aboutsummaryrefslogtreecommitdiff
path: root/clang/lib/Analysis/FlowSensitive/DataflowAnalysisContext.cpp
diff options
context:
space:
mode:
authorLuke Lau <luke@igalia.com>2023-02-10 11:03:55 +0000
committerLuke Lau <luke@igalia.com>2023-03-23 10:33:50 +0000
commit8d16c6809a080947057ae21b9f6165105b4b2ad8 (patch)
treed2246958642fba49d0a547433e80b8b7aa37569e /clang/lib/Analysis/FlowSensitive/DataflowAnalysisContext.cpp
parent61f33def1375a68afc5681627a62ce24446e45e2 (diff)
downloadllvm-8d16c6809a080947057ae21b9f6165105b4b2ad8.zip
llvm-8d16c6809a080947057ae21b9f6165105b4b2ad8.tar.gz
llvm-8d16c6809a080947057ae21b9f6165105b4b2ad8.tar.bz2
[RISCV] Increase default vectorizer LMUL to 2
After some discussion and experimentation, we have seen that changing the default number of vector register bits to LMUL=2 strikes a sweet spot. Whilst we could be clever here and make the vectorizer smarter about dynamically selecting an LMUL that a) Doesn't affect register pressure b) Suitable for the microarchitecture we would need to teach its heuristics about RISC-V register grouping specifics. Instead this just does the easy, pragmatic thing by changing the default to a safe value that doesn't affect register pressure signifcantly[1], but should increase throughput and unlock more interleaving. [1] Register spilling when compiling sqlite at various levels of `-riscv-v-register-bit-width-lmul`: LMUL=1 2573 spills LMUL=2 2583 spills LMUL=4 2819 spills LMUL=8 3256 spills Reviewed By: craig.topper Differential Revision: https://reviews.llvm.org/D143723
Diffstat (limited to 'clang/lib/Analysis/FlowSensitive/DataflowAnalysisContext.cpp')
0 files changed, 0 insertions, 0 deletions