* [dpdk-dev] [PATCH v2 00/17] Add advanced features for Huawei hinic pmd
@ 2019-09-25 14:26 Xiaoyun wang
2019-09-25 14:30 ` Xiaoyun wang
` (18 more replies)
0 siblings, 19 replies; 32+ messages in thread
From: Xiaoyun wang @ 2019-09-25 14:26 UTC (permalink / raw)
To: ferruh.yigit
Cc: dev, xuanziyang2, shahar.belkar, luoxianjun, tanya.brokhman,
zhouguoyang, wulike1, Xiaoyun wang
This patch set adds advanced features for Huawei hinic pmd,
such as VLAN filter and VLAN offload, SR-IOV, FW version get,
set link down and up, Flow director for LACP, VRRP, BGP and so on.
Xiaoyun wang (17):
net/hinic/base: add mbox command channel for SRIOV
net/hinic/base: add HW interfaces for SRIOV
net/hinic: add VF PMD operation interfaces
net/hinic: add VLAN filter and offload
net/hinic: add allmulticast mode and MTU set
net/hinic: add unicast and multicast MAC set
net/hinic: add fdir config interface
net/hinic: add fdir validate flow operations
net/hinic: create and destroy ntuple filter
net/hinic: create and destroy fdir filter
net/hinic: flush fdir filter
net/hinic: set link down and up
net/hinic: support inner L3 checksum offload
net/hinic: support LRO offload
net/hinic: add hinic PMD doc files
net/hinic/base: optimize aeq interfaces
net/hinic: optimize tx&rx performance
doc/guides/nics/features/hinic.ini | 12 +-
doc/guides/nics/hinic.rst | 6 +
doc/guides/rel_notes/release_19_11.rst | 45 +-
drivers/net/hinic/Makefile | 2 +
drivers/net/hinic/base/hinic_compat.h | 62 +-
drivers/net/hinic/base/hinic_csr.h | 29 +-
drivers/net/hinic/base/hinic_pmd_api_cmd.c | 60 +-
drivers/net/hinic/base/hinic_pmd_cfg.c | 35 +
drivers/net/hinic/base/hinic_pmd_cmd.h | 26 +-
drivers/net/hinic/base/hinic_pmd_eqs.c | 245 +--
drivers/net/hinic/base/hinic_pmd_eqs.h | 5 +-
drivers/net/hinic/base/hinic_pmd_hwdev.c | 198 ++-
drivers/net/hinic/base/hinic_pmd_hwdev.h | 9 +-
drivers/net/hinic/base/hinic_pmd_hwif.c | 85 +-
drivers/net/hinic/base/hinic_pmd_hwif.h | 15 +-
drivers/net/hinic/base/hinic_pmd_mbox.c | 938 +++++++++++
drivers/net/hinic/base/hinic_pmd_mbox.h | 93 ++
drivers/net/hinic/base/hinic_pmd_mgmt.c | 83 +-
drivers/net/hinic/base/hinic_pmd_mgmt.h | 2 +-
drivers/net/hinic/base/hinic_pmd_niccfg.c | 777 ++++++++-
drivers/net/hinic/base/hinic_pmd_niccfg.h | 203 +++
drivers/net/hinic/base/hinic_pmd_nicio.c | 15 +-
drivers/net/hinic/base/hinic_pmd_nicio.h | 3 +-
drivers/net/hinic/base/meson.build | 1 +
drivers/net/hinic/hinic_pmd_ethdev.c | 992 ++++++++++--
drivers/net/hinic/hinic_pmd_ethdev.h | 139 +-
drivers/net/hinic/hinic_pmd_flow.c | 2385 ++++++++++++++++++++++++++++
drivers/net/hinic/hinic_pmd_rx.c | 27 +-
drivers/net/hinic/hinic_pmd_rx.h | 11 +
drivers/net/hinic/hinic_pmd_tx.c | 190 ++-
drivers/net/hinic/meson.build | 1 +
31 files changed, 6029 insertions(+), 665 deletions(-)
create mode 100644 drivers/net/hinic/base/hinic_pmd_mbox.c
create mode 100644 drivers/net/hinic/base/hinic_pmd_mbox.h
create mode 100644 drivers/net/hinic/hinic_pmd_flow.c
--
1.8.3.1
^ permalink raw reply [flat|nested] 32+ messages in thread
* [dpdk-dev] [PATCH v2 00/17] Add advanced features for Huawei hinic pmd
2019-09-25 14:26 [dpdk-dev] [PATCH v2 00/17] Add advanced features for Huawei hinic pmd Xiaoyun wang
@ 2019-09-25 14:30 ` Xiaoyun wang
2019-09-25 14:30 ` [dpdk-dev] [PATCH v2 01/17] net/hinic/base: add mbox command channel for SRIOV Xiaoyun wang
` (17 subsequent siblings)
18 siblings, 0 replies; 32+ messages in thread
From: Xiaoyun wang @ 2019-09-25 14:30 UTC (permalink / raw)
To: ferruh.yigit
Cc: dev, xuanziyang2, shahar.belkar, luoxianjun, tanya.brokhman,
zhouguoyang, wulike1, Xiaoyun wang
This patch set adds advanced features for Huawei hinic pmd,
such as VLAN filter and VLAN offload, SR-IOV, FW version get,
set link down and up, Flow director for LACP, VRRP, BGP and so on.
Xiaoyun wang (17):
net/hinic/base: add mbox command channel for SRIOV
net/hinic/base: add HW interfaces for SRIOV
net/hinic: add VF PMD operation interfaces
net/hinic: add VLAN filter and offload
net/hinic: add allmulticast mode and MTU set
net/hinic: add unicast and multicast MAC set
net/hinic: add fdir config interface
net/hinic: add fdir validate flow operations
net/hinic: create and destroy ntuple filter
net/hinic: create and destroy fdir filter
net/hinic: flush fdir filter
net/hinic: set link down and up
net/hinic: support inner L3 checksum offload
net/hinic: support LRO offload
net/hinic: add hinic PMD doc files
net/hinic/base: optimize aeq interfaces
net/hinic: optimize tx&rx performance
doc/guides/nics/features/hinic.ini | 12 +-
doc/guides/nics/hinic.rst | 6 +
doc/guides/rel_notes/release_19_11.rst | 45 +-
drivers/net/hinic/Makefile | 2 +
drivers/net/hinic/base/hinic_compat.h | 62 +-
drivers/net/hinic/base/hinic_csr.h | 29 +-
drivers/net/hinic/base/hinic_pmd_api_cmd.c | 60 +-
drivers/net/hinic/base/hinic_pmd_cfg.c | 35 +
drivers/net/hinic/base/hinic_pmd_cmd.h | 26 +-
drivers/net/hinic/base/hinic_pmd_eqs.c | 245 +--
drivers/net/hinic/base/hinic_pmd_eqs.h | 5 +-
drivers/net/hinic/base/hinic_pmd_hwdev.c | 198 ++-
drivers/net/hinic/base/hinic_pmd_hwdev.h | 9 +-
drivers/net/hinic/base/hinic_pmd_hwif.c | 85 +-
drivers/net/hinic/base/hinic_pmd_hwif.h | 15 +-
drivers/net/hinic/base/hinic_pmd_mbox.c | 938 +++++++++++
drivers/net/hinic/base/hinic_pmd_mbox.h | 93 ++
drivers/net/hinic/base/hinic_pmd_mgmt.c | 83 +-
drivers/net/hinic/base/hinic_pmd_mgmt.h | 2 +-
drivers/net/hinic/base/hinic_pmd_niccfg.c | 777 ++++++++-
drivers/net/hinic/base/hinic_pmd_niccfg.h | 203 +++
drivers/net/hinic/base/hinic_pmd_nicio.c | 15 +-
drivers/net/hinic/base/hinic_pmd_nicio.h | 3 +-
drivers/net/hinic/base/meson.build | 1 +
drivers/net/hinic/hinic_pmd_ethdev.c | 992 ++++++++++--
drivers/net/hinic/hinic_pmd_ethdev.h | 139 +-
drivers/net/hinic/hinic_pmd_flow.c | 2385 ++++++++++++++++++++++++++++
drivers/net/hinic/hinic_pmd_rx.c | 27 +-
drivers/net/hinic/hinic_pmd_rx.h | 11 +
drivers/net/hinic/hinic_pmd_tx.c | 190 ++-
drivers/net/hinic/meson.build | 1 +
31 files changed, 6029 insertions(+), 665 deletions(-)
create mode 100644 drivers/net/hinic/base/hinic_pmd_mbox.c
create mode 100644 drivers/net/hinic/base/hinic_pmd_mbox.h
create mode 100644 drivers/net/hinic/hinic_pmd_flow.c
--
1.8.3.1
^ permalink raw reply [flat|nested] 32+ messages in thread
* [dpdk-dev] [PATCH v2 01/17] net/hinic/base: add mbox command channel for SRIOV
2019-09-25 14:26 [dpdk-dev] [PATCH v2 00/17] Add advanced features for Huawei hinic pmd Xiaoyun wang
2019-09-25 14:30 ` Xiaoyun wang
@ 2019-09-25 14:30 ` Xiaoyun wang
2019-09-25 14:30 ` [dpdk-dev] [PATCH v2 02/17] net/hinic/base: add HW interfaces " Xiaoyun wang
` (16 subsequent siblings)
18 siblings, 0 replies; 32+ messages in thread
From: Xiaoyun wang @ 2019-09-25 14:30 UTC (permalink / raw)
To: ferruh.yigit
Cc: dev, xuanziyang2, shahar.belkar, luoxianjun, tanya.brokhman,
zhouguoyang, wulike1, Xiaoyun wang
Add mbox command channel for SRIOV, which is used to
communicate between VF and VF, VF and PF. This patch
introduces data structures, initialization, interfaces
and commands of mbox channel.
Signed-off-by: Xiaoyun wang <cloud.wangxiaoyun@huawei.com>
---
drivers/net/hinic/Makefile | 1 +
drivers/net/hinic/base/hinic_compat.h | 34 +-
drivers/net/hinic/base/hinic_pmd_hwdev.h | 5 +-
drivers/net/hinic/base/hinic_pmd_mbox.c | 938 +++++++++++++++++++++++++++++++
drivers/net/hinic/base/hinic_pmd_mbox.h | 93 +++
drivers/net/hinic/base/meson.build | 1 +
6 files changed, 1069 insertions(+), 3 deletions(-)
create mode 100644 drivers/net/hinic/base/hinic_pmd_mbox.c
create mode 100644 drivers/net/hinic/base/hinic_pmd_mbox.h
diff --git a/drivers/net/hinic/Makefile b/drivers/net/hinic/Makefile
index 42b4a78..20a338e 100644
--- a/drivers/net/hinic/Makefile
+++ b/drivers/net/hinic/Makefile
@@ -59,6 +59,7 @@ SRCS-$(CONFIG_RTE_LIBRTE_HINIC_PMD) += hinic_pmd_mgmt.c
SRCS-$(CONFIG_RTE_LIBRTE_HINIC_PMD) += hinic_pmd_niccfg.c
SRCS-$(CONFIG_RTE_LIBRTE_HINIC_PMD) += hinic_pmd_nicio.c
SRCS-$(CONFIG_RTE_LIBRTE_HINIC_PMD) += hinic_pmd_wq.c
+SRCS-$(CONFIG_RTE_LIBRTE_HINIC_PMD) += hinic_pmd_mbox.c
SRCS-$(CONFIG_RTE_LIBRTE_HINIC_PMD) += hinic_pmd_ethdev.c
SRCS-$(CONFIG_RTE_LIBRTE_HINIC_PMD) += hinic_pmd_rx.c
diff --git a/drivers/net/hinic/base/hinic_compat.h b/drivers/net/hinic/base/hinic_compat.h
index f599947..db60a51 100644
--- a/drivers/net/hinic/base/hinic_compat.h
+++ b/drivers/net/hinic/base/hinic_compat.h
@@ -227,7 +227,7 @@ static inline u16 ilog2(u32 n)
* hinic_cpu_to_be32 - convert data to big endian 32 bit format
* @data: the data to convert
* @len: length of data to convert, must be Multiple of 4B
- **/
+ */
static inline void hinic_cpu_to_be32(void *data, u32 len)
{
u32 i;
@@ -243,7 +243,7 @@ static inline void hinic_cpu_to_be32(void *data, u32 len)
* hinic_be32_to_cpu - convert data from big endian 32 bit format
* @data: the data to convert
* @len: length of data to convert, must be Multiple of 4B
- **/
+ */
static inline void hinic_be32_to_cpu(void *data, u32 len)
{
u32 i;
@@ -278,4 +278,34 @@ static inline int hinic_mutex_destroy(pthread_mutex_t *pthreadmutex)
return err;
}
+static inline int hinic_mutex_lock(pthread_mutex_t *pthreadmutex)
+{
+ int err;
+
+ err = pthread_mutex_lock(pthreadmutex);
+ if (!err) {
+ return err;
+ } else if (err == EOWNERDEAD) {
+ PMD_DRV_LOG(ERR, "Mutex lock failed. (ErrorNo=%d)", errno);
+#if defined(__GLIBC__)
+#if __GLIBC_PREREQ(2, 12)
+ (void)pthread_mutex_consistent(pthreadmutex);
+#else
+ (void)pthread_mutex_consistent_np(pthreadmutex);
+#endif
+#else
+ (void)pthread_mutex_consistent(pthreadmutex);
+#endif
+ } else {
+ PMD_DRV_LOG(ERR, "Mutex lock failed. (ErrorNo=%d)", errno);
+ }
+
+ return err;
+}
+
+static inline int hinic_mutex_unlock(pthread_mutex_t *pthreadmutex)
+{
+ return pthread_mutex_unlock(pthreadmutex);
+}
+
#endif /* _HINIC_COMPAT_H_ */
diff --git a/drivers/net/hinic/base/hinic_pmd_hwdev.h b/drivers/net/hinic/base/hinic_pmd_hwdev.h
index 6c21c47..fc5bdf9 100644
--- a/drivers/net/hinic/base/hinic_pmd_hwdev.h
+++ b/drivers/net/hinic/base/hinic_pmd_hwdev.h
@@ -7,13 +7,15 @@
#include "hinic_pmd_cmd.h"
-#define HINIC_PAGE_SIZE_MAX 20
+#define HINIC_PAGE_SIZE_MAX 20
#define HINIC_MGMT_CMD_UNSUPPORTED 0xFF
#define HINIC_PF_SET_VF_ALREADY 0x4
#define MAX_PCIE_DFX_BUF_SIZE 1024
+#define HINIC_DEV_BUSY_ACTIVE_FW 0xFE
+
/* dma pool */
struct dma_pool {
u32 inuse;
@@ -436,6 +438,7 @@ struct hinic_hwdev {
struct hinic_hwif *hwif;
struct cfg_mgmt_info *cfg_mgmt;
struct hinic_aeqs *aeqs;
+ struct hinic_mbox_func_to_func *func_to_func;
struct hinic_msg_pf_to_mgmt *pf_to_mgmt;
struct hinic_cmdqs *cmdqs;
struct hinic_nic_io *nic_io;
diff --git a/drivers/net/hinic/base/hinic_pmd_mbox.c b/drivers/net/hinic/base/hinic_pmd_mbox.c
new file mode 100644
index 0000000..7ae8fd0
--- /dev/null
+++ b/drivers/net/hinic/base/hinic_pmd_mbox.c
@@ -0,0 +1,938 @@
+/* SPDX-License-Identifier: BSD-3-Clause
+ * Copyright(c) 2017 Huawei Technologies Co., Ltd
+ */
+
+#include "hinic_compat.h"
+#include "hinic_csr.h"
+#include "hinic_pmd_hwdev.h"
+#include "hinic_pmd_hwif.h"
+#include "hinic_pmd_eqs.h"
+#include "hinic_pmd_mgmt.h"
+#include "hinic_pmd_mbox.h"
+
+#define HINIC_MBOX_INT_DST_FUNC_SHIFT 0
+#define HINIC_MBOX_INT_DST_AEQN_SHIFT 10
+#define HINIC_MBOX_INT_SRC_RESP_AEQN_SHIFT 12
+#define HINIC_MBOX_INT_STAT_DMA_SHIFT 14
+/* The size of data to be send (unit of 4 bytes) */
+#define HINIC_MBOX_INT_TX_SIZE_SHIFT 20
+/* SO_RO(strong order, relax order) */
+#define HINIC_MBOX_INT_STAT_DMA_SO_RO_SHIFT 25
+#define HINIC_MBOX_INT_WB_EN_SHIFT 28
+
+
+#define HINIC_MBOX_INT_DST_FUNC_MASK 0x3FF
+#define HINIC_MBOX_INT_DST_AEQN_MASK 0x3
+#define HINIC_MBOX_INT_SRC_RESP_AEQN_MASK 0x3
+#define HINIC_MBOX_INT_STAT_DMA_MASK 0x3F
+#define HINIC_MBOX_INT_TX_SIZE_MASK 0x1F
+#define HINIC_MBOX_INT_STAT_DMA_SO_RO_MASK 0x3
+#define HINIC_MBOX_INT_WB_EN_MASK 0x1
+
+#define HINIC_MBOX_INT_SET(val, field) \
+ (((val) & HINIC_MBOX_INT_##field##_MASK) << \
+ HINIC_MBOX_INT_##field##_SHIFT)
+
+enum hinic_mbox_tx_status {
+ TX_DONE = 0,
+ TX_IN_PROGRESS,
+};
+
+#define HINIC_MBOX_CTRL_TRIGGER_AEQE_SHIFT 0
+/* specifies the issue request for the message data.
+ * 0 - Tx request is done;
+ * 1 - Tx request is in process.
+ */
+#define HINIC_MBOX_CTRL_TX_STATUS_SHIFT 1
+
+#define HINIC_MBOX_CTRL_TRIGGER_AEQE_MASK 0x1
+#define HINIC_MBOX_CTRL_TX_STATUS_MASK 0x1
+
+#define HINIC_MBOX_CTRL_SET(val, field) \
+ (((val) & HINIC_MBOX_CTRL_##field##_MASK) << \
+ HINIC_MBOX_CTRL_##field##_SHIFT)
+
+#define HINIC_MBOX_HEADER_MSG_LEN_SHIFT 0
+#define HINIC_MBOX_HEADER_MODULE_SHIFT 11
+#define HINIC_MBOX_HEADER_SEG_LEN_SHIFT 16
+#define HINIC_MBOX_HEADER_NO_ACK_SHIFT 22
+#define HINIC_MBOX_HEADER_SEQID_SHIFT 24
+#define HINIC_MBOX_HEADER_LAST_SHIFT 30
+
+#define HINIC_MBOX_HEADER_DIRECTION_SHIFT 31
+#define HINIC_MBOX_HEADER_CMD_SHIFT 32
+#define HINIC_MBOX_HEADER_MSG_ID_SHIFT 40
+#define HINIC_MBOX_HEADER_STATUS_SHIFT 48
+#define HINIC_MBOX_HEADER_SRC_GLB_FUNC_IDX_SHIFT 54
+
+#define HINIC_MBOX_HEADER_MSG_LEN_MASK 0x7FF
+#define HINIC_MBOX_HEADER_MODULE_MASK 0x1F
+#define HINIC_MBOX_HEADER_SEG_LEN_MASK 0x3F
+#define HINIC_MBOX_HEADER_NO_ACK_MASK 0x1
+#define HINIC_MBOX_HEADER_SEQID_MASK 0x3F
+#define HINIC_MBOX_HEADER_LAST_MASK 0x1
+#define HINIC_MBOX_HEADER_DIRECTION_MASK 0x1
+#define HINIC_MBOX_HEADER_CMD_MASK 0xFF
+#define HINIC_MBOX_HEADER_MSG_ID_MASK 0xFF
+#define HINIC_MBOX_HEADER_STATUS_MASK 0x3F
+#define HINIC_MBOX_HEADER_SRC_GLB_FUNC_IDX_MASK 0x3FF
+
+#define HINIC_MBOX_HEADER_GET(val, field) \
+ (((val) >> HINIC_MBOX_HEADER_##field##_SHIFT) & \
+ HINIC_MBOX_HEADER_##field##_MASK)
+#define HINIC_MBOX_HEADER_SET(val, field) \
+ ((u64)((val) & HINIC_MBOX_HEADER_##field##_MASK) << \
+ HINIC_MBOX_HEADER_##field##_SHIFT)
+
+#define HINIC_MBOX_COMP_TIME_MS 8000U
+#define MBOX_MSG_POLLING_TIMEOUT_MS 5000
+
+/* The size unit is Bytes */
+#define HINIC_MBOX_DATA_SIZE 2040
+#define MBOX_MAX_BUF_SZ 2048UL
+#define MBOX_HEADER_SZ 8
+
+/* MBOX size is 64B, 8B for mbox_header, 4B reserved */
+#define MBOX_SEG_LEN 48
+#define MBOX_SEG_LEN_ALIGN 4
+#define MBOX_WB_STATUS_LEN 16UL
+#define MBOX_SIZE 64
+
+/* mbox write back status is 16B, only first 4B is used */
+#define MBOX_WB_STATUS_ERRCODE_MASK 0xFFFF
+#define MBOX_WB_STATUS_MASK 0xFF
+#define MBOX_WB_ERROR_CODE_MASK 0xFF00
+#define MBOX_WB_STATUS_FINISHED_SUCCESS 0xFF
+#define MBOX_WB_STATUS_FINISHED_WITH_ERR 0xFE
+#define MBOX_WB_STATUS_NOT_FINISHED 0x00
+
+#define MBOX_STATUS_FINISHED(wb) \
+ (((wb) & MBOX_WB_STATUS_MASK) != MBOX_WB_STATUS_NOT_FINISHED)
+#define MBOX_STATUS_SUCCESS(wb) \
+ (((wb) & MBOX_WB_STATUS_MASK) == MBOX_WB_STATUS_FINISHED_SUCCESS)
+#define MBOX_STATUS_ERRCODE(wb) \
+ ((wb) & MBOX_WB_ERROR_CODE_MASK)
+
+#define SEQ_ID_START_VAL 0
+
+#define DST_AEQ_IDX_DEFAULT_VAL 0
+#define SRC_AEQ_IDX_DEFAULT_VAL 0
+#define NO_DMA_ATTRIBUTE_VAL 0
+
+#define MBOX_MSG_NO_DATA_LEN 1
+
+#define FUNC_ID_OFF_SET_8B 8
+#define FUNC_ID_OFF_SET_10B 10
+
+#define MBOX_BODY_FROM_HDR(header) ((u8 *)(header) + MBOX_HEADER_SZ)
+#define MBOX_AREA(hwif) \
+ ((hwif)->cfg_regs_base + HINIC_FUNC_CSR_MAILBOX_DATA_OFF)
+
+#define MBOX_RESPONSE_ERROR 0x1
+#define MBOX_MSG_ID_MASK 0xFF
+#define MBOX_MSG_ID(func_to_func) ((func_to_func)->send_msg_id)
+#define MBOX_MSG_ID_INC(func_to_func) (MBOX_MSG_ID(func_to_func) = \
+ (MBOX_MSG_ID(func_to_func) + 1) & MBOX_MSG_ID_MASK)
+
+enum hinic_hwif_direction_type {
+ /* driver send msg to up or up send msg to drier*/
+ HINIC_HWIF_DIRECT_SEND = 0,
+ /* after driver/up send msg to each other, then up/driver ack the msg */
+ HINIC_HWIF_RESPONSE,
+};
+
+enum mbox_send_mod {
+ MBOX_SEND_MSG_POLL = 1
+};
+
+enum mbox_seg_type {
+ NOT_LAST_SEG,
+ LAST_SEG,
+};
+
+enum mbox_ordering_type {
+ STRONG_ORDER,
+ RELAX_ORDER,
+};
+
+enum mbox_write_back_type {
+ NOT_WRITE_BACK = 0,
+ WRITE_BACK,
+};
+
+enum mbox_aeq_trig_type {
+ NOT_TRIGGER,
+ TRIGGER,
+};
+
+static int send_mbox_to_func(struct hinic_mbox_func_to_func *func_to_func,
+ enum hinic_mod_type mod, u16 cmd, void *msg,
+ u16 msg_len, u16 dst_func,
+ enum hinic_hwif_direction_type direction,
+ enum hinic_mbox_ack_type ack_type,
+ struct mbox_msg_info *msg_info);
+
+static int recv_vf_mbox_handler(struct hinic_mbox_func_to_func *func_to_func,
+ struct hinic_recv_mbox *recv_mbox,
+ void *buf_out, u16 *out_size, void *param)
+{
+ int rc = 0;
+ *out_size = 0;
+
+ switch (recv_mbox->mod) {
+ case HINIC_MOD_COMM:
+ hinic_comm_async_event_handle(func_to_func->hwdev,
+ recv_mbox->cmd, recv_mbox->mbox,
+ recv_mbox->mbox_len,
+ buf_out, out_size);
+ break;
+ case HINIC_MOD_L2NIC:
+ hinic_l2nic_async_event_handle(func_to_func->hwdev, param,
+ recv_mbox->cmd, recv_mbox->mbox,
+ recv_mbox->mbox_len,
+ buf_out, out_size);
+ break;
+ default:
+ PMD_DRV_LOG(ERR, "No handler, mod = %d",
+ recv_mbox->mod);
+ rc = HINIC_MBOX_VF_CMD_ERROR;
+ break;
+ }
+
+ return rc;
+}
+
+static void set_mbx_msg_status(struct mbox_msg_info *msg_info, int status)
+{
+ if (status == HINIC_DEV_BUSY_ACTIVE_FW)
+ msg_info->status = HINIC_MBOX_PF_BUSY_ACTIVE_FW;
+ else if (status == HINIC_MBOX_VF_CMD_ERROR)
+ msg_info->status = HINIC_MBOX_VF_CMD_ERROR;
+ else if (status)
+ msg_info->status = HINIC_MBOX_PF_SEND_ERR;
+}
+
+static void recv_func_mbox_handler(struct hinic_mbox_func_to_func *func_to_func,
+ struct hinic_recv_mbox *recv_mbox,
+ u16 src_func_idx, void *param)
+{
+ struct hinic_hwdev *dev = func_to_func->hwdev;
+ struct mbox_msg_info msg_info = { 0 };
+ u16 out_size = MBOX_MAX_BUF_SZ;
+ void *buf_out = recv_mbox->buf_out;
+ int err = 0;
+
+ if (HINIC_IS_VF(dev)) {
+ err = recv_vf_mbox_handler(func_to_func, recv_mbox, buf_out,
+ &out_size, param);
+ } else {
+ err = -EINVAL;
+ PMD_DRV_LOG(ERR, "PMD doesn't support non-VF handle mailbox message");
+ }
+
+ if (!out_size || err)
+ out_size = MBOX_MSG_NO_DATA_LEN;
+
+ if (recv_mbox->ack_type == MBOX_ACK) {
+ msg_info.msg_id = recv_mbox->msg_info.msg_id;
+ set_mbx_msg_status(&msg_info, err);
+ send_mbox_to_func(func_to_func, recv_mbox->mod, recv_mbox->cmd,
+ buf_out, out_size, src_func_idx,
+ HINIC_HWIF_RESPONSE, MBOX_ACK, &msg_info);
+ }
+}
+
+static bool check_mbox_seq_id_and_seg_len(struct hinic_recv_mbox *recv_mbox,
+ u8 seq_id, u8 seg_len)
+{
+ if (seq_id > HINIC_SEQ_ID_MAX_VAL || seg_len > HINIC_MSG_SEG_LEN)
+ return false;
+
+ if (seq_id == 0) {
+ recv_mbox->sed_id = seq_id;
+ } else {
+ if (seq_id != recv_mbox->sed_id + 1) {
+ recv_mbox->sed_id = 0;
+ return false;
+ }
+
+ recv_mbox->sed_id = seq_id;
+ }
+
+ return true;
+}
+
+static void clear_mbox_status(struct hinic_send_mbox *mbox)
+{
+ /* clear mailbox write back status */
+ *mbox->wb_status = 0;
+ rte_wmb();
+}
+
+static void mbox_copy_header(struct hinic_send_mbox *mbox, u64 *header)
+{
+ u32 *data = (u32 *)header;
+ u32 i, idx_max = MBOX_HEADER_SZ / sizeof(u32);
+
+ for (i = 0; i < idx_max; i++)
+ __raw_writel(*(data + i), mbox->data + i * sizeof(u32));
+}
+
+static void
+mbox_copy_send_data(struct hinic_send_mbox *mbox, void *seg, u16 seg_len)
+{
+ u32 *data = (u32 *)seg;
+ u32 data_len, chk_sz = sizeof(u32);
+ u32 i, idx_max;
+ u8 mbox_max_buf[MBOX_SEG_LEN] = {0};
+
+ /* The mbox message should be aligned in 4 bytes. */
+ if (seg_len % chk_sz) {
+ memcpy(mbox_max_buf, seg, seg_len);
+ data = (u32 *)mbox_max_buf;
+ }
+
+ data_len = seg_len;
+ idx_max = ALIGN(data_len, chk_sz) / chk_sz;
+
+ for (i = 0; i < idx_max; i++)
+ __raw_writel(*(data + i),
+ mbox->data + MBOX_HEADER_SZ + i * sizeof(u32));
+}
+
+static void write_mbox_msg_attr(struct hinic_mbox_func_to_func *func_to_func,
+ u16 dst_func, u16 dst_aeqn,
+ __rte_unused u16 seg_len, int poll)
+{
+ u32 mbox_int, mbox_ctrl;
+
+ mbox_int = HINIC_MBOX_INT_SET(dst_func, DST_FUNC) |
+ HINIC_MBOX_INT_SET(dst_aeqn, DST_AEQN) |
+ HINIC_MBOX_INT_SET(HINIC_MBOX_RSP_AEQN, SRC_RESP_AEQN) |
+ HINIC_MBOX_INT_SET(NO_DMA_ATTRIBUTE_VAL, STAT_DMA) |
+ HINIC_MBOX_INT_SET(ALIGN(MBOX_SIZE, MBOX_SEG_LEN_ALIGN) >> 2,
+ TX_SIZE) |
+ HINIC_MBOX_INT_SET(STRONG_ORDER, STAT_DMA_SO_RO) |
+ HINIC_MBOX_INT_SET(WRITE_BACK, WB_EN);
+
+ hinic_hwif_write_reg(func_to_func->hwdev->hwif,
+ HINIC_FUNC_CSR_MAILBOX_INT_OFFSET_OFF, mbox_int);
+
+ rte_wmb();
+ mbox_ctrl = HINIC_MBOX_CTRL_SET(TX_IN_PROGRESS, TX_STATUS);
+
+ if (poll)
+ mbox_ctrl |= HINIC_MBOX_CTRL_SET(NOT_TRIGGER, TRIGGER_AEQE);
+ else
+ mbox_ctrl |= HINIC_MBOX_CTRL_SET(TRIGGER, TRIGGER_AEQE);
+
+ hinic_hwif_write_reg(func_to_func->hwdev->hwif,
+ HINIC_FUNC_CSR_MAILBOX_CONTROL_OFF, mbox_ctrl);
+}
+
+static int init_mbox_info(struct hinic_recv_mbox *mbox_info)
+{
+ int err;
+
+ mbox_info->mbox = kzalloc(MBOX_MAX_BUF_SZ, GFP_KERNEL);
+ if (!mbox_info->mbox) {
+ PMD_DRV_LOG(ERR, "Alloc mbox buf_in mem failed\n");
+ return -ENOMEM;
+ }
+
+ mbox_info->buf_out = kzalloc(MBOX_MAX_BUF_SZ, GFP_KERNEL);
+ if (!mbox_info->buf_out) {
+ PMD_DRV_LOG(ERR, "Alloc mbox buf_out mem failed\n");
+ err = -ENOMEM;
+ goto alloc_buf_out_err;
+ }
+
+ return 0;
+
+alloc_buf_out_err:
+ kfree(mbox_info->mbox);
+
+ return err;
+}
+
+static void clean_mbox_info(struct hinic_recv_mbox *mbox_info)
+{
+ kfree(mbox_info->buf_out);
+ kfree(mbox_info->mbox);
+}
+
+static int alloc_mbox_info(struct hinic_recv_mbox *mbox_info)
+{
+ u16 func_idx, i;
+ int err;
+
+ for (func_idx = 0; func_idx < HINIC_MAX_FUNCTIONS; func_idx++) {
+ err = init_mbox_info(&mbox_info[func_idx]);
+ if (err) {
+ PMD_DRV_LOG(ERR, "Initialize function[%d] mailbox information failed, err: %d",
+ func_idx, err);
+ goto init_mbox_info_err;
+ }
+ }
+
+ return 0;
+
+init_mbox_info_err:
+ for (i = 0; i < func_idx; i++)
+ clean_mbox_info(&mbox_info[i]);
+
+ return err;
+}
+
+static void free_mbox_info(struct hinic_recv_mbox *mbox_info)
+{
+ u16 func_idx;
+
+ for (func_idx = 0; func_idx < HINIC_MAX_FUNCTIONS; func_idx++)
+ clean_mbox_info(&mbox_info[func_idx]);
+}
+
+static void prepare_send_mbox(struct hinic_mbox_func_to_func *func_to_func)
+{
+ struct hinic_send_mbox *send_mbox = &func_to_func->send_mbox;
+
+ send_mbox->data = MBOX_AREA(func_to_func->hwdev->hwif);
+}
+
+static int alloc_mbox_wb_status(struct hinic_mbox_func_to_func *func_to_func)
+{
+ struct hinic_send_mbox *send_mbox = &func_to_func->send_mbox;
+ struct hinic_hwdev *hwdev = func_to_func->hwdev;
+ struct hinic_hwif *hwif = hwdev->hwif;
+ u32 addr_h, addr_l;
+
+ send_mbox->wb_vaddr = dma_zalloc_coherent(hwdev,
+ MBOX_WB_STATUS_LEN,
+ &send_mbox->wb_paddr,
+ GFP_KERNEL);
+ if (!send_mbox->wb_vaddr) {
+ PMD_DRV_LOG(ERR, "Allocating memory for mailbox wb status failed");
+ return -ENOMEM;
+ }
+ send_mbox->wb_status = (volatile u64 *)send_mbox->wb_vaddr;
+
+ addr_h = upper_32_bits(send_mbox->wb_paddr);
+ addr_l = lower_32_bits(send_mbox->wb_paddr);
+ hinic_hwif_write_reg(hwif, HINIC_FUNC_CSR_MAILBOX_RESULT_H_OFF, addr_h);
+ hinic_hwif_write_reg(hwif, HINIC_FUNC_CSR_MAILBOX_RESULT_L_OFF, addr_l);
+
+ return 0;
+}
+
+static void free_mbox_wb_status(struct hinic_mbox_func_to_func *func_to_func)
+{
+ struct hinic_send_mbox *send_mbox = &func_to_func->send_mbox;
+ struct hinic_hwdev *hwdev = func_to_func->hwdev;
+ struct hinic_hwif *hwif = hwdev->hwif;
+
+ hinic_hwif_write_reg(hwif, HINIC_FUNC_CSR_MAILBOX_RESULT_H_OFF, 0);
+ hinic_hwif_write_reg(hwif, HINIC_FUNC_CSR_MAILBOX_RESULT_L_OFF, 0);
+
+ dma_free_coherent(hwdev, MBOX_WB_STATUS_LEN,
+ send_mbox->wb_vaddr, send_mbox->wb_paddr);
+}
+
+static int recv_mbox_handler(struct hinic_mbox_func_to_func *func_to_func,
+ void *header, struct hinic_recv_mbox *recv_mbox, void *param)
+{
+ u64 mbox_header = *((u64 *)header);
+ void *mbox_body = MBOX_BODY_FROM_HDR(header);
+ u16 src_func_idx;
+ enum hinic_hwif_direction_type direction;
+ u8 seq_id, seg_len;
+
+ seq_id = HINIC_MBOX_HEADER_GET(mbox_header, SEQID);
+ seg_len = HINIC_MBOX_HEADER_GET(mbox_header, SEG_LEN);
+ direction = HINIC_MBOX_HEADER_GET(mbox_header, DIRECTION);
+ src_func_idx = HINIC_MBOX_HEADER_GET(mbox_header, SRC_GLB_FUNC_IDX);
+
+ if (!check_mbox_seq_id_and_seg_len(recv_mbox, seq_id, seg_len)) {
+ PMD_DRV_LOG(ERR,
+ "Mailbox sequence and segment check failed, src func id: 0x%x, front id: 0x%x, current id: 0x%x, seg len: 0x%x\n",
+ src_func_idx, recv_mbox->sed_id, seq_id, seg_len);
+ return HINIC_ERROR;
+ }
+
+ memcpy((u8 *)recv_mbox->mbox + seq_id * HINIC_MSG_SEG_LEN,
+ mbox_body, seg_len);
+
+ if (!HINIC_MBOX_HEADER_GET(mbox_header, LAST))
+ return HINIC_ERROR;
+
+ recv_mbox->sed_id = 0;
+ recv_mbox->cmd = HINIC_MBOX_HEADER_GET(mbox_header, CMD);
+ recv_mbox->mod = HINIC_MBOX_HEADER_GET(mbox_header, MODULE);
+ recv_mbox->mbox_len = HINIC_MBOX_HEADER_GET(mbox_header, MSG_LEN);
+ recv_mbox->ack_type = HINIC_MBOX_HEADER_GET(mbox_header, NO_ACK);
+ recv_mbox->msg_info.msg_id = HINIC_MBOX_HEADER_GET(mbox_header, MSG_ID);
+ recv_mbox->msg_info.status = HINIC_MBOX_HEADER_GET(mbox_header, STATUS);
+
+ if (direction == HINIC_HWIF_RESPONSE) {
+ if (recv_mbox->msg_info.msg_id == func_to_func->send_msg_id &&
+ func_to_func->event_flag == EVENT_START) {
+ return HINIC_OK;
+ }
+
+ PMD_DRV_LOG(ERR, "Mbox response timeout, current send msg id(0x%x), recv msg id(0x%x), status(0x%x)",
+ func_to_func->send_msg_id, recv_mbox->msg_info.msg_id,
+ recv_mbox->msg_info.status);
+ return HINIC_ERROR;
+ }
+
+ recv_func_mbox_handler(func_to_func, recv_mbox, src_func_idx, param);
+
+ return HINIC_ERROR;
+}
+
+/**
+ * hinic_mbox_func_aeqe_handler - Process mbox info from func which is
+ * sent by aeqe.
+ *
+ * @param handle
+ * Pointer to hradware nic device.
+ * @param header
+ * Mbox header info.
+ * @param size
+ * The size of aeqe descriptor.
+ * @param param
+ * customized parameter.
+ *
+ * @return
+ * 0 on success, negative error value otherwise.
+ */
+int hinic_mbox_func_aeqe_handler(void *handle, u8 *header,
+ __rte_unused u8 size, void *param)
+{
+ struct hinic_mbox_func_to_func *func_to_func =
+ ((struct hinic_hwdev *)handle)->func_to_func;
+ struct hinic_recv_mbox *recv_mbox;
+ u64 mbox_header = *((u64 *)header);
+ u16 src = HINIC_MBOX_HEADER_GET(mbox_header, SRC_GLB_FUNC_IDX);
+
+ if (src >= HINIC_MAX_FUNCTIONS) {
+ PMD_DRV_LOG(ERR, "Mailbox source function id: %d is invalid",
+ src);
+ return HINIC_ERROR;
+ }
+
+ recv_mbox = (HINIC_MBOX_HEADER_GET(mbox_header, DIRECTION) ==
+ HINIC_HWIF_DIRECT_SEND) ?
+ &func_to_func->mbox_send[src] :
+ &func_to_func->mbox_resp[src];
+
+ return recv_mbox_handler(func_to_func, (u64 *)header, recv_mbox, param);
+}
+
+static u16 get_mbox_status(struct hinic_send_mbox *mbox)
+{
+ /* write back is 16B, but only use first 4B */
+ u64 wb_val = be64_to_cpu(*mbox->wb_status);
+
+ rte_rmb(); /* verify reading before check */
+
+ return (u16)(wb_val & MBOX_WB_STATUS_ERRCODE_MASK);
+}
+
+static void dump_mox_reg(struct hinic_hwdev *hwdev)
+{
+ u32 val;
+
+ val = hinic_hwif_read_reg(hwdev->hwif,
+ HINIC_FUNC_CSR_MAILBOX_CONTROL_OFF);
+ PMD_DRV_LOG(WARNING, "Mailbox control reg: 0x%x", val);
+ val = hinic_hwif_read_reg(hwdev->hwif,
+ HINIC_FUNC_CSR_MAILBOX_INT_OFFSET_OFF);
+ PMD_DRV_LOG(WARNING, "Mailbox interrupt offset: 0x%x", val);
+}
+
+static int send_mbox_seg(struct hinic_mbox_func_to_func *func_to_func,
+ u64 header, u16 dst_func, void *seg, u16 seg_len)
+{
+ struct hinic_send_mbox *send_mbox = &func_to_func->send_mbox;
+ struct hinic_hwdev *hwdev = func_to_func->hwdev;
+ u16 seq_dir = HINIC_MBOX_HEADER_GET(header, DIRECTION);
+ u16 dst_aeqn = (seq_dir == HINIC_HWIF_DIRECT_SEND) ?
+ HINIC_MBOX_RECV_AEQN : HINIC_MBOX_RSP_AEQN;
+ u16 err_code, wb_status = 0;
+ u32 cnt = 0;
+
+ clear_mbox_status(send_mbox);
+
+ mbox_copy_header(send_mbox, &header);
+
+ mbox_copy_send_data(send_mbox, seg, seg_len);
+
+ write_mbox_msg_attr(func_to_func, dst_func, dst_aeqn, seg_len,
+ MBOX_SEND_MSG_POLL);
+
+ rte_wmb();
+
+ while (cnt < MBOX_MSG_POLLING_TIMEOUT_MS) {
+ wb_status = get_mbox_status(send_mbox);
+ if (MBOX_STATUS_FINISHED(wb_status))
+ break;
+
+ rte_delay_ms(1); /* loop every ms */
+ cnt++;
+ }
+
+ if (cnt == MBOX_MSG_POLLING_TIMEOUT_MS) {
+ PMD_DRV_LOG(ERR, "Send mailbox segment timeout, wb status: 0x%x",
+ wb_status);
+ dump_mox_reg(hwdev);
+ return -ETIMEDOUT;
+ }
+
+ if (!MBOX_STATUS_SUCCESS(wb_status)) {
+ PMD_DRV_LOG(ERR, "Send mailbox segment to function %d error, wb status: 0x%x",
+ dst_func, wb_status);
+ /*
+ * err_code: 0 responses no errors, other values can
+ * refer to FS doc.
+ */
+ err_code = MBOX_STATUS_ERRCODE(wb_status);
+ return err_code ? err_code : -EFAULT;
+ }
+
+ return 0;
+}
+
+static void set_mbox_to_func_event(struct hinic_mbox_func_to_func *func_to_func,
+ enum mbox_event_state event_flag)
+{
+ spin_lock(&func_to_func->mbox_lock);
+ func_to_func->event_flag = event_flag;
+ spin_unlock(&func_to_func->mbox_lock);
+}
+
+static int send_mbox_to_func(struct hinic_mbox_func_to_func *func_to_func,
+ enum hinic_mod_type mod, u16 cmd, void *msg,
+ u16 msg_len, u16 dst_func,
+ enum hinic_hwif_direction_type direction,
+ enum hinic_mbox_ack_type ack_type,
+ struct mbox_msg_info *msg_info)
+{
+ struct hinic_hwdev *hwdev = func_to_func->hwdev;
+ int err = 0;
+ u32 seq_id = 0;
+ u16 seg_len = HINIC_MSG_SEG_LEN;
+ u16 left = msg_len;
+ u8 *msg_seg = (u8 *)msg;
+ u64 header = 0;
+
+ err = hinic_mutex_lock(&func_to_func->msg_send_mutex);
+ if (err)
+ return err;
+
+ header = HINIC_MBOX_HEADER_SET(msg_len, MSG_LEN) |
+ HINIC_MBOX_HEADER_SET(mod, MODULE) |
+ HINIC_MBOX_HEADER_SET(seg_len, SEG_LEN) |
+ HINIC_MBOX_HEADER_SET(ack_type, NO_ACK) |
+ HINIC_MBOX_HEADER_SET(SEQ_ID_START_VAL, SEQID) |
+ HINIC_MBOX_HEADER_SET(NOT_LAST_SEG, LAST) |
+ HINIC_MBOX_HEADER_SET(direction, DIRECTION) |
+ HINIC_MBOX_HEADER_SET(cmd, CMD) |
+ HINIC_MBOX_HEADER_SET(msg_info->msg_id, MSG_ID) |
+ HINIC_MBOX_HEADER_SET(msg_info->status, STATUS) |
+ HINIC_MBOX_HEADER_SET(hinic_global_func_id(hwdev),
+ SRC_GLB_FUNC_IDX);
+
+ while (!(HINIC_MBOX_HEADER_GET(header, LAST))) {
+ if (left <= HINIC_MSG_SEG_LEN) {
+ header &=
+ ~(HINIC_MBOX_HEADER_SET(HINIC_MBOX_HEADER_SEG_LEN_MASK,
+ SEG_LEN));
+ header |= HINIC_MBOX_HEADER_SET(left, SEG_LEN);
+ header |= HINIC_MBOX_HEADER_SET(LAST_SEG, LAST);
+
+ seg_len = left;
+ }
+
+ err = send_mbox_seg(func_to_func, header, dst_func, msg_seg,
+ seg_len);
+ if (err) {
+ PMD_DRV_LOG(ERR, "Fail to send mbox seg, seq_id: 0x%lx, err: %d",
+ HINIC_MBOX_HEADER_GET(header, SEQID), err);
+ goto send_err;
+ }
+
+ left -= HINIC_MSG_SEG_LEN;
+ msg_seg += HINIC_MSG_SEG_LEN;
+
+ seq_id++;
+ header &= ~(HINIC_MBOX_HEADER_SET(HINIC_MBOX_HEADER_SEQID_MASK,
+ SEQID));
+ header |= HINIC_MBOX_HEADER_SET(seq_id, SEQID);
+ }
+
+send_err:
+ (void)hinic_mutex_unlock(&func_to_func->msg_send_mutex);
+
+ return err;
+}
+
+static int hinic_mbox_to_func(struct hinic_mbox_func_to_func *func_to_func,
+ enum hinic_mod_type mod, u16 cmd, u16 dst_func,
+ void *buf_in, u16 in_size, void *buf_out, u16 *out_size,
+ u32 timeout)
+{
+ struct hinic_recv_mbox *mbox_for_resp =
+ &func_to_func->mbox_resp[dst_func];
+ struct mbox_msg_info msg_info = {0};
+ u32 time;
+ int err;
+
+ err = hinic_mutex_lock(&func_to_func->mbox_send_mutex);
+ if (err)
+ return err;
+
+ msg_info.msg_id = MBOX_MSG_ID_INC(func_to_func);
+
+ set_mbox_to_func_event(func_to_func, EVENT_START);
+
+ err = send_mbox_to_func(func_to_func, mod, cmd, buf_in, in_size,
+ dst_func, HINIC_HWIF_DIRECT_SEND,
+ MBOX_ACK, &msg_info);
+ if (err)
+ goto send_err;
+
+ time = msecs_to_jiffies(timeout ? timeout : HINIC_MBOX_COMP_TIME_MS);
+ err = hinic_aeq_poll_msg(func_to_func->rsp_aeq, time, NULL);
+ if (err) {
+ set_mbox_to_func_event(func_to_func, EVENT_TIMEOUT);
+ PMD_DRV_LOG(ERR, "Send mailbox message time out");
+ err = -ETIMEDOUT;
+ goto send_err;
+ }
+
+ set_mbox_to_func_event(func_to_func, EVENT_END);
+
+ if (mbox_for_resp->msg_info.status) {
+ err = mbox_for_resp->msg_info.status;
+ if (err != HINIC_MBOX_PF_BUSY_ACTIVE_FW)
+ PMD_DRV_LOG(ERR, "Mailbox response error: 0x%x",
+ mbox_for_resp->msg_info.status);
+ else
+ PMD_DRV_LOG(ERR, "Chip is in active, PF can't process VF message");
+ goto send_err;
+ }
+
+ rte_rmb();
+
+ if (mbox_for_resp->mbox_len && buf_out && out_size) {
+ if (mbox_for_resp->mbox_len <= *out_size) {
+ memcpy(buf_out, mbox_for_resp->mbox,
+ mbox_for_resp->mbox_len);
+ *out_size = mbox_for_resp->mbox_len;
+ } else {
+ PMD_DRV_LOG(ERR, "Mailbox response message len[%u] overflow",
+ mbox_for_resp->mbox_len);
+ err = -ERANGE;
+ }
+ }
+
+send_err:
+ if (err && out_size)
+ *out_size = 0;
+ (void)hinic_mutex_unlock(&func_to_func->mbox_send_mutex);
+
+ return err;
+}
+
+static int
+mbox_func_params_valid(__rte_unused struct hinic_mbox_func_to_func *mbox_obj,
+ void *buf_in, u16 in_size)
+{
+ if (!buf_in || !in_size)
+ return -EINVAL;
+
+ if (in_size > HINIC_MBOX_DATA_SIZE) {
+ PMD_DRV_LOG(ERR, "Mailbox message len(%d) exceed limit(%d)",
+ in_size, HINIC_MBOX_DATA_SIZE);
+ return -EINVAL;
+ }
+
+ return 0;
+}
+
+static u8 hinic_pf_id_of_vf(void *hwdev)
+{
+ struct hinic_hwif *hwif = ((struct hinic_hwdev *)hwdev)->hwif;
+ return hwif->attr.port_to_port_idx;
+}
+
+/**
+ * hinic_mbox_to_pf - Send mbox info to pf and need pf to response.
+ *
+ * @param hwdev
+ * Pointer to hardware nic device.
+ * @param mod
+ * Mode type of hardware.
+ * @param cmd
+ * The command sent to pf.
+ * @param buf_in
+ * Input parameter.
+ * @param in_size
+ * Input parameter size.
+ * @param buf_out
+ * Output parameter.
+ * @param out_size
+ * Output parameter size.
+ * @param timeout
+ * Timeout.
+ *
+ * @return
+ * 0 on success, negative error value otherwise.
+ */
+int hinic_mbox_to_pf(struct hinic_hwdev *hwdev,
+ enum hinic_mod_type mod, u8 cmd, void *buf_in,
+ u16 in_size, void *buf_out, u16 *out_size, u32 timeout)
+{
+ struct hinic_mbox_func_to_func *func_to_func = hwdev->func_to_func;
+ int err;
+
+ err = mbox_func_params_valid(func_to_func, buf_in, in_size);
+ if (err) {
+ PMD_DRV_LOG(ERR, "Mailbox parameters check failed: %d", err);
+ return err;
+ }
+
+ if (!HINIC_IS_VF(hwdev)) {
+ PMD_DRV_LOG(ERR, "Input function type error, func_type: %d",
+ hinic_func_type(hwdev));
+ return -EINVAL;
+ }
+
+ return hinic_mbox_to_func(func_to_func, mod, cmd,
+ hinic_pf_id_of_vf(hwdev), buf_in, in_size,
+ buf_out, out_size, timeout);
+}
+
+/**
+ * hinic_mbox_to_pf_no_ack - Send mbox info to pf and do not need pf to response
+ *
+ * @param hwdev
+ * Pointer to hardware nic device.
+ * @param mod
+ * Mode type of hardware.
+ * @param cmd
+ * The command sent to pf.
+ * @param buf_in
+ * Input parameter.
+ * @param in_size
+ * Input parameter size.
+ *
+ * @return
+ * 0 on success, negative error value otherwise.
+ */
+int hinic_mbox_to_pf_no_ack(struct hinic_hwdev *hwdev, enum hinic_mod_type mod,
+ u8 cmd, void *buf_in, u16 in_size)
+{
+ int err;
+ struct mbox_msg_info msg_info = {0};
+
+ err = hinic_mutex_lock(&hwdev->func_to_func->mbox_send_mutex);
+ if (err)
+ return err;
+
+ err = send_mbox_to_func(hwdev->func_to_func, mod, cmd, buf_in, in_size,
+ hinic_pf_id_of_vf(hwdev), HINIC_HWIF_DIRECT_SEND,
+ MBOX_NO_ACK, &msg_info);
+ if (err)
+ PMD_DRV_LOG(ERR, "Send mailbox no ack failed, err: %d", err);
+
+ (void)hinic_mutex_unlock(&hwdev->func_to_func->mbox_send_mutex);
+
+ return err;
+}
+
+static int hinic_func_to_func_init(struct hinic_hwdev *hwdev)
+{
+ struct hinic_mbox_func_to_func *func_to_func;
+ int err;
+
+ func_to_func = kzalloc(sizeof(*func_to_func), GFP_KERNEL);
+ if (!func_to_func) {
+ PMD_DRV_LOG(ERR, "Allocating memory for func_to_func object failed");
+ return -ENOMEM;
+ }
+ hwdev->func_to_func = func_to_func;
+ func_to_func->hwdev = hwdev;
+ (void)hinic_mutex_init(&func_to_func->mbox_send_mutex, NULL);
+ (void)hinic_mutex_init(&func_to_func->msg_send_mutex, NULL);
+
+ err = alloc_mbox_info(func_to_func->mbox_send);
+ if (err) {
+ PMD_DRV_LOG(ERR, "Allocating memory for mailbox sending failed");
+ goto alloc_mbox_for_send_err;
+ }
+
+ err = alloc_mbox_info(func_to_func->mbox_resp);
+ if (err) {
+ PMD_DRV_LOG(ERR, "Allocating memory for mailbox responsing failed");
+ goto alloc_mbox_for_resp_err;
+ }
+
+ err = alloc_mbox_wb_status(func_to_func);
+ if (err)
+ goto alloc_wb_status_err;
+
+ prepare_send_mbox(func_to_func);
+
+ return 0;
+
+alloc_wb_status_err:
+ free_mbox_info(func_to_func->mbox_resp);
+
+alloc_mbox_for_resp_err:
+ free_mbox_info(func_to_func->mbox_send);
+
+alloc_mbox_for_send_err:
+ kfree(func_to_func);
+
+ return err;
+}
+
+/**
+ * hinic_comm_func_to_func_free - Uninitialize func to func resource.
+ *
+ * @param hwdev
+ * Pointer to hardware nic device.
+ */
+void hinic_comm_func_to_func_free(struct hinic_hwdev *hwdev)
+{
+ struct hinic_mbox_func_to_func *func_to_func = hwdev->func_to_func;
+
+ free_mbox_wb_status(func_to_func);
+ free_mbox_info(func_to_func->mbox_resp);
+ free_mbox_info(func_to_func->mbox_send);
+ (void)hinic_mutex_destroy(&func_to_func->mbox_send_mutex);
+ (void)hinic_mutex_destroy(&func_to_func->msg_send_mutex);
+ kfree(func_to_func);
+}
+
+/**
+ * hinic_comm_func_to_func_init - Initialize func to func resource.
+ *
+ * @param hwdev
+ * Pointer to hardware nic device.
+ */
+int hinic_comm_func_to_func_init(struct hinic_hwdev *hwdev)
+{
+ int rc;
+
+ rc = hinic_func_to_func_init(hwdev);
+ if (rc)
+ return rc;
+
+ hwdev->func_to_func->rsp_aeq = &hwdev->aeqs->aeq[HINIC_MBOX_RSP_AEQN];
+ hwdev->func_to_func->recv_aeq = &hwdev->aeqs->aeq[HINIC_MBOX_RECV_AEQN];
+
+ return 0;
+}
+
diff --git a/drivers/net/hinic/base/hinic_pmd_mbox.h b/drivers/net/hinic/base/hinic_pmd_mbox.h
new file mode 100644
index 0000000..bf7b490
--- /dev/null
+++ b/drivers/net/hinic/base/hinic_pmd_mbox.h
@@ -0,0 +1,93 @@
+/* SPDX-License-Identifier: BSD-3-Clause
+ * Copyright(c) 2017 Huawei Technologies Co., Ltd
+ */
+
+#ifndef _HINIC_PMD_MBOX_H_
+#define _HINIC_PMD_MBOX_H_
+
+#define HINIC_MBOX_RECV_AEQN 0
+#define HINIC_MBOX_RSP_AEQN 2
+
+#define HINIC_MBOX_PF_SEND_ERR 0x1
+#define HINIC_MBOX_PF_BUSY_ACTIVE_FW 0x2
+#define HINIC_MBOX_VF_CMD_ERROR 0x3
+
+/* PFs do not support enable SR-IOV cap when PFs use PMD, VFs just receive
+ * mailbox message from PFs. The max number of PFs is 16, so the max number
+ * of mailbox buffer for functions is also 16.
+ */
+#define HINIC_MAX_FUNCTIONS 16
+#define HINIC_MAX_PF_FUNCS 16
+
+#define HINIC_MGMT_CMD_UNSUPPORTED 0xFF
+
+#define HINIC_SEQ_ID_MAX_VAL 42
+#define HINIC_MSG_SEG_LEN 48
+
+enum hinic_mbox_ack_type {
+ MBOX_ACK,
+ MBOX_NO_ACK,
+};
+
+struct mbox_msg_info {
+ u8 msg_id;
+ u8 status; /*can only use 6 bit*/
+};
+
+struct hinic_recv_mbox {
+ void *mbox;
+ u8 cmd;
+ enum hinic_mod_type mod;
+ u16 mbox_len;
+ void *buf_out;
+ enum hinic_mbox_ack_type ack_type;
+ struct mbox_msg_info msg_info;
+ u8 sed_id;
+};
+
+struct hinic_send_mbox {
+ u8 *data;
+ volatile u64 *wb_status;
+ void *wb_vaddr;
+ dma_addr_t wb_paddr;
+};
+
+enum mbox_event_state {
+ EVENT_START = 0,
+ EVENT_TIMEOUT,
+ EVENT_END,
+};
+
+struct hinic_mbox_func_to_func {
+ struct hinic_hwdev *hwdev;
+
+ pthread_mutex_t mbox_send_mutex;
+ pthread_mutex_t msg_send_mutex;
+
+ struct hinic_send_mbox send_mbox;
+
+ struct hinic_recv_mbox mbox_resp[HINIC_MAX_FUNCTIONS];
+ struct hinic_recv_mbox mbox_send[HINIC_MAX_FUNCTIONS];
+
+ struct hinic_eq *rsp_aeq;
+ struct hinic_eq *recv_aeq;
+
+ u8 send_msg_id;
+ enum mbox_event_state event_flag;
+ spinlock_t mbox_lock; /* lock for mbox event flag */
+};
+
+/*
+ * mbox function prototypes
+ */
+int hinic_comm_func_to_func_init(struct hinic_hwdev *hwdev);
+void hinic_comm_func_to_func_free(struct hinic_hwdev *hwdev);
+int hinic_mbox_func_aeqe_handler(void *handle, u8 *header,
+ u8 size, void *param);
+int hinic_mbox_to_pf(struct hinic_hwdev *hwdev, enum hinic_mod_type mod, u8 cmd,
+ void *buf_in, u16 in_size,
+ void *buf_out, u16 *out_size, u32 timeout);
+int hinic_mbox_to_pf_no_ack(struct hinic_hwdev *hwdev, enum hinic_mod_type mod,
+ u8 cmd, void *buf_in, u16 in_size);
+
+#endif /* _HINIC_PMD_MBOX_H_ */
diff --git a/drivers/net/hinic/base/meson.build b/drivers/net/hinic/base/meson.build
index cde394a..6cf947f 100644
--- a/drivers/net/hinic/base/meson.build
+++ b/drivers/net/hinic/base/meson.build
@@ -12,6 +12,7 @@ sources = [
'hinic_pmd_niccfg.c',
'hinic_pmd_nicio.c',
'hinic_pmd_wq.c',
+ 'hinic_pmd_mbox.c',
]
extra_flags = []
--
1.8.3.1
^ permalink raw reply [flat|nested] 32+ messages in thread
* [dpdk-dev] [PATCH v2 02/17] net/hinic/base: add HW interfaces for SRIOV
2019-09-25 14:26 [dpdk-dev] [PATCH v2 00/17] Add advanced features for Huawei hinic pmd Xiaoyun wang
2019-09-25 14:30 ` Xiaoyun wang
2019-09-25 14:30 ` [dpdk-dev] [PATCH v2 01/17] net/hinic/base: add mbox command channel for SRIOV Xiaoyun wang
@ 2019-09-25 14:30 ` Xiaoyun wang
2019-09-25 14:30 ` [dpdk-dev] [PATCH v2 03/17] net/hinic: add VF PMD operation interfaces Xiaoyun wang
` (15 subsequent siblings)
18 siblings, 0 replies; 32+ messages in thread
From: Xiaoyun wang @ 2019-09-25 14:30 UTC (permalink / raw)
To: ferruh.yigit
Cc: dev, xuanziyang2, shahar.belkar, luoxianjun, tanya.brokhman,
zhouguoyang, wulike1, Xiaoyun wang
This patch introduces some HW interfaces required for SRIOV function,
including: init hwdev, set port state, get default cos,
vf dma attribute table, vf txrx flush and so on.
