From: Andrew Rybchenko <arybchenko@solarflare.com>
To: <dev@dpdk.org>
Cc: Ivan Malov <ivan.malov@oktetlabs.ru>
Subject: [dpdk-dev] [PATCH 31/31] net/sfc: support firmware-assisted TSOv2
Date: Fri, 2 Dec 2016 07:44:51 +0000 [thread overview]
Message-ID: <1480664691-26561-32-git-send-email-arybchenko@solarflare.com> (raw)
In-Reply-To: <1480664691-26561-1-git-send-email-arybchenko@solarflare.com>
From: Ivan Malov <ivan.malov@oktetlabs.ru>
Reviewed-by: Andrew Lee <alee@solarflare.com>
Reviewed-by: Mark Spender <mspender@solarflare.com>
Reviewed-by: Robert Stonehouse <rstonehouse@solarflare.com>
Signed-off-by: Ivan Malov <ivan.malov@oktetlabs.ru>
Signed-off-by: Andrew Rybchenko <arybchenko@solarflare.com>
---
config/common_base | 1 +
doc/guides/nics/features/sfc_efx.ini | 1 +
doc/guides/nics/sfc_efx.rst | 8 ++
drivers/net/sfc/Makefile | 4 +
drivers/net/sfc/sfc.c | 8 ++
drivers/net/sfc/sfc.h | 2 +
drivers/net/sfc/sfc_ethdev.c | 3 +
drivers/net/sfc/sfc_tso.c | 203 +++++++++++++++++++++++++++++++++++
drivers/net/sfc/sfc_tx.c | 88 ++++++++++++++-
drivers/net/sfc/sfc_tx.h | 28 +++++
10 files changed, 343 insertions(+), 3 deletions(-)
create mode 100644 drivers/net/sfc/sfc_tso.c
diff --git a/config/common_base b/config/common_base
index 1eb8eea..8bb2a00 100644
--- a/config/common_base
+++ b/config/common_base
@@ -342,6 +342,7 @@ CONFIG_RTE_LIBRTE_PMD_XENVIRT=n
#
CONFIG_RTE_LIBRTE_SFC_EFX_PMD=y
CONFIG_RTE_LIBRTE_SFC_EFX_DEBUG=n
+CONFIG_RTE_LIBRTE_SFC_EFX_TSO=n
#
# Compile null PMD
diff --git a/doc/guides/nics/features/sfc_efx.ini b/doc/guides/nics/features/sfc_efx.ini
index 07c58d5..3a15baa 100644
--- a/doc/guides/nics/features/sfc_efx.ini
+++ b/doc/guides/nics/features/sfc_efx.ini
@@ -11,6 +11,7 @@ Queue start/stop = Y
MTU update = Y
Jumbo frame = Y
Scattered Rx = Y
+TSO = Y
Promiscuous mode = Y
Allmulticast mode = Y
Multicast MAC filter = Y
diff --git a/doc/guides/nics/sfc_efx.rst b/doc/guides/nics/sfc_efx.rst
index 64489de..f826726 100644
--- a/doc/guides/nics/sfc_efx.rst
+++ b/doc/guides/nics/sfc_efx.rst
@@ -60,6 +60,8 @@ SFC EFX PMD has support for:
- Allmulticast mode
+- TCP segmentation offload (TSO)
+
- Multicast MAC filter
- IPv4/IPv6 TCP/UDP receive checksum offload
@@ -166,6 +168,12 @@ Please note that enabling debugging options may affect system performance.
Enable compilation of the extra run-time consistency checks.
+- ``CONFIG_RTE_LIBRTE_SFC_EFX_TSO`` (default **n**)
+
+ Toggle TCP segmentation offload support.
+ Enabling the feature limits the number of available transmit queues
+ significantly due to the limited number of adapter TSO contexts.
