From: Vineet Gupta <vineet.gupta@linux.dev>
To: sashiko-reviews@lists.linux.dev
Cc: bpf@vger.kernel.org
Subject: Re: [PATCH bpf-next v2 3/3] selftests/bpf: Enable some of the blocked cpuv4 instruction tests for bpf-gcc
Date: Mon, 24 Aug 2026 14:30:57 -0700 [thread overview]
Message-ID: <d94a8330-79d4-485d-ae46-cc93c0ceac5e@linux.dev> (raw)
In-Reply-To: <20260824211401.3C3D91F000E9@smtp.kernel.org>
On 8/24/26 2:14 PM, sashiko-bot@kernel.org wrote:
> Thank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider:
> - [Medium] The newly added `#ifdef __BPF_FEATURE_ADDR_SPACE_CAST` guard unintentionally drops 5 arena tests from the default test_progs flavour for clang.
> --
>
> --- Patch [3]: [PATCH bpf-next v2 3/3] selftests/bpf: Enable some of the blocked cpuv4 instruction tests for bpf-gcc ---
> commit ad6608dfd47ddd6a7068cdc6e0e6d467ec3558b8
> Author: Vineet Gupta <vineet.gupta@linux.dev>
>
> selftests/bpf: Enable some of the blocked cpuv4 instruction tests for bpf-gcc
>
> This commit relaxes clang-only checks in BPF selftests by adding specific
> feature checks like __BPF_FEATURE_LDSX to allow building under bpf-gcc.
> It also wraps arena tests with __BPF_FEATURE_ADDR_SPACE_CAST to skip them
> on bpf-gcc.
>
>> diff --git a/tools/testing/selftests/bpf/progs/verifier_ldsx.c b/tools/testing/selftests/bpf/progs/verifier_ldsx.c
>> index e58ea69e38549..75762e11b443f 100644
>> --- a/tools/testing/selftests/bpf/progs/verifier_ldsx.c
>> +++ b/tools/testing/selftests/bpf/progs/verifier_ldsx.c
> [ ... ]
>> @@ -263,6 +257,14 @@ __naked void ldsx_ctx_8(void)
>> : __clobber_all);
>> }
>>
>> +#ifdef __BPF_FEATURE_ADDR_SPACE_CAST
>> +
>> +struct {
>> + __uint(type, BPF_MAP_TYPE_ARENA);
>> + __uint(map_flags, BPF_F_MMAPABLE);
>> + __uint(max_entries, 1);
>> +} arena SEC(".maps");
>> +
> [Severity: Medium]
> Does this inner guard unintentionally drop the arena tests for clang under
> the default test_progs flavour?
>
> The commit message notes that clang only defines feature macros like this at
> -mcpu=v4, but successfully assembles the inline asm at -mcpu=v3. Because of
> this, you kept the clang version checks on the file's main guard to avoid
> removing tests where they run today.
>
> However, since this inner guard lacks a || __clang_major__ >= 18 fallback,
> clang at -mcpu=v3 evaluates this as false,
Is this just by review or was it actually tested - I'm guessing former ...
> silently dropping the arena map
> and the subsequent 5 arena tests from the default test flavour.
... since clang (at least recent ones, I testet recent and a local jan
build) defines this for all -mcpus
So this is invalid claim.
Thx,
-Vineet
next prev parent reply other threads:[~2026-08-24 21:31 UTC|newest]
Thread overview: 8+ messages / expand[flat|nested] mbox.gz Atom feed top
2026-08-24 21:01 [PATCH bpf-next v2 0/3] selftests/bpf: unblock bpf-gcc test coverage Vineet Gupta
2026-08-24 21:01 ` [PATCH bpf-next v2 1/3] selftests/bpf: name the remaining placeholder programs dummy_test [NFC] Vineet Gupta
2026-08-24 21:01 ` [PATCH bpf-next v2 2/3] selftests/bpf: report placeholder tests as SKIP, not OK Vineet Gupta
2026-08-24 21:42 ` bot+bpf-ci
2026-08-24 21:01 ` [PATCH bpf-next v2 3/3] selftests/bpf: Enable some of the blocked cpuv4 instruction tests for bpf-gcc Vineet Gupta
2026-08-24 21:14 ` sashiko-bot
2026-08-24 21:30 ` Vineet Gupta [this message]
2026-08-26 16:40 ` [PATCH bpf-next v2 0/3] selftests/bpf: unblock bpf-gcc test coverage patchwork-bot+netdevbpf
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=d94a8330-79d4-485d-ae46-cc93c0ceac5e@linux.dev \
--to=vineet.gupta@linux.dev \
--cc=bpf@vger.kernel.org \
--cc=sashiko-reviews@lists.linux.dev \
/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