code)rte_le_to_cpu_32( info.desc.cookie_low); switch (aq_opc) { case i40e_aqc_opc_send_msg_to_vf: - if (msg_opc == I40E_VIRTCHNL_OP_EVENT) + if (msg_opc == VIRTCHNL_OP_EVENT) /* process event*/ i40evf_handle_pf_event(dev, info.msg_buf, info.msg_len); @@ -1591,8 +1590,8 @@ i40evf_dev_configure(struct rte_eth_dev *dev) */ if (!conf->rxmode.hw_strip_crc) { vf = I40EVF_DEV_PRIVATE_TO_VF(dev->data->dev_private); - if ((vf->version_major == I40E_VIRTCHNL_VERSION_MAJOR) && - (vf->version_minor <= I40E_VIRTCHNL_VERSION_MINOR)) { + if ((vf->version_major == VIRTCHNL_VERSION_MAJOR) && + (vf->version_minor <= VIRTCHNL_VERSION_MINOR)) { /* Peer is running non-DPDK PF driver. */ PMD_INIT_LOG(ERR, "VF can't disable HW CRC Strip"); return -EINVAL; @@ -2000,7 +1999,7 @@ i40evf_dev_rx_queue_intr_disable(struct rte_eth_dev *dev, uint16_t queue_id) static void i40evf_add_del_all_mac_addr(struct rte_eth_dev *dev, bool add) { - struct i40e_virtchnl_ether_addr_list *list; + struct virtchnl_ether_addr_list *list; struct i40e_vf *vf = I40EVF_DEV_PRIVATE_TO_VF(dev->data->dev_private); int err, i, j; int next_begin = 0; @@ -2011,11 +2010,11 @@ i40evf_add_del_all_mac_addr(struct rte_eth_dev *dev, bool add) do { j = 0; - len = sizeof(struct i40e_virtchnl_ether_addr_list); + len = sizeof(struct virtchnl_ether_addr_list); for (i = begin; i < I40E_NUM_MACADDR_MAX; i++, next_begin++) { if (is_zero_ether_addr(&dev->data->mac_addrs[i])) continue; - len += sizeof(struct i40e_virtchnl_ether_addr); + len += sizeof(struct virtchnl_ether_addr); if (len >= I40E_AQ_BUF_SZ) { next_begin = i + 1; break; @@ -2042,8 +2041,8 @@ i40evf_add_del_all_mac_addr(struct rte_eth_dev *dev, bool add) } list->vsi_id = vf->vsi_res->vsi_id; list->num_elements = j; - args.ops = add ? I40E_VIRTCHNL_OP_ADD_ETHER_ADDRESS : - I40E_VIRTCHNL_OP_DEL_ETHER_ADDRESS; + args.ops = add ? VIRTCHNL_OP_ADD_ETH_ADDR : + VIRTCHNL_OP_DEL_ETH_ADDR; args.in_args = (uint8_t *)list; args.in_args_size = len; args.out_buffer = vf->aq_resp; diff --git a/drivers/net/i40e/i40e_pf.c b/drivers/net/i40e/i40e_pf.c index 0758503..22ad483 100644 --- a/drivers/net/i40e/i40e_pf.c +++ b/drivers/net/i40e/i40e_pf.c @@ -61,7 +61,7 @@ static int i40e_pf_host_switch_queues(struct i40e_pf_vf *vf, - struct i40e_virtchnl_queue_select *qsel, + struct virtchnl_queue_select *qsel, bool on); /** @@ -128,7 +128,7 @@ i40e_pf_host_vf_reset(struct i40e_pf_vf *vf, bool do_hw_reset) struct i40e_pf *pf; uint16_t vf_id, abs_vf_id, vf_msix_num; int ret; - struct i40e_virtchnl_queue_select qsel; + struct virtchnl_queue_select qsel; if (vf == NULL) return -EINVAL; @@ -139,7 +139,7 @@ i40e_pf_host_vf_reset(struct i40e_pf_vf *vf, bool do_hw_reset) abs_vf_id = vf_id + hw->func_caps.vf_base_id; /* Notify VF that we are in VFR progress */ - I40E_WRITE_REG(hw, I40E_VFGEN_RSTAT1(vf_id), I40E_VFR_INPROGRESS); + I40E_WRITE_REG(hw, I40E_VFGEN_RSTAT1(vf_id), VIRTCHNL_VFR_INPROGRESS); /* * If require a SW VF reset, a VFLR interrupt will be generated, @@ -220,7 +220,7 @@ i40e_pf_host_vf_reset(struct i40e_pf_vf *vf, bool do_hw_reset) } /* Reset done, Set COMPLETE flag and clear reset bit */ - I40E_WRITE_REG(hw, I40E_VFGEN_RSTAT1(vf_id), I40E_VFR_COMPLETED); + I40E_WRITE_REG(hw, I40E_VFGEN_RSTAT1(vf_id), VIRTCHNL_VFR_COMPLETED); val = I40E_READ_REG(hw, I40E_VPGEN_VFRTRIG(vf_id)); val &= ~I40E_VPGEN_VFRTRIG_VFSWR_MASK; I40E_WRITE_REG(hw, I40E_VPGEN_VFRTRIG(vf_id), val); @@ -248,7 +248,7 @@ i40e_pf_host_vf_reset(struct i40e_pf_vf *vf, bool do_hw_reset) return -EFAULT; } - I40E_WRITE_REG(hw, I40E_VFGEN_RSTAT1(vf_id), I40E_VFR_VFACTIVE); + I40E_WRITE_REG(hw, I40E_VFGEN_RSTAT1(vf_id), VIRTCHNL_VFR_VFACTIVE); return ret; } @@ -277,7 +277,7 @@ i40e_pf_host_send_msg_to_vf(struct i40e_pf_vf *vf, static void i40e_pf_host_process_cmd_version(struct i40e_pf_vf *vf, bool b_op) { - struct i40e_virtchnl_version_info info; + struct virtchnl_version_info info; /* Respond like a Linux PF host in order to support both DPDK VF and * Linux VF driver. The expense is original DPDK host specific feature @@ -286,16 +286,16 @@ i40e_pf_host_process_cmd_version(struct i40e_pf_vf *vf, bool b_op) * DPDK VF also can't identify host driver by version number returned. * It always assume talking with Linux PF. */ - info.major = I40E_VIRTCHNL_VERSION_MAJOR; - info.minor = I40E_VIRTCHNL_VERSION_MINOR_NO_VF_CAPS; + info.major = VIRTCHNL_VERSION_MAJOR; + info.minor = VIRTCHNL_VERSION_MINOR_NO_VF_CAPS; if (b_op) - i40e_pf_host_send_msg_to_vf(vf, I40E_VIRTCHNL_OP_VERSION, + i40e_pf_host_send_msg_to_vf(vf, VIRTCHNL_OP_VERSION, I40E_SUCCESS, (uint8_t *)&info, sizeof(info)); else - i40e_pf_host_send_msg_to_vf(vf, I40E_VIRTCHNL_OP_VERSION, + i40e_pf_host_send_msg_to_vf(vf, VIRTCHNL_OP_VERSION, I40E_NOT_SUPPORTED, (uint8_t *)&info, sizeof(info)); @@ -313,22 +313,22 @@ i40e_pf_host_process_cmd_reset_vf(struct i40e_pf_vf *vf) static int i40e_pf_host_process_cmd_get_vf_resource(struct i40e_pf_vf *vf, bool b_op) { - struct i40e_virtchnl_vf_resource *vf_res = NULL; + struct virtchnl_vf_resource *vf_res = NULL; struct i40e_hw *hw = I40E_PF_TO_HW(vf->pf); uint32_t len = 0; int ret = I40E_SUCCESS; if (!b_op) { i40e_pf_host_send_msg_to_vf(vf, - I40E_VIRTCHNL_OP_GET_VF_RESOURCES, + VIRTCHNL_OP_GET_VF_RESOURCES, I40E_NOT_SUPPORTED, NULL, 0); return ret; } /* only have 1 VSI by default */ - len = sizeof(struct i40e_virtchnl_vf_resource) + + len = sizeof(struct virtchnl_vf_resource) + I40E_DEFAULT_VF_VSI_NUM * - sizeof(struct i40e_virtchnl_vsi_resource); + sizeof(struct virtchnl_vsi_resource); vf_res = rte_zmalloc("i40e_vf_res", len, 0); if (vf_res == NULL) { @@ -339,22 +339,22 @@ i40e_pf_host_process_cmd_get_vf_resource(struct i40e_pf_vf *vf, bool b_op) goto send_msg; } - vf_res->vf_offload_flags = I40E_VIRTCHNL_VF_OFFLOAD_L2 | - I40E_VIRTCHNL_VF_OFFLOAD_VLAN; + vf_res->vf_offload_flags = VIRTCHNL_VF_OFFLOAD_L2 | + VIRTCHNL_VF_OFFLOAD_VLAN; vf_res->max_vectors = hw->func_caps.num_msix_vectors_vf; vf_res->num_queue_pairs = vf->vsi->nb_qps; vf_res->num_vsis = I40E_DEFAULT_VF_VSI_NUM; /* Change below setting if PF host can support more VSIs for VF */ - vf_res->vsi_res[0].vsi_type = I40E_VSI_SRIOV; + vf_res->vsi_res[0].vsi_type = VIRTCHNL_VSI_SRIOV; vf_res->vsi_res[0].vsi_id = vf->vsi->vsi_id; vf_res->vsi_res[0].num_queue_pairs = vf->vsi->nb_qps; ether_addr_copy(&vf->mac_addr, (struct ether_addr *)vf_res->vsi_res[0].default_mac_addr); send_msg: - i40e_pf_host_send_msg_to_vf(vf, I40E_VIRTCHNL_OP_GET_VF_RESOURCES, - ret, (uint8_t *)vf_res, len); + i40e_pf_host_send_msg_to_vf(vf, VIRTCHNL_OP_GET_VF_RESOURCES, + ret, (uint8_t *)vf_res, len); rte_free(vf_res); return ret; @@ -363,7 +363,7 @@ i40e_pf_host_process_cmd_get_vf_resource(struct i40e_pf_vf *vf, bool b_op) static int i40e_pf_host_hmc_config_rxq(struct i40e_hw *hw, struct i40e_pf_vf *vf, - struct i40e_virtchnl_rxq_info *rxq, + struct virtchnl_rxq_info *rxq, uint8_t crcstrip) { int err = I40E_SUCCESS; @@ -431,7 +431,7 @@ i40e_vsi_get_tc_of_queue(struct i40e_vsi *vsi, static int i40e_pf_host_hmc_config_txq(struct i40e_hw *hw, struct i40e_pf_vf *vf, - struct i40e_virtchnl_txq_info *txq) + struct virtchnl_txq_info *txq) { int err = I40E_SUCCESS; struct i40e_hmc_obj_txq tx_ctx; @@ -480,14 +480,14 @@ i40e_pf_host_process_cmd_config_vsi_queues(struct i40e_pf_vf *vf, { struct i40e_hw *hw = I40E_PF_TO_HW(vf->pf); struct i40e_vsi *vsi = vf->vsi; - struct i40e_virtchnl_vsi_queue_config_info *vc_vqci = - (struct i40e_virtchnl_vsi_queue_config_info *)msg; - struct i40e_virtchnl_queue_pair_info *vc_qpi; + struct virtchnl_vsi_queue_config_info *vc_vqci = + (struct virtchnl_vsi_queue_config_info *)msg; + struct virtchnl_queue_pair_info *vc_qpi; int i, ret = I40E_SUCCESS; if (!b_op) { i40e_pf_host_send_msg_to_vf(vf, - I40E_VIRTCHNL_OP_CONFIG_VSI_QUEUES, + VIRTCHNL_OP_CONFIG_VSI_QUEUES, I40E_NOT_SUPPORTED, NULL, 0); return ret; } @@ -511,9 +511,9 @@ i40e_pf_host_process_cmd_config_vsi_queues(struct i40e_pf_vf *vf, /* * Apply VF RX queue setting to HMC. - * If the opcode is I40E_VIRTCHNL_OP_CONFIG_VSI_QUEUES_EXT, + * If the opcode is VIRTCHNL_OP_CONFIG_VSI_QUEUES_EXT, * then the extra information of - * 'struct i40e_virtchnl_queue_pair_extra_info' is needed, + * 'struct virtchnl_queue_pair_extra_info' is needed, * otherwise set the last parameter to NULL. */ if (i40e_pf_host_hmc_config_rxq(hw, vf, &vc_qpi[i].rxq, @@ -533,8 +533,8 @@ i40e_pf_host_process_cmd_config_vsi_queues(struct i40e_pf_vf *vf, } send_msg: - i40e_pf_host_send_msg_to_vf(vf, I40E_VIRTCHNL_OP_CONFIG_VSI_QUEUES, - ret, NULL, 0); + i40e_pf_host_send_msg_to_vf(vf, VIRTCHNL_OP_CONFIG_VSI_QUEUES, + ret, NULL, 0); return ret; } @@ -547,15 +547,15 @@ i40e_pf_host_process_cmd_config_vsi_queues_ext(struct i40e_pf_vf *vf, { struct i40e_hw *hw = I40E_PF_TO_HW(vf->pf); struct i40e_vsi *vsi = vf->vsi; - struct i40e_virtchnl_vsi_queue_config_ext_info *vc_vqcei = - (struct i40e_virtchnl_vsi_queue_config_ext_info *)msg; - struct i40e_virtchnl_queue_pair_ext_info *vc_qpei; + struct virtchnl_vsi_queue_config_ext_info *vc_vqcei = + (struct virtchnl_vsi_queue_config_ext_info *)msg; + struct virtchnl_queue_pair_ext_info *vc_qpei; int i, ret = I40E_SUCCESS; if (!b_op) { i40e_pf_host_send_msg_to_vf( vf, - I40E_VIRTCHNL_OP_CONFIG_VSI_QUEUES_EXT, + VIRTCHNL_OP_CONFIG_VSI_QUEUES_EXT, I40E_NOT_SUPPORTED, NULL, 0); return ret; } @@ -578,9 +578,9 @@ i40e_pf_host_process_cmd_config_vsi_queues_ext(struct i40e_pf_vf *vf, } /* * Apply VF RX queue setting to HMC. - * If the opcode is I40E_VIRTCHNL_OP_CONFIG_VSI_QUEUES_EXT, + * If the opcode is VIRTCHNL_OP_CONFIG_VSI_QUEUES_EXT, * then the extra information of - * 'struct i40e_virtchnl_queue_pair_ext_info' is needed, + * 'struct virtchnl_queue_pair_ext_info' is needed, * otherwise set the last parameter to NULL. */ if (i40e_pf_host_hmc_config_rxq(hw, vf, &vc_qpei[i].rxq, @@ -600,15 +600,15 @@ i40e_pf_host_process_cmd_config_vsi_queues_ext(struct i40e_pf_vf *vf, } send_msg: - i40e_pf_host_send_msg_to_vf(vf, I40E_VIRTCHNL_OP_CONFIG_VSI_QUEUES_EXT, - ret, NULL, 0); + i40e_pf_host_send_msg_to_vf(vf, VIRTCHNL_OP_CONFIG_VSI_QUEUES_EXT, + ret, NULL, 0); return ret; } static void i40e_pf_config_irq_link_list(struct i40e_pf_vf *vf, - struct i40e_virtchnl_vector_map *vvm) + struct virtchnl_vector_map *vvm) { #define BITS_PER_CHAR 8 uint64_t linklistmap = 0, tempmap; @@ -711,9 +711,9 @@ i40e_pf_host_process_cmd_config_irq_map(struct i40e_pf_vf *vf, int ret = I40E_SUCCESS; struct i40e_pf *pf = vf->pf; struct i40e_hw *hw = I40E_PF_TO_HW(vf->pf); - struct i40e_virtchnl_irq_map_info *irqmap = - (struct i40e_virtchnl_irq_map_info *)msg; - struct i40e_virtchnl_vector_map *map; + struct virtchnl_irq_map_info *irqmap = + (struct virtchnl_irq_map_info *)msg; + struct virtchnl_vector_map *map; int i; uint16_t vector_id; unsigned long qbit_max; @@ -721,12 +721,12 @@ i40e_pf_host_process_cmd_config_irq_map(struct i40e_pf_vf *vf, if (!b_op) { i40e_pf_host_send_msg_to_vf( vf, - I40E_VIRTCHNL_OP_CONFIG_IRQ_MAP, + VIRTCHNL_OP_CONFIG_IRQ_MAP, I40E_NOT_SUPPORTED, NULL, 0); return ret; } - if (msg == NULL || msglen < sizeof(struct i40e_virtchnl_irq_map_info)) { + if (!msg || msglen < sizeof(struct virtchnl_irq_map_info)) { PMD_DRV_LOG(ERR, "buffer too short"); ret = I40E_ERR_PARAM; goto send_msg; @@ -773,15 +773,15 @@ i40e_pf_host_process_cmd_config_irq_map(struct i40e_pf_vf *vf, } send_msg: - i40e_pf_host_send_msg_to_vf(vf, I40E_VIRTCHNL_OP_CONFIG_IRQ_MAP, - ret, NULL, 0); + i40e_pf_host_send_msg_to_vf(vf, VIRTCHNL_OP_CONFIG_IRQ_MAP, + ret, NULL, 0); return ret; } static int i40e_pf_host_switch_queues(struct i40e_pf_vf *vf, - struct i40e_virtchnl_queue_select *qsel, + struct virtchnl_queue_select *qsel, bool on) { int ret = I40E_SUCCESS; @@ -831,8 +831,8 @@ i40e_pf_host_process_cmd_enable_queues(struct i40e_pf_vf *vf, uint16_t msglen) { int ret = I40E_SUCCESS; - struct i40e_virtchnl_queue_select *q_sel = - (struct i40e_virtchnl_queue_select *)msg; + struct virtchnl_queue_select *q_sel = + (struct virtchnl_queue_select *)msg; if (msg == NULL || msglen != sizeof(*q_sel)) { ret = I40E_ERR_PARAM; @@ -841,8 +841,8 @@ i40e_pf_host_process_cmd_enable_queues(struct i40e_pf_vf *vf, ret = i40e_pf_host_switch_queues(vf, q_sel, true); send_msg: - i40e_pf_host_send_msg_to_vf(vf, I40E_VIRTCHNL_OP_ENABLE_QUEUES, - ret, NULL, 0); + i40e_pf_host_send_msg_to_vf(vf, VIRTCHNL_OP_ENABLE_QUEUES, + ret, NULL, 0); return ret; } @@ -854,13 +854,13 @@ i40e_pf_host_process_cmd_disable_queues(struct i40e_pf_vf *vf, bool b_op) { int ret = I40E_SUCCESS; - struct i40e_virtchnl_queue_select *q_sel = - (struct i40e_virtchnl_queue_select *)msg; + struct virtchnl_queue_select *q_sel = + (struct virtchnl_queue_select *)msg; if (!b_op) { i40e_pf_host_send_msg_to_vf( vf, - I40E_VIRTCHNL_OP_DISABLE_QUEUES, + VIRTCHNL_OP_DISABLE_QUEUES, I40E_NOT_SUPPORTED, NULL, 0); return ret; } @@ -872,8 +872,8 @@ i40e_pf_host_process_cmd_disable_queues(struct i40e_pf_vf *vf, ret = i40e_pf_host_switch_queues(vf, q_sel, false); send_msg: - i40e_pf_host_send_msg_to_vf(vf, I40E_VIRTCHNL_OP_DISABLE_QUEUES, - ret, NULL, 0); + i40e_pf_host_send_msg_to_vf(vf, VIRTCHNL_OP_DISABLE_QUEUES, + ret, NULL, 0); return ret; } @@ -886,8 +886,8 @@ i40e_pf_host_process_cmd_add_ether_address(struct i40e_pf_vf *vf, bool b_op) { int ret = I40E_SUCCESS; - struct i40e_virtchnl_ether_addr_list *addr_list = - (struct i40e_virtchnl_ether_addr_list *)msg; + struct virtchnl_ether_addr_list *addr_list = + (struct virtchnl_ether_addr_list *)msg; struct i40e_mac_filter_info filter; int i; struct ether_addr *mac; @@ -895,7 +895,7 @@ i40e_pf_host_process_cmd_add_ether_address(struct i40e_pf_vf *vf, if (!b_op) { i40e_pf_host_send_msg_to_vf( vf, - I40E_VIRTCHNL_OP_ADD_ETHER_ADDRESS, + VIRTCHNL_OP_ADD_ETH_ADDR, I40E_NOT_SUPPORTED, NULL, 0); return ret; } @@ -920,8 +920,8 @@ i40e_pf_host_process_cmd_add_ether_address(struct i40e_pf_vf *vf, } send_msg: - i40e_pf_host_send_msg_to_vf(vf, I40E_VIRTCHNL_OP_ADD_ETHER_ADDRESS, - ret, NULL, 0); + i40e_pf_host_send_msg_to_vf(vf, VIRTCHNL_OP_ADD_ETH_ADDR, + ret, NULL, 0); return ret; } @@ -933,15 +933,15 @@ i40e_pf_host_process_cmd_del_ether_address(struct i40e_pf_vf *vf, bool b_op) { int ret = I40E_SUCCESS; - struct i40e_virtchnl_ether_addr_list *addr_list = - (struct i40e_virtchnl_ether_addr_list *)msg; + struct virtchnl_ether_addr_list *addr_list = + (struct virtchnl_ether_addr_list *)msg; int i; struct ether_addr *mac; if (!b_op) { i40e_pf_host_send_msg_to_vf( vf, - I40E_VIRTCHNL_OP_DEL_ETHER_ADDRESS, + VIRTCHNL_OP_DEL_ETH_ADDR, I40E_NOT_SUPPORTED, NULL, 0); return ret; } @@ -962,8 +962,8 @@ i40e_pf_host_process_cmd_del_ether_address(struct i40e_pf_vf *vf, } send_msg: - i40e_pf_host_send_msg_to_vf(vf, I40E_VIRTCHNL_OP_DEL_ETHER_ADDRESS, - ret, NULL, 0); + i40e_pf_host_send_msg_to_vf(vf, VIRTCHNL_OP_DEL_ETH_ADDR, + ret, NULL, 0); return ret; } @@ -974,15 +974,15 @@ i40e_pf_host_process_cmd_add_vlan(struct i40e_pf_vf *vf, bool b_op) { int ret = I40E_SUCCESS; - struct i40e_virtchnl_vlan_filter_list *vlan_filter_list = - (struct i40e_virtchnl_vlan_filter_list *)msg; + struct virtchnl_vlan_filter_list *vlan_filter_list = + (struct virtchnl_vlan_filter_list *)msg; int i; uint16_t *vid; if (!b_op) { i40e_pf_host_send_msg_to_vf( vf, - I40E_VIRTCHNL_OP_ADD_VLAN, + VIRTCHNL_OP_ADD_VLAN, I40E_NOT_SUPPORTED, NULL, 0); return ret; } @@ -1002,8 +1002,8 @@ i40e_pf_host_process_cmd_add_vlan(struct i40e_pf_vf *vf, } send_msg: - i40e_pf_host_send_msg_to_vf(vf, I40E_VIRTCHNL_OP_ADD_VLAN, - ret, NULL, 0); + i40e_pf_host_send_msg_to_vf(vf, VIRTCHNL_OP_ADD_VLAN, + ret, NULL, 0); return ret; } @@ -1015,15 +1015,15 @@ i40e_pf_host_process_cmd_del_vlan(struct i40e_pf_vf *vf, bool b_op) { int ret = I40E_SUCCESS; - struct i40e_virtchnl_vlan_filter_list *vlan_filter_list = - (struct i40e_virtchnl_vlan_filter_list *)msg; + struct virtchnl_vlan_filter_list *vlan_filter_list = + (struct virtchnl_vlan_filter_list *)msg; int i; uint16_t *vid; if (!b_op) { i40e_pf_host_send_msg_to_vf( vf, - I40E_VIRTCHNL_OP_DEL_VLAN, + VIRTCHNL_OP_DEL_VLAN, I40E_NOT_SUPPORTED, NULL, 0); return ret; } @@ -1042,8 +1042,8 @@ i40e_pf_host_process_cmd_del_vlan(struct i40e_pf_vf *vf, } send_msg: - i40e_pf_host_send_msg_to_vf(vf, I40E_VIRTCHNL_OP_DEL_VLAN, - ret, NULL, 0); + i40e_pf_host_send_msg_to_vf(vf, VIRTCHNL_OP_DEL_VLAN, + ret, NULL, 0); return ret; } @@ -1056,15 +1056,15 @@ i40e_pf_host_process_cmd_config_promisc_mode( bool b_op) { int ret = I40E_SUCCESS; - struct i40e_virtchnl_promisc_info *promisc = - (struct i40e_virtchnl_promisc_info *)msg; + struct virtchnl_promisc_info *promisc = + (struct virtchnl_promisc_info *)msg; struct i40e_hw *hw = I40E_PF_TO_HW(vf->pf); bool unicast = FALSE, multicast = FALSE; if (!b_op) { i40e_pf_host_send_msg_to_vf( vf, - I40E_VIRTCHNL_OP_CONFIG_PROMISCUOUS_MODE, + VIRTCHNL_OP_CONFIG_PROMISCUOUS_MODE, I40E_NOT_SUPPORTED, NULL, 0); return ret; } @@ -1074,21 +1074,21 @@ i40e_pf_host_process_cmd_config_promisc_mode( goto send_msg; } - if (promisc->flags & I40E_FLAG_VF_UNICAST_PROMISC) + if (promisc->flags & FLAG_VF_UNICAST_PROMISC) unicast = TRUE; ret = i40e_aq_set_vsi_unicast_promiscuous(hw, vf->vsi->seid, unicast, NULL, true); if (ret != I40E_SUCCESS) goto send_msg; - if (promisc->flags & I40E_FLAG_VF_MULTICAST_PROMISC) + if (promisc->flags & FLAG_VF_MULTICAST_PROMISC) multicast = TRUE; ret = i40e_aq_set_vsi_multicast_promiscuous(hw, vf->vsi->seid, multicast, NULL); send_msg: i40e_pf_host_send_msg_to_vf(vf, - I40E_VIRTCHNL_OP_CONFIG_PROMISCUOUS_MODE, ret, NULL, 0); + VIRTCHNL_OP_CONFIG_PROMISCUOUS_MODE, ret, NULL, 0); return ret; } @@ -1099,12 +1099,12 @@ i40e_pf_host_process_cmd_get_stats(struct i40e_pf_vf *vf, bool b_op) i40e_update_vsi_stats(vf->vsi); if (b_op) - i40e_pf_host_send_msg_to_vf(vf, I40E_VIRTCHNL_OP_GET_STATS, + i40e_pf_host_send_msg_to_vf(vf, VIRTCHNL_OP_GET_STATS, I40E_SUCCESS, (uint8_t *)&vf->vsi->eth_stats, sizeof(vf->vsi->eth_stats)); else - i40e_pf_host_send_msg_to_vf(vf, I40E_VIRTCHNL_OP_GET_STATS, + i40e_pf_host_send_msg_to_vf(vf, VIRTCHNL_OP_GET_STATS, I40E_NOT_SUPPORTED, (uint8_t *)&vf->vsi->eth_stats, sizeof(vf->vsi->eth_stats)); @@ -1120,8 +1120,8 @@ i40e_pf_host_process_cmd_cfg_vlan_offload( bool b_op) { int ret = I40E_SUCCESS; - struct i40e_virtchnl_vlan_offload_info *offload = - (struct i40e_virtchnl_vlan_offload_info *)msg; + struct virtchnl_vlan_offload_info *offload = + (struct virtchnl_vlan_offload_info *)msg; if (!b_op) { i40e_pf_host_send_msg_to_vf( @@ -1155,8 +1155,8 @@ i40e_pf_host_process_cmd_cfg_pvid(struct i40e_pf_vf *vf, bool b_op) { int ret = I40E_SUCCESS; - struct i40e_virtchnl_pvid_info *tpid_info = - (struct i40e_virtchnl_pvid_info *)msg; + struct virtchnl_pvid_info *tpid_info = + (struct virtchnl_pvid_info *)msg; if (!b_op) { i40e_pf_host_send_msg_to_vf( @@ -1183,40 +1183,47 @@ i40e_pf_host_process_cmd_cfg_pvid(struct i40e_pf_vf *vf, void i40e_notify_vf_link_status(struct rte_eth_dev *dev, struct i40e_pf_vf *vf) { - struct i40e_virtchnl_pf_event event; + struct virtchnl_pf_event event; - event.event = I40E_VIRTCHNL_EVENT_LINK_CHANGE; + event.event = VIRTCHNL_EVENT_LINK_CHANGE; event.event_data.link_event.link_status = dev->data->dev_link.link_status; - /* need to convert the ETH_SPEED_xxx into I40E_LINK_SPEED_xxx */ + /* need to convert the ETH_SPEED_xxx into VIRTCHNL_LINK_SPEED_xxx */ switch (dev->data->dev_link.link_speed) { case ETH_SPEED_NUM_100M: - event.event_data.link_event.link_speed = I40E_LINK_SPEED_100MB; + event.event_data.link_event.link_speed = + VIRTCHNL_LINK_SPEED_100MB; break; case ETH_SPEED_NUM_1G: - event.event_data.link_event.link_speed = I40E_LINK_SPEED_1GB; + event.event_data.link_event.link_speed = + VIRTCHNL_LINK_SPEED_1GB; break; case ETH_SPEED_NUM_10G: - event.event_data.link_event.link_speed = I40E_LINK_SPEED_10GB; + event.event_data.link_event.link_speed = + VIRTCHNL_LINK_SPEED_10GB; break; case ETH_SPEED_NUM_20G: - event.event_data.link_event.link_speed = I40E_LINK_SPEED_20GB; + event.event_data.link_event.link_speed = + VIRTCHNL_LINK_SPEED_20GB; break; case ETH_SPEED_NUM_25G: - event.event_data.link_event.link_speed = I40E_LINK_SPEED_25GB; + event.event_data.link_event.link_speed = + VIRTCHNL_LINK_SPEED_25GB; break; case ETH_SPEED_NUM_40G: - event.event_data.link_event.link_speed = I40E_LINK_SPEED_40GB; + event.event_data.link_event.link_speed = + VIRTCHNL_LINK_SPEED_40GB; break; default: event.event_data.link_event.link_speed = - I40E_LINK_SPEED_UNKNOWN; + VIRTCHNL_LINK_SPEED_UNKNOWN; break; } - i40e_pf_host_send_msg_to_vf(vf, I40E_VIRTCHNL_OP_EVENT, - I40E_SUCCESS, (uint8_t *)&event, sizeof(event)); + i40e_pf_host_send_msg_to_vf(vf, VIRTCHNL_OP_EVENT, + I40E_SUCCESS, (uint8_t *)&event, + sizeof(event)); } void @@ -1273,71 +1280,71 @@ i40e_pf_host_handle_vf_msg(struct rte_eth_dev *dev, } switch (opcode) { - case I40E_VIRTCHNL_OP_VERSION : + case VIRTCHNL_OP_VERSION: PMD_DRV_LOG(INFO, "OP_VERSION received"); i40e_pf_host_process_cmd_version(vf, b_op); break; - case I40E_VIRTCHNL_OP_RESET_VF : + case VIRTCHNL_OP_RESET_VF: PMD_DRV_LOG(INFO, "OP_RESET_VF received"); i40e_pf_host_process_cmd_reset_vf(vf); break; - case I40E_VIRTCHNL_OP_GET_VF_RESOURCES: + case VIRTCHNL_OP_GET_VF_RESOURCES: PMD_DRV_LOG(INFO, "OP_GET_VF_RESOURCES received"); i40e_pf_host_process_cmd_get_vf_resource(vf, b_op); break; - case I40E_VIRTCHNL_OP_CONFIG_VSI_QUEUES: + case VIRTCHNL_OP_CONFIG_VSI_QUEUES: PMD_DRV_LOG(INFO, "OP_CONFIG_VSI_QUEUES received"); i40e_pf_host_process_cmd_config_vsi_queues(vf, msg, msglen, b_op); break; - case I40E_VIRTCHNL_OP_CONFIG_VSI_QUEUES_EXT: + case VIRTCHNL_OP_CONFIG_VSI_QUEUES_EXT: PMD_DRV_LOG(INFO, "OP_CONFIG_VSI_QUEUES_EXT received"); i40e_pf_host_process_cmd_config_vsi_queues_ext(vf, msg, msglen, b_op); break; - case I40E_VIRTCHNL_OP_CONFIG_IRQ_MAP: + case VIRTCHNL_OP_CONFIG_IRQ_MAP: PMD_DRV_LOG(INFO, "OP_CONFIG_IRQ_MAP received"); i40e_pf_host_process_cmd_config_irq_map(vf, msg, msglen, b_op); break; - case I40E_VIRTCHNL_OP_ENABLE_QUEUES: + case VIRTCHNL_OP_ENABLE_QUEUES: PMD_DRV_LOG(INFO, "OP_ENABLE_QUEUES received"); if (b_op) { i40e_pf_host_process_cmd_enable_queues(vf, msg, msglen); i40e_notify_vf_link_status(dev, vf); } else { i40e_pf_host_send_msg_to_vf( - vf, I40E_VIRTCHNL_OP_ENABLE_QUEUES, + vf, VIRTCHNL_OP_ENABLE_QUEUES, I40E_NOT_SUPPORTED, NULL, 0); } break; - case I40E_VIRTCHNL_OP_DISABLE_QUEUES: + case VIRTCHNL_OP_DISABLE_QUEUES: PMD_DRV_LOG(INFO, "OP_DISABLE_QUEUE received"); i40e_pf_host_process_cmd_disable_queues(vf, msg, msglen, b_op); break; - case I40E_VIRTCHNL_OP_ADD_ETHER_ADDRESS: + case VIRTCHNL_OP_ADD_ETH_ADDR: PMD_DRV_LOG(INFO, "OP_ADD_ETHER_ADDRESS received"); i40e_pf_host_process_cmd_add_ether_address(vf, msg, msglen, b_op); break; - case I40E_VIRTCHNL_OP_DEL_ETHER_ADDRESS: + case VIRTCHNL_OP_DEL_ETH_ADDR: PMD_DRV_LOG(INFO, "OP_DEL_ETHER_ADDRESS received"); i40e_pf_host_process_cmd_del_ether_address(vf, msg, msglen, b_op); break; - case I40E_VIRTCHNL_OP_ADD_VLAN: + case VIRTCHNL_OP_ADD_VLAN: PMD_DRV_LOG(INFO, "OP_ADD_VLAN received"); i40e_pf_host_process_cmd_add_vlan(vf, msg, msglen, b_op); break; - case I40E_VIRTCHNL_OP_DEL_VLAN: + case VIRTCHNL_OP_DEL_VLAN: PMD_DRV_LOG(INFO, "OP_DEL_VLAN received"); i40e_pf_host_process_cmd_del_vlan(vf, msg, msglen, b_op); break; - case I40E_VIRTCHNL_OP_CONFIG_PROMISCUOUS_MODE: + case VIRTCHNL_OP_CONFIG_PROMISCUOUS_MODE: PMD_DRV_LOG(INFO, "OP_CONFIG_PROMISCUOUS_MODE received"); i40e_pf_host_process_cmd_config_promisc_mode(vf, msg, msglen, b_op); break; - case I40E_VIRTCHNL_OP_GET_STATS: + case VIRTCHNL_OP_GET_STATS: PMD_DRV_LOG(INFO, "OP_GET_STATS received"); i40e_pf_host_process_cmd_get_stats(vf, b_op); break; diff --git a/drivers/net/i40e/i40e_pf.h b/drivers/net/i40e/i40e_pf.h index 0961f06..4f93a4d 100644 --- a/drivers/net/i40e/i40e_pf.h +++ b/drivers/net/i40e/i40e_pf.h @@ -35,7 +35,7 @@ #define _I40E_PF_H_ /* VERSION info to exchange between VF and PF host. In case VF works with - * ND kernel driver, it reads I40E_VIRTCHNL_VERSION_MAJOR/MINOR. In + * ND kernel driver, it reads VIRTCHNL_VERSION_MAJOR/MINOR. In * case works with DPDK host, it reads version below. Then VF realize who it * is talking to and use proper language to communicate. * */ @@ -49,45 +49,45 @@ #define I40E_DPDK_OFFSET 0x100 /* DPDK pf driver specific command to VF */ -enum i40e_virtchnl_ops_dpdk { +enum virtchnl_ops_dpdk { /* * Keep some gap between Linux PF commands and * DPDK PF extended commands. */ - I40E_VIRTCHNL_OP_CFG_VLAN_OFFLOAD = I40E_VIRTCHNL_OP_VERSION + + I40E_VIRTCHNL_OP_CFG_VLAN_OFFLOAD = VIRTCHNL_OP_VERSION + I40E_DPDK_OFFSET, I40E_VIRTCHNL_OP_CFG_VLAN_PVID, - I40E_VIRTCHNL_OP_CONFIG_VSI_QUEUES_EXT, + VIRTCHNL_OP_CONFIG_VSI_QUEUES_EXT, }; /* A structure to support extended info of a receive queue. */ -struct i40e_virtchnl_rxq_ext_info { +struct virtchnl_rxq_ext_info { uint8_t crcstrip; }; /* * A structure to support extended info of queue pairs, an additional field - * is added, comparing to original 'struct i40e_virtchnl_queue_pair_info'. + * is added, comparing to original 'struct virtchnl_queue_pair_info'. */ -struct i40e_virtchnl_queue_pair_ext_info { +struct virtchnl_queue_pair_ext_info { /* vsi_id and queue_id should be identical for both rx and tx queues.