Signed-off-by: Xiaoyun wang <cloud.wangxiaoyun@huawei.com>
---
drivers/net/hinic/base/hinic_compat.h | 32 ----
drivers/net/hinic/base/hinic_pmd_api_cmd.c | 60 +++----
drivers/net/hinic/base/hinic_pmd_cfg.c | 35 ++++
drivers/net/hinic/base/hinic_pmd_cmd.h | 14 +-
drivers/net/hinic/base/hinic_pmd_hwdev.c | 198 ++++++++++++++++-----
drivers/net/hinic/base/hinic_pmd_hwdev.h | 4 +
drivers/net/hinic/base/hinic_pmd_hwif.c | 85 ++++++++-
drivers/net/hinic/base/hinic_pmd_hwif.h | 15 +-
drivers/net/hinic/base/hinic_pmd_mgmt.c | 83 +++++----
drivers/net/hinic/base/hinic_pmd_mgmt.h | 2 +-
drivers/net/hinic/base/hinic_pmd_niccfg.c | 268 +++++++++++++++++++++++------
drivers/net/hinic/base/hinic_pmd_niccfg.h | 4 +
drivers/net/hinic/base/hinic_pmd_nicio.c | 15 +-
drivers/net/hinic/base/hinic_pmd_nicio.h | 3 +-
14 files changed, 606 insertions(+), 212 deletions(-)
diff --git a/drivers/net/hinic/base/hinic_compat.h b/drivers/net/hinic/base/hinic_compat.h
index db60a51..85cbb5b 100644
--- a/drivers/net/hinic/base/hinic_compat.h
+++ b/drivers/net/hinic/base/hinic_compat.h
@@ -223,38 +223,6 @@ static inline u16 ilog2(u32 n)
return res;
}
-/**
- * hinic_cpu_to_be32 - convert data to big endian 32 bit format
- * @data: the data to convert
- * @len: length of data to convert, must be Multiple of 4B
- */
-static inline void hinic_cpu_to_be32(void *data, u32 len)
-{
- u32 i;
- u32 *mem = (u32 *)data;
-
- for (i = 0; i < (len >> 2); i++) {
- *mem = cpu_to_be32(*mem);
- mem++;
- }
-}
-
-/**
- * hinic_be32_to_cpu - convert data from big endian 32 bit format
- * @data: the data to convert
- * @len: length of data to convert, must be Multiple of 4B
- */
-static inline void hinic_be32_to_cpu(void *data, u32 len)
-{
- u32 i;
- u32 *mem = (u32 *)data;
-
- for (i = 0; i < (len >> 2); i++) {
- *mem = be32_to_cpu(*mem);
- mem++;
- }
-}
-
static inline int hinic_mutex_init(pthread_mutex_t *pthreadmutex,
const pthread_mutexattr_t *mattr)
{
diff --git a/drivers/net/hinic/base/hinic_pmd_api_cmd.c b/drivers/net/hinic/base/hinic_pmd_api_cmd.c
index 1ea86fa..dbffc2e 100644
--- a/drivers/net/hinic/base/hinic_pmd_api_cmd.c
+++ b/drivers/net/hinic/base/hinic_pmd_api_cmd.c
@@ -120,7 +120,7 @@ static void dump_api_chain_reg(struct hinic_api_cmd_chain *chain)
/**
* chain_busy - check if the chain is still processing last requests
* @chain: chain to check
- **/
+ */
static int chain_busy(struct hinic_api_cmd_chain *chain)
{
switch (chain->chain_type) {
@@ -146,7 +146,7 @@ static int chain_busy(struct hinic_api_cmd_chain *chain)
/**
* get_cell_data_size - get the data size of specific cell type
* @type: chain type
- **/
+ */
static u16 get_cell_data_size(enum hinic_api_cmd_chain_type type,
__rte_unused u16 cmd_size)
{
@@ -170,7 +170,7 @@ static u16 get_cell_data_size(enum hinic_api_cmd_chain_type type,
* prepare_cell_ctrl - prepare the ctrl of the cell for the command
* @cell_ctrl: the control of the cell to set the control into it
* @cell_len: the size of the cell
- **/
+ */
static void prepare_cell_ctrl(u64 *cell_ctrl, u16 cell_len)
{
u64 ctrl;
@@ -202,7 +202,7 @@ static void prepare_cell_ctrl(u64 *cell_ctrl, u16 cell_len)
* @dest: destination node on the card that will receive the command
* @cmd: command data
* @cmd_size: the command size
- **/
+ */
static void prepare_api_cmd(struct hinic_api_cmd_chain *chain,
struct hinic_api_cmd_cell *cell,
enum hinic_node_id dest,
@@ -254,7 +254,7 @@ static void prepare_api_cmd(struct hinic_api_cmd_chain *chain,
* @dest: destination node on the card that will receive the command
* @cmd: command data
* @cmd_size: the command size
- **/
+ */
static void prepare_cell(struct hinic_api_cmd_chain *chain,
enum hinic_node_id dest,
void *cmd, u16 cmd_size)
@@ -283,7 +283,7 @@ static void issue_api_cmd(struct hinic_api_cmd_chain *chain)
/**
* api_cmd_status_update - update the status of the chain
* @chain: chain to update
- **/
+ */
static void api_cmd_status_update(struct hinic_api_cmd_chain *chain)
{
struct hinic_api_cmd_status *wb_status;
@@ -314,7 +314,7 @@ static void api_cmd_status_update(struct hinic_api_cmd_chain *chain)
* wait_for_status_poll - wait for write to mgmt command to complete
* @chain: the chain of the command
* Return: 0 - success, negative - failure
- **/
+ */
static int wait_for_status_poll(struct hinic_api_cmd_chain *chain)
{
unsigned long end;
@@ -340,7 +340,7 @@ static int wait_for_status_poll(struct hinic_api_cmd_chain *chain)
* wait_for_api_cmd_completion - wait for command to complete
* @chain: chain for the command
* Return: 0 - success, negative - failure
- **/
+ */
static int wait_for_api_cmd_completion(struct hinic_api_cmd_chain *chain,
__rte_unused struct hinic_api_cmd_cell_ctxt *ctxt,
__rte_unused void *ack, __rte_unused u16 ack_size)
@@ -384,7 +384,7 @@ static inline void update_api_cmd_ctxt(struct hinic_api_cmd_chain *chain,
* @ack: pointer to messages to response
* @ack_size: the size of ack message
* Return: 0 - success, negative - failure
- **/
+ */
static int api_cmd(struct hinic_api_cmd_chain *chain,
enum hinic_node_id dest,
void *cmd, u16 cmd_size, void *ack, u16 ack_size)
@@ -423,7 +423,7 @@ static int api_cmd(struct hinic_api_cmd_chain *chain,
* @cmd: command data
* @size: the command size
* Return: 0 - success, negative - failure
- **/
+ */
int hinic_api_cmd_write(struct hinic_api_cmd_chain *chain,
enum hinic_node_id dest, void *cmd, u16 size)
{
@@ -434,7 +434,7 @@ int hinic_api_cmd_write(struct hinic_api_cmd_chain *chain,
/**
* api_cmd_hw_restart - restart the chain in the HW
* @chain: the API CMD specific chain to restart
- **/
+ */
static int api_cmd_hw_restart(struct hinic_api_cmd_chain *chain)
{
struct hinic_hwif *hwif = chain->hwdev->hwif;
@@ -470,7 +470,7 @@ static int api_cmd_hw_restart(struct hinic_api_cmd_chain *chain)
/**
* api_cmd_ctrl_init - set the control register of a chain
* @chain: the API CMD specific chain to set control register for
- **/
+ */
static void api_cmd_ctrl_init(struct hinic_api_cmd_chain *chain)
{
struct hinic_hwif *hwif = chain->hwdev->hwif;
@@ -497,7 +497,7 @@ static void api_cmd_ctrl_init(struct hinic_api_cmd_chain *chain)
/**
* api_cmd_set_status_addr - set the status address of a chain in the HW
* @chain: the API CMD specific chain to set status address for
- **/
+ */
static void api_cmd_set_status_addr(struct hinic_api_cmd_chain *chain)
{
struct hinic_hwif *hwif = chain->hwdev->hwif;
@@ -515,7 +515,7 @@ static void api_cmd_set_status_addr(struct hinic_api_cmd_chain *chain)
/**
* api_cmd_set_num_cells - set the number cells of a chain in the HW
* @chain: the API CMD specific chain to set the number of cells for
- **/
+ */
static void api_cmd_set_num_cells(struct hinic_api_cmd_chain *chain)
{
struct hinic_hwif *hwif = chain->hwdev->hwif;
@@ -529,7 +529,7 @@ static void api_cmd_set_num_cells(struct hinic_api_cmd_chain *chain)
/**
* api_cmd_head_init - set the head cell of a chain in the HW
* @chain: the API CMD specific chain to set the head for
- **/
+ */
static void api_cmd_head_init(struct hinic_api_cmd_chain *chain)
{
struct hinic_hwif *hwif = chain->hwdev->hwif;
@@ -548,7 +548,7 @@ static void api_cmd_head_init(struct hinic_api_cmd_chain *chain)
* wait_for_ready_chain - wait for the chain to be ready
* @chain: the API CMD specific chain to wait for
* Return: 0 - success, negative - failure
- **/
+ */
static int wait_for_ready_chain(struct hinic_api_cmd_chain *chain)
{
struct hinic_hwif *hwif = chain->hwdev->hwif;
@@ -580,7 +580,7 @@ static int wait_for_ready_chain(struct hinic_api_cmd_chain *chain)
/**
* api_cmd_chain_hw_clean - clean the HW
* @chain: the API CMD specific chain
- **/
+ */
static void api_cmd_chain_hw_clean(struct hinic_api_cmd_chain *chain)
{
struct hinic_hwif *hwif = chain->hwdev->hwif;
@@ -603,7 +603,7 @@ static void api_cmd_chain_hw_clean(struct hinic_api_cmd_chain *chain)
*(initialize API command csr)
* @chain: the API CMD specific chain to initialize in HW
* Return: 0 - success, negative - failure
- **/
+ */
static int api_cmd_chain_hw_init(struct hinic_api_cmd_chain *chain)
{
api_cmd_chain_hw_clean(chain);
@@ -626,7 +626,7 @@ static int api_cmd_chain_hw_init(struct hinic_api_cmd_chain *chain)
* free_cmd_buf - free the dma buffer of API CMD command
* @chain: the API CMD specific chain of the cmd
* @cell_idx: the cell index of the cmd
- **/
+ */
static void free_cmd_buf(struct hinic_api_cmd_chain *chain, u32 cell_idx)
{
struct hinic_api_cmd_cell_ctxt *cell_ctxt;
@@ -645,7 +645,7 @@ static void free_cmd_buf(struct hinic_api_cmd_chain *chain, u32 cell_idx)
* @cell: the cell in the HW for the cmd
* @cell_idx: the index of the cell
* Return: 0 - success, negative - failure
- **/
+ */
static int alloc_cmd_buf(struct hinic_api_cmd_chain *chain,
struct hinic_api_cmd_cell *cell, u32 cell_idx)
{
@@ -697,7 +697,7 @@ static int alloc_cmd_buf(struct hinic_api_cmd_chain *chain,
* @pre_node: previous cell
* @node_vaddr: the virt addr of the cell
* Return: 0 - success, negative - failure
- **/
+ */
static int api_cmd_create_cell(struct hinic_api_cmd_chain *chain,
u32 cell_idx,
struct hinic_api_cmd_cell *pre_node,
@@ -774,7 +774,7 @@ static int api_cmd_create_cell(struct hinic_api_cmd_chain *chain,
* api_cmd_destroy_cell - destroy API CMD cell of specific chain
* @chain: the API CMD specific chain to destroy its cell
* @cell_idx: the cell to destroy
- **/
+ */
static void api_cmd_destroy_cell(struct hinic_api_cmd_chain *chain,
u32 cell_idx)
{
@@ -808,7 +808,7 @@ static void api_cmd_destroy_cell(struct hinic_api_cmd_chain *chain,
* api_cmd_destroy_cells - destroy API CMD cells of specific chain
* @chain: the API CMD specific chain to destroy its cells
* @num_cells: number of cells to destroy
- **/
+ */
static void api_cmd_destroy_cells(struct hinic_api_cmd_chain *chain,
u32 num_cells)
{
@@ -822,7 +822,7 @@ static void api_cmd_destroy_cells(struct hinic_api_cmd_chain *chain,
* api_cmd_create_cells - create API CMD cells for specific chain
* @chain: the API CMD specific chain
* Return: 0 - success, negative - failure
- **/
+ */
static int api_cmd_create_cells(struct hinic_api_cmd_chain *chain)
{
struct hinic_api_cmd_cell *node = NULL, *pre_node = NULL;
@@ -861,7 +861,7 @@ static int api_cmd_create_cells(struct hinic_api_cmd_chain *chain)
* @chain: the API CMD specific chain to initialize
* @attr: attributes to set in the chain
* Return: 0 - success, negative - failure
- **/
+ */
static int api_chain_init(struct hinic_api_cmd_chain *chain,
struct hinic_api_cmd_chain_attr *attr)
{
@@ -910,7 +910,7 @@ static int api_chain_init(struct hinic_api_cmd_chain *chain,
/**
* api_chain_free - free API CMD specific chain
* @chain: the API CMD specific chain to free
- **/
+ */
static void api_chain_free(struct hinic_api_cmd_chain *chain)
{
void *dev = chain->hwdev;
@@ -925,7 +925,7 @@ static void api_chain_free(struct hinic_api_cmd_chain *chain)
* @cmd_chain: the API CMD specific chain to create
* @attr: attributes to set in the chain
* Return: 0 - success, negative - failure
- **/
+ */
static int api_cmd_create_chain(struct hinic_api_cmd_chain **cmd_chain,
struct hinic_api_cmd_chain_attr *attr)
{
@@ -981,7 +981,7 @@ static int api_cmd_create_chain(struct hinic_api_cmd_chain **cmd_chain,
/**
* api_cmd_destroy_chain - destroy API CMD specific chain
* @chain: the API CMD specific chain to destroy
- **/
+ */
static void api_cmd_destroy_chain(struct hinic_api_cmd_chain *chain)
{
api_cmd_destroy_cells(chain, chain->num_cells);
@@ -994,7 +994,7 @@ static void api_cmd_destroy_chain(struct hinic_api_cmd_chain *chain)
* @hwdev: the hardware interface of a pci function device
* @chain: the API CMD chains that will be initialized
* Return: 0 - success, negative - failure
- **/
+ */
int hinic_api_cmd_init(struct hinic_hwdev *hwdev,
struct hinic_api_cmd_chain **chain)
{
@@ -1031,7 +1031,7 @@ int hinic_api_cmd_init(struct hinic_hwdev *hwdev,
/**
* hinic_api_cmd_free - free the API CMD chains
* @chain: the API CMD chains that will be freed
- **/
+ */
void hinic_api_cmd_free(struct hinic_api_cmd_chain **chain)
{
enum hinic_api_cmd_chain_type chain_type;
diff --git a/drivers/net/hinic/base/hinic_pmd_cfg.c b/drivers/net/hinic/base/hinic_pmd_cfg.c
index 61537b1..aa883e0 100644
--- a/drivers/net/hinic/base/hinic_pmd_cfg.c
+++ b/drivers/net/hinic/base/hinic_pmd_cfg.c
@@ -8,6 +8,7 @@
#include "hinic_pmd_mgmt.h"
#include "hinic_pmd_eqs.h"
#include "hinic_pmd_cfg.h"
+#include "hinic_pmd_mbox.h"
bool hinic_support_nic(struct hinic_hwdev *hwdev, struct nic_service_cap *cap)
{
@@ -122,6 +123,10 @@ static void hinic_parse_pub_res_cap(struct service_cap *cap,
cap->vf_id_start = dev_cap->vf_id_start;
cap->max_sqs = dev_cap->nic_max_sq + 1;
cap->max_rqs = dev_cap->nic_max_rq + 1;
+ } else {
+ cap->max_vf = 0;
+ cap->max_sqs = dev_cap->nic_max_sq;
+ cap->max_rqs = dev_cap->nic_max_rq;
}
cap->chip_svc_type = CFG_SVC_NIC_BIT0;
@@ -180,6 +185,28 @@ static int get_cap_from_fw(struct hinic_hwdev *dev, enum func_type type)
return 0;
}
+static int get_cap_from_pf(struct hinic_hwdev *dev, enum func_type type)
+{
+ int err;
+ u16 in_len, out_len;
+ struct hinic_dev_cap dev_cap;
+
+ memset(&dev_cap, 0, sizeof(dev_cap));
+ in_len = sizeof(dev_cap);
+ out_len = in_len;
+ err = hinic_mbox_to_pf(dev, HINIC_MOD_CFGM, HINIC_CFG_MBOX_CAP,
+ &dev_cap, in_len, &dev_cap, &out_len,
+ CFG_MAX_CMD_TIMEOUT);
+ if (err || dev_cap.mgmt_msg_head.status || !out_len) {
+ PMD_DRV_LOG(ERR, "Get capability from PF failed, err: %d, status: %d, out_len: %d",
+ err, dev_cap.mgmt_msg_head.status, out_len);
+ return -EFAULT;
+ }
+
+ parse_dev_cap(dev, &dev_cap, type);
+ return 0;
+}
+
static int get_dev_cap(struct hinic_hwdev *dev)
{
int err;
@@ -194,6 +221,14 @@ static int get_dev_cap(struct hinic_hwdev *dev)
return err;
}
break;
+ case TYPE_VF:
+ err = get_cap_from_pf(dev, type);
+ if (err) {
+ PMD_DRV_LOG(ERR, "Get VF capability failed, err: %d",
+ err);
+ return err;
+ }
+ break;
default:
PMD_DRV_LOG(ERR, "Unsupported PCI function type");
return -EINVAL;
diff --git a/drivers/net/hinic/base/hinic_pmd_cmd.h b/drivers/net/hinic/base/hinic_pmd_cmd.h
index 7a9e9f6..c8750b8 100644
--- a/drivers/net/hinic/base/hinic_pmd_cmd.h
+++ b/drivers/net/hinic/base/hinic_pmd_cmd.h
@@ -25,6 +25,12 @@ enum hinic_mod_type {
HINIC_MOD_MAX = 15
};
+/* only used by VFD communicating with PFD to register or unregister,
+ * command mode type is HINIC_MOD_L2NIC
+ */
+#define HINIC_PORT_CMD_VF_REGISTER 0x0
+#define HINIC_PORT_CMD_VF_UNREGISTER 0x1
+
/* cmd of mgmt CPU message for NIC module */
enum hinic_port_cmd {
HINIC_PORT_CMD_MGMT_RESET = 0x0,
@@ -416,12 +422,12 @@ enum hinic_pf_status {
};
/* total doorbell or direct wqe size is 512kB, db num: 128, dwqe: 128 */
-#define HINIC_DB_DWQE_SIZE 0x00080000
+#define HINIC_DB_DWQE_SIZE 0x00080000
-/* db/dwqe page size: 4K */
-#define HINIC_DB_PAGE_SIZE 0x00001000ULL
+/* db page size: 4K */
+#define HINIC_DB_PAGE_SIZE 0x00001000ULL
-#define HINIC_DB_MAX_AREAS (HINIC_DB_DWQE_SIZE / HINIC_DB_PAGE_SIZE)
+#define HINIC_DB_MAX_AREAS (HINIC_DB_DWQE_SIZE / HINIC_DB_PAGE_SIZE)
#define HINIC_PCI_MSIX_ENTRY_SIZE 16
#define HINIC_PCI_MSIX_ENTRY_VECTOR_CTRL 12
diff --git a/drivers/net/hinic/base/hinic_pmd_hwdev.c b/drivers/net/hinic/base/hinic_pmd_hwdev.c
index 4f70baf..6480d56 100644
--- a/drivers/net/hinic/base/hinic_pmd_hwdev.c
+++ b/drivers/net/hinic/base/hinic_pmd_hwdev.c
@@ -15,6 +15,7 @@
#include "hinic_pmd_cmdq.h"
#include "hinic_pmd_mgmt.h"
#include "hinic_pmd_niccfg.h"
+#include "hinic_pmd_mbox.h"
#define HINIC_DEAULT_EQ_MSIX_PENDING_LIMIT 0
#define HINIC_DEAULT_EQ_MSIX_COALESC_TIMER_CFG 0xFF
@@ -65,6 +66,52 @@
"Unrecognized module",
};
+struct hinic_vf_dma_attr_table {
+ struct hinic_mgmt_msg_head mgmt_msg_head;
+
+ u16 func_idx;
+ u8 func_dma_entry_num;
+ u8 entry_idx;
+ u8 st;
+ u8 at;
+ u8 ph;
+ u8 no_snooping;
+ u8 tph_en;
+ u8 resv1[3];
+};
+
+/**
+ * hinic_cpu_to_be32 - convert data to big endian 32 bit format
+ * @data: the data to convert
+ * @len: length of data to convert, must be Multiple of 4B
+ */
+void hinic_cpu_to_be32(void *data, u32 len)
+{
+ u32 i;
+ u32 *mem = (u32 *)data;
+
+ for (i = 0; i < (len >> 2); i++) {
+ *mem = cpu_to_be32(*mem);
+ mem++;
+ }
+}
+
+/**
+ * hinic_be32_to_cpu - convert data from big endian 32 bit format
+ * @data: the data to convert
+ * @len: length of data to convert, must be Multiple of 4B
+ */
+void hinic_be32_to_cpu(void *data, u32 len)
+{
+ u32 i;
+ u32 *mem = (u32 *)data;
+
+ for (i = 0; i < (len >> 2); i++) {
+ *mem = be32_to_cpu(*mem);
+ mem++;
+ }
+}
+
static void *
hinic_dma_mem_zalloc(struct hinic_hwdev *hwdev, size_t size,
dma_addr_t *dma_handle, unsigned int flag, unsigned int align)
@@ -332,18 +379,6 @@ void hinic_osdep_deinit(struct hinic_hwdev *hwdev)
}
}
-
-
-
-
-
-
-
-
-
-
-
-
/**
* hinic_set_ci_table - set ci attribute table
* @hwdev: the hardware interface of a nic device
@@ -352,7 +387,7 @@ void hinic_osdep_deinit(struct hinic_hwdev *hwdev)
* @return
* 0 on success and ci attribute table is filled,
* negative error value otherwise.
- **/
+ */
int hinic_set_ci_table(void *hwdev, u16 q_id, struct hinic_sq_attr *attr)
{
struct hinic_cons_idx_attr cons_idx_attr;
@@ -385,7 +420,7 @@ int hinic_set_ci_table(void *hwdev, u16 q_id, struct hinic_sq_attr *attr)
* @return
* 0 on success,
* negative error value otherwise.
- **/
+ */
int hinic_set_pagesize(void *hwdev, u8 page_size)
{
struct hinic_page_size cmd;
@@ -417,7 +452,6 @@ static int wait_for_flr_finish(struct hinic_hwif *hwif)
do {
status = hinic_get_pf_status(hwif);
if (status == HINIC_PF_STATUS_FLR_FINISH_FLAG) {
- hinic_set_pf_status(hwif, HINIC_PF_STATUS_ACTIVE_FLAG);
return 0;
}
@@ -463,13 +497,51 @@ static int wait_cmdq_stop(struct hinic_hwdev *hwdev)
return err;
}
+static int hinic_vf_rx_tx_flush(struct hinic_hwdev *hwdev)
+{
+ struct hinic_clear_resource clr_res;
+ int err;
+
+ err = wait_cmdq_stop(hwdev);
+ if (err) {
+ PMD_DRV_LOG(WARNING, "Cmdq is still working");
+ return err;
+ }
+
+ memset(&clr_res, 0, sizeof(clr_res));
+ clr_res.func_idx = HINIC_HWIF_GLOBAL_IDX(hwdev->hwif);
+ clr_res.ppf_idx = HINIC_HWIF_PPF_IDX(hwdev->hwif);
+ err = hinic_mbox_to_pf_no_ack(hwdev, HINIC_MOD_COMM,
+ HINIC_MGMT_CMD_START_FLR, &clr_res, sizeof(clr_res));
+ if (err)
+ PMD_DRV_LOG(WARNING, "Notice flush message failed");
+
+ /*
+ * PF firstly set VF doorbell flush csr to be disabled. After PF finish
+ * VF resources flush, PF will set VF doorbell flush csr to be enabled.
+ */
+ err = wait_until_doorbell_flush_states(hwdev->hwif, DISABLE_DOORBELL);
+ if (err)
+ PMD_DRV_LOG(WARNING, "Wait doorbell flush disable timeout");
+
+ err = wait_until_doorbell_flush_states(hwdev->hwif, ENABLE_DOORBELL);
+ if (err)
+ PMD_DRV_LOG(WARNING, "Wait doorbell flush enable timeout");
+
+ err = hinic_reinit_cmdq_ctxts(hwdev);
+ if (err)
+ PMD_DRV_LOG(WARNING, "Reinit cmdq failed");
+
+ return err;
+}
+
/**
* hinic_pf_rx_tx_flush - clean up hardware resource
* @hwdev: the hardware interface of a nic device
* @return
* 0 on success,
* negative error value otherwise.
- **/
+ */
static int hinic_pf_rx_tx_flush(struct hinic_hwdev *hwdev)
{
struct hinic_hwif *hwif = hwdev->hwif;
@@ -523,7 +595,10 @@ static int hinic_pf_rx_tx_flush(struct hinic_hwdev *hwdev)
int hinic_func_rx_tx_flush(struct hinic_hwdev *hwdev)
{
- return hinic_pf_rx_tx_flush(hwdev);
+ if (HINIC_FUNC_TYPE(hwdev) == TYPE_VF)
+ return hinic_vf_rx_tx_flush(hwdev);
+ else
+ return hinic_pf_rx_tx_flush(hwdev);
}
/**
@@ -531,7 +606,7 @@ int hinic_func_rx_tx_flush(struct hinic_hwdev *hwdev)
* @hwdev: the hardware interface of a nic device
* @interrupt_info: Information of Interrupt aggregation
* Return: 0 on success, negative error value otherwise.
- **/
+ */
static int hinic_get_interrupt_cfg(struct hinic_hwdev *hwdev,
struct nic_interrupt_info *interrupt_info)
{
@@ -567,7 +642,7 @@ static int hinic_get_interrupt_cfg(struct hinic_hwdev *hwdev,
* @hwdev: the hardware interface of a nic device
* @interrupt_info: Information of Interrupt aggregation
* Return: 0 on success, negative error value otherwise.
- **/
+ */
int hinic_set_interrupt_cfg(struct hinic_hwdev *hwdev,
struct nic_interrupt_info interrupt_info)
{
@@ -623,7 +698,7 @@ int hinic_set_interrupt_cfg(struct hinic_hwdev *hwdev,
* @return
* 0 on success,
* negative error value otherwise.
- **/
+ */
int init_aeqs_msix_attr(void *hwdev)
{
struct hinic_hwdev *nic_hwdev = hwdev;
@@ -662,7 +737,7 @@ int init_aeqs_msix_attr(void *hwdev)
* @ph: PCIE TLP Processing Hint field
* @no_snooping: PCIE TLP No snooping
* @tph_en: PCIE TLP Processing Hint Enable
- **/
+ */
static void set_pf_dma_attr_entry(struct hinic_hwdev *hwdev, u32 entry_idx,
u8 st, u8 at, u8 ph,
enum hinic_pcie_nosnoop no_snooping,
@@ -690,28 +765,74 @@ static void set_pf_dma_attr_entry(struct hinic_hwdev *hwdev, u32 entry_idx,
hinic_hwif_write_reg(hwdev->hwif, addr, val);
}
+static int set_vf_dma_attr_entry(struct hinic_hwdev *hwdev, u8 entry_idx,
+ u8 st, u8 at, u8 ph,
+ enum hinic_pcie_nosnoop no_snooping,
+ enum hinic_pcie_tph tph_en)
+{
+ struct hinic_vf_dma_attr_table attr;
+
+ memset(&attr, 0, sizeof(attr));
+ attr.func_idx = hinic_global_func_id(hwdev);
+ attr.mgmt_msg_head.resp_aeq_num = HINIC_AEQ1;
+ attr.func_dma_entry_num = hinic_dma_attr_entry_num(hwdev);
+ attr.entry_idx = entry_idx;
+ attr.st = st;
+ attr.at = at;
+ attr.ph = ph;
+ attr.no_snooping = no_snooping;
+ attr.tph_en = tph_en;
+
+ return hinic_msg_to_mgmt_sync(hwdev, HINIC_MOD_COMM,
+ HINIC_MGMT_CMD_DMA_ATTR_SET,
+ &attr, sizeof(attr), NULL, NULL, 0);
+}
+
/**
* dma_attr_table_init - initialize the the default dma attributes
* @hwdev: the pointer to the private hardware device object
- **/
-static void dma_attr_table_init(struct hinic_hwdev *hwdev)
+ */
+static int dma_attr_table_init(struct hinic_hwdev *hwdev)
{
+ int err = 0;
+
if (HINIC_IS_VF(hwdev))
- return;
+ err = set_vf_dma_attr_entry(hwdev, PCIE_MSIX_ATTR_ENTRY,
+ HINIC_PCIE_ST_DISABLE, HINIC_PCIE_AT_DISABLE,
+ HINIC_PCIE_PH_DISABLE, HINIC_PCIE_SNOOP,
+ HINIC_PCIE_TPH_DISABLE);
+ else
+ set_pf_dma_attr_entry(hwdev, PCIE_MSIX_ATTR_ENTRY,
+ HINIC_PCIE_ST_DISABLE, HINIC_PCIE_AT_DISABLE,
+ HINIC_PCIE_PH_DISABLE, HINIC_PCIE_SNOOP,
+ HINIC_PCIE_TPH_DISABLE);
- set_pf_dma_attr_entry(hwdev, PCIE_MSIX_ATTR_ENTRY,
- HINIC_PCIE_ST_DISABLE,
- HINIC_PCIE_AT_DISABLE,
- HINIC_PCIE_PH_DISABLE,
- HINIC_PCIE_SNOOP,
- HINIC_PCIE_TPH_DISABLE);
+ return err;
}
+/**
+ * hinic_init_attr_table - init dma and aeq msix attribute table
+ * @hwdev: the pointer to the private hardware device object
+ */
int hinic_init_attr_table(struct hinic_hwdev *hwdev)
{
- dma_attr_table_init(hwdev);
+ int err;
- return init_aeqs_msix_attr(hwdev);
+ err = dma_attr_table_init(hwdev);
+ if (err) {
+ PMD_DRV_LOG(ERR, "Initialize dma attribute table failed, err: %d",
+ err);
+ return err;
+ }
+
+ err = init_aeqs_msix_attr(hwdev);
+ if (err) {
+ PMD_DRV_LOG(ERR, "Initialize aeqs msix attribute failed, err: %d",
+ err);
+ return err;
+ }
+
+ return 0;
}
#define FAULT_SHOW_STR_LEN 16
@@ -826,7 +947,7 @@ static int resources_state_set(struct hinic_hwdev *hwdev,
* @return
* 0 on success,
* negative error value otherwise.
- **/
+ */
int hinic_activate_hwdev_state(struct hinic_hwdev *hwdev)
{
int rc = HINIC_OK;
@@ -834,9 +955,7 @@ int hinic_activate_hwdev_state(struct hinic_hwdev *hwdev)
if (!hwdev)
return -EINVAL;
- if (!HINIC_IS_VF(hwdev))
- hinic_set_pf_status(hwdev->hwif,
- HINIC_PF_STATUS_ACTIVE_FLAG);
+ hinic_set_pf_status(hwdev->hwif, HINIC_PF_STATUS_ACTIVE_FLAG);
rc = resources_state_set(hwdev, HINIC_RES_ACTIVE);
if (rc) {
@@ -851,7 +970,7 @@ int hinic_activate_hwdev_state(struct hinic_hwdev *hwdev)
* hinic_deactivate_hwdev_state - Deactivate host nic state and notify mgmt
* channel that host nic is not ready.
* @hwdev: the pointer to the private hardware device object
- **/
+ */
void hinic_deactivate_hwdev_state(struct hinic_hwdev *hwdev)
{
int rc = HINIC_OK;
@@ -863,8 +982,7 @@ void hinic_deactivate_hwdev_state(struct hinic_hwdev *hwdev)
if (rc)
PMD_DRV_LOG(ERR, "Deinit resources state failed");
- if (!HINIC_IS_VF(hwdev))
- hinic_set_pf_status(hwdev->hwif, HINIC_PF_STATUS_INIT);
+ hinic_set_pf_status(hwdev->hwif, HINIC_PF_STATUS_INIT);
}
int hinic_get_board_info(void *hwdev, struct hinic_board_info *info)
@@ -898,7 +1016,7 @@ int hinic_get_board_info(void *hwdev, struct hinic_board_info *info)
* @return
* 0 on success,
* negative error value otherwise.
- **/
+ */
int hinic_l2nic_reset(struct hinic_hwdev *hwdev)
{
struct hinic_hwif *hwif = hwdev->hwif;
diff --git a/drivers/net/hinic/base/hinic_pmd_hwdev.h b/drivers/net/hinic/base/hinic_pmd_hwdev.h
index fc5bdf9..c37771b 100644
--- a/drivers/net/hinic/base/hinic_pmd_hwdev.h
+++ b/drivers/net/hinic/base/hinic_pmd_hwdev.h
@@ -485,4 +485,8 @@ void hinic_hilink_async_event_handle(struct hinic_hwdev *hwdev, u8 cmd,
int hinic_set_pagesize(void *hwdev, u8 page_size);
+void hinic_cpu_to_be32(void *data, u32 len);
+
+void hinic_be32_to_cpu(void *data, u32 len);
+
#endif /* _HINIC_PMD_HWDEV_H_ */
diff --git a/drivers/net/hinic/base/hinic_pmd_hwif.c b/drivers/net/hinic/base/hinic_pmd_hwif.c
index a5e223b..8a01fbc 100644
--- a/drivers/net/hinic/base/hinic_pmd_hwif.c
+++ b/drivers/net/hinic/base/hinic_pmd_hwif.c
@@ -13,6 +13,9 @@
#define HINIC_INTR_MSI_BAR 2
#define HINIC_DB_MEM_BAR 4
+#define PAGE_SIZE_4K 0x1000
+#define PAGE_SIZE_64K 0x10000
+
#define HINIC_MSIX_CNT_RESEND_TIMER_SHIFT 29
#define HINIC_MSIX_CNT_RESEND_TIMER_MASK 0x7U
@@ -24,17 +27,22 @@
* hwif_ready - test if the HW initialization passed
* @hwdev: the pointer to the private hardware device object
* Return: 0 - success, negative - failure
- **/
+ */
static int hwif_ready(struct hinic_hwdev *hwdev)
{
- u32 addr, attr1;
+ u32 addr, attr0, attr1;
addr = HINIC_CSR_FUNC_ATTR1_ADDR;
attr1 = hinic_hwif_read_reg(hwdev->hwif, addr);
-
if (!HINIC_AF1_GET(attr1, MGMT_INIT_STATUS))
return -EBUSY;
+ addr = HINIC_CSR_FUNC_ATTR0_ADDR;
+ attr0 = hinic_hwif_read_reg(hwdev->hwif, addr);
+ if ((HINIC_AF0_GET(attr0, FUNC_TYPE) == TYPE_VF) &&
+ !HINIC_AF1_GET(attr1, PF_INIT_STATUS))
+ return -EBUSY;
+
return 0;
}
@@ -44,7 +52,7 @@ static int hwif_ready(struct hinic_hwdev *hwdev)
* @attr0: the first attribute that was read from the hw
* @attr1: the second attribute that was read from the hw
* @attr2: the third attribute that was read from the hw
- **/
+ */
static void set_hwif_attr(struct hinic_hwif *hwif, u32 attr0, u32 attr1,
u32 attr2)
{
@@ -68,7 +76,7 @@ static void set_hwif_attr(struct hinic_hwif *hwif, u32 attr0, u32 attr1,
/**
* get_hwif_attr - read and set the attributes as members in hwif
* @hwif: the hardware interface of a pci function device
- **/
+ */
static void get_hwif_attr(struct hinic_hwif *hwif)
{
u32 addr, attr0, attr1, attr2;
@@ -90,6 +98,11 @@ void hinic_set_pf_status(struct hinic_hwif *hwif, enum hinic_pf_status status)
u32 attr5 = HINIC_AF5_SET(status, PF_STATUS);
u32 addr = HINIC_CSR_FUNC_ATTR5_ADDR;
+ if (hwif->attr.func_type == TYPE_VF) {
+ PMD_DRV_LOG(ERR, "VF doesn't support set attr5");
+ return;
+ }
+
hinic_hwif_write_reg(hwif, addr, attr5);
}
@@ -145,7 +158,7 @@ void hinic_disable_doorbell(struct hinic_hwif *hwif)
/**
* set_ppf - try to set hwif as ppf and set the type of hwif in this case
* @hwif: the hardware interface of a pci function device
- **/
+ */
static void set_ppf(struct hinic_hwif *hwif)
{
struct hinic_func_attr *attr = &hwif->attr;
@@ -285,6 +298,30 @@ static void disable_all_msix(struct hinic_hwdev *hwdev)
hinic_set_msix_state(hwdev, i, HINIC_MSIX_DISABLE);
}
+/**
+ * Wait for up enable or disable doorbell flush finished.
+ * @hwif: the hardware interface of a pci function device.
+ * @states: Disable or Enable.
+ */
+int wait_until_doorbell_flush_states(struct hinic_hwif *hwif,
+ enum hinic_doorbell_ctrl states)
+{
+ unsigned long end;
+ enum hinic_doorbell_ctrl db_ctrl;
+
+ end = jiffies +
+ msecs_to_jiffies(HINIC_WAIT_DOORBELL_AND_OUTBOUND_TIMEOUT);
+ do {
+ db_ctrl = hinic_get_doorbell_ctrl_status(hwif);
+ if (db_ctrl == states)
+ return 0;
+
+ rte_delay_ms(1);
+ } while (time_before(jiffies, end));
+
+ return -EFAULT;
+}
+
static int wait_until_doorbell_and_outbound_enabled(struct hinic_hwif *hwif)
{
unsigned long end;
@@ -329,6 +366,17 @@ u8 hinic_ppf_idx(void *hwdev)
}
/**
+ * hinic_dma_attr_entry_num - get number id of DMA attribute table.
+ * @hwdev: the pointer to the private hardware device object.
+ * Return: The number id of DMA attribute table.
+ */
+u8 hinic_dma_attr_entry_num(void *hwdev)
+{
+ struct hinic_hwif *hwif = ((struct hinic_hwdev *)hwdev)->hwif;
+ return hwif->attr.num_dma_attr;
+}
+
+/**
* hinic_init_hwif - initialize the hw interface
* @hwdev: the pointer to the private hardware device object
* @cfg_reg_base: base physical address of configuration registers
@@ -337,7 +385,7 @@ u8 hinic_ppf_idx(void *hwdev)
* @db_base: base virtual address of doorbell registers
* @dwqe_mapping: direct wqe io mapping address
* Return: 0 - success, negative - failure
- **/
+ */
static int hinic_init_hwif(struct hinic_hwdev *hwdev, void *cfg_reg_base,
void *intr_reg_base, u64 db_base_phy,
void *db_base, __rte_unused void *dwqe_mapping)
@@ -371,6 +419,9 @@ static int hinic_init_hwif(struct hinic_hwdev *hwdev, void *cfg_reg_base,
if (!HINIC_IS_VF(hwdev))
set_ppf(hwif);
+ /* disable mgmt cpu report any event */
+ hinic_set_pf_status(hwdev->hwif, HINIC_PF_STATUS_INIT);
+
return 0;
hwif_ready_err:
@@ -403,10 +454,28 @@ static void hinic_get_mmio(struct hinic_hwdev *hwdev, void **cfg_regs_base,
void **intr_base, void **db_base)
{
struct rte_pci_device *pci_dev = hwdev->pcidev_hdl;
+ uint64_t bar0_size;
+ uint64_t bar2_size;
+ uint64_t bar0_phy_addr;
+ uint64_t pagesize = sysconf(_SC_PAGESIZE);
*cfg_regs_base = pci_dev->mem_resource[HINIC_CFG_REGS_BAR].addr;
*intr_base = pci_dev->mem_resource[HINIC_INTR_MSI_BAR].addr;
*db_base = pci_dev->mem_resource[HINIC_DB_MEM_BAR].addr;
+
+ bar0_size = pci_dev->mem_resource[HINIC_CFG_REGS_BAR].len;
+ bar2_size = pci_dev->mem_resource[HINIC_INTR_MSI_BAR].len;
+
+ if (pagesize == PAGE_SIZE_64K && (bar0_size % pagesize != 0)) {
+ bar0_phy_addr =
+ pci_dev->mem_resource[HINIC_CFG_REGS_BAR].phys_addr;
+ if (bar0_phy_addr % pagesize != 0 &&
+ (bar0_size + bar2_size <= pagesize) &&
+ bar2_size >= bar0_size) {
+ *cfg_regs_base = (void *)((uint8_t *)(*intr_base)
+ + bar2_size);
+ }
+ }
}
void hinic_hwif_res_free(struct hinic_hwdev *hwdev)
@@ -459,7 +528,7 @@ int hinic_hwif_res_init(struct hinic_hwdev *hwdev)
* @hwdev: the hardware interface of a nic device
* @msix_idx: Index of msix interrupt
* @clear_resend_en: enable flag of clear resend configuration
- **/
+ */
void hinic_misx_intr_clear_resend_bit(void *hwdev, u16 msix_idx,
u8 clear_resend_en)
{
diff --git a/drivers/net/hinic/base/hinic_pmd_hwif.h b/drivers/net/hinic/base/hinic_pmd_hwif.h
index c1289b5..6752587 100644
--- a/drivers/net/hinic/base/hinic_pmd_hwif.h
+++ b/drivers/net/hinic/base/hinic_pmd_hwif.h
@@ -31,7 +31,7 @@ enum hinic_msix_state {
HINIC_MSIX_DISABLE,
};
-/* Defines the IRQ information structure*/
+/* Defines the IRQ information structure */
struct irq_info {
u16 msix_entry_idx; /* IRQ corresponding index number */
u32 irq_id; /* the IRQ number from OS */
@@ -59,11 +59,11 @@ struct hinic_func_attr {
u8 ppf_idx;
- u16 num_irqs; /* max: 2 ^ 15 */
- u8 num_aeqs; /* max: 2 ^ 3 */
- u8 num_ceqs; /* max: 2 ^ 7 */
+ u16 num_irqs; /* max: 2 ^ 15 */
+ u8 num_aeqs; /* max: 2 ^ 3 */
+ u8 num_ceqs; /* max: 2 ^ 7 */
- u8 num_dma_attr; /* max: 2 ^ 6 */
+ u8 num_dma_attr; /* max: 2 ^ 6 */
u16 global_vf_id_of_pf;
};
@@ -104,6 +104,9 @@ static inline u32 hinic_hwif_read_reg(struct hinic_hwif *hwif, u32 reg)
void hinic_free_db_addr(void *hwdev, void __iomem *db_base);
+int wait_until_doorbell_flush_states(struct hinic_hwif *hwif,
+ enum hinic_doorbell_ctrl states);
+
void hinic_set_msix_state(void *hwdev, u16 msix_idx,
enum hinic_msix_state flag);
@@ -116,4 +119,6 @@ void hinic_misx_intr_clear_resend_bit(void *hwdev, u16 msix_idx,
void hinic_hwif_res_free(struct hinic_hwdev *hwdev);
+u8 hinic_dma_attr_entry_num(void *hwdev);
+
#endif /* _HINIC_PMD_HWIF_H_ */
diff --git a/drivers/net/hinic/base/hinic_pmd_mgmt.c b/drivers/net/hinic/base/hinic_pmd_mgmt.c
index a18e567..eee50a8 100644
--- a/drivers/net/hinic/base/hinic_pmd_mgmt.c
+++ b/drivers/net/hinic/base/hinic_pmd_mgmt.c
@@ -7,6 +7,7 @@
#include "hinic_pmd_hwdev.h"
#include "hinic_pmd_hwif.h"
#include "hinic_pmd_mgmt.h"
+#include "hinic_pmd_mbox.h"
#define BUF_OUT_DEFAULT_SIZE 1
@@ -72,7 +73,7 @@
* mgmt_msg_len - calculate the total message length
* @msg_data_len: the length of the message data
* Return: the total message length
- **/
+ */
static u16 mgmt_msg_len(u16 msg_data_len)
{
/* u64 - the size of the header */
@@ -99,7 +100,7 @@ static u16 mgmt_msg_len(u16 msg_data_len)
* @direction: the direction of the original message
* @cmd: the command to do
* @msg_id: message id
- **/
+ */
static void prepare_header(struct hinic_msg_pf_to_mgmt *pf_to_mgmt,
u64 *header, int msg_len, enum hinic_mod_type mod,
enum hinic_msg_ack_type ack_type,
@@ -128,7 +129,7 @@ static void prepare_header(struct hinic_msg_pf_to_mgmt *pf_to_mgmt,
* @header: pointer of the header to prepare
* @msg: the data of the message
* @msg_len: the length of the message
- **/
+ */
static void prepare_mgmt_cmd(u8 *mgmt_cmd, u64 *header, void *msg,
int msg_len)
{
@@ -149,7 +150,7 @@ static void prepare_mgmt_cmd(u8 *mgmt_cmd, u64 *header, void *msg,
* alloc_recv_msg - allocate received message memory
* @recv_msg: pointer that will hold the allocated data
* Return: 0 - success, negative - failure
- **/
+ */
static int alloc_recv_msg(struct hinic_recv_msg *recv_msg)
{
int err;
@@ -177,7 +178,7 @@ static int alloc_recv_msg(struct hinic_recv_msg *recv_msg)
/**
* free_recv_msg - free received message memory
* @recv_msg: pointer that holds the allocated data
- **/
+ */
static void free_recv_msg(struct hinic_recv_msg *recv_msg)
{
kfree(recv_msg->buf_out);
@@ -188,7 +189,7 @@ static void free_recv_msg(struct hinic_recv_msg *recv_msg)
* alloc_msg_buf - allocate all the message buffers of PF to MGMT channel
* @pf_to_mgmt: PF to MGMT channel
* Return: 0 - success, negative - failure
- **/
+ */
static int alloc_msg_buf(struct hinic_msg_pf_to_mgmt *pf_to_mgmt)
{
int err;
@@ -237,7 +238,7 @@ static int alloc_msg_buf(struct hinic_msg_pf_to_mgmt *pf_to_mgmt)
* free_msg_buf - free all the message buffers of PF to MGMT channel
* @pf_to_mgmt: PF to MGMT channel
* Return: 0 - success, negative - failure
- **/
+ */
static void free_msg_buf(struct hinic_msg_pf_to_mgmt *pf_to_mgmt)
{
kfree(pf_to_mgmt->sync_msg_buf);
@@ -257,7 +258,7 @@ static void free_msg_buf(struct hinic_msg_pf_to_mgmt *pf_to_mgmt)
* @direction: the direction of the original message
* @resp_msg_id: message id of response
* Return: 0 - success, negative - failure
- **/
+ */
static int send_msg_to_mgmt_async(struct hinic_msg_pf_to_mgmt *pf_to_mgmt,
enum hinic_mod_type mod, u8 cmd,
void *msg, u16 msg_len,
@@ -295,7 +296,7 @@ static int send_msg_to_mgmt_async(struct hinic_msg_pf_to_mgmt *pf_to_mgmt,
* @direction: the direction of the original message
* @resp_msg_id: msg id to response for
* Return: 0 - success, negative - failure
- **/
+ */
static int send_msg_to_mgmt_sync(struct hinic_msg_pf_to_mgmt *pf_to_mgmt,
enum hinic_mod_type mod, u8 cmd,
void *msg, u16 msg_len,
@@ -327,7 +328,7 @@ static int send_msg_to_mgmt_sync(struct hinic_msg_pf_to_mgmt *pf_to_mgmt,
* hinic_pf_to_mgmt_init - initialize PF to MGMT channel
* @hwdev: the pointer to the private hardware device object
* Return: 0 - success, negative - failure
- **/
+ */
static int hinic_pf_to_mgmt_init(struct hinic_hwdev *hwdev)
{
struct hinic_msg_pf_to_mgmt *pf_to_mgmt;
@@ -342,7 +343,7 @@ static int hinic_pf_to_mgmt_init(struct hinic_hwdev *hwdev)
hwdev->pf_to_mgmt = pf_to_mgmt;
pf_to_mgmt->hwdev = hwdev;
- err = hinic_mutex_init(&pf_to_mgmt->sync_msg_lock, NULL);
+ err = hinic_mutex_init(&pf_to_mgmt->sync_msg_mutex, NULL);
if (err)
goto mutex_init_err;
@@ -364,7 +365,7 @@ static int hinic_pf_to_mgmt_init(struct hinic_hwdev *hwdev)
free_msg_buf(pf_to_mgmt);
alloc_msg_buf_err:
- hinic_mutex_destroy(&pf_to_mgmt->sync_msg_lock);
+ hinic_mutex_destroy(&pf_to_mgmt->sync_msg_mutex);
mutex_init_err:
kfree(pf_to_mgmt);
@@ -375,14 +376,14 @@ static int hinic_pf_to_mgmt_init(struct hinic_hwdev *hwdev)
/**
* hinic_pf_to_mgmt_free - free PF to MGMT channel
* @hwdev: the pointer to the private hardware device object
- **/
+ */
static void hinic_pf_to_mgmt_free(struct hinic_hwdev *hwdev)
{
struct hinic_msg_pf_to_mgmt *pf_to_mgmt = hwdev->pf_to_mgmt;
hinic_api_cmd_free(pf_to_mgmt->cmd_chain);
free_msg_buf(pf_to_mgmt);
- hinic_mutex_destroy(&pf_to_mgmt->sync_msg_lock);
+ hinic_mutex_destroy(&pf_to_mgmt->sync_msg_mutex);
kfree(pf_to_mgmt);
}
@@ -396,7 +397,9 @@ static void hinic_pf_to_mgmt_free(struct hinic_hwdev *hwdev)
u32 timeo;
int err, i;
- pthread_mutex_lock(&pf_to_mgmt->sync_msg_lock);
+ err = hinic_mutex_lock(&pf_to_mgmt->sync_msg_mutex);
+ if (err)
+ return err;
SYNC_MSG_ID_INC(pf_to_mgmt);
recv_msg = &pf_to_mgmt->recv_resp_msg_from_mgmt;
@@ -455,7 +458,7 @@ static void hinic_pf_to_mgmt_free(struct hinic_hwdev *hwdev)
unlock_sync_msg:
if (err && out_size)
*out_size = 0;
- pthread_mutex_unlock(&pf_to_mgmt->sync_msg_lock);
+ (void)hinic_mutex_unlock(&pf_to_mgmt->sync_msg_mutex);
return err;
}
@@ -464,6 +467,9 @@ static int hinic_get_mgmt_channel_status(void *hwdev)
struct hinic_hwif *hwif = ((struct hinic_hwdev *)hwdev)->hwif;
u32 val;
+ if (hinic_func_type((struct hinic_hwdev *)hwdev) == TYPE_VF)
+ return false;
+
val = hinic_hwif_read_reg(hwif, HINIC_ICPL_RESERVD_ADDR);
return HINIC_GET_MGMT_CHANNEL_STATUS(val, MGMT_CHANNEL_STATUS);
@@ -482,9 +488,13 @@ int hinic_msg_to_mgmt_sync(void *hwdev, enum hinic_mod_type mod, u8 cmd,
if (hinic_get_mgmt_channel_status(hwdev))
return -EPERM;
- rc = hinic_pf_to_mgmt_sync(hwdev, mod, cmd, buf_in,
- in_size, buf_out, out_size,
- timeout);
+ if (hinic_func_type(hwdev) == TYPE_VF) {
+ rc = hinic_mbox_to_pf(hwdev, mod, cmd, buf_in, in_size,
+ buf_out, out_size, timeout);
+ } else {
+ rc = hinic_pf_to_mgmt_sync(hwdev, mod, cmd, buf_in, in_size,
+ buf_out, out_size, timeout);
+ }
return rc;
}
@@ -502,13 +512,15 @@ int hinic_msg_to_mgmt_no_ack(void *hwdev, enum hinic_mod_type mod, u8 cmd,
return err;
}
- pthread_mutex_lock(&pf_to_mgmt->sync_msg_lock);
+ err = hinic_mutex_lock(&pf_to_mgmt->sync_msg_mutex);
+ if (err)
+ return err;
err = send_msg_to_mgmt_sync(pf_to_mgmt, mod, cmd, buf_in, in_size,
HINIC_MSG_NO_ACK, HINIC_MSG_DIRECT_SEND,
MSG_NO_RESP);
- pthread_mutex_unlock(&pf_to_mgmt->sync_msg_lock);
+ (void)hinic_mutex_unlock(&pf_to_mgmt->sync_msg_mutex);
return err;
}
@@ -537,7 +549,7 @@ static bool check_mgmt_seq_id_and_seg_len(struct hinic_recv_msg *recv_msg,
* @pf_to_mgmt: PF to MGMT channel
* @recv_msg: received message details
* @param: customized parameter
- **/
+ */
static void hinic_mgmt_recv_msg_handler(struct hinic_msg_pf_to_mgmt *pf_to_mgmt,
struct hinic_recv_msg *recv_msg,
void *param)
@@ -589,7 +601,7 @@ static void hinic_mgmt_recv_msg_handler(struct hinic_msg_pf_to_mgmt *pf_to_mgmt,
* @param: customized parameter
* Return: 0 when aeq is response message, -1 default result,
* and when wrong message or not last message
- **/
+ */
static int recv_mgmt_msg_handler(struct hinic_msg_pf_to_mgmt *pf_to_mgmt,
u8 *header, struct hinic_recv_msg *recv_msg,
void *param)
@@ -609,7 +621,7 @@ static int recv_mgmt_msg_handler(struct hinic_msg_pf_to_mgmt *pf_to_mgmt,
"func id: 0x%x, front id: 0x%x, current id: 0x%x, seg len: 0x%x",
hinic_global_func_id(pf_to_mgmt->hwdev),
recv_msg->sed_id, seq_id, seq_len);
- return HINIC_RECV_NEXT_AEQE;
+ return HINIC_ERROR;
}
dest_msg = (u8 *)recv_msg->msg + seq_id * HINIC_MSG_SEG_LEN;
@@ -617,7 +629,7 @@ static int recv_mgmt_msg_handler(struct hinic_msg_pf_to_mgmt *pf_to_mgmt,
memcpy(dest_msg, msg_body, seq_len);
if (!HINIC_MSG_HEADER_GET(msg_header, LAST))
- return HINIC_RECV_NEXT_AEQE;
+ return HINIC_ERROR;
recv_msg->cmd = HINIC_MSG_HEADER_GET(msg_header, CMD);
recv_msg->mod = HINIC_MSG_HEADER_GET(msg_header, MODULE);
@@ -627,11 +639,11 @@ static int recv_mgmt_msg_handler(struct hinic_msg_pf_to_mgmt *pf_to_mgmt,
recv_msg->msg_id = HINIC_MSG_HEADER_GET(msg_header, MSG_ID);
if (HINIC_MSG_HEADER_GET(msg_header, DIRECTION) == HINIC_MSG_RESPONSE)
- return HINIC_RECV_DONE;
+ return HINIC_OK;
hinic_mgmt_recv_msg_handler(pf_to_mgmt, recv_msg, param);
- return HINIC_RECV_NEXT_AEQE;
+ return HINIC_ERROR;
}
/**
@@ -642,7 +654,7 @@ static int recv_mgmt_msg_handler(struct hinic_msg_pf_to_mgmt *pf_to_mgmt,
* @param: customized parameter
* Return: 0 when aeq is response message,
* -1 default result, and when wrong message or not last message
- **/
+ */
static int hinic_mgmt_msg_aeqe_handler(void *hwdev, u8 *header,
__rte_unused u8 size, void *param)
{
@@ -667,10 +679,13 @@ static int hinic_handle_aeqe(void *handle, enum hinic_aeq_type event,
case HINIC_MSG_FROM_MGMT_CPU:
rc = hinic_mgmt_msg_aeqe_handler(handle, data, size, param);
break;
+ case HINIC_MBX_FROM_FUNC:
+ rc = hinic_mbox_func_aeqe_handler(handle, data, size, param);
+ break;
default:
PMD_DRV_LOG(ERR, "Unknown event type: 0x%x, size: %d",
event, size);
- rc = HINIC_RECV_NEXT_AEQE;
+ rc = HINIC_ERROR;
break;
}
@@ -685,7 +700,7 @@ static int hinic_handle_aeqe(void *handle, enum hinic_aeq_type event,
* used in polling mode.
* @param: customized parameter
* Return: 0 - Success, EIO - poll timeout, ENODEV - swe not support
- **/
+ */
int hinic_aeq_poll_msg(struct hinic_eq *eq, u32 timeout, void *param)
{
struct hinic_aeq_elem *aeqe_pos;
@@ -753,6 +768,10 @@ int hinic_comm_pf_to_mgmt_init(struct hinic_hwdev *hwdev)
{
int rc;
+ /* VF do not support send msg to mgmt directly */
+ if (hinic_func_type(hwdev) == TYPE_VF)
+ return 0;
+
rc = hinic_pf_to_mgmt_init(hwdev);
if (rc)
return rc;
@@ -764,6 +783,10 @@ int hinic_comm_pf_to_mgmt_init(struct hinic_hwdev *hwdev)
void hinic_comm_pf_to_mgmt_free(struct hinic_hwdev *hwdev)
{
+ /* VF do not support send msg to mgmt directly */
+ if (hinic_func_type(hwdev) == TYPE_VF)
+ return;
+
hinic_pf_to_mgmt_free(hwdev);
}
diff --git a/drivers/net/hinic/base/hinic_pmd_mgmt.h b/drivers/net/hinic/base/hinic_pmd_mgmt.h
index 7804708..cc18843 100644
--- a/drivers/net/hinic/base/hinic_pmd_mgmt.h
+++ b/drivers/net/hinic/base/hinic_pmd_mgmt.h
@@ -82,7 +82,7 @@ struct hinic_msg_pf_to_mgmt {
struct hinic_hwdev *hwdev;
/* mutex for sync message */
- pthread_mutex_t sync_msg_lock;
+ pthread_mutex_t sync_msg_mutex;
void *async_msg_buf;
void *sync_msg_buf;
diff --git a/drivers/net/hinic/base/hinic_pmd_niccfg.c b/drivers/net/hinic/base/hinic_pmd_niccfg.c
index 78012b8..d3bbc3d 100644
--- a/drivers/net/hinic/base/hinic_pmd_niccfg.c
+++ b/drivers/net/hinic/base/hinic_pmd_niccfg.c
@@ -10,6 +10,7 @@
#include "hinic_pmd_mgmt.h"
#include "hinic_pmd_cmdq.h"
#include "hinic_pmd_niccfg.h"
+#include "hinic_pmd_mbox.h"
#define l2nic_msg_to_mgmt_sync(hwdev, cmd, buf_in, \
in_size, buf_out, out_size) \
@@ -17,6 +18,18 @@
buf_in, in_size, \
buf_out, out_size, 0)
+/**
+ * hinic_init_function_table - Initialize function table.
