Roman Lebedev 8217a0ec45 [CVP] After proving that @llvm.with.overflow()/@llvm.sat() don't overflow, also try to prove other no-wrap ... Summary: CVP, unlike InstCombine, does not run till exaustion. It only does a single pass. When dealing with those special binops, if we prove that they can safely be demoted into their usual binop form, we do set the no-wrap we deduced. But when dealing with usual binops, we try to deduce both no-wraps. So if we convert e.g. @llvm.uadd.with.overflow() to `add nuw`, we won't attempt to check whether it can be `add nuw nsw`. This patch proposes to call `processBinOp()` on newly-created binop, which is identical to what we do for div/rem already. Reviewers: nikic, spatel, reames Reviewed By: nikic Subscribers: hiraditya, llvm-commits Tags: #llvm Differential Revision: https://reviews.llvm.org/D69183 llvm-svn: 375273		2019-10-18 19:32:47 +00:00
..
AggressiveInstCombine	[InstCombine] recognize popcount.	2019-10-11 05:13:56 +00:00
Coroutines	[Alignment][NFC] Remove AllocaInst::setAlignment(unsigned)	2019-09-30 13:34:44 +00:00
Hello
InstCombine	[InstCombine] Fix miscompile bug in canEvaluateShuffled	2019-10-18 07:42:02 +00:00
Instrumentation	[Alignment][NFC] Remove dependency on GlobalObject::setAlignment(unsigned)	2019-10-15 11:24:36 +00:00
IPO	Reland: Dead Virtual Function Elimination	2019-10-17 09:58:57 +00:00
ObjCARC	[ObjC][ARC] Skip debug instructions when computing the insert point of	2019-09-19 20:58:51 +00:00
Scalar	[CVP] After proving that @llvm.with.overflow()/@llvm.sat() don't overflow, also try to prove other no-wrap	2019-10-18 19:32:47 +00:00
Utils	[PGO][PGSO] SizeOpts changes.	2019-10-18 16:46:01 +00:00
Vectorize	[SLP] avoid reduction transform on patterns that the backend can load-combine (2nd try)	2019-10-16 18:06:24 +00:00
CMakeLists.txt
LLVMBuild.txt