David Green cf1b18658c [InstCombine] Fold ~A - Min/Max(~A, O) -> Max/Min(A, ~O) - A ...

This is an attempt to get out of a local-minimum that instcombine currently
gets stuck in. We essentially combine two optimisations at once, ~a - ~b = b-a
and min(~a, ~b) = ~max(a, b), only doing the transform if the result is at
least neutral. This involves using IsFreeToInvert, which has been expanded a
little to include selects that can be easily inverted.

This is trying to fix PR35875, using the ideas from Sanjay. It is a large
improvement to one of our rgb to cmy kernels.

Differential Revision: https://reviews.llvm.org/D52177


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@343569 91177308-0d34-0410-b5e6-96231b3b80d8

2018-10-02 09:48:34 +00:00

..

CMakeLists.txt

InstCombine/AMDGPU: Add dimension-aware image intrinsics to SimplifyDemanded

2018-06-21 13:37:31 +00:00

InstCombineAddSub.cpp

[InstCombine] Fold ~A - Min/Max(~A, O) -> Max/Min(A, ~O) - A

2018-10-02 09:48:34 +00:00

InstCombineAndOrXor.cpp

[InstCombine] Fold (xor (min/max X, Y), -1) -> (max/min ~X, ~Y) when X and Y are freely invertible.

2018-09-13 18:52:58 +00:00

InstCombineCalls.cpp

[InstCombine][x86] try even harder to convert blendv intrinsic to generic IR (PR38814)

2018-09-22 14:43:55 +00:00

InstCombineCasts.cpp

Fix InstCombine address space assert

2018-07-31 15:53:03 +00:00

InstCombineCompares.cpp

[InstCombine] Handle vector compares in foldGEPIcmp(), take 2

2018-10-01 14:59:25 +00:00

InstCombineInternal.h

[InstCombine] Fold ~A - Min/Max(~A, O) -> Max/Min(A, ~O) - A

2018-10-02 09:48:34 +00:00

InstCombineLoadStoreAlloca.cpp

[Local] Make DoesKMove required for combineMetadata.

2018-08-24 11:40:04 +00:00

InstCombineMulDivRem.cpp

[InstCombine] refactor mul narrowing folds; NFCI

2018-09-14 22:23:35 +00:00

InstCombinePHI.cpp

[IR] Replace isa<TerminatorInst> with isTerminator().

2018-08-26 09:51:22 +00:00

InstCombineSelect.cpp

[InstCombine] Fold (min/max ~X, Y) -> ~(max/min X, ~Y) when Y is freely invertible

2018-09-22 05:53:27 +00:00

InstCombineShifts.cpp

Remove trailing space

2018-07-30 19:41:25 +00:00

InstCombineSimplifyDemanded.cpp

[InstCombine] enhance vector demanded elements to look at a vector select condition operand

2018-09-11 18:49:00 +00:00

InstCombineTables.td

InstCombine/AMDGPU: Add dimension-aware image intrinsics to SimplifyDemanded

2018-06-21 13:37:31 +00:00

InstCombineVectorOps.cpp

[InstCombine] try to convert vector insert+extract to trunc; 2nd try

2018-10-01 14:40:00 +00:00

InstructionCombining.cpp

[InstCombine] narrow binops on concatenated vectors (PR33026)

2018-09-25 15:57:37 +00:00

LLVMBuild.txt

Update libdeps since TLI was moved from Target to Analysis in r226078.

2015-01-15 05:21:00 +00:00