+ *
+ * @param hwdev
+ * The hardware interface of a nic device.
+ * @param rx_buf_sz
+ * Receive buffer size.
+ *
+ * @return
+ * 0 on success.
+ * negative error value otherwise.
+ */
int hinic_init_function_table(void *hwdev, u16 rx_buf_sz)
{
struct hinic_function_table function_table;
@@ -40,8 +53,8 @@ int hinic_init_function_table(void *hwdev, u16 rx_buf_sz)
&function_table, &out_size, 0);
if (err || function_table.mgmt_msg_head.status || !out_size) {
PMD_DRV_LOG(ERR,
- "Failed to init func table, ret = %d",
- function_table.mgmt_msg_head.status);
+ "Failed to init func table, err: %d, status: 0x%x, out size: 0x%x",
+ err, function_table.mgmt_msg_head.status, out_size);
return -EFAULT;
}
@@ -49,13 +62,17 @@ int hinic_init_function_table(void *hwdev, u16 rx_buf_sz)
}
/**
- * hinic_get_base_qpn - get global number of queue
- * @hwdev: the hardware interface of a nic device
- * @global_qpn: vat page size
+ * hinic_get_base_qpn - Get global queue number.
+ *
+ * @param hwdev
+ * The hardware interface of a nic device.
+ * @param global_qpn
+ * Global queue number.
+ *
* @return
- * 0 on success,
+ * 0 on success.
* negative error value otherwise.
- **/
+ */
int hinic_get_base_qpn(void *hwdev, u16 *global_qpn)
{
struct hinic_cmd_qpn cmd_qpn;
@@ -77,8 +94,8 @@ int hinic_get_base_qpn(void *hwdev, u16 *global_qpn)
&out_size, 0);
if (err || !out_size || cmd_qpn.mgmt_msg_head.status) {
PMD_DRV_LOG(ERR,
- "Failed to get base qpn, status(%d)",
- cmd_qpn.mgmt_msg_head.status);
+ "Failed to get base qpn, err: %d, status: 0x%x, out size: 0x%x",
+ err, cmd_qpn.mgmt_msg_head.status, out_size);
return -EINVAL;
}
@@ -89,12 +106,18 @@ int hinic_get_base_qpn(void *hwdev, u16 *global_qpn)
/**
* hinic_set_mac - Init mac_vlan table in NIC.
- * @hwdev: the hardware interface of a nic device
- * @mac_addr: mac address
- * @vlan_id: set 0 for mac_vlan table initialization
- * @func_id: global function id of NIC
+ *
+ * @param hwdev
+ * The hardware interface of a nic device.
+ * @param mac_addr
+ * MAC address.
+ * @param vlan_id
+ * Set 0 for mac_vlan table initialization.
+ * @param func_id
+ * Global function id of NIC.
+ *
* @return
- * 0 on success and stats is filled,
+ * 0 on success.
* negative error value otherwise.
*/
int hinic_set_mac(void *hwdev, u8 *mac_addr, u16 vlan_id, u16 func_id)
@@ -122,6 +145,7 @@ int hinic_set_mac(void *hwdev, u8 *mac_addr, u16 vlan_id, u16 func_id)
err, mac_info.mgmt_msg_head.status, out_size);
return -EINVAL;
}
+
if (mac_info.mgmt_msg_head.status == HINIC_PF_SET_VF_ALREADY) {
PMD_DRV_LOG(WARNING, "PF has already set vf mac, Ignore set operation.");
return HINIC_PF_SET_VF_ALREADY;
@@ -132,16 +156,21 @@ int hinic_set_mac(void *hwdev, u8 *mac_addr, u16 vlan_id, u16 func_id)
/**
* hinic_del_mac - Uninit mac_vlan table in NIC.
- * @hwdev: the hardware interface of a nic device
- * @mac_addr: mac address
- * @vlan_id: set 0 for mac_vlan table initialization
- * @func_id: global function id of NIC
+ *
+ * @param hwdev
+ * The hardware interface of a nic device.
+ * @param mac_addr
+ * MAC address.
+ * @param vlan_id
+ * Set 0 for mac_vlan table initialization.
+ * @param func_id
+ * Global function id of NIC.
+ *
* @return
- * 0 on success and stats is filled,
+ * 0 on success.
* negative error value otherwise.
*/
-int hinic_del_mac(void *hwdev, u8 *mac_addr, u16 vlan_id,
- u16 func_id)
+int hinic_del_mac(void *hwdev, u8 *mac_addr, u16 vlan_id, u16 func_id)
{
struct hinic_port_mac_set mac_info;
u16 out_size = sizeof(mac_info);
@@ -179,6 +208,18 @@ int hinic_del_mac(void *hwdev, u8 *mac_addr, u16 vlan_id,
return 0;
}
+/**
+ * hinic_get_default_mac - Get default mac address from hardware.
+ *
+ * @param hwdev
+ * The hardware interface of a nic device.
+ * @param mac_addr
+ * MAC address.
+ *
+ * @return
+ * 0 on success.
+ * negative error value otherwise.
+ */
int hinic_get_default_mac(void *hwdev, u8 *mac_addr)
{
struct hinic_port_mac_set mac_info;
@@ -208,6 +249,18 @@ int hinic_get_default_mac(void *hwdev, u8 *mac_addr)
return 0;
}
+/**
+ * hinic_set_port_mtu - Set MTU to port.
+ *
+ * @param hwdev
+ * The hardware interface of a nic device.
+ * @param new_mtu
+ * MTU size.
+ *
+ * @return
+ * 0 on success.
+ * negative error value otherwise.
+ */
int hinic_set_port_mtu(void *hwdev, u32 new_mtu)
{
struct hinic_mtu mtu_info;
@@ -236,6 +289,18 @@ int hinic_set_port_mtu(void *hwdev, u32 new_mtu)
return 0;
}
+/**
+ * hinic_get_link_status - Get link status from hardware.
+ *
+ * @param hwdev
+ * The hardware interface of a nic device.
+ * @param link_state
+ * Link status.
+ *
+ * @return
+ * 0 on success.
+ * negative error value otherwise.
+ */
int hinic_get_link_status(void *hwdev, u8 *link_state)
{
struct hinic_get_link get_link;
@@ -267,10 +332,16 @@ int hinic_get_link_status(void *hwdev, u8 *link_state)
/**
* hinic_set_vport_enable - Notify firmware that driver is ready or not.
- * @hwdev: the hardware interface of a nic device
- * @enable: 1: driver is ready; 0: driver is not ok.
- * Return: 0 on success and state is filled, negative error value otherwise.
- **/
+ *
+ * @param hwdev
+ * The hardware interface of a nic device.
+ * @param enable
+ * 1: driver is ready; 0: driver is not ok.
+ *
+ * @return
+ * 0 on success.
+ * negative error value otherwise.
+ */
int hinic_set_vport_enable(void *hwdev, bool enable)
{
struct hinic_vport_state en_state;
@@ -300,11 +371,15 @@ int hinic_set_vport_enable(void *hwdev, bool enable)
}
/**
- * hinic_set_port_enable - open MAG to receive packets.
- * @hwdev: the hardware interface of a nic device
- * @enable: 1: open MAG; 0: close MAG.
+ * hinic_set_port_enable - Open MAG to receive packets.
+ *
+ * @param hwdev
+ * The hardware interface of a nic device.
+ * @param enable
+ * 1: open MAG; 0: close MAG.
+ *
* @return
- * 0 on success and stats is filled,
+ * 0 on success.
* negative error value otherwise.
*/
int hinic_set_port_enable(void *hwdev, bool enable)
@@ -318,6 +393,9 @@ int hinic_set_port_enable(void *hwdev, bool enable)
return -EINVAL;
}
+ if (HINIC_IS_VF((struct hinic_hwdev *)hwdev))
+ return 0;
+
memset(&en_state, 0, sizeof(en_state));
en_state.mgmt_msg_head.resp_aeq_num = HINIC_AEQ1;
en_state.state = (enable ? HINIC_PORT_ENABLE : HINIC_PORT_DISABLE);
@@ -432,7 +510,7 @@ int hinic_dcb_set_ets(void *hwdev, u8 *up_tc, u8 *pg_bw,
memset(&ets, 0, sizeof(ets));
ets.mgmt_msg_head.resp_aeq_num = HINIC_AEQ1;
- ets.port_id = 0; /* reserved */
+ ets.port_id = 0; /* reserved */
memcpy(ets.up_tc, up_tc, HINIC_DCB_TC_MAX);
memcpy(ets.pg_bw, pg_bw, HINIC_DCB_UP_MAX);
memcpy(ets.pgid, pgid, HINIC_DCB_UP_MAX);
@@ -672,12 +750,17 @@ int hinic_rss_get_template_tbl(void *hwdev, u32 tmpl_idx, u8 *temp)
}
/**
- * hinic_rss_set_hash_engine - Init rss hash function .
- * @hwdev: the hardware interface of a nic device
- * @tmpl_idx: index of rss template from NIC.
- * @type: hash function, such as Toeplitz or XOR.
+ * hinic_rss_set_hash_engine - Init rss hash function.
+ *
+ * @param hwdev
+ * The hardware interface of a nic device.
+ * @param tmpl_idx
+ * Index of rss template from NIC.
+ * @param type
+ * Hash function, such as Toeplitz or XOR.
+ *
* @return
- * 0 on success and stats is filled,
+ * 0 on success.
* negative error value otherwise.
*/
int hinic_rss_set_hash_engine(void *hwdev, u8 tmpl_idx, u8 type)
@@ -845,12 +928,18 @@ int hinic_rss_cfg(void *hwdev, u8 rss_en, u8 tmpl_idx, u8 tc_num, u8 *prio_tc)
}
/**
- * hinic_rss_template_alloc - get rss template id from the chip,
- * all functions share 96 templates.
- * @hwdev: the pointer to the private hardware device object
- * @tmpl_idx: index of rss template from chip.
- * Return: 0 on success and stats is filled, negative error value otherwise.
- **/
+ * hinic_rss_template_alloc - Get rss template id from the chip,
+ * all functions share 96 templates.
+ *
+ * @param hwdev
+ * The hardware interface of a nic device.
+ * @param tmpl_idx
+ * Index of rss template from chip.
+ *
+ * @return
+ * 0 on success.
+ * negative error value otherwise.
+ */
int hinic_rss_template_alloc(void *hwdev, u8 *tmpl_idx)
{
struct hinic_rss_template_mgmt template_mgmt;
@@ -882,11 +971,17 @@ int hinic_rss_template_alloc(void *hwdev, u8 *tmpl_idx)
}
/**
- * hinic_rss_template_alloc - free rss template id to the chip
- * @hwdev: the hardware interface of a nic device
- * @tmpl_idx: index of rss template from NIC.
- * Return: 0 on success and stats is filled, negative error value otherwise.
- **/
+ * hinic_rss_template_free - Free rss template id to the chip.
+ *
+ * @param hwdev
+ * The hardware interface of a nic device.
+ * @param tmpl_idx
+ * Index of rss template from chip.
+ *
+ * @return
+ * 0 on success.
+ * negative error value otherwise.
+ */
int hinic_rss_template_free(void *hwdev, u8 tmpl_idx)
{
struct hinic_rss_template_mgmt template_mgmt;
@@ -917,12 +1012,17 @@ int hinic_rss_template_free(void *hwdev, u8 tmpl_idx)
}
/**
- * hinic_set_rx_vhd_mode - change rx buffer size after initialization,
- * @hwdev: the hardware interface of a nic device
- * @mode: not needed.
- * @rx_buf_sz: receive buffer size.
+ * hinic_set_rx_vhd_mode - Change rx buffer size after initialization.
+ *
+ * @param hwdev
+ * The hardware interface of a nic device.
+ * @param vhd_mode
+ * Not needed.
+ * @param rx_buf_sz
+ * receive buffer size.
+ *
* @return
- * 0 on success and stats is filled,
+ * 0 on success.
* negative error value otherwise.
*/
int hinic_set_rx_vhd_mode(void *hwdev, u16 vhd_mode, u16 rx_buf_sz)
@@ -1100,6 +1200,54 @@ int hinic_reset_port_link_cfg(void *hwdev)
return 0;
}
+/**
+ * hinic_vf_func_init - Register VF to PF.
+ *
+ * @param hwdev
+ * The hardware interface of a nic device.
+ *
+ * @return
+ * 0 on success.
+ * negative error value otherwise.
+ */
+int hinic_vf_func_init(struct hinic_hwdev *hwdev)
+{
+ int err, state = 0;
+
+ if (!HINIC_IS_VF(hwdev))
+ return 0;
+
+ err = hinic_mbox_to_pf(hwdev, HINIC_MOD_L2NIC,
+ HINIC_PORT_CMD_VF_REGISTER, &state, sizeof(state),
+ NULL, NULL, 0);
+ if (err) {
+ PMD_DRV_LOG(ERR, "Fail to register vf");
+ return err;
+ }
+
+ return 0;
+}
+
+/**
+ * hinic_vf_func_free - Unregister VF from PF.
+ *
+ * @param hwdev
+ * The hardware interface of a nic device.
+ */
+void hinic_vf_func_free(struct hinic_hwdev *hwdev)
+{
+ int err;
+
+ if (hinic_func_type(hwdev) != TYPE_VF)
+ return;
+
+ err = hinic_mbox_to_pf(hwdev, HINIC_MOD_L2NIC,
+ HINIC_PORT_CMD_VF_UNREGISTER, &err, sizeof(err),
+ NULL, NULL, 0);
+ if (err)
+ PMD_DRV_LOG(ERR, "Fail to unregister VF, err: %d", err);
+}
+
int hinic_set_fast_recycle_mode(void *hwdev, u8 mode)
{
struct hinic_fast_recycled_mode fast_recycled_mode;
@@ -1199,6 +1347,9 @@ int hinic_set_link_status_follow(void *hwdev,
if (!hwdev)
return -EINVAL;
+ if (HINIC_IS_VF((struct hinic_hwdev *)hwdev))
+ return 0;
+
if (status >= HINIC_LINK_FOLLOW_STATUS_MAX) {
PMD_DRV_LOG(ERR,
"Invalid link follow status: %d", status);
@@ -1254,11 +1405,16 @@ int hinic_get_link_mode(void *hwdev, u32 *supported, u32 *advertised)
}
/**
- * hinic_flush_qp_res - Flush tx && rx chip resources in case of set vport fake
- * failed when device start.
- * @hwdev: the hardware interface of a nic device
- * Return: 0 on success, negative error value otherwise.
- **/
+ * hinic_flush_qp_res - Flush tx && rx chip resources in case of set vport
+ * fake failed when device start.
+ *
+ * @param hwdev
+ * The hardware interface of a nic device.
+ *
+ * @return
+ * 0 on success.
+ * negative error value otherwise.
+ */
int hinic_flush_qp_res(void *hwdev)
{
struct hinic_clear_qp_resource qp_res;
diff --git a/drivers/net/hinic/base/hinic_pmd_niccfg.h b/drivers/net/hinic/base/hinic_pmd_niccfg.h
index 174b40f..dcb9602 100644
--- a/drivers/net/hinic/base/hinic_pmd_niccfg.h
+++ b/drivers/net/hinic/base/hinic_pmd_niccfg.h
@@ -655,4 +655,8 @@ int hinic_set_link_status_follow(void *hwdev,
int hinic_clear_phy_port_stats(struct hinic_hwdev *hwdev);
+int hinic_vf_func_init(struct hinic_hwdev *hwdev);
+
+void hinic_vf_func_free(struct hinic_hwdev *hwdev);
+
#endif /* _HINIC_PMD_NICCFG_H_ */
diff --git a/drivers/net/hinic/base/hinic_pmd_nicio.c b/drivers/net/hinic/base/hinic_pmd_nicio.c
index 248211f..f6cc033 100644
--- a/drivers/net/hinic/base/hinic_pmd_nicio.c
+++ b/drivers/net/hinic/base/hinic_pmd_nicio.c
@@ -442,7 +442,7 @@ static int clean_qp_offload_ctxt(struct hinic_nic_io *nic_io)
* @rx_buf_sz: receive buffer size
* @return
* hw rx buffer size
- **/
+ */
static u16 get_hw_rx_buf_size(u32 rx_buf_sz)
{
u16 num_hw_types = sizeof(hinic_hw_rx_buf_size)
@@ -466,7 +466,7 @@ static u16 get_hw_rx_buf_size(u32 rx_buf_sz)
* @sq_depth: the depth of transmit queue
* @rx_buf_sz: receive buffer size from app
* Return: 0 on success, negative error value otherwise.
- **/
+ */
static int
hinic_set_root_ctxt(void *hwdev, u16 rq_depth, u16 sq_depth, int rx_buf_sz)
{
@@ -495,7 +495,7 @@ static u16 get_hw_rx_buf_size(u32 rx_buf_sz)
* @return
* 0 on success,
* negative error value otherwise.
- **/
+ */
static int hinic_clean_root_ctxt(void *hwdev)
{
struct hinic_root_ctxt root_ctxt;
@@ -618,6 +618,12 @@ static int hinic_init_nic_hwdev(struct hinic_hwdev *hwdev)
goto err_init_nic_hwdev;
}
+ err = hinic_vf_func_init(hwdev);
+ if (err) {
+ PMD_DRV_LOG(ERR, "Failed to init nic mbox");
+ goto err_init_nic_hwdev;
+ }
+
err = hinic_set_fast_recycle_mode(hwdev, RECYCLE_MODE_DPDK);
if (err) {
PMD_DRV_LOG(ERR, "Failed to set fast recycle mode");
@@ -632,6 +638,7 @@ static int hinic_init_nic_hwdev(struct hinic_hwdev *hwdev)
static void hinic_free_nic_hwdev(struct hinic_hwdev *hwdev)
{
+ hinic_vf_func_free(hwdev);
hwdev->nic_io = NULL;
}
@@ -861,7 +868,7 @@ void hinic_deinit_nicio(struct hinic_hwdev *hwdev)
* @return
* 0 on success,
* negative error value otherwise.
- **/
+ */
int hinic_convert_rx_buf_size(u32 rx_buf_sz, u32 *match_sz)
{
u32 i, num_hw_types, best_match_sz;
diff --git a/drivers/net/hinic/base/hinic_pmd_nicio.h b/drivers/net/hinic/base/hinic_pmd_nicio.h
index 487e440..9a487d0 100644
--- a/drivers/net/hinic/base/hinic_pmd_nicio.h
+++ b/drivers/net/hinic/base/hinic_pmd_nicio.h
@@ -217,7 +217,7 @@ struct hinic_nic_io {
u16 vhd_mode;
struct hinic_qp *qps;
- /* sq ci mem base addr of the function*/
+ /* sq ci mem base addr of the function */
void *ci_vaddr_base;
dma_addr_t ci_dma_base;
@@ -229,7 +229,6 @@ struct hinic_sq_db {
u32 db_info;
};
-
int hinic_init_qp_ctxts(struct hinic_hwdev *hwdev);
void hinic_free_qp_ctxts(struct hinic_hwdev *hwdev);
--
1.8.3.1
^ permalink raw reply [flat|nested] 32+ messages in thread
* [dpdk-dev] [PATCH v2 03/17] net/hinic: add VF PMD operation interfaces
2019-09-25 14:26 [dpdk-dev] [PATCH v2 00/17] Add advanced features for Huawei hinic pmd Xiaoyun wang
` (2 preceding siblings ...)
2019-09-25 14:30 ` [dpdk-dev] [PATCH v2 02/17] net/hinic/base: add HW interfaces " Xiaoyun wang
@ 2019-09-25 14:30 ` Xiaoyun wang
2019-09-25 14:30 ` [dpdk-dev] [PATCH v2 04/17] net/hinic: add VLAN filter and offload Xiaoyun wang
` (14 subsequent siblings)
18 siblings, 0 replies; 32+ messages in thread
From: Xiaoyun wang @ 2019-09-25 14:30 UTC (permalink / raw)
To: ferruh.yigit
Cc: dev, xuanziyang2, shahar.belkar, luoxianjun, tanya.brokhman,
zhouguoyang, wulike1, Xiaoyun wang
This patch adds VF PMD operation interfaces to support SRIOV.
Signed-off-by: Xiaoyun wang <cloud.wangxiaoyun@huawei.com>
---
drivers/net/hinic/base/hinic_pmd_niccfg.c | 35 ++++
drivers/net/hinic/base/hinic_pmd_niccfg.h | 14 ++
drivers/net/hinic/hinic_pmd_ethdev.c | 255 ++++++++++++++++++++----------
3 files changed, 223 insertions(+), 81 deletions(-)
diff --git a/drivers/net/hinic/base/hinic_pmd_niccfg.c b/drivers/net/hinic/base/hinic_pmd_niccfg.c
index d3bbc3d..0a20ade 100644
--- a/drivers/net/hinic/base/hinic_pmd_niccfg.c
+++ b/drivers/net/hinic/base/hinic_pmd_niccfg.c
@@ -1436,3 +1436,38 @@ int hinic_flush_qp_res(void *hwdev)
return 0;
}
+
+/**
+ * hinic_vf_get_default_cos - Get default cos of VF.
+ *
+ * @param hwdev
+ * The hardware interface of a nic device.
+ * @param cos_id
+ * Cos value.
+ *
+ * @return
+ * 0 on success.
+ * negative error value otherwise.
+ */
+int hinic_vf_get_default_cos(struct hinic_hwdev *hwdev, u8 *cos_id)
+{
+ struct hinic_vf_default_cos vf_cos;
+ u16 out_size = sizeof(vf_cos);
+ int err;
+
+ memset(&vf_cos, 0, sizeof(vf_cos));
+ vf_cos.mgmt_msg_head.resp_aeq_num = HINIC_AEQ1;
+
+ err = hinic_msg_to_mgmt_sync(hwdev, HINIC_MOD_L2NIC,
+ HINIC_PORT_CMD_GET_VF_COS, &vf_cos,
+ sizeof(vf_cos), &vf_cos,
+ &out_size, 0);
+ if (err || !out_size || vf_cos.mgmt_msg_head.status) {
+ PMD_DRV_LOG(ERR, "Get VF default cos failed, err: %d, status: 0x%x, out size: 0x%x",
+ err, vf_cos.mgmt_msg_head.status, out_size);
+ return -EFAULT;
+ }
+ *cos_id = vf_cos.state.default_cos;
+
+ return 0;
+}
diff --git a/drivers/net/hinic/base/hinic_pmd_niccfg.h b/drivers/net/hinic/base/hinic_pmd_niccfg.h
index dcb9602..94371bb 100644
--- a/drivers/net/hinic/base/hinic_pmd_niccfg.h
+++ b/drivers/net/hinic/base/hinic_pmd_niccfg.h
@@ -562,6 +562,18 @@ struct hinic_clear_qp_resource {
u16 rsvd1;
};
+struct hinic_dcb_state {
+ u8 dcb_on;
+ u8 default_cos;
+ u8 up_cos[8];
+};
+
+struct hinic_vf_default_cos {
+ struct hinic_mgmt_msg_head mgmt_msg_head;
+
+ struct hinic_dcb_state state;
+};
+
/* set physical port Anti-Attack rate */
struct hinic_port_anti_attack_rate {
struct hinic_mgmt_msg_head mgmt_msg_head;
@@ -659,4 +671,6 @@ int hinic_set_link_status_follow(void *hwdev,
void hinic_vf_func_free(struct hinic_hwdev *hwdev);
+int hinic_vf_get_default_cos(struct hinic_hwdev *hwdev, u8 *cos_id);
+
#endif /* _HINIC_PMD_NICCFG_H_ */
diff --git a/drivers/net/hinic/hinic_pmd_ethdev.c b/drivers/net/hinic/hinic_pmd_ethdev.c
index c9a400e..2f413e3 100644
--- a/drivers/net/hinic/hinic_pmd_ethdev.c
+++ b/drivers/net/hinic/hinic_pmd_ethdev.c
@@ -20,6 +20,7 @@
#include "base/hinic_pmd_cmdq.h"
#include "base/hinic_pmd_niccfg.h"
#include "base/hinic_pmd_nicio.h"
+#include "base/hinic_pmd_mbox.h"
#include "hinic_pmd_ethdev.h"
#include "hinic_pmd_tx.h"
#include "hinic_pmd_rx.h"
@@ -29,13 +30,21 @@
/* Hinic devices */
#define HINIC_DEV_ID_PRD 0x1822
+#define HINIC_DEV_ID_VF 0x375E
+#define HINIC_DEV_ID_VF_HV 0x379E
+
+/* Mezz card for Blade Server */
#define HINIC_DEV_ID_MEZZ_25GE 0x0210
#define HINIC_DEV_ID_MEZZ_40GE 0x020D
#define HINIC_DEV_ID_MEZZ_100GE 0x0205
+/* 2*25G and 2*100G card */
+#define HINIC_DEV_ID_1822_DUAL_25GE 0x0206
+#define HINIC_DEV_ID_1822_100GE 0x0200
+
#define HINIC_SERVICE_MODE_NIC 2
-#define HINIC_INTR_CB_UNREG_MAX_RETRIES 10
+#define HINIC_INTR_CB_UNREG_MAX_RETRIES 10
#define DEFAULT_BASE_COS 4
#define NR_MAX_COS 8
@@ -43,7 +52,7 @@
#define HINIC_MIN_RX_BUF_SIZE 1024
#define HINIC_MAX_MAC_ADDRS 1
-/** Driver-specific log messages type. */
+/* Driver-specific log messages type */
int hinic_logtype;
struct hinic_xstats_name_off {
@@ -197,10 +206,16 @@ struct hinic_xstats_name_off {
static int hinic_xstats_calc_num(struct hinic_nic_dev *nic_dev)
{
- return (HINIC_VPORT_XSTATS_NUM +
- HINIC_PHYPORT_XSTATS_NUM +
- HINIC_RXQ_XSTATS_NUM * nic_dev->num_rq +
- HINIC_TXQ_XSTATS_NUM * nic_dev->num_sq);
+ if (HINIC_IS_VF(nic_dev->hwdev)) {
+ return (HINIC_VPORT_XSTATS_NUM +
+ HINIC_RXQ_XSTATS_NUM * nic_dev->num_rq +
+ HINIC_TXQ_XSTATS_NUM * nic_dev->num_sq);
+ } else {
+ return (HINIC_VPORT_XSTATS_NUM +
+ HINIC_PHYPORT_XSTATS_NUM +
+ HINIC_RXQ_XSTATS_NUM * nic_dev->num_rq +
+ HINIC_TXQ_XSTATS_NUM * nic_dev->num_sq);
+ }
}
static const struct rte_eth_desc_lim hinic_rx_desc_lim = {
@@ -599,7 +614,7 @@ static void hinic_reset_tx_queue(struct rte_eth_dev *dev)
txq->wq->delta = txq->q_depth;
txq->wq->mask = txq->q_depth - 1;
- /*clear hardware ci*/
+ /* clear hardware ci */
ci_addr = (volatile u32 *)HINIC_CI_VADDR(nic_io->ci_vaddr_base,
q_id);
*ci_addr = 0;
@@ -888,6 +903,9 @@ static int hinic_dev_start(struct rte_eth_dev *dev)
goto cfg_rxtx_fail;
}
+ /* reactive pf status, so that uP report asyn event */
+ hinic_set_pf_status(nic_dev->hwdev->hwif, HINIC_PF_STATUS_ACTIVE_FLAG);
+
/* open virtual port and ready to start packet receiving */
rc = hinic_set_vport_enable(nic_dev->hwdev, true);
if (rc) {
@@ -915,6 +933,8 @@ static int hinic_dev_start(struct rte_eth_dev *dev)
(void)hinic_set_vport_enable(nic_dev->hwdev, false);
en_vport_fail:
+ hinic_set_pf_status(nic_dev->hwdev->hwif, HINIC_PF_STATUS_INIT);
+
/* Flush tx && rx chip resources in case of set vport fake fail */
(void)hinic_flush_qp_res(nic_dev->hwdev);
rte_delay_ms(100);
@@ -1233,7 +1253,7 @@ static int hinic_dev_stats_reset(struct rte_eth_dev *dev)
*
* @param dev
* Pointer to Ethernet device structure.
- **/
+ */
static int hinic_dev_xstats_reset(struct rte_eth_dev *dev)
{
struct hinic_nic_dev *nic_dev = HINIC_ETH_DEV_TO_PRIVATE_NIC_DEV(dev);
@@ -1697,6 +1717,9 @@ static int hinic_dev_xstats_get(struct rte_eth_dev *dev,
count++;
}
+ if (HINIC_IS_VF(nic_dev->hwdev))
+ return count;
+
/* Get stats from hinic_phy_port_stats */
err = hinic_get_phy_port_stats(nic_dev->hwdev, &port_stats);
if (err)
@@ -1765,6 +1788,9 @@ static int hinic_dev_xstats_get_names(struct rte_eth_dev *dev,
count++;
}
+ if (HINIC_IS_VF(nic_dev->hwdev))
+ return count;
+
/* get phy port stats */
for (i = 0; i < HINIC_PHYPORT_XSTATS_NUM; i++) {
snprintf(xstats_names[count].name,
@@ -1805,18 +1831,39 @@ static int hinic_set_default_dcb_feature(struct hinic_nic_dev *nic_dev)
up_pgid, up_bw, up_strict);
}
-static void hinic_init_default_cos(struct hinic_nic_dev *nic_dev)
+static int hinic_init_default_cos(struct hinic_nic_dev *nic_dev)
{
- nic_dev->default_cos =
- (hinic_global_func_id(nic_dev->hwdev) +
- DEFAULT_BASE_COS) % NR_MAX_COS;
+ u8 cos_id = 0;
+ int err;
+
+ if (!HINIC_IS_VF(nic_dev->hwdev)) {
+ nic_dev->default_cos =
+ (hinic_global_func_id(nic_dev->hwdev) +
+ DEFAULT_BASE_COS) % NR_MAX_COS;
+ } else {
+ err = hinic_vf_get_default_cos(nic_dev->hwdev, &cos_id);
+ if (err) {
+ PMD_DRV_LOG(ERR, "Get VF default cos failed, err: %d",
+ err);
+ return HINIC_ERROR;
+ }
+
+ nic_dev->default_cos = cos_id;
+ }
+
+ return 0;
}
static int hinic_set_default_hw_feature(struct hinic_nic_dev *nic_dev)
{
int err;
- hinic_init_default_cos(nic_dev);
+ err = hinic_init_default_cos(nic_dev);
+ if (err)
+ return err;
+
+ if (hinic_func_type(nic_dev->hwdev) == TYPE_VF)
+ return 0;
/* Restore DCB configure to default status */
err = hinic_set_default_dcb_feature(nic_dev);
@@ -1852,6 +1899,9 @@ static int32_t hinic_card_workmode_check(struct hinic_nic_dev *nic_dev)
struct hinic_board_info info = { 0 };
int rc;
+ if (hinic_func_type(nic_dev->hwdev) == TYPE_VF)
+ return 0;
+
rc = hinic_get_board_info(nic_dev->hwdev, &info);
if (rc)
return rc;
@@ -1867,7 +1917,7 @@ static int hinic_copy_mempool_init(struct hinic_nic_dev *nic_dev)
nic_dev->cpy_mpool =
rte_pktmbuf_pool_create(nic_dev->proc_dev_name,
HINIC_COPY_MEMPOOL_DEPTH,
- RTE_CACHE_LINE_SIZE, 0,
+ 0, 0,
HINIC_COPY_MBUF_SIZE,
rte_socket_id());
if (!nic_dev->cpy_mpool) {
@@ -1978,6 +2028,14 @@ static int hinic_nic_dev_create(struct rte_eth_dev *eth_dev)
goto init_pf_to_mgmt_fail;
}
+ /* init mailbox */
+ rc = hinic_comm_func_to_func_init(nic_dev->hwdev);
+ if (rc) {
+ PMD_DRV_LOG(ERR, "Initialize func_to_func failed, dev_name: %s",
+ eth_dev->data->name);
+ goto init_func_to_func_fail;
+ }
+
rc = hinic_card_workmode_check(nic_dev);
if (rc) {
PMD_DRV_LOG(ERR, "Check card workmode failed, dev_name: %s",
@@ -2077,6 +2135,9 @@ static int hinic_nic_dev_create(struct rte_eth_dev *eth_dev)
init_cmdq_fail:
l2nic_reset_fail:
workmode_check_fail:
+ hinic_comm_func_to_func_free(nic_dev->hwdev);
+
+init_func_to_func_fail:
hinic_comm_pf_to_mgmt_free(nic_dev->hwdev);
init_pf_to_mgmt_fail:
@@ -2110,6 +2171,7 @@ static void hinic_nic_dev_destroy(struct rte_eth_dev *eth_dev)
hinic_deinit_nicio(nic_dev->hwdev);
hinic_deactivate_hwdev_state(nic_dev->hwdev);
hinic_comm_cmdqs_free(nic_dev->hwdev);
+ hinic_comm_func_to_func_free(nic_dev->hwdev);
hinic_comm_pf_to_mgmt_free(nic_dev->hwdev);
hinic_comm_aeqs_free(nic_dev->hwdev);
free_cfg_mgmt(nic_dev->hwdev);
@@ -2119,6 +2181,93 @@ static void hinic_nic_dev_destroy(struct rte_eth_dev *eth_dev)
nic_dev->hwdev = NULL;
}
+/**
+ * DPDK callback to close the device.
+ *
+ * @param dev
+ * Pointer to Ethernet device structure.
+ */
+static void hinic_dev_close(struct rte_eth_dev *dev)
+{
+ struct hinic_nic_dev *nic_dev = HINIC_ETH_DEV_TO_PRIVATE_NIC_DEV(dev);
+
+ if (hinic_test_and_set_bit(HINIC_DEV_CLOSE, &nic_dev->dev_status)) {
+ PMD_DRV_LOG(WARNING, "Device %s already closed",
+ dev->data->name);
+ return;
+ }
+
+ /* stop device first */
+ hinic_dev_stop(dev);
+
+ /* rx_cqe, rx_info */
+ hinic_free_all_rx_resources(dev);
+
+ /* tx_info */
+ hinic_free_all_tx_resources(dev);
+
+ /* free wq, pi_dma_addr */
+ hinic_free_all_rq(nic_dev);
+
+ /* free wq, db_addr */
+ hinic_free_all_sq(nic_dev);
+
+ /* deinit mac vlan tbl */
+ hinic_deinit_mac_addr(dev);
+
+ /* disable hardware and uio interrupt */
+ hinic_disable_interrupt(dev);
+
+ /* deinit nic hardware device */
+ hinic_nic_dev_destroy(dev);
+}
+
+static const struct eth_dev_ops hinic_pmd_ops = {
+ .dev_configure = hinic_dev_configure,
+ .dev_infos_get = hinic_dev_infos_get,
+ .rx_queue_setup = hinic_rx_queue_setup,
+ .tx_queue_setup = hinic_tx_queue_setup,
+ .dev_start = hinic_dev_start,
+ .link_update = hinic_link_update,
+ .rx_queue_release = hinic_rx_queue_release,
+ .tx_queue_release = hinic_tx_queue_release,
+ .dev_stop = hinic_dev_stop,
+ .dev_close = hinic_dev_close,
+ .promiscuous_enable = hinic_dev_promiscuous_enable,
+ .promiscuous_disable = hinic_dev_promiscuous_disable,
+ .rss_hash_update = hinic_rss_hash_update,
+ .rss_hash_conf_get = hinic_rss_conf_get,
+ .reta_update = hinic_rss_indirtbl_update,
+ .reta_query = hinic_rss_indirtbl_query,
+ .stats_get = hinic_dev_stats_get,
+ .stats_reset = hinic_dev_stats_reset,
+ .xstats_get = hinic_dev_xstats_get,
+ .xstats_reset = hinic_dev_xstats_reset,
+ .xstats_get_names = hinic_dev_xstats_get_names,
+};
+
+static const struct eth_dev_ops hinic_pmd_vf_ops = {
+ .dev_configure = hinic_dev_configure,
+ .dev_infos_get = hinic_dev_infos_get,
+ .rx_queue_setup = hinic_rx_queue_setup,
+ .tx_queue_setup = hinic_tx_queue_setup,
+ .dev_start = hinic_dev_start,
+ .link_update = hinic_link_update,
+ .rx_queue_release = hinic_rx_queue_release,
+ .tx_queue_release = hinic_tx_queue_release,
+ .dev_stop = hinic_dev_stop,
+ .dev_close = hinic_dev_close,
+ .rss_hash_update = hinic_rss_hash_update,
+ .rss_hash_conf_get = hinic_rss_conf_get,
+ .reta_update = hinic_rss_indirtbl_update,
+ .reta_query = hinic_rss_indirtbl_query,
+ .stats_get = hinic_dev_stats_get,
+ .stats_reset = hinic_dev_stats_reset,
+ .xstats_get = hinic_dev_xstats_get,
+ .xstats_reset = hinic_dev_xstats_reset,
+ .xstats_get_names = hinic_dev_xstats_get_names,
+};
+
static int hinic_func_init(struct rte_eth_dev *eth_dev)
{
struct rte_pci_device *pci_dev;
@@ -2173,6 +2322,11 @@ static int hinic_func_init(struct rte_eth_dev *eth_dev)
goto create_nic_dev_fail;
}
+ if (HINIC_IS_VF(nic_dev->hwdev))
+ eth_dev->dev_ops = &hinic_pmd_vf_ops;
+ else
+ eth_dev->dev_ops = &hinic_pmd_ops;
+
rc = hinic_init_mac_addr(eth_dev);
if (rc) {
PMD_DRV_LOG(ERR, "Initialize mac table failed, dev_name: %s",
@@ -2214,6 +2368,7 @@ static int hinic_func_init(struct rte_eth_dev *eth_dev)
hinic_deinit_mac_addr(eth_dev);
init_mac_fail:
+ eth_dev->dev_ops = NULL;
hinic_nic_dev_destroy(eth_dev);
create_nic_dev_fail:
@@ -2226,71 +2381,6 @@ static int hinic_func_init(struct rte_eth_dev *eth_dev)
return rc;
}
-/**
- * DPDK callback to close the device.
- *
- * @param dev
- * Pointer to Ethernet device structure.
- */
-static void hinic_dev_close(struct rte_eth_dev *dev)
-{
- struct hinic_nic_dev *nic_dev = HINIC_ETH_DEV_TO_PRIVATE_NIC_DEV(dev);
-
- if (hinic_test_and_set_bit(HINIC_DEV_CLOSE, &nic_dev->dev_status)) {
- PMD_DRV_LOG(WARNING, "Device %s already closed",
- dev->data->name);
- return;
- }
-
- /* stop device first */
- hinic_dev_stop(dev);
-
- /* rx_cqe, rx_info */
- hinic_free_all_rx_resources(dev);
-
- /* tx_info */
- hinic_free_all_tx_resources(dev);
-
- /* free wq, pi_dma_addr */
- hinic_free_all_rq(nic_dev);
-
- /* free wq, db_addr */
- hinic_free_all_sq(nic_dev);
-
- /* deinit mac vlan tbl */
- hinic_deinit_mac_addr(dev);
-
- /* disable hardware and uio interrupt */
- hinic_disable_interrupt(dev);
-
- /* deinit nic hardware device */
- hinic_nic_dev_destroy(dev);
-}
-
-static const struct eth_dev_ops hinic_pmd_ops = {
- .dev_configure = hinic_dev_configure,
- .dev_infos_get = hinic_dev_infos_get,
- .rx_queue_setup = hinic_rx_queue_setup,
- .tx_queue_setup = hinic_tx_queue_setup,
- .dev_start = hinic_dev_start,
- .link_update = hinic_link_update,
- .rx_queue_release = hinic_rx_queue_release,
- .tx_queue_release = hinic_tx_queue_release,
- .dev_stop = hinic_dev_stop,
- .dev_close = hinic_dev_close,
- .promiscuous_enable = hinic_dev_promiscuous_enable,
- .promiscuous_disable = hinic_dev_promiscuous_disable,
- .rss_hash_update = hinic_rss_hash_update,
- .rss_hash_conf_get = hinic_rss_conf_get,
- .reta_update = hinic_rss_indirtbl_update,
- .reta_query = hinic_rss_indirtbl_query,
- .stats_get = hinic_dev_stats_get,
- .stats_reset = hinic_dev_stats_reset,
- .xstats_get = hinic_dev_xstats_get,
- .xstats_reset = hinic_dev_xstats_reset,
- .xstats_get_names = hinic_dev_xstats_get_names,
-};
-
static int hinic_dev_init(struct rte_eth_dev *eth_dev)
{
struct rte_pci_device *pci_dev;
@@ -2303,8 +2393,7 @@ static int hinic_dev_init(struct rte_eth_dev *eth_dev)
(rte_eal_process_type() == RTE_PROC_PRIMARY) ?
"primary" : "secondary");
- /* rte_eth_dev ops, rx_burst and tx_burst */
- eth_dev->dev_ops = &hinic_pmd_ops;
+ /* rte_eth_dev rx_burst and tx_burst */
eth_dev->rx_pkt_burst = hinic_recv_pkts;
eth_dev->tx_pkt_burst = hinic_xmit_pkts;
@@ -2338,6 +2427,10 @@ static int hinic_dev_uninit(struct rte_eth_dev *dev)
{ RTE_PCI_DEVICE(HINIC_HUAWEI_VENDOR_ID, HINIC_DEV_ID_MEZZ_25GE) },
{ RTE_PCI_DEVICE(HINIC_HUAWEI_VENDOR_ID, HINIC_DEV_ID_MEZZ_40GE) },
{ RTE_PCI_DEVICE(HINIC_HUAWEI_VENDOR_ID, HINIC_DEV_ID_MEZZ_100GE) },
+ { RTE_PCI_DEVICE(HINIC_HUAWEI_VENDOR_ID, HINIC_DEV_ID_VF) },
+ { RTE_PCI_DEVICE(HINIC_HUAWEI_VENDOR_ID, HINIC_DEV_ID_VF_HV) },
+ { RTE_PCI_DEVICE(HINIC_HUAWEI_VENDOR_ID, HINIC_DEV_ID_1822_DUAL_25GE) },
+ { RTE_PCI_DEVICE(HINIC_HUAWEI_VENDOR_ID, HINIC_DEV_ID_1822_100GE) },
{.vendor_id = 0},
};
--
1.8.3.1
^ permalink raw reply [flat|nested] 32+ messages in thread
* [dpdk-dev] [PATCH v2 04/17] net/hinic: add VLAN filter and offload
2019-09-25 14:26 [dpdk-dev] [PATCH v2 00/17] Add advanced features for Huawei hinic pmd Xiaoyun wang
` (3 preceding siblings ...)
2019-09-25 14:30 ` [dpdk-dev] [PATCH v2 03/17] net/hinic: add VF PMD operation interfaces Xiaoyun wang
@ 2019-09-25 14:30 ` Xiaoyun wang
2019-09-25 14:30 ` [dpdk-dev] [PATCH v2 05/17] net/hinic: add allmulticast mode and MTU set Xiaoyun wang
` (13 subsequent siblings)
18 siblings, 0 replies; 32+ messages in thread
From: Xiaoyun wang @ 2019-09-25 14:30 UTC (permalink / raw)
To: ferruh.yigit
Cc: dev, xuanziyang2, shahar.belkar, luoxianjun, tanya.brokhman,
zhouguoyang, wulike1, Xiaoyun wang
This patch adds support for VLAN filter and VLAN offload.
Signed-off-by: Xiaoyun wang <cloud.wangxiaoyun@huawei.com>
---
drivers/net/hinic/base/hinic_pmd_cmd.h | 1 +
drivers/net/hinic/base/hinic_pmd_niccfg.c | 136 +++++++++++++++++++++
drivers/net/hinic/base/hinic_pmd_niccfg.h | 29 +++++
drivers/net/hinic/hinic_pmd_ethdev.c | 194 +++++++++++++++++++++++++++++-
drivers/net/hinic/hinic_pmd_ethdev.h | 15 ++-
5 files changed, 369 insertions(+), 6 deletions(-)
diff --git a/drivers/net/hinic/base/hinic_pmd_cmd.h b/drivers/net/hinic/base/hinic_pmd_cmd.h
index c8750b8..5c38b5f 100644
--- a/drivers/net/hinic/base/hinic_pmd_cmd.h
+++ b/drivers/net/hinic/base/hinic_pmd_cmd.h
@@ -140,6 +140,7 @@ enum hinic_port_cmd {
HINIC_PORT_CMD_SET_VHD_CFG = 0xF7,
HINIC_PORT_CMD_SET_LINK_FOLLOW = 0xF8,
+ HINIC_PORT_CMD_SET_VLAN_FILTER = 0xFF
};
/* cmd of mgmt CPU message for HW module */
diff --git a/drivers/net/hinic/base/hinic_pmd_niccfg.c b/drivers/net/hinic/base/hinic_pmd_niccfg.c
index 0a20ade..8bd7ed6 100644
--- a/drivers/net/hinic/base/hinic_pmd_niccfg.c
+++ b/drivers/net/hinic/base/hinic_pmd_niccfg.c
@@ -290,6 +290,142 @@ int hinic_set_port_mtu(void *hwdev, u32 new_mtu)
}
/**
+ * hinic_add_remove_vlan - Add or remove vlan id to vlan elb table.
+ *
+ * @param hwdev
+ * The hardware interface of a nic device.
+ * @param vlan_id
+ * Vlan id.
+ * @param func_id
+ * Global function id of NIC.
+ * @param add
+ * Add or remove operation.
+ *
+ * @return
+ * 0 on success.
+ * negative error value otherwise.
+ */
+int hinic_add_remove_vlan(void *hwdev, u16 vlan_id, u16 func_id, bool add)
+{
+ struct hinic_vlan_config vlan_info;
+ u16 out_size = sizeof(vlan_info);
+ u8 cmd;
+ int err;
+
+ if (!hwdev) {
+ PMD_DRV_LOG(ERR, "Hwdev is NULL");
+ return -EINVAL;
+ }
+
+ cmd = add ? HINIC_PORT_CMD_ADD_VLAN : HINIC_PORT_CMD_DEL_VLAN;
+
+ memset(&vlan_info, 0, sizeof(vlan_info));
+ vlan_info.mgmt_msg_head.resp_aeq_num = HINIC_AEQ1;
+ vlan_info.func_id = func_id;
+ vlan_info.vlan_id = vlan_id;
+
+ err = l2nic_msg_to_mgmt_sync(hwdev, cmd, &vlan_info,
+ sizeof(vlan_info), &vlan_info,
+ &out_size);
+ if (err || !out_size || vlan_info.mgmt_msg_head.status) {
+ PMD_DRV_LOG(ERR,
+ "Failed to %s vlan, err: %d, status: 0x%x, out size: 0x%x\n",
+ add ? "add" : "remove", err,
+ vlan_info.mgmt_msg_head.status, out_size);
+ return -EINVAL;
+ }
+
+ return 0;
+}
+
+/**
+ * hinic_config_vlan_filter - Enable or Disable vlan filter.
+ *
+ * @param hwdev
+ * The hardware interface of a nic device.
+ * @param vlan_filter_ctrl
+ * Enable or Disable.
+ *
+ * @return
+ * 0 on success.
+ * negative error value otherwise.
+ */
+int hinic_config_vlan_filter(void *hwdev, u32 vlan_filter_ctrl)
+{
+ struct hinic_hwdev *nic_hwdev = (struct hinic_hwdev *)hwdev;
+ struct hinic_vlan_filter vlan_filter;
+ u16 out_size = sizeof(vlan_filter);
+ int err;
+
+ if (!hwdev)
+ return -EINVAL;
+
+ memset(&vlan_filter, 0, sizeof(vlan_filter));
+ vlan_filter.mgmt_msg_head.resp_aeq_num = HINIC_AEQ1;
+ vlan_filter.func_id = hinic_global_func_id(nic_hwdev);
+ vlan_filter.vlan_filter_ctrl = vlan_filter_ctrl;
+
+ err = l2nic_msg_to_mgmt_sync(nic_hwdev, HINIC_PORT_CMD_SET_VLAN_FILTER,
+ &vlan_filter, sizeof(vlan_filter),
+ &vlan_filter, &out_size);
+ if (vlan_filter.mgmt_msg_head.status == HINIC_MGMT_CMD_UNSUPPORTED) {
+ err = HINIC_MGMT_CMD_UNSUPPORTED;
+ } else if ((err == HINIC_MBOX_VF_CMD_ERROR) &&
+ (HINIC_IS_VF(nic_hwdev))) {
+ err = HINIC_MGMT_CMD_UNSUPPORTED;
+ } else if (err || !out_size || vlan_filter.mgmt_msg_head.status) {
+ PMD_DRV_LOG(ERR,
+ "Failed to config vlan filter, vlan_filter_ctrl: 0x%x, err: %d, status: 0x%x, out size: 0x%x\n",
+ vlan_filter_ctrl, err,
+ vlan_filter.mgmt_msg_head.status, out_size);
+ err = -EINVAL;
+ }
+
+ return err;
+}
+
+/**
+ * hinic_set_rx_vlan_offload - Enable or Disable vlan offload.
+ *
+ * @param hwdev
+ * The hardware interface of a nic device.
+ * @param en
+ * Enable or Disable.
+ *
+ * @return
+ * 0 on success.
+ * negative error value otherwise.
+ */
+int hinic_set_rx_vlan_offload(void *hwdev, u8 en)
+{
+ struct hinic_vlan_offload vlan_cfg;
+ u16 out_size = sizeof(vlan_cfg);
+ int err;
+
+ if (!hwdev) {
+ PMD_DRV_LOG(ERR, "Hwdev is NULL");
+ return -EINVAL;
+ }
+
+ memset(&vlan_cfg, 0, sizeof(vlan_cfg));
+ vlan_cfg.mgmt_msg_head.resp_aeq_num = HINIC_AEQ1;
+ vlan_cfg.func_id = hinic_global_func_id(hwdev);
+ vlan_cfg.vlan_rx_offload = en;
+
+ err = l2nic_msg_to_mgmt_sync(hwdev, HINIC_PORT_CMD_SET_RX_VLAN_OFFLOAD,
+ &vlan_cfg, sizeof(vlan_cfg),
+ &vlan_cfg, &out_size);
+ if (err || !out_size || vlan_cfg.mgmt_msg_head.status) {
+ PMD_DRV_LOG(ERR,
+ "Failed to set rx vlan offload, err: %d, status: 0x%x, out size: 0x%x\n",
+ err, vlan_cfg.mgmt_msg_head.status, out_size);
+ return -EINVAL;
+ }
+
+ return 0;
+}
+
+/**
* hinic_get_link_status - Get link status from hardware.
*
* @param hwdev
diff --git a/drivers/net/hinic/base/hinic_pmd_niccfg.h b/drivers/net/hinic/base/hinic_pmd_niccfg.h
index 94371bb..d19a834 100644
--- a/drivers/net/hinic/base/hinic_pmd_niccfg.h
+++ b/drivers/net/hinic/base/hinic_pmd_niccfg.h
@@ -380,6 +380,29 @@ struct hinic_mtu {
u32 mtu;
};
+struct hinic_vlan_config {
+ struct hinic_mgmt_msg_head mgmt_msg_head;
+
+ u16 func_id;
+ u16 vlan_id;
+};
+
+struct hinic_vlan_filter {
+ struct hinic_mgmt_msg_head mgmt_msg_head;
+
+ u16 func_id;
+ u8 rsvd1[2];
+ u32 vlan_filter_ctrl;
+};
+
+struct hinic_vlan_offload {
+ struct hinic_mgmt_msg_head mgmt_msg_head;
+
+ u16 func_id;
+ u8 vlan_rx_offload;
+ u8 rsvd1[5];
+};
+
struct hinic_get_link {
struct hinic_mgmt_msg_head mgmt_msg_head;
@@ -597,6 +620,12 @@ int hinic_update_mac(void *hwdev, u8 *old_mac, u8 *new_mac, u16 vlan_id,
int hinic_set_port_mtu(void *hwdev, u32 new_mtu);
+int hinic_add_remove_vlan(void *hwdev, u16 vlan_id, u16 func_id, bool add);
+
+int hinic_config_vlan_filter(void *hwdev, u32 vlan_filter_ctrl);
+
+int hinic_set_rx_vlan_offload(void *hwdev, u8 en);
+
int hinic_set_vport_enable(void *hwdev, bool enable);
int hinic_set_port_enable(void *hwdev, bool enable);
diff --git a/drivers/net/hinic/hinic_pmd_ethdev.c b/drivers/net/hinic/hinic_pmd_ethdev.c
index 2f413e3..4e2a69c 100644
--- a/drivers/net/hinic/hinic_pmd_ethdev.c
+++ b/drivers/net/hinic/hinic_pmd_ethdev.c
@@ -10,6 +10,7 @@
#include <rte_memcpy.h>
#include <rte_mempool.h>
#include <rte_errno.h>
+#include <rte_ether.h>
#include "base/hinic_compat.h"
#include "base/hinic_pmd_hwdev.h"
@@ -52,6 +53,17 @@
#define HINIC_MIN_RX_BUF_SIZE 1024
#define HINIC_MAX_MAC_ADDRS 1
+/*
+ * vlan_id is a 12 bit number.
+ * The VFTA array is actually a 4096 bit array, 128 of 32bit elements.
+ * 2^5 = 32. The val of lower 5 bits specifies the bit in the 32bit element.
+ * The higher 7 bit val specifies VFTA array index.
+ */
+#define HINIC_VFTA_BIT(vlan_id) (1 << ((vlan_id) & 0x1F))
+#define HINIC_VFTA_IDX(vlan_id) ((vlan_id) >> 5)
+
+#define HINIC_VLAN_FILTER_EN (1U << 0)
+
/* Driver-specific log messages type */
int hinic_logtype;
@@ -230,6 +242,7 @@ static int hinic_xstats_calc_num(struct hinic_nic_dev *nic_dev)
.nb_align = HINIC_TXD_ALIGN,
};
+static int hinic_vlan_offload_set(struct rte_eth_dev *dev, int mask);
/**
* Interrupt handler triggered by NIC for handling
@@ -313,6 +326,15 @@ static int hinic_dev_configure(struct rte_eth_dev *dev)
return err;
}
+ /* init vlan offoad */
+ err = hinic_vlan_offload_set(dev,
+ ETH_VLAN_STRIP_MASK | ETH_VLAN_FILTER_MASK);
+ if (err) {
+ PMD_DRV_LOG(ERR, "Initialize vlan filter and strip failed\n");
+ (void)hinic_config_mq_mode(dev, FALSE);
+ return err;
+ }
+
return HINIC_OK;
}
@@ -695,7 +717,8 @@ static void hinic_get_speed_capa(struct rte_eth_dev *dev, uint32_t *speed_capa)
info->rx_offload_capa = DEV_RX_OFFLOAD_VLAN_STRIP |
DEV_RX_OFFLOAD_IPV4_CKSUM |
DEV_RX_OFFLOAD_UDP_CKSUM |
- DEV_RX_OFFLOAD_TCP_CKSUM;
+ DEV_RX_OFFLOAD_TCP_CKSUM |
+ DEV_RX_OFFLOAD_VLAN_FILTER;
info->tx_queue_offload_capa = 0;
info->tx_offload_capa = DEV_TX_OFFLOAD_VLAN_INSERT |
@@ -1353,6 +1376,170 @@ static void hinic_deinit_mac_addr(struct rte_eth_dev *eth_dev)
eth_dev->data->name);
}
+static void hinic_store_vlan_filter(struct hinic_nic_dev *nic_dev,
+ u16 vlan_id, bool on)
+{
+ u32 vid_idx, vid_bit;
+
+ vid_idx = HINIC_VFTA_IDX(vlan_id);
+ vid_bit = HINIC_VFTA_BIT(vlan_id);
+
+ if (on)
+ nic_dev->vfta[vid_idx] |= vid_bit;
+ else
+ nic_dev->vfta[vid_idx] &= ~vid_bit;
+}
+
+static bool hinic_find_vlan_filter(struct hinic_nic_dev *nic_dev,
+ uint16_t vlan_id)
+{
+ u32 vid_idx, vid_bit;
+
+ vid_idx = HINIC_VFTA_IDX(vlan_id);
+ vid_bit = HINIC_VFTA_BIT(vlan_id);
+
+ return (nic_dev->vfta[vid_idx] & vid_bit) ? TRUE : FALSE;
+}
+
+/**
+ * DPDK callback to set vlan filter.
+ *
+ * @param dev
+ * Pointer to Ethernet device structure.
+ * @param vlan_id
+ * vlan id is used to filter vlan packets
+ * @param enable
+ * enable disable or enable vlan filter function
+ */
+static int hinic_vlan_filter_set(struct rte_eth_dev *dev,
+ uint16_t vlan_id, int enable)
+{
+ struct hinic_nic_dev *nic_dev = HINIC_ETH_DEV_TO_PRIVATE_NIC_DEV(dev);
+ int err = 0;
+ u16 func_id;
+
+ if (vlan_id > RTE_ETHER_MAX_VLAN_ID)
+ return -EINVAL;
+
+ func_id = hinic_global_func_id(nic_dev->hwdev);
+
+ if (enable) {
+ /* If vlanid is already set, just return */
+ if (hinic_find_vlan_filter(nic_dev, vlan_id)) {
+ PMD_DRV_LOG(INFO, "Vlan %u has been added, device: %s",
+ vlan_id, nic_dev->proc_dev_name);
+ return 0;
+ }
+
+ err = hinic_add_remove_vlan(nic_dev->hwdev, vlan_id,
+ func_id, TRUE);
+ } else {
+ /* If vlanid can't be found, just return */
+ if (!hinic_find_vlan_filter(nic_dev, vlan_id)) {
+ PMD_DRV_LOG(INFO, "Vlan %u is not in the vlan filter list, device: %s",
+ vlan_id, nic_dev->proc_dev_name);
+ return 0;
+ }
+
+ err = hinic_add_remove_vlan(nic_dev->hwdev, vlan_id,
+ func_id, FALSE);
+ }
+
+ if (err) {
+ PMD_DRV_LOG(ERR, "%s vlan failed, func_id: %d, vlan_id: %d, err: %d",
+ enable ? "Add" : "Remove", func_id, vlan_id, err);
+ return err;
+ }
+
+ hinic_store_vlan_filter(nic_dev, vlan_id, enable);
+
+ PMD_DRV_LOG(INFO, "%s vlan %u succeed, device: %s",
+ enable ? "Add" : "Remove", vlan_id, nic_dev->proc_dev_name);
+ return 0;
+}
+
+/**
+ * DPDK callback to enable or disable vlan offload.
+ *
+ * @param dev
+ * Pointer to Ethernet device structure.
+ * @param mask
+ * Definitions used for VLAN setting
+ */
+static int hinic_vlan_offload_set(struct rte_eth_dev *dev, int mask)
+{
+ struct hinic_nic_dev *nic_dev = HINIC_ETH_DEV_TO_PRIVATE_NIC_DEV(dev);
+ struct rte_eth_rxmode *rxmode = &dev->data->dev_conf.rxmode;
+ bool on;
+ int err;
+
+ /* Enable or disable VLAN filter */
+ if (mask & ETH_VLAN_FILTER_MASK) {
+ on = (rxmode->offloads & DEV_RX_OFFLOAD_VLAN_FILTER) ?
