public inbox for linux-perf-users@vger.kernel.org
 help / color / mirror / Atom feed
From: sashiko-bot@kernel.org
To: "Rong Bao" <rong.bao@csmantle.top>
Cc: linux-perf-users@vger.kernel.org
Subject: Re: [PATCH] perf annotate: Use jump__delete when freeing LoongArch jumps
Date: Sun, 12 Apr 2026 07:17:18 +0000	[thread overview]
Message-ID: <20260412071719.5B339C19424@smtp.kernel.org> (raw)
In-Reply-To: <20260412062828.1734637-1-rong.bao@csmantle.top>

Sashiko AI review found 1 potential issue(s):
- [High] Missing NULL check in `loongarch_call__parse()` causes a segmentation fault on malformed or stripped objdump output.
--

commit 031fe3089ac82ebce8ac1cc8c2f144a5b6abd009
Author: Rong Bao <rong.bao@csmantle.top>

perf annotate: Use jump__delete when freeing LoongArch jumps

This commit fixes an invalid free crash in perf annotate for LoongArch by
assigning jump__delete to the .free field in loongarch_jump_ops. This prevents
fields that share the same buffer from being erroneously freed during cleanup.

> diff --git a/tools/perf/util/annotate-arch/annotate-loongarch.c b/tools/perf/util/annotate-arch/annotate-loongarch.c
> index 950f34e59e5cd..c2addca77320b 100644
> --- a/tools/perf/util/annotate-arch/annotate-loongarch.c
> +++ b/tools/perf/util/annotate-arch/annotate-loongarch.c
> @@ -110,6 +110,7 @@ static int loongarch_jump__parse(const struct arch *arch, struct ins_operands *o
>  }
>  
>  static const struct ins_ops loongarch_jump_ops = {
> +	.free	   = jump__delete,
>  	.parse	   = loongarch_jump__parse,
>  	.scnprintf = jump__scnprintf,
>  	.is_jump   = true,

[ ... ]

While reviewing these changes, I noticed a pre-existing issue in the nearby
loongarch_call__parse() function in the same file:

tools/perf/util/annotate-arch/annotate-loongarch.c:loongarch_call__parse() {
        ...
        ops->target.addr = strtoull(c, &endptr, 16);

        name = strchr(endptr, '<');
        name++;

        if (arch->objdump.skip_functions_char &&
            strchr(name, arch->objdump.skip_functions_char))
                return -1;
        ...
}

Could this cause a segmentation fault if the tool processes a stripped
binary?

If the objdump output for a call instruction lacks the <symbol> suffix
(for example, bl #0x12345), strchr(endptr, '<') would return NULL.

The subsequent name++ would then produce an invalid pointer, leading to a crash
when it is used in the following strchr() and strdup() calls.

-- 
Sashiko AI review · https://sashiko.dev/#/patchset/20260412062828.1734637-1-rong.bao@csmantle.top?part=1

  reply	other threads:[~2026-04-12  7:17 UTC|newest]

Thread overview: 5+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2026-04-12  6:28 [PATCH] perf annotate: Use jump__delete when freeing LoongArch jumps Rong Bao
2026-04-12  7:17 ` sashiko-bot [this message]
2026-04-13  8:33 ` Huacai Chen
2026-04-13  8:49   ` Rong Bao
2026-04-13  9:41     ` Huacai Chen

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=20260412071719.5B339C19424@smtp.kernel.org \
    --to=sashiko-bot@kernel.org \
    --cc=linux-perf-users@vger.kernel.org \
    --cc=rong.bao@csmantle.top \
    --cc=sashiko@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