patches for DPDK stable branches
 help / color / mirror / Atom feed
From: Kevin Traynor <ktraynor@redhat.com>
To: Zhichao Zeng <zhichaox.zeng@intel.com>
Cc: Qi Zhang <qi.z.zhang@intel.com>, dpdk stable <stable@dpdk.org>
Subject: patch 'net/iavf: fix processing VLAN TCI in SSE path' has been queued to stable release 21.11.3
Date: Tue, 25 Oct 2022 16:06:56 +0100	[thread overview]
Message-ID: <20221025150734.142189-61-ktraynor@redhat.com> (raw)
In-Reply-To: <20221025150734.142189-1-ktraynor@redhat.com>

Hi,

FYI, your patch has been queued to stable release 21.11.3

Note it hasn't been pushed to http://dpdk.org/browse/dpdk-stable yet.
It will be pushed if I get no objections before 11/01/22. So please
shout if anyone has objections.

Also note that after the patch there's a diff of the upstream commit vs the
patch applied to the branch. This will indicate if there was any rebasing
needed to apply to the stable branch. If there were code changes for rebasing
(ie: not only metadata diffs), please double check that the rebase was
correctly done.

Queued patches are on a temporary branch at:
https://github.com/kevintraynor/dpdk-stable

This queued commit can be viewed at:
https://github.com/kevintraynor/dpdk-stable/commit/5f0b3890df9cfc9192aa83c8a33f7001a40fd0d9

Thanks.

Kevin

---
From 5f0b3890df9cfc9192aa83c8a33f7001a40fd0d9 Mon Sep 17 00:00:00 2001
From: Zhichao Zeng <zhichaox.zeng@intel.com>
Date: Tue, 13 Sep 2022 17:28:41 +0800
Subject: [PATCH] net/iavf: fix processing VLAN TCI in SSE path

[ upstream commit 1aaacea174ebce7fa28b9fa375442f1ae7de4a5e ]

The SSE RX path does not process the vlan tci correctly when it's stored
in L2TAG2, so the vlan tci could not be extracted from descriptor,
then would not be put into mbuf either.

Add processing when vlan tci is stored in L2TAG2.

Fixes: 1162f5a0ef31 ("net/iavf: support flexible Rx descriptor in SSE path")

Signed-off-by: Zhichao Zeng <zhichaox.zeng@intel.com>
Acked-by: Qi Zhang <qi.z.zhang@intel.com>
---
 drivers/net/iavf/iavf_rxtx_vec_sse.c | 93 +++++++++++++++++++++++-----
 1 file changed, 78 insertions(+), 15 deletions(-)

diff --git a/drivers/net/iavf/iavf_rxtx_vec_sse.c b/drivers/net/iavf/iavf_rxtx_vec_sse.c
index 4b23ca8d82..15cef0d518 100644
--- a/drivers/net/iavf/iavf_rxtx_vec_sse.c
+++ b/drivers/net/iavf/iavf_rxtx_vec_sse.c
@@ -209,7 +209,13 @@ flex_rxd_to_fdir_flags_vec(const __m128i fdir_id0_3)
 }
 
+#ifndef RTE_LIBRTE_IAVF_16BYTE_RX_DESC
+static inline void
+flex_desc_to_olflags_v(struct iavf_rx_queue *rxq, __m128i descs[4], __m128i descs_bh[4],
+		       struct rte_mbuf **rx_pkts)
+#else
 static inline void
 flex_desc_to_olflags_v(struct iavf_rx_queue *rxq, __m128i descs[4],
 		       struct rte_mbuf **rx_pkts)
+#endif
 {
 	const __m128i mbuf_init = _mm_set_epi64x(0, rxq->mbuf_initializer);
@@ -287,4 +293,37 @@ flex_desc_to_olflags_v(struct iavf_rx_queue *rxq, __m128i descs[4],
 	flags = _mm_or_si128(flags, rss_vlan);
 
