Skip to content

AMDGPU: Clean up more real instruction predicate overrides #116868

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
merged 1 commit into from
Nov 19, 2024
Merged

AMDGPU: Clean up more real instruction predicate overrides #116868

Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
42 changes: 16 additions & 26 deletions llvm/lib/Target/AMDGPU/VOP3PInstructions.td
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -669,8 +669,6 @@ class VgprMAIFrag<SDPatternOperator Op> : MAIFrag<Op, MayNotNeedAGPRs> {
let GISelPredicateCode = MayNotNeedAGPRs_gisel;
}

let SubtargetPredicate = HasMAIInsts in {

let isAsCheapAsAMove = 1, isReMaterializable = 1 in {
defm V_ACCVGPR_READ_B32 : VOP3Inst<"v_accvgpr_read_b32", VOPProfileAccRead>;
let isMoveImm = 1 in {
Expand All @@ -680,6 +678,7 @@ let isAsCheapAsAMove = 1, isReMaterializable = 1 in {

class MAIInst<string OpName, VOPProfile P, SDPatternOperator node>
: VOP3InstBase<OpName, P, node> {
let SubtargetPredicate = HasMAIInsts;
Instruction Opcode = !cast<Instruction>(NAME);
bit is_dgemm = 0;
bit is_gfx940_xdl = 0;
Expand All @@ -695,7 +694,7 @@ multiclass MAIInst<string OpName, string P, SDPatternOperator node> {
!if(!or(NoDstOverlap, !eq(node, null_frag)), null_frag, AgprMAIFrag<node>)>,
MFMATable<0, NAME # "_e64">;

let SubtargetPredicate = isGFX90APlus, Mnemonic = OpName in
let OtherPredicates = [isGFX90APlus], Mnemonic = OpName in
def _vgprcd_e64 : MAIInst<OpName # "_vgprcd", !cast<VOPProfileMAI>("VOPProfileMAI_" # P # "_VCD"),
!if(!or(NoDstOverlap, !eq(node, null_frag)), null_frag, VgprMAIFrag<node>)>,
MFMATable<0, NAME # "_vgprcd_e64">;
Expand All @@ -709,7 +708,7 @@ multiclass MAIInst<string OpName, string P, SDPatternOperator node> {
!if(!eq(node, null_frag), null_frag, AgprMAIFrag<node>)>,
MFMATable<1, NAME # "_e64">;

let SubtargetPredicate = isGFX90APlus in
let OtherPredicates = [isGFX90APlus] in
Copy link
Contributor

@Sisyph Sisyph Nov 19, 2024

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Looks like the outer "let SubtargetPredicate = HasMAIInsts in {" would have dominated this, so having isGFX90APlus is an additional restriction now. Is that correct? If so LGTM

Copy link
Contributor Author

@arsenm arsenm Nov 19, 2024

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

The VGPR encoded version was only added in 90a, this was always the restriction. The inner let is still a bit of a hack, but it's at least more local now

def _mac_vgprcd_e64 : MAIInst<OpName # "_mac_vgprcd", !cast<VOPProfileMAI>("VOPProfileMAI_" # P # "_VCD"),
!if(!eq(node, null_frag), null_frag, VgprMAIFrag<node>)>,
MFMATable<1, NAME # "_vgprcd_e64">;
Expand All @@ -735,7 +734,7 @@ defm V_MFMA_F32_32X32X8F16 : MAIInst<"v_mfma_f32_32x32x8f16", "F32_V4F16_X16",
defm V_MFMA_I32_32X32X4I8 : MAIInst<"v_mfma_i32_32x32x4i8", "I32_I32_X32", int_amdgcn_mfma_i32_32x32x4i8>;
}

