https://gcc.gnu.org/bugzilla/show_bug.cgi?id=102838
--- Comment #16 from ro at CeBiTec dot Uni-Bielefeld.DE <ro at CeBiTec dot
Uni-Bielefeld.DE> ---
> --- Comment #15 from Jakub Jelinek <jakub at gcc dot gnu.org> ---
> Ah, bet Solaris aligned_alloc relies on:
> "the value of size shall be an integral multiple of alignment"
> (glibc aligned_alloc doesn't).
Indeed, cf. aligned_alloc(3C):
The alignment argument must be a valid alignment, that is, any power of
two (1, 2, 4, 8, ...) and the size argument must be an integral multi-
ple of alignment.
> Does memalign or posix_memalign rely on that too, or just aligned_alloc?
memalign(3C) has:
The memalign() function allocates size bytes on a specified alignment
boundary and returns a pointer to the allocated block. The value of the
returned address is guaranteed to be an even multiple of alignment. The
value of alignment must be a power of two and must be greater than or
equal to the size of a word.
and posix_memalign(3C):
The posix_memalign() function allocates size bytes aligned on a bound-
ary specified by alignment, and returns a pointer to the allocated mem-
ory in memptr. The value of alignment must be a power of two multiple
of sizeof(void *).
> If just aligned_alloc, we could do incrementally:
> 2021-11-17 Jakub Jelinek <[email protected]>
>
> PR libgomp/102838
> * alloc.c (gomp_aligned_alloc): Prefer _aligned_alloc over
> memalign over posix_memalign over aligned_alloc over fallback
> with malloc instead of aligned_alloc over _aligned_alloc over
> posix_memalign over memalign over fallback with malloc. For
> aligned_alloc, round up size up to multiple of al.
That patch worked just fine (only tried on i386-pc-solaris2.11, 32 and
64-bit): all 64-bit failures are gone again. Thanks.