+#ifndef RTE_LIBRTE_IAVF_16BYTE_RX_DESC
+	if (rxq->rx_flags & IAVF_RX_FLAGS_VLAN_TAG_LOC_L2TAG2_2) {
+		const __m128i l2tag2_mask =
+			_mm_set1_epi32(1 << IAVF_RX_FLEX_DESC_STATUS1_L2TAG2P_S);
+
+		const __m128i vlan_tci0_1 =
+			_mm_unpacklo_epi32(descs_bh[0], descs_bh[1]);
+		const __m128i vlan_tci2_3 =
+			_mm_unpacklo_epi32(descs_bh[2], descs_bh[3]);
+		const __m128i vlan_tci0_3 =
+			_mm_unpacklo_epi64(vlan_tci0_1, vlan_tci2_3);
+
+		__m128i vlan_bits = _mm_and_si128(vlan_tci0_3, l2tag2_mask);
+
+		vlan_bits = _mm_srli_epi32(vlan_bits,
+						IAVF_RX_FLEX_DESC_STATUS1_L2TAG2P_S);
+
+		const __m128i vlan_flags_shuf =
+			_mm_set_epi8(0, 0, 0, 0,
+					0, 0, 0, 0,
+					0, 0, 0, 0,
+					0, 0,
+					RTE_MBUF_F_RX_VLAN |
+					RTE_MBUF_F_RX_VLAN_STRIPPED,
+					0);
+
+		const __m128i vlan_flags = _mm_shuffle_epi8(vlan_flags_shuf, vlan_bits);
+
+		/* merge with vlan_flags */
+		flags = _mm_or_si128(flags, vlan_flags);
+	}
+#endif
+
 	if (rxq->fdir_enabled) {
 		const __m128i fdir_id0_1 =
@@ -749,4 +788,7 @@ _recv_raw_pkts_vec_flex_rxd(struct iavf_rx_queue *rxq,
 	     rxdp += IAVF_VPMD_DESCS_PER_LOOP) {
 		__m128i descs[IAVF_VPMD_DESCS_PER_LOOP];
+#ifndef RTE_LIBRTE_IAVF_16BYTE_RX_DESC
+		__m128i descs_bh[IAVF_VPMD_DESCS_PER_LOOP];
+#endif
 		__m128i pkt_mb0, pkt_mb1, pkt_mb2, pkt_mb3;
 		__m128i staterr, sterr_tmp1, sterr_tmp2;
@@ -807,6 +849,4 @@ _recv_raw_pkts_vec_flex_rxd(struct iavf_rx_queue *rxq,
 		sterr_tmp1 = _mm_unpackhi_epi32(descs[1], descs[0]);
 
-		flex_desc_to_olflags_v(rxq, descs, &rx_pkts[pos]);
-
 		/* D.2 pkt 3,4 set in_port/nb_seg and remove crc */
 		pkt_mb3 = _mm_add_epi16(pkt_mb3, crc_adjust);
@@ -822,22 +862,21 @@ _recv_raw_pkts_vec_flex_rxd(struct iavf_rx_queue *rxq,
 		 * will cause performance drop to get into this context.
 		 */
-		if (offloads & RTE_ETH_RX_OFFLOAD_RSS_HASH) {
+		if (offloads & RTE_ETH_RX_OFFLOAD_RSS_HASH ||
+			rxq->rx_flags & IAVF_RX_FLAGS_VLAN_TAG_LOC_L2TAG2_2) {
 			/* load bottom half of every 32B desc */
-			const __m128i raw_desc_bh3 =
-				_mm_load_si128
+			descs_bh[3] = _mm_load_si128
 					((void *)(&rxdp[3].wb.status_error1));
 			rte_compiler_barrier();
-			const __m128i raw_desc_bh2 =
-				_mm_load_si128
+			descs_bh[2] = _mm_load_si128
 					((void *)(&rxdp[2].wb.status_error1));
 			rte_compiler_barrier();
-			const __m128i raw_desc_bh1 =
-				_mm_load_si128
+			descs_bh[1] = _mm_load_si128
 					((void *)(&rxdp[1].wb.status_error1));
 			rte_compiler_barrier();
-			const __m128i raw_desc_bh0 =
-				_mm_load_si128
+			descs_bh[0] = _mm_load_si128
 					((void *)(&rxdp[0].wb.status_error1));
+		}
 