let Predicates = [isGFX908orGFX90A] in {
let SubtargetPredicate = isGFX908orGFX90A in {
defm V_MFMA_I32_16X16X16I8 : MAIInst<"v_mfma_i32_16x16x16i8", "I32_I32_X4", int_amdgcn_mfma_i32_16x16x16i8>;
defm V_MFMA_I32_32X32X8I8 : MAIInst<"v_mfma_i32_32x32x8i8", "I32_I32_X16", int_amdgcn_mfma_i32_32x32x8i8>;
defm V_MFMA_F32_4X4X2BF16 : MAIInst<"v_mfma_f32_4x4x2bf16", "F32_V2I16_X4", int_amdgcn_mfma_f32_4x4x2bf16>;
Expand All @@ -745,15 +744,13 @@ defm V_MFMA_F32_32X32X2BF16 : MAIInst<"v_mfma_f32_32x32x2bf16", "F32_V2I16_X32",
defm V_MFMA_F32_32X32X4BF16 : MAIInst<"v_mfma_f32_32x32x4bf16", "F32_V2I16_X16", int_amdgcn_mfma_f32_32x32x4bf16>;
}

} // End SubtargetPredicate = HasMAIInsts

let SubtargetPredicate = HasGFX950Insts, is_gfx940_xdl = 1 in {
defm V_MFMA_F32_16X16X32_F16 : MAIInst<"v_mfma_f32_16x16x32f16", "F32_V8F16_X32", int_amdgcn_mfma_f32_16x16x32_f16>;
defm V_MFMA_F32_32X32X16_F16 : MAIInst<"v_mfma_f32_32x32x16f16", "F32_V8F16_X16", int_amdgcn_mfma_f32_32x32x16_f16>;
defm V_MFMA_F32_32X32X16_BF16 : MAIInst<"v_mfma_f32_32x32x16bf16", "F32_V8BF16_X16", int_amdgcn_mfma_f32_32x32x16_bf16>;
}

let Predicates = [isGFX90APlus] in {
let SubtargetPredicate = isGFX90APlus in {
let is_gfx940_xdl = 1 in {
defm V_MFMA_F32_32X32X4BF16_1K : MAIInst<"v_mfma_f32_32x32x4bf16_1k", "F32_V4I16_X32", int_amdgcn_mfma_f32_32x32x4bf16_1k>;
defm V_MFMA_F32_16X16X4BF16_1K : MAIInst<"v_mfma_f32_16x16x4bf16_1k", "F32_V4I16_X16", int_amdgcn_mfma_f32_16x16x4bf16_1k>;
Expand All @@ -766,7 +763,7 @@ let Predicates = [isGFX90APlus] in {
defm V_MFMA_F64_16X16X4F64 : MAIInst<"v_mfma_f64_16x16x4f64", "F64_16X16X4F64", int_amdgcn_mfma_f64_16x16x4f64>;
defm V_MFMA_F64_4X4X4F64 : MAIInst<"v_mfma_f64_4x4x4f64", "F64_4X4X4F64", int_amdgcn_mfma_f64_4x4x4f64>;
}
} // End Predicates = [isGFX90APlus]
} // End SubtargetPredicate = isGFX90APlus

let SubtargetPredicate = isGFX940Plus, is_gfx940_xdl = 1 in {
defm V_MFMA_I32_32X32X16I8 : MAIInst<"v_mfma_i32_32x32x16i8", "I32_I64_X32", int_amdgcn_mfma_i32_32x32x16_i8>;
Expand Down Expand Up @@ -1632,14 +1629,17 @@ multiclass VOP3P_Real_MFMA_gfx940_aliases<string NameFrom, string NameTo, string
VOPProfile Pfl_ACD = PS_ACD.Pfl,
VOPProfile Pfl_VCD = PS_VCD.Pfl> {
if !ne(NameFrom, NameTo) then {
def : InstAlias <NameTo # " " # PS_ACD.AsmOperands,
let SubtargetPredicate = PS_ACD.SubtargetPredicate,
OtherPredicates = PS_ACD.OtherPredicates in {
def : InstAlias <NameTo # " " # PS_ACD.AsmOperands,
(!cast<VOP3P_Real>(Op # "_gfx940_acd") Pfl_ACD.DstRC:$vdst,
Pfl_ACD.Src0RC64:$src0, Pfl_ACD.Src1RC64:$src1, Pfl_ACD.Src2RC64:$src2,
CBSZ:$cbsz, ABID:$abid, blgp:$blgp)>, PredicateControl;
def : InstAlias <NameTo # " " # PS_VCD.AsmOperands,
def : InstAlias <NameTo # " " # PS_VCD.AsmOperands,
(!cast<VOP3P_Real>(Op # "_gfx940_vcd") Pfl_VCD.DstRC:$vdst,
Pfl_VCD.Src0RC64:$src0, Pfl_VCD.Src1RC64:$src1, Pfl_VCD.Src2RC64:$src2,
CBSZ:$cbsz, ABID:$abid, blgp:$blgp)>, PredicateControl;
}
}
}