+
Per-Device Parameters
~~~~~~~~~~~~~~~~~~~~~
diff --git a/drivers/net/sfc/Makefile b/drivers/net/sfc/Makefile
index dd099b2..14d6536 100644
--- a/drivers/net/sfc/Makefile
+++ b/drivers/net/sfc/Makefile
@@ -90,6 +90,8 @@ SRCS-$(CONFIG_RTE_LIBRTE_SFC_EFX_PMD) += sfc_port.c
SRCS-$(CONFIG_RTE_LIBRTE_SFC_EFX_PMD) += sfc_rx.c
SRCS-$(CONFIG_RTE_LIBRTE_SFC_EFX_PMD) += sfc_tx.c
+SRCS-$(CONFIG_RTE_LIBRTE_SFC_EFX_TSO) += sfc_tso.c
+
VPATH += $(SRCDIR)/base
SRCS-$(CONFIG_RTE_LIBRTE_SFC_EFX_PMD) += efx_bootcfg.c
@@ -139,4 +141,6 @@ DEPDIRS-$(CONFIG_RTE_LIBRTE_SFC_EFX_PMD) += lib/librte_ether
DEPDIRS-$(CONFIG_RTE_LIBRTE_SFC_EFX_PMD) += lib/librte_mempool
DEPDIRS-$(CONFIG_RTE_LIBRTE_SFC_EFX_PMD) += lib/librte_mbuf
+DEPDIRS-$(CONFIG_RTE_LIBRTE_SFC_EFX_TSO) += lib/librte_net
+
include $(RTE_SDK)/mk/rte.lib.mk
diff --git a/drivers/net/sfc/sfc.c b/drivers/net/sfc/sfc.c
index e79367d..02a56f7 100644
--- a/drivers/net/sfc/sfc.c
+++ b/drivers/net/sfc/sfc.c
@@ -621,6 +621,14 @@ sfc_attach(struct sfc_adapter *sa)
if (rc != 0)
goto fail_set_rss_defaults;
+#ifdef RTE_LIBRTE_SFC_EFX_TSO
+ sa->tso = efx_nic_cfg_get(sa->nic)->enc_fw_assisted_tso_v2_enabled;
+ if (!sa->tso)
+ sfc_warn(sa, "TSO support isn't available on this adapter");
+#else /* !RTE_LIBRTE_SFC_EFX_TSO */
+ sa->tso = B_FALSE;
+#endif /* RTE_LIBRTE_SFC_EFX_TSO */
+
sfc_log_init(sa, "fini nic");
efx_nic_fini(enp);
diff --git a/drivers/net/sfc/sfc.h b/drivers/net/sfc/sfc.h
index 01dbfb6..df9c20b 100644
--- a/drivers/net/sfc/sfc.h
+++ b/drivers/net/sfc/sfc.h
@@ -195,6 +195,8 @@ struct sfc_adapter {
unsigned int txq_count;
struct sfc_txq_info *txq_info;
+ boolean_t tso;
+
unsigned int rss_channels;
#if EFSYS_OPT_RX_SCALE
diff --git a/drivers/net/sfc/sfc_ethdev.c b/drivers/net/sfc/sfc_ethdev.c
index 09cc46b..1fb763a 100644
--- a/drivers/net/sfc/sfc_ethdev.c
+++ b/drivers/net/sfc/sfc_ethdev.c
@@ -93,6 +93,9 @@ sfc_dev_infos_get(struct rte_eth_dev *dev, struct rte_eth_dev_info *dev_info)
}
#endif
+ if (sa->tso)
+ dev_info->tx_offload_capa |= DEV_TX_OFFLOAD_TCP_TSO;
+
dev_info->rx_desc_lim.nb_max = EFX_RXQ_MAXNDESCS;
dev_info->rx_desc_lim.nb_min = EFX_RXQ_MINNDESCS;
/* The RXQ hardware requires that the descriptor count is a power
diff --git a/drivers/net/sfc/sfc_tso.c b/drivers/net/sfc/sfc_tso.c
new file mode 100644
index 0000000..fa8c04d
--- /dev/null
+++ b/drivers/net/sfc/sfc_tso.c
@@ -0,0 +1,203 @@
+/*-
+ * Copyright (c) 2016 Solarflare Communications Inc.
+ * All rights reserved.
+ *
+ * This software was jointly developed between OKTET Labs (under contract
+ * for Solarflare) and Solarflare Communications, Inc.
+ *
+ * Redistribution and use in source and binary forms, with or without
+ * modification, are permitted provided that the following conditions are met:
+ *
+ * 1. Redistributions of source code must retain the above copyright notice,
+ * this list of conditions and the following disclaimer.
+ * 2. Redistributions in binary form must reproduce the above copyright notice,
+ * this list of conditions and the following disclaimer in the documentation
+ * and/or other materials provided with the distribution.