+		if (offloads & RTE_ETH_RX_OFFLOAD_RSS_HASH) {
 			/**
 			 * to shift the 32b RSS hash value to the
@@ -845,11 +884,11 @@ _recv_raw_pkts_vec_flex_rxd(struct iavf_rx_queue *rxq,
 			 */
 			__m128i rss_hash3 =
-				_mm_slli_epi64(raw_desc_bh3, 32);
+				_mm_slli_epi64(descs_bh[3], 32);
 			__m128i rss_hash2 =
-				_mm_slli_epi64(raw_desc_bh2, 32);
+				_mm_slli_epi64(descs_bh[2], 32);
 			__m128i rss_hash1 =
-				_mm_slli_epi64(raw_desc_bh1, 32);
+				_mm_slli_epi64(descs_bh[1], 32);
 			__m128i rss_hash0 =
-				_mm_slli_epi64(raw_desc_bh0, 32);
+				_mm_slli_epi64(descs_bh[0], 32);
 
 			__m128i rss_hash_msk =
@@ -870,4 +909,28 @@ _recv_raw_pkts_vec_flex_rxd(struct iavf_rx_queue *rxq,
 			pkt_mb0 = _mm_or_si128(pkt_mb0, rss_hash0);
 		} /* if() on RSS hash parsing */
+
+		if (rxq->rx_flags & IAVF_RX_FLAGS_VLAN_TAG_LOC_L2TAG2_2) {
+			/* L2TAG2_2 */
+			__m128i vlan_tci3 = _mm_slli_si128(descs_bh[3], 4);
+			__m128i vlan_tci2 = _mm_slli_si128(descs_bh[2], 4);
+			__m128i vlan_tci1 = _mm_slli_si128(descs_bh[1], 4);
+			__m128i vlan_tci0 = _mm_slli_si128(descs_bh[0], 4);
+
+			const __m128i vlan_tci_msk = _mm_set_epi32(0, 0xFFFF0000, 0, 0);
+
+			vlan_tci3 = _mm_and_si128(vlan_tci3, vlan_tci_msk);
+			vlan_tci2 = _mm_and_si128(vlan_tci2, vlan_tci_msk);
+			vlan_tci1 = _mm_and_si128(vlan_tci1, vlan_tci_msk);
+			vlan_tci0 = _mm_and_si128(vlan_tci0, vlan_tci_msk);
+
+			pkt_mb3 = _mm_or_si128(pkt_mb3, vlan_tci3);
+			pkt_mb2 = _mm_or_si128(pkt_mb2, vlan_tci2);
+			pkt_mb1 = _mm_or_si128(pkt_mb1, vlan_tci1);
+			pkt_mb0 = _mm_or_si128(pkt_mb0, vlan_tci0);
+		}
+
+		flex_desc_to_olflags_v(rxq, descs, descs_bh, &rx_pkts[pos]);
+#else
+		flex_desc_to_olflags_v(rxq, descs, &rx_pkts[pos]);
 #endif
 
-- 
2.37.3

---
  Diff of the applied patch vs upstream commit (please double-check if non-empty:
---
--- -	2022-10-25 14:18:59.927938278 +0100
+++ 0061-net-iavf-fix-processing-VLAN-TCI-in-SSE-path.patch	2022-10-25 14:18:58.461798270 +0100
@@ -1 +1 @@
-From 1aaacea174ebce7fa28b9fa375442f1ae7de4a5e Mon Sep 17 00:00:00 2001
+From 5f0b3890df9cfc9192aa83c8a33f7001a40fd0d9 Mon Sep 17 00:00:00 2001
@@ -5,0 +6,2 @@
+[ upstream commit 1aaacea174ebce7fa28b9fa375442f1ae7de4a5e ]
+
@@ -13 +14,0 @@
-Cc: stable@dpdk.org
@@ -22 +23 @@
-index 4a5232c1d2..17eed21f70 100644
+index 4b23ca8d82..15cef0d518 100644


  parent reply	other threads:[~2022-10-25 15:09 UTC|newest]