+ TRUE : FALSE;
+ err = hinic_config_vlan_filter(nic_dev->hwdev, on);
+ if (err == HINIC_MGMT_CMD_UNSUPPORTED) {
+ PMD_DRV_LOG(WARNING,
+ "Current matching version does not support vlan filter configuration, device: %s, port_id: %d",
+ nic_dev->proc_dev_name, dev->data->port_id);
+ } else if (err) {
+ PMD_DRV_LOG(ERR, "Failed to %s vlan filter, device: %s, port_id: %d, err: %d",
+ on ? "enable" : "disable",
+ nic_dev->proc_dev_name,
+ dev->data->port_id, err);
+ return err;
+ }
+
+ PMD_DRV_LOG(INFO, "%s vlan filter succeed, device: %s, port_id: %d",
+ on ? "Enable" : "Disable",
+ nic_dev->proc_dev_name, dev->data->port_id);
+ }
+
+ /* Enable or disable VLAN stripping */
+ if (mask & ETH_VLAN_STRIP_MASK) {
+ on = (rxmode->offloads & DEV_RX_OFFLOAD_VLAN_STRIP) ?
+ TRUE : FALSE;
+ err = hinic_set_rx_vlan_offload(nic_dev->hwdev, on);
+ if (err) {
+ PMD_DRV_LOG(ERR, "Failed to %s vlan strip, device: %s, port_id: %d, err: %d",
+ on ? "enable" : "disable",
+ nic_dev->proc_dev_name,
+ dev->data->port_id, err);
+ return err;
+ }
+
+ PMD_DRV_LOG(INFO, "%s vlan strip succeed, device: %s, port_id: %d",
+ on ? "Enable" : "Disable",
+ nic_dev->proc_dev_name, dev->data->port_id);
+ }
+
+ if (mask & ETH_VLAN_EXTEND_MASK) {
+ PMD_DRV_LOG(ERR, "Don't support vlan qinq, device: %s, port_id: %d",
+ nic_dev->proc_dev_name, dev->data->port_id);
+ return -ENOTSUP;
+ }
+
+ return 0;
+}
+
+static void hinic_remove_all_vlanid(struct rte_eth_dev *eth_dev)
+{
+ struct hinic_nic_dev *nic_dev =
+ HINIC_ETH_DEV_TO_PRIVATE_NIC_DEV(eth_dev);
+ u16 func_id;
+ int i;
+
+ func_id = hinic_global_func_id(nic_dev->hwdev);
+ for (i = 0; i <= RTE_ETHER_MAX_VLAN_ID; i++) {
+ /* If can't find it, continue */
+ if (!hinic_find_vlan_filter(nic_dev, i))
+ continue;
+
+ (void)hinic_add_remove_vlan(nic_dev->hwdev, i, func_id, FALSE);
+ hinic_store_vlan_filter(nic_dev, i, false);
+ }
+}
+
/**
* DPDK callback to enable promiscuous mode.
*
@@ -2214,6 +2401,7 @@ static void hinic_dev_close(struct rte_eth_dev *dev)
/* deinit mac vlan tbl */
hinic_deinit_mac_addr(dev);
+ hinic_remove_all_vlanid(dev);
/* disable hardware and uio interrupt */
hinic_disable_interrupt(dev);
@@ -2233,6 +2421,8 @@ static void hinic_dev_close(struct rte_eth_dev *dev)
.tx_queue_release = hinic_tx_queue_release,
.dev_stop = hinic_dev_stop,
.dev_close = hinic_dev_close,
+ .vlan_filter_set = hinic_vlan_filter_set,
+ .vlan_offload_set = hinic_vlan_offload_set,
.promiscuous_enable = hinic_dev_promiscuous_enable,
.promiscuous_disable = hinic_dev_promiscuous_disable,
.rss_hash_update = hinic_rss_hash_update,
@@ -2257,6 +2447,8 @@ static void hinic_dev_close(struct rte_eth_dev *dev)
.tx_queue_release = hinic_tx_queue_release,
.dev_stop = hinic_dev_stop,
.dev_close = hinic_dev_close,
+ .vlan_filter_set = hinic_vlan_filter_set,
+ .vlan_offload_set = hinic_vlan_offload_set,
.rss_hash_update = hinic_rss_hash_update,
.rss_hash_conf_get = hinic_rss_conf_get,
.reta_update = hinic_rss_indirtbl_update,
diff --git a/drivers/net/hinic/hinic_pmd_ethdev.h b/drivers/net/hinic/hinic_pmd_ethdev.h
index 4aeddc2..66eaf20 100644
--- a/drivers/net/hinic/hinic_pmd_ethdev.h
+++ b/drivers/net/hinic/hinic_pmd_ethdev.h
@@ -11,12 +11,12 @@
#include "base/hinic_compat.h"
#include "base/hinic_pmd_cfg.h"
-#define HINIC_DEV_NAME_LEN (32)
-#define HINIC_MAX_RX_QUEUES (64)
+#define HINIC_DEV_NAME_LEN 32
+#define HINIC_MAX_RX_QUEUES 64
/* mbuf pool for copy invalid mbuf segs */
-#define HINIC_COPY_MEMPOOL_DEPTH (128)
-#define HINIC_COPY_MBUF_SIZE (4096)
+#define HINIC_COPY_MEMPOOL_DEPTH 128
+#define HINIC_COPY_MBUF_SIZE 4096
#define SIZE_8BYTES(size) (ALIGN((u32)(size), 8) >> 3)
@@ -31,6 +31,9 @@
#define HINIC_TXD_ALIGN 1
#define HINIC_RXD_ALIGN 1
+#define HINIC_UINT32_BIT_SIZE (CHAR_BIT * sizeof(uint32_t))
+#define HINIC_VFTA_SIZE (4096 / HINIC_UINT32_BIT_SIZE)
+
enum hinic_dev_status {
HINIC_DEV_INIT,
HINIC_DEV_CLOSE,
@@ -54,10 +57,12 @@ struct hinic_nic_dev {
u8 num_rss;
u8 rx_queue_list[HINIC_MAX_RX_QUEUES];
+ u32 vfta[HINIC_VFTA_SIZE]; /* VLAN bitmap */
+
/* info */
unsigned int flags;
struct nic_service_cap nic_cap;
- u32 rx_mode_status; /* promisc allmulticast */
+ u32 rx_mode_status; /* promisc or allmulticast */
unsigned long dev_status;
/* dpdk only */
--
1.8.3.1
^ permalink raw reply [flat|nested] 32+ messages in thread
* [dpdk-dev] [PATCH v2 05/17] net/hinic: add allmulticast mode and MTU set
2019-09-25 14:26 [dpdk-dev] [PATCH v2 00/17] Add advanced features for Huawei hinic pmd Xiaoyun wang
` (4 preceding siblings ...)
2019-09-25 14:30 ` [dpdk-dev] [PATCH v2 04/17] net/hinic: add VLAN filter and offload Xiaoyun wang
@ 2019-09-25 14:30 ` Xiaoyun wang
2019-09-26 18:47 ` Ferruh Yigit
2019-09-25 14:30 ` [dpdk-dev] [PATCH v2 06/17] net/hinic: add unicast and multicast MAC set Xiaoyun wang
` (12 subsequent siblings)
18 siblings, 1 reply; 32+ messages in thread
From: Xiaoyun wang @ 2019-09-25 14:30 UTC (permalink / raw)
To: ferruh.yigit
Cc: dev, xuanziyang2, shahar.belkar, luoxianjun, tanya.brokhman,
zhouguoyang, wulike1, Xiaoyun wang
When enable allmulticast mode, all multicast packets
can be received. This patch also adds support for mtu set,
the range of MTU is from 256 to 9600.
Signed-off-by: Xiaoyun wang <cloud.wangxiaoyun@huawei.com>
---
drivers/net/hinic/hinic_pmd_ethdev.c | 108 ++++++++++++++++++++++++++++++++++-
drivers/net/hinic/hinic_pmd_ethdev.h | 3 -
2 files changed, 107 insertions(+), 4 deletions(-)
diff --git a/drivers/net/hinic/hinic_pmd_ethdev.c b/drivers/net/hinic/hinic_pmd_ethdev.c
index 4e2a69c..8eeeaea 100644
--- a/drivers/net/hinic/hinic_pmd_ethdev.c
+++ b/drivers/net/hinic/hinic_pmd_ethdev.c
@@ -64,6 +64,12 @@
#define HINIC_VLAN_FILTER_EN (1U << 0)
+#define HINIC_MTU_TO_PKTLEN(mtu) \
+ ((mtu) + ETH_HLEN + ETH_CRC_LEN)
+
+#define HINIC_PKTLEN_TO_MTU(pktlen) \
+ ((pktlen) - (ETH_HLEN + ETH_CRC_LEN))
+
/* Driver-specific log messages type */
int hinic_logtype;
@@ -718,7 +724,9 @@ static void hinic_get_speed_capa(struct rte_eth_dev *dev, uint32_t *speed_capa)
DEV_RX_OFFLOAD_IPV4_CKSUM |
DEV_RX_OFFLOAD_UDP_CKSUM |
DEV_RX_OFFLOAD_TCP_CKSUM |
- DEV_RX_OFFLOAD_VLAN_FILTER;
+ DEV_RX_OFFLOAD_VLAN_FILTER |
+ DEV_RX_OFFLOAD_SCATTER |
+ DEV_RX_OFFLOAD_JUMBO_FRAME;
info->tx_queue_offload_capa = 0;
info->tx_offload_capa = DEV_TX_OFFLOAD_VLAN_INSERT |
@@ -1376,6 +1384,33 @@ static void hinic_deinit_mac_addr(struct rte_eth_dev *eth_dev)
eth_dev->data->name);
}
+static int hinic_dev_set_mtu(struct rte_eth_dev *dev, uint16_t mtu)
+{
+ int ret = 0;
+ struct hinic_nic_dev *nic_dev = HINIC_ETH_DEV_TO_PRIVATE_NIC_DEV(dev);
+
+ PMD_DRV_LOG(INFO, "Set port mtu, port_id: %d, mtu: %d, max_pkt_len: %d",
+ dev->data->port_id, mtu, HINIC_MTU_TO_PKTLEN(mtu));
+
+ if (mtu < HINIC_MIN_MTU_SIZE || mtu > HINIC_MAX_MTU_SIZE) {
+ PMD_DRV_LOG(ERR, "Invalid mtu: %d, must between %d and %d",
+ mtu, HINIC_MIN_MTU_SIZE, HINIC_MAX_MTU_SIZE);
+ return -EINVAL;
+ }
+
+ ret = hinic_set_port_mtu(nic_dev->hwdev, mtu);
+ if (ret) {
+ PMD_DRV_LOG(ERR, "Set port mtu failed, ret: %d", ret);
+ return ret;
+ }
+
+ /* update max frame size */
+ dev->data->dev_conf.rxmode.max_rx_pkt_len = HINIC_MTU_TO_PKTLEN(mtu);
+ nic_dev->mtu_size = mtu;
+
+ return ret;
+}
+
static void hinic_store_vlan_filter(struct hinic_nic_dev *nic_dev,
u16 vlan_id, bool on)
{
@@ -1540,6 +1575,71 @@ static void hinic_remove_all_vlanid(struct rte_eth_dev *eth_dev)
}
}
+static int hinic_set_dev_allmulticast(struct hinic_nic_dev *nic_dev,
+ bool enable)
+{
+ u32 rx_mode_ctrl = nic_dev->rx_mode_status;
+
+ if (enable)
+ rx_mode_ctrl |= HINIC_RX_MODE_MC_ALL;
+ else
+ rx_mode_ctrl &= (~HINIC_RX_MODE_MC_ALL);
+
+ return hinic_config_rx_mode(nic_dev, rx_mode_ctrl);
+}
+
+/**
+ * DPDK callback to enable allmulticast mode.
+ *
+ * @param dev
+ * Pointer to Ethernet device structure.
+ *
+ * @return
+ * 0 on success,
+ * negative error value otherwise.
+ */
+static int hinic_dev_allmulticast_enable(struct rte_eth_dev *dev)
+{
+ int ret = HINIC_OK;
+ struct hinic_nic_dev *nic_dev = HINIC_ETH_DEV_TO_PRIVATE_NIC_DEV(dev);
+
+ ret = hinic_set_dev_allmulticast(nic_dev, true);
+ if (ret) {
+ PMD_DRV_LOG(ERR, "Enable allmulticast failed, error: %d", ret);
+ return ret;
+ }
+
+ PMD_DRV_LOG(INFO, "Enable allmulticast succeed, nic_dev: %s, port_id: %d",
+ nic_dev->proc_dev_name, dev->data->port_id);
+ return 0;
+}
+
+/**
+ * DPDK callback to disable allmulticast mode.
+ *
+ * @param dev
+ * Pointer to Ethernet device structure.
+ *
+ * @return
+ * 0 on success,
+ * negative error value otherwise.
+ */
+static int hinic_dev_allmulticast_disable(struct rte_eth_dev *dev)
+{
+ int ret = HINIC_OK;
+ struct hinic_nic_dev *nic_dev = HINIC_ETH_DEV_TO_PRIVATE_NIC_DEV(dev);
+
+ ret = hinic_set_dev_allmulticast(nic_dev, false);
+ if (ret) {
+ PMD_DRV_LOG(ERR, "Disable allmulticast failed, error: %d", ret);
+ return ret;
+ }
+
+ PMD_DRV_LOG(INFO, "Disable allmulticast succeed, nic_dev: %s, port_id: %d",
+ nic_dev->proc_dev_name, dev->data->port_id);
+ return 0;
+}
+
/**
* DPDK callback to enable promiscuous mode.
*
@@ -2421,8 +2521,11 @@ static void hinic_dev_close(struct rte_eth_dev *dev)
.tx_queue_release = hinic_tx_queue_release,
.dev_stop = hinic_dev_stop,
.dev_close = hinic_dev_close,
+ .mtu_set = hinic_dev_set_mtu,
.vlan_filter_set = hinic_vlan_filter_set,
.vlan_offload_set = hinic_vlan_offload_set,
+ .allmulticast_enable = hinic_dev_allmulticast_enable,
+ .allmulticast_disable = hinic_dev_allmulticast_disable,
.promiscuous_enable = hinic_dev_promiscuous_enable,
.promiscuous_disable = hinic_dev_promiscuous_disable,
.rss_hash_update = hinic_rss_hash_update,
@@ -2447,8 +2550,11 @@ static void hinic_dev_close(struct rte_eth_dev *dev)
.tx_queue_release = hinic_tx_queue_release,
.dev_stop = hinic_dev_stop,
.dev_close = hinic_dev_close,
+ .mtu_set = hinic_dev_set_mtu,
.vlan_filter_set = hinic_vlan_filter_set,
.vlan_offload_set = hinic_vlan_offload_set,
+ .allmulticast_enable = hinic_dev_allmulticast_enable,
+ .allmulticast_disable = hinic_dev_allmulticast_disable,
.rss_hash_update = hinic_rss_hash_update,
.rss_hash_conf_get = hinic_rss_conf_get,
.reta_update = hinic_rss_indirtbl_update,
diff --git a/drivers/net/hinic/hinic_pmd_ethdev.h b/drivers/net/hinic/hinic_pmd_ethdev.h
index 66eaf20..f7a1167 100644
--- a/drivers/net/hinic/hinic_pmd_ethdev.h
+++ b/drivers/net/hinic/hinic_pmd_ethdev.h
@@ -20,9 +20,6 @@
#define SIZE_8BYTES(size) (ALIGN((u32)(size), 8) >> 3)
-#define HINIC_PKTLEN_TO_MTU(pktlen) \
- ((pktlen) - (ETH_HLEN + ETH_CRC_LEN))
-
#define HINIC_ETH_DEV_TO_PRIVATE_NIC_DEV(dev) \
((struct hinic_nic_dev *)(dev)->data->dev_private)
--
1.8.3.1
^ permalink raw reply [flat|nested] 32+ messages in thread
* [dpdk-dev] [PATCH v2 06/17] net/hinic: add unicast and multicast MAC set
2019-09-25 14:26 [dpdk-dev] [PATCH v2 00/17] Add advanced features for Huawei hinic pmd Xiaoyun wang
` (5 preceding siblings ...)
2019-09-25 14:30 ` [dpdk-dev] [PATCH v2 05/17] net/hinic: add allmulticast mode and MTU set Xiaoyun wang
@ 2019-09-25 14:30 ` Xiaoyun wang
2019-09-25 14:30 ` [dpdk-dev] [PATCH v2 07/17] net/hinic: add fdir config interface Xiaoyun wang
` (11 subsequent siblings)
18 siblings, 0 replies; 32+ messages in thread
From: Xiaoyun wang @ 2019-09-25 14:30 UTC (permalink / raw)
To: ferruh.yigit
Cc: dev, xuanziyang2, shahar.belkar, luoxianjun, tanya.brokhman,
zhouguoyang, wulike1, Xiaoyun wang
This patch adds unicast and multicast set interfaces.
Application can add or remove unicast MAC address, also can set
multicast MAC address, tha maximum multicast list size is 2048.
Signed-off-by: Xiaoyun wang <cloud.wangxiaoyun@huawei.com>
---
drivers/net/hinic/base/hinic_pmd_niccfg.c | 55 +++++++
drivers/net/hinic/hinic_pmd_ethdev.c | 260 +++++++++++++++++++++++++++---
drivers/net/hinic/hinic_pmd_ethdev.h | 2 +
3 files changed, 296 insertions(+), 21 deletions(-)
diff --git a/drivers/net/hinic/base/hinic_pmd_niccfg.c b/drivers/net/hinic/base/hinic_pmd_niccfg.c
index 8bd7ed6..054925c 100644
--- a/drivers/net/hinic/base/hinic_pmd_niccfg.c
+++ b/drivers/net/hinic/base/hinic_pmd_niccfg.c
@@ -250,6 +250,61 @@ int hinic_get_default_mac(void *hwdev, u8 *mac_addr)
}
/**
+* hinic_update_mac - Update mac address to hardware.
+*
+* @param hwdev
+* The hardware interface of a nic device.
+* @param old_mac
+* Old mac address.
+* @param new_mac
+* New mac address.
+* @param vlan_id
+* Set 0 for mac_vlan table initialization.
+* @param func_id
+* Global function id of NIC.
+*
+* @return
+* 0 on success.
+* negative error value otherwise.
+*/
+int hinic_update_mac(void *hwdev, u8 *old_mac, u8 *new_mac, u16 vlan_id,
+ u16 func_id)
+{
+ struct hinic_port_mac_update mac_info;
+ u16 out_size = sizeof(mac_info);
+ int err;
+
+ if (!hwdev || !old_mac || !new_mac) {
+ PMD_DRV_LOG(ERR, "Hwdev, old_mac or new_mac is NULL\n");
+ return -EINVAL;
+ }
+
+ memset(&mac_info, 0, sizeof(mac_info));
+ mac_info.mgmt_msg_head.resp_aeq_num = HINIC_AEQ1;
+ mac_info.func_id = func_id;
+ mac_info.vlan_id = vlan_id;
+ memcpy(mac_info.old_mac, old_mac, ETH_ALEN);
+ memcpy(mac_info.new_mac, new_mac, ETH_ALEN);
+
+ err = l2nic_msg_to_mgmt_sync(hwdev, HINIC_PORT_CMD_UPDATE_MAC,
+ &mac_info, sizeof(mac_info),
+ &mac_info, &out_size);
+ if (err || !out_size ||
+ (mac_info.mgmt_msg_head.status &&
+ mac_info.mgmt_msg_head.status != HINIC_PF_SET_VF_ALREADY)) {
+ PMD_DRV_LOG(ERR, "Failed to update MAC, err: %d, status: 0x%x, out size: 0x%x\n",
+ err, mac_info.mgmt_msg_head.status, out_size);
+ return -EINVAL;
+ }
+ if (mac_info.mgmt_msg_head.status == HINIC_PF_SET_VF_ALREADY) {
+ PMD_DRV_LOG(WARNING, "PF has already set vf mac, Ignore update operation.\n");
+ return HINIC_PF_SET_VF_ALREADY;
+ }
+
+ return 0;
+}
+
+/**
* hinic_set_port_mtu - Set MTU to port.
*
* @param hwdev
diff --git a/drivers/net/hinic/hinic_pmd_ethdev.c b/drivers/net/hinic/hinic_pmd_ethdev.c
index 8eeeaea..16dc1cd 100644
--- a/drivers/net/hinic/hinic_pmd_ethdev.c
+++ b/drivers/net/hinic/hinic_pmd_ethdev.c
@@ -51,8 +51,8 @@
#define NR_MAX_COS 8
#define HINIC_MIN_RX_BUF_SIZE 1024
-#define HINIC_MAX_MAC_ADDRS 1
-
+#define HINIC_MAX_UC_MAC_ADDRS 128
+#define HINIC_MAX_MC_MAC_ADDRS 2048
/*
* vlan_id is a 12 bit number.
* The VFTA array is actually a 4096 bit array, 128 of 32bit elements.
@@ -716,7 +716,7 @@ static void hinic_get_speed_capa(struct rte_eth_dev *dev, uint32_t *speed_capa)
info->max_tx_queues = nic_dev->nic_cap.max_sqs;
info->min_rx_bufsize = HINIC_MIN_RX_BUF_SIZE;
info->max_rx_pktlen = HINIC_MAX_JUMBO_FRAME_SIZE;
- info->max_mac_addrs = HINIC_MAX_MAC_ADDRS;
+ info->max_mac_addrs = HINIC_MAX_UC_MAC_ADDRS;
hinic_get_speed_capa(dev, &info->speed_capa);
info->rx_queue_offload_capa = 0;
@@ -1340,21 +1340,41 @@ static int hinic_init_mac_addr(struct rte_eth_dev *eth_dev)
if (rc)
return rc;
- memmove(eth_dev->data->mac_addrs->addr_bytes,
- addr_bytes, RTE_ETHER_ADDR_LEN);
-
- if (rte_is_zero_ether_addr(eth_dev->data->mac_addrs))
- hinic_gen_random_mac_addr(eth_dev->data->mac_addrs);
+ rte_ether_addr_copy((struct rte_ether_addr *)addr_bytes,
+ ð_dev->data->mac_addrs[0]);
+ if (rte_is_zero_ether_addr(ð_dev->data->mac_addrs[0]))
+ hinic_gen_random_mac_addr(ð_dev->data->mac_addrs[0]);
func_id = hinic_global_func_id(nic_dev->hwdev);
- rc = hinic_set_mac(nic_dev->hwdev, eth_dev->data->mac_addrs->addr_bytes,
- 0, func_id);
+ rc = hinic_set_mac(nic_dev->hwdev,
+ eth_dev->data->mac_addrs[0].addr_bytes,
+ 0, func_id);
if (rc && rc != HINIC_PF_SET_VF_ALREADY)
return rc;
+ rte_ether_addr_copy(ð_dev->data->mac_addrs[0],
+ &nic_dev->default_addr);
+
return 0;
}
+static void hinic_delete_mc_addr_list(struct hinic_nic_dev *nic_dev)
+{
+ u16 func_id;
+ u32 i;
+
+ func_id = hinic_global_func_id(nic_dev->hwdev);
+
+ for (i = 0; i < HINIC_MAX_MC_MAC_ADDRS; i++) {
+ if (rte_is_zero_ether_addr(&nic_dev->mc_list[i]))
+ break;
+
+ hinic_del_mac(nic_dev->hwdev, nic_dev->mc_list[i].addr_bytes,
+ 0, func_id);
+ memset(&nic_dev->mc_list[i], 0, sizeof(struct rte_ether_addr));
+ }
+}
+
/**
* Deinit mac_vlan table in NIC.
*
@@ -1369,19 +1389,29 @@ static void hinic_deinit_mac_addr(struct rte_eth_dev *eth_dev)
{
struct hinic_nic_dev *nic_dev =
HINIC_ETH_DEV_TO_PRIVATE_NIC_DEV(eth_dev);
- int rc;
u16 func_id = 0;
-
- if (rte_is_zero_ether_addr(eth_dev->data->mac_addrs))
- return;
+ int rc;
+ int i;
func_id = hinic_global_func_id(nic_dev->hwdev);
- rc = hinic_del_mac(nic_dev->hwdev,
- eth_dev->data->mac_addrs->addr_bytes,
- 0, func_id);
- if (rc && rc != HINIC_PF_SET_VF_ALREADY)
- PMD_DRV_LOG(ERR, "Delete mac table failed, dev_name: %s",
- eth_dev->data->name);
+
+ for (i = 0; i < HINIC_MAX_UC_MAC_ADDRS; i++) {
+ if (rte_is_zero_ether_addr(ð_dev->data->mac_addrs[i]))
+ continue;
+
+ rc = hinic_del_mac(nic_dev->hwdev,
+ eth_dev->data->mac_addrs[i].addr_bytes,
+ 0, func_id);
+ if (rc && rc != HINIC_PF_SET_VF_ALREADY)
+ PMD_DRV_LOG(ERR, "Delete mac table failed, dev_name: %s",
+ eth_dev->data->name);
+
+ memset(ð_dev->data->mac_addrs[i], 0,
+ sizeof(struct rte_ether_addr));
+ }
+
+ /* delete multicast mac addrs */
+ hinic_delete_mc_addr_list(nic_dev);
}
static int hinic_dev_set_mtu(struct rte_eth_dev *dev, uint16_t mtu)
@@ -2089,6 +2119,169 @@ static int hinic_dev_xstats_get_names(struct rte_eth_dev *dev,
return count;
}
+/**
+ * DPDK callback to set mac address
+ *
+ * @param dev
+ * Pointer to Ethernet device structure.
+ * @param addr
+ * Pointer to mac address
+ * @return
+ * 0 on success, negative error value otherwise.
+ */
+static int hinic_set_mac_addr(struct rte_eth_dev *dev,
+ struct rte_ether_addr *addr)
+{
+ struct hinic_nic_dev *nic_dev = HINIC_ETH_DEV_TO_PRIVATE_NIC_DEV(dev);
+ u16 func_id;
+ int err;
+
+ func_id = hinic_global_func_id(nic_dev->hwdev);
+ err = hinic_update_mac(nic_dev->hwdev, nic_dev->default_addr.addr_bytes,
+ addr->addr_bytes, 0, func_id);
+ if (err)
+ return err;
+
+ rte_ether_addr_copy(addr, &nic_dev->default_addr);
+
+ PMD_DRV_LOG(INFO, "Set new mac address %02x:%02x:%02x:%02x:%02x:%02x\n",
+ addr->addr_bytes[0], addr->addr_bytes[1],
+ addr->addr_bytes[2], addr->addr_bytes[3],
+ addr->addr_bytes[4], addr->addr_bytes[5]);
+
+ return 0;
+}
+
+/**
+ * DPDK callback to remove a MAC address.
+ *
+ * @param dev
+ * Pointer to Ethernet device structure.
+ * @param index
+ * MAC address index.
+ */
+static void hinic_mac_addr_remove(struct rte_eth_dev *dev, uint32_t index)
+{
+ struct hinic_nic_dev *nic_dev = HINIC_ETH_DEV_TO_PRIVATE_NIC_DEV(dev);
+ u16 func_id;
+ int ret;
+
+ if (index >= HINIC_MAX_UC_MAC_ADDRS) {
+ PMD_DRV_LOG(INFO, "Remove mac index(%u) is out of range",
+ index);
+ return;
+ }
+
+ func_id = hinic_global_func_id(nic_dev->hwdev);
+ ret = hinic_del_mac(nic_dev->hwdev,
+ dev->data->mac_addrs[index].addr_bytes, 0, func_id);
+ if (ret)
+ return;
+
+ memset(&dev->data->mac_addrs[index], 0, sizeof(struct rte_ether_addr));
+}
+
+/**
+ * DPDK callback to add a MAC address.
+ *
+ * @param dev
+ * Pointer to Ethernet device structure.
+ * @param mac_addr
+ * MAC address to register.
+ * @param index
+ * MAC address index.
+ * @param vmdq
+ * VMDq pool index to associate address with (ignored).
+ *
+ * @return
+ * 0 on success, a negative errno value otherwise and rte_errno is set.
+ */
+
+static int hinic_mac_addr_add(struct rte_eth_dev *dev,
+ struct rte_ether_addr *mac_addr, uint32_t index,
+ __rte_unused uint32_t vmdq)
+{
+ struct hinic_nic_dev *nic_dev = HINIC_ETH_DEV_TO_PRIVATE_NIC_DEV(dev);
+ unsigned int i;
+ u16 func_id;
+ int ret;
+
+ if (index >= HINIC_MAX_UC_MAC_ADDRS) {
+ PMD_DRV_LOG(INFO, "Add mac index(%u) is out of range,", index);
+ return -EINVAL;
+ }
+
+ /* First, make sure this address isn't already configured. */
+ for (i = 0; (i != HINIC_MAX_UC_MAC_ADDRS); ++i) {
+ /* Skip this index, it's going to be reconfigured. */
+ if (i == index)
+ continue;
+
+ if (memcmp(&dev->data->mac_addrs[i],
+ mac_addr, sizeof(*mac_addr)))
+ continue;
+
+ PMD_DRV_LOG(INFO, "MAC address already configured");
+ return -EADDRINUSE;
+ }
+
+ func_id = hinic_global_func_id(nic_dev->hwdev);
+ ret = hinic_set_mac(nic_dev->hwdev, mac_addr->addr_bytes, 0, func_id);
+ if (ret)
+ return ret;
+
+ dev->data->mac_addrs[index] = *mac_addr;
+ return 0;
+}
+
+/**
+ * DPDK callback to set multicast mac address
+ *
+ * @param dev
+ * Pointer to Ethernet device structure.
+ * @param mc_addr_set
+ * Pointer to multicast mac address
+ * @param nb_mc_addr
+ * mc addr count
+ * @return
+ * 0 on success, negative error value otherwise.
+ */
+static int hinic_set_mc_addr_list(struct rte_eth_dev *dev,
+ struct rte_ether_addr *mc_addr_set,
+ uint32_t nb_mc_addr)
+{
+ struct hinic_nic_dev *nic_dev = HINIC_ETH_DEV_TO_PRIVATE_NIC_DEV(dev);
+ u16 func_id;
+ int ret;
+ u32 i;
+
+ func_id = hinic_global_func_id(nic_dev->hwdev);
+
+ /* delete old multi_cast addrs firstly */
+ hinic_delete_mc_addr_list(nic_dev);
+
+ if (nb_mc_addr > HINIC_MAX_MC_MAC_ADDRS)
+ goto allmulti;
+
+ for (i = 0; i < nb_mc_addr; i++) {
+ ret = hinic_set_mac(nic_dev->hwdev, mc_addr_set[i].addr_bytes,
+ 0, func_id);
+ /* if add mc addr failed, set all multi_cast */
+ if (ret) {
+ hinic_delete_mc_addr_list(nic_dev);
+ goto allmulti;
+ }
+
+ rte_ether_addr_copy(&mc_addr_set[i], &nic_dev->mc_list[i]);
+ }
+
+ return 0;
+
+allmulti:
+ hinic_dev_allmulticast_enable(dev);
+
+ return 0;
+}
static int hinic_set_default_pause_feature(struct hinic_nic_dev *nic_dev)
{
@@ -2537,6 +2730,10 @@ static void hinic_dev_close(struct rte_eth_dev *dev)
.xstats_get = hinic_dev_xstats_get,
.xstats_reset = hinic_dev_xstats_reset,
.xstats_get_names = hinic_dev_xstats_get_names,
+ .mac_addr_set = hinic_set_mac_addr,
+ .mac_addr_remove = hinic_mac_addr_remove,
+ .mac_addr_add = hinic_mac_addr_add,
+ .set_mc_addr_list = hinic_set_mc_addr_list,
};
static const struct eth_dev_ops hinic_pmd_vf_ops = {
@@ -2564,6 +2761,10 @@ static void hinic_dev_close(struct rte_eth_dev *dev)
.xstats_get = hinic_dev_xstats_get,
.xstats_reset = hinic_dev_xstats_reset,
.xstats_get_names = hinic_dev_xstats_get_names,
+ .mac_addr_set = hinic_set_mac_addr,
+ .mac_addr_remove = hinic_mac_addr_remove,
+ .mac_addr_add = hinic_mac_addr_add,
+ .set_mc_addr_list = hinic_set_mc_addr_list,
};
static int hinic_func_init(struct rte_eth_dev *eth_dev)
@@ -2571,6 +2772,7 @@ static int hinic_func_init(struct rte_eth_dev *eth_dev)
struct rte_pci_device *pci_dev;
struct rte_ether_addr *eth_addr;
struct hinic_nic_dev *nic_dev;
+ u32 mac_size;
int rc;
pci_dev = RTE_ETH_DEV_TO_PCI(eth_dev);
@@ -2597,7 +2799,8 @@ static int hinic_func_init(struct rte_eth_dev *eth_dev)
pci_dev->addr.devid, pci_dev->addr.function);
/* alloc mac_addrs */
- eth_addr = rte_zmalloc("hinic_mac", sizeof(*eth_addr), 0);
+ mac_size = HINIC_MAX_UC_MAC_ADDRS * sizeof(struct rte_ether_addr);
+ eth_addr = rte_zmalloc("hinic_mac", mac_size, 0);
if (!eth_addr) {
PMD_DRV_LOG(ERR, "Allocate ethernet addresses' memory failed, dev_name: %s",
eth_dev->data->name);
@@ -2606,6 +2809,15 @@ static int hinic_func_init(struct rte_eth_dev *eth_dev)
}
eth_dev->data->mac_addrs = eth_addr;
+ mac_size = HINIC_MAX_MC_MAC_ADDRS * sizeof(struct rte_ether_addr);
+ nic_dev->mc_list = rte_zmalloc("hinic_mc", mac_size, 0);
+ if (!nic_dev->mc_list) {
+ PMD_DRV_LOG(ERR, "Allocate mcast address' memory failed, dev_name: %s",
+ eth_dev->data->name);
+ rc = -ENOMEM;
+ goto mc_addr_fail;
+ }
+
/*
* Pass the information to the rte_eth_dev_close() that it should also
* release the private port resources.
@@ -2670,6 +2882,10 @@ static int hinic_func_init(struct rte_eth_dev *eth_dev)
hinic_nic_dev_destroy(eth_dev);
create_nic_dev_fail:
+ rte_free(nic_dev->mc_list);
+ nic_dev->mc_list = NULL;
+
+mc_addr_fail:
rte_free(eth_addr);
eth_dev->data->mac_addrs = NULL;
@@ -2714,6 +2930,8 @@ static int hinic_dev_uninit(struct rte_eth_dev *dev)
dev->rx_pkt_burst = NULL;
dev->tx_pkt_burst = NULL;
+ rte_free(nic_dev->mc_list);
+
rte_free(dev->data->mac_addrs);
dev->data->mac_addrs = NULL;
diff --git a/drivers/net/hinic/hinic_pmd_ethdev.h b/drivers/net/hinic/hinic_pmd_ethdev.h
index f7a1167..b4f93ad 100644
--- a/drivers/net/hinic/hinic_pmd_ethdev.h
+++ b/drivers/net/hinic/hinic_pmd_ethdev.h
@@ -56,6 +56,8 @@ struct hinic_nic_dev {
u32 vfta[HINIC_VFTA_SIZE]; /* VLAN bitmap */
+ struct rte_ether_addr default_addr;
+ struct rte_ether_addr *mc_list;
/* info */
unsigned int flags;
struct nic_service_cap nic_cap;
--
1.8.3.1
^ permalink raw reply [flat|nested] 32+ messages in thread
* [dpdk-dev] [PATCH v2 07/17] net/hinic: add fdir config interface
2019-09-25 14:26 [dpdk-dev] [PATCH v2 00/17] Add advanced features for Huawei hinic pmd Xiaoyun wang
` (6 preceding siblings ...)
2019-09-25 14:30 ` [dpdk-dev] [PATCH v2 06/17] net/hinic: add unicast and multicast MAC set Xiaoyun wang
@ 2019-09-25 14:30 ` Xiaoyun wang
2019-09-26 18:48 ` Ferruh Yigit
2019-09-25 14:30 ` [dpdk-dev] [PATCH v2 08/17] net/hinic: add fdir validate flow operations Xiaoyun wang
` (10 subsequent siblings)
18 siblings, 1 reply; 32+ messages in thread
From: Xiaoyun wang @ 2019-09-25 14:30 UTC (permalink / raw)
To: ferruh.yigit
Cc: dev, xuanziyang2, shahar.belkar, luoxianjun, tanya.brokhman,
zhouguoyang, wulike1, Xiaoyun wang
This patch adds fdir config operation, including set fdir filter,
normal filter, set and clear fdir tcam.
Signed-off-by: Xiaoyun wang <cloud.wangxiaoyun@huawei.com>
---
drivers/net/hinic/base/hinic_pmd_cmd.h | 2 +
drivers/net/hinic/base/hinic_pmd_niccfg.c | 205 ++++++++++++++++++++++++++++++
drivers/net/hinic/base/hinic_pmd_niccfg.h | 134 +++++++++++++++++++
3 files changed, 341 insertions(+)
diff --git a/drivers/net/hinic/base/hinic_pmd_cmd.h b/drivers/net/hinic/base/hinic_pmd_cmd.h
index 5c38b5f..6b3dcf3 100644
--- a/drivers/net/hinic/base/hinic_pmd_cmd.h
+++ b/drivers/net/hinic/base/hinic_pmd_cmd.h
@@ -140,6 +140,8 @@ enum hinic_port_cmd {
HINIC_PORT_CMD_SET_VHD_CFG = 0xF7,
HINIC_PORT_CMD_SET_LINK_FOLLOW = 0xF8,
+ HINIC_PORT_CMD_Q_FILTER = 0xFC,
+ HINIC_PORT_CMD_TCAM_FILTER = 0xFE,
HINIC_PORT_CMD_SET_VLAN_FILTER = 0xFF
};
diff --git a/drivers/net/hinic/base/hinic_pmd_niccfg.c b/drivers/net/hinic/base/hinic_pmd_niccfg.c
index 054925c..53d981b 100644
--- a/drivers/net/hinic/base/hinic_pmd_niccfg.c
+++ b/drivers/net/hinic/base/hinic_pmd_niccfg.c
@@ -18,6 +18,23 @@
buf_in, in_size, \
buf_out, out_size, 0)
+
+#define TCAM_SET 0x1
+#define TCAM_CLEAR 0x2
+
+struct hinic_port_qfilter_info {
+ struct hinic_mgmt_msg_head mgmt_msg_head;
+
+ u16 func_id;
+ u8 normal_type_enable;
+ u8 filter_type_enable;
+ u8 filter_enable;
+ u8 filter_type;
+ u8 qid;
+ u8 fdir_flag;
+ u32 key;
+};
+
/**
* hinic_init_function_table - Initialize function table.
*
@@ -1662,3 +1679,191 @@ int hinic_vf_get_default_cos(struct hinic_hwdev *hwdev, u8 *cos_id)
return 0;
}
+
+/**
+ * hinic_set_fdir_filter - Set fdir filter for control path
+ * packet to notify firmware.
+ *
+ * @param hwdev
+ * The hardware interface of a nic device.
+ * @param filter_type
+ * Packet type to filter.
+ * @param qid
+ * Rx qid to filter.
+ * @param type_enable
+ * The status of pkt type filter.
+ * @param enable
+ * Fdir function Enable or Disable.
+ * @return
+ * 0 on success,
+ * negative error value otherwise.
+ */
+int hinic_set_fdir_filter(void *hwdev, u8 filter_type, u8 qid, u8 type_enable,
+ bool enable)
+{
+ struct hinic_port_qfilter_info port_filer_cmd;
+ u16 out_size = sizeof(port_filer_cmd);
+ int err;
+
+ if (!hwdev)
+ return -EINVAL;
+
+ memset(&port_filer_cmd, 0, sizeof(port_filer_cmd));
+ port_filer_cmd.mgmt_msg_head.resp_aeq_num = HINIC_AEQ1;
+ port_filer_cmd.func_id = hinic_global_func_id(hwdev);
+ port_filer_cmd.filter_enable = (u8)enable;
+ port_filer_cmd.filter_type = filter_type;
+ port_filer_cmd.qid = qid;
+ port_filer_cmd.filter_type_enable = type_enable;
+ port_filer_cmd.fdir_flag = 0;
+
+ err = l2nic_msg_to_mgmt_sync(hwdev, HINIC_PORT_CMD_Q_FILTER,
+ &port_filer_cmd, sizeof(port_filer_cmd),
+ &port_filer_cmd, &out_size);
+ if (err || !out_size || port_filer_cmd.mgmt_msg_head.status) {
+ PMD_DRV_LOG(ERR, "Set port Q filter failed, err: %d, status: 0x%x, out size: 0x%x, type: 0x%x,"
+ " enable: 0x%x, qid: 0x%x, filter_type_enable: 0x%x\n",
+ err, port_filer_cmd.mgmt_msg_head.status, out_size,
+ filter_type, enable, qid, type_enable);
+ return -EFAULT;
+ }
+
+ return 0;
+}
+
+/**
+ * hinic_set_normal_filter - Set fdir filter for IO path packet.
+ *
+ * @param hwdev
+ * The hardware interface of a nic device.
+ * @param qid
+ * Rx qid to filter.
+ * @param normal_type_enable
+ * IO path packet function Enable or Disable
+ * @param key
+ * IO path packet filter key value, such as DIP from pkt.
+ * @param enable
+ * Fdir function Enable or Disable.
+ * @param flag
+ * Filter flag, such as dip or others.
+ * @return
+ * 0 on success,
+ * negative error value otherwise.
+ */
+int hinic_set_normal_filter(void *hwdev, u8 qid, u8 normal_type_enable,
+ u32 key, bool enable, u8 flag)
+{
+ struct hinic_port_qfilter_info port_filer_cmd;
+ u16 out_size = sizeof(port_filer_cmd);
+ int err;
+
+ if (!hwdev)
+ return -EINVAL;
+
+ memset(&port_filer_cmd, 0, sizeof(port_filer_cmd));
+ port_filer_cmd.mgmt_msg_head.resp_aeq_num = HINIC_AEQ1;
+ port_filer_cmd.func_id = hinic_global_func_id(hwdev);
+ port_filer_cmd.filter_enable = (u8)enable;
+ port_filer_cmd.qid = qid;
+ port_filer_cmd.normal_type_enable = normal_type_enable;
+ port_filer_cmd.fdir_flag = flag; /* fdir flag: support dip */
+ port_filer_cmd.key = key;
+
+ err = l2nic_msg_to_mgmt_sync(hwdev, HINIC_PORT_CMD_Q_FILTER,
+ &port_filer_cmd, sizeof(port_filer_cmd),
+ &port_filer_cmd, &out_size);
+ if (err || !out_size || port_filer_cmd.mgmt_msg_head.status) {
+ PMD_DRV_LOG(ERR, "Set normal filter failed, err: %d, status: 0x%x, out size: 0x%x, fdir_flag: 0x%x,"
+ " enable: 0x%x, qid: 0x%x, normal_type_enable: 0x%x, key:0x%x\n",
+ err, port_filer_cmd.mgmt_msg_head.status, out_size,
+ flag, enable, qid, normal_type_enable, key);
+ return -EFAULT;
+ }
+
+ return 0;
+}
+
+/**
+ * hinic_set_fdir_tcam - Set fdir filter for control packet
+ * by tcam table to notify hardware.
+ *
+ * @param hwdev
+ * The hardware interface of a nic device.
+ * @param type_mask
+ * Index of TCAM.
+ * @param filter_rule
+ * TCAM rule for control packet, such as lacp or bgp.
+ * @param filter_action
+ * TCAM action for control packet, such as accept or drop.
+ * @return
+ * 0 on success,
+ * negative error value otherwise.
+ */
+int hinic_set_fdir_tcam(void *hwdev, u16 type_mask,
+ struct tag_pa_rule *filter_rule,
+ struct tag_pa_action *filter_action)
+{
+ struct hinic_fdir_tcam_info port_tcam_cmd;
+ u16 out_size = sizeof(port_tcam_cmd);
+ int err;
+
+ if (!hwdev)
+ return -EINVAL;
+
+ memset(&port_tcam_cmd, 0, sizeof(port_tcam_cmd));
+ port_tcam_cmd.mgmt_msg_head.resp_aeq_num = HINIC_AEQ1;
+ port_tcam_cmd.tcam_index = type_mask;
+ port_tcam_cmd.flag = TCAM_SET;
+ memcpy((void *)&port_tcam_cmd.filter_rule,
+ (void *)filter_rule, sizeof(struct tag_pa_rule));
+ memcpy((void *)&port_tcam_cmd.filter_action,
+ (void *)filter_action, sizeof(struct tag_pa_action));
+
+ err = l2nic_msg_to_mgmt_sync(hwdev, HINIC_PORT_CMD_TCAM_FILTER,
+ &port_tcam_cmd, sizeof(port_tcam_cmd),
+ &port_tcam_cmd, &out_size);
+ if (err || !out_size || port_tcam_cmd.mgmt_msg_head.status) {
+ PMD_DRV_LOG(ERR, "Set tcam table failed, err: %d, status: 0x%x, out size: 0x%x\n",
+ err, port_tcam_cmd.mgmt_msg_head.status, out_size);
+ return -EFAULT;
+ }
+
+ return 0;
+}
+
+/**
+ * hinic_clear_fdir_tcam - Clear fdir filter TCAM table for control packet.
+ *
+ * @param hwdev
+ * The hardware interface of a nic device.
+ * @param type_mask
+ * Index of TCAM.
+ * @return
+ * 0 on success,
+ * negative error value otherwise.
+ */
+int hinic_clear_fdir_tcam(void *hwdev, u16 type_mask)
+{
+ struct hinic_fdir_tcam_info port_tcam_cmd;
+ u16 out_size = sizeof(port_tcam_cmd);
+ int err;
+
+ if (!hwdev)
+ return -EINVAL;
+
+ memset(&port_tcam_cmd, 0, sizeof(port_tcam_cmd));
+ port_tcam_cmd.mgmt_msg_head.resp_aeq_num = HINIC_AEQ1;
+ port_tcam_cmd.tcam_index = type_mask;
+ port_tcam_cmd.flag = TCAM_CLEAR;
+
+ err = l2nic_msg_to_mgmt_sync(hwdev, HINIC_PORT_CMD_TCAM_FILTER,
+ &port_tcam_cmd, sizeof(port_tcam_cmd),
+ &port_tcam_cmd, &out_size);
+ if (err || !out_size || port_tcam_cmd.mgmt_msg_head.status) {
+ PMD_DRV_LOG(ERR, "Clear tcam table failed, err: %d, status: 0x%x, out size: 0x%x\n",
+ err, port_tcam_cmd.mgmt_msg_head.status, out_size);
+ return -EFAULT;
+ }
+
+ return 0;
+}
diff --git a/drivers/net/hinic/base/hinic_pmd_niccfg.h b/drivers/net/hinic/base/hinic_pmd_niccfg.h
index d19a834..e8ce332 100644
--- a/drivers/net/hinic/base/hinic_pmd_niccfg.h
+++ b/drivers/net/hinic/base/hinic_pmd_niccfg.h
@@ -609,6 +609,129 @@ struct hinic_port_anti_attack_rate {
u32 xbs; /* eXtended Burst Size */
};
+struct pa_u8_s {
+ u8 val8;
+ u8 mask8;
+};
+
+struct pa_u16_s {
+ u16 val16;
+ u16 mask16;
+};
+
+struct pa_u32_s {
+ u32 val32;
+ u32 mask32;
+};
+
+struct pa_u48_s {
+ u8 val8[6];
+ u8 mask8[6];
+};
+
+struct pa_u64_s {
+ u8 val8[8];
+ u8 mask8[8];
+};
+
+struct tag_pa_eth_ip_header {
+ struct pa_u8_s ip_ver; /* 3bit */
+ struct pa_u8_s ipv4_option_flag; /* 1bit */
+ /* 8bit ipv4 option or ipv6 next header */
+ struct pa_u8_s protocol;
+ struct pa_u8_s dscp; /* 6bit DSCP */
+};
+
+struct tag_pa_common_l2_header {
+ struct pa_u48_s dmac; /* dmac 48bit */
+ struct pa_u16_s eth_type; /* ethernet type/length 16bit */
+ struct pa_u8_s tag_flag; /* tag flag: 4bit */
+ struct pa_u8_s np2np_hdr_qindex; /* NP2NP Header Qindex 4bit */
+ struct pa_u8_s e_tag_pcp; /* 3bit */
+ struct pa_u8_s vlan_layer; /* 2bit */
+ struct pa_u8_s s_tag; /* 3bit */
+ struct pa_u8_s c_tag; /* 3bit */
+ struct pa_u16_s vlan_id; /* 12bit */
+};
+
+struct tag_pa_tcp {
+ struct pa_u16_s sport; /* 16bit */
+ struct pa_u16_s dport; /* 16bit */
+ struct pa_u16_s tcp_flag; /* 6bit */
+};
+
+struct tag_pa_udp {
+ struct pa_u16_s sport; /* 16bit */
+ struct pa_u16_s dport; /* 16bit */
+ /* 8bit :
+ * 1.udp dport=67/68 && ipv4 protocol=0x11
+ * 2.udp dport=546/547 && ipv6 next header=0x11
+ * 3. do not care
+ */
+ struct pa_u8_s dhcp_op_or_msg_type;
+};
+
+/* ICMP:
+ * ipv4 protocol = 0x1
+ * ipv6 next header = 0x3A
+ */
+struct tag_pa_icmp {
+ struct pa_u8_s type; /* 8bit */
+ struct pa_u8_s code; /* 8bit */
+};
+
+/* IGMP:
+ * ipv4 protocol = 0x2
+ */
+struct tag_pa_ipv4_igmp {
+ struct pa_u32_s dip; /* 32bit */
+ struct pa_u8_s type; /* 8bit */
+};
+
+struct tag_pa_rule {
+ struct pa_u8_s ncsi_flag; /* 1bit valid */
+ struct tag_pa_common_l2_header l2_header;
+
+ u8 eth_type;
+
+ struct pa_u64_s eth_other; /* eth_type=other 64bit */
+ struct pa_u8_s eth_roce_opcode; /* eth_type=roce 8bit opcode */
+
+ struct tag_pa_eth_ip_header ip_header; /* eth_type=ip */
+
+ u8 ip_protocol_type;
+
+ struct tag_pa_tcp eth_ip_tcp; /* eth_type=ip && ip_protocol = tcp */
+ struct tag_pa_udp eth_ip_udp; /* eth_type=ip && ip_protocol = udp */
+ struct tag_pa_icmp eth_ip_icmp; /* eth_type=ip && ip_protocol = icmp */
+
+ /* eth_type=ip && ip_protocol = ipv4_igmp */
+ struct tag_pa_ipv4_igmp eth_ipv4_igmp;
+
+ /* eth_type=ip && ip_protocol = sctp;
+ * 16bit ipv4 protocol=0x84 or ipv6 nhr=0x84
+ */
+ struct pa_u16_s eth_ip_sctp;
+};
+
+struct tag_pa_action {
+ u16 pkt_type;
+ u8 err_type;
+ u8 pri;
+ u8 fwd_action;
+ u8 push_len;
+};
+
+struct hinic_fdir_tcam_info {
+ struct hinic_mgmt_msg_head mgmt_msg_head;
+
+ u16 tcam_index;
+ u8 flag; /* clear or set tcam table flag */
+ u8 rsvd1;
+ struct tag_pa_rule filter_rule;
+ struct tag_pa_action filter_action;
+};
+
int hinic_set_mac(void *hwdev, u8 *mac_addr, u16 vlan_id, u16 func_id);
int hinic_del_mac(void *hwdev, u8 *mac_addr, u16 vlan_id, u16 func_id);
@@ -702,4 +825,15 @@ int hinic_set_link_status_follow(void *hwdev,
int hinic_vf_get_default_cos(struct hinic_hwdev *hwdev, u8 *cos_id);
+int hinic_set_fdir_filter(void *hwdev, u8 filter_type, u8 qid,
+ u8 type_enable, bool enable);
+
+int hinic_set_normal_filter(void *hwdev, u8 qid, u8 normal_type_enable,
+ u32 key, bool enable, u8 flag);
+
+int hinic_set_fdir_tcam(void *hwdev, u16 type_mask,
+ struct tag_pa_rule *filter_rule, struct tag_pa_action *filter_action);
+
+int hinic_clear_fdir_tcam(void *hwdev, u16 type_mask);
+
#endif /* _HINIC_PMD_NICCFG_H_ */
--
1.8.3.1
^ permalink raw reply [flat|nested] 32+ messages in thread
* [dpdk-dev] [PATCH v2 08/17] net/hinic: add fdir validate flow operations
2019-09-25 14:26 [dpdk-dev] [PATCH v2 00/17] Add advanced features for Huawei hinic pmd Xiaoyun wang
` (7 preceding siblings ...)
2019-09-25 14:30 ` [dpdk-dev] [PATCH v2 07/17] net/hinic: add fdir config interface Xiaoyun wang
@ 2019-09-25 14:30 ` Xiaoyun wang
2019-09-25 14:30 ` [dpdk-dev] [PATCH v2 09/17] net/hinic: create and destroy ntuple filter Xiaoyun wang
` (9 subsequent siblings)
18 siblings, 0 replies; 32+ messages in thread
From: Xiaoyun wang @ 2019-09-25 14:30 UTC (permalink / raw)
To: ferruh.yigit
Cc: dev, xuanziyang2, shahar.belkar, luoxianjun, tanya.brokhman,
zhouguoyang, wulike1, Xiaoyun wang
This patch is to validate the filter rules, which includes
ntuple filter, ethertype filter and fdir filter. The packets
type that supported are BGP,VRRP,LACP,ARP and ICMP.
Signed-off-by: Xiaoyun wang <cloud.wangxiaoyun@huawei.com>
---
drivers/net/hinic/Makefile | 1 +
drivers/net/hinic/hinic_pmd_ethdev.c | 42 ++
drivers/net/hinic/hinic_pmd_ethdev.h | 24 +
drivers/net/hinic/hinic_pmd_flow.c | 1172 ++++++++++++++++++++++++++++++++++
drivers/net/hinic/meson.build | 1 +
5 files changed, 1240 insertions(+)
create mode 100644 drivers/net/hinic/hinic_pmd_flow.c
diff --git a/drivers/net/hinic/Makefile b/drivers/net/hinic/Makefile
index 20a338e..b78fd8d 100644
--- a/drivers/net/hinic/Makefile
+++ b/drivers/net/hinic/Makefile
@@ -60,6 +60,7 @@ SRCS-$(CONFIG_RTE_LIBRTE_HINIC_PMD) += hinic_pmd_niccfg.c
SRCS-$(CONFIG_RTE_LIBRTE_HINIC_PMD) += hinic_pmd_nicio.c
SRCS-$(CONFIG_RTE_LIBRTE_HINIC_PMD) += hinic_pmd_wq.c
SRCS-$(CONFIG_RTE_LIBRTE_HINIC_PMD) += hinic_pmd_mbox.c
+SRCS-$(CONFIG_RTE_LIBRTE_HINIC_PMD) += hinic_pmd_flow.c
SRCS-$(CONFIG_RTE_LIBRTE_HINIC_PMD) += hinic_pmd_ethdev.c
SRCS-$(CONFIG_RTE_LIBRTE_HINIC_PMD) += hinic_pmd_rx.c
diff --git a/drivers/net/hinic/hinic_pmd_ethdev.c b/drivers/net/hinic/hinic_pmd_ethdev.c
index 16dc1cd..9f7e6f5 100644
--- a/drivers/net/hinic/hinic_pmd_ethdev.c
+++ b/drivers/net/hinic/hinic_pmd_ethdev.c
@@ -2283,6 +2283,46 @@ static int hinic_set_mc_addr_list(struct rte_eth_dev *dev,
return 0;
}
+/**
+ * DPDK callback to manage filter operations
+ *
+ * @param dev
+ * Pointer to Ethernet device structure.
+ * @param filter_type
+ * Filter type.
+ * @param filter_op
+ * Operation to perform.
+ * @param arg
+ * Pointer to operation-specific structure.
+ *
+ * @return
+ * 0 on success, negative errno value on failure.
+ */
+static int hinic_dev_filter_ctrl(struct rte_eth_dev *dev,
+ enum rte_filter_type filter_type,
+ enum rte_filter_op filter_op,
+ void *arg)
+{
+ struct hinic_nic_dev *nic_dev = HINIC_ETH_DEV_TO_PRIVATE_NIC_DEV(dev);
+ int func_id = hinic_global_func_id(nic_dev->hwdev);
+
+ switch (filter_type) {
+ case RTE_ETH_FILTER_GENERIC:
+ if (filter_op != RTE_ETH_FILTER_GET)
+ return -EINVAL;
+ *(const void **)arg = &hinic_flow_ops;
+ break;
+ default:
+ PMD_DRV_LOG(INFO, "Filter type (%d) not supported",
+ filter_type);
+ return -EINVAL;
+ }
+
+ PMD_DRV_LOG(INFO, "Set filter_ctrl succeed, func_id: 0x%x, filter_type: 0x%x,"
+ "filter_op: 0x%x.", func_id, filter_type, filter_op);
+ return 0;
+}
+
static int hinic_set_default_pause_feature(struct hinic_nic_dev *nic_dev)
{
struct nic_pause_config pause_config = {0};
@@ -2734,6 +2774,7 @@ static void hinic_dev_close(struct rte_eth_dev *dev)
.mac_addr_remove = hinic_mac_addr_remove,
.mac_addr_add = hinic_mac_addr_add,
.set_mc_addr_list = hinic_set_mc_addr_list,
+ .filter_ctrl = hinic_dev_filter_ctrl,
};
static const struct eth_dev_ops hinic_pmd_vf_ops = {
@@ -2765,6 +2806,7 @@ static void hinic_dev_close(struct rte_eth_dev *dev)
.mac_addr_remove = hinic_mac_addr_remove,
.mac_addr_add = hinic_mac_addr_add,
.set_mc_addr_list = hinic_set_mc_addr_list,
+ .filter_ctrl = hinic_dev_filter_ctrl,
};
static int hinic_func_init(struct rte_eth_dev *eth_dev)
diff --git a/drivers/net/hinic/hinic_pmd_ethdev.h b/drivers/net/hinic/hinic_pmd_ethdev.h
index b4f93ad..58a38d9 100644
--- a/drivers/net/hinic/hinic_pmd_ethdev.h
+++ b/drivers/net/hinic/hinic_pmd_ethdev.h
@@ -38,6 +38,30 @@ enum hinic_dev_status {
HINIC_DEV_INTR_EN,
};
+/* Information about the fdir mode. */
+struct hinic_hw_fdir_mask {
+ uint32_t src_ipv4_mask;
+ uint32_t dst_ipv4_mask;
+ uint16_t src_port_mask;
+ uint16_t dst_port_mask;
+};
+
+/* Flow Director attribute */
+struct hinic_atr_input {
+ u32 dst_ip;
+ u32 src_ip;
+ u16 src_port;
+ u16 dst_port;
+};
+
+struct hinic_fdir_rule {
+ struct hinic_hw_fdir_mask mask;
+ struct hinic_atr_input hinic_fdir; /* key of fdir filter */
+ uint8_t queue; /* queue assigned when matched */
+};
+
+extern const struct rte_flow_ops hinic_flow_ops;
+
/* hinic nic_device */
struct hinic_nic_dev {
/* hardware device */
diff --git a/drivers/net/hinic/hinic_pmd_flow.c b/drivers/net/hinic/hinic_pmd_flow.c
new file mode 100644
index 0000000..cf9f105
--- /dev/null
+++ b/drivers/net/hinic/hinic_pmd_flow.c
@@ -0,0 +1,1172 @@
+/* SPDX-License-Identifier: BSD-3-Clause
+ * Copyright(c) 2017 Huawei Technologies Co., Ltd
+ */
+
+#include <stdio.h>
+#include <errno.h>
+#include <stdint.h>
+#include <string.h>
+#include <unistd.h>
+
+#include <rte_byteorder.h>
+#include <rte_common.h>
+#include <rte_ether.h>
+#include <rte_ethdev.h>
+#include <rte_malloc.h>
+#include <rte_flow.h>
+#include <rte_flow_driver.h>
+#include "base/hinic_compat.h"
+#include "base/hinic_pmd_hwdev.h"
+#include "base/hinic_pmd_hwif.h"
+#include "base/hinic_pmd_wq.h"
+#include "base/hinic_pmd_cmdq.h"
+#include "base/hinic_pmd_niccfg.h"
+#include "hinic_pmd_ethdev.h"
+
+#ifndef UINT8_MAX
+#define UINT8_MAX (u8)(~((u8)0)) /* 0xFF */
+#define UINT16_MAX (u16)(~((u16)0)) /* 0xFFFF */
+#define UINT32_MAX (u32)(~((u32)0)) /* 0xFFFFFFFF */
+#define UINT64_MAX (u64)(~((u64)0)) /* 0xFFFFFFFFFFFFFFFF */
+#define ASCII_MAX (0x7F)
+#endif
+
+#define HINIC_MIN_N_TUPLE_PRIO 1
+#define HINIC_MAX_N_TUPLE_PRIO 7
+
+/**
+ * Endless loop will never happen with below assumption
+ * 1. there is at least one no-void item(END)
+ * 2. cur is before END.
