Linux Trace Kernel
 help / color / mirror / Atom feed
From: sashiko-bot@kernel.org
To: "Shuai Xue" <xueshuai@linux.alibaba.com>,
	palmer@dabbelt.com, pjw@kernel.org, aou@eecs.berkeley.edu
Cc: alex@ghiti.fr, oleg@redhat.com, rostedt@goodmis.org,
	mhiramat@kernel.org, mark.rutland@arm.com, peterz@infradead.org,
	mingo@redhat.com, acme@kernel.org, namhyung@kernel.org,
	alexander.shishkin@linux.intel.com, jolsa@kernel.org,
	irogers@google.com, adrian.hunter@intel.com,
	james.clark@linaro.org, jpoimboe@kernel.org, jikos@kernel.org,
	mbenes@suse.cz, pmladek@suse.com, joe.lawrence@redhat.com,
	shuah@kernel.org, mpdesouza@suse.com, martin@kaiser.cx,
	kmehltretter@gmail.com, oliver.yang@linux.alibaba.com,
	zhuo.song@linux.alibaba.com, jkchen@linux.alibaba.com,
	linux-trace-kernel@vger.kernel.org,
	linux-perf-users@vger.kernel.org, live-patching@vger.kernel.org
Subject: Re: [PATCH v7 04/10] riscv: stacktrace: Add frame record metadata
Date: Tue, 10 Nov 2026 01:54:33 +0000	[thread overview]
Message-ID: <sashiko-outbox-206003@kernel.org> (raw)
In-Reply-To: <20261110013750.36800-5-xueshuai@linux.alibaba.com>

Thank you for your contribution! Sashiko AI review found 1 potential
issue to consider:
- [Medium] riscv: entry: use call instead of tail for handle_bad_stack()
--

commit ed05329452fecde9e16a00910af1efec91f2922b
Author: Wang Han <wanghan@linux.alibaba.com>

riscv: stacktrace: Add frame record metadata

This commit adds a metadata frame record to struct pt_regs and initializes
it across exception, overflow, boot, and task-creation entry boundaries
for frame-pointer stack unwinding.

> diff --git a/arch/riscv/kernel/entry.S b/arch/riscv/kernel/entry.S
> index 61b7a77166971..4c69132d59785 100644
> --- a/arch/riscv/kernel/entry.S
> +++ b/arch/riscv/kernel/entry.S

[ ... ]

> @@ -354,6 +376,19 @@ SYM_CODE_START_LOCAL(handle_kernel_stack_overflow)
>  	REG_S s3, PT_BADADDR(sp)
>  	REG_S s4, PT_CAUSE(sp)
>  	REG_S s5, PT_TP(sp)
> +
> +	/*
> +	 * Create a metadata frame record for the overflow pt_regs. The
> +	 * overflow path is entered from kernel context, so this is a nested
> +	 * pt_regs boundary and the unwinder can resume from the pre-overflow
> +	 * frame pointer saved in PT_S0.
> +	 */
> +	REG_S zero, (S_STACKFRAME + FRAME_RECORD_FP)(sp)
> +	REG_S zero, (S_STACKFRAME + FRAME_RECORD_RA)(sp)
> +	li t0, FRAME_META_TYPE_PT_REGS
> +	REG_S t0, S_STACKFRAME_TYPE(sp)
> +	addi s0, sp, S_STACKFRAME + STACKFRAME_RECORD_SIZE
> +
>  	move a0, sp
>  	tail handle_bad_stack
>  SYM_CODE_END(handle_kernel_stack_overflow)

[Severity: Medium]
Should this use "call handle_bad_stack" instead of "tail handle_bad_stack"?

Because tail does not update ra, handle_bad_stack()'s prologue saves a
frame record on the overflow stack that pairs the new s0 (pointing at the
overflow pt_regs metadata record) with the interrupted pre-overflow ra.
When panic() unwinds from handle_bad_stack(), the unwinder consumes
regs->ra before reaching the pt_regs metadata record, and if that return
address was hooked by function_graph or kretprobes, recovering it with the
overflow-stack fp fails and aborts the stack trace before reaching the
overflowed task stack.

-- 
Sashiko AI review · https://sashiko.dev/#/patchset/20261110013750.36800-1-xueshuai@linux.alibaba.com?part=4

  reply	other threads:[~2026-11-10  1:54 UTC|newest]

