[PATCH] D50342: Changed how LLVM IR was generated to increase vectorization

emmettneyman updated this revision to Diff 159425. emmettneyman added a comment. Rebased and ready to land Repository: rC Clang https://reviews.llvm.org/D50342 Files: clang/tools/clang-fuzzer/proto-to-llvm/loop_proto_to_llvm.cpp Index: clang/tools/clang-fuzzer/proto-to-llvm/loop_proto_

[PATCH] D50342: Changed how LLVM IR was generated to increase vectorization

This revision was automatically updated to reflect the committed changes. Closed by commit rL339080: Changed how LLVM IR was generated to increase vectorization (authored by emmettneyman, committed by ). Changed prior to commit: https://reviews.llvm.org/D50342?vs=159425&id=159427#toc Repositor

[PATCH] D50482: Added another optimization pass to make vectorizing possible

emmettneyman created this revision. emmettneyman added reviewers: morehouse, kcc. Herald added a subscriber: cfe-commits. I noticed that my code wasn't going deep into the loop vectorizer code so added another pass that makes it go further. Repository: rC Clang https://reviews.llvm.org/D5048

[PATCH] D50482: Added another optimization pass to make vectorizing possible

emmettneyman updated this revision to Diff 159831. emmettneyman added a comment. - minor style fix Repository: rC Clang https://reviews.llvm.org/D50482 Files: clang/tools/clang-fuzzer/handle-llvm/handle_llvm.cpp Index: clang/tools/clang-fuzzer/handle-llvm/handle_llvm.cpp

[PATCH] D50482: Added another optimization pass to make vectorizing possible

emmettneyman added inline comments. Comment at: clang/tools/clang-fuzzer/handle-llvm/handle_llvm.cpp:126 + + auto Is TM guaranteed to be an LLVMTargetMachine? Yes, since the t

[PATCH] D50482: Added another optimization pass to make vectorizing possible

emmettneyman updated this revision to Diff 159834. emmettneyman added a comment. - Inlined function Repository: rC Clang https://reviews.llvm.org/D50482 Files: clang/tools/clang-fuzzer/handle-llvm/handle_llvm.cpp Index: clang/tools/clang-fuzzer/handle-llvm/handle_llvm.cpp ===

[PATCH] D50482: Added another optimization pass to make vectorizing possible

emmettneyman added inline comments. Comment at: clang/tools/clang-fuzzer/handle-llvm/handle_llvm.cpp:103 + Triple ModuleTriple(M->getTargetTriple()); + const TargetOptions Options = InitTargetOptionsFromCodeGenFlags(); morehouse wrote: > I think you can avoid

[PATCH] D50482: Added another optimization pass to make vectorizing possible

emmettneyman updated this revision to Diff 159838. emmettneyman added a comment. Rebase Repository: rC Clang https://reviews.llvm.org/D50482 Files: clang/tools/clang-fuzzer/handle-llvm/handle_llvm.cpp Index: clang/tools/clang-fuzzer/handle-llvm/handle_llvm.cpp =

[PATCH] D50482: Added another optimization pass to make vectorizing possible

This revision was automatically updated to reflect the committed changes. Closed by commit rL339305: Added another optimization pass to make vectorizing possible (authored by emmettneyman, committed by ). Changed prior to commit: https://reviews.llvm.org/D50482?vs=159838&id=159839#toc Reposito

[PATCH] D50530: Added LLVM metadata to generated IR to increase vectorization width

emmettneyman created this revision. emmettneyman added reviewers: morehouse, kcc. Herald added a subscriber: cfe-commits. Edited `loop_proto_to_llvm` to emit metadata at the end of the generated IR. This metadata will increase the vector width when the IR is optimized. Repository: rC Clang h

[PATCH] D50530: Added LLVM metadata to generated IR to increase vectorization width

emmettneyman added inline comments. Comment at: clang/tools/clang-fuzzer/proto-to-llvm/loop_proto_to_llvm.cpp:141 +<< "!1 = !{!\"llvm.loop.vectorize.enable\", i1 true}\n" +<< "!2 = !{!\"llvm.loop.vectorize.width\", i32 " << kArraySize +<< "}\n"

[PATCH] D50530: Added LLVM metadata to generated IR to increase vectorization width

emmettneyman updated this revision to Diff 159993. emmettneyman added a comment. Switched the include statement to the cpp file Repository: rC Clang https://reviews.llvm.org/D50530 Files: clang/tools/clang-fuzzer/proto-to-llvm/loop_proto_to_llvm.cpp Index: clang/tools/clang-fuzzer/proto-

[PATCH] D50530: Added LLVM metadata to generated IR to increase vectorization width

emmettneyman updated this revision to Diff 160016. emmettneyman added a comment. Rebased and ready to land Repository: rC Clang https://reviews.llvm.org/D50530 Files: clang/tools/clang-fuzzer/proto-to-llvm/loop_proto_to_llvm.cpp Index: clang/tools/clang-fuzzer/proto-to-llvm/loop_proto_to

[PATCH] D50530: Added LLVM metadata to generated IR to increase vectorization width

This revision was automatically updated to reflect the committed changes. Closed by commit rL339392: Added LLVM metadata to generated IR to increase vectorization width (authored by emmettneyman, committed by ). Changed prior to commit: https://reviews.llvm.org/D50530?vs=160016&id=160017#toc R

