[AMDGPU] Fix subtarget predicates for some V_MAD, V_FMA and V_DOT instructions.

[kosarev]

Resolves AsmParser ambiguities, e.g., between V_FMA_MIXLO_F16_vi and V_FMA_MIXLO_F16_gfx11.

Part of #69256.

[llvmbot]

@llvm/pr-subscribers-backend-amdgpu

Author: Ivan Kosarev (kosarev)

Changes

Resolves AsmParser ambiguities, e.g., between V_FMA_MIXLO_F16_vi and V_FMA_MIXLO_F16_gfx11.

Part of <#69256>.

---

Full diff: https://github.com/llvm/llvm-project/pull/71194.diff

1 Files Affected:

• (modified) llvm/lib/Target/AMDGPU/VOP3PInstructions.td (+11-18)

diff --git a/llvm/lib/Target/AMDGPU/VOP3PInstructions.td b/llvm/lib/Target/AMDGPU/VOP3PInstructions.td
index 24d7550e2dec496..a6dfac3bc26dc49 100644
--- a/llvm/lib/Target/AMDGPU/VOP3PInstructions.td
+++ b/llvm/lib/Target/AMDGPU/VOP3PInstructions.td
@@ -373,36 +373,31 @@ class SDot2Pat<Instruction Inst> : GCNPat <
 }
 
 let IsDOT = 1 in {
-let SubtargetPredicate = HasDot2Insts in {
-
+let OtherPredicates = [HasDot2Insts] in {
 defm V_DOT2_I32_I16 : VOP3PInst<"v_dot2_i32_i16",
   VOP3P_Profile<VOP_I32_V2I16_V2I16_I32>, int_amdgcn_sdot2, 1>;
 defm V_DOT2_U32_U16 : VOP3PInst<"v_dot2_u32_u16",
   VOP3P_Profile<VOP_I32_V2I16_V2I16_I32>, int_amdgcn_udot2, 1>;
+} // End OtherPredicates = [HasDot2Insts]
 
-} // End SubtargetPredicate = HasDot2Insts
-
-let SubtargetPredicate = HasDot10Insts in
+let OtherPredicates = [HasDot10Insts] in
 defm V_DOT2_F32_F16 : VOP3PInst<"v_dot2_f32_f16",
   VOP3P_Profile<VOP_F32_V2F16_V2F16_F32, VOP3_REGULAR, /*HasDPP*/ 1>,
   AMDGPUfdot2, 1/*ExplicitClamp*/>;
 
-let SubtargetPredicate = HasDot7Insts in {
+let OtherPredicates = [HasDot7Insts] in {
 defm V_DOT4_U32_U8  : VOP3PInst<"v_dot4_u32_u8",
   VOP3P_Profile<VOP_I32_I32_I32_I32, VOP3_PACKED>, int_amdgcn_udot4, 1>;
 defm V_DOT8_U32_U4  : VOP3PInst<"v_dot8_u32_u4",
   VOP3P_Profile<VOP_I32_I32_I32_I32, VOP3_PACKED>, int_amdgcn_udot8, 1>;
+} // End OtherPredicates = [HasDot7Insts]
 
-} // End SubtargetPredicate = HasDot7Insts
-
-let SubtargetPredicate = HasDot1Insts in {
-
+let OtherPredicates = [HasDot1Insts] in {
 defm V_DOT4_I32_I8  : VOP3PInst<"v_dot4_i32_i8",
   VOP3P_Profile<VOP_I32_I32_I32_I32, VOP3_PACKED>, int_amdgcn_sdot4, 1>;
 defm V_DOT8_I32_I4  : VOP3PInst<"v_dot8_i32_i4",
   VOP3P_Profile<VOP_I32_I32_I32_I32, VOP3_PACKED>, int_amdgcn_sdot8, 1>;
-
-} // End SubtargetPredicate = HasDot1Insts
+} // End OtherPredicates = [HasDot1Insts]
 
 def DOT2_BF16_Profile
   : VOP3P_Profile<VOP_F32_V2I16_V2I16_F32, VOP3_REGULAR, /*HasDPP*/ 1> {
@@ -1128,16 +1123,15 @@ defm V_PK_ADD_F16 : VOP3P_Real_vi <0x0f>;
 defm V_PK_MUL_F16 : VOP3P_Real_vi <0x10>;
 defm V_PK_MIN_F16 : VOP3P_Real_vi <0x11>;
 defm V_PK_MAX_F16 : VOP3P_Real_vi <0x12>;
-} // End SubtargetPredicate = isGFX8GFX9
 
-let SubtargetPredicate = HasMadMixInsts in {
+let OtherPredicates = [HasMadMixInsts] in {
 defm V_MAD_MIX_F32 : VOP3P_Real_vi <0x20>;
 defm V_MAD_MIXLO_F16 : VOP3P_Real_vi <0x21>;
 defm V_MAD_MIXHI_F16 : VOP3P_Real_vi <0x22>;
 }
 
-let SubtargetPredicate = HasFmaMixInsts in {
-let DecoderNamespace = "GFX9_DL" in {
+let OtherPredicates = [HasFmaMixInsts],
+    DecoderNamespace = "GFX9_DL" in {
 // The mad_mix instructions were renamed and their behaviors changed,
 // but the opcode stayed the same so we need to put these in a
 // different DecoderNamespace to avoid the ambiguity.
@@ -1145,8 +1139,6 @@ defm V_FMA_MIX_F32 : VOP3P_Real_vi <0x20>;
 defm V_FMA_MIXLO_F16 : VOP3P_Real_vi <0x21>;
 defm V_FMA_MIXHI_F16 : VOP3P_Real_vi <0x22>;
 }
-}
-
 
 defm V_DOT2_I32_I16 : VOP3P_Real_vi <0x26>;
 defm V_DOT2_U32_U16 : VOP3P_Real_vi <0x27>;
@@ -1157,6 +1149,7 @@ defm V_DOT8_U32_U4  : VOP3P_Real_vi <0x2b>;
 
 defm V_DOT4_I32_I8  : VOP3P_Real_vi <0x28>;
 defm V_DOT8_I32_I4  : VOP3P_Real_vi <0x2a>;
+} // End SubtargetPredicate = isGFX8GFX9
 
 let OtherPredicates = [HasMAIInsts] in {
 

[jayfoad]

LGTM.

How do you find these problems? Once they are all fixed can you add some kind of assertion somewhere to stop them reoccuring?

[Sisyph]

LGTM

[kosarev]

How do you find these problems? Once they are all fixed can you add some kind of assertion somewhere to stop them reoccuring?

I use the ambiguity checks in AsmMatcherEmitter, but it also takes making local changes manually teaching it about which of our predicates are mutually exclusive. I do hope to be able to make it catching that sort of problems without manual help at some point later (as part of #69256), but it currently feels like a long way to go.

[kosarev]

Fixed pattern predicates as well to pass codegen tests (must have missed the failures). Please take another look.

[jayfoad]

LGTM.

[jayfoad]

As a follow up maybe change the template argument to VOP_Pseudo Inst to avoid the cast?

[kosarev]

#71802