DPDK patches and discussions
 help / color / mirror / Atom feed
From: scott.k.mitch1@gmail.com
To: dev@dpdk.org
Cc: mb@smartsharesystems.com, stephen@networkplumber.org,
	Scott <scott_mitchell@apple.com>,
	Scott Mitchell <scott.k.mitch1@gmail.com>
Subject: [PATCH v12 3/3] eal/net: add workaround for GCC optimization bug
Date: Fri,  9 Jan 2026 20:56:51 -0500	[thread overview]
Message-ID: <20260110015651.26201-4-scott.k.mitch1@gmail.com> (raw)
In-Reply-To: <20260110015651.26201-1-scott.k.mitch1@gmail.com>

From: Scott <scott_mitchell@apple.com>

GCC has a bug where it incorrectly elides struct initialization in
inline functions when strict aliasing is enabled (-O2/-O3/-Os), causing
reads from uninitialized memory. This affects both designated initializers
and manual field assignment.

Add RTE_FORCE_INIT_BARRIER macro that uses an asm volatile memory barrier
to prevent the compiler from incorrectly optimizing away struct
initialization. Apply the workaround to pseudo-header checksum functions
in rte_ip4.h, rte_ip6.h, hinic driver, and mlx5 driver.

Signed-off-by: Scott Mitchell <scott.k.mitch1@gmail.com>
---
 drivers/net/hinic/hinic_pmd_tx.c |  2 ++
 drivers/net/mlx5/mlx5_flow_dv.c  |  2 ++
 lib/eal/include/rte_common.h     | 14 ++++++++++++++
 lib/net/rte_ip4.h                |  1 +
 lib/net/rte_ip6.h                |  1 +
 5 files changed, 20 insertions(+)

diff --git a/drivers/net/hinic/hinic_pmd_tx.c b/drivers/net/hinic/hinic_pmd_tx.c
index 22fb0bffaf..570715531d 100644
--- a/drivers/net/hinic/hinic_pmd_tx.c
+++ b/drivers/net/hinic/hinic_pmd_tx.c
@@ -725,6 +725,7 @@ hinic_ipv4_phdr_cksum(const struct rte_ipv4_hdr *ipv4_hdr, uint64_t ol_flags)
 		rte_cpu_to_be_16(rte_be_to_cpu_16(ipv4_hdr->total_length) -
 				 rte_ipv4_hdr_len(ipv4_hdr));
 	}
+	RTE_FORCE_INIT_BARRIER(psd_hdr);
 	return rte_raw_cksum(&psd_hdr, sizeof(psd_hdr));
 }
 
@@ -743,6 +744,7 @@ hinic_ipv6_phdr_cksum(const struct rte_ipv6_hdr *ipv6_hdr, uint64_t ol_flags)
 	else
 		psd_hdr.len = ipv6_hdr->payload_len;
 
+	RTE_FORCE_INIT_BARRIER(psd_hdr);
 	sum = __rte_raw_cksum(&ipv6_hdr->src_addr,
 		sizeof(ipv6_hdr->src_addr) + sizeof(ipv6_hdr->dst_addr), 0);
 	sum = __rte_raw_cksum(&psd_hdr, sizeof(psd_hdr), sum);
diff --git a/drivers/net/mlx5/mlx5_flow_dv.c b/drivers/net/mlx5/mlx5_flow_dv.c
index 47f6d28410..1eeeb6747f 100644
--- a/drivers/net/mlx5/mlx5_flow_dv.c
+++ b/drivers/net/mlx5/mlx5_flow_dv.c
@@ -4445,6 +4445,8 @@ __flow_encap_decap_resource_register(struct rte_eth_dev *dev,
 			.reserve = 0,
 		}
 	};
+	RTE_FORCE_INIT_BARRIER(encap_decap_key);
+
 	struct mlx5_flow_cb_ctx ctx = {
 		.error = error,
 		.data = resource,
diff --git a/lib/eal/include/rte_common.h b/lib/eal/include/rte_common.h
index 00d428e295..d58e856d96 100644
--- a/lib/eal/include/rte_common.h
+++ b/lib/eal/include/rte_common.h
@@ -555,6 +555,20 @@ static void __attribute__((destructor(RTE_PRIO(prio)), used)) func(void)
 #define __rte_no_ubsan_alignment
 #endif
 
