- User Since
- Jul 21 2014, 12:07 PM (148 w, 23 h)
This is a good idea. LGTM too.
Thanks for improving this!
Thu, May 18
Wed, May 17
Please write a new patch with all these improvements and add me as a reviewer. Thanks.
Mon, May 15
Fri, May 12
Would be good to include the analysis for the failure in the log/description.
Thu, May 11
Thanks. I've added opt-remarks to SLP in rL302811. Hopefully you can use those for your analysis too.
Wed, May 10
Tue, May 9
Address Kristof's comments. Thanks, Kristof!
Mon, May 8
Updated according to Kristof's idea: rather than whitelist, blacklist
subtargets (Qualcomm, Cavium) that didn't get a chance to benchmark this yet.
Wed, May 3
Tue, May 2
Thanks @evandro, let me know.
Apr 20 2017
Apr 13 2017
Apr 12 2017
Apr 11 2017
This is actually showing the difference with 64-bit SLP enabled as prepared by opt-diff. New remarks are prefixed with '+'.
Apr 10 2017
Apr 7 2017
Apr 6 2017
Apr 5 2017
The counterpart of fused multiply-and-sub was committed under rL299572.
Apr 4 2017
Thanks, Aaron! @rjmccall, does it look good to you too?
Apr 3 2017
Address John's comment.
Apr 1 2017
Ping. This is the only patch in the series now that wasn't approved.
Mar 31 2017
Mar 30 2017
Add a new test requested by Sanjay.
Also add 'contract' for CompoundAssignOperator (+= and -=). For l-values,
these don't go through the expr-visitor in ScalarExprEmitter::Visit. This
again piggybacks on how debug-locations are added.
Mar 29 2017
Address review comments. Also test with all target enabled now. Sorry again
for the silly mistake.
Sorry about the delay on this but I was working on something related for ARM that may benefit from this as well. What I need for ARM is something that can communicate to the SLPVectorizer that load-pair and store-pair (of two registers) is efficiently supported on the target. I am wondering if we can combine the two things if your new hook would take the type and the vectorization width.
Mar 28 2017
(Sorry about updating the title/description back and forth but arcanist is buggy)
Rename pragma from #pragma clang fast_math contract_fast(on/off) -> #pragma clang fp contract(on/fast/off)
Does this look OK now?
Mar 27 2017
Address Sanjay's comments
Address Matt's comment