llvm-project

History

AMDGPU: Handle cvt_scale F32/F16->F4/F8 gfx950 hazard (#117844 )

gfx950 SP changes doc says:
No 4 clk forwarding on opcodes that convert from
F32/F16->F8 or F32/F16->F4. Must insert a NOP or
instruction writing some other destination VREG
after a conversion to F4/F8 since it writes either
low/high half or bytes.

Co-authored-by: Pravin Jagtap <Pravin.Jagtap@amd.com>
Co-authored-by: Jeffrey Byrnes <Jeffrey.Byrnes@amd.com>

2024-12-02 09:23:17 -05:00

AMDGPUAsmUtils.cpp

[AMDGPU][MC] Implement fft and rotate modes for ds_swizzle_b32 (#108064 )

2024-09-27 10:18:34 -07:00

AMDGPUAsmUtils.h

[AMDGPU][MC] Support UC_VERSION_* constants. (#95618 )

2024-06-18 15:44:14 +01:00

AMDGPUBaseInfo.cpp

AMDGPU: Handle cvt_scale F32/F16->F4/F8 gfx950 hazard (#117844 )

2024-12-02 09:23:17 -05:00

AMDGPUBaseInfo.h

AMDGPU: Handle cvt_scale F32/F16->F4/F8 gfx950 hazard (#117844 )