- User Since
- Jun 18 2015, 1:48 PM (330 w, 1 d)
Feb 17 2021
Feb 16 2021
Updated patch with the cleaner approach of creating a getInitiationInterval() function.
Feb 10 2021
Jan 29 2021
Oct 23 2020
Apr 23 2018
Apr 20 2018
Sep 28 2016
May 27 2016
Updated diff adding more checks in dag-combine-ldst.ll to make sure the %r0 comes from the right place.
May 25 2016
Is my last change correct? I ran the opt -instnamer and added a new line as asked.
May 19 2016
May 16 2016
Here is a link to the RFC thread: https://groups.google.com/forum/#!topic/llvm-dev/JRM2tlYlJ6I. The idea doesn't seem very popular.
May 13 2016
Sorry, I haven't started it yet. I will do that right now!
Added a test case showing that the optimization is done in /O1 and not in /O0. Also added the fix to 2 test cases that were broken by this patch since the code generated changed slightly in /O0.
May 12 2016
May 10 2016
Ok then, it sounds good to me, at least for the test part. But I don't think I am the right person to accept this commit. I don't know why there was a threshold of 50 for optimize for size or why there was another threshold for partial unrolling. There might frontends other than clang that used it?
May 4 2016
The test have been fixed otherwise in rev 268523.
Modified the patch not to affect /O1 optimization level.
May 3 2016
- Merged test cases into one file.
- Added a case for minsize.
- Modified the test that shows the behavior of optsize when no unroll factor is specified.
Ok, thanks! I will modify the test cases and merge them in one file.
- I am not sure it is possible to merge the optsize tests in one call to opt, since one of them use the -unroll-count parameter and not the other. Do you see another way to do that?
I think the blog comment is right. The pragma should make the loop unroll even in /Os. I think it is essential to allow the user to optimize some specific loops even if he generally wants to optimize for size the rest of the code. I will add tests that show the behavior of the loop unroll pass when optnone or optsize are specified.
May 2 2016
Apr 26 2016
Apr 20 2016
Apr 19 2016
Apr 14 2016
Mar 30 2016
Mar 23 2016
nit: in summary, consiring -> considering ?
Mar 11 2016
Mar 10 2016
Mar 9 2016
Thanks! Could someone commit it for me? I don't have commit access.
Replaced the test by a new case in lang/c/anonymous test, as suggested by Jim. Is this correct?
Sure. Would you have a suggestion of how to call it? Maybe something like variable_flat since it is the command used? Also, is it the right place to add it in functionalities or should it be elsewhere?
Fixed test header, removed useless include.
Feb 26 2016
Are the changes correct? And if so, could someone commit it for me? I don't have commit access.
Feb 19 2016
I have updated the patch to add the assertion.
Feb 17 2016
Feb 11 2016
Feb 9 2016
I don't have commit access. It would be great if you can push it for me. Thanks !
Feb 8 2016
Added a small unit test for scalar right shift operator, which invokes the >>= operator.
Scalar::operator<<= works well as-is because it uses APInt &operator<<=(unsigned shiftAmt), whereas the right shift equivalent is not implemented. Should I add APInt &operator>>=, or should I change Scalar::operator<<= for consistency ?