Expand All @@ -1656,7 +1656,10 @@ multiclass VOP3P_Real_MFMA_gfx940<bits<7> op, string Name = !cast<VOP3_Pseudo>(N
VOP3Pe_MAI <op, PS_VCD.Pfl, 0>;
} // End AssemblerPredicate = isGFX940Plus, DecoderNamespace = "GFX940"

let SubtargetPredicate = isGFX940Plus in {
let SubtargetPredicate = PS_ACD.SubtargetPredicate,
OtherPredicates = PS_ACD.OtherPredicates,
AssemblerPredicate = isGFX940Plus
in {
defm : VOP3P_Real_MFMA_gfx940_aliases<Name, PS_ACD.Mnemonic, NAME>;

if !ne(!subst("_1k", "", PS_ACD.Mnemonic), PS_ACD.Mnemonic) then
Expand Down Expand Up @@ -1703,7 +1706,6 @@ multiclass VOP3P_Real_SMFMAC<bits<7> op, string alias> {
}
}

let SubtargetPredicate = isGFX8GFX9 in {
defm V_PK_MAD_I16 : VOP3P_Real_vi <0x00>;
defm V_PK_MUL_LO_U16 : VOP3P_Real_vi <0x01>;
defm V_PK_ADD_I16 : VOP3P_Real_vi <0x02>;
Expand All @@ -1725,11 +1727,9 @@ 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>;

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 OtherPredicates = [HasFmaMixInsts],
DecoderNamespace = "GFX9_DL" in {
Expand All @@ -1750,9 +1750,6 @@ 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 {

defm V_ACCVGPR_READ_B32 : VOP3P_Real_MAI <0x58>;
defm V_ACCVGPR_WRITE_B32 : VOP3P_Real_MAI <0x59>;
Expand All @@ -1778,8 +1775,6 @@ defm V_MFMA_F32_4X4X2BF16 : VOP3P_Real_MFMA_vi_gfx90a <0x6b>;
defm V_MFMA_F32_32X32X4BF16 : VOP3P_Real_MFMA_vi_gfx90a <0x6c>;
defm V_MFMA_F32_16X16X8BF16 : VOP3P_Real_MFMA_vi_gfx90a <0x6d>;

} // End OtherPredicates = [HasMAIInsts]

defm V_MFMA_F32_32X32X4BF16_1K : VOP3P_Real_MFMA_gfx90a <0x63>;
defm V_MFMA_F32_16X16X4BF16_1K : VOP3P_Real_MFMA_gfx90a <0x64>;
defm V_MFMA_F32_4X4X4BF16_1K : VOP3P_Real_MFMA_gfx90a <0x65>;
Expand All @@ -1794,11 +1789,9 @@ defm V_MFMA_F32_32X32X16_BF16 : VOP3P_Real_MFMA_gfx950 <0x37, "v_mfma_f32_32x

defm V_MFMA_I32_32X32X16I8 : VOP3P_Real_MFMA_gfx940 <0x56, "v_mfma_i32_32x32x16_i8">;
defm V_MFMA_I32_16X16X32I8 : VOP3P_Real_MFMA_gfx940 <0x57, "v_mfma_i32_16x16x32_i8">;
let SubtargetPredicate = HasXF32Insts in {
defm V_MFMA_F32_16X16X8XF32 : VOP3P_Real_MFMA_gfx940 <0x3e, "v_mfma_f32_16x16x8_xf32">;
defm V_MFMA_F32_32X32X4XF32 : VOP3P_Real_MFMA_gfx940 <0x3f, "v_mfma_f32_32x32x4_xf32">;
} // End SubtargetPredicate = HasXF32Insts
let SubtargetPredicate = HasFP8Insts in {