*/ - struct i40e_virtchnl_txq_info txq; - struct i40e_virtchnl_rxq_info rxq; - struct i40e_virtchnl_rxq_ext_info rxq_ext; + struct virtchnl_txq_info txq; + struct virtchnl_rxq_info rxq; + struct virtchnl_rxq_ext_info rxq_ext; }; /* * A structure to support extended info of VSI queue pairs, - * 'struct i40e_virtchnl_queue_pair_ext_info' is used, see its original - * of 'struct i40e_virtchnl_queue_pair_info'. + * 'struct virtchnl_queue_pair_ext_info' is used, see its original + * of 'struct virtchnl_queue_pair_info'. */ -struct i40e_virtchnl_vsi_queue_config_ext_info { +struct virtchnl_vsi_queue_config_ext_info { uint16_t vsi_id; uint16_t num_queue_pairs; - struct i40e_virtchnl_queue_pair_ext_info qpair[0]; + struct virtchnl_queue_pair_ext_info qpair[0]; }; -struct i40e_virtchnl_vlan_offload_info { +struct virtchnl_vlan_offload_info { uint16_t vsi_id; uint8_t enable_vlan_strip; uint8_t reserved; @@ -106,7 +106,7 @@ struct i40e_virtchnl_vlan_offload_info { * enable op, needs to specify the pvid. PF returns status * code in retval. */ -struct i40e_virtchnl_pvid_info { +struct virtchnl_pvid_info { uint16_t vsi_id; struct i40e_vsi_vlan_pvid_info info; }; diff --git a/drivers/net/i40e/rte_pmd_i40e.h b/drivers/net/i40e/rte_pmd_i40e.h index 1efb2c4..cbe742e 100644 --- a/drivers/net/i40e/rte_pmd_i40e.h +++ b/drivers/net/i40e/rte_pmd_i40e.h @@ -59,7 +59,7 @@ enum rte_pmd_i40e_mb_event_rsp { */ struct rte_pmd_i40e_mb_event_param { uint16_t vfid; /**< Virtual Function number */ - uint16_t msg_type; /**< VF to PF message type, see i40e_virtchnl_ops */ + uint16_t msg_type; /**< VF to PF message type, see virtchnl_ops */ uint16_t retval; /**< return value */ void *msg; /**< pointer to message */ uint16_t msglen; /**< length of the message */ -- 2.4.11 ^ permalink raw reply [flat|nested] 69+ messages in thread
* [dpdk-dev] [PATCH v4 02/17] net/i40e/base: sync nvmupdate command and adminq subtask 2017-06-27 14:00 ` [dpdk-dev] [PATCH v4 00/17] net/i40e: base code update Jingjing Wu 2017-06-27 14:00 ` [dpdk-dev] [PATCH v4 01/17] net/i40e/base: use new virtchnl header file Jingjing Wu @ 2017-06-27 14:00 ` Jingjing Wu 2017-06-27 14:00 ` [dpdk-dev] [PATCH v4 03/17] net/i40e/base: add AQ command for read/write PHY registers Jingjing Wu ` (15 subsequent siblings) 17 siblings, 0 replies; 69+ messages in thread From: Jingjing Wu @ 2017-06-27 14:00 UTC (permalink / raw) To: dev; +Cc: jingjing.wu, helin.zhang During NVMupdate, state machine gets into unrecoverable state because i40e_clean_adminq_subtask can get scheduled after the admin queue command but before other state variables are updated. This patch adds locking around admin queue command and update of state variables so that adminq_subtask will have accurate information whenever it gets scheduled. Signed-off-by: Jingjing Wu <jingjing.wu@intel.com> --- drivers/net/i40e/base/i40e_nvm.c | 6 ++++++ 1 file changed, 6 insertions(+) diff --git a/drivers/net/i40e/base/i40e_nvm.c b/drivers/net/i40e/base/i40e_nvm.c index e896502..90521fa 100644 --- a/drivers/net/i40e/base/i40e_nvm.c +++ b/drivers/net/i40e/base/i40e_nvm.c @@ -899,6 +899,11 @@ enum i40e_status_code i40e_nvmupd_command(struct i40e_hw *hw, hw->nvmupd_state = I40E_NVMUPD_STATE_INIT; } + /* Acquire lock to prevent race condition where adminq_task + * can execute after i40e_nvmupd_nvm_read/write but before state + * variables (nvm_wait_opcode, nvm_release_on_done) are updated + */ + i40e_acquire_spinlock(&hw->aq.arq_spinlock); switch (hw->nvmupd_state) { case I40E_NVMUPD_STATE_INIT: status = i40e_nvmupd_state_init(hw, cmd, bytes, perrno); @@ -934,6 +939,7 @@ enum i40e_status_code i40e_nvmupd_command(struct i40e_hw *hw, *perrno = -ESRCH; break; } + i40e_release_spinlock(&hw->aq.arq_spinlock); return status; } -- 2.4.11 ^ permalink raw reply [flat|nested] 69+ messages in thread
* [dpdk-dev] [PATCH v4 03/17] net/i40e/base: add AQ command for read/write PHY registers 2017-06-27 14:00 ` [dpdk-dev] [PATCH v4 00/17] net/i40e: base code update Jingjing Wu 2017-06-27 14:00 ` [dpdk-dev] [PATCH v4 01/17] net/i40e/base: use new virtchnl header file Jingjing Wu 2017-06-27 14:00 ` [dpdk-dev] [PATCH v4 02/17] net/i40e/base: sync nvmupdate command and adminq subtask Jingjing Wu @ 2017-06-27 14:00 ` Jingjing Wu 2017-06-27 14:00 ` [dpdk-dev] [PATCH v4 04/17] net/i40e/base: add support for Adaptive Virtual Function Jingjing Wu ` (14 subsequent siblings) 17 siblings, 0 replies; 69+ messages in thread From: Jingjing Wu @ 2017-06-27 14:00 UTC (permalink / raw) To: dev; +Cc: jingjing.wu, helin.zhang This patch adds new additional command for accessing to PHY registers. Signed-off-by: Jingjing Wu <jingjing.wu@intel.com> --- drivers/net/i40e/base/i40e_adminq_cmd.h | 18 +++++++++ drivers/net/i40e/base/i40e_common.c | 70 +++++++++++++++++++++++++++++++++ drivers/net/i40e/base/i40e_prototype.h | 9 +++++ 3 files changed, 97 insertions(+) diff --git a/drivers/net/i40e/base/i40e_adminq_cmd.h b/drivers/net/i40e/base/i40e_adminq_cmd.h index 09f5bf5..83b28f8 100644 --- a/drivers/net/i40e/base/i40e_adminq_cmd.h +++ b/drivers/net/i40e/base/i40e_adminq_cmd.h @@ -245,6 +245,8 @@ enum i40e_admin_queue_opc { i40e_aqc_opc_set_phy_debug = 0x0622, i40e_aqc_opc_upload_ext_phy_fm = 0x0625, i40e_aqc_opc_run_phy_activity = 0x0626, + i40e_aqc_opc_set_phy_register = 0x0628, + i40e_aqc_opc_get_phy_register = 0x0629, /* NVM commands */ i40e_aqc_opc_nvm_read = 0x0701, @@ -2128,6 +2130,22 @@ struct i40e_aqc_run_phy_activity { I40E_CHECK_CMD_LENGTH(i40e_aqc_run_phy_activity); +/* Set PHY Register command (0x0628) */ +/* Get PHY Register command (0x0629) */ +struct i40e_aqc_phy_register_access { + u8 phy_interface; +#define I40E_AQ_PHY_REG_ACCESS_INTERNAL 0 +#define I40E_AQ_PHY_REG_ACCESS_EXTERNAL 1 +#define I40E_AQ_PHY_REG_ACCESS_EXTERNAL_MODULE 2 + u8 dev_addres; + u8 reserved1[2]; + u32 reg_address; + u32 reg_value; + u8 reserved2[4]; +}; + +I40E_CHECK_CMD_LENGTH(i40e_aqc_phy_register_access); + /* NVM Read command (indirect 0x0701) * NVM Erase commands (direct 0x0702) * NVM Update commands (indirect 0x0703) diff --git a/drivers/net/i40e/base/i40e_common.c b/drivers/net/i40e/base/i40e_common.c index 6c09c27..280d9da 100644 --- a/drivers/net/i40e/base/i40e_common.c +++ b/drivers/net/i40e/base/i40e_common.c @@ -6863,6 +6863,76 @@ void i40e_write_rx_ctl(struct i40e_hw *hw, u32 reg_addr, u32 reg_val) if (status || use_register) wr32(hw, reg_addr, reg_val); } + +/** + * i40e_aq_set_phy_register + * @hw: pointer to the hw struct + * @phy_select: select which phy should be accessed + * @dev_addr: PHY device address + * @reg_addr: PHY register address + * @reg_val: new register value + * @cmd_details: pointer to command details structure or NULL + * + * Write the external PHY register. + **/ +enum i40e_status_code i40e_aq_set_phy_register(struct i40e_hw *hw, + u8 phy_select, u8 dev_addr, + u32 reg_addr, u32 reg_val, + struct i40e_asq_cmd_details *cmd_details) +{ + struct i40e_aq_desc desc; + struct i40e_aqc_phy_register_access *cmd = + (struct i40e_aqc_phy_register_access *)&desc.params.raw; + enum i40e_status_code status; + + i40e_fill_default_direct_cmd_desc(&desc, + i40e_aqc_opc_set_phy_register); + + cmd->phy_interface = phy_select; + cmd->dev_addres = dev_addr; + cmd->reg_address = reg_addr; + cmd->reg_value = reg_val; + + status = i40e_asq_send_command(hw, &desc, NULL, 0, cmd_details); + + return status; +} + +/** + * i40e_aq_get_phy_register + * @hw: pointer to the hw struct + * @phy_select: select which phy should be accessed + * @dev_addr: PHY device address + * @reg_addr: PHY register address + * @reg_val: read register value + * @cmd_details: pointer to command details structure or NULL + * + * Read the external PHY register. + **/ +enum i40e_status_code i40e_aq_get_phy_register(struct i40e_hw *hw, + u8 phy_select, u8 dev_addr, + u32 reg_addr, u32 *reg_val, + struct i40e_asq_cmd_details *cmd_details) +{ + struct i40e_aq_desc desc; + struct i40e_aqc_phy_register_access *cmd = + (struct i40e_aqc_phy_register_access *)&desc.params.raw; + enum i40e_status_code status; + + i40e_fill_default_direct_cmd_desc(&desc, + i40e_aqc_opc_get_phy_register); + + cmd->phy_interface = phy_select; + cmd->dev_addres = dev_addr; + cmd->reg_address = reg_addr; + + status = i40e_asq_send_command(hw, &desc, NULL, 0, cmd_details); + if (!status) + *reg_val = cmd->reg_value; + + return status; +} + #ifdef VF_DRIVER /** diff --git a/drivers/net/i40e/base/i40e_prototype.h b/drivers/net/i40e/base/i40e_prototype.h index 6ec4304..54b6750 100644 --- a/drivers/net/i40e/base/i40e_prototype.h +++ b/drivers/net/i40e/base/i40e_prototype.h @@ -533,6 +533,15 @@ enum i40e_status_code i40e_aq_rx_ctl_write_register(struct i40e_hw *hw, u32 reg_addr, u32 reg_val, struct i40e_asq_cmd_details *cmd_details); void i40e_write_rx_ctl(struct i40e_hw *hw, u32 reg_addr, u32 reg_val); +enum i40e_status_code i40e_aq_set_phy_register(struct i40e_hw *hw, + u8 phy_select, u8 dev_addr, + u32 reg_addr, u32 reg_val, + struct i40e_asq_cmd_details *cmd_details); +enum i40e_status_code i40e_aq_get_phy_register(struct i40e_hw *hw, + u8 phy_select, u8 dev_addr, + u32 reg_addr, u32 *reg_val, + struct i40e_asq_cmd_details *cmd_details); + enum i40e_status_code i40e_aq_set_arp_proxy_config(struct i40e_hw *hw, struct i40e_aqc_arp_proxy_data *proxy_config, struct i40e_asq_cmd_details *cmd_details); -- 2.4.11 ^ permalink raw reply [flat|nested] 69+ messages in thread
* [dpdk-dev] [PATCH v4 04/17] net/i40e/base: add support for Adaptive Virtual Function 2017-06-27 14:00 ` [dpdk-dev] [PATCH v4 00/17] net/i40e: base code update Jingjing Wu ` (2 preceding siblings ...) 2017-06-27 14:00 ` [dpdk-dev] [PATCH v4 03/17] net/i40e/base: add AQ command for read/write PHY registers Jingjing Wu @ 2017-06-27 14:00 ` Jingjing Wu 2017-06-27 14:00 ` [dpdk-dev] [PATCH v4 05/17] net/i40e/base: store the requested FEC information Jingjing Wu ` (13 subsequent siblings) 17 siblings, 0 replies; 69+ messages in thread From: Jingjing Wu @ 2017-06-27 14:00 UTC (permalink / raw) To: dev; +Cc: jingjing.wu, helin.zhang Add device id define and mac_type assignment needed for Adaptive Virtual Function. Signed-off-by: Jingjing Wu <jingjing.wu@intel.com> --- drivers/net/i40e/base/i40e_common.c | 1 + drivers/net/i40e/base/i40e_devids.h | 1 + 2 files changed, 2 insertions(+) diff --git a/drivers/net/i40e/base/i40e_common.c b/drivers/net/i40e/base/i40e_common.c index 280d9da..5f6dd7b 100644 --- a/drivers/net/i40e/base/i40e_common.c +++ b/drivers/net/i40e/base/i40e_common.c @@ -93,6 +93,7 @@ STATIC enum i40e_status_code i40e_set_mac_type(struct i40e_hw *hw) #if defined(INTEGRATED_VF) || defined(VF_DRIVER) case I40E_DEV_ID_VF: case I40E_DEV_ID_VF_HV: + case I40E_DEV_ID_ADAPTIVE_VF: hw->mac.type = I40E_MAC_VF; break; #endif diff --git a/drivers/net/i40e/base/i40e_devids.h b/drivers/net/i40e/base/i40e_devids.h index 4546689..f4a8784 100644 --- a/drivers/net/i40e/base/i40e_devids.h +++ b/drivers/net/i40e/base/i40e_devids.h @@ -54,6 +54,7 @@ POSSIBILITY OF SUCH DAMAGE. #if defined(INTEGRATED_VF) || defined(VF_DRIVER) || defined(I40E_NDIS_SUPPORT) #define I40E_DEV_ID_VF 0x154C #define I40E_DEV_ID_VF_HV 0x1571 +#define I40E_DEV_ID_ADAPTIVE_VF 0x1889 #endif /* VF_DRIVER */ #ifdef X722_A0_SUPPORT #define I40E_DEV_ID_X722_A0 0x374C -- 2.4.11 ^ permalink raw reply [flat|nested] 69+ messages in thread
