diff options
Diffstat (limited to 'llvm/lib/Target/X86/X86InstrSSE.td')
| -rw-r--r-- | llvm/lib/Target/X86/X86InstrSSE.td | 17 |
1 files changed, 17 insertions, 0 deletions
diff --git a/llvm/lib/Target/X86/X86InstrSSE.td b/llvm/lib/Target/X86/X86InstrSSE.td index 561ba99db4af..f8660a9fa123 100644 --- a/llvm/lib/Target/X86/X86InstrSSE.td +++ b/llvm/lib/Target/X86/X86InstrSSE.td @@ -577,20 +577,37 @@ let Predicates = [HasAVX, NoVLX] in { def : Pat<(alignedloadv8f16 addr:$src), (VMOVAPSrm addr:$src)>; + def : Pat<(alignedloadv8bf16 addr:$src), + (VMOVAPSrm addr:$src)>; def : Pat<(loadv8f16 addr:$src), (VMOVUPSrm addr:$src)>; + def : Pat<(loadv8bf16 addr:$src), + (VMOVUPSrm addr:$src)>; def : Pat<(alignedstore (v8f16 VR128:$src), addr:$dst), (VMOVAPSmr addr:$dst, VR128:$src)>; + def : Pat<(alignedstore (v8bf16 VR128:$src), addr:$dst), + (VMOVAPSmr addr:$dst, VR128:$src)>; def : Pat<(store (v8f16 VR128:$src), addr:$dst), (VMOVUPSmr addr:$dst, VR128:$src)>; + def : Pat<(store (v8bf16 VR128:$src), addr:$dst), + (VMOVUPSmr addr:$dst, VR128:$src)>; + def : Pat<(alignedloadv16f16 addr:$src), (VMOVAPSYrm addr:$src)>; + def : Pat<(alignedloadv16bf16 addr:$src), + (VMOVAPSYrm addr:$src)>; def : Pat<(loadv16f16 addr:$src), (VMOVUPSYrm addr:$src)>; + def : Pat<(loadv16bf16 addr:$src), + (VMOVUPSYrm addr:$src)>; def : Pat<(alignedstore (v16f16 VR256:$src), addr:$dst), (VMOVAPSYmr addr:$dst, VR256:$src)>; + def : Pat<(alignedstore (v16bf16 VR256:$src), addr:$dst), + (VMOVAPSYmr addr:$dst, VR256:$src)>; def : Pat<(store (v16f16 VR256:$src), addr:$dst), (VMOVUPSYmr addr:$dst, VR256:$src)>; + def : Pat<(store (v16bf16 VR256:$src), addr:$dst), + (VMOVUPSYmr addr:$dst, VR256:$src)>; } // Use movaps / movups for SSE integer load / store (one byte shorter). |