Changes in shuffle-insert_subvector.ll

4455c5cd - [CostModel][X86] Update RUN -passes=* to double quotes to appease update scripts on windows

Fri, 18 Mar 2022 11:44:07 +0000

[CostModel][X86] Update RUN -passes=* to double quotes to appease update scripts on windows List of files: /llvm-project-15.0.7/llvm/test/Analysis/CostModel/X86/shuffle-insert_subvector.ll

15ba588d - [test] Migrate '-analyze -cost-model' to '-passes=print'

Wed, 09 Feb 2022 23:25:18 +0000

[test] Migrate '-analyze -cost-model' to '-passes=print' List of files: /llvm-project-15.0.7/llvm/test/Analysis/CostModel/X86/shuffle-insert_subvector.ll

872a9500 - [CostModel] Treat 'widen subvector' patterns as zero cost

Mon, 02 Aug 2021 10:42:10 +0000

[CostModel] Treat 'widen subvector' patterns as zero costAs discussed on D107228, widening a subvector by inserting the whole subvector into the bottom a larger undef vector should always be cheap enough that we can treat it as zero cost.NOTE: If this proves to cause issues we have the option of introducing a "SK_WidenSubvector" shuffle kind enum that targets could override the zero cost, but that doesn't seem necessary atm.Differential Revision: https://reviews.llvm.org/D107228 List of files: /llvm-project-15.0.7/llvm/test/Analysis/CostModel/X86/shuffle-insert_subvector.ll

7397dcb4 - [TTI] Add basic SK_InsertSubvector shuffle mask recognition

Mon, 02 Aug 2021 10:21:42 +0000

[TTI] Add basic SK_InsertSubvector shuffle mask recognitionThis patch adds an initial ShuffleVectorInst::isInsertSubvectorMask helper to recognize 2-op shuffles where the lowest elements of one of the sources are being inserted into the "in-place" other operand, this includes "concat_vectors" patterns as can be seen in the Arm shuffle cost changes. This also helped fix a x86 issue with irregular/length-changing SK_InsertSubvector costs - I'm hoping this will help with D107188This doesn't currently attempt to work with 1-op shuffles that could either be a "widening" shuffle or a self-insertion.The self-insertion case is tricky, but we currently always match this with the existing SK_PermuteSingleSrc logic.The widening case will be addressed in a follow up patch that treats the cost as 0.Masks with a high number of undef elts will still struggle to match optimal subvector widths - its currently bounded by minimum-width possible insertion, whilst some cases would benefit from wider (pow2?) subvectors.Differential Revision: https://reviews.llvm.org/D107228 List of files: /llvm-project-15.0.7/llvm/test/Analysis/CostModel/X86/shuffle-insert_subvector.ll

4ca86074 - [InstructionCost] Don't conflate Invalid costs with Unknown costs.

Tue, 30 Mar 2021 07:54:59 +0000

[InstructionCost] Don't conflate Invalid costs with Unknown costs.We previously made a change to getUserCost to return a Invalid costwhen one of the TTI costs returned '-1' (meaning 'unknown' or'infinitely expensive'). It makes no sense to say that: shufflevector <2 x i8> %x, <2 x i8> %y, <4 x i32> has an invalid cost. Perhaps the cost is not known, but the IR is validand can be code-generated. Invalid should only be used for IR thatcannot possibly be code-generated and where a cost is nonsensical.With more passes now asserting that the cost must be valid, it is possiblethat those assertions will fail for perfectly valid IR. An incompletecost-model probably shouldn't be a reason for the compiler to break.It's better to consider these costs as 'very expensive' and ignore themfor other reasons. At some point, we should consider replacing -1 withsome other mechanism.Reviewed By: paulwalker-arm, dmgreenDifferential Revision: https://reviews.llvm.org/D99502 List of files: /llvm-project-15.0.7/llvm/test/Analysis/CostModel/X86/shuffle-insert_subvector.ll

fe9403df - [CostModel][X86] Remove unused check-prefixes

Tue, 10 Nov 2020 12:48:02 +0000

[CostModel][X86] Remove unused check-prefixes List of files: /llvm-project-15.0.7/llvm/test/Analysis/CostModel/X86/shuffle-insert_subvector.ll

ec24e505 - [CostModel][X86] add CostModel for SK_Select(v8f64, v8i64, v16f32, v16i32, v32i16, v64i8)

