diff options
| author | Farhana Aleen <farhana.aleen@gmail.com> | 2018-07-16 18:19:59 +0000 |
|---|---|---|
| committer | Farhana Aleen <farhana.aleen@gmail.com> | 2018-07-16 18:19:59 +0000 |
| commit | c370d7b33d0aba8b80f3a0b633f4f8d92c448833 (patch) | |
| tree | b4e3f2030a2771b0aba9849c07db42611c322fb4 /llvm/lib/Target/AMDGPU/VOP3PInstructions.td | |
| parent | 7f01d209939c3a08c7d15ea145d7056dc138c23a (diff) | |
| download | bcm5719-llvm-c370d7b33d0aba8b80f3a0b633f4f8d92c448833.tar.gz bcm5719-llvm-c370d7b33d0aba8b80f3a0b633f4f8d92c448833.zip | |
[AMDGPU] [AMDGPU] Support a fdot2 pattern.
Summary: Optimize fma((float)S0.x, (float)S1.x fma((float)S0.y, (float)S1.y, z))
-> fdot2((v2f16)S0, (v2f16)S1, (float)z)
Author: FarhanaAleen
Reviewed By: rampitec, b-sumner
Subscribers: AMDGPU
Differential Revision: https://reviews.llvm.org/D49146
llvm-svn: 337198
Diffstat (limited to 'llvm/lib/Target/AMDGPU/VOP3PInstructions.td')
| -rw-r--r-- | llvm/lib/Target/AMDGPU/VOP3PInstructions.td | 2 |
1 files changed, 1 insertions, 1 deletions
diff --git a/llvm/lib/Target/AMDGPU/VOP3PInstructions.td b/llvm/lib/Target/AMDGPU/VOP3PInstructions.td index 8e2eff13d6d..5c78ada3211 100644 --- a/llvm/lib/Target/AMDGPU/VOP3PInstructions.td +++ b/llvm/lib/Target/AMDGPU/VOP3PInstructions.td @@ -167,7 +167,7 @@ defm : MadFmaMixPats<fma, V_FMA_MIX_F32, V_FMA_MIXLO_F16, V_FMA_MIXHI_F16>; let SubtargetPredicate = HasDLInsts in { -def V_DOT2_F32_F16 : VOP3PInst<"v_dot2_f32_f16", VOP3_Profile<VOP_F32_V2F16_V2F16_F32>, int_amdgcn_fdot2>; +def V_DOT2_F32_F16 : VOP3PInst<"v_dot2_f32_f16", VOP3_Profile<VOP_F32_V2F16_V2F16_F32>, AMDGPUfdot2>; def V_DOT2_I32_I16 : VOP3PInst<"v_dot2_i32_i16", VOP3_Profile<VOP_I32_V2I16_V2I16_I32>, int_amdgcn_sdot2>; def V_DOT2_U32_U16 : VOP3PInst<"v_dot2_u32_u16", VOP3_Profile<VOP_I32_V2I16_V2I16_I32>, int_amdgcn_udot2>; def V_DOT4_I32_I8 : VOP3PInst<"v_dot4_i32_i8", VOP3_Profile<VOP_I32_I32_I32_I32, VOP3_PACKED>, int_amdgcn_sdot4>; |
