Linux Perf Users
 help / color / mirror / Atom feed
From: sashiko-bot@kernel.org
To: "Ian Rogers" <irogers@google.com>
Cc: linux-perf-users@vger.kernel.org
Subject: Re: [PATCH v6 2/2] perf inject: Fix itrace branch stack synthesis
Date: Mon, 18 May 2026 19:45:59 +0000	[thread overview]
Message-ID: <20260518194559.D8D7BC2BCB7@smtp.kernel.org> (raw)
In-Reply-To: <20260518184939.2901713-3-irogers@google.com>

Thank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider:
- [Medium] The patch introduces a reliable NULL pointer dereference in `intel_pt_do_synth_pebs_sample` when branch stack injection is requested via `perf inject --itrace=L`.
--

commit 2c77e921627255ed5c1e2aee4ca756b993e7acc3
Author: Ian Rogers <irogers@google.com>

perf inject: Fix itrace branch stack synthesis

When using "perf inject --itrace=L" to synthesize branch stacks from AUX data,
this commit fixes several issues, including missing PERF_SAMPLE_BRANCH_STACK
flags, incorrect sample types early in __cmd_inject, and other validation and
memory problems in sample repiping.

> diff --git a/tools/perf/util/intel-pt.c b/tools/perf/util/intel-pt.c
> index dd2637678b405..7153b48cfe63a 100644
> --- a/tools/perf/util/intel-pt.c
> +++ b/tools/perf/util/intel-pt.c

[ ... ]

> @@ -2505,7 +2505,7 @@ static int intel_pt_do_synth_pebs_sample(struct intel_pt_queue *ptq, struct evse
>  		intel_pt_add_xmm(intr_regs, pos, items, regs_mask);
>  	}
>  
> -	if (sample_type & PERF_SAMPLE_BRANCH_STACK) {
> +	if ((sample_type | evsel->synth_sample_type) & PERF_SAMPLE_BRANCH_STACK) {
>  		if (items->mask[INTEL_PT_LBR_0_POS] ||
>  		    items->mask[INTEL_PT_LBR_1_POS] ||
>  		    items->mask[INTEL_PT_LBR_2_POS]) {
> 			intel_pt_add_lbrs(ptq->last_branch, items);
> 		} else if (pt->synth_opts.last_branch) {
> 			thread_stack__br_sample(ptq->thread, ptq->cpu,
> 						ptq->last_branch,
> 						pt->br_stack_sz);
> 		} else {
> 			ptq->last_branch->nr = 0;
> 		}
> 		sample.branch_stack = ptq->last_branch;
> 	}

Could this lead to a NULL pointer dereference when running perf inject
--itrace=L?

When branch stack injection is requested, pt->synth_opts.add_last_branch is
true, but pt->synth_opts.last_branch and pt->synth_opts.other_events remain
false. This bypasses the allocation of ptq->last_branch in
intel_pt_alloc_queue(), leaving it as NULL.

Simultaneously, intel_pt_br_stack_init() sets the PERF_SAMPLE_BRANCH_STACK
flag in evsel->synth_sample_type for all evsels. By checking
(sample_type | evsel->synth_sample_type) & PERF_SAMPLE_BRANCH_STACK,
execution is forced into this branch stack handling block for PEBS samples.

Inside this block, the code dereferences ptq->last_branch either in
intel_pt_add_lbrs() or by assigning ptq->last_branch->nr = 0. Since
ptq->last_branch is NULL, this leads to a segmentation fault.

