diff options
| author | Craig Topper <craig.topper@sifive.com> | 2024-12-12 16:17:21 -0800 |
|---|---|---|
| committer | GitHub <noreply@github.com> | 2024-12-12 16:17:21 -0800 |
| commit | 88c18da37dfb10d570414bcb92ad075241f1b7c3 (patch) | |
| tree | 401c57eeba8f783db14eaded10badc184c6647ee /llvm/lib/CodeGen/MachineBasicBlock.cpp | |
| parent | d01c11df04ae45a3d5b08e69bb683c760bbddd54 (diff) | |
| download | llvm-88c18da37dfb10d570414bcb92ad075241f1b7c3.zip llvm-88c18da37dfb10d570414bcb92ad075241f1b7c3.tar.gz llvm-88c18da37dfb10d570414bcb92ad075241f1b7c3.tar.bz2 | |
[RISCV] Rename suffixes on VCPOP/VMSBF/VMSET/etc pseudos. NFC (#119785)
These are suffixed with B1, B2, B4, B8, B16, B32, or B64 which I think
these were supposed to match the naming of the vbool types from C where
the number should be SEW/LMUL. So the smallest mask is 64 and the
largest is 1. This provides a compact syntax for describing the 7
possible ratios between LMUL and SEW.
We had the instruction names in the opposite order.
Diffstat (limited to 'llvm/lib/CodeGen/MachineBasicBlock.cpp')
0 files changed, 0 insertions, 0 deletions