From: Fang Xieyan <fangxy@xiaopeng.com>
To: netdev@vger.kernel.org, Jamal Hadi Salim <jhs@mojatatu.com>,
Jiri Pirko <jiri@resnulli.us>,
"David S . Miller" <davem@davemloft.net>,
Eric Dumazet <edumazet@google.com>,
Jakub Kicinski <kuba@kernel.org>, Paolo Abeni <pabeni@redhat.com>
Cc: Simon Horman <horms@kernel.org>, Cong Wang <xiyou.wangcong@gmail.com>
Subject: [PATCH net] net/sched: act_ife: validate metadata length before decoding
Date: Sun, 20 Sep 2026 15:42:45 +0800 [thread overview]
Message-ID: <20260920074245.79965-1-fangxy@xiaopeng.com> (raw)
skbmark_decode(), skbprio_decode() and skbtcindex_decode() read fixed-size
values from the TLV payload without validating its length.
A malformed IFE frame can declare a shorter payload, causing the decoders
to consume bytes beyond the declared metadata value:
[TLV type=IFE_META_SKBMARK len=4]
-> dlen == 0, but decode reads 4 bytes
The decoder may therefore set skb metadata from unintended input.
Validate the payload length before decoding and return -EINVAL for
invalid lengths. The caller treats decoder failures as per-TLV errors
and continues decoding the remaining metadata.
Fixes: 084e2f6566d2 ("Support to encoding decoding skb mark on IFE action")
Fixes: 200e10f46936 ("Support to encoding decoding skb prio on IFE action")
Fixes: 408fbc22ef1e ("net sched ife action: Introduce skb tcindex metadata encap decap")
Assisted-by: Hawkeye:GLM-5.3-flash
Assisted-by: Qoder:Qwen3.8-Max
Signed-off-by: Fang Xieyan <fangxy@xiaopeng.com>
---
Found by auditing the IFE decode path at v6.18-rc7. The malformed
metadata cases were reproduced with a userspace sanitizer model of
the decode path, which reported reads of size 4/2 for dlen values
smaller than the expected metadata size.
Compile-tested on x86_64 with defconfig and NET_ACT_IFE=y: act_ife.o
and the three act_meta_*.o build warning-free.
net/sched/act_meta_mark.c | 6 +++++-
net/sched/act_meta_skbprio.c | 6 +++++-
net/sched/act_meta_skbtcindex.c | 6 +++++-
3 files changed, 15 insertions(+), 3 deletions(-)
diff --git a/net/sched/act_meta_mark.c b/net/sched/act_meta_mark.c
index ea0573cb8b2d..3a7d568df0da 100644
--- a/net/sched/act_meta_mark.c
+++ b/net/sched/act_meta_mark.c
@@ -28,8 +28,12 @@ static int skbmark_encode(struct sk_buff *skb, void *skbdata,
static int skbmark_decode(struct sk_buff *skb, void *data, u16 len)
{
- u32 ifemark = *(u32 *)data;
+ u32 ifemark;
+ if (len != sizeof(ifemark))
+ return -EINVAL;
+
+ ifemark = *(u32 *)data;
skb->mark = ntohl(ifemark);
return 0;
}
diff --git a/net/sched/act_meta_skbprio.c b/net/sched/act_meta_skbprio.c
index 2df3133ce5ad..fa4ba4269ca0 100644
--- a/net/sched/act_meta_skbprio.c
+++ b/net/sched/act_meta_skbprio.c
@@ -33,8 +33,12 @@ static int skbprio_encode(struct sk_buff *skb, void *skbdata,
static int skbprio_decode(struct sk_buff *skb, void *data, u16 len)
{
- u32 ifeprio = *(u32 *)data;
+ u32 ifeprio;
+ if (len != sizeof(ifeprio))
+ return -EINVAL;
+
+ ifeprio = *(u32 *)data;
skb->priority = ntohl(ifeprio);
return 0;
}
diff --git a/net/sched/act_meta_skbtcindex.c b/net/sched/act_meta_skbtcindex.c
index 44547caead46..48f001ada8f6 100644
--- a/net/sched/act_meta_skbtcindex.c
+++ b/net/sched/act_meta_skbtcindex.c
@@ -28,8 +28,12 @@ static int skbtcindex_encode(struct sk_buff *skb, void *skbdata,
static int skbtcindex_decode(struct sk_buff *skb, void *data, u16 len)
{
- u16 ifetc_index = *(u16 *)data;
+ u16 ifetc_index;
+ if (len != sizeof(ifetc_index))
+ return -EINVAL;
+
+ ifetc_index = *(u16 *)data;
skb->tc_index = ntohs(ifetc_index);
return 0;
}
--
2.50.1
next reply other threads:[~2026-09-20 7:42 UTC|newest]
Thread overview: 2+ messages / expand[flat|nested] mbox.gz Atom feed top
2026-09-20 7:42 Fang Xieyan [this message]
2026-09-21 7:45 ` [PATCH net] net/sched: act_ife: validate metadata length before decoding netdev-bot+sashiko
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=20260920074245.79965-1-fangxy@xiaopeng.com \
--to=fangxy@xiaopeng.com \
--cc=davem@davemloft.net \
--cc=edumazet@google.com \
--cc=horms@kernel.org \
--cc=jhs@mojatatu.com \
--cc=jiri@resnulli.us \
--cc=kuba@kernel.org \
--cc=netdev@vger.kernel.org \
--cc=pabeni@redhat.com \
--cc=xiyou.wangcong@gmail.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