+ */
+static inline const struct rte_flow_item *
+next_no_void_pattern(const struct rte_flow_item pattern[],
+ const struct rte_flow_item *cur)
+{
+ const struct rte_flow_item *next =
+ cur ? cur + 1 : &pattern[0];
+ while (1) {
+ if (next->type != RTE_FLOW_ITEM_TYPE_VOID)
+ return next;
+ next++;
+ }
+}
+
+static inline const struct rte_flow_action *
+next_no_void_action(const struct rte_flow_action actions[],
+ const struct rte_flow_action *cur)
+{
+ const struct rte_flow_action *next =
+ cur ? cur + 1 : &actions[0];
+ while (1) {
+ if (next->type != RTE_FLOW_ACTION_TYPE_VOID)
+ return next;
+ next++;
+ }
+}
+
+static int hinic_check_ethertype_attr_ele(const struct rte_flow_attr *attr,
+ struct rte_flow_error *error)
+{
+ /* Must be input direction */
+ if (!attr->ingress) {
+ rte_flow_error_set(error, EINVAL,
+ RTE_FLOW_ERROR_TYPE_ATTR_INGRESS,
+ attr, "Only support ingress.");
+ return -rte_errno;
+ }
+
+ if (attr->egress) {
+ rte_flow_error_set(error, EINVAL,
+ RTE_FLOW_ERROR_TYPE_ATTR_EGRESS,
+ attr, "Not support egress.");
+ return -rte_errno;
+ }
+
+ if (attr->priority) {
+ rte_flow_error_set(error, EINVAL,
+ RTE_FLOW_ERROR_TYPE_ATTR_PRIORITY,
+ attr, "Not support priority.");
+ return -rte_errno;
+ }
+
+ if (attr->group) {
+ rte_flow_error_set(error, EINVAL,
+ RTE_FLOW_ERROR_TYPE_ATTR_GROUP,
+ attr, "Not support group.");
+ return -rte_errno;
+ }
+
+ return 0;
+}
+
+static int hinic_check_filter_arg(const struct rte_flow_attr *attr,
+ const struct rte_flow_item *pattern,
+ const struct rte_flow_action *actions,
+ struct rte_flow_error *error)
+{
+ if (!pattern) {
+ rte_flow_error_set(error, EINVAL, RTE_FLOW_ERROR_TYPE_ITEM_NUM,
+ NULL, "NULL pattern.");
+ return -rte_errno;
+ }
+
+ if (!actions) {
+ rte_flow_error_set(error, EINVAL,
+ RTE_FLOW_ERROR_TYPE_ACTION_NUM,
+ NULL, "NULL action.");
+ return -rte_errno;
+ }
+
+ if (!attr) {
+ rte_flow_error_set(error, EINVAL, RTE_FLOW_ERROR_TYPE_ATTR,
+ NULL, "NULL attribute.");
+ return -rte_errno;
+ }
+
+ return 0;
+}
+
+static int hinic_check_ethertype_first_item(const struct rte_flow_item *item,
+ struct rte_flow_error *error)
+{
+ /* The first non-void item should be MAC */
+ if (item->type != RTE_FLOW_ITEM_TYPE_ETH) {
+ rte_flow_error_set(error, EINVAL,
+ RTE_FLOW_ERROR_TYPE_ITEM,
+ item, "Not supported by ethertype filter");
+ return -rte_errno;
+ }
+
+ /* Not supported last point for range */
+ if (item->last) {
+ rte_flow_error_set(error, EINVAL,
+ RTE_FLOW_ERROR_TYPE_UNSPECIFIED,
+ item, "Not supported last point for range");
+ return -rte_errno;
+ }
+
+ /* Get the MAC info. */
+ if (!item->spec || !item->mask) {
+ rte_flow_error_set(error, EINVAL,
+ RTE_FLOW_ERROR_TYPE_ITEM,
+ item, "Not supported by ethertype filter");
+ return -rte_errno;
+ }
+ return 0;
+}
+
+static int
+hinic_parse_ethertype_aciton(const struct rte_flow_action *actions,
+ const struct rte_flow_action *act,
+ const struct rte_flow_action_queue *act_q,
+ struct rte_eth_ethertype_filter *filter,
+ struct rte_flow_error *error)
+{
+ /* Parse action */
+ act = next_no_void_action(actions, NULL);
+ if (act->type != RTE_FLOW_ACTION_TYPE_QUEUE &&
+ act->type != RTE_FLOW_ACTION_TYPE_DROP) {
+ rte_flow_error_set(error, EINVAL,
+ RTE_FLOW_ERROR_TYPE_ACTION,
+ act, "Not supported action.");
+ return -rte_errno;
+ }
+
+ if (act->type == RTE_FLOW_ACTION_TYPE_QUEUE) {
+ act_q = (const struct rte_flow_action_queue *)act->conf;
+ filter->queue = act_q->index;
+ } else {
+ filter->flags |= RTE_ETHTYPE_FLAGS_DROP;
+ }
+
+ /* Check if the next non-void item is END */
+ act = next_no_void_action(actions, act);
+ if (act->type != RTE_FLOW_ACTION_TYPE_END) {
+ rte_flow_error_set(error, EINVAL,
+ RTE_FLOW_ERROR_TYPE_ACTION,
+ act, "Not supported action.");
+ return -rte_errno;
+ }
+
+ return 0;
+}
+
+/**
+ * Parse the rule to see if it is a ethertype rule.
+ * And get the ethertype filter info BTW.
+ * pattern:
+ * The first not void item can be ETH.
+ * The next not void item must be END.
+ * action:
+ * The first not void action should be QUEUE.
+ * The next not void action should be END.
+ * pattern example:
+ * ITEM Spec Mask
+ * ETH type 0x0807 0xFFFF
+ * END
+ * other members in mask and spec should set to 0x00.
+ * item->last should be NULL.
+ */
+static int
+cons_parse_ethertype_filter(const struct rte_flow_attr *attr,
+ const struct rte_flow_item *pattern,
+ const struct rte_flow_action *actions,
+ struct rte_eth_ethertype_filter *filter,
+ struct rte_flow_error *error)
+{
+ const struct rte_flow_item *item;
+ const struct rte_flow_action *act = NULL;
+ const struct rte_flow_item_eth *eth_spec;
+ const struct rte_flow_item_eth *eth_mask;
+ const struct rte_flow_action_queue *act_q = NULL;
+
+ if (hinic_check_filter_arg(attr, pattern, actions, error))
+ return -rte_errno;
+
+ item = next_no_void_pattern(pattern, NULL);
+ if (hinic_check_ethertype_first_item(item, error))
+ return -rte_errno;
+
+ eth_spec = (const struct rte_flow_item_eth *)item->spec;
+ eth_mask = (const struct rte_flow_item_eth *)item->mask;
+
+ /*
+ * Mask bits of source MAC address must be full of 0.
+ * Mask bits of destination MAC address must be full
+ * of 1 or full of 0.
+ */
+ if (!rte_is_zero_ether_addr(ð_mask->src) ||
+ (!rte_is_zero_ether_addr(ð_mask->dst) &&
+ !rte_is_broadcast_ether_addr(ð_mask->dst))) {
+ rte_flow_error_set(error, EINVAL, RTE_FLOW_ERROR_TYPE_ITEM,
+ item, "Invalid ether address mask");
+ return -rte_errno;
+ }
+
+ if ((eth_mask->type & UINT16_MAX) != UINT16_MAX) {
+ rte_flow_error_set(error, EINVAL, RTE_FLOW_ERROR_TYPE_ITEM,
+ item, "Invalid ethertype mask");
+ return -rte_errno;
+ }
+
+ /*
+ * If mask bits of destination MAC address
+ * are full of 1, set RTE_ETHTYPE_FLAGS_MAC.
+ */
+ if (rte_is_broadcast_ether_addr(ð_mask->dst)) {
+ filter->mac_addr = eth_spec->dst;
+ filter->flags |= RTE_ETHTYPE_FLAGS_MAC;
+ } else {
+ filter->flags &= ~RTE_ETHTYPE_FLAGS_MAC;
+ }
+ filter->ether_type = rte_be_to_cpu_16(eth_spec->type);
+
+ /* Check if the next non-void item is END. */
+ item = next_no_void_pattern(pattern, item);
+ if (item->type != RTE_FLOW_ITEM_TYPE_END) {
+ rte_flow_error_set(error, EINVAL, RTE_FLOW_ERROR_TYPE_ITEM,
+ item, "Not supported by ethertype filter.");
+ return -rte_errno;
+ }
+
+ if (hinic_parse_ethertype_aciton(actions, act, act_q, filter, error))
+ return -rte_errno;
+
+ if (hinic_check_ethertype_attr_ele(attr, error))
+ return -rte_errno;
+
+ return 0;
+}
+
+static int
+hinic_parse_ethertype_filter(struct rte_eth_dev *dev,
+ const struct rte_flow_attr *attr,
+ const struct rte_flow_item pattern[],
+ const struct rte_flow_action actions[],
+ struct rte_eth_ethertype_filter *filter,
+ struct rte_flow_error *error)
+{
+ if (cons_parse_ethertype_filter(attr, pattern, actions, filter, error))
+ return -rte_errno;
+
+ /* NIC doesn't support MAC address. */
+ if (filter->flags & RTE_ETHTYPE_FLAGS_MAC) {
+ memset(filter, 0, sizeof(struct rte_eth_ethertype_filter));
+ rte_flow_error_set(error, EINVAL,
+ RTE_FLOW_ERROR_TYPE_ITEM,
+ NULL, "Not supported by ethertype filter");
+ return -rte_errno;
+ }
+
+ if (filter->queue >= dev->data->nb_rx_queues) {
+ memset(filter, 0, sizeof(struct rte_eth_ethertype_filter));
+ rte_flow_error_set(error, EINVAL,
+ RTE_FLOW_ERROR_TYPE_ITEM,
+ NULL, "Queue index much too big");
+ return -rte_errno;
+ }
+
+ if (filter->ether_type == RTE_ETHER_TYPE_IPV4 ||
+ filter->ether_type == RTE_ETHER_TYPE_IPV6) {
+ memset(filter, 0, sizeof(struct rte_eth_ethertype_filter));
+ rte_flow_error_set(error, EINVAL,
+ RTE_FLOW_ERROR_TYPE_ITEM,
+ NULL, "IPv4/IPv6 not supported by ethertype filter");
+ return -rte_errno;
+ }
+
+ if (filter->flags & RTE_ETHTYPE_FLAGS_DROP) {
+ memset(filter, 0, sizeof(struct rte_eth_ethertype_filter));
+ rte_flow_error_set(error, EINVAL,
+ RTE_FLOW_ERROR_TYPE_ITEM,
+ NULL, "Drop option is unsupported");
+ return -rte_errno;
+ }
+
+ /* Hinic only support LACP/ARP for ether type */
+ if (filter->ether_type != RTE_ETHER_TYPE_SLOW &&
+ filter->ether_type != RTE_ETHER_TYPE_ARP) {
+ memset(filter, 0, sizeof(struct rte_eth_ethertype_filter));
+ rte_flow_error_set(error, EINVAL,
+ RTE_FLOW_ERROR_TYPE_ITEM, NULL,
+ "only lacp/arp type supported by ethertype filter");
+ return -rte_errno;
+ }
+
+ return 0;
+}
+
+static int hinic_check_ntuple_attr_ele(const struct rte_flow_attr *attr,
+ struct rte_eth_ntuple_filter *filter,
+ struct rte_flow_error *error)
+{
+ /* Must be input direction */
+ if (!attr->ingress) {
+ memset(filter, 0, sizeof(struct rte_eth_ntuple_filter));
+ rte_flow_error_set(error, EINVAL,
+ RTE_FLOW_ERROR_TYPE_ATTR_INGRESS,
+ attr, "Only support ingress.");
+ return -rte_errno;
+ }
+
+ if (attr->egress) {
+ memset(filter, 0, sizeof(struct rte_eth_ntuple_filter));
+ rte_flow_error_set(error, EINVAL,
+ RTE_FLOW_ERROR_TYPE_ATTR_EGRESS,
+ attr, "Not support egress.");
+ return -rte_errno;
+ }
+
+ if (attr->priority > 0xFFFF) {
+ memset(filter, 0, sizeof(struct rte_eth_ntuple_filter));
+ rte_flow_error_set(error, EINVAL,
+ RTE_FLOW_ERROR_TYPE_ATTR_PRIORITY,
+ attr, "Error priority.");
+ return -rte_errno;
+ }
+
+ if (attr->priority < HINIC_MIN_N_TUPLE_PRIO ||
+ attr->priority > HINIC_MAX_N_TUPLE_PRIO)
+ filter->priority = 1;
+ else
+ filter->priority = (uint16_t)attr->priority;
+
+ return 0;
+}
+
+static int
+hinic_check_ntuple_act_ele(__rte_unused const struct rte_flow_item *item,
+ const struct rte_flow_action actions[],
+ struct rte_eth_ntuple_filter *filter,
+ struct rte_flow_error *error)
+{
+ const struct rte_flow_action *act;
+ /*
+ * n-tuple only supports forwarding,
+ * check if the first not void action is QUEUE.
+ */
+ act = next_no_void_action(actions, NULL);
+ if (act->type != RTE_FLOW_ACTION_TYPE_QUEUE) {
+ memset(filter, 0, sizeof(struct rte_eth_ntuple_filter));
+ rte_flow_error_set(error, EINVAL,
+ RTE_FLOW_ERROR_TYPE_ACTION,
+ act, "Flow action type is not QUEUE.");
+ return -rte_errno;
+ }
+ filter->queue =
+ ((const struct rte_flow_action_queue *)act->conf)->index;
+
+ /* Check if the next not void item is END */
+ act = next_no_void_action(actions, act);
+ if (act->type != RTE_FLOW_ACTION_TYPE_END) {
+ memset(filter, 0, sizeof(struct rte_eth_ntuple_filter));
+ rte_flow_error_set(error, EINVAL,
+ RTE_FLOW_ERROR_TYPE_ACTION,
+ act, "Next not void item is not END.");
+ return -rte_errno;
+ }
+
+ return 0;
+}
+
+static int hinic_ntuple_item_check_ether(const struct rte_flow_item **ipv4_item,
+ const struct rte_flow_item pattern[],
+ struct rte_flow_error *error)
+{
+ const struct rte_flow_item *item;
+
+ /* The first not void item can be MAC or IPv4 */
+ item = next_no_void_pattern(pattern, NULL);
+
+ if (item->type != RTE_FLOW_ITEM_TYPE_ETH &&
+ item->type != RTE_FLOW_ITEM_TYPE_IPV4) {
+ rte_flow_error_set(error, EINVAL,
+ RTE_FLOW_ERROR_TYPE_ITEM,
+ item, "Not supported by ntuple filter");
+ return -rte_errno;
+ }
+
+ /* Skip Ethernet */
+ if (item->type == RTE_FLOW_ITEM_TYPE_ETH) {
+ /* Not supported last point for range */
+ if (item->last) {
+ rte_flow_error_set(error,
+ EINVAL,
+ RTE_FLOW_ERROR_TYPE_UNSPECIFIED,
+ item, "Not supported last point for range");
+ return -rte_errno;
+ }
+ /* if the first item is MAC, the content should be NULL */
+ if (item->spec || item->mask) {
+ rte_flow_error_set(error, EINVAL,
+ RTE_FLOW_ERROR_TYPE_ITEM,
+ item, "Not supported by ntuple filter");
+ return -rte_errno;
+ }
+ /* check if the next not void item is IPv4 */
+ item = next_no_void_pattern(pattern, item);
+ if (item->type != RTE_FLOW_ITEM_TYPE_IPV4) {
+ rte_flow_error_set(error,
+ EINVAL, RTE_FLOW_ERROR_TYPE_ITEM,
+ item, "Not supported by ntuple filter");
+ return -rte_errno;
+ }
+ }
+
+ *ipv4_item = item;
+ return 0;
+}
+
+static int
+hinic_ntuple_item_check_ipv4(const struct rte_flow_item **in_out_item,
+ const struct rte_flow_item pattern[],
+ struct rte_eth_ntuple_filter *filter,
+ struct rte_flow_error *error)
+{
+ const struct rte_flow_item_ipv4 *ipv4_spec;
+ const struct rte_flow_item_ipv4 *ipv4_mask;
+ const struct rte_flow_item *item = *in_out_item;
+
+ /* Get the IPv4 info */
+ if (!item->spec || !item->mask) {
+ rte_flow_error_set(error, EINVAL,
+ RTE_FLOW_ERROR_TYPE_ITEM,
+ item, "Invalid ntuple mask");
+ return -rte_errno;
+ }
+ /* Not supported last point for range */
+ if (item->last) {
+ rte_flow_error_set(error, EINVAL,
+ RTE_FLOW_ERROR_TYPE_UNSPECIFIED,
+ item, "Not supported last point for range");
+ return -rte_errno;
+ }
+
+ ipv4_mask = (const struct rte_flow_item_ipv4 *)item->mask;
+ /*
+ * Only support src & dst addresses, protocol,
+ * others should be masked.
+ */
+ if (ipv4_mask->hdr.version_ihl ||
+ ipv4_mask->hdr.type_of_service ||
+ ipv4_mask->hdr.total_length ||
+ ipv4_mask->hdr.packet_id ||
+ ipv4_mask->hdr.fragment_offset ||
+ ipv4_mask->hdr.time_to_live ||
+ ipv4_mask->hdr.hdr_checksum ||
+ !ipv4_mask->hdr.next_proto_id) {
+ rte_flow_error_set(error,
+ EINVAL, RTE_FLOW_ERROR_TYPE_ITEM,
+ item, "Not supported by ntuple filter");
+ return -rte_errno;
+ }
+
+ filter->dst_ip_mask = ipv4_mask->hdr.dst_addr;
+ filter->src_ip_mask = ipv4_mask->hdr.src_addr;
+ filter->proto_mask = ipv4_mask->hdr.next_proto_id;
+
+ ipv4_spec = (const struct rte_flow_item_ipv4 *)item->spec;
+ filter->dst_ip = ipv4_spec->hdr.dst_addr;
+ filter->src_ip = ipv4_spec->hdr.src_addr;
+ filter->proto = ipv4_spec->hdr.next_proto_id;
+
+ /* Get next no void item */
+ *in_out_item = next_no_void_pattern(pattern, item);
+ return 0;
+}
+
+static int hinic_ntuple_item_check_l4(const struct rte_flow_item **in_out_item,
+ const struct rte_flow_item pattern[],
+ struct rte_eth_ntuple_filter *filter,
+ struct rte_flow_error *error)
+{
+ const struct rte_flow_item_tcp *tcp_spec;
+ const struct rte_flow_item_tcp *tcp_mask;
+ const struct rte_flow_item_icmp *icmp_mask;
+ const struct rte_flow_item *item = *in_out_item;
+ u32 ntuple_filter_size = sizeof(struct rte_eth_ntuple_filter);
+
+ if (item->type == RTE_FLOW_ITEM_TYPE_END)
+ return 0;
+
+ /* Get TCP or UDP info */
+ if (item->type != RTE_FLOW_ITEM_TYPE_END &&
+ (!item->spec || !item->mask)) {
+ memset(filter, 0, ntuple_filter_size);
+ rte_flow_error_set(error, EINVAL,
+ RTE_FLOW_ERROR_TYPE_ITEM,
+ item, "Invalid ntuple mask");
+ return -rte_errno;
+ }
+
+ /* Not supported last point for range */
+ if (item->last) {
+ memset(filter, 0, ntuple_filter_size);
+ rte_flow_error_set(error, EINVAL,
+ RTE_FLOW_ERROR_TYPE_UNSPECIFIED,
+ item, "Not supported last point for range");
+ return -rte_errno;
+ }
+
+ if (item->type == RTE_FLOW_ITEM_TYPE_TCP) {
+ tcp_mask = (const struct rte_flow_item_tcp *)item->mask;
+
+ /*
+ * Only support src & dst ports, tcp flags,
+ * others should be masked.
+ */
+ if (tcp_mask->hdr.sent_seq ||
+ tcp_mask->hdr.recv_ack ||
+ tcp_mask->hdr.data_off ||
+ tcp_mask->hdr.rx_win ||
+ tcp_mask->hdr.cksum ||
+ tcp_mask->hdr.tcp_urp) {
+ memset(filter, 0, ntuple_filter_size);
+ rte_flow_error_set(error, EINVAL,
+ RTE_FLOW_ERROR_TYPE_ITEM,
+ item, "Not supported by ntuple filter");
+ return -rte_errno;
+ }
+
+ filter->dst_port_mask = tcp_mask->hdr.dst_port;
+ filter->src_port_mask = tcp_mask->hdr.src_port;
+ if (tcp_mask->hdr.tcp_flags == 0xFF) {
+ filter->flags |= RTE_NTUPLE_FLAGS_TCP_FLAG;
+ } else if (!tcp_mask->hdr.tcp_flags) {
+ filter->flags &= ~RTE_NTUPLE_FLAGS_TCP_FLAG;
+ } else {
+ memset(filter, 0, ntuple_filter_size);
+ rte_flow_error_set(error, EINVAL,
+ RTE_FLOW_ERROR_TYPE_ITEM,
+ item, "Not supported by ntuple filter");
+ return -rte_errno;
+ }
+
+ tcp_spec = (const struct rte_flow_item_tcp *)item->spec;
+ filter->dst_port = tcp_spec->hdr.dst_port;
+ filter->src_port = tcp_spec->hdr.src_port;
+ filter->tcp_flags = tcp_spec->hdr.tcp_flags;
+ } else if (item->type == RTE_FLOW_ITEM_TYPE_ICMP) {
+ icmp_mask = (const struct rte_flow_item_icmp *)item->mask;
+
+ /* ICMP all should be masked. */
+ if (icmp_mask->hdr.icmp_cksum ||
+ icmp_mask->hdr.icmp_ident ||
+ icmp_mask->hdr.icmp_seq_nb ||
+ icmp_mask->hdr.icmp_type ||
+ icmp_mask->hdr.icmp_code) {
+ memset(filter, 0, ntuple_filter_size);
+ rte_flow_error_set(error, EINVAL,
+ RTE_FLOW_ERROR_TYPE_ITEM,
+ item, "Not supported by ntuple filter");
+ return -rte_errno;
+ }
+ }
+
+ /* Get next no void item */
+ *in_out_item = next_no_void_pattern(pattern, item);
+ return 0;
+}
+
+static int hinic_ntuple_item_check_end(const struct rte_flow_item *item,
+ struct rte_eth_ntuple_filter *filter,
+ struct rte_flow_error *error)
+{
+ /* Check if the next not void item is END */
+ if (item->type != RTE_FLOW_ITEM_TYPE_END) {
+ memset(filter, 0, sizeof(struct rte_eth_ntuple_filter));
+ rte_flow_error_set(error, EINVAL,
+ RTE_FLOW_ERROR_TYPE_ITEM,
+ item, "Not supported by ntuple filter");
+ return -rte_errno;
+ }
+ return 0;
+}
+
+static int hinic_check_ntuple_item_ele(const struct rte_flow_item *item,
+ const struct rte_flow_item pattern[],
+ struct rte_eth_ntuple_filter *filter,
+ struct rte_flow_error *error)
+{
+ if (hinic_ntuple_item_check_ether(&item, pattern, error) ||
+ hinic_ntuple_item_check_ipv4(&item, pattern, filter, error) ||
+ hinic_ntuple_item_check_l4(&item, pattern, filter, error) ||
+ hinic_ntuple_item_check_end(item, filter, error))
+ return -rte_errno;
+
+ return 0;
+}
+
+/**
+ * Parse the rule to see if it is a n-tuple rule.
+ * And get the n-tuple filter info BTW.
+ * pattern:
+ * The first not void item can be ETH or IPV4.
+ * The second not void item must be IPV4 if the first one is ETH.
+ * The third not void item must be UDP or TCP.
+ * The next not void item must be END.
+ * action:
+ * The first not void action should be QUEUE.
+ * The next not void action should be END.
+ * pattern example:
+ * ITEM Spec Mask
+ * ETH NULL NULL
+ * IPV4 src_addr 192.168.1.20 0xFFFFFFFF
+ * dst_addr 192.167.3.50 0xFFFFFFFF
+ * next_proto_id 17 0xFF
+ * UDP/TCP/ src_port 80 0xFFFF
+ * SCTP dst_port 80 0xFFFF
+ * END
+ * other members in mask and spec should set to 0x00.
+ * item->last should be NULL.
+ * Please aware there's an asumption for all the parsers.
+ * rte_flow_item is using big endian, rte_flow_attr and
+ * rte_flow_action are using CPU order.
+ * Because the pattern is used to describe the packets,
+ * normally the packets should use network order.
+ */
+static int
+cons_parse_ntuple_filter(const struct rte_flow_attr *attr,
+ const struct rte_flow_item pattern[],
+ const struct rte_flow_action actions[],
+ struct rte_eth_ntuple_filter *filter,
+ struct rte_flow_error *error)
+{
+ const struct rte_flow_item *item = NULL;
+
+ if (hinic_check_filter_arg(attr, pattern, actions, error))
+ return -rte_errno;
+
+ if (hinic_check_ntuple_item_ele(item, pattern, filter, error))
+ return -rte_errno;
+
+ if (hinic_check_ntuple_act_ele(item, actions, filter, error))
+ return -rte_errno;
+
+ if (hinic_check_ntuple_attr_ele(attr, filter, error))
+ return -rte_errno;
+
+ return 0;
+}
+
+static int
+hinic_parse_ntuple_filter(struct rte_eth_dev *dev,
+ const struct rte_flow_attr *attr,
+ const struct rte_flow_item pattern[],
+ const struct rte_flow_action actions[],
+ struct rte_eth_ntuple_filter *filter,
+ struct rte_flow_error *error)
+{
+ int ret;
+
+ ret = cons_parse_ntuple_filter(attr, pattern, actions, filter, error);
+ if (ret)
+ return ret;
+
+ /* Hinic doesn't support tcp flags */
+ if (filter->flags & RTE_NTUPLE_FLAGS_TCP_FLAG) {
+ memset(filter, 0, sizeof(struct rte_eth_ntuple_filter));
+ rte_flow_error_set(error, EINVAL,
+ RTE_FLOW_ERROR_TYPE_ITEM,
+ NULL, "Not supported by ntuple filter");
+ return -rte_errno;
+ }
+
+ /* Hinic doesn't support many priorities */
+ if (filter->priority < HINIC_MIN_N_TUPLE_PRIO ||
+ filter->priority > HINIC_MAX_N_TUPLE_PRIO) {
+ memset(filter, 0, sizeof(struct rte_eth_ntuple_filter));
+ rte_flow_error_set(error, EINVAL,
+ RTE_FLOW_ERROR_TYPE_ITEM,
+ NULL, "Priority not supported by ntuple filter");
+ return -rte_errno;
+ }
+
+ if (filter->queue >= dev->data->nb_rx_queues)
+ return -rte_errno;
+
+ /* Fixed value for hinic */
+ filter->flags = RTE_5TUPLE_FLAGS;
+ return 0;
+}
+
+static int hinic_normal_item_check_ether(const struct rte_flow_item **ip_item,
+ const struct rte_flow_item pattern[],
+ struct rte_flow_error *error)
+{
+ const struct rte_flow_item *item;
+
+ /* The first not void item can be MAC or IPv4 or TCP or UDP */
+ item = next_no_void_pattern(pattern, NULL);
+
+ if (item->type != RTE_FLOW_ITEM_TYPE_ETH &&
+ item->type != RTE_FLOW_ITEM_TYPE_IPV4 &&
+ item->type != RTE_FLOW_ITEM_TYPE_TCP &&
+ item->type != RTE_FLOW_ITEM_TYPE_UDP) {
+ rte_flow_error_set(error, EINVAL,
+ RTE_FLOW_ERROR_TYPE_ITEM, item,
+ "Not supported by fdir filter,support mac,ipv4,tcp,udp");
+ return -rte_errno;
+ }
+
+ /* Not supported last point for range */
+ if (item->last) {
+ rte_flow_error_set(error, EINVAL,
+ RTE_FLOW_ERROR_TYPE_UNSPECIFIED, item,
+ "Not supported last point for range");
+ return -rte_errno;
+ }
+
+ /* Skip Ethernet */
+ if (item->type == RTE_FLOW_ITEM_TYPE_ETH) {
+ /* All should be masked. */
+ if (item->spec || item->mask) {
+ rte_flow_error_set(error, EINVAL,
+ RTE_FLOW_ERROR_TYPE_ITEM,
+ item, "Not supported by fdir filter,support mac");
+ return -rte_errno;
+ }
+ /* Check if the next not void item is IPv4 */
+ item = next_no_void_pattern(pattern, item);
+ if (item->type != RTE_FLOW_ITEM_TYPE_IPV4) {
+ rte_flow_error_set(error, EINVAL,
+ RTE_FLOW_ERROR_TYPE_ITEM, item,
+ "Not supported by fdir filter,support mac,ipv4");
+ return -rte_errno;
+ }
+ }
+
+ *ip_item = item;
+ return 0;
+}
+
+static int hinic_normal_item_check_ip(const struct rte_flow_item **in_out_item,
+ const struct rte_flow_item pattern[],
+ struct hinic_fdir_rule *rule,
+ struct rte_flow_error *error)
+{
+ const struct rte_flow_item_ipv4 *ipv4_spec;
+ const struct rte_flow_item_ipv4 *ipv4_mask;
+ const struct rte_flow_item *item = *in_out_item;
+
+ /* Get the IPv4 info */
+ if (item->type == RTE_FLOW_ITEM_TYPE_IPV4) {
+ /* Not supported last point for range */
+ if (item->last) {
+ rte_flow_error_set(error, EINVAL,
+ RTE_FLOW_ERROR_TYPE_UNSPECIFIED,
+ item, "Not supported last point for range");
+ return -rte_errno;
+ }
+
+ if (!item->mask) {
+ memset(rule, 0, sizeof(struct hinic_fdir_rule));
+ rte_flow_error_set(error, EINVAL,
+ RTE_FLOW_ERROR_TYPE_ITEM,
+ item, "Invalid fdir filter mask");
+ return -rte_errno;
+ }
+
+ ipv4_mask = (const struct rte_flow_item_ipv4 *)item->mask;
+ /*
+ * Only support src & dst addresses,
+ * others should be masked.
+ */
+ if (ipv4_mask->hdr.version_ihl ||
+ ipv4_mask->hdr.type_of_service ||
+ ipv4_mask->hdr.total_length ||
+ ipv4_mask->hdr.packet_id ||
+ ipv4_mask->hdr.fragment_offset ||
+ ipv4_mask->hdr.time_to_live ||
+ ipv4_mask->hdr.next_proto_id ||
+ ipv4_mask->hdr.hdr_checksum) {
+ rte_flow_error_set(error,
+ EINVAL, RTE_FLOW_ERROR_TYPE_ITEM, item,
+ "Not supported by fdir filter, support src,dst ip");
+ return -rte_errno;
+ }
+
+ rule->mask.dst_ipv4_mask = ipv4_mask->hdr.dst_addr;
+ rule->mask.src_ipv4_mask = ipv4_mask->hdr.src_addr;
+
+ if (item->spec) {
+ ipv4_spec =
+ (const struct rte_flow_item_ipv4 *)item->spec;
+ rule->hinic_fdir.dst_ip = ipv4_spec->hdr.dst_addr;
+ rule->hinic_fdir.src_ip = ipv4_spec->hdr.src_addr;
+ }
+
+ /*
+ * Check if the next not void item is
+ * TCP or UDP or END.
+ */
+ item = next_no_void_pattern(pattern, item);
+ if (item->type != RTE_FLOW_ITEM_TYPE_TCP &&
+ item->type != RTE_FLOW_ITEM_TYPE_UDP &&
+ item->type != RTE_FLOW_ITEM_TYPE_END) {
+ memset(rule, 0, sizeof(struct hinic_fdir_rule));
+ rte_flow_error_set(error, EINVAL,
+ RTE_FLOW_ERROR_TYPE_ITEM, item,
+ "Not supported by fdir filter, support tcp, udp, end");
+ return -rte_errno;
+ }
+ }
+
+ *in_out_item = item;
+ return 0;
+}
+
+static int hinic_normal_item_check_l4(const struct rte_flow_item **in_out_item,
+ const struct rte_flow_item pattern[],
+ struct hinic_fdir_rule *rule,
+ struct rte_flow_error *error)
+{
+ const struct rte_flow_item_tcp *tcp_spec;
+ const struct rte_flow_item_tcp *tcp_mask;
+ const struct rte_flow_item_udp *udp_spec;
+ const struct rte_flow_item_udp *udp_mask;
+ const struct rte_flow_item *item = *in_out_item;
+
+ if (item->type != RTE_FLOW_ITEM_TYPE_END) {
+ /* Not supported last point for range */
+ if (item->last) {
+ rte_flow_error_set(error, EINVAL,
+ RTE_FLOW_ERROR_TYPE_UNSPECIFIED,
+ item, "Not supported last point for range");
+ return -rte_errno;
+ }
+
+ /* Get TCP/UDP info */
+ if (item->type == RTE_FLOW_ITEM_TYPE_TCP) {
+ /*
+ * Only care about src & dst ports,
+ * others should be masked.
+ */
+ if (!item->mask) {
+ memset(rule, 0, sizeof(struct hinic_fdir_rule));
+ rte_flow_error_set(error, EINVAL,
+ RTE_FLOW_ERROR_TYPE_ITEM, item,
+ "Not supported by fdir filter,support src,dst ports");
+ return -rte_errno;
+ }
+
+ tcp_mask = (const struct rte_flow_item_tcp *)item->mask;
+ if (tcp_mask->hdr.sent_seq ||
+ tcp_mask->hdr.recv_ack ||
+ tcp_mask->hdr.data_off ||
+ tcp_mask->hdr.tcp_flags ||
+ tcp_mask->hdr.rx_win ||
+ tcp_mask->hdr.cksum ||
+ tcp_mask->hdr.tcp_urp) {
+ memset(rule, 0, sizeof(struct hinic_fdir_rule));
+ rte_flow_error_set(error, EINVAL,
+ RTE_FLOW_ERROR_TYPE_ITEM,
+ item, "Not supported by fdir filter,support tcp");
+ return -rte_errno;
+ }
+
+ rule->mask.src_port_mask = tcp_mask->hdr.src_port;
+ rule->mask.dst_port_mask = tcp_mask->hdr.dst_port;
+
+ if (item->spec) {
+ tcp_spec =
+ (const struct rte_flow_item_tcp *)
+ item->spec;
+ rule->hinic_fdir.src_port =
+ tcp_spec->hdr.src_port;
+ rule->hinic_fdir.dst_port =
+ tcp_spec->hdr.dst_port;
+ }
+
+ } else if (item->type == RTE_FLOW_ITEM_TYPE_UDP) {
+ /*
+ * Only care about src & dst ports,
+ * others should be masked.
+ */
+ if (!item->mask) {
+ memset(rule, 0, sizeof(struct hinic_fdir_rule));
+ rte_flow_error_set(error, EINVAL,
+ RTE_FLOW_ERROR_TYPE_ITEM,
+ item, "Not supported by fdir filter,support src,dst ports");
+ return -rte_errno;
+ }
+
+ udp_mask = (const struct rte_flow_item_udp *)item->mask;
+ if (udp_mask->hdr.dgram_len ||
+ udp_mask->hdr.dgram_cksum) {
+ memset(rule, 0, sizeof(struct hinic_fdir_rule));
+ rte_flow_error_set(error, EINVAL,
+ RTE_FLOW_ERROR_TYPE_ITEM,
+ item, "Not supported by fdir filter,support udp");
+ return -rte_errno;
+ }
+ rule->mask.src_port_mask = udp_mask->hdr.src_port;
+ rule->mask.dst_port_mask = udp_mask->hdr.dst_port;
+
+ if (item->spec) {
+ udp_spec =
+ (const struct rte_flow_item_udp *)
+ item->spec;
+ rule->hinic_fdir.src_port =
+ udp_spec->hdr.src_port;
+ rule->hinic_fdir.dst_port =
+ udp_spec->hdr.dst_port;
+ }
+ } else {
+ memset(rule, 0, sizeof(struct hinic_fdir_rule));
+ rte_flow_error_set(error, EINVAL,
+ RTE_FLOW_ERROR_TYPE_ITEM,
+ item, "Not supported by fdir filter,support tcp/udp");
+ return -rte_errno;
+ }
+
+ /* Get next no void item */
+ *in_out_item = next_no_void_pattern(pattern, item);
+ }
+
+ return 0;
+}
+
+static int hinic_normal_item_check_end(const struct rte_flow_item *item,
+ struct hinic_fdir_rule *rule,
+ struct rte_flow_error *error)
+{
+ /* Check if the next not void item is END */
+ if (item->type != RTE_FLOW_ITEM_TYPE_END) {
+ memset(rule, 0, sizeof(struct hinic_fdir_rule));
+ rte_flow_error_set(error, EINVAL,
+ RTE_FLOW_ERROR_TYPE_ITEM,
+ item, "Not supported by fdir filter,support end");
+ return -rte_errno;
+ }
+
+ return 0;
+}
+
+static int hinic_check_normal_item_ele(const struct rte_flow_item *item,
+ const struct rte_flow_item pattern[],
+ struct hinic_fdir_rule *rule,
+ struct rte_flow_error *error)
+{
+ if (hinic_normal_item_check_ether(&item, pattern, error) ||
+ hinic_normal_item_check_ip(&item, pattern, rule, error) ||
+ hinic_normal_item_check_l4(&item, pattern, rule, error) ||
+ hinic_normal_item_check_end(item, rule, error))
+ return -rte_errno;
+
+ return 0;
+}
+
+static int hinic_check_normal_attr_ele(const struct rte_flow_attr *attr,
+ struct hinic_fdir_rule *rule,
+ struct rte_flow_error *error)
+{
+ /* Must be input direction */
+ if (!attr->ingress) {
+ memset(rule, 0, sizeof(struct hinic_fdir_rule));
+ rte_flow_error_set(error, EINVAL,
+ RTE_FLOW_ERROR_TYPE_ATTR_INGRESS,
+ attr, "Only support ingress.");
+ return -rte_errno;
+ }
+
+ /* Not supported */
+ if (attr->egress) {
+ memset(rule, 0, sizeof(struct hinic_fdir_rule));
+ rte_flow_error_set(error, EINVAL,
+ RTE_FLOW_ERROR_TYPE_ATTR_EGRESS,
+ attr, "Not support egress.");
+ return -rte_errno;
+ }
+
+ /* Not supported */
+ if (attr->priority) {
+ memset(rule, 0, sizeof(struct hinic_fdir_rule));
+ rte_flow_error_set(error, EINVAL,
+ RTE_FLOW_ERROR_TYPE_ATTR_PRIORITY,
+ attr, "Not support priority.");
+ return -rte_errno;
+ }
+
+ return 0;
+}
+
+static int hinic_check_normal_act_ele(const struct rte_flow_item *item,
+ const struct rte_flow_action actions[],
+ struct hinic_fdir_rule *rule,
+ struct rte_flow_error *error)
+{
+ const struct rte_flow_action *act;
+
+ /* Check if the first not void action is QUEUE */
+ act = next_no_void_action(actions, NULL);
+ if (act->type != RTE_FLOW_ACTION_TYPE_QUEUE) {
+ memset(rule, 0, sizeof(struct hinic_fdir_rule));
+ rte_flow_error_set(error, EINVAL, RTE_FLOW_ERROR_TYPE_ACTION,
+ item, "Not supported action.");
+ return -rte_errno;
+ }
+
+ rule->queue = ((const struct rte_flow_action_queue *)act->conf)->index;
+
+ /* Check if the next not void item is END */
+ act = next_no_void_action(actions, act);
+ if (act->type != RTE_FLOW_ACTION_TYPE_END) {
+ memset(rule, 0, sizeof(struct hinic_fdir_rule));
+ rte_flow_error_set(error, EINVAL, RTE_FLOW_ERROR_TYPE_ACTION,
+ act, "Not supported action.");
+ return -rte_errno;
+ }
+
+ return 0;
+}
+
+/**
+ * Parse the rule to see if it is a IP or MAC VLAN flow director rule.
+ * And get the flow director filter info BTW.
+ * UDP/TCP/SCTP PATTERN:
+ * The first not void item can be ETH or IPV4 or IPV6
+ * The second not void item must be IPV4 or IPV6 if the first one is ETH.
+ * The next not void item could be UDP or TCP(optional)
+ * The next not void item must be END.
+ * ACTION:
+ * The first not void action should be QUEUE.
+ * The second not void optional action should be MARK,
+ * mark_id is a uint32_t number.
+ * The next not void action should be END.
+ * UDP/TCP pattern example:
+ * ITEM Spec Mask
+ * ETH NULL NULL
+ * IPV4 src_addr 1.2.3.6 0xFFFFFFFF
+ * dst_addr 1.2.3.5 0xFFFFFFFF
+ * UDP/TCP src_port 80 0xFFFF
+ * dst_port 80 0xFFFF
+ * END
+ * Other members in mask and spec should set to 0x00.
+ * Item->last should be NULL.
+ */
+static int
+hinic_parse_fdir_filter_normal(const struct rte_flow_attr *attr,
+ const struct rte_flow_item pattern[],
+ const struct rte_flow_action actions[],
+ struct hinic_fdir_rule *rule,
+ struct rte_flow_error *error)
+{
+ const struct rte_flow_item *item = NULL;
+
+ if (hinic_check_filter_arg(attr, pattern, actions, error))
+ return -rte_errno;
+
+ if (hinic_check_normal_item_ele(item, pattern, rule, error))
+ return -rte_errno;
+
+ if (hinic_check_normal_attr_ele(attr, rule, error))
+ return -rte_errno;
+
+ if (hinic_check_normal_act_ele(item, actions, rule, error))
+ return -rte_errno;
+
+ return 0;
+}
+
+static int
+hinic_parse_fdir_filter(struct rte_eth_dev *dev,
+ const struct rte_flow_attr *attr,
+ const struct rte_flow_item pattern[],
+ const struct rte_flow_action actions[],
+ struct hinic_fdir_rule *rule,
+ struct rte_flow_error *error)
+{
+ int ret;
+
+ ret = hinic_parse_fdir_filter_normal(attr, pattern,
+ actions, rule, error);
+ if (ret)
+ return ret;
+
+ if (rule->queue >= dev->data->nb_rx_queues)
+ return -ENOTSUP;
+
+ return ret;
+}
+
+/**
+ * Check if the flow rule is supported by nic.
+ * It only checkes the format. Don't guarantee the rule can be programmed into
+ * the HW. Because there can be no enough room for the rule.
+ */
+static int hinic_flow_validate(struct rte_eth_dev *dev,
+ const struct rte_flow_attr *attr,
+ const struct rte_flow_item pattern[],
+ const struct rte_flow_action actions[],
+ struct rte_flow_error *error)
+{
+ struct rte_eth_ethertype_filter ethertype_filter;
+ struct rte_eth_ntuple_filter ntuple_filter;
+ struct hinic_fdir_rule fdir_rule;
+ int ret;
+
+ memset(&ntuple_filter, 0, sizeof(struct rte_eth_ntuple_filter));
+ ret = hinic_parse_ntuple_filter(dev, attr, pattern,
+ actions, &ntuple_filter, error);
+ if (!ret)
+ return 0;
+
+ memset(ðertype_filter, 0, sizeof(struct rte_eth_ethertype_filter));
+ ret = hinic_parse_ethertype_filter(dev, attr, pattern,
+ actions, ðertype_filter, error);
+
+ if (!ret)
+ return 0;
+
+ memset(&fdir_rule, 0, sizeof(struct hinic_fdir_rule));
+ ret = hinic_parse_fdir_filter(dev, attr, pattern,
+ actions, &fdir_rule, error);
+
+ return ret;
+}
+
+const struct rte_flow_ops hinic_flow_ops = {
+ .validate = hinic_flow_validate,
+};
diff --git a/drivers/net/hinic/meson.build b/drivers/net/hinic/meson.build
index 87c8d16..bc7e246 100644
--- a/drivers/net/hinic/meson.build
+++ b/drivers/net/hinic/meson.build
@@ -8,6 +8,7 @@ sources = files(
'hinic_pmd_ethdev.c',
'hinic_pmd_rx.c',
'hinic_pmd_tx.c',
+ 'hinic_pmd_flow.c',
)
includes += include_directories('base')
--
1.8.3.1
^ permalink raw reply [flat|nested] 32+ messages in thread
* [dpdk-dev] [PATCH v2 09/17] net/hinic: create and destroy ntuple filter
2019-09-25 14:26 [dpdk-dev] [PATCH v2 00/17] Add advanced features for Huawei hinic pmd Xiaoyun wang
` (8 preceding siblings ...)
2019-09-25 14:30 ` [dpdk-dev] [PATCH v2 08/17] net/hinic: add fdir validate flow operations Xiaoyun wang
@ 2019-09-25 14:30 ` Xiaoyun wang
2019-09-25 14:30 ` [dpdk-dev] [PATCH v2 10/17] net/hinic: create and destroy fdir filter Xiaoyun wang
` (8 subsequent siblings)
18 siblings, 0 replies; 32+ messages in thread
From: Xiaoyun wang @ 2019-09-25 14:30 UTC (permalink / raw)
To: ferruh.yigit
Cc: dev, xuanziyang2, shahar.belkar, luoxianjun, tanya.brokhman,
zhouguoyang, wulike1, Xiaoyun wang
Add support for creating/destroying ntuple filter.
The filter rule includes BGP and VRRP packet type.
Signed-off-by: Xiaoyun wang <cloud.wangxiaoyun@huawei.com>
---
drivers/net/hinic/hinic_pmd_ethdev.h | 75 ++++
drivers/net/hinic/hinic_pmd_flow.c | 684 +++++++++++++++++++++++++++++++++++
2 files changed, 759 insertions(+)
diff --git a/drivers/net/hinic/hinic_pmd_ethdev.h b/drivers/net/hinic/hinic_pmd_ethdev.h
index 58a38d9..93cd6a7 100644
--- a/drivers/net/hinic/hinic_pmd_ethdev.h
+++ b/drivers/net/hinic/hinic_pmd_ethdev.h
@@ -38,6 +38,57 @@ enum hinic_dev_status {
HINIC_DEV_INTR_EN,
};
+#define HINIC_MAX_Q_FILTERS 64 /* hinic just support 64 filter types */
+#define HINIC_PKT_TYPE_FIND_ID(pkt_type) ((pkt_type) - HINIC_MAX_Q_FILTERS)
+
+/* 5tuple filter info */
+struct hinic_5tuple_filter_info {
+ uint32_t dst_ip;
+ uint32_t src_ip;
+ uint16_t dst_port;
+ uint16_t src_port;
+ uint8_t proto; /* l4 protocol. */
+ /*
+ * seven levels (001b-111b), 111b is highest,
+ * used when more than one filter matches.
+ */
+ uint8_t priority;
+ uint8_t dst_ip_mask:1, /* if mask is 1b, do not compare dst ip. */
+ src_ip_mask:1, /* if mask is 1b, do not compare src ip. */
+ dst_port_mask:1, /* if mask is 1b, do not compare dst port. */
+ src_port_mask:1, /* if mask is 1b, do not compare src port. */
+ proto_mask:1; /* if mask is 1b, do not compare protocol. */
+};
+
+/* 5tuple filter structure */
+struct hinic_5tuple_filter {
+ TAILQ_ENTRY(hinic_5tuple_filter) entries;
+ uint16_t index; /* the index of 5tuple filter */
+ struct hinic_5tuple_filter_info filter_info;
+ uint16_t queue; /* rx queue assigned to */
+};
+
+TAILQ_HEAD(hinic_5tuple_filter_list, hinic_5tuple_filter);
+
+/*
+ * If this filter is added by configuration,
+ * it should not be removed.
+ */
+struct hinic_pkt_filter {
+ uint16_t pkt_proto;
+ uint8_t qid;
+ bool enable;
+};
+
+/* Structure to store filters' info. */
+struct hinic_filter_info {
+ uint8_t pkt_type;
+ uint8_t qid;
+ uint64_t type_mask; /* Bit mask for every used filter */
+ struct hinic_5tuple_filter_list fivetuple_list;
+ struct hinic_pkt_filter pkt_filters[HINIC_MAX_Q_FILTERS];
+};
+
/* Information about the fdir mode. */
struct hinic_hw_fdir_mask {
uint32_t src_ipv4_mask;
@@ -60,6 +111,26 @@ struct hinic_fdir_rule {
uint8_t queue; /* queue assigned when matched */
};
+/* ntuple filter list structure */
+struct hinic_ntuple_filter_ele {
+ TAILQ_ENTRY(hinic_ntuple_filter_ele) entries;
+ struct rte_eth_ntuple_filter filter_info;
+};
+
+struct rte_flow {
+ enum rte_filter_type filter_type;
+ void *rule;
+};
+
+/* hinic_flow memory list structure */
+struct hinic_flow_mem {
+ TAILQ_ENTRY(hinic_flow_mem) entries;
+ struct rte_flow *flow;
+};
+
+TAILQ_HEAD(hinic_ntuple_filter_list, hinic_ntuple_filter_ele);
+TAILQ_HEAD(hinic_flow_mem_list, hinic_flow_mem);
+
extern const struct rte_flow_ops hinic_flow_ops;
/* hinic nic_device */
@@ -94,6 +165,10 @@ struct hinic_nic_dev {
* vf: the same with associate pf
*/
u32 default_cos;
+
+ struct hinic_filter_info filter;
+ struct hinic_ntuple_filter_list filter_ntuple_list;
+ struct hinic_flow_mem_list hinic_flow_list;
};
#endif /* _HINIC_PMD_ETHDEV_H_ */
diff --git a/drivers/net/hinic/hinic_pmd_flow.c b/drivers/net/hinic/hinic_pmd_flow.c
index cf9f105..73a336a 100644
--- a/drivers/net/hinic/hinic_pmd_flow.c
+++ b/drivers/net/hinic/hinic_pmd_flow.c
@@ -23,6 +23,8 @@
#include "base/hinic_pmd_niccfg.h"
#include "hinic_pmd_ethdev.h"
+#define HINIC_MAX_RX_QUEUE_NUM 64
+
#ifndef UINT8_MAX
#define UINT8_MAX (u8)(~((u8)0)) /* 0xFF */
#define UINT16_MAX (u16)(~((u16)0)) /* 0xFFFF */
@@ -31,9 +33,56 @@
#define ASCII_MAX (0x7F)
#endif
+/* IPSURX MACRO */
+#define PA_ETH_TYPE_ROCE 0
+#define PA_ETH_TYPE_IPV4 1
+#define PA_ETH_TYPE_IPV6 2
+#define PA_ETH_TYPE_OTHER 3
+
+#define PA_IP_PROTOCOL_TYPE_TCP 1
+#define PA_IP_PROTOCOL_TYPE_UDP 2
+#define PA_IP_PROTOCOL_TYPE_ICMP 3
+#define PA_IP_PROTOCOL_TYPE_IPV4_IGMP 4
+#define PA_IP_PROTOCOL_TYPE_SCTP 5
+#define PA_IP_PROTOCOL_TYPE_VRRP 112
+
+#define IP_HEADER_PROTOCOL_TYPE_TCP 6
+
#define HINIC_MIN_N_TUPLE_PRIO 1
#define HINIC_MAX_N_TUPLE_PRIO 7
+/* TCAM type mask in hardware */
+#define TCAM_PKT_BGP_SPORT 1
+#define TCAM_PKT_VRRP 2
+#define TCAM_PKT_BGP_DPORT 3
+#define TCAM_PKT_LACP 4
+
+#define BGP_DPORT_ID 179
+#define IPPROTO_VRRP 112
+
+/* Packet type defined in hardware to perform filter */
+#define PKT_IGMP_IPV4_TYPE 64
+#define PKT_ICMP_IPV4_TYPE 65
+#define PKT_ICMP_IPV6_TYPE 66
+#define PKT_ICMP_IPV6RS_TYPE 67
+#define PKT_ICMP_IPV6RA_TYPE 68
+#define PKT_ICMP_IPV6NS_TYPE 69
+#define PKT_ICMP_IPV6NA_TYPE 70
+#define PKT_ICMP_IPV6RE_TYPE 71
+#define PKT_DHCP_IPV4_TYPE 72
+#define PKT_DHCP_IPV6_TYPE 73
+#define PKT_LACP_TYPE 74
+#define PKT_ARP_REQ_TYPE 79
+#define PKT_ARP_REP_TYPE 80
+#define PKT_ARP_TYPE 81
+#define PKT_BGPD_DPORT_TYPE 83
+#define PKT_BGPD_SPORT_TYPE 84
+#define PKT_VRRP_TYPE 85
+
+#define HINIC_DEV_PRIVATE_TO_FILTER_INFO(nic_dev) \
+ (&((struct hinic_nic_dev *)nic_dev)->filter)
+
+
/**
* Endless loop will never happen with below assumption
* 1. there is at least one no-void item(END)
@@ -1167,6 +1216,641 @@ static int hinic_flow_validate(struct rte_eth_dev *dev,
return ret;
}
+static inline int
+ntuple_ip_filter(struct rte_eth_ntuple_filter *filter,
+ struct hinic_5tuple_filter_info *filter_info)
+{
+ switch (filter->dst_ip_mask) {
+ case UINT32_MAX:
+ filter_info->dst_ip_mask = 0;
+ filter_info->dst_ip = filter->dst_ip;
+ break;
+ case 0:
+ filter_info->dst_ip_mask = 1;
+ filter_info->dst_ip = 0;
+ break;
+ default:
+ PMD_DRV_LOG(ERR, "Invalid dst_ip mask.");
+ return -EINVAL;
+ }
+
+ switch (filter->src_ip_mask) {
+ case UINT32_MAX:
+ filter_info->src_ip_mask = 0;
+ filter_info->src_ip = filter->src_ip;
+ break;
+ case 0:
+ filter_info->src_ip_mask = 1;
+ filter_info->src_ip = 0;
+ break;
+ default:
+ PMD_DRV_LOG(ERR, "Invalid src_ip mask.");
+ return -EINVAL;
+ }
+ return 0;
+}
+
+static inline int
+ntuple_port_filter(struct rte_eth_ntuple_filter *filter,
+ struct hinic_5tuple_filter_info *filter_info)
+{
+ switch (filter->dst_port_mask) {
+ case UINT16_MAX:
+ filter_info->dst_port_mask = 0;
+ filter_info->dst_port = filter->dst_port;
+ break;
+ case 0:
+ filter_info->dst_port_mask = 1;
+ filter_info->dst_port = 0;
+ break;
+ default:
+ PMD_DRV_LOG(ERR, "Invalid dst_port mask.");
+ return -EINVAL;
+ }
+
+ switch (filter->src_port_mask) {
+ case UINT16_MAX:
+ filter_info->src_port_mask = 0;
+ filter_info->src_port = filter->src_port;
+ break;
+ case 0:
+ filter_info->src_port_mask = 1;
+ filter_info->src_port = 0;
+ break;
+ default:
+ PMD_DRV_LOG(ERR, "Invalid src_port mask.");
+ return -EINVAL;
+ }
+
+ return 0;
+}
+
+static inline int
+ntuple_proto_filter(struct rte_eth_ntuple_filter *filter,
+ struct hinic_5tuple_filter_info *filter_info)
+{
+ switch (filter->proto_mask) {
+ case UINT8_MAX:
+ filter_info->proto_mask = 0;
+ filter_info->proto = filter->proto;
+ break;
+ case 0:
+ filter_info->proto_mask = 1;
+ filter_info->proto = 0;
+ break;
+ default:
+ PMD_DRV_LOG(ERR, "Invalid protocol mask.");
+ return -EINVAL;
+ }
+
+ return 0;
+}
+
+static inline int
+ntuple_filter_to_5tuple(struct rte_eth_ntuple_filter *filter,
+ struct hinic_5tuple_filter_info *filter_info)
+{
+ if (filter->queue >= HINIC_MAX_RX_QUEUE_NUM ||
+ filter->priority > HINIC_MAX_N_TUPLE_PRIO ||
+ filter->priority < HINIC_MIN_N_TUPLE_PRIO)
+ return -EINVAL;
+
+ if (ntuple_ip_filter(filter, filter_info) ||
+ ntuple_port_filter(filter, filter_info) ||
+ ntuple_proto_filter(filter, filter_info))
+ return -EINVAL;
+
+ filter_info->priority = (uint8_t)filter->priority;
+ return 0;
+}
+
+static inline struct hinic_5tuple_filter *
+hinic_5tuple_filter_lookup(struct hinic_5tuple_filter_list *filter_list,
+ struct hinic_5tuple_filter_info *key)
+{
+ struct hinic_5tuple_filter *it;
+
+ TAILQ_FOREACH(it, filter_list, entries) {
+ if (memcmp(key, &it->filter_info,
+ sizeof(struct hinic_5tuple_filter_info)) == 0) {
+ return it;
+ }
+ }
+
+ return NULL;
+}
+
+static int hinic_set_bgp_dport_tcam(struct hinic_nic_dev *nic_dev)
+{
+ struct tag_pa_rule bgp_rule;
+ struct tag_pa_action bgp_action;
+
+ memset(&bgp_rule, 0, sizeof(bgp_rule));
+ memset(&bgp_action, 0, sizeof(bgp_action));
+ /* BGP TCAM rule */
+ bgp_rule.eth_type = PA_ETH_TYPE_IPV4; /* Eth type is IPV4 */
+ bgp_rule.ip_header.protocol.val8 = IP_HEADER_PROTOCOL_TYPE_TCP;
+ bgp_rule.ip_header.protocol.mask8 = UINT8_MAX;
+ bgp_rule.ip_protocol_type = PA_IP_PROTOCOL_TYPE_TCP;
+ bgp_rule.eth_ip_tcp.dport.val16 = BGP_DPORT_ID; /* Dport is 179 */
+ bgp_rule.eth_ip_tcp.dport.mask16 = UINT16_MAX;
+
+ /* BGP TCAM action */
+ bgp_action.err_type = 0x3f; /* err from ipsu, not convert */
+ bgp_action.fwd_action = 0x7; /* 0x3:drop; 0x7: not convert */
+ bgp_action.pkt_type = PKT_BGPD_DPORT_TYPE; /* bgp_dport: 83 */
+ bgp_action.pri = 0xf; /* pri of BGP is 0xf, result from ipsu parse
+ * results, not need to convert
+ */
+ bgp_action.push_len = 0xf; /* push_len:0xf, not convert */
+
+ return hinic_set_fdir_tcam(nic_dev->hwdev,
+ TCAM_PKT_BGP_DPORT, &bgp_rule, &bgp_action);
+}
+
+static int hinic_set_bgp_sport_tcam(struct hinic_nic_dev *nic_dev)
+{
+ struct tag_pa_rule bgp_rule;
+ struct tag_pa_action bgp_action;
+
+ memset(&bgp_rule, 0, sizeof(bgp_rule));
+ memset(&bgp_action, 0, sizeof(bgp_action));
+ /* BGP TCAM rule */
+ bgp_rule.eth_type = PA_ETH_TYPE_IPV4;
+ bgp_rule.ip_header.protocol.val8 = IP_HEADER_PROTOCOL_TYPE_TCP;
+ bgp_rule.ip_header.protocol.mask8 = UINT8_MAX;
+ bgp_rule.ip_protocol_type = PA_IP_PROTOCOL_TYPE_TCP;
+ bgp_rule.eth_ip_tcp.sport.val16 = BGP_DPORT_ID;
+ bgp_rule.eth_ip_tcp.sport.mask16 = UINT16_MAX;
+
+ /* BGP TCAM action */
+ bgp_action.err_type = 0x3f; /* err from ipsu, not convert */
+ bgp_action.fwd_action = 0x7; /* 0x3:drop; 0x7: not convert */
+ bgp_action.pkt_type = PKT_BGPD_SPORT_TYPE; /* bgp:sport: 84 */
+ bgp_action.pri = 0xf; /* pri of BGP is 0xf, result from ipsu parse
+ * results, not need to convert
+ */
+ bgp_action.push_len = 0xf; /* push_len:0xf, not convert */
+
+ return hinic_set_fdir_tcam(nic_dev->hwdev, TCAM_PKT_BGP_SPORT,
+ &bgp_rule, &bgp_action);
+}
+
+static int hinic_set_vrrp_tcam(struct hinic_nic_dev *nic_dev)
+{
+ struct tag_pa_rule vrrp_rule;
+ struct tag_pa_action vrrp_action;
+
+ memset(&vrrp_rule, 0, sizeof(vrrp_rule));
+ memset(&vrrp_action, 0, sizeof(vrrp_action));
+ /* VRRP TCAM rule */
+ vrrp_rule.eth_type = PA_ETH_TYPE_IPV4;
+ vrrp_rule.ip_protocol_type = PA_IP_PROTOCOL_TYPE_TCP;
+ vrrp_rule.ip_header.protocol.mask8 = 0xff;
+ vrrp_rule.ip_header.protocol.val8 = PA_IP_PROTOCOL_TYPE_VRRP;
+
+ /* VRRP TCAM action */
+ vrrp_action.err_type = 0x3f;
+ vrrp_action.fwd_action = 0x7;
+ vrrp_action.pkt_type = PKT_VRRP_TYPE; /* VRRP: 85 */
+ vrrp_action.pri = 0xf;
+ vrrp_action.push_len = 0xf;
+
+ return hinic_set_fdir_tcam(nic_dev->hwdev, TCAM_PKT_VRRP,
+ &vrrp_rule, &vrrp_action);
+}
+
+static int
+hinic_filter_info_init(struct hinic_5tuple_filter *filter,
+ struct hinic_filter_info *filter_info)
+{
+ switch (filter->filter_info.proto) {
+ case IPPROTO_TCP:
+ /* Filter type is bgp type if dst_port or src_port is 179 */
+ if (filter->filter_info.dst_port == RTE_BE16(BGP_DPORT_ID) &&
+ !(filter->filter_info.dst_port_mask)) {
+ filter_info->pkt_type = PKT_BGPD_DPORT_TYPE;
+ } else if (filter->filter_info.src_port ==
+ RTE_BE16(BGP_DPORT_ID) &&
+ !(filter->filter_info.src_port_mask)) {
+ filter_info->pkt_type = PKT_BGPD_SPORT_TYPE;
+ } else {
+ PMD_DRV_LOG(INFO, "TCP PROTOCOL:5tuple filters"
+ " just support BGP now, proto:0x%x, "
+ "dst_port:0x%x, dst_port_mask:0x%x."