defm V_MFMA_F32_16X16X32_BF8_BF8 : VOP3P_Real_MFMA_gfx940 <0x70>;
defm V_MFMA_F32_16X16X32_BF8_FP8 : VOP3P_Real_MFMA_gfx940 <0x71>;
defm V_MFMA_F32_16X16X32_FP8_BF8 : VOP3P_Real_MFMA_gfx940 <0x72>;
Expand All @@ -1807,7 +1800,6 @@ defm V_MFMA_F32_32X32X16_BF8_BF8 : VOP3P_Real_MFMA_gfx940 <0x74>;
defm V_MFMA_F32_32X32X16_BF8_FP8 : VOP3P_Real_MFMA_gfx940 <0x75>;
defm V_MFMA_F32_32X32X16_FP8_BF8 : VOP3P_Real_MFMA_gfx940 <0x76>;
defm V_MFMA_F32_32X32X16_FP8_FP8 : VOP3P_Real_MFMA_gfx940 <0x77>;
} // End SubtargetPredicate = HasFP8Insts

defm V_MFMA_F32_32X32X4BF16_1K : VOP3P_Real_MFMA_gfx940 <0x5d, "v_mfma_f32_32x32x4_2b_bf16">;
defm V_MFMA_F32_16X16X4BF16_1K : VOP3P_Real_MFMA_gfx940 <0x5e, "v_mfma_f32_16x16x4_4b_bf16">;
Expand All @@ -1824,7 +1816,6 @@ defm V_SMFMAC_F32_16X16X32_BF16 : VOP3P_Real_SMFMAC <0x66, "v_smfmac_f32_16x1
defm V_SMFMAC_F32_32X32X16_BF16 : VOP3P_Real_SMFMAC <0x68, "v_smfmac_f32_32x32x16bf16">;
defm V_SMFMAC_I32_16X16X64_I8 : VOP3P_Real_SMFMAC <0x6a, "v_smfmac_i32_16x16x64i8">;
defm V_SMFMAC_I32_32X32X32_I8 : VOP3P_Real_SMFMAC <0x6c, "v_smfmac_i32_32x32x32i8">;
let SubtargetPredicate = HasFP8Insts in {
defm V_SMFMAC_F32_16X16X64_BF8_BF8 : VOP3P_Real_SMFMAC <0x78, "v_smfmac_f32_16x16x64bf8bf8">;
defm V_SMFMAC_F32_16X16X64_BF8_FP8 : VOP3P_Real_SMFMAC <0x79, "v_smfmac_f32_16x16x64bf8fp8">;
defm V_SMFMAC_F32_16X16X64_FP8_BF8 : VOP3P_Real_SMFMAC <0x7a, "v_smfmac_f32_16x16x64fp8bf8">;
Expand All @@ -1833,7 +1824,6 @@ defm V_SMFMAC_F32_32X32X32_BF8_BF8 : VOP3P_Real_SMFMAC <0x7c, "v_smfmac_f32_32x3
defm V_SMFMAC_F32_32X32X32_BF8_FP8 : VOP3P_Real_SMFMAC <0x7d, "v_smfmac_f32_32x32x32bf8fp8">;
defm V_SMFMAC_F32_32X32X32_FP8_BF8 : VOP3P_Real_SMFMAC <0x7e, "v_smfmac_f32_32x32x32fp8bf8">;
defm V_SMFMAC_F32_32X32X32_FP8_FP8 : VOP3P_Real_SMFMAC <0x7f, "v_smfmac_f32_32x32x32fp8fp8">;
} // End SubtargetPredicate = HasFP8Insts

defm V_PK_FMA_F32 : VOP3P_Real_vi <0x30>;
defm V_PK_MUL_F32 : VOP3P_Real_vi <0x31>;
Expand Down
Loading