Re: [PATCH v2 1/4] meson: hide tsan related warnings

[Peter Maydell]

On Thu, 15 Aug 2024 at 12:05, Daniel P. Berrangé <berrange@redhat.com> wrote:
>
> On Thu, Aug 15, 2024 at 11:12:39AM +0100, Peter Maydell wrote:
> > On Wed, 14 Aug 2024 at 23:42, Pierrick Bouvier
> > <pierrick.bouvier@linaro.org> wrote:
> > >
> > > When building with gcc-12 -fsanitize=thread, gcc reports some
> > > constructions not supported with tsan.
> > > Found on debian stable.
> > >
> > > qemu/include/qemu/atomic.h:36:52: error: ‘atomic_thread_fence’ is not 
> > > supported with ‘-fsanitize=thread’ [-Werror=tsan]
> > >    36 | #define smp_mb()                     ({ barrier(); 
> > > __atomic_thread_fence(__ATOMIC_SEQ_CST); })
> > >       |                                                    
> > > ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
> > >
> > > Signed-off-by: Pierrick Bouvier <pierrick.bouvier@linaro.org>
> > > ---
> > >  meson.build | 10 +++++++++-
> > >  1 file changed, 9 insertions(+), 1 deletion(-)
> > >
> > > diff --git a/meson.build b/meson.build
> > > index 81ecd4bae7c..52e5aa95cc0 100644
> > > --- a/meson.build
> > > +++ b/meson.build
> > > @@ -499,7 +499,15 @@ if get_option('tsan')
> > >                           prefix: '#include <sanitizer/tsan_interface.h>')
> > >      error('Cannot enable TSAN due to missing fiber annotation interface')
> > >    endif
> > > -  qemu_cflags = ['-fsanitize=thread'] + qemu_cflags
> > > +  tsan_warn_suppress = []
> > > +  # gcc (>=11) will report constructions not supported by tsan:
> > > +  # "error: ‘atomic_thread_fence’ is not supported with 
> > > ‘-fsanitize=thread’"
> > > +  # https://gcc.gnu.org/gcc-11/changes.html
> > > +  # However, clang does not support this warning and this triggers an 
> > > error.
> > > +  if cc.has_argument('-Wno-tsan')
> > > +    tsan_warn_suppress = ['-Wno-tsan']
> > > +  endif
> >
> > That last part sounds like a clang bug -- -Wno-foo is supposed
> > to not be an error on compilers that don't implement -Wfoo for
> > any value of foo (unless some other warning/error would also
> > be emitted).
>
> -Wno-foo isn't an error, but it is a warning... which we then
> turn into an error due to -Werror, unless we pass -Wno-unknown-warning-option
> to clang.

Which is irritating if you want to be able to blanket say
'-Wno-silly-compiler-warning' and not see any of that
warning regardless of compiler version. That's why the
gcc behaviour is the way it is (i.e. -Wno-such-thingy
is neither a warning nor an error if it would be the only
warning/error), and if clang doesn't match it that's a shame.

thanks
-- PMM