selftests/bpf: Fix null pointer check in skb_pkt_end.c
Subsystems (3)
PATCH 1/1 selftests/bpf: Fix null pointer check in skb_pkt_end.c
Ensure that 'tcp' is checked for NULL before dereferencing. This resolves a potential null pointer dereference warning reported by static analysis.
--- tools/testing/selftests/bpf/progs/skb_pkt_end.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-)
tools/testing/selftests/bpf/progs/skb_pkt_end.c +2 -2
@@ -45,10 +45,10 @@ int main_prog(struct __sk_buff *skb) goto out; tcp = (void *)(ip +1 ); -if (tcp->dest !=0 ) - goto out;if (!tcp) goto out; +if (tcp->dest !=0 ) + goto out; urg_ptr = tcp->urg_ptr;
Signature · 1 line
2.34.1
Martin KaFai Lau
reviewerQuoted 13 lines · Prabhav Kumar Vaish
Ensure that 'tcp' is checked for NULL before dereferencing. This resolves a potential null pointer dereference warning reported by static analysis.
Signed-off-by Prabhav Kumar Vaish <[email protected]>--- tools/testing/selftests/bpf/progs/skb_pkt_end.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-)tools/testing/selftests/bpf/progs/skb_pkt_end.c +0
-0
tools/testing/selftests/bpf/progs/skb_pkt_end.c int main_prog(struct __sk_buff *skb) · line 48
Quoted 6 lines
tools/testing/selftests/bpf/progs/skb_pkt_end.c +0
-2@@ -45,10 +45,10 @@ int main_prog(struct __sk_buff *skb) goto out; tcp = (void *)(ip +1 ); -if (tcp->dest !=0 ) - goto out;if (!tcp)
This case will never be hit, so this change is not doing anything other than silencing the static checker. Take a look at commit 9cc873e85800 ("selftests/bpf: Add skb_pkt_end test"). The test was written to have a specific llvm generated code. You will need to check the generated code is still testing what it is supposed to test.
pw-bot: cr
Quoted 6 lines
tools/testing/selftests/bpf/progs/skb_pkt_end.c +2
-0goto out; +if (tcp->dest !=0 ) + goto out; urg_ptr = tcp->urg_ptr;
Re: [PATCH] selftests/bpf: Fix potential null pointer dereference in skb_pkt_end.c
From: Prabhav Kumar <[email protected]>
Hi Martin,
Thank you for the feedback.
You're right — I reviewed commit 9cc873e85800 and now understand that the test is designed to validate specific LLVM code generation, not just runtime behavior. I see how my change, although addressing the warning, could unintentionally alter the generated instructions and defeat the purpose of the test.
I'll drop this patch to preserve the original test intent. Thanks again for the context!
Best Regards,
Prabhav