summaryrefslogtreecommitdiff
path: root/llvm/lib/Target/AArch64/AArch64TargetTransformInfo.cpp
AgeCommit message (Expand)Author
2025-11-19[TTI] Use MemIntrinsicCostAttributes for getMaskedMemoryOpCost (#168029)Shih-Po Hung
2025-11-18[AArch64] - Improve costing for Identity shuffles for SVE targets. (#165375)Pawan Nirpal
2025-11-17[CostModel][AArch64] Remove promotion cost for SVE bfloat arith supported wit...Benjamin Maxwell
2025-11-17[AArch64] Allow forcing unrolling of small loops (#167488)Vladi Krapp
2025-11-15[AArch64] Move AArch64SMEAttributes out of Utils library to fix layering. NFC...Craig Topper
2025-11-06Revert "[InterleavedAccess] Construct interleaved access store with shuffles"Martin Storsjö
2025-11-05[InterleavedAccess] Construct interleaved access store with shufflesRamkrishnan
2025-11-04[LAA] Only use inbounds/nusw in isNoWrap if the GEP is dereferenced. (#161445)Florian Hahn
2025-11-04[AArch64] Improve the cost model for extending mull (#125651)David Green
2025-11-03[AArch64] Sink mismatching wide extends to mul (#164986)David Green
2025-11-02[llvm] Remove redundant typename (NFC) (#166087)Kazu Hirata
2025-11-01[Analysis, CodeGen] Use ArrayRef instead of const ArrayRef (NFC) (#166026)Kazu Hirata
2025-10-31[CostModel][AArch64] Model cost of extract.last.active intrinsic (clastb) (#1...Graham Hunter
2025-10-24AArch64: Use RuntimeLibcallsInfo in SMEAttributes (NFC) (#164968)Matt Arsenault
2025-10-24[AArch64][CostModel] Reduce cost of wider than legal get.active.lane.mask (#1...Kerry McLaughlin
2025-10-23[LLVM][InstCombine][SVE] Improve isAllActivePredicate by looking through from...Paul Walker
2025-10-20[AArch64][CostModel] Add constraints on which partial reductions are (#163728)Sushant Gokhale
2025-10-06[AArch64] Return Invalid partial reduction cost for i128 accumulator. (#162066)Sander de Smalen
2025-10-03[AArch64] Refactor and refine cost-model for partial reductions (#158641)Sander de Smalen
2025-09-12[AArch64][InstCombine] Canonicalize whilelo intrinsic (#151553)Matthew Devereau
2025-09-12[AArch64][SME] Improve codegen for aarch64.sme.cnts* when not in streaming mo...Kerry McLaughlin
2025-09-09[AArch64] Enable RT and partial unrolling with reductions for Apple CPUs. (#1...Florian Hahn
2025-09-04[AArch64] Give a higher cost for more expensive SVE FCMP instructions (#153816)David Green
2025-09-04[LV][AArch64] Prefer epilogue with fixed-width over scalable VF. (#155546)Hassnaa Hamdi
2025-09-03[AArch64] Update cost model for extracting halves from 128+ bit vectors (#155...Gaëtan Bossu
2025-09-01[LV] Bundle sub reductions into VPExpressionRecipe (#147255)Sam Tebbs
2025-08-21[AArch64] [CostModel] Fix cost modelling for saturating arithmetic intrinsics...Mary Kassayova
2025-08-20[AArch64][SME] Lower aarch64.sme.cnts* to vscale when in streaming mode (#154...Kerry McLaughlin
2025-08-19[LV][TTI] Calculate cost of extracting last index in a scalable vector (#144086)David Sherwood
2025-08-18Reland "[AArch64][SME] Port all SME routines to RuntimeLibcalls" (#153417)Benjamin Maxwell
2025-08-18Reorder checks to speed up getAppleRuntimeUnrollPreferences() (#154010)Ahmad Yasin
2025-08-14[AArch64] Change the cost of fma and fmuladd to match fmul. (#152963)David Green
2025-08-14[TTI] Add cost kind to getAddressComputationCost(). NFC. (#153342)Elvis Wang
2025-08-13Revert "[AArch64][SME] Port all SME routines to RuntimeLibcalls" (#153392)Nikita Popov
2025-08-13[AArch64] Tune unrolling prefs for more patterns on Apple CPUs (#149358)Ahmad Yasin
2025-08-13[AArch64][SME] Port all SME routines to RuntimeLibcalls (#152505)Benjamin Maxwell
2025-08-12[LV] Create in-loop sub reductions (#147026)Sam Tebbs
2025-08-11[TTI] Consistently pass the pointer type to getAddressComputationCost. NFCI (...Luke Lau
2025-08-08[AArch64] Rename Cost -> PromotedCost to avoid shadowing errorDavid Green
2025-08-08[AArch64] Generalize costing for FP16 instructions (#150033)David Green
2025-08-08[CostModel] Provide a default model for histogram intrinsics (#149348)Graham Hunter
2025-08-05[LLVM][CGP] Allow finer control for sinking compares. (#151366)Paul Walker
2025-08-04[AArch64] Add better fcmp costs for expanded predicates (#147940)David Green
2025-08-04[AArch64] Add sve bf16 fpext and fpround costs. (#150485)David Green
2025-08-01[AArch64] Dont inline streaming fn into non-streaming caller (#150595)Sander de Smalen
2025-07-31[AArch64] Allow unrolling of scalar epilogue loops (#151164)John Brawn
2025-07-23[NFC][Clang][FMV] Make FMV priority data type future proof. (#150079)Alexandros Lamprineas
2025-07-16[AArch64] Reduce the costs of and/or/xor reductions (#148553)David Green
2025-07-15[DebugInfo][RemoveDIs] Suppress getNextNonDebugInfoInstruction (#144383)Jeremy Morse
2025-07-15[AArch64,TTI] Remove RealUse check for vector insert/extract costs. (#146526)Florian Hahn
generated by cgit v1.2.3 (git 2.25.1) at 2026-06-19 06:58:13 +0000