+ "src_port:0x%x, src_port_mask:0x%x.",
+ filter->filter_info.proto,
+ filter->filter_info.dst_port,
+ filter->filter_info.dst_port_mask,
+ filter->filter_info.src_port,
+ filter->filter_info.src_port_mask);
+ return -EINVAL;
+ }
+ break;
+
+ case IPPROTO_VRRP:
+ filter_info->pkt_type = PKT_VRRP_TYPE;
+ break;
+
+ case IPPROTO_ICMP:
+ filter_info->pkt_type = PKT_ICMP_IPV4_TYPE;
+ break;
+
+ case IPPROTO_ICMPV6:
+ filter_info->pkt_type = PKT_ICMP_IPV6_TYPE;
+ break;
+
+ default:
+ PMD_DRV_LOG(ERR, "5tuple filters just support BGP/VRRP/ICMP now, "
+ "proto: 0x%x, dst_port: 0x%x, dst_port_mask: 0x%x."
+ "src_port: 0x%x, src_port_mask: 0x%x.",
+ filter->filter_info.proto, filter->filter_info.dst_port,
+ filter->filter_info.dst_port_mask,
+ filter->filter_info.src_port,
+ filter->filter_info.src_port_mask);
+ return -EINVAL;
+ }
+
+ return 0;
+}
+
+static int
+hinic_lookup_new_filter(struct hinic_5tuple_filter *filter,
+ struct hinic_filter_info *filter_info,
+ int *index)
+{
+ int type_id;
+
+ type_id = HINIC_PKT_TYPE_FIND_ID(filter_info->pkt_type);
+
+ if (type_id > HINIC_MAX_Q_FILTERS - 1) {
+ PMD_DRV_LOG(ERR, "Pkt filters only support 64 filter type.");
+ return -EINVAL;
+ }
+
+ if (!(filter_info->type_mask & (1 << type_id))) {
+ filter_info->type_mask |= 1 << type_id;
+ filter->index = type_id;
+ filter_info->pkt_filters[type_id].enable = true;
+ filter_info->pkt_filters[type_id].pkt_proto =
+ filter->filter_info.proto;
+ TAILQ_INSERT_TAIL(&filter_info->fivetuple_list,
+ filter, entries);
+ } else {
+ PMD_DRV_LOG(ERR, "Filter type: %d exists.", type_id);
+ return -EIO;
+ }
+
+ *index = type_id;
+ return 0;
+}
+
+/*
+ * Add a 5tuple filter
+ *
+ * @param dev:
+ * Pointer to struct rte_eth_dev.
+ * @param filter:
+ * Pointer to the filter that will be added.
+ * @return
+ * - On success, zero.
+ * - On failure, a negative value.
+ */
+static int
+hinic_add_5tuple_filter(struct rte_eth_dev *dev,
+ struct hinic_5tuple_filter *filter)
+{
+ struct hinic_filter_info *filter_info =
+ HINIC_DEV_PRIVATE_TO_FILTER_INFO(dev->data->dev_private);
+ int i, ret_fw;
+ struct hinic_nic_dev *nic_dev = HINIC_ETH_DEV_TO_PRIVATE_NIC_DEV(dev);
+
+ if (hinic_filter_info_init(filter, filter_info) ||
+ hinic_lookup_new_filter(filter, filter_info, &i))
+ return -EFAULT;
+
+ ret_fw = hinic_set_fdir_filter(nic_dev->hwdev, filter_info->pkt_type,
+ filter_info->qid,
+ filter_info->pkt_filters[i].enable,
+ true);
+ if (ret_fw) {
+ PMD_DRV_LOG(ERR, "Set fdir filter failed, type: 0x%x, qid: 0x%x, enable: 0x%x",
+ filter_info->pkt_type, filter->queue,
+ filter_info->pkt_filters[i].enable);
+ return -EFAULT;
+ }
+
+ PMD_DRV_LOG(INFO, "Add 5tuple succeed, type: 0x%x, qid: 0x%x, enable: 0x%x",
+ filter_info->pkt_type, filter_info->qid,
+ filter_info->pkt_filters[filter->index].enable);
+
+ switch (filter->filter_info.proto) {
+ case IPPROTO_TCP:
+ if (filter->filter_info.dst_port == RTE_BE16(BGP_DPORT_ID)) {
+ ret_fw = hinic_set_bgp_dport_tcam(nic_dev);
+ if (ret_fw) {
+ PMD_DRV_LOG(ERR, "Set dport bgp failed, "
+ "type: 0x%x, qid: 0x%x, enable: 0x%x",
+ filter_info->pkt_type, filter->queue,
+ filter_info->pkt_filters[i].enable);
+ return -EFAULT;
+ }
+
+ PMD_DRV_LOG(INFO, "Set dport bgp succeed, qid: 0x%x, enable: 0x%x",
+ filter->queue,
+ filter_info->pkt_filters[i].enable);
+ } else if (filter->filter_info.src_port ==
+ RTE_BE16(BGP_DPORT_ID)) {
+ ret_fw = hinic_set_bgp_sport_tcam(nic_dev);
+ if (ret_fw) {
+ PMD_DRV_LOG(ERR, "Set sport bgp failed, "
+ "type: 0x%x, qid: 0x%x, enable: 0x%x",
+ filter_info->pkt_type, filter->queue,
+ filter_info->pkt_filters[i].enable);
+ return -EFAULT;
+ }
+
+ PMD_DRV_LOG(INFO, "Set sport bgp succeed, qid: 0x%x, enable: 0x%x",
+ filter->queue,
+ filter_info->pkt_filters[i].enable);
+ }
+
+ break;
+
+ case IPPROTO_VRRP:
+ ret_fw = hinic_set_vrrp_tcam(nic_dev);
+ if (ret_fw) {
+ PMD_DRV_LOG(ERR, "Set VRRP failed, "
+ "type: 0x%x, qid: 0x%x, enable: 0x%x",
+ filter_info->pkt_type, filter->queue,
+ filter_info->pkt_filters[i].enable);
+ return -EFAULT;
+ }
+ PMD_DRV_LOG(INFO, "Set VRRP succeed, qid: 0x%x, enable: 0x%x",
+ filter->queue,
+ filter_info->pkt_filters[i].enable);
+ break;
+
+ default:
+ break;
+ }
+
+ return 0;
+}
+
+/*
+ * Remove a 5tuple filter
+ *
+ * @param dev
+ * Pointer to struct rte_eth_dev.
+ * @param filter
+ * The pointer of the filter will be removed.
+ */
+static void
+hinic_remove_5tuple_filter(struct rte_eth_dev *dev,
+ struct hinic_5tuple_filter *filter)
+{
+ struct hinic_filter_info *filter_info =
+ HINIC_DEV_PRIVATE_TO_FILTER_INFO(dev->data->dev_private);
+ struct hinic_nic_dev *nic_dev = HINIC_ETH_DEV_TO_PRIVATE_NIC_DEV(dev);
+
+ switch (filter->filter_info.proto) {
+ case IPPROTO_VRRP:
+ (void)hinic_clear_fdir_tcam(nic_dev->hwdev, TCAM_PKT_VRRP);
+ break;
+
+ case IPPROTO_TCP:
+ if (filter->filter_info.dst_port == RTE_BE16(BGP_DPORT_ID))
+ (void)hinic_clear_fdir_tcam(nic_dev->hwdev,
+ TCAM_PKT_BGP_DPORT);
+ else if (filter->filter_info.src_port == RTE_BE16(BGP_DPORT_ID))
+ (void)hinic_clear_fdir_tcam(nic_dev->hwdev,
+ TCAM_PKT_BGP_SPORT);
+ break;
+
+ default:
+ break;
+ }
+
+ hinic_filter_info_init(filter, filter_info);
+
+ filter_info->pkt_filters[filter->index].enable = false;
+ filter_info->pkt_filters[filter->index].pkt_proto = 0;
+
+ PMD_DRV_LOG(INFO, "Del 5tuple succeed, type: 0x%x, qid: 0x%x, enable: 0x%x",
+ filter_info->pkt_type,
+ filter_info->pkt_filters[filter->index].qid,
+ filter_info->pkt_filters[filter->index].enable);
+ (void)hinic_set_fdir_filter(nic_dev->hwdev, filter_info->pkt_type,
+ filter_info->pkt_filters[filter->index].qid,
+ filter_info->pkt_filters[filter->index].enable,
+ true);
+
+ filter_info->pkt_type = 0;
+ filter_info->qid = 0;
+ filter_info->pkt_filters[filter->index].qid = 0;
+ filter_info->type_mask &= ~(1 << (filter->index));
+ TAILQ_REMOVE(&filter_info->fivetuple_list, filter, entries);
+
+ rte_free(filter);
+}
+
+/*
+ * Add or delete a ntuple filter
+ *
+ * @param dev
+ * Pointer to struct rte_eth_dev.
+ * @param ntuple_filter
+ * Pointer to struct rte_eth_ntuple_filter
+ * @param add
+ * If true, add filter; if false, remove filter
+ * @return
+ * - On success, zero.
+ * - On failure, a negative value.
+ */
+static int hinic_add_del_ntuple_filter(struct rte_eth_dev *dev,
+ struct rte_eth_ntuple_filter *ntuple_filter,
+ bool add)
+{
+ struct hinic_filter_info *filter_info =
+ HINIC_DEV_PRIVATE_TO_FILTER_INFO(dev->data->dev_private);
+ struct hinic_5tuple_filter_info filter_5tuple;
+ struct hinic_5tuple_filter *filter;
+ int ret;
+
+ if (ntuple_filter->flags != RTE_5TUPLE_FLAGS) {
+ PMD_DRV_LOG(ERR, "Only 5tuple is supported.");
+ return -EINVAL;
+ }
+
+ memset(&filter_5tuple, 0, sizeof(struct hinic_5tuple_filter_info));
+ ret = ntuple_filter_to_5tuple(ntuple_filter, &filter_5tuple);
+ if (ret < 0)
+ return ret;
+
+ filter = hinic_5tuple_filter_lookup(&filter_info->fivetuple_list,
+ &filter_5tuple);
+ if (filter != NULL && add) {
+ PMD_DRV_LOG(ERR, "Filter exists.");
+ return -EEXIST;
+ }
+ if (filter == NULL && !add) {
+ PMD_DRV_LOG(ERR, "Filter doesn't exist.");
+ return -ENOENT;
+ }
+
+ if (add) {
+ filter = rte_zmalloc("hinic_5tuple_filter",
+ sizeof(struct hinic_5tuple_filter), 0);
+ if (filter == NULL)
+ return -ENOMEM;
+ rte_memcpy(&filter->filter_info, &filter_5tuple,
+ sizeof(struct hinic_5tuple_filter_info));
+ filter->queue = ntuple_filter->queue;
+
+ filter_info->qid = ntuple_filter->queue;
+
+ ret = hinic_add_5tuple_filter(dev, filter);
+ if (ret)
+ rte_free(filter);
+
+ return ret;
+ }
+
+ hinic_remove_5tuple_filter(dev, filter);
+
+ return 0;
+}
+
+/**
+ * Create or destroy a flow rule.
+ * Theorically one rule can match more than one filters.
+ * We will let it use the filter which it hitt first.
+ * So, the sequence matters.
+ */
+static struct rte_flow *hinic_flow_create(struct rte_eth_dev *dev,
+ const struct rte_flow_attr *attr,
+ const struct rte_flow_item pattern[],
+ const struct rte_flow_action actions[],
+ struct rte_flow_error *error)
+{
+ int ret;
+ struct rte_eth_ntuple_filter ntuple_filter;
+ struct rte_flow *flow = NULL;
+ struct hinic_ntuple_filter_ele *ntuple_filter_ptr;
+ struct hinic_flow_mem *hinic_flow_mem_ptr;
+ struct hinic_nic_dev *nic_dev = HINIC_ETH_DEV_TO_PRIVATE_NIC_DEV(dev);
+
+ flow = rte_zmalloc("hinic_rte_flow", sizeof(struct rte_flow), 0);
+ if (!flow) {
+ PMD_DRV_LOG(ERR, "Failed to allocate flow memory");
+ return NULL;
+ }
+
+ hinic_flow_mem_ptr = rte_zmalloc("hinic_flow_mem",
+ sizeof(struct hinic_flow_mem), 0);
+ if (!hinic_flow_mem_ptr) {
+ PMD_DRV_LOG(ERR, "Failed to allocate hinic_flow_mem_ptr");
+ rte_free(flow);
+ return NULL;
+ }
+
+ hinic_flow_mem_ptr->flow = flow;
+ TAILQ_INSERT_TAIL(&nic_dev->hinic_flow_list, hinic_flow_mem_ptr,
+ entries);
+
+ /* add ntuple filter */
+ memset(&ntuple_filter, 0, sizeof(struct rte_eth_ntuple_filter));
+ ret = hinic_parse_ntuple_filter(dev, attr, pattern,
+ actions, &ntuple_filter, error);
+ if (ret)
+ goto out;
+
+ ret = hinic_add_del_ntuple_filter(dev, &ntuple_filter, TRUE);
+ if (ret)
+ goto out;
+ ntuple_filter_ptr = rte_zmalloc("hinic_ntuple_filter",
+ sizeof(struct hinic_ntuple_filter_ele), 0);
+ rte_memcpy(&ntuple_filter_ptr->filter_info,
+ &ntuple_filter,
+ sizeof(struct rte_eth_ntuple_filter));
+ TAILQ_INSERT_TAIL(&nic_dev->filter_ntuple_list,
+ ntuple_filter_ptr, entries);
+ flow->rule = ntuple_filter_ptr;
+ flow->filter_type = RTE_ETH_FILTER_NTUPLE;
+
+ PMD_DRV_LOG(INFO, "Create flow ntuple succeed, func_id: 0x%x",
+ hinic_global_func_id(nic_dev->hwdev));
+ return flow;
+
+out:
+ TAILQ_REMOVE(&nic_dev->hinic_flow_list, hinic_flow_mem_ptr, entries);
+ rte_flow_error_set(error, -ret,
+ RTE_FLOW_ERROR_TYPE_HANDLE, NULL,
+ "Failed to create flow.");
+ rte_free(hinic_flow_mem_ptr);
+ rte_free(flow);
+ return NULL;
+}
+
+/* Destroy a flow rule on hinic. */
+static int hinic_flow_destroy(struct rte_eth_dev *dev,
+ struct rte_flow *flow,
+ struct rte_flow_error *error)
+{
+ int ret;
+ struct rte_flow *pmd_flow = flow;
+ enum rte_filter_type filter_type = pmd_flow->filter_type;
+ struct rte_eth_ntuple_filter ntuple_filter;
+ struct hinic_ntuple_filter_ele *ntuple_filter_ptr;
+ struct hinic_flow_mem *hinic_flow_mem_ptr;
+ struct hinic_nic_dev *nic_dev = HINIC_ETH_DEV_TO_PRIVATE_NIC_DEV(dev);
+
+ switch (filter_type) {
+ case RTE_ETH_FILTER_NTUPLE:
+ ntuple_filter_ptr = (struct hinic_ntuple_filter_ele *)
+ pmd_flow->rule;
+ rte_memcpy(&ntuple_filter, &ntuple_filter_ptr->filter_info,
+ sizeof(struct rte_eth_ntuple_filter));
+ ret = hinic_add_del_ntuple_filter(dev, &ntuple_filter, FALSE);
+ if (!ret) {
+ TAILQ_REMOVE(&nic_dev->filter_ntuple_list,
+ ntuple_filter_ptr, entries);
+ rte_free(ntuple_filter_ptr);
+ }
+ break;
+ default:
+ PMD_DRV_LOG(WARNING, "Filter type (%d) not supported",
+ filter_type);
+ ret = -EINVAL;
+ break;
+ }
+
+ if (ret) {
+ rte_flow_error_set(error, EINVAL,
+ RTE_FLOW_ERROR_TYPE_HANDLE,
+ NULL, "Failed to destroy flow");
+ return ret;
+ }
+
+ TAILQ_FOREACH(hinic_flow_mem_ptr, &nic_dev->hinic_flow_list, entries) {
+ if (hinic_flow_mem_ptr->flow == pmd_flow) {
+ TAILQ_REMOVE(&nic_dev->hinic_flow_list,
+ hinic_flow_mem_ptr, entries);
+ rte_free(hinic_flow_mem_ptr);
+ break;
+ }
+ }
+ rte_free(flow);
+
+ PMD_DRV_LOG(INFO, "Destroy flow succeed, func_id: 0x%x",
+ hinic_global_func_id(nic_dev->hwdev));
+
+ return ret;
+}
+
const struct rte_flow_ops hinic_flow_ops = {
.validate = hinic_flow_validate,
+ .create = hinic_flow_create,
+ .destroy = hinic_flow_destroy,
};
--
1.8.3.1
^ permalink raw reply [flat|nested] 32+ messages in thread
* [dpdk-dev] [PATCH v2 10/17] net/hinic: create and destroy fdir filter
2019-09-25 14:26 [dpdk-dev] [PATCH v2 00/17] Add advanced features for Huawei hinic pmd Xiaoyun wang
` (9 preceding siblings ...)
2019-09-25 14:30 ` [dpdk-dev] [PATCH v2 09/17] net/hinic: create and destroy ntuple filter Xiaoyun wang
@ 2019-09-25 14:30 ` Xiaoyun wang
2019-09-25 14:30 ` [dpdk-dev] [PATCH v2 11/17] net/hinic: flush " Xiaoyun wang
` (7 subsequent siblings)
18 siblings, 0 replies; 32+ messages in thread
From: Xiaoyun wang @ 2019-09-25 14:30 UTC (permalink / raw)
To: ferruh.yigit
Cc: dev, xuanziyang2, shahar.belkar, luoxianjun, tanya.brokhman,
zhouguoyang, wulike1, Xiaoyun wang
Add support for creating/destroying fdir filter.
The filter rule includes LACP,ARP and ICMP packet type.
Signed-off-by: Xiaoyun wang <cloud.wangxiaoyun@huawei.com>
---
drivers/net/hinic/hinic_pmd_ethdev.h | 16 ++
drivers/net/hinic/hinic_pmd_flow.c | 423 +++++++++++++++++++++++++++++++++--
2 files changed, 422 insertions(+), 17 deletions(-)
diff --git a/drivers/net/hinic/hinic_pmd_ethdev.h b/drivers/net/hinic/hinic_pmd_ethdev.h
index 93cd6a7..b7fcef6 100644
--- a/drivers/net/hinic/hinic_pmd_ethdev.h
+++ b/drivers/net/hinic/hinic_pmd_ethdev.h
@@ -117,6 +117,18 @@ struct hinic_ntuple_filter_ele {
struct rte_eth_ntuple_filter filter_info;
};
+/* ethertype filter list structure */
+struct hinic_ethertype_filter_ele {
+ TAILQ_ENTRY(hinic_ethertype_filter_ele) entries;
+ struct rte_eth_ethertype_filter filter_info;
+};
+
+/* fdir filter list structure */
+struct hinic_fdir_rule_ele {
+ TAILQ_ENTRY(hinic_fdir_rule_ele) entries;
+ struct hinic_fdir_rule filter_info;
+};
+
struct rte_flow {
enum rte_filter_type filter_type;
void *rule;
@@ -129,6 +141,8 @@ struct hinic_flow_mem {
};
TAILQ_HEAD(hinic_ntuple_filter_list, hinic_ntuple_filter_ele);
+TAILQ_HEAD(hinic_ethertype_filter_list, hinic_ethertype_filter_ele);
+TAILQ_HEAD(hinic_fdir_rule_filter_list, hinic_fdir_rule_ele);
TAILQ_HEAD(hinic_flow_mem_list, hinic_flow_mem);
extern const struct rte_flow_ops hinic_flow_ops;
@@ -168,6 +182,8 @@ struct hinic_nic_dev {
struct hinic_filter_info filter;
struct hinic_ntuple_filter_list filter_ntuple_list;
+ struct hinic_ethertype_filter_list filter_ethertype_list;
+ struct hinic_fdir_rule_filter_list filter_fdir_rule_list;
struct hinic_flow_mem_list hinic_flow_list;
};
diff --git a/drivers/net/hinic/hinic_pmd_flow.c b/drivers/net/hinic/hinic_pmd_flow.c
index 73a336a..b2eff2e 100644
--- a/drivers/net/hinic/hinic_pmd_flow.c
+++ b/drivers/net/hinic/hinic_pmd_flow.c
@@ -82,6 +82,19 @@
#define HINIC_DEV_PRIVATE_TO_FILTER_INFO(nic_dev) \
(&((struct hinic_nic_dev *)nic_dev)->filter)
+enum hinic_atr_flow_type {
+ HINIC_ATR_FLOW_TYPE_IPV4_DIP = 0x1,
+ HINIC_ATR_FLOW_TYPE_IPV4_SIP = 0x2,
+ HINIC_ATR_FLOW_TYPE_DPORT = 0x3,
+ HINIC_ATR_FLOW_TYPE_SPORT = 0x4,
+};
+
+/* Structure to store fdir's info. */
+struct hinic_fdir_info {
+ uint8_t fdir_flag;
+ uint8_t qid;
+ uint32_t fdir_key;
+};
/**
* Endless loop will never happen with below assumption
@@ -1340,6 +1353,29 @@ static int hinic_flow_validate(struct rte_eth_dev *dev,
return NULL;
}
+static int hinic_set_lacp_tcam(struct hinic_nic_dev *nic_dev)
+{
+ struct tag_pa_rule lacp_rule;
+ struct tag_pa_action lacp_action;
+
+ memset(&lacp_rule, 0, sizeof(lacp_rule));
+ memset(&lacp_action, 0, sizeof(lacp_action));
+ /* LACP TCAM rule */
+ lacp_rule.eth_type = PA_ETH_TYPE_OTHER;
+ lacp_rule.l2_header.eth_type.val16 = 0x8809;
+ lacp_rule.l2_header.eth_type.mask16 = 0xffff;
+
+ /* LACP TCAM action */
+ lacp_action.err_type = 0x3f; /* err from ipsu, not convert */
+ lacp_action.fwd_action = 0x7; /* 0x3:drop; 0x7: not convert */
+ lacp_action.pkt_type = PKT_LACP_TYPE;
+ lacp_action.pri = 0x0;
+ lacp_action.push_len = 0xf; /* push_len:0xf, not convert */
+
+ return hinic_set_fdir_tcam(nic_dev->hwdev, TCAM_PKT_LACP,
+ &lacp_rule, &lacp_action);
+}
+
static int hinic_set_bgp_dport_tcam(struct hinic_nic_dev *nic_dev)
{
struct tag_pa_rule bgp_rule;
@@ -1722,6 +1758,276 @@ static int hinic_add_del_ntuple_filter(struct rte_eth_dev *dev,
return 0;
}
+static inline int
+hinic_check_ethertype_filter(struct rte_eth_ethertype_filter *filter)
+{
+ if (filter->queue >= HINIC_MAX_RX_QUEUE_NUM)
+ return -EINVAL;
+
+ if (filter->ether_type == RTE_ETHER_TYPE_IPV4 ||
+ filter->ether_type == RTE_ETHER_TYPE_IPV6) {
+ PMD_DRV_LOG(ERR, "Unsupported ether_type(0x%04x) in"
+ " ethertype filter", filter->ether_type);
+ return -EINVAL;
+ }
+
+ if (filter->flags & RTE_ETHTYPE_FLAGS_MAC) {
+ PMD_DRV_LOG(ERR, "Mac compare is not supported");
+ return -EINVAL;
+ }
+ if (filter->flags & RTE_ETHTYPE_FLAGS_DROP) {
+ PMD_DRV_LOG(ERR, "Drop option is not supported");
+ return -EINVAL;
+ }
+
+ return 0;
+}
+
+static inline int
+hinic_ethertype_filter_lookup(struct hinic_filter_info *filter_info,
+ struct hinic_pkt_filter *ethertype_filter)
+{
+ switch (ethertype_filter->pkt_proto) {
+ case RTE_ETHER_TYPE_SLOW:
+ filter_info->pkt_type = PKT_LACP_TYPE;
+ break;
+
+ case RTE_ETHER_TYPE_ARP:
+ filter_info->pkt_type = PKT_ARP_TYPE;
+ break;
+
+ default:
+ PMD_DRV_LOG(ERR, "Just support LACP/ARP for ethertype filters");
+ return -EIO;
+ }
+
+ return HINIC_PKT_TYPE_FIND_ID(filter_info->pkt_type);
+}
+
+static inline int
+hinic_ethertype_filter_insert(struct hinic_filter_info *filter_info,
+ struct hinic_pkt_filter *ethertype_filter)
+{
+ int id;
+
+ /* Find LACP or VRRP type id */
+ id = hinic_ethertype_filter_lookup(filter_info, ethertype_filter);
+ if (id < 0)
+ return -EINVAL;
+
+ if (!(filter_info->type_mask & (1 << id))) {
+ filter_info->type_mask |= 1 << id;
+ filter_info->pkt_filters[id].pkt_proto =
+ ethertype_filter->pkt_proto;
+ filter_info->pkt_filters[id].enable = ethertype_filter->enable;
+ filter_info->qid = ethertype_filter->qid;
+ return id;
+ }
+
+ PMD_DRV_LOG(ERR, "Filter type: %d exists", id);
+ return -EINVAL;
+}
+
+static inline void
+hinic_ethertype_filter_remove(struct hinic_filter_info *filter_info,
+ uint8_t idx)
+{
+ if (idx >= HINIC_MAX_Q_FILTERS)
+ return;
+
+ filter_info->pkt_type = 0;
+ filter_info->type_mask &= ~(1 << idx);
+ filter_info->pkt_filters[idx].pkt_proto = (uint16_t)0;
+ filter_info->pkt_filters[idx].enable = FALSE;
+ filter_info->pkt_filters[idx].qid = 0;
+}
+
+static inline int
+hinic_add_del_ethertype_filter(struct rte_eth_dev *dev,
+ struct rte_eth_ethertype_filter *filter,
+ bool add)
+{
+ struct hinic_nic_dev *nic_dev = HINIC_ETH_DEV_TO_PRIVATE_NIC_DEV(dev);
+ struct hinic_filter_info *filter_info =
+ HINIC_DEV_PRIVATE_TO_FILTER_INFO(dev->data->dev_private);
+ struct hinic_pkt_filter ethertype_filter;
+ int i;
+ int ret_fw;
+
+ if (hinic_check_ethertype_filter(filter))
+ return -EINVAL;
+
+ if (add) {
+ ethertype_filter.pkt_proto = filter->ether_type;
+ ethertype_filter.enable = TRUE;
+ ethertype_filter.qid = (u8)filter->queue;
+ i = hinic_ethertype_filter_insert(filter_info,
+ ðertype_filter);
+ if (i < 0)
+ return -ENOSPC;
+
+ ret_fw = hinic_set_fdir_filter(nic_dev->hwdev,
+ filter_info->pkt_type, filter_info->qid,
+ filter_info->pkt_filters[i].enable, true);
+ if (ret_fw) {
+ PMD_DRV_LOG(ERR, "add ethertype failed, type: 0x%x, qid: 0x%x, enable: 0x%x",
+ filter_info->pkt_type, filter->queue,
+ filter_info->pkt_filters[i].enable);
+
+ hinic_ethertype_filter_remove(filter_info, i);
+ return -ENOENT;
+ }
+ PMD_DRV_LOG(INFO, "Add ethertype succeed, type: 0x%x, qid: 0x%x, enable: 0x%x",
+ filter_info->pkt_type, filter->queue,
+ filter_info->pkt_filters[i].enable);
+
+ switch (ethertype_filter.pkt_proto) {
+ case RTE_ETHER_TYPE_SLOW:
+ ret_fw = hinic_set_lacp_tcam(nic_dev);
+ if (ret_fw) {
+ PMD_DRV_LOG(ERR, "Add lacp tcam failed");
+ hinic_ethertype_filter_remove(filter_info, i);
+ return -ENOENT;
+ }
+
+ PMD_DRV_LOG(INFO, "Add lacp tcam succeed");
+ break;
+ default:
+ break;
+ }
+
+ } else {
+ ethertype_filter.pkt_proto = filter->ether_type;
+ i = hinic_ethertype_filter_lookup(filter_info,
+ ðertype_filter);
+
+ if ((filter_info->type_mask & (1 << i))) {
+ filter_info->pkt_filters[i].enable = FALSE;
+ (void)hinic_set_fdir_filter(nic_dev->hwdev,
+ filter_info->pkt_type,
+ filter_info->pkt_filters[i].qid,
+ filter_info->pkt_filters[i].enable,
+ true);
+
+ PMD_DRV_LOG(INFO, "Del ethertype succeed, type: 0x%x, qid: 0x%x, enable: 0x%x",
+ filter_info->pkt_type,
+ filter_info->pkt_filters[i].qid,
+ filter_info->pkt_filters[i].enable);
+
+ switch (ethertype_filter.pkt_proto) {
+ case RTE_ETHER_TYPE_SLOW:
+ (void)hinic_clear_fdir_tcam(nic_dev->hwdev,
+ TCAM_PKT_LACP);
+ PMD_DRV_LOG(INFO,
+ "Del lacp tcam succeed");
+ break;
+ default:
+ break;
+ }
+
+ hinic_ethertype_filter_remove(filter_info, i);
+
+ } else {
+ PMD_DRV_LOG(ERR, "Ethertype doesn't exist, type: 0x%x, qid: 0x%x, enable: 0x%x",
+ filter_info->pkt_type, filter->queue,
+ filter_info->pkt_filters[i].enable);
+ return -ENOENT;
+ }
+ }
+
+ return 0;
+}
+
+static int
+hinic_fdir_info_init(struct hinic_fdir_rule *rule,
+ struct hinic_fdir_info *fdir_info)
+{
+ switch (rule->mask.src_ipv4_mask) {
+ case UINT32_MAX:
+ fdir_info->fdir_flag = HINIC_ATR_FLOW_TYPE_IPV4_SIP;
+ fdir_info->qid = rule->queue;
+ fdir_info->fdir_key = rule->hinic_fdir.src_ip;
+ return 0;
+
+ case 0:
+ break;
+
+ default:
+ PMD_DRV_LOG(ERR, "Invalid src_ip mask.");
+ return -EINVAL;
+ }
+
+ switch (rule->mask.dst_ipv4_mask) {
+ case UINT32_MAX:
+ fdir_info->fdir_flag = HINIC_ATR_FLOW_TYPE_IPV4_DIP;
+ fdir_info->qid = rule->queue;
+ fdir_info->fdir_key = rule->hinic_fdir.dst_ip;
+ return 0;
+
+ case 0:
+ break;
+
+ default:
+ PMD_DRV_LOG(ERR, "Invalid dst_ip mask.");
+ return -EINVAL;
+ }
+
+ if (fdir_info->fdir_flag == 0) {
+ PMD_DRV_LOG(ERR, "All support mask is NULL.");
+ return -EINVAL;
+ }
+
+ return 0;
+}
+
+static inline int
+hinic_add_del_fdir_filter(struct rte_eth_dev *dev,
+ struct hinic_fdir_rule *rule,
+ bool add)
+{
+ struct hinic_nic_dev *nic_dev = HINIC_ETH_DEV_TO_PRIVATE_NIC_DEV(dev);
+ struct hinic_fdir_info fdir_info;
+ int ret;
+
+ memset(&fdir_info, 0, sizeof(struct hinic_fdir_info));
+
+ ret = hinic_fdir_info_init(rule, &fdir_info);
+ if (ret) {
+ PMD_DRV_LOG(ERR, "Init hinic fdir info failed!");
+ return ret;
+ }
+
+ if (add) {
+ ret = hinic_set_normal_filter(nic_dev->hwdev, fdir_info.qid,
+ true, fdir_info.fdir_key,
+ true, fdir_info.fdir_flag);
+ if (ret) {
+ PMD_DRV_LOG(ERR, "Add fdir filter failed, flag: 0x%x, qid: 0x%x, key: 0x%x",
+ fdir_info.fdir_flag, fdir_info.qid,
+ fdir_info.fdir_key);
+ return -ENOENT;
+ }
+ PMD_DRV_LOG(INFO, "Add fdir filter succeed, flag: 0x%x, qid: 0x%x, key: 0x%x",
+ fdir_info.fdir_flag, fdir_info.qid,
+ fdir_info.fdir_key);
+ } else {
+ ret = hinic_set_normal_filter(nic_dev->hwdev, fdir_info.qid,
+ false, fdir_info.fdir_key, true,
+ fdir_info.fdir_flag);
+ if (ret) {
+ PMD_DRV_LOG(ERR, "Del fdir filter ailed, flag: 0x%x, qid: 0x%x, key: 0x%x",
+ fdir_info.fdir_flag, fdir_info.qid,
+ fdir_info.fdir_key);
+ return -ENOENT;
+ }
+ PMD_DRV_LOG(INFO, "Del fdir filter succeed, flag: 0x%x, qid: 0x%x, key: 0x%x",
+ fdir_info.fdir_flag, fdir_info.qid,
+ fdir_info.fdir_key);
+ }
+
+ return 0;
+}
+
/**
* Create or destroy a flow rule.
* Theorically one rule can match more than one filters.
@@ -1736,8 +2042,12 @@ static struct rte_flow *hinic_flow_create(struct rte_eth_dev *dev,
{
int ret;
struct rte_eth_ntuple_filter ntuple_filter;
+ struct rte_eth_ethertype_filter ethertype_filter;
+ struct hinic_fdir_rule fdir_rule;
struct rte_flow *flow = NULL;
+ struct hinic_ethertype_filter_ele *ethertype_filter_ptr;
struct hinic_ntuple_filter_ele *ntuple_filter_ptr;
+ struct hinic_fdir_rule_ele *fdir_rule_ptr;
struct hinic_flow_mem *hinic_flow_mem_ptr;
struct hinic_nic_dev *nic_dev = HINIC_ETH_DEV_TO_PRIVATE_NIC_DEV(dev);
@@ -1759,29 +2069,78 @@ static struct rte_flow *hinic_flow_create(struct rte_eth_dev *dev,
TAILQ_INSERT_TAIL(&nic_dev->hinic_flow_list, hinic_flow_mem_ptr,
entries);
- /* add ntuple filter */
+ /* Add ntuple filter */
memset(&ntuple_filter, 0, sizeof(struct rte_eth_ntuple_filter));
ret = hinic_parse_ntuple_filter(dev, attr, pattern,
actions, &ntuple_filter, error);
- if (ret)
+ if (!ret) {
+ ret = hinic_add_del_ntuple_filter(dev, &ntuple_filter, TRUE);
+ if (!ret) {
+ ntuple_filter_ptr = rte_zmalloc("hinic_ntuple_filter",
+ sizeof(struct hinic_ntuple_filter_ele), 0);
+ rte_memcpy(&ntuple_filter_ptr->filter_info,
+ &ntuple_filter,
+ sizeof(struct rte_eth_ntuple_filter));
+ TAILQ_INSERT_TAIL(&nic_dev->filter_ntuple_list,
+ ntuple_filter_ptr, entries);
+ flow->rule = ntuple_filter_ptr;
+ flow->filter_type = RTE_ETH_FILTER_NTUPLE;
+
+ PMD_DRV_LOG(INFO, "Create flow ntuple succeed, func_id: 0x%x",
+ hinic_global_func_id(nic_dev->hwdev));
+ return flow;
+ }
+ goto out;
+ }
+
+ /* Add ethertype filter */
+ memset(ðertype_filter, 0, sizeof(struct rte_eth_ethertype_filter));
+ ret = hinic_parse_ethertype_filter(dev, attr, pattern, actions,
+ ðertype_filter, error);
+ if (!ret) {
+ ret = hinic_add_del_ethertype_filter(dev, ðertype_filter,
+ TRUE);
+ if (!ret) {
+ ethertype_filter_ptr =
+ rte_zmalloc("hinic_ethertype_filter",
+ sizeof(struct hinic_ethertype_filter_ele), 0);
+ rte_memcpy(ðertype_filter_ptr->filter_info,
+ ðertype_filter,
+ sizeof(struct rte_eth_ethertype_filter));
+ TAILQ_INSERT_TAIL(&nic_dev->filter_ethertype_list,
+ ethertype_filter_ptr, entries);
+ flow->rule = ethertype_filter_ptr;
+ flow->filter_type = RTE_ETH_FILTER_ETHERTYPE;
+
+ PMD_DRV_LOG(INFO, "Create flow ethertype succeed, func_id: 0x%x",
+ hinic_global_func_id(nic_dev->hwdev));
+ return flow;
+ }
goto out;
+ }
- ret = hinic_add_del_ntuple_filter(dev, &ntuple_filter, TRUE);
- if (ret)
+ /* Add fdir filter */
+ memset(&fdir_rule, 0, sizeof(struct hinic_fdir_rule));
+ ret = hinic_parse_fdir_filter(dev, attr, pattern,
+ actions, &fdir_rule, error);
+ if (!ret) {
+ ret = hinic_add_del_fdir_filter(dev, &fdir_rule, TRUE);
+ if (!ret) {
+ fdir_rule_ptr = rte_zmalloc("hinic_fdir_rule",
+ sizeof(struct hinic_fdir_rule_ele), 0);
+ rte_memcpy(&fdir_rule_ptr->filter_info, &fdir_rule,
+ sizeof(struct hinic_fdir_rule));
+ TAILQ_INSERT_TAIL(&nic_dev->filter_fdir_rule_list,
+ fdir_rule_ptr, entries);
+ flow->rule = fdir_rule_ptr;
+ flow->filter_type = RTE_ETH_FILTER_FDIR;
+
+ PMD_DRV_LOG(INFO, "Create flow fdir rule succeed, func_id : 0x%x",
+ hinic_global_func_id(nic_dev->hwdev));
+ return flow;
+ }
goto out;
- ntuple_filter_ptr = rte_zmalloc("hinic_ntuple_filter",
- sizeof(struct hinic_ntuple_filter_ele), 0);
- rte_memcpy(&ntuple_filter_ptr->filter_info,
- &ntuple_filter,
- sizeof(struct rte_eth_ntuple_filter));
- TAILQ_INSERT_TAIL(&nic_dev->filter_ntuple_list,
- ntuple_filter_ptr, entries);
- flow->rule = ntuple_filter_ptr;
- flow->filter_type = RTE_ETH_FILTER_NTUPLE;
-
- PMD_DRV_LOG(INFO, "Create flow ntuple succeed, func_id: 0x%x",
- hinic_global_func_id(nic_dev->hwdev));
- return flow;
+ }
out:
TAILQ_REMOVE(&nic_dev->hinic_flow_list, hinic_flow_mem_ptr, entries);
@@ -1802,7 +2161,11 @@ static int hinic_flow_destroy(struct rte_eth_dev *dev,
struct rte_flow *pmd_flow = flow;
enum rte_filter_type filter_type = pmd_flow->filter_type;
struct rte_eth_ntuple_filter ntuple_filter;
+ struct rte_eth_ethertype_filter ethertype_filter;
+ struct hinic_fdir_rule fdir_rule;
struct hinic_ntuple_filter_ele *ntuple_filter_ptr;
+ struct hinic_ethertype_filter_ele *ethertype_filter_ptr;
+ struct hinic_fdir_rule_ele *fdir_rule_ptr;
struct hinic_flow_mem *hinic_flow_mem_ptr;
struct hinic_nic_dev *nic_dev = HINIC_ETH_DEV_TO_PRIVATE_NIC_DEV(dev);
@@ -1819,6 +2182,32 @@ static int hinic_flow_destroy(struct rte_eth_dev *dev,
rte_free(ntuple_filter_ptr);
}
break;
+ case RTE_ETH_FILTER_ETHERTYPE:
+ ethertype_filter_ptr = (struct hinic_ethertype_filter_ele *)
+ pmd_flow->rule;
+ rte_memcpy(ðertype_filter,
+ ðertype_filter_ptr->filter_info,
+ sizeof(struct rte_eth_ethertype_filter));
+ ret = hinic_add_del_ethertype_filter(dev,
+ ðertype_filter, FALSE);
+ if (!ret) {
+ TAILQ_REMOVE(&nic_dev->filter_ethertype_list,
+ ethertype_filter_ptr, entries);
+ rte_free(ethertype_filter_ptr);
+ }
+ break;
+ case RTE_ETH_FILTER_FDIR:
+ fdir_rule_ptr = (struct hinic_fdir_rule_ele *)pmd_flow->rule;
+ rte_memcpy(&fdir_rule,
+ &fdir_rule_ptr->filter_info,
+ sizeof(struct hinic_fdir_rule));
+ ret = hinic_add_del_fdir_filter(dev, &fdir_rule, FALSE);
+ if (!ret) {
+ TAILQ_REMOVE(&nic_dev->filter_fdir_rule_list,
+ fdir_rule_ptr, entries);
+ rte_free(fdir_rule_ptr);
+ }
+ break;
default:
PMD_DRV_LOG(WARNING, "Filter type (%d) not supported",
filter_type);
--
1.8.3.1
^ permalink raw reply [flat|nested] 32+ messages in thread
* [dpdk-dev] [PATCH v2 11/17] net/hinic: flush fdir filter
2019-09-25 14:26 [dpdk-dev] [PATCH v2 00/17] Add advanced features for Huawei hinic pmd Xiaoyun wang
` (10 preceding siblings ...)
2019-09-25 14:30 ` [dpdk-dev] [PATCH v2 10/17] net/hinic: create and destroy fdir filter Xiaoyun wang
@ 2019-09-25 14:30 ` Xiaoyun wang
2019-09-25 14:30 ` [dpdk-dev] [PATCH v2 12/17] net/hinic: set link down and up Xiaoyun wang
` (6 subsequent siblings)
18 siblings, 0 replies; 32+ messages in thread
From: Xiaoyun wang @ 2019-09-25 14:30 UTC (permalink / raw)
To: ferruh.yigit
Cc: dev, xuanziyang2, shahar.belkar, luoxianjun, tanya.brokhman,
zhouguoyang, wulike1, Xiaoyun wang
Supports to flush fdir filter.
Destroy all flow rules associated with a port on hinic.
Signed-off-by: Xiaoyun wang <cloud.wangxiaoyun@huawei.com>
---
drivers/net/hinic/hinic_pmd_ethdev.c | 16 ++++
drivers/net/hinic/hinic_pmd_ethdev.h | 3 +-
drivers/net/hinic/hinic_pmd_flow.c | 140 +++++++++++++++++++++++++++++++++++
3 files changed, 158 insertions(+), 1 deletion(-)
diff --git a/drivers/net/hinic/hinic_pmd_ethdev.c b/drivers/net/hinic/hinic_pmd_ethdev.c
index 9f7e6f5..2c329d1 100644
--- a/drivers/net/hinic/hinic_pmd_ethdev.c
+++ b/drivers/net/hinic/hinic_pmd_ethdev.c
@@ -341,6 +341,9 @@ static int hinic_dev_configure(struct rte_eth_dev *dev)
return err;
}
+ /*clear fdir filter flag in function table*/
+ hinic_free_fdir_filter(nic_dev);
+
return HINIC_OK;
}
@@ -1113,6 +1116,8 @@ static void hinic_dev_stop(struct rte_eth_dev *dev)
/* clean root context */
hinic_free_qp_ctxts(nic_dev->hwdev);
+ hinic_free_fdir_filter(nic_dev);
+
/* free mbuf */
hinic_free_all_rx_mbuf(dev);
hinic_free_all_tx_mbuf(dev);
@@ -2814,6 +2819,7 @@ static int hinic_func_init(struct rte_eth_dev *eth_dev)
struct rte_pci_device *pci_dev;
struct rte_ether_addr *eth_addr;
struct hinic_nic_dev *nic_dev;
+ struct hinic_filter_info *filter_info;
u32 mac_size;
int rc;
@@ -2905,6 +2911,16 @@ static int hinic_func_init(struct rte_eth_dev *eth_dev)
}
hinic_set_bit(HINIC_DEV_INTR_EN, &nic_dev->dev_status);
+ /* initialize filter info */
+ filter_info = &nic_dev->filter;
+ memset(filter_info, 0, sizeof(struct hinic_filter_info));
+ /* initialize 5tuple filter list */
+ TAILQ_INIT(&filter_info->fivetuple_list);
+ TAILQ_INIT(&nic_dev->filter_ntuple_list);
+ TAILQ_INIT(&nic_dev->filter_ethertype_list);
+ TAILQ_INIT(&nic_dev->filter_fdir_rule_list);
+ TAILQ_INIT(&nic_dev->hinic_flow_list);
+
hinic_set_bit(HINIC_DEV_INIT, &nic_dev->dev_status);
PMD_DRV_LOG(INFO, "Initialize %s in primary successfully",
eth_dev->data->name);
diff --git a/drivers/net/hinic/hinic_pmd_ethdev.h b/drivers/net/hinic/hinic_pmd_ethdev.h
index b7fcef6..dd96667 100644
--- a/drivers/net/hinic/hinic_pmd_ethdev.h
+++ b/drivers/net/hinic/hinic_pmd_ethdev.h
@@ -173,7 +173,6 @@ struct hinic_nic_dev {
u32 rx_mode_status; /* promisc or allmulticast */
unsigned long dev_status;
- /* dpdk only */
char proc_dev_name[HINIC_DEV_NAME_LEN];
/* PF0->COS4, PF1->COS5, PF2->COS6, PF3->COS7,
* vf: the same with associate pf
@@ -187,4 +186,6 @@ struct hinic_nic_dev {
struct hinic_flow_mem_list hinic_flow_list;
};
+void hinic_free_fdir_filter(struct hinic_nic_dev *nic_dev);
+
#endif /* _HINIC_PMD_ETHDEV_H_ */
diff --git a/drivers/net/hinic/hinic_pmd_flow.c b/drivers/net/hinic/hinic_pmd_flow.c
index b2eff2e..49c9d87 100644
--- a/drivers/net/hinic/hinic_pmd_flow.c
+++ b/drivers/net/hinic/hinic_pmd_flow.c
@@ -1456,6 +1456,40 @@ static int hinic_set_vrrp_tcam(struct hinic_nic_dev *nic_dev)
&vrrp_rule, &vrrp_action);
}
+/**
+ * Clear all fdir configuration.
+ *
+ * @param nic_dev
+ * The hardware interface of a Ethernet device.
+ *
+ * @return
+ * 0 on success,
+ * negative error value otherwise.
+ */
+void hinic_free_fdir_filter(struct hinic_nic_dev *nic_dev)
+{
+ struct hinic_filter_info *filter_info =
+ HINIC_DEV_PRIVATE_TO_FILTER_INFO(nic_dev);
+
+ if (filter_info->type_mask &
+ (1 << HINIC_PKT_TYPE_FIND_ID(PKT_BGPD_DPORT_TYPE)))
+ hinic_clear_fdir_tcam(nic_dev->hwdev, TCAM_PKT_BGP_DPORT);
+
+ if (filter_info->type_mask &
+ (1 << HINIC_PKT_TYPE_FIND_ID(PKT_BGPD_SPORT_TYPE)))
+ hinic_clear_fdir_tcam(nic_dev->hwdev, TCAM_PKT_BGP_SPORT);
+
+ if (filter_info->type_mask &
+ (1 << HINIC_PKT_TYPE_FIND_ID(PKT_VRRP_TYPE)))
+ hinic_clear_fdir_tcam(nic_dev->hwdev, TCAM_PKT_VRRP);
+
+ if (filter_info->type_mask &
+ (1 << HINIC_PKT_TYPE_FIND_ID(PKT_LACP_TYPE)))
+ hinic_clear_fdir_tcam(nic_dev->hwdev, TCAM_PKT_LACP);
+
+ hinic_set_fdir_filter(nic_dev->hwdev, 0, 0, 0, false);
+}
+
static int
hinic_filter_info_init(struct hinic_5tuple_filter *filter,
struct hinic_filter_info *filter_info)
@@ -2238,8 +2272,114 @@ static int hinic_flow_destroy(struct rte_eth_dev *dev,
return ret;
}
+/* Remove all the n-tuple filters */
+static void hinic_clear_all_ntuple_filter(struct rte_eth_dev *dev)
+{
+ struct hinic_filter_info *filter_info =
+ HINIC_DEV_PRIVATE_TO_FILTER_INFO(dev->data->dev_private);
+ struct hinic_5tuple_filter *p_5tuple;
+
+ while ((p_5tuple = TAILQ_FIRST(&filter_info->fivetuple_list)))
+ hinic_remove_5tuple_filter(dev, p_5tuple);
+}
+
+/* Remove all the ether type filters */
+static void hinic_clear_all_ethertype_filter(struct rte_eth_dev *dev)
+{
+ struct hinic_nic_dev *nic_dev = HINIC_ETH_DEV_TO_PRIVATE_NIC_DEV(dev);
+ struct hinic_filter_info *filter_info =
+ HINIC_DEV_PRIVATE_TO_FILTER_INFO(nic_dev);
+ int ret = 0;
+
+ if (filter_info->type_mask &
+ (1 << HINIC_PKT_TYPE_FIND_ID(PKT_LACP_TYPE))) {
+ hinic_ethertype_filter_remove(filter_info,
+ HINIC_PKT_TYPE_FIND_ID(PKT_LACP_TYPE));
+ ret = hinic_set_fdir_filter(nic_dev->hwdev, PKT_LACP_TYPE,
+ filter_info->qid, false, true);
+
+ (void)hinic_clear_fdir_tcam(nic_dev->hwdev, TCAM_PKT_LACP);
+ }
+
+ if (filter_info->type_mask &
+ (1 << HINIC_PKT_TYPE_FIND_ID(PKT_ARP_TYPE))) {
+ hinic_ethertype_filter_remove(filter_info,
+ HINIC_PKT_TYPE_FIND_ID(PKT_ARP_TYPE));
+ ret = hinic_set_fdir_filter(nic_dev->hwdev, PKT_ARP_TYPE,
+ filter_info->qid, false, true);
+ }
+
+ if (ret)
+ PMD_DRV_LOG(ERR, "Clear ethertype failed, filter type: 0x%x",
+ filter_info->pkt_type);
+}
+
+/* Remove all the ether type filters */
+static void hinic_clear_all_fdir_filter(struct rte_eth_dev *dev)
+{
+ struct hinic_nic_dev *nic_dev = HINIC_ETH_DEV_TO_PRIVATE_NIC_DEV(dev);
+
+ (void)hinic_set_fdir_filter(nic_dev->hwdev, 0, 0, 0, false);
+}
+
+static void hinic_filterlist_flush(struct rte_eth_dev *dev)
+{
+ struct hinic_ntuple_filter_ele *ntuple_filter_ptr;
+ struct hinic_ethertype_filter_ele *ethertype_filter_ptr;
+ struct hinic_fdir_rule_ele *fdir_rule_ptr;
+ struct hinic_flow_mem *hinic_flow_mem_ptr;
+ struct hinic_nic_dev *nic_dev = HINIC_ETH_DEV_TO_PRIVATE_NIC_DEV(dev);
+
+ while ((ntuple_filter_ptr =
+ TAILQ_FIRST(&nic_dev->filter_ntuple_list))) {
+ TAILQ_REMOVE(&nic_dev->filter_ntuple_list, ntuple_filter_ptr,
+ entries);
+ rte_free(ntuple_filter_ptr);
+ }
+
+ while ((ethertype_filter_ptr =
+ TAILQ_FIRST(&nic_dev->filter_ethertype_list))) {
+ TAILQ_REMOVE(&nic_dev->filter_ethertype_list,
+ ethertype_filter_ptr,
+ entries);
+ rte_free(ethertype_filter_ptr);
+ }
+
+ while ((fdir_rule_ptr =
+ TAILQ_FIRST(&nic_dev->filter_fdir_rule_list))) {
+ TAILQ_REMOVE(&nic_dev->filter_fdir_rule_list, fdir_rule_ptr,
+ entries);
+ rte_free(fdir_rule_ptr);
+ }
+
+ while ((hinic_flow_mem_ptr =
+ TAILQ_FIRST(&nic_dev->hinic_flow_list))) {
+ TAILQ_REMOVE(&nic_dev->hinic_flow_list, hinic_flow_mem_ptr,
+ entries);
+ rte_free(hinic_flow_mem_ptr->flow);
+ rte_free(hinic_flow_mem_ptr);
+ }
+}
+
+/* Destroy all flow rules associated with a port on hinic. */
+static int hinic_flow_flush(struct rte_eth_dev *dev,
+ __rte_unused struct rte_flow_error *error)
+{
+ struct hinic_nic_dev *nic_dev = HINIC_ETH_DEV_TO_PRIVATE_NIC_DEV(dev);
+
+ hinic_clear_all_ntuple_filter(dev);
+ hinic_clear_all_ethertype_filter(dev);
+ hinic_clear_all_fdir_filter(dev);
+ hinic_filterlist_flush(dev);
+
+ PMD_DRV_LOG(INFO, "Flush flow succeed, func_id: 0x%x",
+ hinic_global_func_id(nic_dev->hwdev));
+ return 0;
+}
+
const struct rte_flow_ops hinic_flow_ops = {
.validate = hinic_flow_validate,
.create = hinic_flow_create,
.destroy = hinic_flow_destroy,
+ .flush = hinic_flow_flush,
};
--
1.8.3.1
^ permalink raw reply [flat|nested] 32+ messages in thread
* [dpdk-dev] [PATCH v2 12/17] net/hinic: set link down and up
2019-09-25 14:26 [dpdk-dev] [PATCH v2 00/17] Add advanced features for Huawei hinic pmd Xiaoyun wang
` (11 preceding siblings ...)
2019-09-25 14:30 ` [dpdk-dev] [PATCH v2 11/17] net/hinic: flush " Xiaoyun wang
@ 2019-09-25 14:30 ` Xiaoyun wang
2019-09-26 18:47 ` Ferruh Yigit
2019-09-25 14:30 ` [dpdk-dev] [PATCH v2 13/17] net/hinic: support inner L3 checksum offload Xiaoyun wang
` (5 subsequent siblings)
18 siblings, 1 reply; 32+ messages in thread
From: Xiaoyun wang @ 2019-09-25 14:30 UTC (permalink / raw)
To: ferruh.yigit
Cc: dev, xuanziyang2, shahar.belkar, luoxianjun, tanya.brokhman,
zhouguoyang, wulike1, Xiaoyun wang
This patch supports setting link down and up,
Also adds support for geting firmware version.
Signed-off-by: Xiaoyun wang <cloud.wangxiaoyun@huawei.com>
---
drivers/net/hinic/base/hinic_pmd_cmd.h | 4 ++
drivers/net/hinic/base/hinic_pmd_niccfg.c | 78 ++++++++++++++++++++++++++++
drivers/net/hinic/base/hinic_pmd_niccfg.h | 22 ++++++++
drivers/net/hinic/hinic_pmd_ethdev.c | 85 +++++++++++++++++++++++++++++++
4 files changed, 189 insertions(+)
diff --git a/drivers/net/hinic/base/hinic_pmd_cmd.h b/drivers/net/hinic/base/hinic_pmd_cmd.h
index 6b3dcf3..e0633bd 100644
--- a/drivers/net/hinic/base/hinic_pmd_cmd.h
+++ b/drivers/net/hinic/base/hinic_pmd_cmd.h
@@ -85,6 +85,8 @@ enum hinic_port_cmd {
HINIC_PORT_CMD_GET_JUMBO_FRAME_SIZE = 0x4a,
HINIC_PORT_CMD_SET_JUMBO_FRAME_SIZE,
+ HINIC_PORT_CMD_GET_MGMT_VERSION = 0x58,
+
HINIC_PORT_CMD_GET_PORT_TYPE = 0x5b,
HINIC_PORT_CMD_GET_VPORT_ENABLE = 0x5c,
@@ -119,6 +121,8 @@ enum hinic_port_cmd {
HINIC_PORT_CMD_SET_IPSU_MAC = 0xcb,
HINIC_PORT_CMD_GET_IPSU_MAC = 0xcc,
+ HINIC_PORT_CMD_SET_XSFP_STATUS = 0xD4,
+
HINIC_PORT_CMD_GET_LINK_MODE = 0xD9,
HINIC_PORT_CMD_SET_SPEED = 0xDA,
HINIC_PORT_CMD_SET_AUTONEG = 0xDB,
diff --git a/drivers/net/hinic/base/hinic_pmd_niccfg.c b/drivers/net/hinic/base/hinic_pmd_niccfg.c
index 53d981b..2dc431e 100644
--- a/drivers/net/hinic/base/hinic_pmd_niccfg.c
+++ b/drivers/net/hinic/base/hinic_pmd_niccfg.c
@@ -1293,6 +1293,46 @@ int hinic_set_rx_mode(void *hwdev, u32 enable)
return 0;
}
+/**
+ * hinic_get_mgmt_version - Get mgmt module version from chip.
