This is the mail archive of the
mailing list for the GCC project.
Re: graphite in -O3
- From: Richard Biener <richard dot guenther at gmail dot com>
- To: David Edelsohn <dje dot gcc at gmail dot com>
- Cc: Andi Kleen <andi at firstfloor dot org>, GCC Development <gcc at gcc dot gnu dot org>
- Date: Mon, 17 Nov 2014 11:45:15 +0100
- Subject: Re: graphite in -O3
- Authentication-results: sourceware.org; auth=none
- References: <20141116191059 dot GC12538 at two dot firstfloor dot org> <CAGWvnynV7YRMUKptxDObMwtQdF+ZU=HEGZr1Co6P7jZByXY2hw at mail dot gmail dot com>
On Sun, Nov 16, 2014 at 10:08 PM, David Edelsohn <email@example.com> wrote:
> Because they have not shown general performance benefit.
Also they very often cause very long compile-times.
> I hope that the renewed attention and development effort will allow
> them to be enabled by default eventually.
Yeah - though I'd like to have a group flag that enables the mature
ones. Mature here means transforms that guard themselves properly
for long compile-time and ones that have a cost model (may be
a runtime check even).
> Thanks, David
> On Sun, Nov 16, 2014 at 2:10 PM, Andi Kleen <firstname.lastname@example.org> wrote:
>> Is there any specific reason why none of the graphite loop optimizations
>> (loop-block, loop-interchange, loop-strip-mine, loop-jam)
>> are enabled with -O3 or -Ofast?
>> I assume doing so would make them much more widely used.
>> Perhaps would be something to consider for 5.0?