Search results

[InstCombine] Match range check pattern with SExt

llvm:instcombineCovers the InstCombine, InstSimplify and AggressiveInstCombine passesCovers the InstCombine, InstSimplify and AggressiveInstCombine passes

llvm:transforms

Status: Merged (completed).

llvm/llvm-projectnumber 118910

#118910 In llvm/llvm-project;

· by MatzeB was closed on Dec 10, 2024

[InstCombine] Optimistically allow multiple shufflevector uses in foldOpPhi

llvm:instcombineCovers the InstCombine, InstSimplify and AggressiveInstCombine passesCovers the InstCombine, InstSimplify and AggressiveInstCombine passes

llvm:transforms

Status: Merged (completed).

llvm/llvm-projectnumber 114278

#114278 In llvm/llvm-project;

· by MatzeB was closed on Dec 13, 2024

X86: Do not return invalid cost for fp16 conversion

backend:X86

llvm:transforms

Status: Merged (completed).

llvm/llvm-projectnumber 114128

#114128 In llvm/llvm-project;

· by MatzeB was closed on Oct 30, 2024

InstCombine: Fold shufflevector(select) and shufflevector(phi)

llvm:transforms

Status: Merged (completed).

llvm/llvm-projectnumber 113746

#113746 In llvm/llvm-project;

· by MatzeB was closed on Oct 28, 2024

InstCombine: Order shufflevector operands by complexity

backend:X86

clangClang issues not falling into any other categoryClang issues not falling into any other category

llvm:analysisIncludes value tracking, cost tables and constant foldingIncludes value tracking, cost tables and constant folding

llvm:transforms

Status: Closed (abandoned).

llvm/llvm-projectnumber 113212

#113212 In llvm/llvm-project;

· by MatzeB was closed on Apr 7, 2025

X86: Improve cost model of fp16 conversion

backend:X86

llvm:transforms

Status: Merged (completed).

llvm/llvm-projectnumber 113195

#113195 In llvm/llvm-project;

· by MatzeB was closed on Oct 25, 2024

GlobalOpt: Handle threadlocal.address intrinsic

llvm:transforms

Status: Merged (completed).

llvm/llvm-projectnumber 88454

#88454 In llvm/llvm-project;

· by MatzeB was closed on Apr 29, 2024

InstCombine: Increase threadlocal.address alignment if pointee is more aligned

llvm:analysisIncludes value tracking, cost tables and constant foldingIncludes value tracking, cost tables and constant folding

llvm:transforms

Status: Merged (completed).

llvm/llvm-projectnumber 88435

#88435 In llvm/llvm-project;

· by MatzeB was closed on Apr 16, 2024

CodeGenPrepare: Add support for llvm.threadlocal.address address-mode sinking

backend:X86

llvm:transforms

Status: Merged (completed).

llvm/llvm-projectnumber 87844

#87844 In llvm/llvm-project;

· by MatzeB was closed on Apr 17, 2024

Verify threadlocal_address constraints

llvm:ir

llvm:transforms

mlir

mlir:llvm

Status: Merged (completed).

llvm/llvm-projectnumber 87841

#87841 In llvm/llvm-project;

· by MatzeB was closed on Apr 9, 2024

LoopVectorize: Add better heuristic for vectorized epilogue skip test

llvm:transforms

vectorizers

Status: Merged (completed).

llvm/llvm-projectnumber 72589

#72589 In llvm/llvm-project;

· by MatzeB was closed on Nov 20, 2023

LoopVectorize: Set branch_weight for conditional branches

llvm:ir

llvm:transforms

PGOProfile Guided OptimizationsProfile Guided Optimizations

vectorizers

Status: Merged (completed).

llvm/llvm-projectnumber 72450

#72450 In llvm/llvm-project;

· by MatzeB was closed on Nov 16, 2023