* [dpdk-dev] [PATCH v4 05/17] net/i40e/base: store the requested FEC information 2017-06-27 14:00 ` [dpdk-dev] [PATCH v4 00/17] net/i40e: base code update Jingjing Wu ` (3 preceding siblings ...) 2017-06-27 14:00 ` [dpdk-dev] [PATCH v4 04/17] net/i40e/base: add support for Adaptive Virtual Function Jingjing Wu @ 2017-06-27 14:00 ` Jingjing Wu 2017-06-27 14:00 ` [dpdk-dev] [PATCH v4 06/17] net/i40e/base: add new phy types for 25G AOC and ACC Jingjing Wu ` (12 subsequent siblings) 17 siblings, 0 replies; 69+ messages in thread From: Jingjing Wu @ 2017-06-27 14:00 UTC (permalink / raw) To: dev; +Cc: jingjing.wu, helin.zhang Store information about FEC modes, that were requested. It will be used in printing link status information function and this way there is no need to call admin queue there. Signed-off-by: Jingjing Wu <jingjing.wu@intel.com> --- drivers/net/i40e/base/i40e_common.c | 4 ++++ drivers/net/i40e/base/i40e_type.h | 1 + 2 files changed, 5 insertions(+) diff --git a/drivers/net/i40e/base/i40e_common.c b/drivers/net/i40e/base/i40e_common.c index 5f6dd7b..7d59b59 100644 --- a/drivers/net/i40e/base/i40e_common.c +++ b/drivers/net/i40e/base/i40e_common.c @@ -2826,6 +2826,10 @@ enum i40e_status_code i40e_update_link_info(struct i40e_hw *hw) if (status) return status; + hw->phy.link_info.req_fec_info = + abilities.fec_cfg_curr_mod_ext_info & + (I40E_AQ_REQUEST_FEC_KR | I40E_AQ_REQUEST_FEC_RS); + i40e_memcpy(hw->phy.link_info.module_type, &abilities.module_type, sizeof(hw->phy.link_info.module_type), I40E_NONDMA_TO_NONDMA); } diff --git a/drivers/net/i40e/base/i40e_type.h b/drivers/net/i40e/base/i40e_type.h index 52a114a..af5347b 100644 --- a/drivers/net/i40e/base/i40e_type.h +++ b/drivers/net/i40e/base/i40e_type.h @@ -271,6 +271,7 @@ struct i40e_link_status { enum i40e_aq_link_speed link_speed; u8 link_info; u8 an_info; + u8 req_fec_info; u8 fec_info; u8 ext_info; u8 loopback; -- 2.4.11 ^ permalink raw reply [flat|nested] 69+ messages in thread
* [dpdk-dev] [PATCH v4 06/17] net/i40e/base: add new phy types for 25G AOC and ACC 2017-06-27 14:00 ` [dpdk-dev] [PATCH v4 00/17] net/i40e: base code update Jingjing Wu ` (4 preceding siblings ...) 2017-06-27 14:00 ` [dpdk-dev] [PATCH v4 05/17] net/i40e/base: store the requested FEC information Jingjing Wu @ 2017-06-27 14:00 ` Jingjing Wu 2017-06-27 14:00 ` [dpdk-dev] [PATCH v4 07/17] net/i40e/base: report supported link modes Jingjing Wu ` (11 subsequent siblings) 17 siblings, 0 replies; 69+ messages in thread From: Jingjing Wu @ 2017-06-27 14:00 UTC (permalink / raw) To: dev; +Cc: jingjing.wu, helin.zhang This patch adds new phy types for 25G Active Optical Cables (AOC) and Active Copper Cables (ACC) support. Signed-off-by: Jingjing Wu <jingjing.wu@intel.com> --- drivers/net/i40e/base/i40e_adminq_cmd.h | 4 ++++ drivers/net/i40e/base/i40e_common.c | 2 ++ 2 files changed, 6 insertions(+) diff --git a/drivers/net/i40e/base/i40e_adminq_cmd.h b/drivers/net/i40e/base/i40e_adminq_cmd.h index 83b28f8..70079a0 100644 --- a/drivers/net/i40e/base/i40e_adminq_cmd.h +++ b/drivers/net/i40e/base/i40e_adminq_cmd.h @@ -1849,6 +1849,8 @@ enum i40e_aq_phy_type { I40E_PHY_TYPE_25GBASE_CR = 0x20, I40E_PHY_TYPE_25GBASE_SR = 0x21, I40E_PHY_TYPE_25GBASE_LR = 0x22, + I40E_PHY_TYPE_25GBASE_AOC = 0x23, + I40E_PHY_TYPE_25GBASE_ACC = 0x24, I40E_PHY_TYPE_MAX }; @@ -1906,6 +1908,8 @@ struct i40e_aq_get_phy_abilities_resp { #define I40E_AQ_PHY_TYPE_EXT_25G_CR 0x02 #define I40E_AQ_PHY_TYPE_EXT_25G_SR 0x04 #define I40E_AQ_PHY_TYPE_EXT_25G_LR 0x08 +#define I40E_AQ_PHY_TYPE_EXT_25G_AOC 0x10 +#define I40E_AQ_PHY_TYPE_EXT_25G_ACC 0x20 u8 fec_cfg_curr_mod_ext_info; #define I40E_AQ_ENABLE_FEC_KR 0x01 #define I40E_AQ_ENABLE_FEC_RS 0x02 diff --git a/drivers/net/i40e/base/i40e_common.c b/drivers/net/i40e/base/i40e_common.c index 7d59b59..004c062 100644 --- a/drivers/net/i40e/base/i40e_common.c +++ b/drivers/net/i40e/base/i40e_common.c @@ -1296,6 +1296,8 @@ STATIC enum i40e_media_type i40e_get_media_type(struct i40e_hw *hw) case I40E_PHY_TYPE_40GBASE_AOC: case I40E_PHY_TYPE_10GBASE_AOC: case I40E_PHY_TYPE_25GBASE_CR: + case I40E_PHY_TYPE_25GBASE_AOC: + case I40E_PHY_TYPE_25GBASE_ACC: media = I40E_MEDIA_TYPE_DA; break; case I40E_PHY_TYPE_1000BASE_KX: -- 2.4.11 ^ permalink raw reply [flat|nested] 69+ messages in thread
* [dpdk-dev] [PATCH v4 07/17] net/i40e/base: report supported link modes 2017-06-27 14:00 ` [dpdk-dev] [PATCH v4 00/17] net/i40e: base code update Jingjing Wu ` (5 preceding siblings ...) 2017-06-27 14:00 ` [dpdk-dev] [PATCH v4 06/17] net/i40e/base: add new phy types for 25G AOC and ACC Jingjing Wu @ 2017-06-27 14:00 ` Jingjing Wu 2017-06-27 14:00 ` [dpdk-dev] [PATCH v4 08/17] net/i40e/base: track id can be 0 Jingjing Wu ` (10 subsequent siblings) 17 siblings, 0 replies; 69+ messages in thread From: Jingjing Wu @ 2017-06-27 14:00 UTC (permalink / raw) To: dev; +Cc: jingjing.wu, helin.zhang Signed-off-by: Jingjing Wu <jingjing.wu@intel.com> --- drivers/net/i40e/base/i40e_adminq_cmd.h | 22 +++++++++++++++++++--- drivers/net/i40e/base/i40e_common.c | 18 +++++++++++++++--- 2 files changed, 34 insertions(+), 6 deletions(-) diff --git a/drivers/net/i40e/base/i40e_adminq_cmd.h b/drivers/net/i40e/base/i40e_adminq_cmd.h index 70079a0..c1ca290 100644 --- a/drivers/net/i40e/base/i40e_adminq_cmd.h +++ b/drivers/net/i40e/base/i40e_adminq_cmd.h @@ -1831,6 +1831,8 @@ enum i40e_aq_phy_type { I40E_PHY_TYPE_10GBASE_CR1_CU = 0xB, I40E_PHY_TYPE_10GBASE_AOC = 0xC, I40E_PHY_TYPE_40GBASE_AOC = 0xD, + I40E_PHY_TYPE_UNRECOGNIZED = 0xE, + I40E_PHY_TYPE_UNSUPPORTED = 0xF, I40E_PHY_TYPE_100BASE_TX = 0x11, I40E_PHY_TYPE_1000BASE_T = 0x12, I40E_PHY_TYPE_10GBASE_T = 0x13, @@ -1851,7 +1853,9 @@ enum i40e_aq_phy_type { I40E_PHY_TYPE_25GBASE_LR = 0x22, I40E_PHY_TYPE_25GBASE_AOC = 0x23, I40E_PHY_TYPE_25GBASE_ACC = 0x24, - I40E_PHY_TYPE_MAX + I40E_PHY_TYPE_MAX, + I40E_PHY_TYPE_EMPTY = 0xFE, + I40E_PHY_TYPE_DEFAULT = 0xFF, }; #define I40E_LINK_SPEED_100MB_SHIFT 0x1 @@ -2039,19 +2043,31 @@ struct i40e_aqc_get_link_status { #define I40E_AQ_25G_SERDES_UCODE_ERR 0X04 #define I40E_AQ_25G_NIMB_UCODE_ERR 0X05 u8 loopback; /* use defines from i40e_aqc_set_lb_mode */ +/* Since firmware API 1.7 loopback field keeps power class info as well */ +#define I40E_AQ_LOOPBACK_MASK 0x07 +#define I40E_AQ_PWR_CLASS_SHIFT_LB 6 +#define I40E_AQ_PWR_CLASS_MASK_LB (0x03 << I40E_AQ_PWR_CLASS_SHIFT_LB) __le16 max_frame_size; u8 config; #define I40E_AQ_CONFIG_FEC_KR_ENA 0x01 #define I40E_AQ_CONFIG_FEC_RS_ENA 0x02 #define I40E_AQ_CONFIG_CRC_ENA 0x04 #define I40E_AQ_CONFIG_PACING_MASK 0x78 - u8 power_desc; + union { + struct { + u8 power_desc; #define I40E_AQ_LINK_POWER_CLASS_1 0x00 #define I40E_AQ_LINK_POWER_CLASS_2 0x01 #define I40E_AQ_LINK_POWER_CLASS_3 0x02 #define I40E_AQ_LINK_POWER_CLASS_4 0x03 #define I40E_AQ_PWR_CLASS_MASK 0x03 - u8 reserved[4]; + u8 reserved[4]; + }; + struct { + u8 link_type[4]; + u8 link_type_ext; + }; + }; }; I40E_CHECK_CMD_LENGTH(i40e_aqc_get_link_status); diff --git a/drivers/net/i40e/base/i40e_common.c b/drivers/net/i40e/base/i40e_common.c index 004c062..bb4b7eb 100644 --- a/drivers/net/i40e/base/i40e_common.c +++ b/drivers/net/i40e/base/i40e_common.c @@ -1692,8 +1692,14 @@ enum i40e_status_code i40e_aq_get_phy_capabilities(struct i40e_hw *hw, status = I40E_ERR_UNKNOWN_PHY; if (report_init) { - hw->phy.phy_types = LE32_TO_CPU(abilities->phy_type); - hw->phy.phy_types |= ((u64)abilities->phy_type_ext << 32); + if (hw->aq.api_maj_ver == I40E_FW_API_VERSION_MAJOR && + hw->aq.api_min_ver >= 7) { + status = i40e_aq_get_link_info(hw, true, NULL, NULL); + } else { + hw->phy.phy_types = LE32_TO_CPU(abilities->phy_type); + hw->phy.phy_types |= + ((u64)abilities->phy_type_ext << 32); + } } return status; @@ -1955,7 +1961,7 @@ enum i40e_status_code i40e_aq_get_link_info(struct i40e_hw *hw, hw_link_info->fec_info = resp->config & (I40E_AQ_CONFIG_FEC_KR_ENA | I40E_AQ_CONFIG_FEC_RS_ENA); hw_link_info->ext_info = resp->ext_info; - hw_link_info->loopback = resp->loopback; + hw_link_info->loopback = resp->loopback & I40E_AQ_LOOPBACK_MASK; hw_link_info->max_frame_size = LE16_TO_CPU(resp->max_frame_size); hw_link_info->pacing = resp->config & I40E_AQ_CONFIG_PACING_MASK; @@ -1986,6 +1992,12 @@ enum i40e_status_code i40e_aq_get_link_info(struct i40e_hw *hw, hw->aq.fw_min_ver < 40)) && hw_link_info->phy_type == 0xE) hw_link_info->phy_type = I40E_PHY_TYPE_10GBASE_SFPP_CU; + if (hw->aq.api_maj_ver == I40E_FW_API_VERSION_MAJOR && + hw->aq.api_min_ver >= 7) { + hw->phy.phy_types = LE32_TO_CPU(*(__le32 *)resp->link_type); + hw->phy.phy_types |= ((u64)resp->link_type_ext << 32); + } + /* save link status information */ if (link) i40e_memcpy(link, hw_link_info, sizeof(*hw_link_info), -- 2.4.11 ^ permalink raw reply [flat|nested] 69+ messages in thread
* [dpdk-dev] [PATCH v4 08/17] net/i40e/base: track id can be 0 2017-06-27 14:00 ` [dpdk-dev] [PATCH v4 00/17] net/i40e: base code update Jingjing Wu ` (6 preceding siblings ...) 2017-06-27 14:00 ` [dpdk-dev] [PATCH v4 07/17] net/i40e/base: report supported link modes Jingjing Wu @ 2017-06-27 14:00 ` Jingjing Wu 2017-06-27 14:00 ` [dpdk-dev] [PATCH v4 09/17] net/i40e/base: update FW AQ API version to 1.7 Jingjing Wu ` (9 subsequent siblings) 17 siblings, 0 replies; 69+ messages in thread From: Jingjing Wu @ 2017-06-27 14:00 UTC (permalink / raw) To: dev; +Cc: jingjing.wu, helin.zhang track_id == 0 is valid for “read only” profiles when profile does not have any “write” commands. Signed-off-by: Jingjing Wu <jingjing.wu@intel.com> --- drivers/net/i40e/base/i40e_common.c | 7 +------ drivers/net/i40e/base/i40e_prototype.h | 12 +++++++----- 2 files changed, 8 insertions(+), 11 deletions(-) diff --git a/drivers/net/i40e/base/i40e_common.c b/drivers/net/i40e/base/i40e_common.c index bb4b7eb..38c30a1 100644 --- a/drivers/net/i40e/base/i40e_common.c +++ b/drivers/net/i40e/base/i40e_common.c @@ -7370,11 +7370,6 @@ i40e_write_profile(struct i40e_hw *hw, struct i40e_profile_segment *profile, u32 offset = 0, info = 0; u32 i; - if (!track_id) { - i40e_debug(hw, I40E_DEBUG_PACKAGE, "Track_id can't be 0."); - return I40E_NOT_SUPPORTED; - } - dev_cnt = profile->device_table_count; for (i = 0; i < dev_cnt; i++) { @@ -7449,6 +7444,6 @@ i40e_add_pinfo_to_list(struct i40e_hw *hw, memcpy(pinfo->name, profile->name, I40E_DDP_NAME_SIZE); status = i40e_aq_write_ddp(hw, (void *)sec, sec->data_end, - track_id, &offset, &info, NULL); + track_id, &offset, &info, NULL); return status; } diff --git a/drivers/net/i40e/base/i40e_prototype.h b/drivers/net/i40e/base/i40e_prototype.h index 54b6750..9171e97 100644 --- a/drivers/net/i40e/base/i40e_prototype.h +++ b/drivers/net/i40e/base/i40e_prototype.h @@ -575,12 +575,14 @@ u8 i40e_get_phy_address(struct i40e_hw *hw, u8 dev_num); enum i40e_status_code i40e_blink_phy_link_led(struct i40e_hw *hw, u32 time, u32 interval); enum i40e_status_code i40e_aq_write_ddp(struct i40e_hw *hw, void *buff, - u16 buff_size, u32 track_id, - u32 *error_offset, u32 *error_info, - struct i40e_asq_cmd_details *cmd_details); + u16 buff_size, u32 track_id, + u32 *error_offset, u32 *error_info, + struct i40e_asq_cmd_details * + cmd_details); enum i40e_status_code i40e_aq_get_ddp_list(struct i40e_hw *hw, void *buff, - u16 buff_size, u8 flags, - struct i40e_asq_cmd_details *cmd_details); + u16 buff_size, u8 flags, + struct i40e_asq_cmd_details * + cmd_details); struct i40e_generic_seg_header * i40e_find_segment_in_package(u32 segment_type, struct i40e_package_header *pkg_header); -- 2.4.11 ^ permalink raw reply [flat|nested] 69+ messages in thread
