Search results

[InstCombine] Rewrite multi-use GEPs when simplifying comparison

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

llvm:transforms

Status: Open (in progress).

llvm/llvm-projectnumber 146100

#146100 In llvm/llvm-project;

· nikic opened on Jun 27, 2025

[InstCombine] Optimize binOp(select(),min/max()).

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

llvm:transforms

Status: Open (in progress).

llvm/llvm-projectnumber 145830

#145830 In llvm/llvm-project;

· vzakhari opened on Jun 26, 2025

[InstCombine] smin(smax(X, -1), 1) -> scmp(X, 0) and smax(smin(X, 1), -1) -> scmp(X, 0)

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

llvm:transforms

Status: Open (in progress).

llvm/llvm-projectnumber 145736

#145736 In llvm/llvm-project;

· AZero13 opened on Jun 25, 2025

[InstCombine] Treat identical operands as one in pushFreezeToPreventPoisonFromPropagating

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

llvm:transforms

Status: Open (in progress).

llvm/llvm-projectnumber 145348

#145348 In llvm/llvm-project;

· c-rhodes opened on Jun 23, 2025

[Reland][ValueTracking] Improve Bitcast handling to match SDAG

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

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

llvm:transforms

Status: Open (in progress).

llvm/llvm-projectnumber 145223

#145223 In llvm/llvm-project;

· abhishek-kaushik22 opened on Jun 22, 2025

[AMDGPU] Optimize image sample followed by llvm.amdgcn.cvt.pkrtz into d16 variant

backend:AMDGPU

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

llvm:transforms

Status: Open (in progress).

llvm/llvm-projectnumber 145203

#145203 In llvm/llvm-project;

· harrisonGPU opened on Jun 22, 2025

[InstCombine] Fold reconstruction across select

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

llvm:transforms

Status: Open (in progress).

llvm/llvm-projectnumber 145102

#145102 In llvm/llvm-project;

· macsencasaus opened on Jun 20, 2025

[LangRef] Require that vscale be a power of two

backend:AArch64

backend:RISC-V

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

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

llvm:ir

llvm:SelectionDAGSelectionDAGISel as wellSelectionDAGISel as well

llvm:transforms

vectorizers

Status: Open (in progress).

llvm/llvm-projectnumber 145098

#145098 In llvm/llvm-project;

· preames opened on Jun 20, 2025

[InstCombine] Pull unary shuffles through fneg/fabs

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

llvm:transforms

Status: Open (in progress).

llvm/llvm-projectnumber 144933

#144933 In llvm/llvm-project;

· lukel97 opened on Jun 19, 2025

[InstCombine] Generate better code for <code>std::bit_floor</code> from libstdc++

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

llvm:transforms

Status: Open (in progress).

llvm/llvm-projectnumber 144842

#144842 In llvm/llvm-project;

· el-ev opened on Jun 19, 2025

[VectorCombine] Recognise reduction patterns from extract+binop+shuffle chains

llvm::vectorcombineCost-based vector combine passCost-based vector combine pass

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

missed-optimization

Status: Open.

Feature

#144654 In llvm/llvm-project;

· RKSimon opened on Jun 18, 2025

Allow folding icmp eq (add X, C2), C when there is more than one-use when we can compute the range

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

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

llvm:ir

llvm:transforms

Status: Open (in progress).

llvm/llvm-projectnumber 144566

#144566 In llvm/llvm-project;

· AZero13 opened on Jun 17, 2025