Thread overview: 101+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2022-10-25 15:05 patch 'build: enable developer mode for all working trees' " Kevin Traynor
2022-10-25 15:05 ` patch 'net: accept unaligned data in checksum routines' " Kevin Traynor
2022-10-25 15:05 ` patch 'eal: fix side effect in some pointer arithmetic macros' " Kevin Traynor
2022-10-25 15:05 ` patch 'app/testpmd: restore ixgbe bypass commands' " Kevin Traynor
2022-10-25 15:06 ` patch 'net/bonding: fix array overflow in Rx burst' " Kevin Traynor
2022-10-25 15:06 ` patch 'net/bonding: fix double slave link status query' " Kevin Traynor
2022-10-25 15:06 ` patch 'net/failsafe: fix interrupt handle leak' " Kevin Traynor
2022-10-25 15:06 ` patch 'net/nfp: compose firmware file name with new hwinfo' " Kevin Traynor
2022-10-25 15:06 ` patch 'net/axgbe: fix scattered Rx' " Kevin Traynor
2022-10-25 15:06 ` patch 'net/axgbe: fix mbuf lengths in " Kevin Traynor
2022-10-25 15:06 ` patch 'net/axgbe: fix length of each segment " Kevin Traynor
2022-10-25 15:06 ` patch 'net/axgbe: fix checksum and RSS " Kevin Traynor
2022-10-25 15:06 ` patch 'net/axgbe: optimise " Kevin Traynor
2022-10-25 15:06 ` patch 'net/axgbe: remove freeing buffer in " Kevin Traynor
2022-10-25 15:06 ` patch 'net/nfp: improve HW info header log readability' " Kevin Traynor
2022-10-25 15:06 ` patch 'net/txgbe: fix IPv6 flow rule' " Kevin Traynor
2022-10-25 15:06 ` patch 'net/txgbe: remove semaphore between SW/FW' " Kevin Traynor
2022-10-25 15:06 ` patch 'net/txgbe: rename some extended statistics' " Kevin Traynor
2022-10-25 15:06 ` patch 'net/ngbe: " Kevin Traynor
2022-10-25 15:06 ` patch 'net/ngbe: remove semaphore between SW/FW' " Kevin Traynor
2022-10-25 15:06 ` patch 'net/ngbe: fix maximum frame size' " Kevin Traynor
2022-10-25 15:06 ` patch 'common/cnxk: fix log level during MCAM allocation' " Kevin Traynor
2022-10-25 15:06 ` patch 'net/mvneta: fix build with GCC 12' " Kevin Traynor
2022-10-25 15:06 ` patch 'common/cnxk: fix missing flow counter reset' " Kevin Traynor
2022-10-25 15:06 ` patch 'common/cnxk: fix printing disabled MKEX registers' " Kevin Traynor
2022-10-25 15:06 ` patch 'malloc: fix storage size for some allocations' " Kevin Traynor
2022-10-25 15:06 ` patch 'event/dsw: fix flow migration' " Kevin Traynor
2022-10-25 15:06 ` patch 'event/sw: fix device name in dump' " Kevin Traynor
2022-10-25 15:06 ` patch 'eventdev/eth_tx: add spinlock for adapter start/stop' " Kevin Traynor
2022-10-25 15:06 ` patch 'eventdev/eth_tx: fix adapter stop' " Kevin Traynor
2022-10-25 15:06 ` patch 'test/ipsec: skip if no compatible device' " Kevin Traynor
2022-10-25 15:06 ` patch 'examples/ipsec-secgw: use Tx checksum offload conditionally' " Kevin Traynor
2022-10-25 15:06 ` patch 'test/crypto: fix debug messages' " Kevin Traynor
2022-10-25 15:06 ` patch 'common/qat: fix VF to PF answer' " Kevin Traynor