+ *
+ * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS"
+ * AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO,
+ * THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR
+ * PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR
+ * CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL,
+ * EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO,
+ * PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS;
+ * OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY,
+ * WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR
+ * OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE,
+ * EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+ */
+
+#include <rte_ip.h>
+#include <rte_tcp.h>
+
+#include "sfc.h"
+#include "sfc_debug.h"
+#include "sfc_tx.h"
+#include "sfc_ev.h"
+
+/** Standard TSO header length */
+#define SFC_TSOH_STD_LEN 256
+
+/** The number of TSO option descriptors that precede the packet descriptors */
+#define SFC_TSO_OPDESCS_IDX_SHIFT 2
+
+int
+sfc_tso_alloc_tsoh_objs(struct sfc_adapter *sa, unsigned int sw_index,
+ struct sfc_txq *txq, unsigned int socket_id)
+{
+ unsigned int i;
+ struct sfc_txq_info *txq_info = &sa->txq_info[sw_index];
+
+ for (i = 0; i < txq_info->entries; ++i) {
+ txq->sw_ring[i].tsoh = rte_malloc_socket("sfc-txq-tsoh-obj",
+ SFC_TSOH_STD_LEN,
+ SFC_TX_SEG_BOUNDARY,
+ socket_id);
+ if (txq->sw_ring[i].tsoh == NULL)
+ goto fail_alloc_tsoh_objs;
+ }
+
+ return 0;
+
+fail_alloc_tsoh_objs:
+ while (i > 0)
+ rte_free(txq->sw_ring[--i].tsoh);
+
+ return ENOMEM;
+}
+
+void
+sfc_tso_free_tsoh_objs(struct sfc_adapter *sa, unsigned int sw_index)
+{
+ unsigned int txds;
+ struct sfc_txq_info *txq_info = &sa->txq_info[sw_index];
+ struct sfc_txq *txq = txq_info->txq;
+
+ for (txds = 0; txds < txq_info->entries; ++txds) {
+ rte_free(txq->sw_ring[txds].tsoh);
+ txq->sw_ring[txds].tsoh = NULL;
+ }
+}
+
+static void
+sfc_tso_prepare_header(struct sfc_txq *txq, struct rte_mbuf **in_seg,
+ size_t *in_off, unsigned int idx, size_t bytes_left)
+{
+ struct rte_mbuf *m = *in_seg;
+ size_t bytes_to_copy = 0;
+ uint8_t *tsoh = txq->sw_ring[idx & txq->ptr_mask].tsoh;
+
+ do {
+ bytes_to_copy = MIN(bytes_left, m->data_len);
+
+ rte_memcpy(tsoh, rte_pktmbuf_mtod(m, uint8_t *),
+ bytes_to_copy);
+
+ bytes_left -= bytes_to_copy;
+ tsoh += bytes_to_copy;
+
+ if (bytes_left > 0) {
+ m = m->next;
+ SFC_ASSERT(m != NULL);
+ }
+ } while (bytes_left > 0);
+
+ if (bytes_to_copy == m->data_len) {
+ *in_seg = m->next;
+ *in_off = 0;
+ } else {
+ *in_seg = m;
+ *in_off = bytes_to_copy;
+ }
+}
+
+int
+sfc_tso_do(struct sfc_txq *txq, unsigned int idx, struct rte_mbuf **in_seg,
+ size_t *in_off, efx_desc_t **pend, unsigned int *pkt_descs,
+ size_t *pkt_len)
+{
+ uint8_t *tsoh;
+ const struct tcp_hdr *th;
+ efsys_dma_addr_t header_paddr;
+ efsys_dma_addr_t paddr_next_frag;
+ uint16_t packet_id;
+ uint32_t sent_seq;
+ struct rte_mbuf *m = *in_seg;
+ size_t nh_off = m->l2_len; /* IP header offset */
+ size_t tcph_off = m->l2_len + m->l3_len; /* TCP header offset */
+ size_t header_len = m->l2_len + m->l3_len + m->l4_len;
+ const efx_nic_cfg_t *encp = efx_nic_cfg_get(txq->evq->sa->nic);
+
+ idx += SFC_TSO_OPDESCS_IDX_SHIFT;
+
+ /* Packets which have too big headers should be discarded */
+ if (unlikely(header_len > SFC_TSOH_STD_LEN))
+ return EMSGSIZE;
+
+ /*
+ * The TCP header must start at most 208 bytes into the frame.
+ * If it starts later than this then the NIC won't realise
+ * it's a TCP packet and TSO edits won't be applied
+ */
+ if (unlikely(tcph_off > encp->enc_tx_tso_tcp_header_offset_limit))
+ return EMSGSIZE;
+
+ header_paddr = rte_pktmbuf_mtophys(m);
+ paddr_next_frag = P2ROUNDUP(header_paddr + 1, SFC_TX_SEG_BOUNDARY);
+
+ /*
+ * Sometimes headers may be split across multiple mbufs. In such cases
+ * we need to glue those pieces and store them in some temporary place.
