shader_recompiler: Implement most integer image atomics, workgroup barriers and shared memory load/store (#231)

* shader_recompiler: Add LDEXP

* shader_recompiler: Add most image integer atomic ops

* shader_recompiler: Implement shared memory load/store

* shader_recompiler: More image atomics

* externals: Update sirit

* clang format

* cmake: Add missing files

* shader_recompiler: Fix some atomic bugs

* shader_recompiler: Vs outputs

* shader_recompiler: Shared mem has side-effects, fix format component order

* shader_recompiler: Inline constant buffer impl

* video_core: Fix regressions

* Work

* Fixup a few things
This commit is contained in:
TheTurtle 2024-07-05 00:15:44 +03:00 committed by GitHub
parent af3bbc33e9
commit 6ceab6dfac
No known key found for this signature in database
GPG key ID: B5690EEEBB952194
69 changed files with 1597 additions and 310 deletions

View file

@ -33,16 +33,21 @@ void EmitDeviceMemoryBarrier(EmitContext& ctx);
void EmitGetScc(EmitContext& ctx);
void EmitGetExec(EmitContext& ctx);
void EmitGetVcc(EmitContext& ctx);
void EmitGetSccLo(EmitContext& ctx);
void EmitGetVccLo(EmitContext& ctx);
void EmitGetVccHi(EmitContext& ctx);
void EmitSetScc(EmitContext& ctx);
void EmitSetExec(EmitContext& ctx);
void EmitSetVcc(EmitContext& ctx);
void EmitSetSccLo(EmitContext& ctx);
void EmitSetVccLo(EmitContext& ctx);
void EmitSetVccHi(EmitContext& ctx);
void EmitPrologue(EmitContext& ctx);
void EmitEpilogue(EmitContext& ctx);
void EmitDiscard(EmitContext& ctx);
void EmitBarrier(EmitContext& ctx);
void EmitWorkgroupMemoryBarrier(EmitContext& ctx);
void EmitDeviceMemoryBarrier(EmitContext& ctx);
Id EmitGetUserData(EmitContext& ctx, IR::ScalarReg reg);
void EmitGetThreadBitScalarReg(EmitContext& ctx);
void EmitSetThreadBitScalarReg(EmitContext& ctx);
@ -82,12 +87,13 @@ Id EmitUndefU8(EmitContext& ctx);
Id EmitUndefU16(EmitContext& ctx);
Id EmitUndefU32(EmitContext& ctx);
Id EmitUndefU64(EmitContext& ctx);
Id EmitReadSharedU8(EmitContext& ctx, Id offset);
Id EmitReadSharedS8(EmitContext& ctx, Id offset);
Id EmitReadSharedU16(EmitContext& ctx, Id offset);
Id EmitReadSharedS16(EmitContext& ctx, Id offset);
Id EmitReadSharedU32(EmitContext& ctx, Id offset);
Id EmitReadSharedU64(EmitContext& ctx, Id offset);
Id EmitLoadSharedU8(EmitContext& ctx, Id offset);
Id EmitLoadSharedS8(EmitContext& ctx, Id offset);
Id EmitLoadSharedU16(EmitContext& ctx, Id offset);
Id EmitLoadSharedS16(EmitContext& ctx, Id offset);
Id EmitLoadSharedU32(EmitContext& ctx, Id offset);
Id EmitLoadSharedU64(EmitContext& ctx, Id offset);
Id EmitLoadSharedU128(EmitContext& ctx, Id offset);
void EmitWriteSharedU8(EmitContext& ctx, Id offset, Id value);
void EmitWriteSharedU16(EmitContext& ctx, Id offset, Id value);
void EmitWriteSharedU32(EmitContext& ctx, Id offset, Id value);
@ -140,7 +146,7 @@ Id EmitSelectF64(EmitContext& ctx, Id cond, Id true_value, Id false_value);
void EmitBitCastU16F16(EmitContext& ctx);
Id EmitBitCastU32F32(EmitContext& ctx, Id value);