Wed, 23 Sep 2020 02:13:03 +0000

[CostModel][X86] add CostModel for SK_Select(v8f64, v8i64, v16f32, v16i32, v32i16, v64i8)add CostModel for SK_Select(v8f64, v8i64, v16f32, v16i32, v32i16, v64i8)Reviewed By: RKSimonDifferential Revision: https://reviews.llvm.org/D87884 List of files: /llvm-project-15.0.7/llvm/test/Analysis/CostModel/X86/shuffle-insert_subvector.ll

ff669190 - [X86][CostModel] Bump the cost of vpermw/vpermt2b/vperm2w

Thu, 30 Apr 2020 17:52:35 +0000

[X86][CostModel] Bump the cost of vpermw/vpermt2b/vperm2wvpermw is 2 uops. vpermt2b/vpermt2w are two shuffle uops and a port 015 uop. Weirdly vpermb is a single uop.This patch bumps the cost to 2 for these operations. Maybe should go to 3 for the vpermt2*, but I've started conservative.I've also removed a few entries that were now the same as earlier subtargets or that I didn't think we really did. Like I don't think we extend v32i8 to v32i16, shuffle, and then truncate.Differential Revision: https://reviews.llvm.org/D79148 List of files: /llvm-project-15.0.7/llvm/test/Analysis/CostModel/X86/shuffle-insert_subvector.ll

12c629ec - [CostModel][X86] Add shuffle costs for some common sub-128bit vectors

Thu, 09 Apr 2020 18:55:51 +0000

[CostModel][X86] Add shuffle costs for some common sub-128bit vectorsv2i8/v4i8/v8i8 + v2i16/v4i16 all show up in vectorizer code and by just using the legalized types (v16i8/v8i16) we're highly exaggerating the actual cost of the shuffle. List of files: /llvm-project-15.0.7/llvm/test/Analysis/CostModel/X86/shuffle-insert_subvector.ll

be84d2b5 - [CostModel][X86] Add some insert subvector cost tests for vXf32/vXi32/vXi16/vXi8 types

Sat, 04 Apr 2020 21:46:36 +0000

[CostModel][X86] Add some insert subvector cost tests for vXf32/vXi32/vXi16/vXi8 types List of files: /llvm-project-15.0.7/llvm/test/Analysis/CostModel/X86/shuffle-insert_subvector.ll

168a44a7 - [CostModel][X86] Improve extract/insert element costs (PR43605)

Thu, 27 Feb 2020 15:36:13 +0000

[CostModel][X86] Improve extract/insert element costs (PR43605)This tries to improve the accuracy of extract/insert element costs by accounting for subvector extraction/insertion for >128-bit vectors and the shuffling of elements to/from the 0'th index.It also adds INSERTPS for f32 types and PINSR/PEXTR costs for integer types (at the moment we assume the same cost as MOVD/MOVQ - which isn't always true).Differential Revision: https://reviews.llvm.org/D74976 List of files: /llvm-project-15.0.7/llvm/test/Analysis/CostModel/X86/shuffle-insert_subvector.ll

eaa41e10 - [CostModel][X86] Try to check against common prefixes before using target-specific cpu checks

Mon, 24 Feb 2020 11:58:50 +0000

[CostModel][X86] Try to check against common prefixes before using target-specific cpu checksSLM/GLM is still a mess so not all of them have been updated yet. List of files: /llvm-project-15.0.7/llvm/test/Analysis/CostModel/X86/shuffle-insert_subvector.ll

acd10b9f - [CostModel][X86] Add some initial extract/insert subvector shuffle cost tests

Sat, 20 Oct 2018 17:38:33 +0000

[CostModel][X86] Add some initial extract/insert subvector shuffle cost testsJust f64/i64 tests initially to demonstrate PR39368llvm-svn: 344857 List of files: /llvm-project-15.0.7/llvm/test/Analysis/CostModel/X86/shuffle-insert_subvector.ll