Re: [PATCH v2 04/27] locking/atomic: make atomic*_{cmp,}xchg optional

[Guenter Roeck]

On 7/8/23 06:07, Linux regression tracking (Thorsten Leemhuis) wrote:
> On 27.06.23 19:07, Guenter Roeck wrote:
> > On Mon, Jun 05, 2023 at 08:01:01AM +0100, Mark Rutland wrote:
> > > Most architectures define the atomic/atomic64 xchg and cmpxchg
> > > operations in terms of arch_xchg and arch_cmpxchg respectfully.
> > >
> > > Add fallbacks for these cases and remove the trivial cases from arch
> > > code. On some architectures the existing definitions are kept as these
> > > are used to build other arch_atomic*() operations.
> > >
> > > Signed-off-by: Mark Rutland <mark.rutland@xxxxxxx>
> > [...]
> >
> > This patch results in:
> >
> > ERROR: modpost: "__xchg_called_with_bad_pointer" [lib/atomic64_test.ko] undefined!
> >
> > when trying to build sparc64:allmodconfig.
>
> Guenter, please correct me if I'm wrong:
>
> This is fixed by Arnd's patch "sparc: mark __arch_xchg() as
> __always_inline", but that afaics sadly hasn't even reached -next yet.
> https://lore.kernel.org/all/20230628094938.2318171-1-arnd@xxxxxxxxxx/
>
> Hence adding it to the tracking now that the end of the merge window is
> near:
>
> #regzbot ^introduced d12157efc8e083c7
> #regzbot title locking/atomic: build error on sparc64:allmodconfig
> #regzbot ignore-activity

Yes, this is correct.

Guenter