* [dpdk-dev] [PATCH v4 09/17] net/i40e/base: update FW AQ API version to 1.7 2017-06-27 14:00 ` [dpdk-dev] [PATCH v4 00/17] net/i40e: base code update Jingjing Wu ` (7 preceding siblings ...) 2017-06-27 14:00 ` [dpdk-dev] [PATCH v4 08/17] net/i40e/base: track id can be 0 Jingjing Wu @ 2017-06-27 14:00 ` Jingjing Wu 2017-06-27 14:00 ` [dpdk-dev] [PATCH v4 10/17] net/i40e/base: add support for switch parameters Jingjing Wu ` (8 subsequent siblings) 17 siblings, 0 replies; 69+ messages in thread From: Jingjing Wu @ 2017-06-27 14:00 UTC (permalink / raw) To: dev; +Cc: jingjing.wu, helin.zhang Signed-off-by: Jingjing Wu <jingjing.wu@intel.com> --- drivers/net/i40e/base/i40e_adminq_cmd.h | 10 +++++++++- drivers/net/i40e/base/i40e_common.c | 5 +++-- drivers/net/i40e/base/i40e_type.h | 4 ---- 3 files changed, 12 insertions(+), 7 deletions(-) diff --git a/drivers/net/i40e/base/i40e_adminq_cmd.h b/drivers/net/i40e/base/i40e_adminq_cmd.h index c1ca290..dd9fb55 100644 --- a/drivers/net/i40e/base/i40e_adminq_cmd.h +++ b/drivers/net/i40e/base/i40e_adminq_cmd.h @@ -41,7 +41,15 @@ POSSIBILITY OF SUCH DAMAGE. */ #define I40E_FW_API_VERSION_MAJOR 0x0001 -#define I40E_FW_API_VERSION_MINOR 0x0005 +#define I40E_FW_API_VERSION_MINOR_X722 0x0005 +#define I40E_FW_API_VERSION_MINOR_X710 0x0007 + +#define I40E_FW_MINOR_VERSION(_h) ((_h)->mac.type == I40E_MAC_XL710 ? \ + I40E_FW_API_VERSION_MINOR_X710 : \ + I40E_FW_API_VERSION_MINOR_X722) + +/* API version 1.7 implements additional link and PHY-specific APIs */ +#define I40E_MINOR_VER_GET_LINK_INFO_XL710 0x0007 struct i40e_aq_desc { __le16 flags; diff --git a/drivers/net/i40e/base/i40e_common.c b/drivers/net/i40e/base/i40e_common.c index 38c30a1..9895a77 100644 --- a/drivers/net/i40e/base/i40e_common.c +++ b/drivers/net/i40e/base/i40e_common.c @@ -1692,8 +1692,9 @@ enum i40e_status_code i40e_aq_get_phy_capabilities(struct i40e_hw *hw, status = I40E_ERR_UNKNOWN_PHY; if (report_init) { - if (hw->aq.api_maj_ver == I40E_FW_API_VERSION_MAJOR && - hw->aq.api_min_ver >= 7) { + if (hw->mac.type == I40E_MAC_XL710 && + hw->aq.api_maj_ver == I40E_FW_API_VERSION_MAJOR && + hw->aq.api_min_ver >= I40E_MINOR_VER_GET_LINK_INFO_XL710) { status = i40e_aq_get_link_info(hw, true, NULL, NULL); } else { hw->phy.phy_types = LE32_TO_CPU(abilities->phy_type); diff --git a/drivers/net/i40e/base/i40e_type.h b/drivers/net/i40e/base/i40e_type.h index af5347b..6ca1a0f 100644 --- a/drivers/net/i40e/base/i40e_type.h +++ b/drivers/net/i40e/base/i40e_type.h @@ -199,10 +199,6 @@ enum i40e_memcpy_type { I40E_DMA_TO_NONDMA }; -#define I40E_FW_API_VERSION_MINOR_X722 0x0005 -#define I40E_FW_API_VERSION_MINOR_X710 0x0005 - - /* These are structs for managing the hardware information and the operations. * The structures of function pointers are filled out at init time when we * know for sure exactly which hardware we're working with. This gives us the -- 2.4.11 ^ permalink raw reply [flat|nested] 69+ messages in thread
* [dpdk-dev] [PATCH v4 10/17] net/i40e/base: add support for switch parameters 2017-06-27 14:00 ` [dpdk-dev] [PATCH v4 00/17] net/i40e: base code update Jingjing Wu ` (8 preceding siblings ...) 2017-06-27 14:00 ` [dpdk-dev] [PATCH v4 09/17] net/i40e/base: update FW AQ API version to 1.7 Jingjing Wu @ 2017-06-27 14:00 ` Jingjing Wu 2017-06-27 14:00 ` [dpdk-dev] [PATCH v4 11/17] net/i40e/base: use admin queue for setting LEDs behavior Jingjing Wu ` (7 subsequent siblings) 17 siblings, 0 replies; 69+ messages in thread From: Jingjing Wu @ 2017-06-27 14:00 UTC (permalink / raw) To: dev; +Cc: jingjing.wu, helin.zhang Adds double VLAN tagging ethertype fields to Set Switch Parameters AQ command. These were added in firmware API 1.7. Callers of i40e_aq_set_switch_config() can specify the ethertypes to use by filling out the corresponding fields in struct i40e_hw. Signed-off-by: Jingjing Wu <jingjing.wu@intel.com> --- drivers/net/i40e/base/i40e_adminq.c | 6 ++++++ drivers/net/i40e/base/i40e_adminq_cmd.h | 17 ++++++++++++++++- drivers/net/i40e/base/i40e_common.c | 6 +++++- drivers/net/i40e/base/i40e_type.h | 6 ++++++ 4 files changed, 33 insertions(+), 2 deletions(-) diff --git a/drivers/net/i40e/base/i40e_adminq.c b/drivers/net/i40e/base/i40e_adminq.c index a60292a..19871c7 100644 --- a/drivers/net/i40e/base/i40e_adminq.c +++ b/drivers/net/i40e/base/i40e_adminq.c @@ -682,6 +682,12 @@ enum i40e_status_code i40e_init_adminq(struct i40e_hw *hw) &oem_lo); hw->nvm.oem_ver = ((u32)oem_hi << 16) | oem_lo; + /* The ability to RX (not drop) 802.1ad frames was added in API 1.7 */ + if ((hw->aq.api_maj_ver > 1) || + ((hw->aq.api_maj_ver == 1) && + (hw->aq.api_min_ver >= 7))) + hw->flags |= I40E_HW_FLAG_802_1AD_CAPABLE; + if (hw->aq.api_maj_ver > I40E_FW_API_VERSION_MAJOR) { ret_code = I40E_ERR_FIRMWARE_API_VERSION; goto init_adminq_free_arq; diff --git a/drivers/net/i40e/base/i40e_adminq_cmd.h b/drivers/net/i40e/base/i40e_adminq_cmd.h index dd9fb55..c36da2a 100644 --- a/drivers/net/i40e/base/i40e_adminq_cmd.h +++ b/drivers/net/i40e/base/i40e_adminq_cmd.h @@ -787,7 +787,22 @@ struct i40e_aqc_set_switch_config { #define I40E_AQ_SET_SWITCH_CFG_L2_FILTER 0x0002 #define I40E_AQ_SET_SWITCH_CFG_HW_ATR_EVICT 0x0004 __le16 valid_flags; - u8 reserved[12]; + /* The ethertype in switch_tag is dropped on ingress and used + * internally by the switch. Set this to zero for the default + * of 0x88a8 (802.1ad). Should be zero for firmware API + * versions lower than 1.7. + */ + __le16 switch_tag; + /* The ethertypes in first_tag and second_tag are used to + * match the outer and inner VLAN tags (respectively) when HW + * double VLAN tagging is enabled via the set port parameters + * AQ command. Otherwise these are both ignored. Set them to + * zero for their defaults of 0x8100 (802.1Q). Should be zero + * for firmware API versions lower than 1.7. + */ + __le16 first_tag; + __le16 second_tag; + u8 reserved[6]; }; I40E_CHECK_CMD_LENGTH(i40e_aqc_set_switch_config); diff --git a/drivers/net/i40e/base/i40e_common.c b/drivers/net/i40e/base/i40e_common.c index 9895a77..b4901ef 100644 --- a/drivers/net/i40e/base/i40e_common.c +++ b/drivers/net/i40e/base/i40e_common.c @@ -2695,7 +2695,11 @@ enum i40e_status_code i40e_aq_set_switch_config(struct i40e_hw *hw, i40e_aqc_opc_set_switch_config); scfg->flags = CPU_TO_LE16(flags); scfg->valid_flags = CPU_TO_LE16(valid_flags); - + if (hw->flags & I40E_HW_FLAG_802_1AD_CAPABLE) { + scfg->switch_tag = CPU_TO_LE16(hw->switch_tag); + scfg->first_tag = CPU_TO_LE16(hw->first_tag); + scfg->second_tag = CPU_TO_LE16(hw->second_tag); + } status = i40e_asq_send_command(hw, &desc, NULL, 0, cmd_details); return status; diff --git a/drivers/net/i40e/base/i40e_type.h b/drivers/net/i40e/base/i40e_type.h index 6ca1a0f..152b4a7 100644 --- a/drivers/net/i40e/base/i40e_type.h +++ b/drivers/net/i40e/base/i40e_type.h @@ -700,8 +700,14 @@ struct i40e_hw { u16 wol_proxy_vsi_seid; #define I40E_HW_FLAG_AQ_SRCTL_ACCESS_ENABLE BIT_ULL(0) +#define I40E_HW_FLAG_802_1AD_CAPABLE BIT_ULL(1) u64 flags; + /* Used in set switch config AQ command */ + u16 switch_tag; + u16 first_tag; + u16 second_tag; + /* debug mask */ u32 debug_mask; char err_str[16]; -- 2.4.11 ^ permalink raw reply [flat|nested] 69+ messages in thread
* [dpdk-dev] [PATCH v4 11/17] net/i40e/base: use admin queue for setting LEDs behavior 2017-06-27 14:00 ` [dpdk-dev] [PATCH v4 00/17] net/i40e: base code update Jingjing Wu ` (9 preceding siblings ...) 2017-06-27 14:00 ` [dpdk-dev] [PATCH v4 10/17] net/i40e/base: add support for switch parameters Jingjing Wu @ 2017-06-27 14:00 ` Jingjing Wu 2017-06-27 14:00 ` [dpdk-dev] [PATCH v4 12/17] net/i40e/base: avoid potential null pointer dereference Jingjing Wu ` (6 subsequent siblings) 17 siblings, 0 replies; 69+ messages in thread From: Jingjing Wu @ 2017-06-27 14:00 UTC (permalink / raw) To: dev; +Cc: jingjing.wu, helin.zhang Instead of accessing register directly, use newly added AQC in order to blink LEDs. Introduce and utilize a new flag to prevent excessive API version checking. Signed-off-by: Jingjing Wu <jingjing.wu@intel.com> --- drivers/net/i40e/base/i40e_adminq.c | 6 ++ drivers/net/i40e/base/i40e_common.c | 148 ++++++++++++++++++++++++++-------- drivers/net/i40e/base/i40e_register.h | 2 +- drivers/net/i40e/base/i40e_type.h | 1 + 4 files changed, 121 insertions(+), 36 deletions(-) diff --git a/drivers/net/i40e/base/i40e_adminq.c b/drivers/net/i40e/base/i40e_adminq.c index 19871c7..8cc8c5e 100644 --- a/drivers/net/i40e/base/i40e_adminq.c +++ b/drivers/net/i40e/base/i40e_adminq.c @@ -688,6 +688,12 @@ enum i40e_status_code i40e_init_adminq(struct i40e_hw *hw) (hw->aq.api_min_ver >= 7))) hw->flags |= I40E_HW_FLAG_802_1AD_CAPABLE; + if (hw->mac.type == I40E_MAC_XL710 && + hw->aq.api_maj_ver == I40E_FW_API_VERSION_MAJOR && + hw->aq.api_min_ver >= I40E_MINOR_VER_GET_LINK_INFO_XL710) { + hw->flags |= I40E_HW_FLAG_AQ_PHY_ACCESS_CAPABLE; + } + if (hw->aq.api_maj_ver > I40E_FW_API_VERSION_MAJOR) { ret_code = I40E_ERR_FIRMWARE_API_VERSION; goto init_adminq_free_arq; diff --git a/drivers/net/i40e/base/i40e_common.c b/drivers/net/i40e/base/i40e_common.c index b4901ef..3f3b275 100644 --- a/drivers/net/i40e/base/i40e_common.c +++ b/drivers/net/i40e/base/i40e_common.c @@ -6678,24 +6678,38 @@ enum i40e_status_code i40e_led_get_phy(struct i40e_hw *hw, u16 *led_addr, u16 temp_addr; u8 port_num; u32 i; - - temp_addr = I40E_PHY_LED_PROV_REG_1; - i = rd32(hw, I40E_PFGEN_PORTNUM); - port_num = (u8)(i & I40E_PFGEN_PORTNUM_PORT_NUM_MASK); - phy_addr = i40e_get_phy_address(hw, port_num); - - for (gpio_led_port = 0; gpio_led_port < 3; gpio_led_port++, - temp_addr++) { - status = i40e_read_phy_register_clause45(hw, + u32 reg_val_aq; + + if (hw->flags & I40E_HW_FLAG_AQ_PHY_ACCESS_CAPABLE) { + status = + i40e_aq_get_phy_register(hw, + I40E_AQ_PHY_REG_ACCESS_EXTERNAL, + I40E_PHY_COM_REG_PAGE, + I40E_PHY_LED_PROV_REG_1, + ®_val_aq, NULL); + if (status) + return status; + *val = (u16)reg_val_aq; + } else { + temp_addr = I40E_PHY_LED_PROV_REG_1; + i = rd32(hw, I40E_PFGEN_PORTNUM); + port_num = (u8)(i & I40E_PFGEN_PORTNUM_PORT_NUM_MASK); + phy_addr = i40e_get_phy_address(hw, port_num); + + for (gpio_led_port = 0; gpio_led_port < 3; gpio_led_port++, + temp_addr++) { + status = + i40e_read_phy_register_clause45(hw, I40E_PHY_COM_REG_PAGE, temp_addr, phy_addr, ®_val); - if (status) - return status; - *val = reg_val; - if (reg_val & I40E_PHY_LED_LINK_MODE_MASK) { - *led_addr = temp_addr; - break; + if (status) + return status; + *val = reg_val; + if (reg_val & I40E_PHY_LED_LINK_MODE_MASK) { + *led_addr = temp_addr; + break; + } } } return status; @@ -6713,51 +6727,115 @@ enum i40e_status_code i40e_led_set_phy(struct i40e_hw *hw, bool on, u16 led_addr, u32 mode) { enum i40e_status_code status = I40E_SUCCESS; - u16 led_ctl = 0; - u16 led_reg = 0; + u32 led_ctl = 0; + u32 led_reg = 0; u8 phy_addr = 0; u8 port_num; u32 i; - i = rd32(hw, I40E_PFGEN_PORTNUM); - port_num = (u8)(i & I40E_PFGEN_PORTNUM_PORT_NUM_MASK); - phy_addr = i40e_get_phy_address(hw, port_num); - status = i40e_read_phy_register_clause45(hw, I40E_PHY_COM_REG_PAGE, - led_addr, phy_addr, &led_reg); + if (hw->flags & I40E_HW_FLAG_AQ_PHY_ACCESS_CAPABLE) { + status = + i40e_aq_get_phy_register(hw, + I40E_AQ_PHY_REG_ACCESS_EXTERNAL, + I40E_PHY_COM_REG_PAGE, + I40E_PHY_LED_PROV_REG_1, + &led_reg, NULL); + } else { + i = rd32(hw, I40E_PFGEN_PORTNUM); + port_num = (u8)(i & I40E_PFGEN_PORTNUM_PORT_NUM_MASK); + phy_addr = i40e_get_phy_address(hw, port_num); + status = i40e_read_phy_register_clause45(hw, + I40E_PHY_COM_REG_PAGE, + led_addr, phy_addr, + (u16 *)&led_reg); + } if (status) return status; led_ctl = led_reg; if (led_reg & I40E_PHY_LED_LINK_MODE_MASK) { led_reg = 0; - status = i40e_write_phy_register_clause45(hw, - I40E_PHY_COM_REG_PAGE, - led_addr, phy_addr, - led_reg); + if (hw->aq.api_maj_ver == I40E_FW_API_VERSION_MAJOR && + hw->aq.api_min_ver >= I40E_MINOR_VER_GET_LINK_INFO_XL710) { + status = i40e_aq_set_phy_register(hw, + I40E_AQ_PHY_REG_ACCESS_EXTERNAL, + I40E_PHY_COM_REG_PAGE, + I40E_PHY_LED_PROV_REG_1, + led_reg, NULL); + } else { + status = i40e_write_phy_register_clause45(hw, + I40E_PHY_COM_REG_PAGE, + led_addr, phy_addr, + (u16)led_reg); + } if (status) return status; } - status = i40e_read_phy_register_clause45(hw, I40E_PHY_COM_REG_PAGE, - led_addr, phy_addr, &led_reg); + if (hw->flags & I40E_HW_FLAG_AQ_PHY_ACCESS_CAPABLE) { + status = + i40e_aq_get_phy_register(hw, + I40E_AQ_PHY_REG_ACCESS_EXTERNAL, + I40E_PHY_COM_REG_PAGE, + I40E_PHY_LED_PROV_REG_1, + &led_reg, NULL); + } else { + status = i40e_read_phy_register_clause45(hw, + I40E_PHY_COM_REG_PAGE, + led_addr, phy_addr, + (u16 *)&led_reg); + } if (status) goto restore_config; if (on) led_reg = I40E_PHY_LED_MANUAL_ON; else led_reg = 0; - status = i40e_write_phy_register_clause45(hw, I40E_PHY_COM_REG_PAGE, - led_addr, phy_addr, led_reg); + + if (hw->flags & I40E_HW_FLAG_AQ_PHY_ACCESS_CAPABLE) { + status = + i40e_aq_set_phy_register(hw, + I40E_AQ_PHY_REG_ACCESS_EXTERNAL, + I40E_PHY_COM_REG_PAGE, + I40E_PHY_LED_PROV_REG_1, + led_reg, NULL); + } else { + status = + i40e_write_phy_register_clause45(hw, I40E_PHY_COM_REG_PAGE, + led_addr, phy_addr, + (u16)led_reg); + } if (status) goto restore_config; if (mode & I40E_PHY_LED_MODE_ORIG) { led_ctl = (mode & I40E_PHY_LED_MODE_MASK); - status = i40e_write_phy_register_clause45(hw, - I40E_PHY_COM_REG_PAGE, - led_addr, phy_addr, led_ctl); + if (hw->flags & I40E_HW_FLAG_AQ_PHY_ACCESS_CAPABLE) { + status = i40e_aq_set_phy_register(hw, + I40E_AQ_PHY_REG_ACCESS_EXTERNAL, + I40E_PHY_COM_REG_PAGE, + I40E_PHY_LED_PROV_REG_1, + led_ctl, NULL); + } else { + status = i40e_write_phy_register_clause45(hw, + I40E_PHY_COM_REG_PAGE, + led_addr, phy_addr, + (u16)led_ctl); + } } return status; restore_config: - status = i40e_write_phy_register_clause45(hw, I40E_PHY_COM_REG_PAGE, - led_addr, phy_addr, led_ctl); + if (hw->flags & I40E_HW_FLAG_AQ_PHY_ACCESS_CAPABLE) { + status = + i40e_aq_set_phy_register(hw, + I40E_AQ_PHY_REG_ACCESS_EXTERNAL, + I40E_PHY_COM_REG_PAGE, + I40E_PHY_LED_PROV_REG_1, + led_ctl, NULL); + } else { + status = + i40e_write_phy_register_clause45(hw, + I40E_PHY_COM_REG_PAGE, + led_addr, phy_addr, + (u16)led_ctl); + } return status; } #endif /* PF_DRIVER */ diff --git a/drivers/net/i40e/base/i40e_register.h b/drivers/net/i40e/base/i40e_register.h index b150fbd..a482ab9 100644 --- a/drivers/net/i40e/base/i40e_register.h +++ b/drivers/net/i40e/base/i40e_register.h @@ -2805,7 +2805,7 @@ POSSIBILITY OF SUCH DAMAGE. #define I40E_GLV_RUPP_MAX_INDEX 383 #define I40E_GLV_RUPP_RUPP_SHIFT 0 #define I40E_GLV_RUPP_RUPP_MASK I40E_MASK(0xFFFFFFFF, I40E_GLV_RUPP_RUPP_SHIFT) -#define I40E_GLV_TEPC(_VSI) (0x00344000 + ((_VSI) * 8)) /* _i=0...383 */ /* Reset: CORER */ +#define I40E_GLV_TEPC(_i) (0x00344000 + ((_i) * 8)) /* _i=0...383 */ /* Reset: CORER */ #define I40E_GLV_TEPC_MAX_INDEX 383 #define I40E_GLV_TEPC_TEPC_SHIFT 0 #define I40E_GLV_TEPC_TEPC_MASK I40E_MASK(0xFFFFFFFF, I40E_GLV_TEPC_TEPC_SHIFT) diff --git a/drivers/net/i40e/base/i40e_type.h b/drivers/net/i40e/base/i40e_type.h index 152b4a7..9d7b1a2 100644 --- a/drivers/net/i40e/base/i40e_type.h +++ b/drivers/net/i40e/base/i40e_type.h @@ -701,6 +701,7 @@ struct i40e_hw { #define I40E_HW_FLAG_AQ_SRCTL_ACCESS_ENABLE BIT_ULL(0) #define I40E_HW_FLAG_802_1AD_CAPABLE BIT_ULL(1) +#define I40E_HW_FLAG_AQ_PHY_ACCESS_CAPABLE BIT_ULL(2) u64 flags; /* Used in set switch config AQ command */ -- 2.4.11 ^ permalink raw reply [flat|nested] 69+ messages in thread
* [dpdk-dev] [PATCH v4 12/17] net/i40e/base: avoid potential null pointer dereference 2017-06-27 14:00 ` [dpdk-dev] [PATCH v4 00/17] net/i40e: base code update Jingjing Wu ` (10 preceding siblings ...) 2017-06-27 14:00 ` [dpdk-dev] [PATCH v4 11/17] net/i40e/base: use admin queue for setting LEDs behavior Jingjing Wu @ 2017-06-27 14:00 ` Jingjing Wu 2017-06-27 14:00 ` [dpdk-dev] [PATCH v4 13/17] net/i40e/base: avoid reset timeout issue Jingjing Wu ` (5 subsequent siblings) 17 siblings, 0 replies; 69+ messages in thread From: Jingjing Wu @ 2017-06-27 14:00 UTC (permalink / raw) To: dev; +Cc: jingjing.wu, helin.zhang Signed-off-by: Jingjing Wu <jingjing.wu@intel.com> --- drivers/net/i40e/base/i40e_common.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) diff --git a/drivers/net/i40e/base/i40e_common.c b/drivers/net/i40e/base/i40e_common.c index 3f3b275..78ed2a8 100644 --- a/drivers/net/i40e/base/i40e_common.c +++ b/drivers/net/i40e/base/i40e_common.c @@ -330,13 +330,15 @@ void i40e_debug_aq(struct i40e_hw *hw, enum i40e_debug_mask mask, void *desc, void *buffer, u16 buf_len) { struct i40e_aq_desc *aq_desc = (struct i40e_aq_desc *)desc; - u16 len = LE16_TO_CPU(aq_desc->datalen); u8 *buf = (u8 *)buffer; + u16 len; u16 i = 0; if ((!(mask & hw->debug_mask)) || (desc == NULL)) return; + len = LE16_TO_CPU(aq_desc->datalen); + i40e_debug(hw, mask, "AQ CMD: opcode 0x%04X, flags 0x%04X, datalen 0x%04X, retval 0x%04X\n", LE16_TO_CPU(aq_desc->opcode), -- 2.4.11 ^ permalink raw reply [flat|nested] 69+ messages in thread
* [dpdk-dev] [PATCH v4 13/17] net/i40e/base: avoid reset timeout issue 2017-06-27 14:00 ` [dpdk-dev] [PATCH v4 00/17] net/i40e: base code update Jingjing Wu ` (11 preceding siblings ...) 2017-06-27 14:00 ` [dpdk-dev] [PATCH v4 12/17] net/i40e/base: avoid potential null pointer dereference Jingjing Wu @ 2017-06-27 14:00 ` Jingjing Wu 2017-06-27 14:00 ` [dpdk-dev] [PATCH v4 14/17] net/i40e/base: add EEPROM checksum verification Jingjing Wu ` (4 subsequent siblings) 17 siblings, 0 replies; 69+ messages in thread From: Jingjing Wu @ 2017-06-27 14:00 UTC (permalink / raw) To: dev; +Cc: jingjing.wu, helin.zhang This patch allows detection of upcoming core reset in case NIC gets stuck while performing FLR reset. The i40e_pf_reset() function returns I40E_ERR_NOT_READY when global reset was detected. Signed-off-by: Jingjing Wu <jingjing.wu@intel.com> --- drivers/net/i40e/base/i40e_common.c | 8 ++++++++ 1 file changed, 8 insertions(+) diff --git a/drivers/net/i40e/base/i40e_common.c b/drivers/net/i40e/base/i40e_common.c index 78ed2a8..ed2e01a 100644 --- a/drivers/net/i40e/base/i40e_common.c +++ b/drivers/net/i40e/base/i40e_common.c @@ -1382,6 +1382,8 @@ enum i40e_status_code i40e_pf_reset(struct i40e_hw *hw) * we don't need to do the PF Reset */ if (!cnt) { + u32 reg2 = 0; + reg = rd32(hw, I40E_PFGEN_CTRL); wr32(hw, I40E_PFGEN_CTRL, (reg | I40E_PFGEN_CTRL_PFSWR_MASK)); @@ -1389,6 +1391,12 @@ enum i40e_status_code i40e_pf_reset(struct i40e_hw *hw) reg = rd32(hw, I40E_PFGEN_CTRL); if (!(reg & I40E_PFGEN_CTRL_PFSWR_MASK)) break; + reg2 = rd32(hw, I40E_GLGEN_RSTAT); + if (reg2 & I40E_GLGEN_RSTAT_DEVSTATE_MASK) { + DEBUGOUT("Core reset upcoming.\n"); + DEBUGOUT1("I40E_GLGEN_RSTAT = 0x%x\n", reg2); + return I40E_ERR_NOT_READY; + } i40e_msec_delay(1); } if (reg & I40E_PFGEN_CTRL_PFSWR_MASK) { -- 2.4.11 ^ permalink raw reply [flat|nested] 69+ messages in thread
* [dpdk-dev] [PATCH v4 14/17] net/i40e/base: add EEPROM checksum verification 2017-06-27 14:00 ` [dpdk-dev] [PATCH v4 00/17] net/i40e: base code update Jingjing Wu ` (12 preceding siblings ...) 2017-06-27 14:00 ` [dpdk-dev] [PATCH v4 13/17] net/i40e/base: avoid reset timeout issue Jingjing Wu @ 2017-06-27 14:00 ` Jingjing Wu 2017-06-27 14:00 ` [dpdk-dev] [PATCH v4 15/17] net/i40e/base: extend processing of DDP Jingjing Wu ` (3 subsequent siblings) 17 siblings, 0 replies; 69+ messages in thread From: Jingjing Wu @ 2017-06-27 14:00 UTC (permalink / raw) To: dev; +Cc: jingjing.wu, helin.zhang This patch ensures PFs mutually exclusive access to NVM. Signed-off-by: Jingjing Wu <jingjing.wu@intel.com> --- drivers/net/i40e/base/i40e_nvm.c | 14 ++++++++++---- 1 file changed, 10 insertions(+), 4 deletions(-) diff --git a/drivers/net/i40e/base/i40e_nvm.c b/drivers/net/i40e/base/i40e_nvm.c index 90521fa..a1e7830 100644 --- a/drivers/net/i40e/base/i40e_nvm.c +++ b/drivers/net/i40e/base/i40e_nvm.c @@ -749,12 +749,18 @@ enum i40e_status_code i40e_validate_nvm_checksum(struct i40e_hw *hw, DEBUGFUNC("i40e_validate_nvm_checksum"); - if (hw->flags & I40E_HW_FLAG_AQ_SRCTL_ACCESS_ENABLE) - ret_code = i40e_acquire_nvm(hw, I40E_RESOURCE_READ); + /* acquire_nvm provides exclusive NVM lock to synchronize access across + * PFs. X710 uses i40e_read_nvm_word_srctl which polls for done bit + * twice (first time to be able to write address to I40E_GLNVM_SRCTL + * register, second to read data from I40E_GLNVM_SRDATA. One PF can see + * done bit and try to write address, while another one will interpret + * it as a good time to read data. It will cause invalid data to be + * read. + */ + ret_code = i40e_acquire_nvm(hw, I40E_RESOURCE_READ); if (!ret_code) { ret_code = i40e_calc_nvm_checksum(hw, &checksum_local); - if (hw->flags & I40E_HW_FLAG_AQ_SRCTL_ACCESS_ENABLE) - i40e_release_nvm(hw); + i40e_release_nvm(hw); if (ret_code != I40E_SUCCESS) goto i40e_validate_nvm_checksum_exit; } else { -- 2.4.11 ^ permalink raw reply [flat|nested] 69+ messages in thread
* [dpdk-dev] [PATCH v4 15/17] net/i40e/base: extend processing of DDP 2017-06-27 14:00 ` [dpdk-dev] [PATCH v4 00/17] net/i40e: base code update Jingjing Wu ` (13 preceding siblings ...) 2017-06-27 14:00 ` [dpdk-dev] [PATCH v4 14/17] net/i40e/base: add EEPROM checksum verification Jingjing Wu @ 2017-06-27 14:00 ` Jingjing Wu 2017-06-27 14:00 ` [dpdk-dev] [PATCH v4 16/17] net/i40e: use set switch aq instead of register setting Jingjing Wu ` (2 subsequent siblings) 17 siblings, 0 replies; 69+ messages in thread From: Jingjing Wu @ 2017-06-27 14:00 UTC (permalink / raw) To: dev; +Cc: jingjing.wu, helin.zhang, Andrey Chilikin, Beilei Xing This patch adds extended processing of DDP packages: - Execution of adminq command sections to support AQ-depended profiles, for example, for programming cloud filters types. - Ability to write a profile without registering it in the list of applied profiles, to be used for AQ-depended profiles. - Profile rollback is implemented to support restoration of original parser/analyzer configuration without the need of core reset, for example, for deploying new profile without resetting device. - Search for a specific section in a profile, to be used by driver to access metadata sections with description of PCTYPE/PTYPEs defined in the profile. Signed-off-by: Andrey Chilikin <andrey.chilikin@intel.com> Signed-off-by: Beilei Xing <beilei.xing@intel.com> Signed-off-by: Jingjing Wu <jingjing.wu@intel.com> --- drivers/net/i40e/base/i40e_common.c | 258 ++++++++++++++++++++++++++++++--- drivers/net/i40e/base/i40e_prototype.h | 6 + drivers/net/i40e/base/i40e_type.h | 25 +++- 3 files changed, 265 insertions(+), 24 deletions(-) diff --git a/drivers/net/i40e/base/i40e_common.c b/drivers/net/i40e/base/i40e_common.c index ed2e01a..900d379 100644 --- a/drivers/net/i40e/base/i40e_common.c +++ b/drivers/net/i40e/base/i40e_common.c @@ -7441,6 +7441,165 @@ i40e_find_segment_in_package(u32 segment_type, return NULL; } +/* Get section table in profile */ +#define I40E_SECTION_TABLE(profile, sec_tbl) \ + do { \ + struct i40e_profile_segment *p = (profile); \ + u32 count; \ + u32 *nvm; \ + count = p->device_table_count; \ + nvm = (u32 *)&p->device_table[count]; \ + sec_tbl = (struct i40e_section_table *)&nvm[nvm[0] + 1]; \ + } while (0) + +/* Get section header in profile */ +#define I40E_SECTION_HEADER(profile, offset) \ + (struct i40e_profile_section_header *)((u8 *)(profile) + (offset)) + +/** + * i40e_find_section_in_profile + * @section_type: the section type to search for (i.e., SECTION_TYPE_NOTE) + * @profile: pointer to the i40e segment header to be searched + * + * This function searches i40e segment for a particular section type. On + * success it returns a pointer to the section header, otherwise it will + * return NULL. + **/ +struct i40e_profile_section_header * +i40e_find_section_in_profile(u32 section_type, + struct i40e_profile_segment *profile) +{ + struct i40e_profile_section_header *sec; + struct i40e_section_table *sec_tbl; + u32 sec_off; + u32 i; + + if (profile->header.type != SEGMENT_TYPE_I40E) + return NULL; + + I40E_SECTION_TABLE(profile, sec_tbl); + + for (i = 0; i < sec_tbl->section_count; i++) { + sec_off = sec_tbl->section_offset[i]; + sec = I40E_SECTION_HEADER(profile, sec_off); + if (sec->section.type == section_type) + return sec; + } + + return NULL; +} + +/** + * i40e_ddp_exec_aq_section - Execute generic AQ for DDP + * @hw: pointer to the hw struct + * @aq: command buffer containing all data to execute AQ + **/ +STATIC enum +i40e_status_code i40e_ddp_exec_aq_section(struct i40e_hw *hw, + struct i40e_profile_aq_section *aq) +{ + enum i40e_status_code status; + struct i40e_aq_desc desc; + u8 *msg = NULL; + u16 msglen; + + i40e_fill_default_direct_cmd_desc(&desc, aq->opcode); + desc.flags |= CPU_TO_LE16(aq->flags); + i40e_memcpy(desc.params.raw, aq->param, sizeof(desc.params.raw), + I40E_NONDMA_TO_NONDMA); + + msglen = aq->datalen; + if (msglen) { + desc.flags |= CPU_TO_LE16((u16)(I40E_AQ_FLAG_BUF | + I40E_AQ_FLAG_RD)); + if (msglen > I40E_AQ_LARGE_BUF) + desc.flags |= CPU_TO_LE16((u16)I40E_AQ_FLAG_LB); + desc.datalen = CPU_TO_LE16(msglen); + msg = &aq->data[0]; + } + + status = i40e_asq_send_command(hw, &desc, msg, msglen, NULL); + + if (status != I40E_SUCCESS) { + i40e_debug(hw, I40E_DEBUG_PACKAGE, + "unable to exec DDP AQ opcode %u, error %d\n", + aq->opcode, status); + return status; + } + + /* copy returned desc to aq_buf */ + i40e_memcpy(aq->param, desc.params.raw, sizeof(desc.params.raw), + I40E_NONDMA_TO_NONDMA); + + return I40E_SUCCESS; +} + +/** + * i40e_validate_profile + * @hw: pointer to the hardware structure + * @profile: pointer to the profile segment of the package to be validated + * @track_id: package tracking id + * @rollback: flag if the profile is for rollback. + * + * Validates supported devices and profile's sections. + */ +STATIC enum i40e_status_code +i40e_validate_profile(struct i40e_hw *hw, struct i40e_profile_segment *profile, + u32 track_id, bool rollback) +{ + struct i40e_profile_section_header *sec = NULL; + enum i40e_status_code status = I40E_SUCCESS; + struct i40e_section_table *sec_tbl; + u32 vendor_dev_id; + u32 dev_cnt; + u32 sec_off; + u32 i; + + if (track_id == I40E_DDP_TRACKID_INVALID) { + i40e_debug(hw, I40E_DEBUG_PACKAGE, "Invalid track_id\n"); + return I40E_NOT_SUPPORTED; + } + + dev_cnt = profile->device_table_count; + for (i = 0; i < dev_cnt; i++) { + vendor_dev_id = profile->device_table[i].vendor_dev_id; + if ((vendor_dev_id >> 16) == I40E_INTEL_VENDOR_ID && + hw->device_id == (vendor_dev_id & 0xFFFF)) + break; + } + if (dev_cnt && (i == dev_cnt)) { + i40e_debug(hw, I40E_DEBUG_PACKAGE, + "Device doesn't support DDP\n"); + return I40E_ERR_DEVICE_NOT_SUPPORTED; + } + + I40E_SECTION_TABLE(profile, sec_tbl); + + /* Validate sections types */ + for (i = 0; i < sec_tbl->section_count; i++) { + sec_off = sec_tbl->section_offset[i]; + sec = I40E_SECTION_HEADER(profile, sec_off); + if (rollback) { + if (sec->section.type == SECTION_TYPE_MMIO || + sec->section.type == SECTION_TYPE_AQ || + sec->section.type == SECTION_TYPE_RB_AQ) { + i40e_debug(hw, I40E_DEBUG_PACKAGE, + "Not a roll-back package\n"); + return I40E_NOT_SUPPORTED; + } + } else { + if (sec->section.type == SECTION_TYPE_RB_AQ || + sec->section.type == SECTION_TYPE_RB_MMIO) { + i40e_debug(hw, I40E_DEBUG_PACKAGE, + "Not an original package\n"); + return I40E_NOT_SUPPORTED; + } + } + } + + return status; +} + /** * i40e_write_profile * @hw: pointer to the hardware structure @@ -7456,47 +7615,99 @@ i40e_write_profile(struct i40e_hw *hw, struct i40e_profile_segment *profile, enum i40e_status_code status = I40E_SUCCESS; struct i40e_section_table *sec_tbl; struct i40e_profile_section_header *sec = NULL; - u32 dev_cnt; - u32 vendor_dev_id; - u32 *nvm; + struct i40e_profile_aq_section *ddp_aq; u32 section_size = 0; u32 offset = 0, info = 0; + u32 sec_off; u32 i; - dev_cnt = profile->device_table_count; + status = i40e_validate_profile(hw, profile, track_id, false); + if (status) + return status; - for (i = 0; i < dev_cnt; i++) { - vendor_dev_id = profile->device_table[i].vendor_dev_id; - if ((vendor_dev_id >> 16) == I40E_INTEL_VENDOR_ID) - if (hw->device_id == (vendor_dev_id & 0xFFFF)) + I40E_SECTION_TABLE(profile, sec_tbl); + + for (i = 0; i < sec_tbl->section_count; i++) { + sec_off = sec_tbl->section_offset[i]; + sec = I40E_SECTION_HEADER(profile, sec_off); + /* Process generic admin command */ + if (sec->section.type == SECTION_TYPE_AQ) { + ddp_aq = (struct i40e_profile_aq_section *)&sec[1]; + status = i40e_ddp_exec_aq_section(hw, ddp_aq); + if (status) { + i40e_debug(hw, I40E_DEBUG_PACKAGE, + "Failed to execute aq: section %d, opcode %u\n", + i, ddp_aq->opcode); break; + } + sec->section.type = SECTION_TYPE_RB_AQ; + } + + /* Skip any non-mmio sections */ + if (sec->section.type != SECTION_TYPE_MMIO) + continue; + + section_size = sec->section.size + + sizeof(struct i40e_profile_section_header); + + /* Write MMIO section */ + status = i40e_aq_write_ddp(hw, (void *)sec, (u16)section_size, + track_id, &offset, &info, NULL); + if (status) { + i40e_debug(hw, I40E_DEBUG_PACKAGE, + "Failed to write profile: section %d, offset %d, info %d\n", + i, offset, info); + break; + } } - if (i == dev_cnt) { - i40e_debug(hw, I40E_DEBUG_PACKAGE, "Device doesn't support DDP"); - return I40E_ERR_DEVICE_NOT_SUPPORTED; - } + return status; +} + +/** + * i40e_rollback_profile + * @hw: pointer to the hardware structure + * @profile: pointer to the profile segment of the package to be removed + * @track_id: package tracking id + * + * Rolls back previously loaded package. + */ +enum i40e_status_code +i40e_rollback_profile(struct i40e_hw *hw, struct i40e_profile_segment *profile, + u32 track_id) +{ + struct i40e_profile_section_header *sec = NULL; + enum i40e_status_code status = I40E_SUCCESS; + struct i40e_section_table *sec_tbl; + u32 offset = 0, info = 0; + u32 section_size = 0; + u32 sec_off; + int i; - nvm = (u32 *)&profile->device_table[dev_cnt]; - sec_tbl = (struct i40e_section_table *)&nvm[nvm[0] + 1]; + status = i40e_validate_profile(hw, profile, track_id, true); + if (status) + return status; - for (i = 0; i < sec_tbl->section_count; i++) { - sec = (struct i40e_profile_section_header *)((u8 *)profile + - sec_tbl->section_offset[i]); + I40E_SECTION_TABLE(profile, sec_tbl); - /* Skip 'AQ', 'note' and 'name' sections */ - if (sec->section.type != SECTION_TYPE_MMIO) + /* For rollback write sections in reverse */ + for (i = sec_tbl->section_count - 1; i >= 0; i--) { + sec_off = sec_tbl->section_offset[i]; + sec = I40E_SECTION_HEADER(profile, sec_off); + + /* Skip any non-rollback sections */ + if (sec->section.type != SECTION_TYPE_RB_MMIO) continue; section_size = sec->section.size + sizeof(struct i40e_profile_section_header); - /* Write profile */ + /* Write roll-back MMIO section */ status = i40e_aq_write_ddp(hw, (void *)sec, (u16)section_size, track_id, &offset, &info, NULL); if (status) { i40e_debug(hw, I40E_DEBUG_PACKAGE, - "Failed to write profile: offset %d, info %d", - offset, info); + "Failed to write profile: section %d, offset %d, info %d\n", + i, offset, info); break; } } @@ -7534,7 +7745,8 @@ i40e_add_pinfo_to_list(struct i40e_hw *hw, pinfo->track_id = track_id; pinfo->version = profile->version; pinfo->op = I40E_DDP_ADD_TRACKID; - memcpy(pinfo->name, profile->name, I40E_DDP_NAME_SIZE); + i40e_memcpy(pinfo->name, profile->name, I40E_DDP_NAME_SIZE, + I40E_NONDMA_TO_NONDMA); status = i40e_aq_write_ddp(hw, (void *)sec, sec->data_end, track_id, &offset, &info, NULL); diff --git a/drivers/net/i40e/base/i40e_prototype.h b/drivers/net/i40e/base/i40e_prototype.h index 9171e97..acb2023 100644 --- a/drivers/net/i40e/base/i40e_prototype.h +++ b/drivers/net/i40e/base/i40e_prototype.h @@ -586,10 +586,16 @@ enum i40e_status_code i40e_aq_get_ddp_list(struct i40e_hw *hw, void *buff, struct i40e_generic_seg_header * i40e_find_segment_in_package(u32 segment_type, struct i40e_package_header *pkg_header); +struct i40e_profile_section_header * +i40e_find_section_in_profile(u32 section_type, + struct i40e_profile_segment *profile); enum i40e_status_code i40e_write_profile(struct i40e_hw *hw, struct i40e_profile_segment *i40e_seg, u32 track_id); enum i40e_status_code +i40e_rollback_profile(struct i40e_hw *hw, struct i40e_profile_segment *i40e_seg, + u32 track_id); +enum i40e_status_code i40e_add_pinfo_to_list(struct i40e_hw *hw, struct i40e_profile_segment *profile, u8 *profile_info_sec, u32 track_id); diff --git a/drivers/net/i40e/base/i40e_type.h b/drivers/net/i40e/base/i40e_type.h index 9d7b1a2..dca725a 100644 --- a/drivers/net/i40e/base/i40e_type.h +++ b/drivers/net/i40e/base/i40e_type.h @@ -1918,8 +1918,10 @@ struct i40e_generic_seg_header { struct i40e_metadata_segment { struct i40e_generic_seg_header header; struct i40e_ddp_version version; +#define I40E_DDP_TRACKID_RDONLY 0 +#define I40E_DDP_TRACKID_INVALID 0xFFFFFFFF u32 track_id; - char name[I40E_DDP_NAME_SIZE]; + char name[I40E_DDP_NAME_SIZE]; }; struct i40e_device_id_entry { @@ -1946,15 +1948,36 @@ struct i40e_profile_section_header { struct { #define SECTION_TYPE_INFO 0x00000010 #define SECTION_TYPE_MMIO 0x00000800 +#define SECTION_TYPE_RB_MMIO 0x00001800 #define SECTION_TYPE_AQ 0x00000801 +#define SECTION_TYPE_RB_AQ 0x00001801 #define SECTION_TYPE_NOTE 0x80000000 #define SECTION_TYPE_NAME 0x80000001 +#define SECTION_TYPE_PROTO 0x80000002 +#define SECTION_TYPE_PCTYPE 0x80000003 +#define SECTION_TYPE_PTYPE 0x80000004 u32 type; u32 offset; u32 size; } section; }; +struct i40e_profile_tlv_section_record { + u8 rtype; + u8 type; + u16 len; + u8 data[12]; +}; + +/* Generic AQ section in proflie */ +struct i40e_profile_aq_section { + u16 opcode; + u16 flags; + u8 param[16]; + u16 datalen; + u8 data[1]; +}; + struct i40e_profile_info { u32 track_id; struct i40e_ddp_version version; -- 2.4.11 ^ permalink raw reply [flat|nested] 69+ messages in thread
* [dpdk-dev] [PATCH v4 16/17] net/i40e: use set switch aq instead of register setting 2017-06-27 14:00 ` [dpdk-dev] [PATCH v4 00/17] net/i40e: base code update Jingjing Wu ` (14 preceding siblings ...) 2017-06-27 14:00 ` [dpdk-dev] [PATCH v4 15/17] net/i40e/base: extend processing of DDP Jingjing Wu @ 2017-06-27 14:00 ` Jingjing Wu 2017-06-27 14:00 ` [dpdk-dev] [PATCH v4 17/17] net/i40e/base: update base code info Jingjing Wu 2017-06-27 14:25 ` [dpdk-dev] [PATCH v4 00/17] net/i40e: base code update Ferruh Yigit 17 siblings, 0 replies; 69+ messages in thread From: Jingjing Wu @ 2017-06-27 14:00 UTC (permalink / raw) To: dev; +Cc: jingjing.wu, helin.zhang TPID can be set by set_switch_config aq, change the TPID setting by set_switch_config on new FW release. Signed-off-by: Jingjing Wu <jingjing.wu@intel.com> --- drivers/net/i40e/i40e_ethdev.c | 102 +++++++++++++++++++++++++---------------- 1 file changed, 62 insertions(+), 40 deletions(-) diff --git a/drivers/net/i40e/i40e_ethdev.c b/drivers/net/i40e/i40e_ethdev.c index 4ee1113..b1706a0 100644 --- a/drivers/net/i40e/i40e_ethdev.c +++ b/drivers/net/i40e/i40e_ethdev.c @@ -2973,71 +2973,93 @@ i40e_vlan_filter_set(struct rte_eth_dev *dev, uint16_t vlan_id, int on) } static int -i40e_vlan_tpid_set(struct rte_eth_dev *dev, - enum rte_vlan_type vlan_type, - uint16_t tpid) +i40e_vlan_tpid_set_by_registers(struct rte_eth_dev *dev, + enum rte_vlan_type vlan_type, + uint16_t tpid, int qinq) { struct i40e_hw *hw = I40E_DEV_PRIVATE_TO_HW(dev->data->dev_private); - uint64_t reg_r = 0, reg_w = 0; - uint16_t reg_id = 0; - int ret = 0; - int qinq = dev->data->dev_conf.rxmode.hw_vlan_extend; + uint64_t reg_r = 0; + uint64_t reg_w = 0; + uint16_t reg_id = 3; + int ret; - switch (vlan_type) { - case ETH_VLAN_TYPE_OUTER: - if (qinq) + if (qinq) { + if (vlan_type == ETH_VLAN_TYPE_OUTER) reg_id = 2; - else - reg_id = 3; - break; - case ETH_VLAN_TYPE_INNER: - if (qinq) - reg_id = 3; - else { - ret = -EINVAL; - PMD_DRV_LOG(ERR, - "Unsupported vlan type in single vlan."); - return ret; - } - break; - default: - ret = -EINVAL; - PMD_DRV_LOG(ERR, "Unsupported vlan type %d", vlan_type); - return ret; } + ret = i40e_aq_debug_read_register(hw, I40E_GL_SWT_L2TAGCTRL(reg_id), ®_r, NULL); if (ret != I40E_SUCCESS) { PMD_DRV_LOG(ERR, "Fail to debug read from I40E_GL_SWT_L2TAGCTRL[%d]", reg_id); - ret = -EIO; - return ret; + return -EIO; } PMD_DRV_LOG(DEBUG, - "Debug read from I40E_GL_SWT_L2TAGCTRL[%d]: 0x%08"PRIx64, - reg_id, reg_r); + "Debug read from I40E_GL_SWT_L2TAGCTRL[%d]: 0x%08"PRIx64, + reg_id, reg_r); reg_w = reg_r & (~(I40E_GL_SWT_L2TAGCTRL_ETHERTYPE_MASK)); reg_w |= ((uint64_t)tpid << I40E_GL_SWT_L2TAGCTRL_ETHERTYPE_SHIFT); if (reg_r == reg_w) { - ret = 0; PMD_DRV_LOG(DEBUG, "No need to write"); - return ret; + return 0; } ret = i40e_aq_debug_write_register(hw, I40E_GL_SWT_L2TAGCTRL(reg_id), reg_w, NULL); if (ret != I40E_SUCCESS) { - ret = -EIO; PMD_DRV_LOG(ERR, - "Fail to debug write to I40E_GL_SWT_L2TAGCTRL[%d]", - reg_id); - return ret; + "Fail to debug write to I40E_GL_SWT_L2TAGCTRL[%d]", + reg_id); + return -EIO; } PMD_DRV_LOG(DEBUG, - "Debug write 0x%08"PRIx64" to I40E_GL_SWT_L2TAGCTRL[%d]", - reg_w, reg_id); + "Debug write 0x%08"PRIx64" to I40E_GL_SWT_L2TAGCTRL[%d]", + reg_w, reg_id); + + return 0; +} + +static int +i40e_vlan_tpid_set(struct rte_eth_dev *dev, + enum rte_vlan_type vlan_type, + uint16_t tpid) +{ + struct i40e_hw *hw = I40E_DEV_PRIVATE_TO_HW(dev->data->dev_private); + int qinq = dev->data->dev_conf.rxmode.hw_vlan_extend; + int ret = 0; + + if ((vlan_type != ETH_VLAN_TYPE_INNER && + vlan_type != ETH_VLAN_TYPE_OUTER) || + (!qinq && vlan_type == ETH_VLAN_TYPE_INNER)) { + PMD_DRV_LOG(ERR, + "Unsupported vlan type."); + return -EINVAL; + } + /* 802.1ad frames ability is added in NVM API 1.7*/ + if (hw->flags & I40E_HW_FLAG_802_1AD_CAPABLE) { + if (qinq) { + if (vlan_type == ETH_VLAN_TYPE_OUTER) + hw->first_tag = rte_cpu_to_le_16(tpid); + else if (vlan_type == ETH_VLAN_TYPE_INNER) + hw->second_tag = rte_cpu_to_le_16(tpid); + } else { + if (vlan_type == ETH_VLAN_TYPE_OUTER) + hw->second_tag = rte_cpu_to_le_16(tpid); + } + ret = i40e_aq_set_switch_config(hw, 0, 0, NULL); + if (ret != I40E_SUCCESS) { + PMD_DRV_LOG(ERR, + "Set switch config failed aq_err: %d", + hw->aq.asq_last_status); + ret = -EIO; + } + } else + /* If NVM API < 1.7, keep the register setting */ + ret = i40e_vlan_tpid_set_by_registers(dev, vlan_type, + tpid, qinq); return ret; } @@ -3066,7 +3088,7 @@ i40e_vlan_offload_set(struct rte_eth_dev *dev, int mask) if (mask & ETH_VLAN_EXTEND_MASK) { if (dev->data->dev_conf.rxmode.hw_vlan_extend) { i40e_vsi_config_double_vlan(vsi, TRUE); - /* Set global registers with default ether type value */ + /* Set global registers with default ethertype. */ i40e_vlan_tpid_set(dev, ETH_VLAN_TYPE_OUTER, ETHER_TYPE_VLAN); i40e_vlan_tpid_set(dev, ETH_VLAN_TYPE_INNER, -- 2.4.11 ^ permalink raw reply [flat|nested] 69+ messages in thread
