From 2a3e0d7e76079289e2b007a15c311c51218f0b89 Mon Sep 17 00:00:00 2001 From: Tom Stellard Date: Wed, 23 Jan 2013 21:39:47 +0000 Subject: R600: Simplify stream outputs intrinsic Patch by: Vincent Lejeune Reviewed-by: Tom Stellard git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@173296 91177308-0d34-0410-b5e6-96231b3b80d8 --- lib/Target/R600/R600ISelLowering.cpp | 32 ----------------------------- lib/Target/R600/R600Instructions.td | 24 +++++++++++----------- lib/Target/R600/R600Intrinsics.td | 2 +- lib/Target/R600/R600MachineFunctionInfo.cpp | 1 - lib/Target/R600/R600MachineFunctionInfo.h | 1 - 5 files changed, 13 insertions(+), 47 deletions(-) (limited to 'lib/Target') diff --git a/lib/Target/R600/R600ISelLowering.cpp b/lib/Target/R600/R600ISelLowering.cpp index 69ca3f5..3434d7e 100644 --- a/lib/Target/R600/R600ISelLowering.cpp +++ b/lib/Target/R600/R600ISelLowering.cpp @@ -386,39 +386,7 @@ SDValue R600TargetLowering::LowerOperation(SDValue Op, SelectionDAG &DAG) const Chain); } - case AMDGPUIntrinsic::R600_store_stream_output : { - MachineFunction &MF = DAG.getMachineFunction(); - R600MachineFunctionInfo *MFI = MF.getInfo(); - int64_t RegIndex = cast(Op.getOperand(3))->getZExtValue(); - int64_t BufIndex = cast(Op.getOperand(4))->getZExtValue(); - - SDNode **OutputsMap = MFI->StreamOutputs[BufIndex]; - unsigned Inst; - switch (cast(Op.getOperand(4))->getZExtValue() ) { - // STREAM3 - case 3: - Inst = 4; - break; - // STREAM2 - case 2: - Inst = 3; - break; - // STREAM1 - case 1: - Inst = 2; - break; - // STREAM0 - case 0: - Inst = 1; - break; - default: - llvm_unreachable("Wrong buffer id for stream outputs !"); - } - return InsertScalarToRegisterExport(DAG, Op.getDebugLoc(), OutputsMap, - RegIndex / 4, RegIndex % 4, Inst, 0, Op.getOperand(2), - Chain); - } // default for switch(IntrinsicID) default: break; } diff --git a/lib/Target/R600/R600Instructions.td b/lib/Target/R600/R600Instructions.td index a6c3910..3e069da 100644 --- a/lib/Target/R600/R600Instructions.td +++ b/lib/Target/R600/R600Instructions.td @@ -604,24 +604,24 @@ multiclass ExportPattern cf_inst> { multiclass SteamOutputExportPattern buf0inst, bits<8> buf1inst, bits<8> buf2inst, bits<8> buf3inst> { // Stream0 - def : Pat<(EXPORT (v4f32 R600_Reg128:$src), (i32 1), - (i32 imm:$type), (i32 imm:$arraybase), (i32 imm:$mask)), - (ExportInst R600_Reg128:$src, imm:$type, imm:$arraybase, + def : Pat<(int_R600_store_stream_output (v4f32 R600_Reg128:$src), + (i32 imm:$arraybase), (i32 0), (i32 imm:$mask)), + (ExportInst R600_Reg128:$src, 0, imm:$arraybase, 4095, imm:$mask, buf0inst, 0)>; // Stream1 - def : Pat<(EXPORT (v4f32 R600_Reg128:$src), (i32 2), - (i32 imm:$type), (i32 imm:$arraybase), (i32 imm:$mask)), - (ExportInst R600_Reg128:$src, imm:$type, imm:$arraybase, + def : Pat<(int_R600_store_stream_output (v4f32 R600_Reg128:$src), + (i32 imm:$arraybase), (i32 1), (i32 imm:$mask)), + (ExportInst R600_Reg128:$src, 0, imm:$arraybase, 4095, imm:$mask, buf1inst, 0)>; // Stream2 - def : Pat<(EXPORT (v4f32 R600_Reg128:$src), (i32 3), - (i32 imm:$type), (i32 imm:$arraybase), (i32 imm:$mask)), - (ExportInst R600_Reg128:$src, imm:$type, imm:$arraybase, + def : Pat<(int_R600_store_stream_output (v4f32 R600_Reg128:$src), + (i32 imm:$arraybase), (i32 2), (i32 imm:$mask)), + (ExportInst R600_Reg128:$src, 0, imm:$arraybase, 4095, imm:$mask, buf2inst, 0)>; // Stream3 - def : Pat<(EXPORT (v4f32 R600_Reg128:$src), (i32 4), - (i32 imm:$type), (i32 imm:$arraybase), (i32 imm:$mask)), - (ExportInst R600_Reg128:$src, imm:$type, imm:$arraybase, + def : Pat<(int_R600_store_stream_output (v4f32 R600_Reg128:$src), + (i32 imm:$arraybase), (i32 3), (i32 imm:$mask)), + (ExportInst R600_Reg128:$src, 0, imm:$arraybase, 4095, imm:$mask, buf3inst, 0)>; } diff --git a/lib/Target/R600/R600Intrinsics.td b/lib/Target/R600/R600Intrinsics.td index 3825bc4..06a7341 100644 --- a/lib/Target/R600/R600Intrinsics.td +++ b/lib/Target/R600/R600Intrinsics.td @@ -20,7 +20,7 @@ let TargetPrefix = "R600", isTarget = 1 in { def int_R600_load_input_linear : Intrinsic<[llvm_float_ty], [llvm_i32_ty], [IntrReadMem]>; def int_R600_store_stream_output : - Intrinsic<[], [llvm_float_ty, llvm_i32_ty, llvm_i32_ty], []>; + Intrinsic<[], [llvm_v4f32_ty, llvm_i32_ty, llvm_i32_ty, llvm_i32_ty], []>; def int_R600_store_pixel_color : Intrinsic<[], [llvm_float_ty, llvm_i32_ty], []>; def int_R600_store_pixel_depth : diff --git a/lib/Target/R600/R600MachineFunctionInfo.cpp b/lib/Target/R600/R600MachineFunctionInfo.cpp index 4eb5efa..bcb7f94 100644 --- a/lib/Target/R600/R600MachineFunctionInfo.cpp +++ b/lib/Target/R600/R600MachineFunctionInfo.cpp @@ -17,7 +17,6 @@ R600MachineFunctionInfo::R600MachineFunctionInfo(const MachineFunction &MF) HasLinearInterpolation(false), HasPerspectiveInterpolation(false) { memset(Outputs, 0, sizeof(Outputs)); - memset(StreamOutputs, 0, sizeof(StreamOutputs)); } unsigned R600MachineFunctionInfo::GetIJPerspectiveIndex() const { diff --git a/lib/Target/R600/R600MachineFunctionInfo.h b/lib/Target/R600/R600MachineFunctionInfo.h index e97fb5b..91f9de2 100644 --- a/lib/Target/R600/R600MachineFunctionInfo.h +++ b/lib/Target/R600/R600MachineFunctionInfo.h @@ -25,7 +25,6 @@ public: R600MachineFunctionInfo(const MachineFunction &MF); std::vector ReservedRegs; SDNode *Outputs[16]; - SDNode *StreamOutputs[64][4]; bool HasLinearInterpolation; bool HasPerspectiveInterpolation; -- cgit v1.1