Details
- Reviewers
craig.topper MaskRay asb luismarques
Diff Detail
Unit Tests
Time | Test | |
---|---|---|
1,240 ms | x64 debian > libFuzzer.libFuzzer::fuzzer-finalstats.test |
Event Timeline
llvm/test/CodeGen/RISCV/rv32zbs.ll | ||
---|---|---|
805 | Everything should always be optimised, but you mean a specific optimisation doesn't apply |
llvm/test/CodeGen/RISCV/rv32zbs.ll | ||
---|---|---|
805 | This patch will show how lui r1, xxx addi r1, r1, yyy or r0, r0, r1 will be optimized to ori r0, r0, zzz bseti r0, 11 bseti r0, hi An extra register is saved. |
llvm/test/CodeGen/RISCV/rv32zbs.ll | ||
---|---|---|
805 | Yes, the comment is complete nonsense when you read the file, it only makes some amount of sense when looking specifically at this diff, but even then is not the best way to express that. |
Sorry. I mean to replace the previous patch with new tests. And I will submit the optimization patch later.
llvm/test/CodeGen/RISCV/rv32zbs.ll | ||
---|---|---|
805 | I see. I will pay attention my way in the future. |
Everything should always be optimised, but you mean a specific optimisation doesn't apply