Yaxun Liu 3c42f1c3c9 LoopUnroll: respect pragma unroll when AllowRemainder is disabled ...

Currently when AllowRemainder is disabled, pragma unroll count is not
respected even though there is no remainder. This bug causes a loop
fully unrolled in many cases even though the user specifies a unroll
count. Especially it affects OpenCL/CUDA since in many cases a loop
contains convergent instructions and currently AllowRemainder is
disabled for such loops.

Differential Revision: https://reviews.llvm.org/D43826

llvm-svn: 326585

2018-03-02 16:22:32 +00:00

..

AggressiveInstCombine

[AggressiveInstCombine] Fixed TruncCombine class to handle TruncInst leaf node correctly.

2018-01-31 22:39:05 +00:00

Coroutines

[Coroutines] Move debug statement before assert

2018-02-19 20:50:09 +00:00

Hello

…

InstCombine

[InstCombine] allow fmul fold with less than 'fast'

2018-03-02 00:14:51 +00:00

Instrumentation

[InstrProfiling] Emit the runtime hook when no counters are lowered

2018-02-28 19:00:08 +00:00

IPO

[ArgumentPromotion] don't break musttail invariant PR36543

2018-03-02 00:59:27 +00:00

ObjCARC

[ObjCARC] Do not turn a call to @objc_autoreleaseReturnValue into a call

2018-01-19 23:51:13 +00:00

Scalar

LoopUnroll: respect pragma unroll when AllowRemainder is disabled

2018-03-02 16:22:32 +00:00

Utils

[SimplifyLibCalls] Update an obviously copy and pasted header comment to match this file. NFC

2018-03-01 20:05:09 +00:00

Vectorize

[LV][CFG] Add irreducible CFG detection for outer loops

2018-03-02 12:24:25 +00:00

CMakeLists.txt

Another try to commit 323321 (aggressive instruction combine).

2018-01-25 12:06:32 +00:00

LLVMBuild.txt

Another try to commit 323321 (aggressive instruction combine).

2018-01-25 12:06:32 +00:00