+ *
+ * @param hwdev
+ * The hardware interface of a nic device.
+ * @param fw
+ * Firmware version.
+ *
+ * @return
+ * 0 on success.
+ * negative error value otherwise.
+ */
+int hinic_get_mgmt_version(void *hwdev, char *fw)
+{
+ struct hinic_version_info fw_ver;
+ u16 out_size = sizeof(fw_ver);
+ int err;
+
+ if (!hwdev || !fw) {
+ PMD_DRV_LOG(ERR, "Hwdev or fw is NULL");
+ return -EINVAL;
+ }
+
+ memset(&fw_ver, 0, sizeof(fw_ver));
+ fw_ver.mgmt_msg_head.resp_aeq_num = HINIC_AEQ1;
+
+ err = l2nic_msg_to_mgmt_sync(hwdev, HINIC_PORT_CMD_GET_MGMT_VERSION,
+ &fw_ver, sizeof(fw_ver), &fw_ver,
+ &out_size);
+ if (err || !out_size || fw_ver.mgmt_msg_head.status) {
+ PMD_DRV_LOG(ERR, "Failed to get mgmt version, err: %d, status: 0x%x, out size: 0x%x\n",
+ err, fw_ver.mgmt_msg_head.status, out_size);
+ return -EINVAL;
+ }
+
+ snprintf(fw, HINIC_MGMT_VERSION_MAX_LEN, "%s", fw_ver.ver);
+
+ return 0;
+}
+
int hinic_set_rx_csum_offload(void *hwdev, u32 en)
{
struct hinic_checksum_offload rx_csum_cfg;
@@ -1613,6 +1653,44 @@ int hinic_get_link_mode(void *hwdev, u32 *supported, u32 *advertised)
}
/**
+ * hinic_set_xsfp_tx_status - Enable or disable the fiber in
+ * tx direction when set link up or down.
+ *
+ * @param hwdev
+ * The hardware interface of a nic device.
+ * @param enable
+ * Enable or Disable.
+ *
+ * @return
+ * 0 on success.
+ * negative error value otherwise.
+ */
+int hinic_set_xsfp_tx_status(void *hwdev, bool enable)
+{
+ struct hinic_set_xsfp_status xsfp_status;
+ u16 out_size = sizeof(struct hinic_set_xsfp_status);
+ int err;
+
+ memset(&xsfp_status, 0, sizeof(xsfp_status));
+ xsfp_status.mgmt_msg_head.resp_aeq_num = HINIC_AEQ1;
+ xsfp_status.port_id = hinic_global_func_id(hwdev);
+ xsfp_status.xsfp_tx_dis = ((enable == 0) ? 1 : 0);
+
+ err = l2nic_msg_to_mgmt_sync(hwdev, HINIC_PORT_CMD_SET_XSFP_STATUS,
+ &xsfp_status, sizeof(struct hinic_set_xsfp_status),
+ &xsfp_status, &out_size);
+ if (err || !out_size || xsfp_status.mgmt_msg_head.status) {
+ PMD_DRV_LOG(ERR,
+ "Failed to %s port xsfp status, err: %d, status: 0x%x, out size: 0x%x\n",
+ enable ? "Disable" : "Enable", err,
+ xsfp_status.mgmt_msg_head.status, out_size);
+ return -EFAULT;
+ }
+
+ return 0;
+}
+
+/**
* hinic_flush_qp_res - Flush tx && rx chip resources in case of set vport
* fake failed when device start.
*
diff --git a/drivers/net/hinic/base/hinic_pmd_niccfg.h b/drivers/net/hinic/base/hinic_pmd_niccfg.h
index e8ce332..b9e037e 100644
--- a/drivers/net/hinic/base/hinic_pmd_niccfg.h
+++ b/drivers/net/hinic/base/hinic_pmd_niccfg.h
@@ -471,6 +471,17 @@ struct hinic_rx_mode_config {
u32 rx_mode;
};
+#define HINIC_MGMT_VERSION_MAX_LEN 32
+#define HINIC_COMPILE_TIME_LEN 20
+#define HINIC_FW_VERSION_NAME 16
+
+struct hinic_version_info {
+ struct hinic_mgmt_msg_head mgmt_msg_head;
+
+ u8 ver[HINIC_FW_VERSION_NAME];
+ u8 time[HINIC_COMPILE_TIME_LEN];
+};
+
/* rss */
struct nic_rss_indirect_tbl {
u32 group_index;
@@ -578,6 +589,13 @@ struct hinic_link_mode_cmd {
u16 advertised;
};
+struct hinic_set_xsfp_status {
+ struct hinic_mgmt_msg_head mgmt_msg_head;
+
+ u32 port_id;
+ u32 xsfp_tx_dis; /* 0: tx enable; 1: tx disable */
+};
+
struct hinic_clear_qp_resource {
struct hinic_mgmt_msg_head mgmt_msg_head;
@@ -800,6 +818,8 @@ int hinic_get_rss_type(void *hwdev, u32 tmpl_idx,
int hinic_set_rx_mode(void *hwdev, u32 enable);
+int hinic_get_mgmt_version(void *hwdev, char *fw);
+
int hinic_set_rx_csum_offload(void *hwdev, u32 en);
int hinic_set_link_status_follow(void *hwdev,
@@ -807,6 +827,8 @@ int hinic_set_link_status_follow(void *hwdev,
int hinic_get_link_mode(void *hwdev, u32 *supported, u32 *advertised);
+int hinic_set_xsfp_tx_status(void *hwdev, bool enable);
+
int hinic_flush_qp_res(void *hwdev);
int hinic_init_function_table(void *hwdev, u16 rx_buf_sz);
diff --git a/drivers/net/hinic/hinic_pmd_ethdev.c b/drivers/net/hinic/hinic_pmd_ethdev.c
index 2c329d1..8ea5998 100644
--- a/drivers/net/hinic/hinic_pmd_ethdev.c
+++ b/drivers/net/hinic/hinic_pmd_ethdev.c
@@ -750,6 +750,27 @@ static void hinic_get_speed_capa(struct rte_eth_dev *dev, uint32_t *speed_capa)
return 0;
}
+static int hinic_fw_version_get(struct rte_eth_dev *dev, char *fw_version,
+ size_t fw_size)
+{
+ struct hinic_nic_dev *nic_dev = HINIC_ETH_DEV_TO_PRIVATE_NIC_DEV(dev);
+ char fw_ver[HINIC_MGMT_VERSION_MAX_LEN] = {0};
+ int err;
+
+ err = hinic_get_mgmt_version(nic_dev->hwdev, fw_ver);
+ if (err) {
+ PMD_DRV_LOG(ERR, "Failed to get fw version\n");
+ return -EINVAL;
+ }
+
+ if (fw_size < strlen(fw_ver) + 1)
+ return (strlen(fw_ver) + 1);
+
+ snprintf(fw_version, fw_size, "%s", fw_ver);
+
+ return 0;
+}
+
static int hinic_config_rx_mode(struct hinic_nic_dev *nic_dev, u32 rx_mode_ctrl)
{
int err;
@@ -881,6 +902,66 @@ static int hinic_link_update(struct rte_eth_dev *dev, int wait_to_complete)
}
/**
+ * DPDK callback to bring the link UP.
+ *
+ * @param dev
+ * Pointer to Ethernet device structure.
+ *
+ * @return
+ * 0 on success, negative errno value on failure.
+ */
+static int hinic_dev_set_link_up(struct rte_eth_dev *dev)
+{
+ struct hinic_nic_dev *nic_dev = HINIC_ETH_DEV_TO_PRIVATE_NIC_DEV(dev);
+ int ret;
+
+ ret = hinic_set_xsfp_tx_status(nic_dev->hwdev, true);
+ if (ret) {
+ PMD_DRV_LOG(ERR, "Enable port tx xsfp failed, dev_name: %s, port_id: %d",
+ nic_dev->proc_dev_name, dev->data->port_id);
+ return ret;
+ }
+
+ /* link status follow phy port status, up will open pma */
+ ret = hinic_set_port_enable(nic_dev->hwdev, true);
+ if (ret)
+ PMD_DRV_LOG(ERR, "Set mac link up failed, dev_name: %s, port_id: %d",
+ nic_dev->proc_dev_name, dev->data->port_id);
+
+ return ret;
+}
+
+/**
+ * DPDK callback to bring the link DOWN.
+ *
+ * @param dev
+ * Pointer to Ethernet device structure.
+ *
+ * @return
+ * 0 on success, negative errno value on failure.
+ */
+static int hinic_dev_set_link_down(struct rte_eth_dev *dev)
+{
+ struct hinic_nic_dev *nic_dev = HINIC_ETH_DEV_TO_PRIVATE_NIC_DEV(dev);
+ int ret;
+
+ ret = hinic_set_xsfp_tx_status(nic_dev->hwdev, false);
+ if (ret) {
+ PMD_DRV_LOG(ERR, "Disable port tx xsfp failed, dev_name: %s, port_id: %d",
+ nic_dev->proc_dev_name, dev->data->port_id);
+ return ret;
+ }
+
+ /* link status follow phy port status, up will close pma */
+ ret = hinic_set_port_enable(nic_dev->hwdev, false);
+ if (ret)
+ PMD_DRV_LOG(ERR, "Set mac link down failed, dev_name: %s, port_id: %d",
+ nic_dev->proc_dev_name, dev->data->port_id);
+
+ return ret;
+}
+
+/**
* DPDK callback to start the device.
*
* @param dev
@@ -2751,9 +2832,12 @@ static void hinic_dev_close(struct rte_eth_dev *dev)
static const struct eth_dev_ops hinic_pmd_ops = {
.dev_configure = hinic_dev_configure,
.dev_infos_get = hinic_dev_infos_get,
+ .fw_version_get = hinic_fw_version_get,
.rx_queue_setup = hinic_rx_queue_setup,
.tx_queue_setup = hinic_tx_queue_setup,
.dev_start = hinic_dev_start,
+ .dev_set_link_up = hinic_dev_set_link_up,
+ .dev_set_link_down = hinic_dev_set_link_down,
.link_update = hinic_link_update,
.rx_queue_release = hinic_rx_queue_release,
.tx_queue_release = hinic_tx_queue_release,
@@ -2785,6 +2869,7 @@ static void hinic_dev_close(struct rte_eth_dev *dev)
static const struct eth_dev_ops hinic_pmd_vf_ops = {
.dev_configure = hinic_dev_configure,
.dev_infos_get = hinic_dev_infos_get,
+ .fw_version_get = hinic_fw_version_get,
.rx_queue_setup = hinic_rx_queue_setup,
.tx_queue_setup = hinic_tx_queue_setup,
.dev_start = hinic_dev_start,
--
1.8.3.1
^ permalink raw reply [flat|nested] 32+ messages in thread
* [dpdk-dev] [PATCH v2 13/17] net/hinic: support inner L3 checksum offload
2019-09-25 14:26 [dpdk-dev] [PATCH v2 00/17] Add advanced features for Huawei hinic pmd Xiaoyun wang
` (12 preceding siblings ...)
2019-09-25 14:30 ` [dpdk-dev] [PATCH v2 12/17] net/hinic: set link down and up Xiaoyun wang
@ 2019-09-25 14:30 ` Xiaoyun wang
2019-09-25 14:30 ` [dpdk-dev] [PATCH v2 14/17] net/hinic: support LRO offload Xiaoyun wang
` (4 subsequent siblings)
18 siblings, 0 replies; 32+ messages in thread
From: Xiaoyun wang @ 2019-09-25 14:30 UTC (permalink / raw)
To: ferruh.yigit
Cc: dev, xuanziyang2, shahar.belkar, luoxianjun, tanya.brokhman,
zhouguoyang, wulike1, Xiaoyun wang
This patch supports inner L3 checksum offload,
modifies rx checksum offload.
Signed-off-by: Xiaoyun wang <cloud.wangxiaoyun@huawei.com>
---
drivers/net/hinic/hinic_pmd_ethdev.h | 1 +
drivers/net/hinic/hinic_pmd_rx.c | 10 +-
drivers/net/hinic/hinic_pmd_tx.c | 190 ++++++++++++++++++++++-------------
3 files changed, 127 insertions(+), 74 deletions(-)
diff --git a/drivers/net/hinic/hinic_pmd_ethdev.h b/drivers/net/hinic/hinic_pmd_ethdev.h
index dd96667..3e3f3b3 100644
--- a/drivers/net/hinic/hinic_pmd_ethdev.h
+++ b/drivers/net/hinic/hinic_pmd_ethdev.h
@@ -178,6 +178,7 @@ struct hinic_nic_dev {
* vf: the same with associate pf
*/
u32 default_cos;
+ u32 rx_csum_en;
struct hinic_filter_info filter;
struct hinic_ntuple_filter_list filter_ntuple_list;
diff --git a/drivers/net/hinic/hinic_pmd_rx.c b/drivers/net/hinic/hinic_pmd_rx.c
index 08e02ae..37b4f5c 100644
--- a/drivers/net/hinic/hinic_pmd_rx.c
+++ b/drivers/net/hinic/hinic_pmd_rx.c
@@ -658,7 +658,6 @@ int hinic_rx_configure(struct rte_eth_dev *dev)
struct hinic_nic_dev *nic_dev = HINIC_ETH_DEV_TO_PRIVATE_NIC_DEV(dev);
struct rte_eth_rss_conf rss_conf =
dev->data->dev_conf.rx_adv_conf.rss_conf;
- u32 csum_en = 0;
int err;
if (nic_dev->flags & ETH_MQ_RX_RSS_FLAG) {
@@ -678,9 +677,10 @@ int hinic_rx_configure(struct rte_eth_dev *dev)
/* Enable both L3/L4 rx checksum offload */
if (dev->data->dev_conf.rxmode.offloads & DEV_RX_OFFLOAD_CHECKSUM)
- csum_en = HINIC_RX_CSUM_OFFLOAD_EN;
+ nic_dev->rx_csum_en = HINIC_RX_CSUM_OFFLOAD_EN;
- err = hinic_set_rx_csum_offload(nic_dev->hwdev, csum_en);
+ err = hinic_set_rx_csum_offload(nic_dev->hwdev,
+ HINIC_RX_CSUM_OFFLOAD_EN);
if (err)
goto rx_csum_ofl_err;
@@ -781,6 +781,10 @@ static inline uint64_t hinic_rx_csum(uint32_t status, struct hinic_rxq *rxq)
{
uint32_t checksum_err;
uint64_t flags;
+ struct hinic_nic_dev *nic_dev = rxq->nic_dev;
+
+ if (unlikely(!(nic_dev->rx_csum_en & HINIC_RX_CSUM_OFFLOAD_EN)))
+ return PKT_RX_IP_CKSUM_UNKNOWN;
/* most case checksum is ok */
checksum_err = HINIC_GET_RX_CSUM_ERR(status);
diff --git a/drivers/net/hinic/hinic_pmd_tx.c b/drivers/net/hinic/hinic_pmd_tx.c
index 0ef7add..26f481f 100644
--- a/drivers/net/hinic/hinic_pmd_tx.c
+++ b/drivers/net/hinic/hinic_pmd_tx.c
@@ -20,6 +20,9 @@
#include "hinic_pmd_tx.h"
/* packet header and tx offload info */
+#define ETHER_LEN_NO_VLAN 14
+#define ETHER_LEN_WITH_VLAN 18
+#define HEADER_LEN_OFFSET 2
#define VXLANLEN 8
#define MAX_PLD_OFFSET 221
#define MAX_SINGLE_SGE_SIZE 65536
@@ -34,6 +37,9 @@
#define HINIC_TSO_PKT_MAX_SGE 127 /* tso max sge 127 */
#define HINIC_TSO_SEG_NUM_INVALID(num) ((num) > HINIC_TSO_PKT_MAX_SGE)
+#define HINIC_TX_OUTER_CHECKSUM_FLAG_SET 1
+#define HINIC_TX_OUTER_CHECKSUM_FLAG_NO_SET 0
+
/* sizeof(struct hinic_sq_bufdesc) == 16, shift 4 */
#define HINIC_BUF_DESC_SIZE(nr_descs) (SIZE_8BYTES(((u32)nr_descs) << 4))
@@ -476,16 +482,16 @@ static inline bool hinic_is_tso_sge_valid(struct rte_mbuf *mbuf,
hinic_set_l4_csum_info(struct hinic_sq_task *task,
u32 *queue_info, struct hinic_tx_offload_info *poff_info)
{
- u32 tcp_udp_cs, sctp;
+ u32 tcp_udp_cs, sctp = 0;
u16 l2hdr_len;
- sctp = 0;
if (unlikely(poff_info->inner_l4_type == SCTP_OFFLOAD_ENABLE))
sctp = 1;
tcp_udp_cs = poff_info->inner_l4_tcp_udp;
- if (poff_info->tunnel_type == TUNNEL_UDP_NO_CSUM) {
+ if (poff_info->tunnel_type == TUNNEL_UDP_CSUM ||
+ poff_info->tunnel_type == TUNNEL_UDP_NO_CSUM) {
l2hdr_len = poff_info->outer_l2_len;
task->pkt_info2 |=
@@ -665,50 +671,6 @@ static inline void hinic_xmit_mbuf_cleanup(struct hinic_txq *txq)
return (struct hinic_sq_wqe *)WQ_WQE_ADDR(wq, cur_pi);
}
-static inline int
-hinic_validate_tx_offload(const struct rte_mbuf *m)
-{
- uint64_t ol_flags = m->ol_flags;
- uint64_t inner_l3_offset = m->l2_len;
-
- /* just support vxlan offload */
- if ((ol_flags & PKT_TX_TUNNEL_MASK) &&
- !(ol_flags & PKT_TX_TUNNEL_VXLAN))
- return -ENOTSUP;
-
- if (ol_flags & PKT_TX_OUTER_IP_CKSUM)
- inner_l3_offset += m->outer_l2_len + m->outer_l3_len;
-
- /* Headers are fragmented */
- if (rte_pktmbuf_data_len(m) < inner_l3_offset + m->l3_len + m->l4_len)
- return -ENOTSUP;
-
- /* IP checksum can be counted only for IPv4 packet */
- if ((ol_flags & PKT_TX_IP_CKSUM) && (ol_flags & PKT_TX_IPV6))
- return -EINVAL;
-
- /* IP type not set when required */
- if (ol_flags & (PKT_TX_L4_MASK | PKT_TX_TCP_SEG)) {
- if (!(ol_flags & (PKT_TX_IPV4 | PKT_TX_IPV6)))
- return -EINVAL;
- }
-
- /* Check requirements for TSO packet */
- if (ol_flags & PKT_TX_TCP_SEG) {
- if (m->tso_segsz == 0 ||
- ((ol_flags & PKT_TX_IPV4) &&
- !(ol_flags & PKT_TX_IP_CKSUM)))
- return -EINVAL;
- }
-
- /* PKT_TX_OUTER_IP_CKSUM set for non outer IPv4 packet. */
- if ((ol_flags & PKT_TX_OUTER_IP_CKSUM) &&
- !(ol_flags & PKT_TX_OUTER_IPV4))
- return -EINVAL;
-
- return 0;
-}
-
static inline uint16_t
hinic_ipv4_phdr_cksum(const struct rte_ipv4_hdr *ipv4_hdr, uint64_t ol_flags)
{
@@ -760,6 +722,65 @@ static inline void hinic_xmit_mbuf_cleanup(struct hinic_txq *txq)
return __rte_raw_cksum_reduce(sum);
}
+static inline void
+hinic_get_pld_offset(struct rte_mbuf *m, struct hinic_tx_offload_info *off_info,
+ int outer_cs_flag)
+{
+ uint64_t ol_flags = m->ol_flags;
+
+ if (outer_cs_flag == 1) {
+ if ((ol_flags & PKT_TX_UDP_CKSUM) == PKT_TX_UDP_CKSUM) {
+ off_info->payload_offset = m->outer_l2_len +
+ m->outer_l3_len + m->l2_len + m->l3_len;
+ } else if ((ol_flags & PKT_TX_TCP_CKSUM) ||
+ (ol_flags & PKT_TX_TCP_SEG)) {
+ off_info->payload_offset = m->outer_l2_len +
+ m->outer_l3_len + m->l2_len +
+ m->l3_len + m->l4_len;
+ }
+ } else {
+ if ((ol_flags & PKT_TX_UDP_CKSUM) == PKT_TX_UDP_CKSUM) {
+ off_info->payload_offset = m->l2_len + m->l3_len;
+ } else if ((ol_flags & PKT_TX_TCP_CKSUM) ||
+ (ol_flags & PKT_TX_TCP_SEG)) {
+ off_info->payload_offset = m->l2_len + m->l3_len +
+ m->l4_len;
+ }
+ }
+}
+
+static inline void
+hinic_analyze_tx_info(struct rte_mbuf *mbuf,
+ struct hinic_tx_offload_info *off_info)
+{
+ struct rte_ether_hdr *eth_hdr;
+ struct rte_vlan_hdr *vlan_hdr;
+ struct rte_ipv4_hdr *ip4h;
+ u16 pkt_type;
+ u8 *hdr;
+
+ hdr = (u8 *)rte_pktmbuf_mtod(mbuf, u8*);
+ eth_hdr = (struct rte_ether_hdr *)hdr;
+ pkt_type = rte_be_to_cpu_16(eth_hdr->ether_type);
+
+ if (pkt_type == RTE_ETHER_TYPE_VLAN) {
+ off_info->outer_l2_len = ETHER_LEN_WITH_VLAN;
+ vlan_hdr = (struct rte_vlan_hdr *)(hdr + 1);
+ pkt_type = rte_be_to_cpu_16(vlan_hdr->eth_proto);
+ } else {
+ off_info->outer_l2_len = ETHER_LEN_NO_VLAN;
+ }
+
+ if (pkt_type == RTE_ETHER_TYPE_IPV4) {
+ ip4h = (struct rte_ipv4_hdr *)(hdr + off_info->outer_l2_len);
+ off_info->outer_l3_len = (ip4h->version_ihl & 0xf) <<
+ HEADER_LEN_OFFSET;
+ } else if (pkt_type == RTE_ETHER_TYPE_IPV6) {
+ /* not support ipv6 extension header */
+ off_info->outer_l3_len = sizeof(struct rte_ipv6_hdr);
+ }
+}
+
static inline int
hinic_tx_offload_pkt_prepare(struct rte_mbuf *m,
struct hinic_tx_offload_info *off_info)
@@ -771,42 +792,66 @@ static inline void hinic_xmit_mbuf_cleanup(struct hinic_txq *txq)
struct rte_ether_hdr *eth_hdr;
struct rte_vlan_hdr *vlan_hdr;
u16 eth_type = 0;
- uint64_t inner_l3_offset = m->l2_len;
+ uint64_t inner_l3_offset;
uint64_t ol_flags = m->ol_flags;
- /* Does packet set any of available offloads */
+ /* Check if the packets set available offload flags */
if (!(ol_flags & HINIC_TX_CKSUM_OFFLOAD_MASK))
return 0;
- if (unlikely(hinic_validate_tx_offload(m)))
+ /* Support only vxlan offload */
+ if ((ol_flags & PKT_TX_TUNNEL_MASK) &&
+ !(ol_flags & PKT_TX_TUNNEL_VXLAN))
+ return -ENOTSUP;
+
+#ifdef RTE_LIBRTE_ETHDEV_DEBUG
+ if (rte_validate_tx_offload(m) != 0)
return -EINVAL;
+#endif
- if ((ol_flags & PKT_TX_OUTER_IP_CKSUM) ||
- (ol_flags & PKT_TX_OUTER_IPV6) ||
- (ol_flags & PKT_TX_TUNNEL_VXLAN)) {
- inner_l3_offset += m->outer_l2_len + m->outer_l3_len;
- off_info->outer_l2_len = m->outer_l2_len;
- off_info->outer_l3_len = m->outer_l3_len;
- /* just support vxlan tunneling pkt */
- off_info->inner_l2_len = m->l2_len - VXLANLEN -
- sizeof(struct rte_udp_hdr);
- off_info->inner_l3_len = m->l3_len;
- off_info->inner_l4_len = m->l4_len;
- off_info->tunnel_length = m->l2_len;
- off_info->payload_offset = m->outer_l2_len +
- m->outer_l3_len + m->l2_len + m->l3_len;
- off_info->tunnel_type = TUNNEL_UDP_NO_CSUM;
+ if (ol_flags & PKT_TX_TUNNEL_VXLAN) {
+ if ((ol_flags & PKT_TX_OUTER_IP_CKSUM) ||
+ (ol_flags & PKT_TX_OUTER_IPV6)) {
+ inner_l3_offset = m->l2_len + m->outer_l2_len +
+ m->outer_l3_len;
+ off_info->outer_l2_len = m->outer_l2_len;
+ off_info->outer_l3_len = m->outer_l3_len;
+ /* just support vxlan tunneling pkt */
+ off_info->inner_l2_len = m->l2_len - VXLANLEN -
+ sizeof(*udp_hdr);
+ off_info->inner_l3_len = m->l3_len;
+ off_info->inner_l4_len = m->l4_len;
+ off_info->tunnel_length = m->l2_len;
+ off_info->tunnel_type = TUNNEL_UDP_NO_CSUM;
+
+ hinic_get_pld_offset(m, off_info,
+ HINIC_TX_OUTER_CHECKSUM_FLAG_SET);
+ } else {
+ inner_l3_offset = m->l2_len;
+ hinic_analyze_tx_info(m, off_info);
+ /* just support vxlan tunneling pkt */
+ off_info->inner_l2_len = m->l2_len - VXLANLEN -
+ sizeof(*udp_hdr) - off_info->outer_l2_len -
+ off_info->outer_l3_len;
+ off_info->inner_l3_len = m->l3_len;
+ off_info->inner_l4_len = m->l4_len;
+ off_info->tunnel_length = m->l2_len -
+ off_info->outer_l2_len - off_info->outer_l3_len;
+ off_info->tunnel_type = TUNNEL_UDP_NO_CSUM;
+
+ hinic_get_pld_offset(m, off_info,
+ HINIC_TX_OUTER_CHECKSUM_FLAG_NO_SET);
+ }
} else {
+ inner_l3_offset = m->l2_len;
off_info->inner_l2_len = m->l2_len;
off_info->inner_l3_len = m->l3_len;
off_info->inner_l4_len = m->l4_len;
off_info->tunnel_type = NOT_TUNNEL;
- off_info->payload_offset = m->l2_len + m->l3_len;
- }
- if (((ol_flags & PKT_TX_L4_MASK) != PKT_TX_SCTP_CKSUM) &&
- ((ol_flags & PKT_TX_L4_MASK) != PKT_TX_UDP_CKSUM))
- off_info->payload_offset += m->l4_len;
+ hinic_get_pld_offset(m, off_info,
+ HINIC_TX_OUTER_CHECKSUM_FLAG_NO_SET);
+ }
/* invalid udp or tcp header */
if (unlikely(off_info->payload_offset > MAX_PLD_OFFSET))
@@ -855,6 +900,10 @@ static inline void hinic_xmit_mbuf_cleanup(struct hinic_txq *txq)
udp_hdr->dgram_cksum =
hinic_ipv6_phdr_cksum(ipv6_hdr, ol_flags);
}
+ } else if (ol_flags & PKT_TX_OUTER_IPV4) {
+ off_info->tunnel_type = TUNNEL_UDP_NO_CSUM;
+ off_info->inner_l4_tcp_udp = 1;
+ off_info->outer_l3_type = IPV4_PKT_NO_CHKSUM_OFFLOAD;
}
if (ol_flags & PKT_TX_IPV4)
@@ -892,7 +941,6 @@ static inline void hinic_xmit_mbuf_cleanup(struct hinic_txq *txq)
off_info->inner_l4_type = UDP_OFFLOAD_ENABLE;
off_info->inner_l4_tcp_udp = 1;
- off_info->inner_l4_len = sizeof(struct rte_udp_hdr);
} else if (((ol_flags & PKT_TX_L4_MASK) == PKT_TX_TCP_CKSUM) ||
(ol_flags & PKT_TX_TCP_SEG)) {
if (ol_flags & PKT_TX_IPV4) {
--
1.8.3.1
^ permalink raw reply [flat|nested] 32+ messages in thread
* [dpdk-dev] [PATCH v2 14/17] net/hinic: support LRO offload
2019-09-25 14:26 [dpdk-dev] [PATCH v2 00/17] Add advanced features for Huawei hinic pmd Xiaoyun wang
` (13 preceding siblings ...)
2019-09-25 14:30 ` [dpdk-dev] [PATCH v2 13/17] net/hinic: support inner L3 checksum offload Xiaoyun wang
@ 2019-09-25 14:30 ` Xiaoyun wang
2019-09-25 14:30 ` [dpdk-dev] [PATCH v2 15/17] net/hinic: add hinic PMD doc files Xiaoyun wang
` (3 subsequent siblings)
18 siblings, 0 replies; 32+ messages in thread
From: Xiaoyun wang @ 2019-09-25 14:30 UTC (permalink / raw)
To: ferruh.yigit
Cc: dev, xuanziyang2, shahar.belkar, luoxianjun, tanya.brokhman,
zhouguoyang, wulike1, Xiaoyun wang
This patch supports LRO offload.
Signed-off-by: Xiaoyun wang <cloud.wangxiaoyun@huawei.com>
---
drivers/net/hinic/hinic_pmd_ethdev.c | 24 ++++++++++++++++++------
1 file changed, 18 insertions(+), 6 deletions(-)
diff --git a/drivers/net/hinic/hinic_pmd_ethdev.c b/drivers/net/hinic/hinic_pmd_ethdev.c
index 8ea5998..6f3eb18 100644
--- a/drivers/net/hinic/hinic_pmd_ethdev.c
+++ b/drivers/net/hinic/hinic_pmd_ethdev.c
@@ -70,6 +70,9 @@
#define HINIC_PKTLEN_TO_MTU(pktlen) \
((pktlen) - (ETH_HLEN + ETH_CRC_LEN))
+/* lro numer limit for one packet */
+#define HINIC_LRO_WQE_NUM_DEFAULT 8
+
/* Driver-specific log messages type */
int hinic_logtype;
@@ -729,7 +732,8 @@ static void hinic_get_speed_capa(struct rte_eth_dev *dev, uint32_t *speed_capa)
DEV_RX_OFFLOAD_TCP_CKSUM |
DEV_RX_OFFLOAD_VLAN_FILTER |
DEV_RX_OFFLOAD_SCATTER |
- DEV_RX_OFFLOAD_JUMBO_FRAME;
+ DEV_RX_OFFLOAD_JUMBO_FRAME |
+ DEV_RX_OFFLOAD_TCP_LRO;
info->tx_queue_offload_capa = 0;
info->tx_offload_capa = DEV_TX_OFFLOAD_VLAN_INSERT |
@@ -790,6 +794,7 @@ static int hinic_rxtx_configure(struct rte_eth_dev *dev)
{
int err;
struct hinic_nic_dev *nic_dev = HINIC_ETH_DEV_TO_PRIVATE_NIC_DEV(dev);
+ bool lro_en;
/* rx configure, if rss enable, need to init default configuration */
err = hinic_rx_configure(dev);
@@ -806,6 +811,18 @@ static int hinic_rxtx_configure(struct rte_eth_dev *dev)
goto set_rx_mode_fail;
}
+ /* config lro */
+ lro_en = dev->data->dev_conf.rxmode.offloads & DEV_RX_OFFLOAD_TCP_LRO ?
+ true : false;
+
+ err = hinic_set_rx_lro(nic_dev->hwdev, lro_en, lro_en,
+ HINIC_LRO_WQE_NUM_DEFAULT);
+ if (err) {
+ PMD_DRV_LOG(ERR, "%s lro failed, err: %d",
+ lro_en ? "Enable" : "Disable", err);
+ goto set_rx_mode_fail;
+ }
+
return HINIC_OK;
set_rx_mode_fail:
@@ -2476,11 +2493,6 @@ static int hinic_set_default_hw_feature(struct hinic_nic_dev *nic_dev)
if (err)
return err;
- /* disable LRO */
- err = hinic_set_rx_lro(nic_dev->hwdev, 0, 0, (u8)0);
- if (err)
- return err;
-
/* Set pause enable, and up will disable pfc. */
err = hinic_set_default_pause_feature(nic_dev);
if (err)
--
1.8.3.1
^ permalink raw reply [flat|nested] 32+ messages in thread
* [dpdk-dev] [PATCH v2 15/17] net/hinic: add hinic PMD doc files
2019-09-25 14:26 [dpdk-dev] [PATCH v2 00/17] Add advanced features for Huawei hinic pmd Xiaoyun wang
` (14 preceding siblings ...)
2019-09-25 14:30 ` [dpdk-dev] [PATCH v2 14/17] net/hinic: support LRO offload Xiaoyun wang
@ 2019-09-25 14:30 ` Xiaoyun wang
2019-09-26 18:51 ` Ferruh Yigit
2019-09-25 14:30 ` [dpdk-dev] [PATCH v2 16/17] net/hinic/base: optimize aeq interfaces Xiaoyun wang
` (2 subsequent siblings)
18 siblings, 1 reply; 32+ messages in thread
From: Xiaoyun wang @ 2019-09-25 14:30 UTC (permalink / raw)
To: ferruh.yigit
Cc: dev, xuanziyang2, shahar.belkar, luoxianjun, tanya.brokhman,
zhouguoyang, wulike1, Xiaoyun wang
[-- Warning: decoded text below may be mangled, UTF-8 assumed --]
[-- Attachment #1: Type: text/plain; charset="UTF-8", Size: 5042 bytes --]
Add doc files about new features and modification.
Signed-off-by: Xiaoyun wang <cloud.wangxiaoyun@huawei.com>
---
doc/guides/nics/features/hinic.ini | 12 ++++++++-
doc/guides/nics/hinic.rst | 6 +++++
doc/guides/rel_notes/release_19_11.rst | 45 ++++++----------------------------
3 files changed, 25 insertions(+), 38 deletions(-)
diff --git a/doc/guides/nics/features/hinic.ini b/doc/guides/nics/features/hinic.ini
index fe063d6..dc02b4b 100644
--- a/doc/guides/nics/features/hinic.ini
+++ b/doc/guides/nics/features/hinic.ini
@@ -9,16 +9,22 @@ Link status = Y
Link status event = Y
Free Tx mbuf on demand = Y
Queue start/stop = Y
-Jumbo frame = N
+MTU update = Y
+Jumbo frame = Y
Scattered Rx = Y
TSO = Y
+LRO = Y
Promiscuous mode = Y
+Allmulticast mode = Y
Unicast MAC filter = Y
Multicast MAC filter = Y
RSS hash = Y
RSS key update = Y
RSS reta update = Y
Inner RSS = Y
+SR-IOV = Y
+VLAN filter = Y
+VLAN offload = Y
CRC offload = Y
L3 checksum offload = Y
L4 checksum offload = Y
@@ -27,6 +33,10 @@ Inner L4 checksum = Y
Basic stats = Y
Extended stats = Y
Stats per queue = Y
+Flow director = Y
+Flow control = Y
+FW version = Y
+Multiprocess aware = Y
Linux UIO = Y
Linux VFIO = Y
BSD nic_uio = N
diff --git a/doc/guides/nics/hinic.rst b/doc/guides/nics/hinic.rst
index c9329bc..f036fc5 100644
--- a/doc/guides/nics/hinic.rst
+++ b/doc/guides/nics/hinic.rst
@@ -24,6 +24,12 @@ Features
- Link state information
- Link flow control
- Scattered and gather for TX and RX
+- SR¨CIOV - Partially supported at this point, VFIO only
+- Allmulticast mode
+- Unicast MAC filter
+- Multicast MAC filter
+- FW version
+- Flow director
Prerequisites
-------------
diff --git a/doc/guides/rel_notes/release_19_11.rst b/doc/guides/rel_notes/release_19_11.rst
index 65361c4..6c6f27f 100644
--- a/doc/guides/rel_notes/release_19_11.rst
+++ b/doc/guides/rel_notes/release_19_11.rst
@@ -56,11 +56,15 @@ New Features
Also, make sure to start the actual text at the margin.
=========================================================
-* **Updated the Intel ice driver.**
+* **Updated the Huawei hinic driver.**
- Updated the Intel ice driver with new features and improvements, including:
+ Updated the Huawei hinic driver with new features and improvements, including:
- * Added support for device-specific DDP package loading.
+ * Enabled SR-IOV - Partially supported at this point, VFIO only.
+ * Supported VLAN filter and VLAN offload.
+ * Supported Unicast MAC filter and Multicast MAC filter.
+ * Supported FW version get.
+ * Supported Flow director for LACP, VRRP, BGP and so on.
Removed Items
-------------
@@ -99,30 +103,6 @@ API Changes
Also, make sure to start the actual text at the margin.
=========================================================
-* ethdev: changed ``rte_eth_dev_infos_get`` return value from ``void`` to
- ``int`` to provide a way to report various error conditions.
-
-* ethdev: changed ``rte_eth_promiscuous_enable`` and
- ``rte_eth_promiscuous_disable`` return value from ``void`` to ``int`` to
- provide a way to report various error conditions.
-
-* ethdev: changed ``rte_eth_allmulticast_enable`` and
- ``rte_eth_allmulticast_disable`` return value from ``void`` to ``int`` to
- provide a way to report various error conditions.
-
-* ethdev: changed ``rte_eth_dev_xstats_reset`` return value from ``void`` to
- ``int`` to provide a way to report various error conditions.
-
-* ethdev: changed ``rte_eth_link_get`` and ``rte_eth_link_get_nowait``
- return value from ``void`` to ``int`` to provide a way to report various
- error conditions.
-
-* ethdev: changed ``rte_eth_macaddr_get`` return value from ``void`` to
- ``int`` to provide a way to report various error conditions.
-
-* ethdev: changed ``rte_eth_dev_owner_delete`` return value from ``void`` to
- ``int`` to provide a way to report various error conditions.
-
ABI Changes
-----------
@@ -174,7 +154,7 @@ The libraries prepended with a plus sign were incremented in this version.
librte_distributor.so.1
librte_eal.so.11
librte_efd.so.1
- + librte_ethdev.so.13
+ librte_ethdev.so.12
librte_eventdev.so.7
librte_flow_classify.so.1
librte_gro.so.1
@@ -252,12 +232,3 @@ Tested Platforms
Also, make sure to start the actual text at the margin.
=========================================================
-* **Updated Mellanox mlx5 driver.**
-
- Updated Mellanox mlx5 driver with new features and improvements, including:
-
- * Added support for VLAN pop flow offload command.
- * Added support for VLAN push flow offload command.
- * Added support for VLAN set PCP offload command.
- * Added support for VLAN set VID offload command.
-
--
1.8.3.1
^ permalink raw reply [flat|nested] 32+ messages in thread
* [dpdk-dev] [PATCH v2 16/17] net/hinic/base: optimize aeq interfaces
2019-09-25 14:26 [dpdk-dev] [PATCH v2 00/17] Add advanced features for Huawei hinic pmd Xiaoyun wang
` (15 preceding siblings ...)
2019-09-25 14:30 ` [dpdk-dev] [PATCH v2 15/17] net/hinic: add hinic PMD doc files Xiaoyun wang
@ 2019-09-25 14:30 ` Xiaoyun wang
2019-09-25 14:30 ` [dpdk-dev] [PATCH v2 17/17] net/hinic: optimize tx&rx performance Xiaoyun wang
2019-09-26 18:51 ` [dpdk-dev] [PATCH v2 00/17] Add advanced features for Huawei hinic pmd Ferruh Yigit
18 siblings, 0 replies; 32+ messages in thread
From: Xiaoyun wang @ 2019-09-25 14:30 UTC (permalink / raw)
To: ferruh.yigit
Cc: dev, xuanziyang2, shahar.belkar, luoxianjun, tanya.brokhman,
zhouguoyang, wulike1, Xiaoyun wang
This patch deletes ceq interfaces that not needed,
fix aeq buges in some scenarios.
Signed-off-by: Xiaoyun wang <cloud.wangxiaoyun@huawei.com>
---
drivers/net/hinic/base/hinic_csr.h | 29 +---
drivers/net/hinic/base/hinic_pmd_cmd.h | 5 +-
drivers/net/hinic/base/hinic_pmd_eqs.c | 245 +++++++++------------------------
drivers/net/hinic/base/hinic_pmd_eqs.h | 5 +-
4 files changed, 67 insertions(+), 217 deletions(-)
diff --git a/drivers/net/hinic/base/hinic_csr.h b/drivers/net/hinic/base/hinic_csr.h
index b63e52b..2626f69 100644
--- a/drivers/net/hinic/base/hinic_csr.h
+++ b/drivers/net/hinic/base/hinic_csr.h
@@ -48,16 +48,12 @@
/* EQ registers */
#define HINIC_AEQ_MTT_OFF_BASE_ADDR 0x200
-#define HINIC_CEQ_MTT_OFF_BASE_ADDR 0x400
#define HINIC_EQ_MTT_OFF_STRIDE 0x40
#define HINIC_CSR_AEQ_MTT_OFF(id) \
(HINIC_AEQ_MTT_OFF_BASE_ADDR + (id) * HINIC_EQ_MTT_OFF_STRIDE)
-#define HINIC_CSR_CEQ_MTT_OFF(id) \
- (HINIC_CEQ_MTT_OFF_BASE_ADDR + (id) * HINIC_EQ_MTT_OFF_STRIDE)
-
#define HINIC_CSR_EQ_PAGE_OFF_STRIDE 8
#define HINIC_AEQ_HI_PHYS_ADDR_REG(q_id, pg_num) \
@@ -68,23 +64,11 @@
(HINIC_CSR_AEQ_MTT_OFF(q_id) + \
(pg_num) * HINIC_CSR_EQ_PAGE_OFF_STRIDE + 4)
-#define HINIC_CEQ_HI_PHYS_ADDR_REG(q_id, pg_num) \
- (HINIC_CSR_CEQ_MTT_OFF(q_id) + \
- (pg_num) * HINIC_CSR_EQ_PAGE_OFF_STRIDE)
-
-#define HINIC_CEQ_LO_PHYS_ADDR_REG(q_id, pg_num) \
- (HINIC_CSR_CEQ_MTT_OFF(q_id) + \
- (pg_num) * HINIC_CSR_EQ_PAGE_OFF_STRIDE + 4)
-
#define HINIC_EQ_HI_PHYS_ADDR_REG(type, q_id, pg_num) \
- ((u32)((type == HINIC_AEQ) ? \
- HINIC_AEQ_HI_PHYS_ADDR_REG(q_id, pg_num) : \
- HINIC_CEQ_HI_PHYS_ADDR_REG(q_id, pg_num)))
+ ((u32)(HINIC_AEQ_HI_PHYS_ADDR_REG(q_id, pg_num)))
#define HINIC_EQ_LO_PHYS_ADDR_REG(type, q_id, pg_num) \
- ((u32)((type == HINIC_AEQ) ? \
- HINIC_AEQ_LO_PHYS_ADDR_REG(q_id, pg_num) : \
- HINIC_CEQ_LO_PHYS_ADDR_REG(q_id, pg_num)))
+ ((u32)(HINIC_AEQ_LO_PHYS_ADDR_REG(q_id, pg_num)))
#define HINIC_AEQ_CTRL_0_ADDR_BASE 0xE00
#define HINIC_AEQ_CTRL_1_ADDR_BASE 0xE04
@@ -105,15 +89,6 @@
#define HINIC_CSR_AEQ_PROD_IDX_ADDR(idx) \
(HINIC_AEQ_CONS_IDX_1_ADDR_BASE + (idx) * HINIC_EQ_OFF_STRIDE)
-#define HINIC_CEQ_CONS_IDX_0_ADDR_BASE 0x1008
-#define HINIC_CEQ_CONS_IDX_1_ADDR_BASE 0x100C
-
-#define HINIC_CSR_CEQ_CONS_IDX_ADDR(idx) \
- (HINIC_CEQ_CONS_IDX_0_ADDR_BASE + (idx) * HINIC_EQ_OFF_STRIDE)
-
-#define HINIC_CSR_CEQ_PROD_IDX_ADDR(idx) \
- (HINIC_CEQ_CONS_IDX_1_ADDR_BASE + (idx) * HINIC_EQ_OFF_STRIDE)
-
/* API CMD registers */
#define HINIC_CSR_API_CMD_BASE 0xF000
diff --git a/drivers/net/hinic/base/hinic_pmd_cmd.h b/drivers/net/hinic/base/hinic_pmd_cmd.h
index e0633bd..c025851 100644
--- a/drivers/net/hinic/base/hinic_pmd_cmd.h
+++ b/drivers/net/hinic/base/hinic_pmd_cmd.h
@@ -5,10 +5,7 @@
#ifndef _HINIC_PORT_CMD_H_
#define _HINIC_PORT_CMD_H_
-enum hinic_eq_type {
- HINIC_AEQ,
- HINIC_CEQ
-};
+#define HINIC_AEQ 0
enum hinic_resp_aeq_num {
HINIC_AEQ0 = 0,
diff --git a/drivers/net/hinic/base/hinic_pmd_eqs.c b/drivers/net/hinic/base/hinic_pmd_eqs.c
index 8d216cf..abe0dae 100644
--- a/drivers/net/hinic/base/hinic_pmd_eqs.c
+++ b/drivers/net/hinic/base/hinic_pmd_eqs.c
@@ -43,32 +43,6 @@
((val) & (~(AEQ_CTRL_1_##member##_MASK \
<< AEQ_CTRL_1_##member##_SHIFT)))
-#define CEQ_CTRL_0_INTR_IDX_SHIFT 0
-#define CEQ_CTRL_0_DMA_ATTR_SHIFT 12
-#define CEQ_CTRL_0_LIMIT_KICK_SHIFT 20
-#define CEQ_CTRL_0_PCI_INTF_IDX_SHIFT 24
-#define CEQ_CTRL_0_INTR_MODE_SHIFT 31
-
-#define CEQ_CTRL_0_INTR_IDX_MASK 0x3FFU
-#define CEQ_CTRL_0_DMA_ATTR_MASK 0x3FU
-#define CEQ_CTRL_0_LIMIT_KICK_MASK 0xFU
-#define CEQ_CTRL_0_PCI_INTF_IDX_MASK 0x3U
-#define CEQ_CTRL_0_INTR_MODE_MASK 0x1U
-
-#define CEQ_CTRL_0_SET(val, member) \
- (((val) & CEQ_CTRL_0_##member##_MASK) << \
- CEQ_CTRL_0_##member##_SHIFT)
-
-#define CEQ_CTRL_1_LEN_SHIFT 0
-#define CEQ_CTRL_1_PAGE_SIZE_SHIFT 28
-
-#define CEQ_CTRL_1_LEN_MASK 0x1FFFFFU
-#define CEQ_CTRL_1_PAGE_SIZE_MASK 0xFU
-
-#define CEQ_CTRL_1_SET(val, member) \
- (((val) & CEQ_CTRL_1_##member##_MASK) << \
- CEQ_CTRL_1_##member##_SHIFT)
-
#define EQ_CONS_IDX_CONS_IDX_SHIFT 0
#define EQ_CONS_IDX_XOR_CHKSUM_SHIFT 24
#define EQ_CONS_IDX_INT_ARMED_SHIFT 31
@@ -90,13 +64,11 @@
#define EQ_CONS_IDX(eq) ((eq)->cons_idx | \
((u32)(eq)->wrapped << EQ_WRAPPED_SHIFT))
-#define EQ_CONS_IDX_REG_ADDR(eq) (((eq)->type == HINIC_AEQ) ? \
- HINIC_CSR_AEQ_CONS_IDX_ADDR((eq)->q_id) :\
- HINIC_CSR_CEQ_CONS_IDX_ADDR((eq)->q_id))
+#define EQ_CONS_IDX_REG_ADDR(eq) \
+ (HINIC_CSR_AEQ_CONS_IDX_ADDR((eq)->q_id))
-#define EQ_PROD_IDX_REG_ADDR(eq) (((eq)->type == HINIC_AEQ) ? \
- HINIC_CSR_AEQ_PROD_IDX_ADDR((eq)->q_id) :\
- HINIC_CSR_CEQ_PROD_IDX_ADDR((eq)->q_id))
+#define EQ_PROD_IDX_REG_ADDR(eq) \
+ (HINIC_CSR_AEQ_PROD_IDX_ADDR((eq)->q_id))
#define GET_EQ_NUM_PAGES(eq, size) \
((u16)(ALIGN((eq)->eq_len * (u32)(eq)->elem_size, (size)) \
@@ -111,9 +83,6 @@
#define EQ_SET_HW_ELEM_SIZE_VAL(eq) ((u32)ilog2(ELEMENT_SIZE_IN_32B(eq)))
#define AEQ_DMA_ATTR_DEFAULT 0
-#define CEQ_DMA_ATTR_DEFAULT 0
-
-#define CEQ_LMT_KICK_DEFAULT 0
#define EQ_WRAPPED_SHIFT 20
@@ -137,7 +106,7 @@ static u8 eq_cons_idx_checksum_set(u32 val)
* set_eq_cons_idx - write the cons idx to the hw
* @eq: The event queue to update the cons idx for
* @arm_state: indicate whether report interrupts when generate eq element
- **/
+ */
static void set_eq_cons_idx(struct hinic_eq *eq, u32 arm_state)
{
u32 eq_cons_idx, eq_wrap_ci, val;
@@ -172,125 +141,60 @@ static void set_eq_cons_idx(struct hinic_eq *eq, u32 arm_state)
/**
* eq_update_ci - update the cons idx of event queue
* @eq: the event queue to update the cons idx for
- **/
+ */
void eq_update_ci(struct hinic_eq *eq)
{
set_eq_cons_idx(eq, HINIC_EQ_ARMED);
}
-struct hinic_ceq_ctrl_reg {
- struct hinic_mgmt_msg_head mgmt_msg_head;
-
- u16 func_id;
- u16 q_id;
- u32 ctrl0;
- u32 ctrl1;
-};
-
-static int set_ceq_ctrl_reg(struct hinic_hwdev *hwdev, u16 q_id,
- u32 ctrl0, u32 ctrl1)
-{
- struct hinic_ceq_ctrl_reg ceq_ctrl;
- u16 in_size = sizeof(ceq_ctrl);
-
- memset(&ceq_ctrl, 0, in_size);
- ceq_ctrl.mgmt_msg_head.resp_aeq_num = HINIC_AEQ1;
- ceq_ctrl.func_id = hinic_global_func_id(hwdev);
- ceq_ctrl.q_id = q_id;
- ceq_ctrl.ctrl0 = ctrl0;
- ceq_ctrl.ctrl1 = ctrl1;
-
- return hinic_msg_to_mgmt_sync(hwdev, HINIC_MOD_COMM,
- HINIC_MGMT_CMD_CEQ_CTRL_REG_WR_BY_UP,
- &ceq_ctrl, in_size, NULL, NULL, 0);
-}
-
/**
* set_eq_ctrls - setting eq's ctrls registers
* @eq: the event queue for setting
- **/
-static int set_eq_ctrls(struct hinic_eq *eq)
+ */
+static void set_aeq_ctrls(struct hinic_eq *eq)
{
- enum hinic_eq_type type = eq->type;
struct hinic_hwif *hwif = eq->hwdev->hwif;
struct irq_info *eq_irq = &eq->eq_irq;
u32 addr, val, ctrl0, ctrl1, page_size_val, elem_size;
u32 pci_intf_idx = HINIC_PCI_INTF_IDX(hwif);
- int ret = 0;
-
- if (type == HINIC_AEQ) {
- /* set ctrl0 */
- addr = HINIC_CSR_AEQ_CTRL_0_ADDR(eq->q_id);
-
- val = hinic_hwif_read_reg(hwif, addr);
-
- val = AEQ_CTRL_0_CLEAR(val, INTR_IDX) &
- AEQ_CTRL_0_CLEAR(val, DMA_ATTR) &
- AEQ_CTRL_0_CLEAR(val, PCI_INTF_IDX) &
- AEQ_CTRL_0_CLEAR(val, INTR_MODE);
-
- ctrl0 = AEQ_CTRL_0_SET(eq_irq->msix_entry_idx, INTR_IDX) |
- AEQ_CTRL_0_SET(AEQ_DMA_ATTR_DEFAULT, DMA_ATTR) |
- AEQ_CTRL_0_SET(pci_intf_idx, PCI_INTF_IDX) |
- AEQ_CTRL_0_SET(HINIC_INTR_MODE_ARMED, INTR_MODE);
-
- val |= ctrl0;
- hinic_hwif_write_reg(hwif, addr, val);
+ /* set ctrl0 */
+ addr = HINIC_CSR_AEQ_CTRL_0_ADDR(eq->q_id);
- /* set ctrl1 */
- addr = HINIC_CSR_AEQ_CTRL_1_ADDR(eq->q_id);
+ val = hinic_hwif_read_reg(hwif, addr);
- page_size_val = EQ_SET_HW_PAGE_SIZE_VAL(eq);
- elem_size = EQ_SET_HW_ELEM_SIZE_VAL(eq);
+ val = AEQ_CTRL_0_CLEAR(val, INTR_IDX) &
+ AEQ_CTRL_0_CLEAR(val, DMA_ATTR) &
+ AEQ_CTRL_0_CLEAR(val, PCI_INTF_IDX) &
+ AEQ_CTRL_0_CLEAR(val, INTR_MODE);
- ctrl1 = AEQ_CTRL_1_SET(eq->eq_len, LEN) |
- AEQ_CTRL_1_SET(elem_size, ELEM_SIZE) |
- AEQ_CTRL_1_SET(page_size_val, PAGE_SIZE);
+ ctrl0 = AEQ_CTRL_0_SET(eq_irq->msix_entry_idx, INTR_IDX) |
+ AEQ_CTRL_0_SET(AEQ_DMA_ATTR_DEFAULT, DMA_ATTR) |
+ AEQ_CTRL_0_SET(pci_intf_idx, PCI_INTF_IDX) |
+ AEQ_CTRL_0_SET(HINIC_INTR_MODE_ARMED, INTR_MODE);
- hinic_hwif_write_reg(hwif, addr, ctrl1);
- } else {
- ctrl0 = CEQ_CTRL_0_SET(eq_irq->msix_entry_idx, INTR_IDX) |
- CEQ_CTRL_0_SET(CEQ_DMA_ATTR_DEFAULT, DMA_ATTR) |
- CEQ_CTRL_0_SET(CEQ_LMT_KICK_DEFAULT, LIMIT_KICK) |
- CEQ_CTRL_0_SET(pci_intf_idx, PCI_INTF_IDX) |
- CEQ_CTRL_0_SET(HINIC_INTR_MODE_ARMED, INTR_MODE);
+ val |= ctrl0;
- page_size_val = EQ_SET_HW_PAGE_SIZE_VAL(eq);
+ hinic_hwif_write_reg(hwif, addr, val);
- ctrl1 = CEQ_CTRL_1_SET(eq->eq_len, LEN) |
- CEQ_CTRL_1_SET(page_size_val, PAGE_SIZE);
+ /* set ctrl1 */
+ addr = HINIC_CSR_AEQ_CTRL_1_ADDR(eq->q_id);
- /* set ceq ctrl reg through mgmt cpu */
- ret = set_ceq_ctrl_reg(eq->hwdev, eq->q_id, ctrl0, ctrl1);
- }
-
- return ret;
-}
-
-/**
- * ceq_elements_init - Initialize all the elements in the ceq
- * @eq: the event queue
- * @init_val: value to init with it the elements
- **/
-static void ceq_elements_init(struct hinic_eq *eq, u32 init_val)
-{
- u16 i;
- u32 *ceqe;
+ page_size_val = EQ_SET_HW_PAGE_SIZE_VAL(eq);
+ elem_size = EQ_SET_HW_ELEM_SIZE_VAL(eq);
- for (i = 0; i < eq->eq_len; i++) {
- ceqe = GET_CEQ_ELEM(eq, i);
- *(ceqe) = cpu_to_be32(init_val);
- }
+ ctrl1 = AEQ_CTRL_1_SET(eq->eq_len, LEN) |
+ AEQ_CTRL_1_SET(elem_size, ELEM_SIZE) |
+ AEQ_CTRL_1_SET(page_size_val, PAGE_SIZE);
- rte_wmb(); /* Write the init values */
+ hinic_hwif_write_reg(hwif, addr, ctrl1);
}
/**
* aeq_elements_init - initialize all the elements in the aeq
* @eq: the event queue
* @init_val: value to init with it the elements
- **/
+ */
static void aeq_elements_init(struct hinic_eq *eq, u32 init_val)
{
struct hinic_aeq_elem *aeqe;
@@ -307,7 +211,7 @@ static void aeq_elements_init(struct hinic_eq *eq, u32 init_val)
/**
* alloc_eq_pages - allocate the pages for the queue
* @eq: the event queue
- **/
+ */
static int alloc_eq_pages(struct hinic_eq *eq)
{
struct hinic_hwif *hwif = eq->hwdev->hwif;
@@ -355,10 +259,7 @@ static int alloc_eq_pages(struct hinic_eq *eq)
init_val = EQ_WRAPPED(eq);
- if (eq->type == HINIC_AEQ)
- aeq_elements_init(eq, init_val);
- else
- ceq_elements_init(eq, init_val);
+ aeq_elements_init(eq, init_val);
return 0;
@@ -375,7 +276,7 @@ static int alloc_eq_pages(struct hinic_eq *eq)
/**
* free_eq_pages - free the pages of the queue
* @eq: the event queue
- **/
+ */
static void free_eq_pages(struct hinic_eq *eq)
{
struct hinic_hwdev *hwdev = eq->hwdev;
@@ -393,7 +294,7 @@ static void free_eq_pages(struct hinic_eq *eq)
#define MSIX_ENTRY_IDX_0 (0)
/**
- * init_eq - initialize eq
+ * init_aeq - initialize aeq
* @eq: the event queue
* @hwdev: the pointer to the private hardware device object
* @q_id: Queue id number
@@ -402,36 +303,31 @@ static void free_eq_pages(struct hinic_eq *eq)
* @page_size: the page size of the event queue
* @entry: msix entry associated with the event queue
* Return: 0 - Success, Negative - failure
- **/
-static int init_eq(struct hinic_eq *eq, struct hinic_hwdev *hwdev, u16 q_id,
- u16 q_len, enum hinic_eq_type type, u32 page_size,
+ */
+static int init_aeq(struct hinic_eq *eq, struct hinic_hwdev *hwdev, u16 q_id,
+ u16 q_len, u32 page_size,
__rte_unused struct irq_info *entry)
{
int err = 0;
eq->hwdev = hwdev;
eq->q_id = q_id;
- eq->type = type;
+ eq->type = HINIC_AEQ;
eq->page_size = page_size;
eq->eq_len = q_len;
/* clear eq_len to force eqe drop in hardware */
- if (eq->type == HINIC_AEQ) {
- hinic_hwif_write_reg(eq->hwdev->hwif,
- HINIC_CSR_AEQ_CTRL_1_ADDR(eq->q_id), 0);
- } else {
- err = set_ceq_ctrl_reg(eq->hwdev, eq->q_id, 0, 0);
- if (err) {
- PMD_DRV_LOG(ERR, "Set ceq control registers ctrl0[0] ctrl1[0] failed");
- return err;
- }
- }
+ hinic_hwif_write_reg(eq->hwdev->hwif,
+ HINIC_CSR_AEQ_CTRL_1_ADDR(eq->q_id), 0);
+
+ /* Clear PI and CI, also clear the ARM bit */
+ hinic_hwif_write_reg(eq->hwdev->hwif, EQ_CONS_IDX_REG_ADDR(eq), 0);
+ hinic_hwif_write_reg(eq->hwdev->hwif, EQ_PROD_IDX_REG_ADDR(eq), 0);
eq->cons_idx = 0;
eq->wrapped = 0;
- eq->elem_size = (type == HINIC_AEQ) ?
