This is the mail archive of the
gcc-patches@gcc.gnu.org
mailing list for the GCC project.
Re: [PATCH] PR60092, basic support for posix_malloc
- From: Jakub Jelinek <jakub at redhat dot com>
- To: Richard Biener <rguenther at suse dot de>
- Cc: gcc-patches at gcc dot gnu dot org
- Date: Thu, 6 Feb 2014 14:48:45 +0100
- Subject: Re: [PATCH] PR60092, basic support for posix_malloc
- Authentication-results: sourceware.org; auth=none
- References: <alpine dot LSU dot 2 dot 11 dot 1402061419500 dot 29326 at zhemvz dot fhfr dot qr>
- Reply-to: Jakub Jelinek <jakub at redhat dot com>
On Thu, Feb 06, 2014 at 02:21:01PM +0100, Richard Biener wrote:
> + /* We marking allocated storage local, we deal with it becoming
> + global by escaping and setting of vars_contains_escaped_heap. */
Did you mean By marking ..., or something else?
> + extern int posix_memalign(void **memptr,
> + __SIZE_TYPE__ alignment, __SIZE_TYPE__ size);
> +
> + int foo (int *p)
> + {
> + int res = *p;
> + int *q;
> + posix_memalign ((void **)&q, 128, 128 * sizeof (int));
Do you really want to have strict aliasing violations in the testcase?
I think one has to take address of a void * variable and if you want
int *, then cast to int * afterwards. Also, I think for posix_memalign
you really should be checking return value of the function.
Otherwise LGTM.
Jakub