void EmitBitCastU64F64(EmitContext& ctx);
void EmitBitCastF16U16(EmitContext&);
Id EmitBitCastF16U16(EmitContext& ctx, Id value);
Id EmitBitCastF32U32(EmitContext& ctx, Id value);
void EmitBitCastF64U64(EmitContext& ctx);
Id EmitPackUint2x32(EmitContext& ctx, Id value);
@ -172,6 +178,7 @@ Id EmitFPNeg64(EmitContext& ctx, Id value);
Id EmitFPSin(EmitContext& ctx, Id value);
Id EmitFPCos(EmitContext& ctx, Id value);
Id EmitFPExp2(EmitContext& ctx, Id value);
Id EmitFPLdexp(EmitContext& ctx, Id value, Id exp);
Id EmitFPLog2(EmitContext& ctx, Id value);
Id EmitFPRecip32(EmitContext& ctx, Id value);
Id EmitFPRecip64(EmitContext& ctx, Id value);
@ -236,8 +243,11 @@ Id EmitFPUnordGreaterThanEqual64(EmitContext& ctx, Id lhs, Id rhs);
Id EmitFPIsNan16(EmitContext& ctx, Id value);
Id EmitFPIsNan32(EmitContext& ctx, Id value);
Id EmitFPIsNan64(EmitContext& ctx, Id value);
Id EmitFPIsInf32(EmitContext& ctx, Id value);
Id EmitFPIsInf64(EmitContext& ctx, Id value);
Id EmitIAdd32(EmitContext& ctx, IR::Inst* inst, Id a, Id b);
Id EmitIAdd64(EmitContext& ctx, Id a, Id b);
Id EmitIAddCary32(EmitContext& ctx, Id a, Id b);
Id EmitISub32(EmitContext& ctx, Id a, Id b);
Id EmitISub64(EmitContext& ctx, Id a, Id b);
Id EmitSMulExt(EmitContext& ctx, Id a, Id b);
@ -333,6 +343,7 @@ Id EmitConvertF64U8(EmitContext& ctx, Id value);
Id EmitConvertF64U16(EmitContext& ctx, Id value);
Id EmitConvertF64U32(EmitContext& ctx, Id value);
Id EmitConvertF64U64(EmitContext& ctx, Id value);
Id EmitConvertU16U32(EmitContext& ctx, Id value);
Id EmitImageSampleImplicitLod(EmitContext& ctx, IR::Inst* inst, u32 handle, Id coords, Id bias_lc,
Id offset);
@ -355,6 +366,18 @@ Id EmitImageGradient(EmitContext& ctx, IR::Inst* inst, const IR::Value& index, I
Id EmitImageRead(EmitContext& ctx, IR::Inst* inst, const IR::Value& index, Id coords);
void EmitImageWrite(EmitContext& ctx, IR::Inst* inst, u32 handle, Id coords, Id color);
Id EmitImageAtomicIAdd32(EmitContext& ctx, IR::Inst* inst, u32 handle, Id coords, Id value);
Id EmitImageAtomicSMin32(EmitContext& ctx, IR::Inst* inst, u32 handle, Id coords, Id value);
Id EmitImageAtomicUMin32(EmitContext& ctx, IR::Inst* inst, u32 handle, Id coords, Id value);
Id EmitImageAtomicSMax32(EmitContext& ctx, IR::Inst* inst, u32 handle, Id coords, Id value);
Id EmitImageAtomicUMax32(EmitContext& ctx, IR::Inst* inst, u32 handle, Id coords, Id value);
Id EmitImageAtomicInc32(EmitContext& ctx, IR::Inst* inst, u32 handle, Id coords, Id value);
Id EmitImageAtomicDec32(EmitContext& ctx, IR::Inst* inst, u32 handle, Id coords, Id value);
Id EmitImageAtomicAnd32(EmitContext& ctx, IR::Inst* inst, u32 handle, Id coords, Id value);
Id EmitImageAtomicOr32(EmitContext& ctx, IR::Inst* inst, u32 handle, Id coords, Id value);
Id EmitImageAtomicXor32(EmitContext& ctx, IR::Inst* inst, u32 handle, Id coords, Id value);
Id EmitImageAtomicExchange32(EmitContext& ctx, IR::Inst* inst, u32 handle, Id coords, Id value);
Id EmitLaneId(EmitContext& ctx);
Id EmitQuadShuffle(EmitContext& ctx, Id value, Id index);