+ * Also, packet headers must be contiguous in memory, so that
+ * they can be referred to with a single DMA descriptor. Hence, handle
+ * the case where the original header crosses a 4K memory boundary
+ */
+ if ((m->data_len < header_len) ||
+ ((paddr_next_frag - header_paddr) < header_len)) {
+ sfc_tso_prepare_header(txq, in_seg, in_off, idx, header_len);
+ tsoh = txq->sw_ring[idx & txq->ptr_mask].tsoh;
+
+ header_paddr = rte_malloc_virt2phy((void *)tsoh);
+ } else {
+ if (m->data_len == header_len) {
+ *in_off = 0;
+ *in_seg = m->next;
+ } else {
+ *in_off = header_len;
+ }
+
+ tsoh = rte_pktmbuf_mtod(m, uint8_t *);
+ }
+
+ /* Handle IP header */
+ if (m->ol_flags & PKT_TX_IPV4) {
+ const struct ipv4_hdr *iphe4;
+
+ iphe4 = (const struct ipv4_hdr *)(tsoh + nh_off);
+ rte_memcpy(&packet_id, &iphe4->packet_id, sizeof(uint16_t));
+ packet_id = rte_be_to_cpu_16(packet_id);
+ } else if (m->ol_flags & PKT_TX_IPV6) {
+ packet_id = 0;
+ } else {
+ return EINVAL;
+ }
+
+ /* Handle TCP header */
+ th = (const struct tcp_hdr *)(tsoh + tcph_off);
+
+ rte_memcpy(&sent_seq, &th->sent_seq, sizeof(uint32_t));
+ sent_seq = rte_be_to_cpu_32(sent_seq);
+
+ efx_tx_qdesc_tso2_create(txq->common, packet_id, sent_seq, m->tso_segsz,
+ *pend, EFX_TX_FATSOV2_OPT_NDESCS);
+
+ *pend += EFX_TX_FATSOV2_OPT_NDESCS;
+ *pkt_descs += EFX_TX_FATSOV2_OPT_NDESCS;
+
+ efx_tx_qdesc_dma_create(txq->common, header_paddr, header_len,
+ B_FALSE, (*pend)++);
+ (*pkt_descs)++;
+ *pkt_len -= header_len;
+
+ return 0;
+}
diff --git a/drivers/net/sfc/sfc_tx.c b/drivers/net/sfc/sfc_tx.c
index 86bcfec..cb21f78 100644
--- a/drivers/net/sfc/sfc_tx.c
+++ b/drivers/net/sfc/sfc_tx.c
@@ -184,6 +184,12 @@ sfc_tx_qinit(struct sfc_adapter *sa, unsigned int sw_index,
if (txq->sw_ring == NULL)
goto fail_desc_alloc;
+ if (sa->tso) {
+ rc = sfc_tso_alloc_tsoh_objs(sa, sw_index, txq, socket_id);
+ if (rc != 0)
+ goto fail_alloc_tsoh_objs;
+ }
+
txq->state = SFC_TXQ_INITIALIZED;
txq->ptr_mask = txq_info->entries - 1;
txq->free_thresh = (tx_conf->tx_free_thresh) ? tx_conf->tx_free_thresh :
@@ -199,6 +205,9 @@ sfc_tx_qinit(struct sfc_adapter *sa, unsigned int sw_index,
return 0;
+fail_alloc_tsoh_objs:
+ rte_free(txq->sw_ring);
+
fail_desc_alloc:
rte_free(txq->pend_desc);
@@ -234,6 +243,8 @@ sfc_tx_qfini(struct sfc_adapter *sa, unsigned int sw_index)
SFC_ASSERT(txq != NULL);
SFC_ASSERT(txq->state == SFC_TXQ_INITIALIZED);
+ sfc_tso_free_tsoh_objs(sa, sw_index);
+
txq_info->txq = NULL;
txq_info->entries = 0;
@@ -300,6 +311,11 @@ sfc_tx_init(struct sfc_adapter *sa)
sa->txq_count = sa->eth_dev->data->nb_tx_queues;
+ if (sa->tso)
+ sa->txq_count = MIN(sa->txq_count,
