Re: [PATCH 2/2] x86: make __get_wchan() use arch_stack_walk()

[Qi Zheng]

On 2023/4/12 21:23, Peter Zijlstra wrote:
> On Wed, Apr 12, 2023 at 03:15:33PM +0200, Peter Zijlstra wrote:
> > On Fri, Apr 07, 2023 at 10:08:22PM -0700, Josh Poimboeuf wrote:
> > > On Thu, Mar 30, 2023 at 04:15:52PM +0800, Qi Zheng wrote:
> > > > Make __get_wchan() use arch_stack_walk() directly to
> > > > avoid open-coding of unwind logic.
> > > >
> > > > Signed-off-by: Qi Zheng <zhengqi.arch@xxxxxxxxxxxxx>
> > >
> > > Can we just have a shared version of __get_wchan() for all
> > > CONFIG_ARCH_STACKWALK arches?
> >
> > Didn't I do that a while back ? I can't seem to actually find the
> > patch-set though :/
>
> Could be this series:
>
>    https://lkml.kernel.org/r/20211022150933.883959987@xxxxxxxxxxxxx

Oh, I vaguely remember the beginning because I was trying to fix
get_wchan() not supporting ORC unwinder on x86 [1], and then you sent a
patch set, and the patch [2] in this patch set tried to implement the
shared version of __get_wchan().

[1]. https://lore.kernel.org/all/20211008111626.271115116@xxxxxxxxxxxxx/
[2]. https://lore.kernel.org/all/20211008111626.392918519@xxxxxxxxxxxxx/

> And this here:
>
>    https://lore.kernel.org/lkml/CAHk-=wjHbKfck1Ws4Y0pUZ7bxdjU9eh2WK0EFsv65utfeVkT9Q@xxxxxxxxxxxxxx/
>
> might be why I dropped it.. I can't remember.

Didn't realize I had replied to this email before.

But I also don't see why you dropped it. Looks like you have fixed the
UAF problem.

So do we still need to implement a shared version of __get_wchan()?
If we still need it, do I need to send it again? Or just pick your
previous patch directly? Both are fine to me. :)

Thanks,
Qi