On Fri, 2013-05-17 at 01:11 -0400, David Edelsohn wrote:
> On Thu, May 16, 2013 at 10:40 PM, Bill Schmidt
> wrote:
> > This removes two degradations in CPU2006 for 32-bit PowerPC due to lost
> > vectorization opportunities. Previously, GCC treated malloc'd arrays as
> > only guaranteeing 4-byte
On Fri, May 17, 2013 at 11:17:26AM +0200, Richard Biener wrote:
> Yes - note that it's called MALLOC_*ABI*_ALIGNMENT for a reason - it
> is supposed to be the alignment that is required for conforming to the C ABI
> on the target. For different allocators I'd rather have a new function
> attribute
On Fri, May 17, 2013 at 10:47 AM, Jakub Jelinek wrote:
> On Fri, May 17, 2013 at 10:32:11AM +0200, Richard Biener wrote:
>> On Fri, May 17, 2013 at 4:40 AM, Bill Schmidt
>> wrote:
>> > This removes two degradations in CPU2006 for 32-bit PowerPC due to lost
>> > vectorization opportunities. Previ
On Fri, May 17, 2013 at 04:55:20AM -0400, David Edelsohn wrote:
> As Bill wrote earlier, 2 * POINTER_SIZE causes a different performance
> regression for 16 byte alignment on PPC64. 8 bytes is a work-around
> for now.
I was talking about the config/linux.h definition, PPC64 can of course
override
On Fri, May 17, 2013 at 4:47 AM, Jakub Jelinek wrote:
> On Fri, May 17, 2013 at 10:32:11AM +0200, Richard Biener wrote:
>> On Fri, May 17, 2013 at 4:40 AM, Bill Schmidt
>> wrote:
>> > This removes two degradations in CPU2006 for 32-bit PowerPC due to lost
>> > vectorization opportunities. Previo
On Fri, May 17, 2013 at 10:32:11AM +0200, Richard Biener wrote:
> On Fri, May 17, 2013 at 4:40 AM, Bill Schmidt
> wrote:
> > This removes two degradations in CPU2006 for 32-bit PowerPC due to lost
> > vectorization opportunities. Previously, GCC treated malloc'd arrays as
> > only guaranteeing 4-
On Fri, May 17, 2013 at 4:40 AM, Bill Schmidt
wrote:
> This removes two degradations in CPU2006 for 32-bit PowerPC due to lost
> vectorization opportunities. Previously, GCC treated malloc'd arrays as
> only guaranteeing 4-byte alignment, even though the glibc implementation
> guarantees 8-byte a
On Thu, May 16, 2013 at 10:40 PM, Bill Schmidt
wrote:
> This removes two degradations in CPU2006 for 32-bit PowerPC due to lost
> vectorization opportunities. Previously, GCC treated malloc'd arrays as
> only guaranteeing 4-byte alignment, even though the glibc implementation
> guarantees 8-byte
This removes two degradations in CPU2006 for 32-bit PowerPC due to lost
vectorization opportunities. Previously, GCC treated malloc'd arrays as
only guaranteeing 4-byte alignment, even though the glibc implementation
guarantees 8-byte alignment. This raises the guarantee to 8 bytes,
which is suff