+ efx_nic_cfg_get(sa->nic)->enc_fw_assisted_tso_v2_n_contexts /
+ efx_nic_cfg_get(sa->nic)->enc_hw_pf_count);
+
sa->txq_info = rte_calloc_socket("sfc-txqs", sa->txq_count,
sizeof(sa->txq_info[0]), 0,
sa->socket_id);
@@ -373,17 +389,25 @@ sfc_tx_qstart(struct sfc_adapter *sa, unsigned int sw_index)
* hence, we always enable it here
*/
if ((txq->flags & ETH_TXQ_FLAGS_NOXSUMTCP) ||
- (txq->flags & ETH_TXQ_FLAGS_NOXSUMUDP))
+ (txq->flags & ETH_TXQ_FLAGS_NOXSUMUDP)) {
flags = EFX_TXQ_CKSUM_IPV4;
- else
+ } else {
flags = EFX_TXQ_CKSUM_IPV4 | EFX_TXQ_CKSUM_TCPUDP;
+ if (sa->tso)
+ flags |= EFX_TXQ_FATSOV2;
+ }
+
rc = efx_tx_qcreate(sa->nic, sw_index, 0, &txq->mem,
txq_info->entries, 0 /* not used on EF10 */,
flags, evq->common,
&txq->common, &desc_index);
- if (rc != 0)
+ if (rc != 0) {
+ if (sa->tso && (rc == ENOSPC))
+ sfc_err(sa, "ran out of TSO contexts");
+
goto fail_tx_qcreate;
+ }
txq->added = txq->pending = txq->completed = desc_index;
txq->hw_vlan_tci = 0;
@@ -494,6 +518,13 @@ sfc_tx_start(struct sfc_adapter *sa)
sfc_log_init(sa, "txq_count = %u", sa->txq_count);
+ if (sa->tso) {
+ if (!efx_nic_cfg_get(sa->nic)->enc_fw_assisted_tso_v2_enabled) {
+ sfc_warn(sa, "TSO support was unable to be restored");
+ sa->tso = B_FALSE;
+ }
+ }
+
rc = efx_tx_init(sa->nic);
if (rc != 0)
goto fail_efx_tx_init;
@@ -607,6 +638,7 @@ sfc_xmit_pkts(void *tx_queue, struct rte_mbuf **tx_pkts, uint16_t nb_pkts)
struct rte_mbuf *m_seg = *pktp;
size_t pkt_len = m_seg->pkt_len;
unsigned int pkt_descs = 0;
+ size_t in_off = 0;
/*
* Here VLAN TCI is expected to be zero in case if no
@@ -617,6 +649,46 @@ sfc_xmit_pkts(void *tx_queue, struct rte_mbuf **tx_pkts, uint16_t nb_pkts)
*/
pkt_descs += sfc_tx_maybe_insert_tag(txq, m_seg, &pend);
+#ifdef RTE_LIBRTE_SFC_EFX_TSO
+ if (m_seg->ol_flags & PKT_TX_TCP_SEG) {
+ /*
+ * We expect correct 'pkt->l[2, 3, 4]_len' values
+ * to be set correctly by the caller
+ */
+ if (sfc_tso_do(txq, added, &m_seg, &in_off, &pend,
+ &pkt_descs, &pkt_len) != 0) {
+ /* We may have reached this place for
+ * one of the following reasons:
+ *
+ * 1) Packet header length is greater
+ * than SFC_TSOH_STD_LEN
+ * 2) TCP header starts at more then
+ * 208 bytes into the frame
+ *
+ * We will deceive RTE saying that we have sent
+ * the packet, but we will actually drop it.
+ * Hence, we should revert 'pend' to the
+ * previous state (in case we have added
+ * VLAN descriptor) and start processing
+ * another one packet. But the original
+ * mbuf shouldn't be orphaned
+ */
+ pend -= pkt_descs;
+
+ rte_pktmbuf_free(*pktp);
+
+ continue;
+ }
+
+ /*
+ * We've only added 2 FATSOv2 option descriptors
+ * and 1 descriptor for the linearized packet header.