- HINIC_AEQE_SIZE : HINIC_CEQE_SIZE;
+ eq->elem_size = HINIC_AEQE_SIZE;
eq->num_pages = GET_EQ_NUM_PAGES(eq, page_size);
eq->num_elem_in_pg = GET_EQ_NUM_ELEMS(eq, page_size);
@@ -452,16 +348,10 @@ static int init_eq(struct hinic_eq *eq, struct hinic_hwdev *hwdev, u16 q_id,
return err;
}
- /* pmd use MSIX_ENTRY_IDX_0*/
+ /* pmd use MSIX_ENTRY_IDX_0 */
eq->eq_irq.msix_entry_idx = MSIX_ENTRY_IDX_0;
- err = set_eq_ctrls(eq);
- if (err) {
- PMD_DRV_LOG(ERR, "Init eq control registers failed");
- goto init_eq_ctrls_err;
- }
-
- hinic_hwif_write_reg(eq->hwdev->hwif, EQ_PROD_IDX_REG_ADDR(eq), 0);
+ set_aeq_ctrls(eq);
set_eq_cons_idx(eq, HINIC_EQ_ARMED);
if (eq->q_id == 0)
@@ -470,32 +360,23 @@ static int init_eq(struct hinic_eq *eq, struct hinic_hwdev *hwdev, u16 q_id,
eq->poll_retry_nr = HINIC_RETRY_NUM;
return 0;
-
-init_eq_ctrls_err:
- free_eq_pages(eq);
-
- return err;
}
/**
- * remove_eq - remove eq
+ * remove_aeq - remove aeq
* @eq: the event queue
- **/
-static void remove_eq(struct hinic_eq *eq)
+ */
+static void remove_aeq(struct hinic_eq *eq)
{
struct irq_info *entry = &eq->eq_irq;
- if (eq->type == HINIC_AEQ) {
- if (eq->q_id == 0)
- hinic_set_msix_state(eq->hwdev, entry->msix_entry_idx,
- HINIC_MSIX_DISABLE);
+ if (eq->q_id == 0)
+ hinic_set_msix_state(eq->hwdev, entry->msix_entry_idx,
+ HINIC_MSIX_DISABLE);
- /* clear eq_len to avoid hw access host memory */
- hinic_hwif_write_reg(eq->hwdev->hwif,
- HINIC_CSR_AEQ_CTRL_1_ADDR(eq->q_id), 0);
- } else {
- (void)set_ceq_ctrl_reg(eq->hwdev, eq->q_id, 0, 0);
- }
+ /* clear eq_len to avoid hw access host memory */
+ hinic_hwif_write_reg(eq->hwdev->hwif,
+ HINIC_CSR_AEQ_CTRL_1_ADDR(eq->q_id), 0);
/* update cons_idx to avoid invalid interrupt */
eq->cons_idx = (u16)hinic_hwif_read_reg(eq->hwdev->hwif,
@@ -511,7 +392,7 @@ static void remove_eq(struct hinic_eq *eq)
* @num_aeqs: number of aeq
* @msix_entries: msix entries associated with the event queues
* Return: 0 - Success, Negative - failure
- **/
+ */
static int
hinic_aeqs_init(struct hinic_hwdev *hwdev, u16 num_aeqs,
struct irq_info *msix_entries)
@@ -529,9 +410,9 @@ static void remove_eq(struct hinic_eq *eq)
aeqs->num_aeqs = num_aeqs;
for (q_id = HINIC_AEQN_START; q_id < num_aeqs; q_id++) {
- err = init_eq(&aeqs->aeq[q_id], hwdev, q_id,
- HINIC_DEFAULT_AEQ_LEN, HINIC_AEQ,
- HINIC_EQ_PAGE_SIZE, &msix_entries[q_id]);
+ err = init_aeq(&aeqs->aeq[q_id], hwdev, q_id,
+ HINIC_DEFAULT_AEQ_LEN, HINIC_EQ_PAGE_SIZE,
+ &msix_entries[q_id]);
if (err) {
PMD_DRV_LOG(ERR, "Init aeq %d failed", q_id);
goto init_aeq_err;
@@ -542,7 +423,7 @@ static void remove_eq(struct hinic_eq *eq)
init_aeq_err:
for (i = 0; i < q_id; i++)
- remove_eq(&aeqs->aeq[i]);
+ remove_aeq(&aeqs->aeq[i]);
kfree(aeqs);
@@ -552,7 +433,7 @@ static void remove_eq(struct hinic_eq *eq)
/**
* hinic_aeqs_free - free all the aeqs
* @hwdev: the pointer to the private hardware device object
- **/
+ */
static void hinic_aeqs_free(struct hinic_hwdev *hwdev)
{
struct hinic_aeqs *aeqs = hwdev->aeqs;
@@ -560,7 +441,7 @@ static void hinic_aeqs_free(struct hinic_hwdev *hwdev)
/* hinic pmd use aeq[1~3], aeq[0] used in kernel only */
for (q_id = HINIC_AEQN_START; q_id < aeqs->num_aeqs ; q_id++)
- remove_eq(&aeqs->aeq[q_id]);
+ remove_aeq(&aeqs->aeq[q_id]);
kfree(aeqs);
}
diff --git a/drivers/net/hinic/base/hinic_pmd_eqs.h b/drivers/net/hinic/base/hinic_pmd_eqs.h
index fdb9854..16046ec 100644
--- a/drivers/net/hinic/base/hinic_pmd_eqs.h
+++ b/drivers/net/hinic/base/hinic_pmd_eqs.h
@@ -21,9 +21,6 @@
#define HINIC_DEFAULT_AEQ_LEN 64
-#define HINIC_RECV_NEXT_AEQE HINIC_ERROR
-#define HINIC_RECV_DONE HINIC_OK
-
#define GET_EQ_ELEMENT(eq, idx) \
(((u8 *)(eq)->virt_addr[(idx) / (eq)->num_elem_in_pg]) + \
(((u32)(idx) & ((eq)->num_elem_in_pg - 1)) * (eq)->elem_size))
@@ -58,7 +55,7 @@ enum hinic_aeq_type {
struct hinic_eq {
struct hinic_hwdev *hwdev;
u16 q_id;
- enum hinic_eq_type type;
+ u16 type;
u32 page_size;
u16 eq_len;
--
1.8.3.1
^ permalink raw reply [flat|nested] 32+ messages in thread
* [dpdk-dev] [PATCH v2 17/17] net/hinic: optimize tx&rx performance
2019-09-25 14:26 [dpdk-dev] [PATCH v2 00/17] Add advanced features for Huawei hinic pmd Xiaoyun wang
` (16 preceding siblings ...)
2019-09-25 14:30 ` [dpdk-dev] [PATCH v2 16/17] net/hinic/base: optimize aeq interfaces Xiaoyun wang
@ 2019-09-25 14:30 ` Xiaoyun wang
2019-09-27 2:08 ` Gavin Hu (Arm Technology China)
2019-09-26 18:51 ` [dpdk-dev] [PATCH v2 00/17] Add advanced features for Huawei hinic pmd Ferruh Yigit
18 siblings, 1 reply; 32+ messages in thread
From: Xiaoyun wang @ 2019-09-25 14:30 UTC (permalink / raw)
To: ferruh.yigit
Cc: dev, xuanziyang2, shahar.belkar, luoxianjun, tanya.brokhman,
zhouguoyang, wulike1, Xiaoyun wang
This patch optimizes receive packets performance
in arm platform.
Signed-off-by: Xiaoyun wang <cloud.wangxiaoyun@huawei.com>
---
drivers/net/hinic/hinic_pmd_rx.c | 17 +++++++++++++++++
drivers/net/hinic/hinic_pmd_rx.h | 11 +++++++++++
2 files changed, 28 insertions(+)
diff --git a/drivers/net/hinic/hinic_pmd_rx.c b/drivers/net/hinic/hinic_pmd_rx.c
index 37b4f5c..94071ee 100644
--- a/drivers/net/hinic/hinic_pmd_rx.c
+++ b/drivers/net/hinic/hinic_pmd_rx.c
@@ -950,6 +950,19 @@ void hinic_rx_alloc_pkts(struct hinic_rxq *rxq)
}
}
+#if defined(__ARM64_NEON__)
+static inline uint32_t __attribute__((always_inline))
+hinic_read_cqe_status(uintptr_t addr)
+{
+ uint32_t val;
+
+ asm volatile("ldar %x[val], [%x[addr]]"
+ : [val] "=r" (val)
+ : [addr] "r" (addr));
+ return val;
+}
+#endif
+
u16 hinic_recv_pkts(void *rx_queue, struct rte_mbuf **rx_pkts, u16 nb_pkts)
{
struct rte_mbuf *rxm;
@@ -972,7 +985,11 @@ u16 hinic_recv_pkts(void *rx_queue, struct rte_mbuf **rx_pkts, u16 nb_pkts)
while (pkts < nb_pkts) {
/* 2. current ci is done */
rx_cqe = &rxq->rx_cqe[sw_ci];
+#if defined(__X86_64_SSE__)
status = rx_cqe->status;
+#elif defined(__ARM64_NEON__)
+ status = hinic_read_cqe_status((uintptr_t)&rxq->rx_cqe[sw_ci]);
+#endif
if (!HINIC_GET_RX_DONE_BE(status))
break;
diff --git a/drivers/net/hinic/hinic_pmd_rx.h b/drivers/net/hinic/hinic_pmd_rx.h
index fe2735b..fa27e91 100644
--- a/drivers/net/hinic/hinic_pmd_rx.h
+++ b/drivers/net/hinic/hinic_pmd_rx.h
@@ -28,6 +28,7 @@ struct hinic_rq_ctrl {
u32 ctrl_fmt;
};
+#if defined(__X86_64_SSE__)
struct hinic_rq_cqe {
u32 status;
u32 vlan_len;
@@ -36,6 +37,16 @@ struct hinic_rq_cqe {
u32 rsvd[4];
};
+#elif defined(__ARM64_NEON__)
+struct hinic_rq_cqe {
+ u32 status;
+ u32 vlan_len;
+ u32 offload_type;
+ u32 rss_hash;
+
+ u32 rsvd[4];
+} __rte_cache_aligned;
+#endif
struct hinic_rq_cqe_sect {
struct hinic_sge sge;
--
1.8.3.1
^ permalink raw reply [flat|nested] 32+ messages in thread
* Re: [dpdk-dev] [PATCH v2 12/17] net/hinic: set link down and up
2019-09-25 14:30 ` [dpdk-dev] [PATCH v2 12/17] net/hinic: set link down and up Xiaoyun wang
@ 2019-09-26 18:47 ` Ferruh Yigit
2019-09-28 8:59 ` [dpdk-dev] 答复: " Wangxiaoyun (Cloud, Network Chip Application Development Dept)
2019-09-30 8:39 ` [dpdk-dev] " Wangxiaoyun (Cloud, Network Chip Application Development Dept)
0 siblings, 2 replies; 32+ messages in thread
From: Ferruh Yigit @ 2019-09-26 18:47 UTC (permalink / raw)
To: Xiaoyun wang
Cc: dev, xuanziyang2, shahar.belkar, luoxianjun, tanya.brokhman,
zhouguoyang, wulike1
On 9/25/2019 3:30 PM, Xiaoyun wang wrote:
> This patch supports setting link down and up,
> Also adds support for geting firmware version.
>
> Signed-off-by: Xiaoyun wang <cloud.wangxiaoyun@huawei.com>
<...>
> @@ -2751,9 +2832,12 @@ static void hinic_dev_close(struct rte_eth_dev *dev)
> static const struct eth_dev_ops hinic_pmd_ops = {
> .dev_configure = hinic_dev_configure,
> .dev_infos_get = hinic_dev_infos_get,
> + .fw_version_get = hinic_fw_version_get,
> .rx_queue_setup = hinic_rx_queue_setup,
> .tx_queue_setup = hinic_tx_queue_setup,
> .dev_start = hinic_dev_start,
> + .dev_set_link_up = hinic_dev_set_link_up,
> + .dev_set_link_down = hinic_dev_set_link_down,
> .link_update = hinic_link_update,
> .rx_queue_release = hinic_rx_queue_release,
> .tx_queue_release = hinic_tx_queue_release,
> @@ -2785,6 +2869,7 @@ static void hinic_dev_close(struct rte_eth_dev *dev)
> static const struct eth_dev_ops hinic_pmd_vf_ops = {
> .dev_configure = hinic_dev_configure,
> .dev_infos_get = hinic_dev_infos_get,
> + .fw_version_get = hinic_fw_version_get,
> .rx_queue_setup = hinic_rx_queue_setup,
> .tx_queue_setup = hinic_tx_queue_setup,
> .dev_start = hinic_dev_start,
>
Patch is also updating fw version get support, can you please mention from it in
the commit log? Or separate the patch..
^ permalink raw reply [flat|nested] 32+ messages in thread
* Re: [dpdk-dev] [PATCH v2 05/17] net/hinic: add allmulticast mode and MTU set
2019-09-25 14:30 ` [dpdk-dev] [PATCH v2 05/17] net/hinic: add allmulticast mode and MTU set Xiaoyun wang
@ 2019-09-26 18:47 ` Ferruh Yigit
2019-09-30 8:43 ` Wangxiaoyun (Cloud, Network Chip Application Development Dept)
0 siblings, 1 reply; 32+ messages in thread
From: Ferruh Yigit @ 2019-09-26 18:47 UTC (permalink / raw)
To: Xiaoyun wang
Cc: dev, xuanziyang2, shahar.belkar, luoxianjun, tanya.brokhman,
zhouguoyang, wulike1
On 9/25/2019 3:30 PM, Xiaoyun wang wrote:
> When enable allmulticast mode, all multicast packets
> can be received. This patch also adds support for mtu set,
> the range of MTU is from 256 to 9600.
>
> Signed-off-by: Xiaoyun wang <cloud.wangxiaoyun@huawei.com>
<...>
> +static int hinic_dev_set_mtu(struct rte_eth_dev *dev, uint16_t mtu)
> +{
> + int ret = 0;
> + struct hinic_nic_dev *nic_dev = HINIC_ETH_DEV_TO_PRIVATE_NIC_DEV(dev);
> +
> + PMD_DRV_LOG(INFO, "Set port mtu, port_id: %d, mtu: %d, max_pkt_len: %d",
> + dev->data->port_id, mtu, HINIC_MTU_TO_PKTLEN(mtu));
> +
> + if (mtu < HINIC_MIN_MTU_SIZE || mtu > HINIC_MAX_MTU_SIZE) {
> + PMD_DRV_LOG(ERR, "Invalid mtu: %d, must between %d and %d",
> + mtu, HINIC_MIN_MTU_SIZE, HINIC_MAX_MTU_SIZE);
> + return -EINVAL;
> + }
> +
> + ret = hinic_set_port_mtu(nic_dev->hwdev, mtu);
> + if (ret) {
> + PMD_DRV_LOG(ERR, "Set port mtu failed, ret: %d", ret);
> + return ret;
> + }
> +
> + /* update max frame size */
> + dev->data->dev_conf.rxmode.max_rx_pkt_len = HINIC_MTU_TO_PKTLEN(mtu);
> + nic_dev->mtu_size = mtu;
> +
> + return ret;
> +}
There are (relatively) new 'min_mtu' & 'max_mtu' in 'struct rte_eth_dev_info',
since MTU is enabled you may prefer to set them in "hinic_dev_infos_get()'
^ permalink raw reply [flat|nested] 32+ messages in thread
* Re: [dpdk-dev] [PATCH v2 07/17] net/hinic: add fdir config interface
2019-09-25 14:30 ` [dpdk-dev] [PATCH v2 07/17] net/hinic: add fdir config interface Xiaoyun wang
@ 2019-09-26 18:48 ` Ferruh Yigit
2019-09-30 14:08 ` Wangxiaoyun (Cloud, Network Chip Application Development Dept)
0 siblings, 1 reply; 32+ messages in thread
From: Ferruh Yigit @ 2019-09-26 18:48 UTC (permalink / raw)
To: Xiaoyun wang
Cc: dev, xuanziyang2, shahar.belkar, luoxianjun, tanya.brokhman,
zhouguoyang, wulike1
On 9/25/2019 3:30 PM, Xiaoyun wang wrote:
> This patch adds fdir config operation, including set fdir filter,
> normal filter, set and clear fdir tcam.
>
> Signed-off-by: Xiaoyun wang <cloud.wangxiaoyun@huawei.com>
> ---
> drivers/net/hinic/base/hinic_pmd_cmd.h | 2 +
> drivers/net/hinic/base/hinic_pmd_niccfg.c | 205 ++++++++++++++++++++++++++++++
> drivers/net/hinic/base/hinic_pmd_niccfg.h | 134 +++++++++++++++++++
This looks like base code update to prepare driver for rte_flow, may be better
to use "net/hinic/base:" in the patch title.
^ permalink raw reply [flat|nested] 32+ messages in thread
* Re: [dpdk-dev] [PATCH v2 15/17] net/hinic: add hinic PMD doc files
2019-09-25 14:30 ` [dpdk-dev] [PATCH v2 15/17] net/hinic: add hinic PMD doc files Xiaoyun wang
@ 2019-09-26 18:51 ` Ferruh Yigit
2019-09-30 14:15 ` Wangxiaoyun (Cloud, Network Chip Application Development Dept)
0 siblings, 1 reply; 32+ messages in thread
From: Ferruh Yigit @ 2019-09-26 18:51 UTC (permalink / raw)
To: Xiaoyun wang
Cc: dev, xuanziyang2, shahar.belkar, luoxianjun, tanya.brokhman,
zhouguoyang, wulike1
On 9/25/2019 3:30 PM, Xiaoyun wang wrote:
> Add doc files about new features and modification.
>
> Signed-off-by: Xiaoyun wang <cloud.wangxiaoyun@huawei.com>
> ---
> doc/guides/nics/features/hinic.ini | 12 ++++++++-
> doc/guides/nics/hinic.rst | 6 +++++
> doc/guides/rel_notes/release_19_11.rst | 45 ++++++----------------------------
> 3 files changed, 25 insertions(+), 38 deletions(-)
>
> diff --git a/doc/guides/nics/features/hinic.ini b/doc/guides/nics/features/hinic.ini
> index fe063d6..dc02b4b 100644
> --- a/doc/guides/nics/features/hinic.ini
> +++ b/doc/guides/nics/features/hinic.ini
> @@ -9,16 +9,22 @@ Link status = Y
> Link status event = Y
> Free Tx mbuf on demand = Y
> Queue start/stop = Y
> -Jumbo frame = N
> +MTU update = Y
> +Jumbo frame = Y
> Scattered Rx = Y
> TSO = Y
> +LRO = Y
> Promiscuous mode = Y
> +Allmulticast mode = Y
> Unicast MAC filter = Y
> Multicast MAC filter = Y
> RSS hash = Y
> RSS key update = Y
> RSS reta update = Y
> Inner RSS = Y
> +SR-IOV = Y
> +VLAN filter = Y
> +VLAN offload = Y
> CRC offload = Y
> L3 checksum offload = Y
> L4 checksum offload = Y
> @@ -27,6 +33,10 @@ Inner L4 checksum = Y
> Basic stats = Y
> Extended stats = Y
> Stats per queue = Y
> +Flow director = Y
> +Flow control = Y
> +FW version = Y
> +Multiprocess aware = Y
> Linux UIO = Y
> Linux VFIO = Y
> BSD nic_uio = N
> diff --git a/doc/guides/nics/hinic.rst b/doc/guides/nics/hinic.rst
> index c9329bc..f036fc5 100644
> --- a/doc/guides/nics/hinic.rst
> +++ b/doc/guides/nics/hinic.rst
> @@ -24,6 +24,12 @@ Features
> - Link state information
> - Link flow control
> - Scattered and gather for TX and RX
> +- SR�CIOV - Partially supported at this point, VFIO only
Can you fix the char is SR-IOV?
> +- Allmulticast mode
> +- Unicast MAC filter
> +- Multicast MAC filter
> +- FW version
> +- Flow director
>
> Prerequisites
> -------------
> diff --git a/doc/guides/rel_notes/release_19_11.rst b/doc/guides/rel_notes/release_19_11.rst
> index 65361c4..6c6f27f 100644
> --- a/doc/guides/rel_notes/release_19_11.rst
> +++ b/doc/guides/rel_notes/release_19_11.rst
> @@ -56,11 +56,15 @@ New Features
> Also, make sure to start the actual text at the margin.
> =========================================================
>
> -* **Updated the Intel ice driver.**
> +* **Updated the Huawei hinic driver.**
>
> - Updated the Intel ice driver with new features and improvements, including:
> + Updated the Huawei hinic driver with new features and improvements, including:
>
> - * Added support for device-specific DDP package loading.
> + * Enabled SR-IOV - Partially supported at this point, VFIO only.
> + * Supported VLAN filter and VLAN offload.
> + * Supported Unicast MAC filter and Multicast MAC filter.
> + * Supported FW version get.
> + * Supported Flow director for LACP, VRRP, BGP and so on.
Can you please distribute the doc patches in to the related patches that
introduces the feature, for all three document, it helps by documenting what has
been added in the patch.
>
> Removed Items
> -------------
> @@ -99,30 +103,6 @@ API Changes
> Also, make sure to start the actual text at the margin.
> =========================================================
>
> -* ethdev: changed ``rte_eth_dev_infos_get`` return value from ``void`` to
> - ``int`` to provide a way to report various error conditions.
> -
> -* ethdev: changed ``rte_eth_promiscuous_enable`` and
> - ``rte_eth_promiscuous_disable`` return value from ``void`` to ``int`` to
> - provide a way to report various error conditions.
> -
> -* ethdev: changed ``rte_eth_allmulticast_enable`` and
> - ``rte_eth_allmulticast_disable`` return value from ``void`` to ``int`` to
> - provide a way to report various error conditions.
> -
> -* ethdev: changed ``rte_eth_dev_xstats_reset`` return value from ``void`` to
> - ``int`` to provide a way to report various error conditions.
> -
> -* ethdev: changed ``rte_eth_link_get`` and ``rte_eth_link_get_nowait``
> - return value from ``void`` to ``int`` to provide a way to report various
> - error conditions.
> -
> -* ethdev: changed ``rte_eth_macaddr_get`` return value from ``void`` to
> - ``int`` to provide a way to report various error conditions.
> -
> -* ethdev: changed ``rte_eth_dev_owner_delete`` return value from ``void`` to
> - ``int`` to provide a way to report various error conditions.
> -
>
> ABI Changes
> -----------
> @@ -174,7 +154,7 @@ The libraries prepended with a plus sign were incremented in this version.
> librte_distributor.so.1
> librte_eal.so.11
> librte_efd.so.1
> - + librte_ethdev.so.13
> + librte_ethdev.so.12
> librte_eventdev.so.7
> librte_flow_classify.so.1
> librte_gro.so.1
> @@ -252,12 +232,3 @@ Tested Platforms
> Also, make sure to start the actual text at the margin.
> =========================================================
>
> -* **Updated Mellanox mlx5 driver.**
> -
> - Updated Mellanox mlx5 driver with new features and improvements, including:
> -
> - * Added support for VLAN pop flow offload command.
> - * Added support for VLAN push flow offload command.
> - * Added support for VLAN set PCP offload command.
> - * Added support for VLAN set VID offload command.
> -
>
I guess above changes are git mistake, please check in next version.
^ permalink raw reply [flat|nested] 32+ messages in thread
* Re: [dpdk-dev] [PATCH v2 00/17] Add advanced features for Huawei hinic pmd
2019-09-25 14:26 [dpdk-dev] [PATCH v2 00/17] Add advanced features for Huawei hinic pmd Xiaoyun wang
` (17 preceding siblings ...)
2019-09-25 14:30 ` [dpdk-dev] [PATCH v2 17/17] net/hinic: optimize tx&rx performance Xiaoyun wang
@ 2019-09-26 18:51 ` Ferruh Yigit
2019-09-30 14:19 ` Wangxiaoyun (Cloud, Network Chip Application Development Dept)
18 siblings, 1 reply; 32+ messages in thread
From: Ferruh Yigit @ 2019-09-26 18:51 UTC (permalink / raw)
To: Xiaoyun wang
Cc: dev, xuanziyang2, shahar.belkar, luoxianjun, tanya.brokhman,
zhouguoyang, wulike1
On 9/25/2019 3:30 PM, Xiaoyun wang wrote:
> This patch set adds advanced features for Huawei hinic pmd,
> such as VLAN filter and VLAN offload, SR-IOV, FW version get,
> set link down and up, Flow director for LACP, VRRP, BGP and so on.
>
> Xiaoyun wang (17):
> net/hinic/base: add mbox command channel for SRIOV
> net/hinic/base: add HW interfaces for SRIOV
> net/hinic: add VF PMD operation interfaces
> net/hinic: add VLAN filter and offload
> net/hinic: add allmulticast mode and MTU set
> net/hinic: add unicast and multicast MAC set
> net/hinic: add fdir config interface
> net/hinic: add fdir validate flow operations
> net/hinic: create and destroy ntuple filter
> net/hinic: create and destroy fdir filter
> net/hinic: flush fdir filter
> net/hinic: set link down and up
> net/hinic: support inner L3 checksum offload
> net/hinic: support LRO offload
> net/hinic: add hinic PMD doc files
> net/hinic/base: optimize aeq interfaces
> net/hinic: optimize tx&rx performance
Hi Xiaoyun,
Overall patchset lgtm, only there are a few minor comments, also can you please
send a new version on top of latest next-net because of the ethdev API updates?
Thanks,
ferruh
^ permalink raw reply [flat|nested] 32+ messages in thread
* Re: [dpdk-dev] [PATCH v2 17/17] net/hinic: optimize tx&rx performance
2019-09-25 14:30 ` [dpdk-dev] [PATCH v2 17/17] net/hinic: optimize tx&rx performance Xiaoyun wang
@ 2019-09-27 2:08 ` Gavin Hu (Arm Technology China)
2019-09-30 14:41 ` Wangxiaoyun (Cloud, Network Chip Application Development Dept)
0 siblings, 1 reply; 32+ messages in thread
From: Gavin Hu (Arm Technology China) @ 2019-09-27 2:08 UTC (permalink / raw)
To: Xiaoyun wang, ferruh.yigit
Cc: dev, xuanziyang2, shahar.belkar, luoxianjun, tanya.brokhman,
zhouguoyang, wulike1, nd
Hi Xiaoyun,
> -----Original Message-----
> From: dev <dev-bounces@dpdk.org> On Behalf Of Xiaoyun wang
> Sent: Wednesday, September 25, 2019 10:31 PM
> To: ferruh.yigit@intel.com
> Cc: dev@dpdk.org; xuanziyang2@huawei.com; shahar.belkar@huawei.com;
> luoxianjun@huawei.com; tanya.brokhman@huawei.com;
> zhouguoyang@huawei.com; wulike1@huawei.com; Xiaoyun wang
> <cloud.wangxiaoyun@huawei.com>
> Subject: [dpdk-dev] [PATCH v2 17/17] net/hinic: optimize tx&rx
> performance
>
> This patch optimizes receive packets performance
> in arm platform.
>
> Signed-off-by: Xiaoyun wang <cloud.wangxiaoyun@huawei.com>
> ---
> drivers/net/hinic/hinic_pmd_rx.c | 17 +++++++++++++++++
> drivers/net/hinic/hinic_pmd_rx.h | 11 +++++++++++
> 2 files changed, 28 insertions(+)
>
> diff --git a/drivers/net/hinic/hinic_pmd_rx.c
> b/drivers/net/hinic/hinic_pmd_rx.c
> index 37b4f5c..94071ee 100644
> --- a/drivers/net/hinic/hinic_pmd_rx.c
> +++ b/drivers/net/hinic/hinic_pmd_rx.c
> @@ -950,6 +950,19 @@ void hinic_rx_alloc_pkts(struct hinic_rxq *rxq)
> }
> }
>
> +#if defined(__ARM64_NEON__)
No NEON intrinsics used, maybe RTE_ARCH_ARM64 is better.
In the following line __rte_always_inline is commonly used in DPDK, the effect is same.
/Gavin
> +static inline uint32_t __attribute__((always_inline))
> +hinic_read_cqe_status(uintptr_t addr)
> +{
> + uint32_t val;
> +
> + asm volatile("ldar %x[val], [%x[addr]]"
> + : [val] "=r" (val)
> + : [addr] "r" (addr));
> + return val;
> +}
> +#endif
I understand your intention is the reading of the status is observed before the following reads.
This can be fulfilled by __atomic_load_n(...) with __ATOMIC_ACQUIRE semantics.
This C11 way applies to all the arches, and you don't need the differentiation of arches.
/Gavin
> +
> u16 hinic_recv_pkts(void *rx_queue, struct rte_mbuf **rx_pkts, u16
> nb_pkts)
> {
> struct rte_mbuf *rxm;
> @@ -972,7 +985,11 @@ u16 hinic_recv_pkts(void *rx_queue, struct
> rte_mbuf **rx_pkts, u16 nb_pkts)
> while (pkts < nb_pkts) {
> /* 2. current ci is done */
> rx_cqe = &rxq->rx_cqe[sw_ci];
> +#if defined(__X86_64_SSE__)
> status = rx_cqe->status;
> +#elif defined(__ARM64_NEON__)
> + status = hinic_read_cqe_status((uintptr_t)&rxq-
> >rx_cqe[sw_ci]);
> +#endif
> if (!HINIC_GET_RX_DONE_BE(status))
> break;
>
> diff --git a/drivers/net/hinic/hinic_pmd_rx.h
> b/drivers/net/hinic/hinic_pmd_rx.h
> index fe2735b..fa27e91 100644
> --- a/drivers/net/hinic/hinic_pmd_rx.h
> +++ b/drivers/net/hinic/hinic_pmd_rx.h
> @@ -28,6 +28,7 @@ struct hinic_rq_ctrl {
> u32 ctrl_fmt;
> };
>
> +#if defined(__X86_64_SSE__)
> struct hinic_rq_cqe {
> u32 status;
> u32 vlan_len;
> @@ -36,6 +37,16 @@ struct hinic_rq_cqe {
>
> u32 rsvd[4];
> };
> +#elif defined(__ARM64_NEON__)
> +struct hinic_rq_cqe {
> + u32 status;
> + u32 vlan_len;
> + u32 offload_type;
> + u32 rss_hash;
> +
> + u32 rsvd[4];
> +} __rte_cache_aligned;
> +#endif
>
> struct hinic_rq_cqe_sect {
> struct hinic_sge sge;
> --
> 1.8.3.1
^ permalink raw reply [flat|nested] 32+ messages in thread
* [dpdk-dev] 答复: [PATCH v2 12/17] net/hinic: set link down and up
2019-09-26 18:47 ` Ferruh Yigit
@ 2019-09-28 8:59 ` Wangxiaoyun (Cloud, Network Chip Application Development Dept)
2019-09-30 8:39 ` [dpdk-dev] " Wangxiaoyun (Cloud, Network Chip Application Development Dept)
1 sibling, 0 replies; 32+ messages in thread
From: Wangxiaoyun (Cloud, Network Chip Application Development Dept) @ 2019-09-28 8:59 UTC (permalink / raw)
To: Ferruh Yigit, dev
Cc: Luoxianjun, Tanya Brokhman, zhouguoyang, Wulike (Collin),
Xuanziyang (William),
Shahar Belkar
-----邮件原件-----
发件人: Ferruh Yigit [mailto:ferruh.yigit@intel.com]
发送时间: 2019年9月27日 2:48
收件人: Wangxiaoyun (Cloud, Network Chip Application Development Dept)
抄送: dev@dpdk.org; Xuanziyang (William); Shahar Belkar; Luoxianjun; Tanya Brokhman; zhouguoyang; Wulike (Collin)
主题: Re: [PATCH v2 12/17] net/hinic: set link down and up
On 9/25/2019 3:30 PM, Xiaoyun wang wrote:
> This patch supports setting link down and up,
> Also adds support for geting firmware version.
>
> Signed-off-by: Xiaoyun wang <cloud.wangxiaoyun@huawei.com>
<...>
> @@ -2751,9 +2832,12 @@ static void hinic_dev_close(struct rte_eth_dev *dev)
> static const struct eth_dev_ops hinic_pmd_ops = {
> .dev_configure = hinic_dev_configure,
> .dev_infos_get = hinic_dev_infos_get,
> + .fw_version_get = hinic_fw_version_get,
> .rx_queue_setup = hinic_rx_queue_setup,
> .tx_queue_setup = hinic_tx_queue_setup,
> .dev_start = hinic_dev_start,
> + .dev_set_link_up = hinic_dev_set_link_up,
> + .dev_set_link_down = hinic_dev_set_link_down,
> .link_update = hinic_link_update,
> .rx_queue_release = hinic_rx_queue_release,
> .tx_queue_release = hinic_tx_queue_release,
> @@ -2785,6 +2869,7 @@ static void hinic_dev_close(struct rte_eth_dev *dev)
> static const struct eth_dev_ops hinic_pmd_vf_ops = {
> .dev_configure = hinic_dev_configure,
> .dev_infos_get = hinic_dev_infos_get,
> + .fw_version_get = hinic_fw_version_get,
> .rx_queue_setup = hinic_rx_queue_setup,
> .tx_queue_setup = hinic_tx_queue_setup,
> .dev_start = hinic_dev_start,
>
Patch is also updating fw version get support, can you please mention from it in
the commit log? Or separate the patch..
Hi, Ferruh,
Thanks for your comments. I will separate this patch to 2 patches, one is "set link down and up" , the other is "get firmware version".
And I will fix it by patch V3.
Best regards
Xiaoyun Wang
^ permalink raw reply [flat|nested] 32+ messages in thread
* Re: [dpdk-dev] [PATCH v2 12/17] net/hinic: set link down and up
2019-09-26 18:47 ` Ferruh Yigit
2019-09-28 8:59 ` [dpdk-dev] 答复: " Wangxiaoyun (Cloud, Network Chip Application Development Dept)
@ 2019-09-30 8:39 ` Wangxiaoyun (Cloud, Network Chip Application Development Dept)
1 sibling, 0 replies; 32+ messages in thread
From: Wangxiaoyun (Cloud, Network Chip Application Development Dept) @ 2019-09-30 8:39 UTC (permalink / raw)
To: Ferruh Yigit
Cc: dev, xuanziyang2, shahar.belkar, luoxianjun, tanya.brokhman,
zhouguoyang, wulike1
Thanks for your comments, I have splited this patch into 2 patched, one is for link down and up,the other is for getting fw version.
在 2019/9/27 2:47, Ferruh Yigit 写道:
> On 9/25/2019 3:30 PM, Xiaoyun wang wrote:
>> This patch supports setting link down and up,
>> Also adds support for geting firmware version.
>>
>> Signed-off-by: Xiaoyun wang <cloud.wangxiaoyun@huawei.com>
> <...>
>
>> @@ -2751,9 +2832,12 @@ static void hinic_dev_close(struct rte_eth_dev *dev)
>> static const struct eth_dev_ops hinic_pmd_ops = {
>> .dev_configure = hinic_dev_configure,
>> .dev_infos_get = hinic_dev_infos_get,
>> + .fw_version_get = hinic_fw_version_get,
>> .rx_queue_setup = hinic_rx_queue_setup,
>> .tx_queue_setup = hinic_tx_queue_setup,
>> .dev_start = hinic_dev_start,
>> + .dev_set_link_up = hinic_dev_set_link_up,
>> + .dev_set_link_down = hinic_dev_set_link_down,
>> .link_update = hinic_link_update,
>> .rx_queue_release = hinic_rx_queue_release,
>> .tx_queue_release = hinic_tx_queue_release,
>> @@ -2785,6 +2869,7 @@ static void hinic_dev_close(struct rte_eth_dev *dev)
>> static const struct eth_dev_ops hinic_pmd_vf_ops = {
>> .dev_configure = hinic_dev_configure,
>> .dev_infos_get = hinic_dev_infos_get,
>> + .fw_version_get = hinic_fw_version_get,
>> .rx_queue_setup = hinic_rx_queue_setup,
>> .tx_queue_setup = hinic_tx_queue_setup,
>> .dev_start = hinic_dev_start,
>>
> Patch is also updating fw version get support, can you please mention from it in
> the commit log? Or separate the patch..
>
>
^ permalink raw reply [flat|nested] 32+ messages in thread
* Re: [dpdk-dev] [PATCH v2 05/17] net/hinic: add allmulticast mode and MTU set
2019-09-26 18:47 ` Ferruh Yigit
@ 2019-09-30 8:43 ` Wangxiaoyun (Cloud, Network Chip Application Development Dept)
0 siblings, 0 replies; 32+ messages in thread
From: Wangxiaoyun (Cloud, Network Chip Application Development Dept) @ 2019-09-30 8:43 UTC (permalink / raw)
To: Ferruh Yigit
Cc: dev, xuanziyang2, shahar.belkar, luoxianjun, tanya.brokhman,
zhouguoyang, wulike1
Hi Ferruh,
Thanks for your comments, I have added min_mtu & max_mtu to hinic_dev_infos_get interface, which will be shown in Patch V3.
Best regards
Xiaoyun Wang
在 2019/9/27 2:47, Ferruh Yigit 写道:
> On 9/25/2019 3:30 PM, Xiaoyun wang wrote:
>> When enable allmulticast mode, all multicast packets
>> can be received. This patch also adds support for mtu set,
>> the range of MTU is from 256 to 9600.
>>
>> Signed-off-by: Xiaoyun wang <cloud.wangxiaoyun@huawei.com>
> <...>
>
>> +static int hinic_dev_set_mtu(struct rte_eth_dev *dev, uint16_t mtu)
>> +{
>> + int ret = 0;
>> + struct hinic_nic_dev *nic_dev = HINIC_ETH_DEV_TO_PRIVATE_NIC_DEV(dev);
>> +
>> + PMD_DRV_LOG(INFO, "Set port mtu, port_id: %d, mtu: %d, max_pkt_len: %d",
>> + dev->data->port_id, mtu, HINIC_MTU_TO_PKTLEN(mtu));
>> +
>> + if (mtu < HINIC_MIN_MTU_SIZE || mtu > HINIC_MAX_MTU_SIZE) {
>> + PMD_DRV_LOG(ERR, "Invalid mtu: %d, must between %d and %d",
>> + mtu, HINIC_MIN_MTU_SIZE, HINIC_MAX_MTU_SIZE);
>> + return -EINVAL;
>> + }
>> +
>> + ret = hinic_set_port_mtu(nic_dev->hwdev, mtu);
>> + if (ret) {
>> + PMD_DRV_LOG(ERR, "Set port mtu failed, ret: %d", ret);
>> + return ret;
>> + }
>> +
>> + /* update max frame size */
>> + dev->data->dev_conf.rxmode.max_rx_pkt_len = HINIC_MTU_TO_PKTLEN(mtu);
>> + nic_dev->mtu_size = mtu;
>> +
>> + return ret;
>> +}
> There are (relatively) new 'min_mtu' & 'max_mtu' in 'struct rte_eth_dev_info',
> since MTU is enabled you may prefer to set them in "hinic_dev_infos_get()'
>
^ permalink raw reply [flat|nested] 32+ messages in thread
* Re: [dpdk-dev] [PATCH v2 07/17] net/hinic: add fdir config interface
2019-09-26 18:48 ` Ferruh Yigit
@ 2019-09-30 14:08 ` Wangxiaoyun (Cloud, Network Chip Application Development Dept)
0 siblings, 0 replies; 32+ messages in thread
From: Wangxiaoyun (Cloud, Network Chip Application Development Dept) @ 2019-09-30 14:08 UTC (permalink / raw)
To: Ferruh Yigit
Cc: dev, xuanziyang2, shahar.belkar, luoxianjun, tanya.brokhman,
zhouguoyang, wulike1
Hi Ferruh,
Thanks for your comments, I have fixed the patch comments with "net/hinic/base: add fdir config interface", and send a new patch V3 to you.
Best regards
Xiaoyun Wang
在 2019/9/27 2:48, Ferruh Yigit 写道:
> On 9/25/2019 3:30 PM, Xiaoyun wang wrote:
>> This patch adds fdir config operation, including set fdir filter,
>> normal filter, set and clear fdir tcam.
>>
>> Signed-off-by: Xiaoyun wang <cloud.wangxiaoyun@huawei.com>
>> ---
>> drivers/net/hinic/base/hinic_pmd_cmd.h | 2 +
>> drivers/net/hinic/base/hinic_pmd_niccfg.c | 205 ++++++++++++++++++++++++++++++
>> drivers/net/hinic/base/hinic_pmd_niccfg.h | 134 +++++++++++++++++++
> This looks like base code update to prepare driver for rte_flow, may be better
> to use "net/hinic/base:" in the patch title.
>
>
^ permalink raw reply [flat|nested] 32+ messages in thread
* Re: [dpdk-dev] [PATCH v2 15/17] net/hinic: add hinic PMD doc files
2019-09-26 18:51 ` Ferruh Yigit
@ 2019-09-30 14:15 ` Wangxiaoyun (Cloud, Network Chip Application Development Dept)
0 siblings, 0 replies; 32+ messages in thread
From: Wangxiaoyun (Cloud, Network Chip Application Development Dept) @ 2019-09-30 14:15 UTC (permalink / raw)
To: Ferruh Yigit
Cc: dev, xuanziyang2, shahar.belkar, luoxianjun, tanya.brokhman,
zhouguoyang, wulike1
Hi Ferruh,
Thanks for your comments. I send to you a new patch V3 with the following review comments fixed.
1) I have fixed SR-IOV with unix style;
2) I splited the hinic.ini and hinic.rst into releated feature patches;
3) I have updated the release_19_11.rst with next version.
Best regards
Xiaoyun Wang
在 2019/9/27 2:51, Ferruh Yigit 写道:
> On 9/25/2019 3:30 PM, Xiaoyun wang wrote:
>> Add doc files about new features and modification.
>>
>> Signed-off-by: Xiaoyun wang <cloud.wangxiaoyun@huawei.com>
>> ---
>> doc/guides/nics/features/hinic.ini | 12 ++++++++-
>> doc/guides/nics/hinic.rst | 6 +++++
>> doc/guides/rel_notes/release_19_11.rst | 45 ++++++----------------------------
>> 3 files changed, 25 insertions(+), 38 deletions(-)
>>
>> diff --git a/doc/guides/nics/features/hinic.ini b/doc/guides/nics/features/hinic.ini
>> index fe063d6..dc02b4b 100644
>> --- a/doc/guides/nics/features/hinic.ini
>> +++ b/doc/guides/nics/features/hinic.ini
>> @@ -9,16 +9,22 @@ Link status = Y
>> Link status event = Y
>> Free Tx mbuf on demand = Y
>> Queue start/stop = Y
>> -Jumbo frame = N
>> +MTU update = Y
>> +Jumbo frame = Y
>> Scattered Rx = Y
>> TSO = Y
>> +LRO = Y
>> Promiscuous mode = Y
>> +Allmulticast mode = Y
>> Unicast MAC filter = Y
>> Multicast MAC filter = Y
>> RSS hash = Y
>> RSS key update = Y
>> RSS reta update = Y
>> Inner RSS = Y
>> +SR-IOV = Y
>> +VLAN filter = Y
>> +VLAN offload = Y
>> CRC offload = Y
>> L3 checksum offload = Y
>> L4 checksum offload = Y
>> @@ -27,6 +33,10 @@ Inner L4 checksum = Y
>> Basic stats = Y
>> Extended stats = Y
>> Stats per queue = Y
>> +Flow director = Y
>> +Flow control = Y
>> +FW version = Y
>> +Multiprocess aware = Y
>> Linux UIO = Y
>> Linux VFIO = Y
>> BSD nic_uio = N
>> diff --git a/doc/guides/nics/hinic.rst b/doc/guides/nics/hinic.rst
>> index c9329bc..f036fc5 100644
>> --- a/doc/guides/nics/hinic.rst
>> +++ b/doc/guides/nics/hinic.rst
>> @@ -24,6 +24,12 @@ Features
>> - Link state information
>> - Link flow control
>> - Scattered and gather for TX and RX
>> +- SR�CIOV - Partially supported at this point, VFIO only
> Can you fix the char is SR-IOV?
>
>> +- Allmulticast mode
>> +- Unicast MAC filter
>> +- Multicast MAC filter
>> +- FW version
>> +- Flow director
>>
>> Prerequisites
>> -------------
>> diff --git a/doc/guides/rel_notes/release_19_11.rst b/doc/guides/rel_notes/release_19_11.rst
>> index 65361c4..6c6f27f 100644
>> --- a/doc/guides/rel_notes/release_19_11.rst
>> +++ b/doc/guides/rel_notes/release_19_11.rst
>> @@ -56,11 +56,15 @@ New Features
>> Also, make sure to start the actual text at the margin.
>> =========================================================
>>
>> -* **Updated the Intel ice driver.**
>> +* **Updated the Huawei hinic driver.**
>>
>> - Updated the Intel ice driver with new features and improvements, including:
>> + Updated the Huawei hinic driver with new features and improvements, including:
>>
>> - * Added support for device-specific DDP package loading.
>> + * Enabled SR-IOV - Partially supported at this point, VFIO only.
>> + * Supported VLAN filter and VLAN offload.
>> + * Supported Unicast MAC filter and Multicast MAC filter.
>> + * Supported FW version get.
>> + * Supported Flow director for LACP, VRRP, BGP and so on.
>
> Can you please distribute the doc patches in to the related patches that
> introduces the feature, for all three document, it helps by documenting what has
> been added in the patch.
>
>>
>> Removed Items
>> -------------
>> @@ -99,30 +103,6 @@ API Changes
>> Also, make sure to start the actual text at the margin.
>> =========================================================
>>
>> -* ethdev: changed ``rte_eth_dev_infos_get`` return value from ``void`` to
>> - ``int`` to provide a way to report various error conditions.
>> -
>> -* ethdev: changed ``rte_eth_promiscuous_enable`` and
>> - ``rte_eth_promiscuous_disable`` return value from ``void`` to ``int`` to
>> - provide a way to report various error conditions.
>> -
>> -* ethdev: changed ``rte_eth_allmulticast_enable`` and
>> - ``rte_eth_allmulticast_disable`` return value from ``void`` to ``int`` to
>> - provide a way to report various error conditions.
>> -
>> -* ethdev: changed ``rte_eth_dev_xstats_reset`` return value from ``void`` to
>> - ``int`` to provide a way to report various error conditions.
>> -
>> -* ethdev: changed ``rte_eth_link_get`` and ``rte_eth_link_get_nowait``
>> - return value from ``void`` to ``int`` to provide a way to report various
>> - error conditions.
>> -
>> -* ethdev: changed ``rte_eth_macaddr_get`` return value from ``void`` to
>> - ``int`` to provide a way to report various error conditions.
>> -
>> -* ethdev: changed ``rte_eth_dev_owner_delete`` return value from ``void`` to
>> - ``int`` to provide a way to report various error conditions.
>> -
>>
>> ABI Changes
>> -----------
>> @@ -174,7 +154,7 @@ The libraries prepended with a plus sign were incremented in this version.
>> librte_distributor.so.1
>> librte_eal.so.11
>> librte_efd.so.1
>> - + librte_ethdev.so.13
>> + librte_ethdev.so.12
>> librte_eventdev.so.7
>> librte_flow_classify.so.1
>> librte_gro.so.1
>> @@ -252,12 +232,3 @@ Tested Platforms
>> Also, make sure to start the actual text at the margin.
>> =========================================================
>>
>> -* **Updated Mellanox mlx5 driver.**
>> -
>> - Updated Mellanox mlx5 driver with new features and improvements, including:
>> -
>> - * Added support for VLAN pop flow offload command.
>> - * Added support for VLAN push flow offload command.
>> - * Added support for VLAN set PCP offload command.
>> - * Added support for VLAN set VID offload command.
>> -
>>
> I guess above changes are git mistake, please check in next version.
>
> .
^ permalink raw reply [flat|nested] 32+ messages in thread
* Re: [dpdk-dev] [PATCH v2 00/17] Add advanced features for Huawei hinic pmd
2019-09-26 18:51 ` [dpdk-dev] [PATCH v2 00/17] Add advanced features for Huawei hinic pmd Ferruh Yigit
@ 2019-09-30 14:19 ` Wangxiaoyun (Cloud, Network Chip Application Development Dept)
0 siblings, 0 replies; 32+ messages in thread
From: Wangxiaoyun (Cloud, Network Chip Application Development Dept) @ 2019-09-30 14:19 UTC (permalink / raw)
To: Ferruh Yigit
Cc: dev, xuanziyang2, shahar.belkar, luoxianjun, tanya.brokhman,
zhouguoyang, wulike1
Hi Ferruh,
Thanks for your comments, I have sent a new patch V3 with new API changes, and add comments
for every patch version, which is show in cover-letter patch, please check it.
Best regards
Xiaoyun Wang
在 2019/9/27 2:51, Ferruh Yigit 写道:
> On 9/25/2019 3:30 PM, Xiaoyun wang wrote:
>> This patch set adds advanced features for Huawei hinic pmd,
>> such as VLAN filter and VLAN offload, SR-IOV, FW version get,
>> set link down and up, Flow director for LACP, VRRP, BGP and so on.
>>
>> Xiaoyun wang (17):
>> net/hinic/base: add mbox command channel for SRIOV
>> net/hinic/base: add HW interfaces for SRIOV
>> net/hinic: add VF PMD operation interfaces
>> net/hinic: add VLAN filter and offload
>> net/hinic: add allmulticast mode and MTU set
>> net/hinic: add unicast and multicast MAC set
>> net/hinic: add fdir config interface
>> net/hinic: add fdir validate flow operations
>> net/hinic: create and destroy ntuple filter
>> net/hinic: create and destroy fdir filter
>> net/hinic: flush fdir filter
>> net/hinic: set link down and up
>> net/hinic: support inner L3 checksum offload
>> net/hinic: support LRO offload
>> net/hinic: add hinic PMD doc files
>> net/hinic/base: optimize aeq interfaces
>> net/hinic: optimize tx&rx performance
> Hi Xiaoyun,
>
> Overall patchset lgtm, only there are a few minor comments, also can you please
> send a new version on top of latest next-net because of the ethdev API updates?
>
> Thanks,
> ferruh
>
^ permalink raw reply [flat|nested] 32+ messages in thread
* Re: [dpdk-dev] [PATCH v2 17/17] net/hinic: optimize tx&rx performance
2019-09-27 2:08 ` Gavin Hu (Arm Technology China)
@ 2019-09-30 14:41 ` Wangxiaoyun (Cloud, Network Chip Application Development Dept)
0 siblings, 0 replies; 32+ messages in thread
From: Wangxiaoyun (Cloud, Network Chip Application Development Dept) @ 2019-09-30 14:41 UTC (permalink / raw)
To: Gavin Hu (Arm Technology China), ferruh.yigit
Cc: dev, xuanziyang2, shahar.belkar, luoxianjun, tanya.brokhman,
zhouguoyang, wulike1, nd
Hi Gavin,
Thanks for your comments.
+#if defined(__ARM64_NEON__)
No NEON intrinsics used, maybe RTE_ARCH_ARM64 is better.
In the following line __rte_always_inline is commonly used in DPDK, the effect is same.
/Gavin
For this patch, we don't use NEON intrinsics, but for tx& rx process, we use NEON intrinsics
for wqebb bigend conversion on arm platform, so in order to keep ingres, all intrinsics
optimization on arm we use __ARM64_NEON__ definitation.
I understand your intention is the reading of the status is observed before the following reads.
This can be fulfilled by __atomic_load_n(...) with __ATOMIC_ACQUIRE semantics.
This C11 way applies to all the arches, and you don't need the differentiation of arches.
/Gavin
Thanks, i have changed it to __atomic_load_n(...) with __ATOMIC_ACQUIRE semantics, and send a new patch V3.
Best regards
Xiaoyun Wang
在 2019/9/27 10:08, Gavin Hu (Arm Technology China) 写道:
> Hi Xiaoyun,
>
>> -----Original Message-----
>> From: dev <dev-bounces@dpdk.org> On Behalf Of Xiaoyun wang
>> Sent: Wednesday, September 25, 2019 10:31 PM
>> To: ferruh.yigit@intel.com
>> Cc: dev@dpdk.org; xuanziyang2@huawei.com; shahar.belkar@huawei.com;
>> luoxianjun@huawei.com; tanya.brokhman@huawei.com;
>> zhouguoyang@huawei.com; wulike1@huawei.com; Xiaoyun wang
>> <cloud.wangxiaoyun@huawei.com>
>> Subject: [dpdk-dev] [PATCH v2 17/17] net/hinic: optimize tx&rx
>> performance
>>
>> This patch optimizes receive packets performance
>> in arm platform.
>>
>> Signed-off-by: Xiaoyun wang <cloud.wangxiaoyun@huawei.com>
>> ---
>> drivers/net/hinic/hinic_pmd_rx.c | 17 +++++++++++++++++
>> drivers/net/hinic/hinic_pmd_rx.h | 11 +++++++++++
>> 2 files changed, 28 insertions(+)
>>
>> diff --git a/drivers/net/hinic/hinic_pmd_rx.c
>> b/drivers/net/hinic/hinic_pmd_rx.c
>> index 37b4f5c..94071ee 100644
>> --- a/drivers/net/hinic/hinic_pmd_rx.c
>> +++ b/drivers/net/hinic/hinic_pmd_rx.c
>> @@ -950,6 +950,19 @@ void hinic_rx_alloc_pkts(struct hinic_rxq *rxq)
>> }
>> }
>>
>> +#if defined(__ARM64_NEON__)
> No NEON intrinsics used, maybe RTE_ARCH_ARM64 is better.
> In the following line __rte_always_inline is commonly used in DPDK, the effect is same.
> /Gavin
>
>> +static inline uint32_t __attribute__((always_inline))
>> +hinic_read_cqe_status(uintptr_t addr)
>> +{
>> + uint32_t val;
>> +
>> + asm volatile("ldar %x[val], [%x[addr]]"
>> + : [val] "=r" (val)
>> + : [addr] "r" (addr));
>> + return val;
>> +}
>> +#endif
> I understand your intention is the reading of the status is observed before the following reads.
> This can be fulfilled by __atomic_load_n(...) with __ATOMIC_ACQUIRE semantics.
> This C11 way applies to all the arches, and you don't need the differentiation of arches.
> /Gavin
>> +
>> u16 hinic_recv_pkts(void *rx_queue, struct rte_mbuf **rx_pkts, u16
>> nb_pkts)
>> {
>> struct rte_mbuf *rxm;
>> @@ -972,7 +985,11 @@ u16 hinic_recv_pkts(void *rx_queue, struct
>> rte_mbuf **rx_pkts, u16 nb_pkts)
>> while (pkts < nb_pkts) {
>> /* 2. current ci is done */
>> rx_cqe = &rxq->rx_cqe[sw_ci];
>> +#if defined(__X86_64_SSE__)
>> status = rx_cqe->status;
>> +#elif defined(__ARM64_NEON__)
>> + status = hinic_read_cqe_status((uintptr_t)&rxq-
>>> rx_cqe[sw_ci]);
>> +#endif
>> if (!HINIC_GET_RX_DONE_BE(status))
>> break;
>>
>> diff --git a/drivers/net/hinic/hinic_pmd_rx.h
>> b/drivers/net/hinic/hinic_pmd_rx.h
>> index fe2735b..fa27e91 100644
>> --- a/drivers/net/hinic/hinic_pmd_rx.h
>> +++ b/drivers/net/hinic/hinic_pmd_rx.h
>> @@ -28,6 +28,7 @@ struct hinic_rq_ctrl {
>> u32 ctrl_fmt;
>> };
>>
>> +#if defined(__X86_64_SSE__)
>> struct hinic_rq_cqe {
>> u32 status;
>> u32 vlan_len;
>> @@ -36,6 +37,16 @@ struct hinic_rq_cqe {
>>
>> u32 rsvd[4];
>> };
>> +#elif defined(__ARM64_NEON__)
>> +struct hinic_rq_cqe {
>> + u32 status;
>> + u32 vlan_len;
>> + u32 offload_type;
>> + u32 rss_hash;
>> +
>> + u32 rsvd[4];
>> +} __rte_cache_aligned;
>> +#endif
>>
>> struct hinic_rq_cqe_sect {
>> struct hinic_sge sge;
>> --
>> 1.8.3.1
>
^ permalink raw reply [flat|nested] 32+ messages in thread
end of thread, other threads:[~2019-09-30 14:41 UTC | newest]
Thread overview: 32+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2019-09-25 14:26 [dpdk-dev] [PATCH v2 00/17] Add advanced features for Huawei hinic pmd Xiaoyun wang
2019-09-25 14:30 ` Xiaoyun wang
2019-09-25 14:30 ` [dpdk-dev] [PATCH v2 01/17] net/hinic/base: add mbox command channel for SRIOV Xiaoyun wang
2019-09-25 14:30 ` [dpdk-dev] [PATCH v2 02/17] net/hinic/base: add HW interfaces " Xiaoyun wang
2019-09-25 14:30 ` [dpdk-dev] [PATCH v2 03/17] net/hinic: add VF PMD operation interfaces Xiaoyun wang
2019-09-25 14:30 ` [dpdk-dev] [PATCH v2 04/17] net/hinic: add VLAN filter and offload Xiaoyun wang
2019-09-25 14:30 ` [dpdk-dev] [PATCH v2 05/17] net/hinic: add allmulticast mode and MTU set Xiaoyun wang
2019-09-26 18:47 ` Ferruh Yigit
2019-09-30 8:43 ` Wangxiaoyun (Cloud, Network Chip Application Development Dept)
2019-09-25 14:30 ` [dpdk-dev] [PATCH v2 06/17] net/hinic: add unicast and multicast MAC set Xiaoyun wang
2019-09-25 14:30 ` [dpdk-dev] [PATCH v2 07/17] net/hinic: add fdir config interface Xiaoyun wang
2019-09-26 18:48 ` Ferruh Yigit
2019-09-30 14:08 ` Wangxiaoyun (Cloud, Network Chip Application Development Dept)
2019-09-25 14:30 ` [dpdk-dev] [PATCH v2 08/17] net/hinic: add fdir validate flow operations Xiaoyun wang
2019-09-25 14:30 ` [dpdk-dev] [PATCH v2 09/17] net/hinic: create and destroy ntuple filter Xiaoyun wang
2019-09-25 14:30 ` [dpdk-dev] [PATCH v2 10/17] net/hinic: create and destroy fdir filter Xiaoyun wang
2019-09-25 14:30 ` [dpdk-dev] [PATCH v2 11/17] net/hinic: flush " Xiaoyun wang
2019-09-25 14:30 ` [dpdk-dev] [PATCH v2 12/17] net/hinic: set link down and up Xiaoyun wang
2019-09-26 18:47 ` Ferruh Yigit
2019-09-28 8:59 ` [dpdk-dev] 答复: " Wangxiaoyun (Cloud, Network Chip Application Development Dept)
2019-09-30 8:39 ` [dpdk-dev] " Wangxiaoyun (Cloud, Network Chip Application Development Dept)
2019-09-25 14:30 ` [dpdk-dev] [PATCH v2 13/17] net/hinic: support inner L3 checksum offload Xiaoyun wang
2019-09-25 14:30 ` [dpdk-dev] [PATCH v2 14/17] net/hinic: support LRO offload Xiaoyun wang
2019-09-25 14:30 ` [dpdk-dev] [PATCH v2 15/17] net/hinic: add hinic PMD doc files Xiaoyun wang
2019-09-26 18:51 ` Ferruh Yigit
2019-09-30 14:15 ` Wangxiaoyun (Cloud, Network Chip Application Development Dept)
2019-09-25 14:30 ` [dpdk-dev] [PATCH v2 16/17] net/hinic/base: optimize aeq interfaces Xiaoyun wang
2019-09-25 14:30 ` [dpdk-dev] [PATCH v2 17/17] net/hinic: optimize tx&rx performance Xiaoyun wang
2019-09-27 2:08 ` Gavin Hu (Arm Technology China)
2019-09-30 14:41 ` Wangxiaoyun (Cloud, Network Chip Application Development Dept)
2019-09-26 18:51 ` [dpdk-dev] [PATCH v2 00/17] Add advanced features for Huawei hinic pmd Ferruh Yigit
2019-09-30 14:19 ` Wangxiaoyun (Cloud, Network Chip Application Development Dept)
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).