[PATCH] D50670: Implementation of multiple loops in cxx_loop_proto

emmettneyman created this revision. emmettneyman added reviewers: morehouse, kcc. Herald added a subscriber: cfe-commits. Extended `cxx_loop_proto` to have multiple for loops. Modified `loop_proto_to_llvm` and `loop_proto_to_cxx` to handle the new protos. In `loop_proto_to_llvm`, I only translat

[PATCH] D50670: Implementation of multiple loops in cxx_loop_proto

emmettneyman added a comment. No, it doesn't actually. I thought it would try to combine the separate loops into one block of vectorized instructions but after looking at the coverage of multiple loops vs single loop, they cover exactly the same parts of the Loop Vectorizer. Should I switch my

[PATCH] D50670: Implementation of multiple loops in cxx_loop_proto

emmettneyman updated this revision to Diff 160740. emmettneyman added a comment. Changed the multiloop protos to nested loop protos. All the protos have an inner loop and an outer loop. Repository: rC Clang https://reviews.llvm.org/D50670 Files: clang/tools/clang-fuzzer/cxx_loop_proto.pro

[PATCH] D50670: Implementation of nested loops in cxx_loop_proto

emmettneyman added a comment. In https://reviews.llvm.org/D50670#1200784, @morehouse wrote: > Does this hit new coverage in the vectorizer? Yes, this does hit new coverage in the loop vectorizer code. Comment at: clang/tools/clang-fuzzer/proto-to-llvm/loop_proto_to_llvm.cpp:

[PATCH] D50670: Implementation of nested loops in cxx_loop_proto

emmettneyman updated this revision to Diff 160866. emmettneyman edited the summary of this revision. emmettneyman added a comment. Small changes to generated IR Repository: rC Clang https://reviews.llvm.org/D50670 Files: clang/tools/clang-fuzzer/cxx_loop_proto.proto clang/tools/clang-fuz

[PATCH] D50670: Implementation of nested loops in cxx_loop_proto

emmettneyman updated this revision to Diff 160870. emmettneyman added a comment. Small changes to generated IR, my last change hadn't saved Repository: rC Clang https://reviews.llvm.org/D50670 Files: clang/tools/clang-fuzzer/cxx_loop_proto.proto clang/tools/clang-fuzzer/proto-to-cxx/loop

[PATCH] D50670: Implementation of nested loops in cxx_loop_proto

emmettneyman added inline comments. Comment at: clang/tools/clang-fuzzer/proto-to-llvm/loop_proto_to_llvm.cpp:132 + os << "target triple = \"x86_64-pc-linux-gnu\"\n" + << "define void @foo(i32* %a, i32* %b, i32* noalias %c, i64 %s) {\n" + << "%cmp = icmp sgt i64 %s, 0\n"

[PATCH] D50670: Implementation of nested loops in cxx_loop_proto

emmettneyman updated this revision to Diff 160920. emmettneyman added a comment. Made the inner loop optional Repository: rC Clang https://reviews.llvm.org/D50670 Files: clang/tools/clang-fuzzer/cxx_loop_proto.proto clang/tools/clang-fuzzer/proto-to-cxx/loop_proto_to_cxx.cpp clang/tool

[PATCH] D50670: Implementation of nested loops in cxx_loop_proto

emmettneyman updated this revision to Diff 160932. emmettneyman added a comment. Changed modifying global var to scoped class Repository: rC Clang https://reviews.llvm.org/D50670 Files: clang/tools/clang-fuzzer/cxx_loop_proto.proto clang/tools/clang-fuzzer/proto-to-cxx/loop_proto_to_cxx.

[PATCH] D50670: Implementation of nested loops in cxx_loop_proto

emmettneyman updated this revision to Diff 160936. emmettneyman added a comment. Updated comments, rebased, and ready to land Repository: rC Clang https://reviews.llvm.org/D50670 Files: clang/tools/clang-fuzzer/cxx_loop_proto.proto clang/tools/clang-fuzzer/proto-to-cxx/loop_proto_to_cxx.

[PATCH] D50670: Implementation of nested loops in cxx_loop_proto

This revision was automatically updated to reflect the committed changes. Closed by commit rC339832: Implementation of nested loops in cxx_loop_proto (authored by emmettneyman, committed by ). Changed prior to commit: https://reviews.llvm.org/D50670?vs=160936&id=160937#toc Repository: rC Cla

[PATCH] D50829: Update README and Dockerfile to include llvm-proto-fuzzer

emmettneyman created this revision. emmettneyman added reviewers: morehouse, kcc. Herald added a subscriber: cfe-commits. Added commands to Dockerfile to build llvm-proto-fuzzer and the other related tools. Also added a section to the bottom of the README describing what llvm-proto-fuzzer does a

[PATCH] D50829: Update README and Dockerfile to include llvm-proto-fuzzer

emmettneyman added a comment. Is there anything I should add to the documentation? Anything I should remove? Repository: rC Clang https://reviews.llvm.org/D50829 ___ cfe-commits mailing list cfe-commits@lists.llvm.org http://lists.llvm.org/cgi-bi

<    1   2