[Patch] Memory fencing fixes

[Mike Frysinger]

On Monday 04 February 2013 16:39:05 Jeff Law wrote:
> I think this incorporates all the input from the list.  If someone could
> take a looksie at the ARM stuff it'd be appreciated.  While I think I
> stitched the bits from Carlos & Joseph together correctly, sanity
> checking helps.

just style nits below.  common & ia64 part looks good.

for future work (i think this has expanded scope enough), i wonder if we 
should try to libraryize the unwind logic more.

> --- a/ports/sysdeps/unix/sysv/linux/arm/nptl/unwind-forcedunwind.c
> +++ b/ports/sysdeps/unix/sysv/linux/arm/nptl/unwind-forcedunwind.c
>  
> +static __always_inline void
> +_maybe_pthread_cancel_init ()
> +{
> +  if (__glibc_unlikely (libgcc_s_handle == NULL, 0))
> +    pthread_cancel_init ();
> +  else
> +    atomic_read_barrier ();
> +
> +}

no need for that empty line before the }

> --- a/sysdeps/gnu/unwind-resume.c
> +++ b/sysdeps/gnu/unwind-resume.c
>  
> +static __always_inline void
> +_maybe_init (void)
> +{
> +  if (__glibc_unlikely (libgcc_s_handle == NULL))
> +    init ();
> +  else
> +    atomic_read_barrier ();
> +}
> +
> +
>  void

only need one blank line between funcs
-mike
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 836 bytes
Desc: This is a digitally signed message part.
URL: <http://sourceware.org/pipermail/libc-alpha/attachments/20130219/8503688e/attachment.sig>