-- 
Sashiko AI review · https://sashiko.dev/#/patchset/20260518184939.2901713-1-irogers@google.com?part=2

  reply	other threads:[~2026-05-18 19:46 UTC|newest]

Thread overview: 35+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2026-04-28  7:03 [PATCH v1 0/2] perf inject intel-PT LBR/brstack synthesis fixes Ian Rogers
2026-04-28  7:03 ` [PATCH v1 1/2] perf event: Fix size of synthesized sample with branch stacks Ian Rogers
2026-04-28 23:19   ` Namhyung Kim
2026-04-28  7:03 ` [PATCH v1 2/2] perf inject: Fix itrace branch stack synthesis Ian Rogers
2026-04-28 20:20   ` sashiko-bot
2026-04-29 18:11 ` [PATCH v2 0/2] perf inject intel-PT LBR/brstack synthesis fixes Ian Rogers
2026-04-29 18:11   ` [PATCH v2 1/2] perf event: Fix size of synthesized sample with branch stacks Ian Rogers
2026-04-29 20:51     ` sashiko-bot
2026-04-29 18:11   ` [PATCH v2 2/2] perf inject: Fix itrace branch stack synthesis Ian Rogers
2026-04-29 21:18     ` sashiko-bot
2026-05-18  6:12   ` [PATCH v3 0/2] perf inject intel-PT LBR/brstack synthesis fixes Ian Rogers
2026-05-18  6:12     ` [PATCH v3 1/2] perf event: Fix size of synthesized sample with branch stacks Ian Rogers
2026-05-18  6:39       ` sashiko-bot
2026-05-18  6:12     ` [PATCH v3 2/2] perf inject: Fix itrace branch stack synthesis Ian Rogers
2026-05-18  7:07       ` sashiko-bot
2026-05-18 15:37     ` [PATCH v4 0/2] perf inject intel-PT LBR/brstack synthesis fixes Ian Rogers
2026-05-18 15:37       ` [PATCH v4 1/2] perf event: Fix size of synthesized sample with branch stacks Ian Rogers
2026-05-18 15:37       ` [PATCH v4 2/2] perf inject: Fix itrace branch stack synthesis Ian Rogers
2026-05-18 16:47         ` sashiko-bot
2026-05-18 17:12       ` [PATCH v5 0/2] perf inject intel-PT LBR/brstack synthesis fixes Ian Rogers
2026-05-18 17:12         ` [PATCH v5 1/2] perf event: Fix size of synthesized sample with branch stacks Ian Rogers
2026-05-18 17:12         ` [PATCH v5 2/2] perf inject: Fix itrace branch stack synthesis Ian Rogers
2026-05-18 18:12           ` sashiko-bot
2026-05-18 18:49         ` [PATCH v6 0/2] perf inject intel-PT LBR/brstack synthesis fixes Ian Rogers
2026-05-18 18:49           ` [PATCH v6 1/2] perf event: Fix size of synthesized sample with branch stacks Ian Rogers
2026-05-18 18:49           ` [PATCH v6 2/2] perf inject: Fix itrace branch stack synthesis Ian Rogers
2026-05-18 19:45             ` sashiko-bot [this message]
2026-05-18 20:38           ` [PATCH v7 0/2] perf inject intel-PT LBR/brstack synthesis fixes Ian Rogers
2026-05-18 20:38             ` [PATCH v7 1/2] perf event: Fix size of synthesized sample with branch stacks Ian Rogers
2026-05-18 20:38             ` [PATCH v7 2/2] perf inject: Fix itrace branch stack synthesis Ian Rogers
2026-05-18 21:37               ` sashiko-bot
2026-05-18 22:43             ` [PATCH v8 0/2] perf inject intel-PT LBR/brstack synthesis fixes Ian Rogers
2026-05-18 22:43               ` [PATCH v8 1/2] perf event: Fix size of synthesized sample with branch stacks Ian Rogers
2026-05-18 22:43               ` [PATCH v8 2/2] perf inject: Fix itrace branch stack synthesis Ian Rogers
2026-05-20 19:13               ` [PATCH v8 0/2] perf inject intel-PT LBR/brstack synthesis fixes Arnaldo Carvalho de Melo

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=20260518194559.D8D7BC2BCB7@smtp.kernel.org \
    --to=sashiko-bot@kernel.org \
    --cc=irogers@google.com \
    --cc=linux-perf-users@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