From: luca.boccassi@gmail.com
To: Dariusz Sosnowski <dsosnowski@nvidia.com>
Cc: Bing Zhao <bingz@nvidia.com>, dpdk stable <stable@dpdk.org>
Subject: patch 'net/mlx5: fix header modify action on group 0' has been queued to stable release 22.11.9
Date: Thu, 12 Jun 2025 22:06:36 +0100 [thread overview]
Message-ID: <20250612210733.2506558-19-luca.boccassi@gmail.com> (raw)
In-Reply-To: <20250612210733.2506558-1-luca.boccassi@gmail.com>
Hi,
FYI, your patch has been queued to stable release 22.11.9
Note it hasn't been pushed to http://dpdk.org/browse/dpdk-stable yet.
It will be pushed if I get no objections before 06/14/25. 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/bluca/dpdk-stable
This queued commit can be viewed at:
https://github.com/bluca/dpdk-stable/commit/0f7cba98cac32c3592272d5d493398df38cfa538
Thanks.
Luca Boccassi
---
From 0f7cba98cac32c3592272d5d493398df38cfa538 Mon Sep 17 00:00:00 2001
From: Dariusz Sosnowski <dsosnowski@nvidia.com>
Date: Fri, 25 Apr 2025 21:32:57 +0200
Subject: [PATCH] net/mlx5: fix header modify action on group 0
[ upstream commit f1ebb26d0d9388b6df65d493b06ce3b55adfb893 ]
HW modify header commands generated for multiple modify field
flow actions, which modify/access the same packet fields
do not have to be separated by NOPs when used on group 0.
This is because:
- On group > 0, HW uses Modify Header Pattern objects
which require NOP explicitly.
- On group 0, modify field action is implemented using
Modify Header Context object managed by FW.
FW inserts requires NOPs internally.
mlx5 PMD inserted NOP always, which caused flow/table creation
failures on group 0 flow rules.
This patch addresses that.
Fixes: 0f4aa72b99da ("net/mlx5: support flow modify field with HWS")
Signed-off-by: Dariusz Sosnowski <dsosnowski@nvidia.com>
Acked-by: Bing Zhao <bingz@nvidia.com>
---
drivers/net/mlx5/mlx5_flow_hw.c | 21 ++++++++++++++++-----
1 file changed, 16 insertions(+), 5 deletions(-)
diff --git a/drivers/net/mlx5/mlx5_flow_hw.c b/drivers/net/mlx5/mlx5_flow_hw.c
index 0f36700143..7040dd863d 100644
--- a/drivers/net/mlx5/mlx5_flow_hw.c
+++ b/drivers/net/mlx5/mlx5_flow_hw.c
@@ -856,7 +856,8 @@ flow_hw_action_modify_field_is_shared(const struct rte_flow_action *action,
static __rte_always_inline bool
flow_hw_should_insert_nop(const struct mlx5_hw_modify_header_action *mhdr,
- const struct mlx5_modification_cmd *cmd)
+ const struct mlx5_modification_cmd *cmd,
+ const struct rte_flow_attr *attr)
{
struct mlx5_modification_cmd last_cmd = { { 0 } };
struct mlx5_modification_cmd new_cmd = { { 0 } };
@@ -864,6 +865,15 @@ flow_hw_should_insert_nop(const struct mlx5_hw_modify_header_action *mhdr,
unsigned int last_type;
bool should_insert = false;
+ /*
+ * Modify header action list does not require NOPs in root table,
+ * because different type of underlying object is used:
+ * - in root table - MODIFY_HEADER_CONTEXT (does not require NOPs),
+ * - in non-root - either inline modify action or based on Modify Header Pattern
+ * (which requires NOPs).