+/**
+ * Force struct initialization to prevent GCC optimization bug.
+ * GCC has a bug where it incorrectly elides struct initialization in
+ * inline functions when strict aliasing is enabled, causing reads from
+ * uninitialized memory. This memory barrier prevents the misoptimization.
+ */
+#ifdef RTE_CC_GCC
+#define RTE_FORCE_INIT_BARRIER(var) do {	\
+	asm volatile("" : "+m" (var));		\
+} while (0)
+#else
+#define RTE_FORCE_INIT_BARRIER(var)
+#endif
+
 /*********** Macros for pointer arithmetic ********/
 
 /**
diff --git a/lib/net/rte_ip4.h b/lib/net/rte_ip4.h
index 822a660cfb..a0839c584e 100644
--- a/lib/net/rte_ip4.h
+++ b/lib/net/rte_ip4.h
@@ -238,6 +238,7 @@ rte_ipv4_phdr_cksum(const struct rte_ipv4_hdr *ipv4_hdr, uint64_t ol_flags)
 		psd_hdr.len = rte_cpu_to_be_16((uint16_t)(l3_len -
 			rte_ipv4_hdr_len(ipv4_hdr)));
 	}
+	RTE_FORCE_INIT_BARRIER(psd_hdr);
 	return rte_raw_cksum(&psd_hdr, sizeof(psd_hdr));
 }
 
diff --git a/lib/net/rte_ip6.h b/lib/net/rte_ip6.h
index d1abf1f5d5..902f100a44 100644
--- a/lib/net/rte_ip6.h
+++ b/lib/net/rte_ip6.h
@@ -572,6 +572,7 @@ rte_ipv6_phdr_cksum(const struct rte_ipv6_hdr *ipv6_hdr, uint64_t ol_flags)
 	else
 		psd_hdr.len = ipv6_hdr->payload_len;
 
+	RTE_FORCE_INIT_BARRIER(psd_hdr);
 	sum = __rte_raw_cksum(&ipv6_hdr->src_addr,
 		sizeof(ipv6_hdr->src_addr) + sizeof(ipv6_hdr->dst_addr),
 		0);
-- 
2.39.5 (Apple Git-154)


  parent reply	other threads:[~2026-01-10  1:57 UTC|newest]

Thread overview: 10+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2026-01-10  1:56 [PATCH v12 0/3] net: optimize raw checksum computation scott.k.mitch1
2026-01-10  1:56 ` [PATCH v12 1/3] net: optimize __rte_raw_cksum and add tests scott.k.mitch1
2026-01-10  2:28   ` Scott Mitchell
2026-01-10 14:47   ` Morten Brørup
2026-01-10  1:56 ` [PATCH v12 2/3] eal: add workaround for UBSAN alignment false positive scott.k.mitch1
2026-01-10 15:02   ` Morten Brørup
2026-01-10  1:56 ` scott.k.mitch1 [this message]
2026-01-10 15:29   ` [PATCH v12 3/3] eal/net: add workaround for GCC optimization bug Morten Brørup
2026-01-11  6:21     ` Scott Mitchell
2026-01-10 16:59 ` [PATCH v12 0/3] net: optimize raw checksum computation Stephen Hemminger

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=20260110015651.26201-4-scott.k.mitch1@gmail.com \
    --to=scott.k.mitch1@gmail.com \
    --cc=dev@dpdk.org \
    --cc=mb@smartsharesystems.com \
    --cc=scott_mitchell@apple.com \
    --cc=stephen@networkplumber.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).