On 12/6/2013, 11:28 AM, Michael Meissner wrote:
On Thu, Dec 05, 2013 at 12:40:17PM -0500, Vladimir Makarov wrote:
The following patch fixes two GCC testsuite failures for LRA. The
patch makes swap through registers instead of memory for the test
cases when LRA is used.
There are differences in reload and LRA constraint matching
algorithm which results in different alternative choices when the
original pattern is used.
Actually my first proposed solution variant used one pattern which
is now for LRA in this patch. But some doubt arose that it may
affect reload pass in some bad way.
Ok to commit?
I must admit to not remembering why I used ??&r. I know I wanted it to prefer
doing the memory patterns. I would think we should try just the pattern
without the ??.
I tried it about 2 months ago. I did not see any problems of such
change for reload and LRA. There were no regressions on GCC testsuite.
So, Mike, if you don't see any compelling reason to keep ??, probably
we should remove them.
If you don't mind, I'll make the patch and test again and after that
submit it for approval.