+ */
+ if (attr->group == 0)
+ return false;
if (cmds_num == 0)
return false;
last_cmd = *(&mhdr->mhdr_cmds[cmds_num - 1]);
@@ -939,7 +949,8 @@ flow_hw_mhdr_cmd_append(struct mlx5_hw_modify_header_action *mhdr,
static __rte_always_inline int
flow_hw_converted_mhdr_cmds_append(struct mlx5_hw_modify_header_action *mhdr,
- struct mlx5_flow_dv_modify_hdr_resource *resource)
+ struct mlx5_flow_dv_modify_hdr_resource *resource,
+ const struct rte_flow_attr *attr)
{
uint32_t idx;
int ret;
@@ -947,7 +958,7 @@ flow_hw_converted_mhdr_cmds_append(struct mlx5_hw_modify_header_action *mhdr,
for (idx = 0; idx < resource->actions_num; ++idx) {
struct mlx5_modification_cmd *src = &resource->actions[idx];
- if (flow_hw_should_insert_nop(mhdr, src)) {
+ if (flow_hw_should_insert_nop(mhdr, src, attr)) {
ret = flow_hw_mhdr_cmd_nop_append(mhdr);
if (ret)
return ret;
@@ -1057,14 +1068,14 @@ flow_hw_modify_field_compile(struct rte_eth_dev *dev,
* This NOP command will not be a part of action's command range used to update commands
* on rule creation.
*/
- if (flow_hw_should_insert_nop(mhdr, &resource->actions[0])) {
+ if (flow_hw_should_insert_nop(mhdr, &resource->actions[0], attr)) {
ret = flow_hw_mhdr_cmd_nop_append(mhdr);
if (ret)
return rte_flow_error_set(error, ret, RTE_FLOW_ERROR_TYPE_UNSPECIFIED,
NULL, "too many modify field operations specified");
}
cmds_start = mhdr->mhdr_cmds_num;
- ret = flow_hw_converted_mhdr_cmds_append(mhdr, resource);
+ ret = flow_hw_converted_mhdr_cmds_append(mhdr, resource, attr);
if (ret)
return rte_flow_error_set(error, ret, RTE_FLOW_ERROR_TYPE_UNSPECIFIED,
NULL, "too many modify field operations specified");
--
2.47.2
---
Diff of the applied patch vs upstream commit (please double-check if non-empty:
---
--- - 2025-06-12 22:06:24.592330271 +0100
+++ 0019-net-mlx5-fix-header-modify-action-on-group-0.patch 2025-06-12 22:06:23.814043456 +0100
@@ -1 +1 @@
-From f1ebb26d0d9388b6df65d493b06ce3b55adfb893 Mon Sep 17 00:00:00 2001
+From 0f7cba98cac32c3592272d5d493398df38cfa538 Mon Sep 17 00:00:00 2001
@@ -5,0 +6,2 @@
+[ upstream commit f1ebb26d0d9388b6df65d493b06ce3b55adfb893 ]
+
@@ -22 +23,0 @@
-Cc: stable@dpdk.org
@@ -31 +32 @@
-index 2f7b769aec..cbe6bd3ef9 100644
+index 0f36700143..7040dd863d 100644
@@ -34 +35 @@
-@@ -1422,7 +1422,8 @@ flow_hw_action_modify_field_is_shared(const struct rte_flow_action *action,
+@@ -856,7 +856,8 @@ flow_hw_action_modify_field_is_shared(const struct rte_flow_action *action,
@@ -44 +45 @@
-@@ -1430,6 +1431,15 @@ flow_hw_should_insert_nop(const struct mlx5_hw_modify_header_action *mhdr,
+@@ -864,6 +865,15 @@ flow_hw_should_insert_nop(const struct mlx5_hw_modify_header_action *mhdr,
@@ -60 +61 @@
-@@ -1508,7 +1518,8 @@ flow_hw_mhdr_cmd_append(struct mlx5_hw_modify_header_action *mhdr,
+@@ -939,7 +949,8 @@ flow_hw_mhdr_cmd_append(struct mlx5_hw_modify_header_action *mhdr,
@@ -70 +71 @@
-@@ -1516,7 +1527,7 @@ flow_hw_converted_mhdr_cmds_append(struct mlx5_hw_modify_header_action *mhdr,
+@@ -947,7 +958,7 @@ flow_hw_converted_mhdr_cmds_append(struct mlx5_hw_modify_header_action *mhdr,
@@ -79 +80 @@
-@@ -1639,14 +1650,14 @@ flow_hw_modify_field_compile(struct rte_eth_dev *dev,
+@@ -1057,14 +1068,14 @@ flow_hw_modify_field_compile(struct rte_eth_dev *dev,