+ * The outstanding work will be done in the same manner
+ * as for the usual non-TSO path
+ */
+ }
+#endif /* RTE_LIBRTE_SFC_EFX_TSO */
+
for (; m_seg != NULL; m_seg = m_seg->next) {
efsys_dma_addr_t next_frag;
size_t seg_len;
@@ -624,6 +696,16 @@ sfc_xmit_pkts(void *tx_queue, struct rte_mbuf **tx_pkts, uint16_t nb_pkts)
seg_len = m_seg->data_len;
next_frag = rte_mbuf_data_dma_addr(m_seg);
+ /*
+ * If we've started TSO transaction few steps earlier,
+ * we'll skip packet header using an offset in the
+ * current segment (which has been set to the
+ * first one containing payload)
+ */
+ seg_len -= in_off;
+ next_frag += in_off;
+ in_off = 0;
+
do {
efsys_dma_addr_t frag_addr = next_frag;
size_t frag_len;
diff --git a/drivers/net/sfc/sfc_tx.h b/drivers/net/sfc/sfc_tx.h
index 4d25c6a..d4298e4 100644
--- a/drivers/net/sfc/sfc_tx.h
+++ b/drivers/net/sfc/sfc_tx.h
@@ -50,6 +50,9 @@ struct sfc_evq;
struct sfc_tx_sw_desc {
struct rte_mbuf *mbuf;
+#ifdef RTE_LIBRTE_SFC_EFX_TSO
+ uint8_t *tsoh; /* Buffer to store TSO header */
+#endif /* RTE_LIBRTE_SFC_EFX_TSO */
};
enum sfc_txq_state_bit {
@@ -113,6 +116,31 @@ void sfc_tx_stop(struct sfc_adapter *sa);
uint16_t sfc_xmit_pkts(void *tx_queue, struct rte_mbuf **tx_pkts,
uint16_t nb_pkts);
+#ifdef RTE_LIBRTE_SFC_EFX_TSO
+/* From 'sfc_tso.c' */
+int sfc_tso_alloc_tsoh_objs(struct sfc_adapter *sa, unsigned int sw_index,
+ struct sfc_txq *txq, unsigned int socket_id);
+void sfc_tso_free_tsoh_objs(struct sfc_adapter *sa, unsigned int sw_index);
+int sfc_tso_do(struct sfc_txq *txq, unsigned int idx, struct rte_mbuf **in_seg,
+ size_t *in_off, efx_desc_t **pend, unsigned int *pkt_descs,
+ size_t *pkt_len);
+#else /* !RTE_LIBRTE_SFC_EFX_TSO */
+static inline int
+sfc_tso_alloc_tsoh_objs(__rte_unused struct sfc_adapter *sa,
+ __rte_unused unsigned int sw_index,
+ __rte_unused struct sfc_txq *txq,
+ __rte_unused unsigned int socket_id)
+{
+ return 0;
+}
+
+static inline void
+sfc_tso_free_tsoh_objs(__rte_unused struct sfc_adapter *sa,
+ __rte_unused unsigned int sw_index)
+{
+}
+#endif /* RTE_LIBRTE_SFC_EFX_TSO */
+
#ifdef __cplusplus
}
#endif
--
2.5.5
next prev parent reply other threads:[~2016-12-02 7:45 UTC|newest]
Thread overview: 70+ messages / expand[flat|nested] mbox.gz Atom feed top
2016-12-02 7:44 [dpdk-dev] [PATCH 00/31] Support more features in Solarflare PMD Andrew Rybchenko
2016-12-02 7:44 ` [dpdk-dev] [PATCH 01/31] net/sfc: implement MCDI logging callback Andrew Rybchenko
2016-12-02 7:44 ` [dpdk-dev] [PATCH 02/31] net/sfc: support parameter to choose performance profile Andrew Rybchenko
2016-12-02 7:44 ` [dpdk-dev] [PATCH 03/31] net/sfc: implement ethdev hook to get basic statistics Andrew Rybchenko
2016-12-02 7:44 ` [dpdk-dev] [PATCH 04/31] net/sfc: support extended statistics Andrew Rybchenko
2016-12-02 7:44 ` [dpdk-dev] [PATCH 05/31] net/sfc: support flow control settings get/set Andrew Rybchenko
2016-12-02 7:44 ` [dpdk-dev] [PATCH 06/31] net/sfc: support link status change interrupt Andrew Rybchenko
2016-12-02 7:44 ` [dpdk-dev] [PATCH 07/31] net/sfc: implement device operation to change MTU Andrew Rybchenko
2016-12-02 7:44 ` [dpdk-dev] [PATCH 08/31] net/sfc: support link speed and duplex settings Andrew Rybchenko
2016-12-02 7:44 ` [dpdk-dev] [PATCH 09/31] net/sfc: support link up/down Andrew Rybchenko
