RE: [PATCH v1 0/8] Update to C11, fix signal undefined behavior

[David Laight]

From: Ian Rogers
> Sent: 24 October 2022 19:12
> 
> On Mon, Oct 24, 2022 at 10:59 AM Ian Rogers <irogers@xxxxxxxxxx> wrote:
> >
> > On Mon, Oct 24, 2022 at 10:51 AM Arnaldo Carvalho de Melo
> > <acme@xxxxxxxxxx> wrote:
> > >
> > > Em Mon, Oct 24, 2022 at 10:35:15AM -0700, Ian Rogers escreveu:
> > > > The use of C11 is mainstream in the kernel [1]. There was some
> > > > confusion on volatile and signal handlers in [2]. Switch to using
> > > > stdatomic.h (requires C11) and sig_atomic_t as per [3]. Thanks to Leo
> > > > Yan <leo.yan@xxxxxxxxxx> for the suggestions.
> > > >
> > > > [1] https://lore.kernel.org/lkml/CAHk-=whWbENRz-
> vLY6vpESDLj6kGUTKO3khGtVfipHqwewh2HQ@xxxxxxxxxxxxxx/
> > > > [2] https://lore.kernel.org/lkml/20221024011024.462518-1-irogers@xxxxxxxxxx/
> > > > [3] https://wiki.sei.cmu.edu/confluence/display/c/SIG31-
> C.+Do+not+access+shared+objects+in+signal+handlers
> > >
> > > I think I'll apply this to perf/core, i.e. for 6.3, ok?
> >
> > Sounds good to me. 6.3 or 6.2? I suspect there is more cleanup like
> > this and to the iterators (from C11) that can be done.
> >
> > Thanks,
> > Ian
> 
> So I noticed a few changes missing #include-ing stdatomic.h and
> sig_atomic_t is actually in signal.h. I'm not sure we need the C11
> change then, but it seems like the right thing to do anyway. I'll do a
> v2 to drop the unneeded (currently) include of stdatomic.h.

While the C standard requires you use sig_atomic_t (to avoid
wider RMW being done for writes) the kernel very much requires
that volatiles accesses are atomic.
So the compilers used will do that even when the C standard
would allow them to do otherwise.

Pretty much the last mainstream cpu that couldn't do atomic
byte writes was an old alpha.

So for anything that Linux is going to run on these changes
aren't really needed.

	David

-
Registered Address Lakeside, Bramley Road, Mount Farm, Milton Keynes, MK1 1PT, UK
Registration No: 1397386 (Wales)