next prev parent reply other threads:[~2025-06-12 21:08 UTC|newest]
Thread overview: 72+ messages / expand[flat|nested] mbox.gz Atom feed top
2025-06-12 21:06 patch 'net/i40e/base: remove unused variables' " luca.boccassi
2025-06-12 21:06 ` patch 'ci: bump tested distributions in GHA' " luca.boccassi
2025-06-12 21:06 ` patch 'net/fm10k/base: fix compilation warnings' " luca.boccassi
2025-06-12 21:06 ` patch 'net/ixgbe/base: correct definition of endianness macro' " luca.boccassi
2025-06-12 21:06 ` patch 'net/ixgbe/base: fix compilation warnings' " luca.boccassi
2025-06-12 21:06 ` patch 'net/i40e/base: fix unused value " luca.boccassi
2025-06-12 21:06 ` patch 'net/i40e/base: fix compiler " luca.boccassi
2025-06-12 21:06 ` patch 'acl: fix build with GCC 15 on aarch64' " luca.boccassi
2025-06-12 21:06 ` patch 'eal/linux: improve ASLR check' " luca.boccassi
2025-06-12 21:06 ` patch 'net/e1000: fix xstats name' " luca.boccassi
2025-06-12 21:06 ` patch 'net/e1000: fix EEPROM dump' " luca.boccassi
2025-06-12 21:06 ` patch 'net/ixgbe: fix port mask default value in filter' " luca.boccassi
2025-06-12 21:06 ` patch 'net/e1000: fix igb Tx queue offloads capability' " luca.boccassi
2025-06-12 21:06 ` patch 'vhost/crypto: fix cipher data length' " luca.boccassi
2025-06-12 21:06 ` patch 'crypto/virtio: fix cipher data source " luca.boccassi
2025-06-12 21:06 ` patch 'app/crypto-perf: fix AAD offset alignment' " luca.boccassi
2025-06-12 21:06 ` patch 'crypto/qat: fix out-of-place header bytes in AEAD raw API' " luca.boccassi
2025-06-12 21:06 ` patch 'crypto/qat: fix out-of-place chain/cipher/auth headers' " luca.boccassi
2025-06-12 21:06 ` luca.boccassi [this message]
2025-06-12 21:06 ` patch 'net/mlx5: validate GTP PSC QFI width' " luca.boccassi
2025-06-12 21:06 ` patch 'net/mlx5: fix counter service cleanup on init failure' " luca.boccassi
2025-06-12 21:06 ` patch 'net/mlx5: remove unsupported flow meter action in HWS' " luca.boccassi
2025-06-12 21:06 ` patch 'net/mlx5: fix maximal queue size query' " luca.boccassi
2025-06-12 21:06 ` patch 'net/mlx5: align PF and VF/SF MAC address handling' " luca.boccassi
2025-06-12 21:06 ` patch 'app/testpmd: fix RSS hash key update' " luca.boccassi
2025-06-12 21:06 ` patch 'net/af_xdp: fix use after free in zero-copy Tx' " luca.boccassi
2025-06-12 21:06 ` patch 'net/hns3: fix integer overflow in interrupt unmap' " luca.boccassi
2025-06-12 21:06 ` patch 'net/hns3: fix memory leak on failure' " luca.boccassi
2025-06-12 21:06 ` patch 'net/hns3: fix extra wait for link up' " luca.boccassi
2025-06-12 21:06 ` patch 'net/hns3: fix memory leak for indirect flow action' " luca.boccassi
2025-06-12 21:06 ` patch 'net/hns3: fix interrupt rollback' " luca.boccassi
2025-06-12 21:06 ` patch 'net/hns3: fix divide by zero' " luca.boccassi
2025-06-12 21:06 ` patch 'net/hns3: fix resources release on reset' " luca.boccassi
2025-06-12 21:06 ` patch 'net/qede: fix use after free' " luca.boccassi
2025-06-12 21:06 ` patch 'bus/fslmc: " luca.boccassi
2025-06-12 21:06 ` patch 'net/null: fix packet copy' " luca.boccassi
2025-06-12 21:06 ` patch 'bus/vmbus: align ring buffer data to page boundary' " luca.boccassi