2016-12-02 7:44 ` [dpdk-dev] [PATCH 10/31] net/sfc: support promiscuous and all-multicast control Andrew Rybchenko
2016-12-02 7:44 ` [dpdk-dev] [PATCH 11/31] net/sfc: support main (the first) MAC address change Andrew Rybchenko
2016-12-02 7:44 ` [dpdk-dev] [PATCH 12/31] net/sfc: support multicast addresses list controls Andrew Rybchenko
2016-12-02 7:44 ` [dpdk-dev] [PATCH 13/31] net/sfc: support checksum offloads on receive Andrew Rybchenko
2016-12-02 7:44 ` [dpdk-dev] [PATCH 14/31] net/sfc: handle received packet type info provided by HW Andrew Rybchenko
2016-12-02 7:44 ` [dpdk-dev] [PATCH 15/31] net/sfc: support callback to get receive queue information Andrew Rybchenko
2016-12-02 7:44 ` [dpdk-dev] [PATCH 16/31] net/sfc: support Rx free threshold Andrew Rybchenko
2016-12-02 7:44 ` [dpdk-dev] [PATCH 17/31] net/sfc: add callback to get RxQ pending descriptors count Andrew Rybchenko
2016-12-02 7:44 ` [dpdk-dev] [PATCH 18/31] net/sfc: add RxQ descriptor done callback Andrew Rybchenko
2016-12-02 7:44 ` [dpdk-dev] [PATCH 19/31] net/sfc: support scattered Rx DMA Andrew Rybchenko
2016-12-02 7:44 ` [dpdk-dev] [PATCH 20/31] net/sfc: support deferred start of receive queues Andrew Rybchenko
2016-12-02 7:44 ` [dpdk-dev] [PATCH 21/31] net/sfc: add callback to get transmit queue information Andrew Rybchenko
2016-12-02 7:44 ` [dpdk-dev] [PATCH 22/31] net/sfc: support Tx free threshold Andrew Rybchenko
2016-12-02 7:44 ` [dpdk-dev] [PATCH 23/31] net/sfc: support deferred start of transmit queues Andrew Rybchenko
2016-12-02 7:44 ` [dpdk-dev] [PATCH 24/31] net/sfc: support VLAN offload on transmit path Andrew Rybchenko
2016-12-02 7:44 ` [dpdk-dev] [PATCH 25/31] net/sfc: add basic stubs for RSS support on driver attach Andrew Rybchenko
2016-12-02 7:44 ` [dpdk-dev] [PATCH 26/31] net/sfc: support RSS hash offload Andrew Rybchenko
2016-12-02 7:44 ` [dpdk-dev] [PATCH 27/31] net/sfc: add callback to query RSS key and hash types config Andrew Rybchenko
2016-12-02 7:44 ` [dpdk-dev] [PATCH 28/31] net/sfc: add callback to set " Andrew Rybchenko
2016-12-02 7:44 ` [dpdk-dev] [PATCH 29/31] net/sfc: add callback to query RSS redirection table Andrew Rybchenko
2016-12-02 7:44 ` [dpdk-dev] [PATCH 30/31] net/sfc: add callback to update " Andrew Rybchenko
2016-12-02 7:44 ` Andrew Rybchenko [this message]
2016-12-09 17:34 ` [dpdk-dev] [PATCH 00/31] Support more features in Solarflare PMD Ferruh Yigit
2016-12-15 12:50 ` Andrew Rybchenko
2016-12-15 12:50 ` [dpdk-dev] [PATCH v2 00/32] " Andrew Rybchenko
2016-12-15 12:50 ` [dpdk-dev] [PATCH v2 01/32] net/sfc: implement MCDI logging callback Andrew Rybchenko
2016-12-15 12:50 ` [dpdk-dev] [PATCH v2 02/32] net/sfc: support parameter to choose performance profile Andrew Rybchenko
2016-12-15 12:50 ` [dpdk-dev] [PATCH v2 03/32] net/sfc: implement ethdev hook to get basic statistics Andrew Rybchenko
2016-12-15 12:50 ` [dpdk-dev] [PATCH v2 04/32] net/sfc: support extended statistics Andrew Rybchenko
2016-12-15 12:50 ` [dpdk-dev] [PATCH v2 05/32] net/sfc: support flow control settings get/set Andrew Rybchenko
2016-12-15 12:50 ` [dpdk-dev] [PATCH v2 06/32] net/sfc: support link status change interrupt Andrew Rybchenko
2016-12-15 12:50 ` [dpdk-dev] [PATCH v2 07/32] net/sfc: implement device operation to change MTU Andrew Rybchenko