2022-10-25 15:06 ` patch 'test/ipsec: fix build with GCC 12' " Kevin Traynor
2022-10-25 15:06 ` patch 'ipsec: " Kevin Traynor
2022-10-25 15:06 ` patch 'crypto/qat: " Kevin Traynor
2022-10-25 15:06 ` patch 'cryptodev: fix missing SHA3 algorithm strings' " Kevin Traynor
2022-10-25 15:06 ` patch 'eventdev: fix name of Rx conf type in documentation' " Kevin Traynor
2022-10-25 15:06 ` patch 'net/i40e: fix VF representor release' " Kevin Traynor
2022-10-25 15:06 ` patch 'net/ice: fix RSS hash update' " Kevin Traynor
2022-10-25 15:06 ` patch 'net/ice/base: fix inner symmetric RSS hash in raw flow' " Kevin Traynor
2022-10-25 15:06 ` patch 'net/iavf: fix L3 checksum Tx offload flag' " Kevin Traynor
2022-10-25 15:06 ` patch 'net/iavf: fix VLAN insertion' " Kevin Traynor
2022-10-25 15:06 ` patch 'net/iavf: fix pattern check for flow director parser' " Kevin Traynor
2022-10-25 15:06 ` patch 'net/iavf: fix Tx done descriptors cleanup' " Kevin Traynor
2022-10-25 15:06 ` patch 'net/iavf: update IPsec ESN values when updating session' " Kevin Traynor
2022-10-25 15:06 ` patch 'common/iavf: avoid copy in async mode' " Kevin Traynor
2022-10-25 15:06 ` patch 'net/ice/base: fix division during E822 PTP init' " Kevin Traynor
2022-10-25 15:06 ` patch 'net/ice/base: fix 100M speed capability' " Kevin Traynor
2022-10-25 15:06 ` patch 'net/ice/base: fix DSCP PFC TLV creation' " Kevin Traynor
2022-10-25 15:06 ` patch 'net/ice/base: fix media type of PHY 10G SFI C2C' " Kevin Traynor
2022-10-25 15:06 ` patch 'net/ice/base: fix function descriptions for parser' " Kevin Traynor
2022-10-25 15:06 ` patch 'net/ice/base: fix endian format' " Kevin Traynor
2022-10-25 15:06 ` patch 'net/ice/base: fix array overflow in add switch recipe' " Kevin Traynor
2022-10-25 15:06 ` patch 'net/ice/base: fix bit finding range over ptype bitmap' " Kevin Traynor
2022-10-25 15:06 ` patch 'net/ice/base: fix add MAC rule' " Kevin Traynor
2022-10-25 15:06 ` patch 'net/ice/base: fix double VLAN in promiscuous mode' " Kevin Traynor
2022-10-25 15:06 ` patch 'net/ice/base: ignore promiscuous already exist' " Kevin Traynor
2022-10-25 15:06 ` patch 'net/ice/base: fix input set of GTPoGRE' " Kevin Traynor
2022-10-25 15:06 ` Kevin Traynor [this message]
2022-10-25 15:06 ` patch 'net/iavf: fix outer checksum flags' " Kevin Traynor
2022-10-25 15:06 ` patch 'net/virtio: fix crash when configured twice' " Kevin Traynor
2022-10-25 15:06 ` patch 'net/mlx4: fix Verbs FD leak in secondary process' " Kevin Traynor
2022-10-25 15:07 ` patch 'net/mlx5: " Kevin Traynor
2022-10-25 15:07 ` patch 'net/mlx5: fix check for orphan wait descriptor' " Kevin Traynor
2022-10-25 15:07 ` patch 'net/mlx5: fix single not inline packet storing' " Kevin Traynor
2022-10-25 15:07 ` patch 'net/mlx5: fix inline length exceeding descriptor limit' " Kevin Traynor
2022-10-25 15:07 ` patch 'net/mlx5: fix Tx check for hardware descriptor length' " Kevin Traynor
