Until mischeduler is clever enough to avoid spilling in a vectorized loop with many (scalar) DLRs it is better to avoid high vectorization factors (8 and above). llvm-svn: 344129 |
||
|---|---|---|
| .. | ||
| AArch64 | ||
| AMDGPU | ||
| ARM | ||
| PowerPC | ||
| SystemZ | ||
| X86 | ||
| no_info.ll | ||
Until mischeduler is clever enough to avoid spilling in a vectorized loop with many (scalar) DLRs it is better to avoid high vectorization factors (8 and above). llvm-svn: 344129 |
||
|---|---|---|
| .. | ||
| AArch64 | ||
| AMDGPU | ||
| ARM | ||
| PowerPC | ||
| SystemZ | ||
| X86 | ||
| no_info.ll | ||