2025-06-12 21:06 ` patch 'bus/vmbus: use Hyper-V page size' " luca.boccassi
2025-06-12 21:06 ` patch 'net/netvsc: " luca.boccassi
2025-06-12 21:06 ` patch 'net/netvsc: add stats counters from VF' " luca.boccassi
2025-06-12 21:06 ` patch 'app/testpmd: relax number of TCs in DCB command' " luca.boccassi
2025-06-12 21:06 ` patch 'net/mana: check vendor ID when probing RDMA device' " luca.boccassi
2025-06-12 21:07 ` patch 'net/hns3: fix CRC data segment' " luca.boccassi
2025-06-12 21:07 ` patch 'net/tap: fix qdisc add failure handling' " luca.boccassi
2025-06-12 21:07 ` patch 'net/mlx5: fix VLAN stripping on hairpin queue' " luca.boccassi
2025-06-12 21:07 ` patch 'mem: fix lockup on address space shortage' " luca.boccassi
2025-06-12 21:07 ` patch 'test/malloc: improve resiliency' " luca.boccassi
2025-06-12 21:07 ` patch 'trace: fix overflow in per-lcore trace buffer' " luca.boccassi
2025-06-12 21:07 ` patch 'common/cnxk: fix E-tag pattern parsing' " luca.boccassi
2025-06-12 21:07 ` patch 'common/cnxk: fix CQ tail drop' " luca.boccassi
2025-06-12 21:07 ` patch 'net/cnxk: fix descriptor count update on reconfig' " luca.boccassi
2025-06-12 21:07 ` patch 'ethdev: fix error struct in flow configure' " luca.boccassi
2025-06-12 21:07 ` patch 'net/ice/base: fix integer overflow' " luca.boccassi
2025-06-12 21:07 ` patch 'net/ice/base: fix typo in device ID description' " luca.boccassi
2025-06-12 21:07 ` patch 'common/dpaax: fix PDCP key command race condition' " luca.boccassi
2025-06-12 21:07 ` patch 'common/dpaax: fix PDCP AES only 12-bit SN' " luca.boccassi
2025-06-12 21:07 ` patch 'crypto/virtio: add request check on request side' " luca.boccassi
2025-06-12 21:07 ` patch 'crypto/virtio: fix driver cleanup' " luca.boccassi
2025-06-12 21:07 ` patch 'ethdev: keep promiscuous/allmulti value before disabling' " luca.boccassi
2025-06-12 21:07 ` patch 'eal: fix return value of lcore role' " luca.boccassi
2025-06-12 21:07 ` patch 'eal: warn if no lcore is available' " luca.boccassi
2025-06-12 21:07 ` patch 'bus: cleanup device lists' " luca.boccassi
2025-06-12 21:07 ` patch 'eal/linux: unregister alarm callback before free' " luca.boccassi
2025-06-12 21:07 ` patch 'eal/freebsd: " luca.boccassi
2025-06-12 21:07 ` patch 'bus/pci/bsd: fix device existence check' " luca.boccassi
2025-06-12 21:07 ` patch 'power/intel_uncore: fix crash closing uninitialized driver' " luca.boccassi
2025-06-12 21:07 ` patch 'pcapng: fix null dereference in close' " luca.boccassi
2025-06-12 21:07 ` patch 'net/mlx5: avoid setting kernel MTU if not needed' " luca.boccassi
2025-06-12 21:07 ` patch 'net/mlx5: fix hypervisor detection in VLAN workaround' " luca.boccassi
2025-06-12 21:07 ` patch 'net/hns3: check requirement for hardware GRO' " luca.boccassi
2025-06-12 21:07 ` patch 'net/hns3: allow Tx vector when fast free not enabled' " luca.boccassi
2025-06-12 21:07 ` patch 'net/hns3: allow Rx vector mode with VLAN filter' " luca.boccassi
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=20250612210733.2506558-19-luca.boccassi@gmail.com \
--to=luca.boccassi@gmail.com \
--cc=bingz@nvidia.com \
--cc=dsosnowski@nvidia.com \
--cc=stable@dpdk.org \
/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).