Re: [PATCH v5] kprobes: unpoison stack in jprobe_return() for KASAN
From: Dmitry Vyukov
Date: Sat Oct 15 2016 - 04:01:10 EST
On Sat, Oct 15, 2016 at 8:30 AM, Ingo Molnar <mingo@xxxxxxxxxx> wrote:
> * Dmitry Vyukov <dvyukov@xxxxxxxxxx> wrote:
>> KASAN stack instrumentation poisons stack redzones on function entry
>> and unpoisons them on function exit. If a function exits abnormally
>> (e.g. with a longjmp like jprobe_return()), stack redzones are left
>> poisoned. Later this leads to random KASAN false reports.
>> Unpoison stack redzones in the frames we are going to jump over
>> before doing actual longjmp in jprobe_return().
> Does this affect any other architecture besides arm64? If not then it might make
> the most sense to merge this via the arm64 tree.
This is mostly for x86_64. This patch fixes KASAN false positives
related to jprobe on x86_64.
Arm64 related part is only a function rename. As I introduce a
function similar to an existing one, Mark asked to me rename the
existing function to clarify the difference between the two.