[PATCH] testsuite: add dg-require-atomic-cmpxchg-word to chrono tests

[Jonathan Wakely]

On Fri, 26 Jun 2026 at 07:16, Torbjorn SVENSSON
<torbjorn.svensson@foss.st.com> wrote:
>
>
>
> On 2026-06-25 14:52, Jonathan Wakely wrote:
> > On Thu, 25 Jun 2026 at 13:20, Torbjorn SVENSSON
> > <torbjorn.svensson@foss.st.com> wrote:
> >>
> >>
> >> On 2026-06-25 13:41, Jonathan Wakely wrote:
> >>> On Mon, 22 Jun 2026 at 10:56, Torbjorn SVENSSON
> >>> <torbjorn.svensson@foss.st.com> wrote:
> >>>>
> >>>> Hi Jonathan,
> >>>>
> >>>> Did you find any way to get this to work without atomics?
> >>>
> >>> https://gcc.gnu.org/pipermail/gcc-patches/2026-June/721613.html should
> >>> solve this.
> >>>
> >>> Are you able to check that, or should I push it to the forge and ask
> >>> the linaro ci-bot to check it?
> >>
> >> I can give it a try. Will let you know about the outcome tomorrow.
> >
> > Great, thanks. I'll be revising the patch slightly, but it would still
> > be good to know if the general approach solves the problem.
> >
>
> I can confirm that your patch resolves the failures on arm-none-eabi that I targeted with my original patch.
> I can also confirm that your patch does not cause any regressions for any of the arm-none-eabi permutations that I test.

Great, thanks for confirming it.

> Thank you for fixing this a better way than mine!

This way the functionality actually works on Cortex-M0 (and similar
targets without full atomics).