2016-12-15 12:50 ` [dpdk-dev] [PATCH v2 08/32] net/sfc: support link speed and duplex settings Andrew Rybchenko
2016-12-15 12:51 ` [dpdk-dev] [PATCH v2 09/32] net/sfc: support link up/down Andrew Rybchenko
2016-12-15 12:51 ` [dpdk-dev] [PATCH v2 10/32] net/sfc: support promiscuous and all-multicast control Andrew Rybchenko
2016-12-15 12:51 ` [dpdk-dev] [PATCH v2 11/32] net/sfc: support main (the first) MAC address change Andrew Rybchenko
2016-12-15 12:51 ` [dpdk-dev] [PATCH v2 12/32] net/sfc: support multicast addresses list controls Andrew Rybchenko
2016-12-15 12:51 ` [dpdk-dev] [PATCH v2 13/32] net/sfc: support checksum offloads on receive Andrew Rybchenko
2016-12-15 12:51 ` [dpdk-dev] [PATCH v2 14/32] net/sfc: handle received packet type info provided by HW Andrew Rybchenko
2016-12-15 12:51 ` [dpdk-dev] [PATCH v2 15/32] net/sfc: support callback to get receive queue information Andrew Rybchenko
2016-12-15 12:51 ` [dpdk-dev] [PATCH v2 16/32] net/sfc: support Rx free threshold Andrew Rybchenko
2016-12-15 12:51 ` [dpdk-dev] [PATCH v2 17/32] net/sfc: add callback to get RxQ pending descriptors count Andrew Rybchenko
2016-12-15 12:51 ` [dpdk-dev] [PATCH v2 18/32] net/sfc: add RxQ descriptor done callback Andrew Rybchenko
2016-12-15 12:51 ` [dpdk-dev] [PATCH v2 19/32] net/sfc: support scattered Rx DMA Andrew Rybchenko
2016-12-15 12:51 ` [dpdk-dev] [PATCH v2 20/32] net/sfc: support deferred start of receive queues Andrew Rybchenko
2016-12-15 12:51 ` [dpdk-dev] [PATCH v2 21/32] net/sfc: add callback to get transmit queue information Andrew Rybchenko
2016-12-15 12:51 ` [dpdk-dev] [PATCH v2 22/32] net/sfc: support Tx free threshold Andrew Rybchenko
2016-12-15 12:51 ` [dpdk-dev] [PATCH v2 23/32] net/sfc: support deferred start of transmit queues Andrew Rybchenko
2016-12-15 12:51 ` [dpdk-dev] [PATCH v2 24/32] net/sfc: support VLAN offload on transmit path Andrew Rybchenko
2016-12-15 12:51 ` [dpdk-dev] [PATCH v2 25/32] net/sfc/base: do not use enum type when values are bitmask Andrew Rybchenko
2017-01-18 11:30 ` Ferruh Yigit
2017-01-18 11:52 ` Andrew Rybchenko
2016-12-15 12:51 ` [dpdk-dev] [PATCH v2 26/32] net/sfc: add basic stubs for RSS support on driver attach Andrew Rybchenko
2016-12-15 12:51 ` [dpdk-dev] [PATCH v2 27/32] net/sfc: support RSS hash offload Andrew Rybchenko
2016-12-15 12:51 ` [dpdk-dev] [PATCH v2 28/32] net/sfc: add callback to query RSS key and hash types config Andrew Rybchenko
2016-12-15 12:51 ` [dpdk-dev] [PATCH v2 29/32] net/sfc: add callback to set " Andrew Rybchenko
2016-12-15 12:51 ` [dpdk-dev] [PATCH v2 30/32] net/sfc: add callback to query RSS redirection table Andrew Rybchenko
2016-12-15 12:51 ` [dpdk-dev] [PATCH v2 31/32] net/sfc: add callback to update " Andrew Rybchenko
2016-12-15 12:51 ` [dpdk-dev] [PATCH v2 32/32] net/sfc: support firmware-assisted TSOv2 Andrew Rybchenko
2016-12-16 9:57 ` [dpdk-dev] [PATCH v2 00/32] Support more features in Solarflare PMD Ferruh Yigit
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=1480664691-26561-32-git-send-email-arybchenko@solarflare.com \
--to=arybchenko@solarflare.com \
--cc=dev@dpdk.org \
--cc=ivan.malov@oktetlabs.ru \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
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).