* [dpdk-dev] [PATCH v4 17/17] net/i40e/base: update base code info 2017-06-27 14:00 ` [dpdk-dev] [PATCH v4 00/17] net/i40e: base code update Jingjing Wu ` (15 preceding siblings ...) 2017-06-27 14:00 ` [dpdk-dev] [PATCH v4 16/17] net/i40e: use set switch aq instead of register setting Jingjing Wu @ 2017-06-27 14:00 ` Jingjing Wu 2017-06-27 14:25 ` [dpdk-dev] [PATCH v4 00/17] net/i40e: base code update Ferruh Yigit 17 siblings, 0 replies; 69+ messages in thread From: Jingjing Wu @ 2017-06-27 14:00 UTC (permalink / raw) To: dev; +Cc: jingjing.wu, helin.zhang Signed-off-by: Jingjing Wu <jingjing.wu@intel.com> --- drivers/net/i40e/base/README | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/net/i40e/base/README b/drivers/net/i40e/base/README index 0da9f67..59e76c2 100644 --- a/drivers/net/i40e/base/README +++ b/drivers/net/i40e/base/README @@ -34,7 +34,7 @@ Intel® I40E driver ================== This directory contains source code of FreeBSD i40e driver of version -cid-i40e.2017.03.21.tar.gz released by the team which develops +cid-i40e.2017.06.23.tar.gz released by the team which develops basic drivers for any i40e NIC. The directory of base/ contains the original source package. This driver is valid for the product(s) listed below -- 2.4.11 ^ permalink raw reply [flat|nested] 69+ messages in thread
* Re: [dpdk-dev] [PATCH v4 00/17] net/i40e: base code update 2017-06-27 14:00 ` [dpdk-dev] [PATCH v4 00/17] net/i40e: base code update Jingjing Wu ` (16 preceding siblings ...) 2017-06-27 14:00 ` [dpdk-dev] [PATCH v4 17/17] net/i40e/base: update base code info Jingjing Wu @ 2017-06-27 14:25 ` Ferruh Yigit 17 siblings, 0 replies; 69+ messages in thread From: Ferruh Yigit @ 2017-06-27 14:25 UTC (permalink / raw) To: Jingjing Wu, dev; +Cc: helin.zhang On 6/27/2017 3:00 PM, Jingjing Wu wrote: > v4 changes: > - fix more checkpatch warnings > > v3 changes: > - extra patch to update README > - remove unnecessary code > - fix checkpatch warning > > v2 changes: > - fix compile issue > - update from package from 2017.05.24 to 2017.06.23 > > i40e base code upate. The main changes are: > - use virtchnl.h instead of i40e_virtchnl.h > - add support for Adaptive Virtual Function > - add new AQ commands for read/write PHY registers > - add new phy types for 25G > - extend processing of DDP > > Jingjing Wu (17): > net/i40e/base: use new virtchnl header file > net/i40e/base: sync nvmupdate command and adminq subtask > net/i40e/base: add AQ command for read/write PHY registers > net/i40e/base: add support for Adaptive Virtual Function > net/i40e/base: store the requested FEC information > net/i40e/base: add new phy types for 25G AOC and ACC > net/i40e/base: report supported link modes > net/i40e/base: track id can be 0 > net/i40e/base: update FW AQ API version to 1.7 > net/i40e/base: add support for switch parameters > net/i40e/base: use admin queue for setting LEDs behavior > net/i40e/base: avoid potential null pointer dereference > net/i40e/base: avoid reset timeout issue > net/i40e/base: add EEPROM checksum verification > net/i40e/base: extend processing of DDP > net/i40e: use set switch aq instead of register setting > net/i40e/base: update base code info Series applied to dpdk-next-net/master, thanks. ^ permalink raw reply [flat|nested] 69+ messages in thread
end of thread, other threads:[~2017-06-27 14:25 UTC | newest] Thread overview: 69+ messages (download: mbox.gz / follow: Atom feed) -- links below jump to the message on this page -- 2017-05-27 3:47 [dpdk-dev] [PATCH 00/10] net/i40e: base code update Jingjing Wu 2017-05-27 3:47 ` [dpdk-dev] [PATCH 01/10] net/i40e/base: use new virtchnl header file Jingjing Wu 2017-05-27 3:47 ` [dpdk-dev] [PATCH 02/10] net/i40e/base: sync nvmupdate command and adminq subtask Jingjing Wu 2017-05-27 3:47 ` [dpdk-dev] [PATCH 03/10] net/i40e/base: add AQ command for read/write PHY registers Jingjing Wu 2017-05-27 3:47 ` [dpdk-dev] [PATCH 04/10] net/i40e/base: add support for Adaptive Virtual Function Jingjing Wu 2017-05-27 3:47 ` [dpdk-dev] [PATCH 05/10] net/i40e/base: store the requested FEC information Jingjing Wu 2017-05-27 3:47 ` [dpdk-dev] [PATCH 06/10] net/i40e/base: add new phy types for 25G AOC and ACC Jingjing Wu 2017-05-27 3:47 ` [dpdk-dev] [PATCH 07/10] net/i40e/base: report supported link modes Jingjing Wu 2017-05-27 3:47 ` [dpdk-dev] [PATCH 08/10] net/i40e/base: track id can be 0 Jingjing Wu 2017-05-27 3:47 ` [dpdk-dev] [PATCH 09/10] net/i40e/base: update FW AQ API version to 1.7 Jingjing Wu 2017-05-27 3:47 ` [dpdk-dev] [PATCH 10/10] net/i40e/base: update base code info Jingjing Wu 2017-05-30 11:10 ` [dpdk-dev] [PATCH 00/10] net/i40e: base code update Ferruh Yigit 2017-06-02 2:24 ` Wu, Jingjing 2017-06-26 9:48 ` [dpdk-dev] [PATCH v2 00/16] " Jingjing Wu 2017-06-26 9:48 ` [dpdk-dev] [PATCH v2 01/16] net/i40e/base: use new virtchnl header file Jingjing Wu 2017-06-26 9:48 ` [dpdk-dev] [PATCH v2 02/16] net/i40e/base: sync nvmupdate command and adminq subtask Jingjing Wu 2017-06-26 9:48 ` [dpdk-dev] [PATCH v2 03/16] net/i40e/base: add AQ command for read/write PHY registers Jingjing Wu 2017-06-26 9:48 ` [dpdk-dev] [PATCH v2 04/16] net/i40e/base: add support for Adaptive Virtual Function Jingjing Wu 2017-06-26 9:48 ` [dpdk-dev] [PATCH v2 05/16] net/i40e/base: store the requested FEC information Jingjing Wu 2017-06-26 9:48 ` [dpdk-dev] [PATCH v2 06/16] net/i40e/base: add new phy types for 25G AOC and ACC Jingjing Wu 2017-06-26 9:48 ` [dpdk-dev] [PATCH v2 07/16] net/i40e/base: report supported link modes Jingjing Wu 2017-06-26 9:48 ` [dpdk-dev] [PATCH v2 08/16] net/i40e/base: track id can be 0 Jingjing Wu 2017-06-26 9:48 ` [dpdk-dev] [PATCH v2 09/16] net/i40e/base: update FW AQ API version to 1.7 Jingjing Wu 2017-06-26 9:48 ` [dpdk-dev] [PATCH v2 10/16] net/i40e/base: add support for switch parameters Jingjing Wu 2017-06-26 9:48 ` [dpdk-dev] [PATCH v2 11/16] net/i40e/base: use admin queue for setting LEDs behavior Jingjing Wu 2017-06-26 9:48 ` [dpdk-dev] [PATCH v2 12/16] net/i40e/base: avoid potential null pointer dereference Jingjing Wu 2017-06-26 9:48 ` [dpdk-dev] [PATCH v2 13/16] net/i40e/base: avoid reset timeout issue Jingjing Wu 2017-06-26 9:48 ` [dpdk-dev] [PATCH v2 14/16] net/i40e/base: add EEPROM checksum verification Jingjing Wu 2017-06-26 9:48 ` [dpdk-dev] [PATCH v2 15/16] net/i40e/base: extend processing of DDP Jingjing Wu 2017-06-26 13:18 ` Ferruh Yigit 2017-06-26 9:48 ` [dpdk-dev] [PATCH v2 16/16] net/i40e: use set switch aq instead of register setting Jingjing Wu 2017-06-26 13:20 ` Ferruh Yigit 2017-06-27 13:29 ` [dpdk-dev] [PATCH v3 00/17] net/i40e: base code update Jingjing Wu 2017-06-27 13:29 ` [dpdk-dev] [PATCH v3 01/17] net/i40e/base: use new virtchnl header file Jingjing Wu 2017-06-27 13:29 ` [dpdk-dev] [PATCH v3 02/17] net/i40e/base: sync nvmupdate command and adminq subtask Jingjing Wu 2017-06-27 13:29 ` [dpdk-dev] [PATCH v3 03/17] net/i40e/base: add AQ command for read/write PHY registers Jingjing Wu 2017-06-27 13:29 ` [dpdk-dev] [PATCH v3 04/17] net/i40e/base: add support for Adaptive Virtual Function Jingjing Wu 2017-06-27 13:29 ` [dpdk-dev] [PATCH v3 05/17] net/i40e/base: store the requested FEC information Jingjing Wu 2017-06-27 13:29 ` [dpdk-dev] [PATCH v3 06/17] net/i40e/base: add new phy types for 25G AOC and ACC Jingjing Wu 2017-06-27 13:29 ` [dpdk-dev] [PATCH v3 07/17] net/i40e/base: report supported link modes Jingjing Wu 2017-06-27 13:29 ` [dpdk-dev] [PATCH v3 08/17] net/i40e/base: track id can be 0 Jingjing Wu 2017-06-27 13:29 ` [dpdk-dev] [PATCH v3 09/17] net/i40e/base: update FW AQ API version to 1.7 Jingjing Wu 2017-06-27 13:29 ` [dpdk-dev] [PATCH v3 10/17] net/i40e/base: add support for switch parameters Jingjing Wu 2017-06-27 13:29 ` [dpdk-dev] [PATCH v3 11/17] net/i40e/base: use admin queue for setting LEDs behavior Jingjing Wu 2017-06-27 13:29 ` [dpdk-dev] [PATCH v3 12/17] net/i40e/base: avoid potential null pointer dereference Jingjing Wu 2017-06-27 13:29 ` [dpdk-dev] [PATCH v3 13/17] net/i40e/base: avoid reset timeout issue Jingjing Wu 2017-06-27 13:29 ` [dpdk-dev] [PATCH v3 14/17] net/i40e/base: add EEPROM checksum verification Jingjing Wu 2017-06-27 13:29 ` [dpdk-dev] [PATCH v3 15/17] net/i40e/base: extend processing of DDP Jingjing Wu 2017-06-27 13:29 ` [dpdk-dev] [PATCH v3 16/17] net/i40e: use set switch aq instead of register setting Jingjing Wu 2017-06-27 13:29 ` [dpdk-dev] [PATCH v3 17/17] net/i40e/base: update base code info Jingjing Wu 2017-06-27 14:00 ` [dpdk-dev] [PATCH v4 00/17] net/i40e: base code update Jingjing Wu 2017-06-27 14:00 ` [dpdk-dev] [PATCH v4 01/17] net/i40e/base: use new virtchnl header file Jingjing Wu 2017-06-27 14:00 ` [dpdk-dev] [PATCH v4 02/17] net/i40e/base: sync nvmupdate command and adminq subtask Jingjing Wu 2017-06-27 14:00 ` [dpdk-dev] [PATCH v4 03/17] net/i40e/base: add AQ command for read/write PHY registers Jingjing Wu 2017-06-27 14:00 ` [dpdk-dev] [PATCH v4 04/17] net/i40e/base: add support for Adaptive Virtual Function Jingjing Wu 2017-06-27 14:00 ` [dpdk-dev] [PATCH v4 05/17] net/i40e/base: store the requested FEC information Jingjing Wu 2017-06-27 14:00 ` [dpdk-dev] [PATCH v4 06/17] net/i40e/base: add new phy types for 25G AOC and ACC Jingjing Wu 2017-06-27 14:00 ` [dpdk-dev] [PATCH v4 07/17] net/i40e/base: report supported link modes Jingjing Wu 2017-06-27 14:00 ` [dpdk-dev] [PATCH v4 08/17] net/i40e/base: track id can be 0 Jingjing Wu 2017-06-27 14:00 ` [dpdk-dev] [PATCH v4 09/17] net/i40e/base: update FW AQ API version to 1.7 Jingjing Wu 2017-06-27 14:00 ` [dpdk-dev] [PATCH v4 10/17] net/i40e/base: add support for switch parameters Jingjing Wu 2017-06-27 14:00 ` [dpdk-dev] [PATCH v4 11/17] net/i40e/base: use admin queue for setting LEDs behavior Jingjing Wu 2017-06-27 14:00 ` [dpdk-dev] [PATCH v4 12/17] net/i40e/base: avoid potential null pointer dereference Jingjing Wu 2017-06-27 14:00 ` [dpdk-dev] [PATCH v4 13/17] net/i40e/base: avoid reset timeout issue Jingjing Wu 2017-06-27 14:00 ` [dpdk-dev] [PATCH v4 14/17] net/i40e/base: add EEPROM checksum verification Jingjing Wu 2017-06-27 14:00 ` [dpdk-dev] [PATCH v4 15/17] net/i40e/base: extend processing of DDP Jingjing Wu 2017-06-27 14:00 ` [dpdk-dev] [PATCH v4 16/17] net/i40e: use set switch aq instead of register setting Jingjing Wu 2017-06-27 14:00 ` [dpdk-dev] [PATCH v4 17/17] net/i40e/base: update base code info Jingjing Wu 2017-06-27 14:25 ` [dpdk-dev] [PATCH v4 00/17] net/i40e: base code update Ferruh Yigit
This is a public inbox, see mirroring instructions for how to clone and mirror all data and code used for this inbox; as well as URLs for NNTP newsgroup(s).