2022-10-25 15:07 ` patch 'net/mlx5: fix modify action with tunnel decapsulation' " Kevin Traynor
2022-10-25 15:07 ` patch 'net/mlx5: fix meter profile delete after disable' " Kevin Traynor
2022-10-25 15:07 ` patch 'net/iavf: check illegal packet sizes' " Kevin Traynor
2022-10-25 15:07 ` patch 'net/ice: " Kevin Traynor
2022-10-25 15:07 ` patch 'net/cnxk: fix DF bit in vector mode' " Kevin Traynor
2022-10-25 15:07 ` patch 'net/axgbe: reset end of packet in scattered Rx' " Kevin Traynor
2022-10-25 15:07 ` patch 'net/axgbe: clear buffer on scattered Rx chaining failure' " Kevin Traynor
2022-10-25 15:07 ` patch 'net/axgbe: save segment data in scattered Rx' " Kevin Traynor
2022-10-25 15:07 ` patch 'common/sfc_efx/base: fix maximum Tx data count' " Kevin Traynor
2022-10-25 15:07 ` patch 'event/cnxk: fix missing xstats operations' " Kevin Traynor
2022-10-25 15:07 ` patch 'cryptodev: fix unduly newlines in logs' " Kevin Traynor
2022-10-25 15:07 ` patch 'net/bnxt: fix null pointer dereference in LED config' " Kevin Traynor
2022-11-17  9:05   ` 答复: " Mao,Yingming
2022-11-17 10:08     ` Kevin Traynor
2022-10-25 15:07 ` patch 'net/bnxt: fix error code during MTU change' " Kevin Traynor
2022-10-25 15:07 ` patch 'net/bnxt: remove unnecessary check' " Kevin Traynor
2022-10-25 15:07 ` patch 'net/bnxt: fix representor info freeing' " Kevin Traynor
2022-10-25 15:07 ` patch 'net/bnxt: fix build with GCC 13' " Kevin Traynor
2022-10-25 15:07 ` patch 'mem: fix API doc about allocation on secondary processes' " Kevin Traynor
2022-10-25 15:07 ` patch 'examples/vm_power_manager: use safe list iterator' " Kevin Traynor
2022-10-25 15:07 ` patch 'timer: fix stopping all timers' " Kevin Traynor
2022-10-25 15:07 ` patch 'vhost: fix build with GCC 12' " Kevin Traynor
2022-10-25 15:07 ` patch 'net/i40e: fix build with MinGW " Kevin Traynor
2022-10-25 15:07 ` patch 'net/qede/base: fix 32-bit build with " Kevin Traynor
2022-10-25 15:07 ` patch 'net/tap: fix overflow of network interface index' " Kevin Traynor
2022-10-25 15:07 ` patch 'net/memif: fix crash with different number of Rx/Tx queues' " Kevin Traynor
2022-10-25 15:07 ` patch 'common/sfc_efx/base: remove VQ index check during VQ start' " Kevin Traynor
2022-10-25 15:07 ` patch 'net/hns3: fix Rx with PTP' " Kevin Traynor
2022-10-25 15:07 ` patch 'net/hns3: fix crash in SVE Tx' " Kevin Traynor
2022-10-25 15:07 ` patch 'net/hns3: fix next-to-use overflow " Kevin Traynor
2022-10-25 15:07 ` patch 'net/hns3: fix next-to-use overflow in simple " Kevin Traynor
2022-10-25 15:07 ` patch 'net/hns3: fix crash when secondary process access FW' " Kevin Traynor

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=20221025150734.142189-61-ktraynor@redhat.com \
    --to=ktraynor@redhat.com \
    --cc=qi.z.zhang@intel.com \
    --cc=stable@dpdk.org \
    --cc=zhichaox.zeng@intel.com \
    /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).