https://gcc.gnu.org/bugzilla/show_bug.cgi?id=39840
--- Comment #10 from Andrew Pinski ---
(In reply to Andrew Pinski from comment #9)
> GCC has support turning on/off target specific extensions since at least GCC
> 5, maybe earlier. So closing as fixed.
I Mean on specific on a per function
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=39840
Andrew Pinski changed:
What|Removed |Added
Target Milestone|--- |5.0
Status|UNCONFIRMED
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=39840
Bug 39840 depends on bug 37565, which changed state.
Bug 37565 Summary: __optimize__ attribute doesn't work correctly
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=37565
What|Removed |Added
--- Comment #7 from rguenth at gcc dot gnu dot org 2009-04-22 09:36 ---
The problem with different instruction sets in different BBs is also how to
avoid code motion across them. IMNSHO this is a bad idea.
--
rguenth at gcc dot gnu dot org changed:
What|Removed
--- Comment #8 from hjl dot tools at gmail dot com 2009-04-22 13:58 ---
(In reply to comment #7)
The problem with different instruction sets in different BBs is also how to
avoid code motion across them. IMNSHO this is a bad idea.
I agree. There are too many issues with it. I'd
--- Comment #1 from hjl dot tools at gmail dot com 2009-04-21 19:07 ---
Please provide some sample code which can be compiled.
--
hjl dot tools at gmail dot com changed:
What|Removed |Added
--- Comment #2 from drepper at redhat dot com 2009-04-21 19:37 ---
[I couldn't attach the code as an attachment, bugzilla has a bug.]
The program below has to be compiled with -mavx to allow the AVX intrinsics
being used. But this also triggers using the use of the vmovss instruction
Gcc 4.4 and above supports different target options on the function
level but not on a basic block level. So you can create an interneral
version for AVX.
Sent from my iPhone
On Apr 21, 2009, at 12:37 PM, drepper at redhat dot com gcc-bugzi...@gcc.gnu.org
wrote:
--- Comment #2
--- Comment #3 from pinskia at gmail dot com 2009-04-21 19:41 ---
Subject: Re: Non-optimal (or wrong) implementation of SSE intrinsics
Gcc 4.4 and above supports different target options on the function
level but not on a basic block level. So you can create an interneral
version
--- Comment #4 from drepper at redhat dot com 2009-04-21 19:51 ---
(In reply to comment #3)
Gcc 4.4 and above supports different target options on the function
level but not on a basic block level. So you can create an interneral
version for AVX.
This doesn't work either. Aside
--- Comment #5 from hjl dot tools at gmail dot com 2009-04-21 20:34 ---
Created an attachment (id=17667)
-- (http://gcc.gnu.org/bugzilla/attachment.cgi?id=17667action=view)
An example
I am enclosing a modified example which can be compiled with both
icc and gcc. I also included
--- Comment #6 from hjl dot tools at gmail dot com 2009-04-21 21:56 ---
Created an attachment (id=17668)
-- (http://gcc.gnu.org/bugzilla/attachment.cgi?id=17668action=view)
An eample
Here is an example for gcc 4.4. If function level optimization works,
we don't need separate files for
12 matches
Mail list logo