[PATCH] D92762: [clang][AArch64][SVE] Avoid going through memory for coerced VLST arguments

This revision was automatically updated to reflect the committed changes. Closed by commit rG3d5b18a3fdf4: [clang][AArch64][SVE] Avoid going through memory for coerced VLST arguments (authored by joechrisellis). Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.ll

[PATCH] D92762: [clang][AArch64][SVE] Avoid going through memory for coerced VLST arguments

c-rhodes added inline comments. Comment at: clang/lib/CodeGen/CGCall.cpp:2674 + // perform the conversion. + if (Ty->getAs()) { +auto *Coerced = Fn->getArg(FirstIRArg); joechrisellis wrote: > c-rhodes wrote: > > Do we want to check `VT->getVecto

[PATCH] D92762: [clang][AArch64][SVE] Avoid going through memory for coerced VLST arguments

joechrisellis marked an inline comment as done. joechrisellis added inline comments. Comment at: clang/lib/CodeGen/CGCall.cpp:2671-2673 + // See if this is a VLST coerced to a VLAT at the function boundary and + // the types match up. If so, use llvm.experimental.vector

[PATCH] D92762: [clang][AArch64][SVE] Avoid going through memory for coerced VLST arguments

joechrisellis updated this revision to Diff 314365. joechrisellis added a comment. Address @c-rhodes's comment. - Use name `castFixedSve` instead of `castScalableSve`. Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D92762/new/ https://reviews.llvm.o

[PATCH] D92762: [clang][AArch64][SVE] Avoid going through memory for coerced VLST arguments

c-rhodes accepted this revision. c-rhodes added inline comments. This revision is now accepted and ready to land. Comment at: clang/lib/CodeGen/CGCall.cpp:2706 + ArgVals.push_back(ParamValue::forDirect(Builder.CreateExtractVector( +VecTyTo, Coerced, Ze

[PATCH] D92762: [clang][AArch64][SVE] Avoid going through memory for coerced VLST arguments

joechrisellis updated this revision to Diff 314354. joechrisellis marked 2 inline comments as done. joechrisellis added a comment. Address @c-rhodes's comments. - Make comment clearer. - Change structure of the if statements for better readability. - Drop superfluous `this->`. Repository: rG

[PATCH] D92762: [clang][AArch64][SVE] Avoid going through memory for coerced VLST arguments

c-rhodes added inline comments. Comment at: clang/lib/CodeGen/CGCall.cpp:2671-2673 + // See if this is a VLST coerced to a VLAT at the function boundary and + // the types match up. If so, use llvm.experimental.vector.extract to + // perform the conversion. ---

[PATCH] D92762: [clang][AArch64][SVE] Avoid going through memory for coerced VLST arguments

joechrisellis updated this revision to Diff 310532. joechrisellis added a comment. Herald added a subscriber: NickHung. Rebase. Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D92762/new/ https://reviews.llvm.org/D92762 Files: clang/lib/CodeGen/CGC

[PATCH] D92762: [clang][AArch64][SVE] Avoid going through memory for coerced VLST arguments

joechrisellis updated this revision to Diff 310185. joechrisellis added a comment. Do not try to generate `llvm.experimental.vector.extract` instructions if the vector element types differ. Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D92762/new/

[PATCH] D92762: [clang][AArch64][SVE] Avoid going through memory for coerced VLST arguments

joechrisellis created this revision. joechrisellis added reviewers: peterwaller-arm, DavidTruby. Herald added subscribers: psnobl, kristof.beyls, tschuett. Herald added a reviewer: efriedma. joechrisellis requested review of this revision. Herald added a project: clang. Herald added a subscriber: c