[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <CACRpkdZsdggUptPkPhzzu8vv9moSO+rjvZxZRb1VqJ2xWjY=HQ@mail.gmail.com>
Date: Mon, 22 Aug 2022 13:20:32 +0200
From: Linus Walleij <linus.walleij@...aro.org>
To: Chen Zhongjin <chenzhongjin@...wei.com>
Cc: linux-arm-kernel@...ts.infradead.org, linux-kernel@...r.kernel.org,
linux-arch@...r.kernel.org, linux@...linux.org.uk, arnd@...db.de,
ardb@...nel.org, rmk+kernel@...linux.org.uk, rostedt@...dmis.org,
nick.hawkins@....com, john@...ozen.org, mhiramat@...nel.org
Subject: Re: [RESEND PATCH] ARM: Recover kretprobes return address for EABI
stack unwinder
On Mon, Aug 15, 2022 at 1:06 PM Chen Zhongjin <chenzhongjin@...wei.com> wrote:
> 'fed240d9c974 ("ARM: Recover kretprobe modified return address in stacktrace")'
> has implemented kretprobes return address recovery for FP
> unwinder, this patch makes it works for EABI unwinder.
>
> It saves __kretprobe_trampoline address in LR on stack to identify
> and recover the correct return address in EABI unwinder.
>
> Since EABI doesn't use r11 as frame pointer, we need to use SP to
> identify different kretprobes addresses. Here the value of SP has fixed
> distance to conventional FP position so it's fine to use it.
>
> Passed kunit kprobes_test on QEMU.
>
> Signed-off-by: Chen Zhongjin <chenzhongjin@...wei.com>
This looks correct to me FWIW I understand the assembly
but I'm vaguely unfamiliar with the kprobe API, but anyway:
Reviewed-by: Linus Walleij <linus.walleij@...aro.org>
Would you please drop it into Russell's patch tracker?
Yours,
Linus Walleij
Powered by blists - more mailing lists