Re: [PATCH v1] powerpc: Make setjmp/longjump signature standard

From: Nathan Chancellor
Date: Fri Mar 27 2020 - 13:27:10 EST


On Fri, Mar 27, 2020 at 10:10:44AM -0700, Nick Desaulniers wrote:
> On Fri, Mar 27, 2020 at 3:08 AM Clement Courbet <courbet@xxxxxxxxxx> wrote:
> >
> > Declaring setjmp()/longjmp() as taking longs makes the signature
> > non-standard, and makes clang complain. In the past, this has been
> > worked around by adding -ffreestanding to the compile flags.
> >
> > The implementation looks like it only ever propagates the value
> > (in longjmp) or sets it to 1 (in setjmp), and we only call longjmp
> > with integer parameters.
> >
> > This allows removing -ffreestanding from the compilation flags.
> >
> > Context:
> > https://lore.kernel.org/patchwork/patch/1214060
> > https://lore.kernel.org/patchwork/patch/1216174
> >
> > Signed-off-by: Clement Courbet <courbet@xxxxxxxxxx>

Thanks for fixing this properly, not really sure why I did not think of
this in the first place. I guess my thought was the warning makes it
seem like clang is going to ignore the kernel's implementation of
setjmp/longjmp but I can't truly remember.

> Hi Clement, thanks for the patch! Would you mind sending a V2 that
> included a similar fix to arch/powerpc/xmon/Makefile?

Agreed.

> For context, this was the original patch:
> https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?id=aea447141c7e7824b81b49acd1bc78
> which was then modified to:
> https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?id=c9029ef9c95765e7b63c4d9aa780674447db1ec0
>
> So on your V2, if you include in the commit message, the line:
>
> Fixes c9029ef9c957 ("powerpc: Avoid clang warnings around setjmp and longjmp")
>
> then that will help our LTS branch maintainers back port it to the
> appropriate branches.

The tags should be:

Cc: stable@xxxxxxxxxxxxxxx # v4.14+
Fixes: c9029ef9c957 ("powerpc: Avoid clang warnings around setjmp and longjmp")

that way it explicitly gets picked up for stable, rather than Sasha's
AUTOSEL process, which could miss it.

With the xmon/Makefile -ffreestanding removed and the tags updated,
consider this:

Reviewed-by: Nathan Chancellor <natechancellor@xxxxxxxxx>
Tested-by: Nathan Chancellor <natechancellor@xxxxxxxxx>

Cheers,
Nathan