Thread overview: 39+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2026-09-14  9:26 [PATCH v6 0/7] riscv: Add reliable stack unwinding for livepatch Shuai Xue
2026-09-14  9:26 ` [PATCH v6 1/7] riscv: stacktrace: Add frame record metadata Shuai Xue
2026-09-14  9:40   ` sashiko-bot
2026-09-14  9:26 ` [PATCH v6 2/7] riscv: stacktrace: disable KASAN and KCOV instrumentation for stacktrace.o Shuai Xue
2026-09-14  9:32   ` sashiko-bot
2026-09-14  9:26 ` [PATCH v6 3/7] riscv: ftrace: always preserve s0 in dynamic ftrace register frame Shuai Xue
2026-09-14  9:32   ` sashiko-bot
2026-09-14  9:26 ` [PATCH v6 4/7] riscv: stacktrace: introduce stack-bound tracking helpers Shuai Xue
2026-09-14  9:33   ` sashiko-bot
2026-09-14  9:26 ` [PATCH v6 5/7] riscv: stacktrace: switch to frame-pointer based unwinder Shuai Xue
2026-09-14  9:39   ` sashiko-bot
2026-10-09 18:59   ` Karl Mehltretter
2026-10-10  3:38     ` Shuai Xue
2026-10-13 20:18   ` Martin Kaiser
2026-10-14 13:09     ` Shuai Xue
2026-09-14  9:26 ` [PATCH v6 6/7] riscv: Kconfig: enable HAVE_RELIABLE_STACKTRACE and HAVE_LIVEPATCH Shuai Xue
2026-09-14  9:38   ` sashiko-bot
2026-09-14  9:26 ` [PATCH v6 7/7] selftests/livepatch: Add RISC-V syscall wrapper prefix Shuai Xue
2026-09-14  9:38   ` sashiko-bot
2026-09-29  7:28 ` [PATCH v6 0/7] riscv: Add reliable stack unwinding for livepatch Martin Kaiser
2026-10-08 11:42   ` Shuai Xue
2026-10-12  8:44     ` Martin Kaiser
2026-11-10  1:37 ` [PATCH v7 00/10] " Shuai Xue
2026-11-10  1:37   ` [PATCH v7 01/10] riscv: stacktrace: do not report the walker's own frame from arch_stack_walk() Shuai Xue
2026-11-10  1:52     ` sashiko-bot
2026-11-10  1:37   ` [PATCH v7 02/10] riscv: Add KUnit test for stack trace start frames Shuai Xue
2026-11-10  1:50     ` sashiko-bot
2026-11-10  1:37   ` [PATCH v7 03/10] riscv: ftrace: use the frame pointer as the function graph return-address cookie Shuai Xue
2026-11-10  1:37   ` [PATCH v7 04/10] riscv: stacktrace: Add frame record metadata Shuai Xue
2026-11-10  1:54     ` sashiko-bot [this message]
2026-11-10  1:37   ` [PATCH v7 05/10] riscv: stacktrace: disable KASAN and KCOV instrumentation for stacktrace.o Shuai Xue
2026-11-10  1:37   ` [PATCH v7 06/10] riscv: ftrace: always preserve s0 in dynamic ftrace register frame Shuai Xue
2026-11-10  1:37   ` [PATCH v7 07/10] riscv: stacktrace: introduce stack-bound tracking helpers Shuai Xue
2026-11-10  1:37   ` [PATCH v7 08/10] riscv: stacktrace: switch to frame-pointer based unwinder Shuai Xue
2026-11-10  2:20     ` sashiko-bot
2026-11-10  1:37   ` [PATCH v7 09/10] riscv: Kconfig: enable HAVE_RELIABLE_STACKTRACE and HAVE_LIVEPATCH Shuai Xue
2026-11-10  2:10     ` sashiko-bot
2026-11-10  1:37   ` [PATCH v7 10/10] selftests/livepatch: Add RISC-V syscall wrapper prefix Shuai Xue
2026-11-10  6:35   ` [PATCH v7 00/10] riscv: Add reliable stack unwinding for livepatch Karl Mehltretter

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=sashiko-outbox-206003@kernel.org \
    --to=sashiko-bot@kernel.org \
    --cc=acme@kernel.org \
    --cc=adrian.hunter@intel.com \
    --cc=alex@ghiti.fr \
    --cc=alexander.shishkin@linux.intel.com \
    --cc=aou@eecs.berkeley.edu \
    --cc=irogers@google.com \
    --cc=james.clark@linaro.org \
    --cc=jikos@kernel.org \
    --cc=jkchen@linux.alibaba.com \
    --cc=joe.lawrence@redhat.com \
    --cc=jolsa@kernel.org \
    --cc=jpoimboe@kernel.org \
    --cc=kmehltretter@gmail.com \
    --cc=linux-perf-users@vger.kernel.org \
    --cc=linux-trace-kernel@vger.kernel.org \
    --cc=live-patching@vger.kernel.org \
    --cc=mark.rutland@arm.com \
    --cc=martin@kaiser.cx \
    --cc=mbenes@suse.cz \
    --cc=mhiramat@kernel.org \
    --cc=mingo@redhat.com \
    --cc=mpdesouza@suse.com \
    --cc=namhyung@kernel.org \
    --cc=oleg@redhat.com \
    --cc=oliver.yang@linux.alibaba.com \
    --cc=palmer@dabbelt.com \
    --cc=peterz@infradead.org \
    --cc=pjw@kernel.org \
    --cc=pmladek@suse.com \
    --cc=rostedt@goodmis.org \
    --cc=sashiko-reviews@lists.linux.dev \
    --cc=shuah@kernel.org \
    --cc=xueshuai@linux.alibaba.com \
    --cc=zhuo.song@linux.alibaba.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