Re: New warning in linus/master

[Christophe Leroy (CS GROUP)]

Le 22/04/2026 à 07:56, Thomas Weißschuh a écrit :
> Hi Ricardo,
>
> On Wed, Apr 22, 2026 at 11:51:45AM +0800, Ricardo Ribalda wrote:
> > Media-CI has found a couple of new warnings in the latest kernel
> > version for aarch64 and powerpc. They get fixed with this patch and
> > before moving I wanted to know if this was under your radar.
>
> Thanks for the report. I was not aware of these so far.
>
> > diff --git a/arch/arm64/kernel/vdso/Makefile b/arch/arm64/kernel/vdso/Makefile
> > index 7dec05dd33b7..65914842fae0 100644
> > --- a/arch/arm64/kernel/vdso/Makefile
> > +++ b/arch/arm64/kernel/vdso/Makefile
> > @@ -50,7 +50,7 @@ CFLAGS_vgettimeofday.o = $(CC_FLAGS_ADD_VDSO)
> >   CFLAGS_vgetrandom.o = $(CC_FLAGS_ADD_VDSO)
> >
> >   ifneq ($(c-gettimeofday-y),)
> > -  CFLAGS_vgettimeofday.o += -include $(c-gettimeofday-y)
> > +  CFLAGS_vgettimeofday.o += -include $(c-gettimeofday-y)
> > -Wno-maybe-uninitialized
> >   endif
>
> (...)
>
> I'd like to know exactly what is going on before suppressing the warning.
> It is a non-standard warning, only enabled by *some* of the vDSO builds
> for some reason.
>
> > https://eur01.safelinks.protection.outlook.com/?url=https%3A%2F%2Fgitlab.freedesktop.org%2Flinux-media%2Fusers%2Fribalda%2F-%2Fpipelines%2F1649144%2Ftest_report%3Fjob_name%3Dcross-gcc&data=05%7C02%7Cchristophe.leroy2%40cs-soprasteria.com%7C5c0f15ee20a44072f47108dea0340304%7C8b87af7d86474dc78df45f69a2011bb5%7C0%7C0%7C639124342424606113%7CUnknown%7CTWFpbGZsb3d8eyJFbXB0eU1hcGkiOnRydWUsIlYiOiIwLjAuMDAwMCIsIlAiOiJXaW4zMiIsIkFOIjoiTWFpbCIsIldUIjoyfQ%3D%3D%7C0%7C%7C%7C&sdata=9Tvp2aYfB02ijxVGLzCOYcMwZlPvqVDufxEporLl1mc%3D&reserved=0
>
> While I was able to download a configuration from this job and also use the
> same container image, I can not reproduce the issue. Is the configuration the
> full one or only the template?
>
> Could you provide full reproduction steps?
>
> > CONFIG_CC_VERSION_TEXT="aarch64-linux-gnu-gcc (Debian 15.2.0-16) 15.2.0"
>
> The linked test log shows GCC 14.2, not 15.2.
> Not that I could reproduce it with either one...
>
> >    SYNC    include/config/auto.conf
> >    CC      arch/arm64/kernel/vdso/vgettimeofday.o
> > In file included from <command-line>:
> > In function ‘vdso_set_timespec’,
> >      inlined from ‘do_aux’ at /workdir/lib/vdso/gettimeofday.c:266:2,
> >      inlined from ‘__cvdso_clock_gettime_common’ at
> > /workdir/lib/vdso/gettimeofday.c:293:10,
> >      inlined from ‘__cvdso_clock_gettime_data.constprop’ at
> > /workdir/lib/vdso/gettimeofday.c:306:7:
> > /workdir/lib/vdso/gettimeofday.c:104:26: warning: ‘sec’ may be used
> > uninitialized [-Wmaybe-uninitialized]
> >    104 |         ts->tv_sec = sec + __iter_div_u64_rem(ns, NSEC_PER_SEC, &ns);
> >        |                      ~~~~^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
> > /workdir/lib/vdso/gettimeofday.c: In function
> > ‘__cvdso_clock_gettime_data.constprop’:
> > /workdir/lib/vdso/gettimeofday.c:242:13: note: ‘sec’ was declared here
> >    242 |         u64 sec, ns;
> >        |             ^~~
>
> The same pattern is used in some other do_ handlers, which do not generate this
> warning. I also can't immediately see what is wrong.

You are likely in a special case where vdso_get_timestamp() returns 
false without setting &sec, and some versions of GCC and/or CLANG fail 
to see that vdso_set_timespec() is not called when vdso_get_timestamp() 
returns false.

Christophe