From 1f10c5ea943b0f13df9d61ca22dc964760ecc45f Mon Sep 17 00:00:00 2001 From: "Michael J. Spencer" Date: Thu, 1 May 2014 22:12:39 +0000 Subject: [IR] Make {extract,insert}element accept an index of any integer type. Given the following C code llvm currently generates suboptimal code for x86-64: __m128 bss4( const __m128 *ptr, size_t i, size_t j ) { float f = ptr[i][j]; return (__m128) { f, f, f, f }; } ================================================= define <4 x float> @_Z4bss4PKDv4_fmm(<4 x float>* nocapture readonly %ptr, i64 %i, i64 %j) #0 { %a1 = getelementptr inbounds <4 x float>* %ptr, i64 %i %a2 = load <4 x float>* %a1, align 16, !tbaa !1 %a3 = trunc i64 %j to i32 %a4 = extractelement <4 x float> %a2, i32 %a3 %a5 = insertelement <4 x float> undef, float %a4, i32 0 %a6 = insertelement <4 x float> %a5, float %a4, i32 1 %a7 = insertelement <4 x float> %a6, float %a4, i32 2 %a8 = insertelement <4 x float> %a7, float %a4, i32 3 ret <4 x float> %a8 } ================================================= shlq $4, %rsi addq %rdi, %rsi movslq %edx, %rax vbroadcastss (%rsi,%rax,4), %xmm0 retq ================================================= The movslq is uneeded, but is present because of the trunc to i32 and then sext back to i64 that the backend adds for vbroadcastss. We can't remove it because it changes the meaning. The IR that clang generates is already suboptimal. What clang really should emit is: %a4 = extractelement <4 x float> %a2, i64 %j This patch makes that legal. A separate patch will teach clang to do it. Differential Revision: http://reviews.llvm.org/D3519 llvm-svn: 207801 --- llvm/lib/Bitcode/Writer/BitcodeWriter.cpp | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-) (limited to 'llvm/lib/Bitcode/Writer/BitcodeWriter.cpp') diff --git a/llvm/lib/Bitcode/Writer/BitcodeWriter.cpp b/llvm/lib/Bitcode/Writer/BitcodeWriter.cpp index 92965fa..2337487 100644 --- a/llvm/lib/Bitcode/Writer/BitcodeWriter.cpp +++ b/llvm/lib/Bitcode/Writer/BitcodeWriter.cpp @@ -1087,12 +1087,14 @@ static void WriteConstants(unsigned FirstVal, unsigned LastVal, Code = bitc::CST_CODE_CE_EXTRACTELT; Record.push_back(VE.getTypeID(C->getOperand(0)->getType())); Record.push_back(VE.getValueID(C->getOperand(0))); + Record.push_back(VE.getTypeID(C->getOperand(1)->getType())); Record.push_back(VE.getValueID(C->getOperand(1))); break; case Instruction::InsertElement: Code = bitc::CST_CODE_CE_INSERTELT; Record.push_back(VE.getValueID(C->getOperand(0))); Record.push_back(VE.getValueID(C->getOperand(1))); + Record.push_back(VE.getTypeID(C->getOperand(2)->getType())); Record.push_back(VE.getValueID(C->getOperand(2))); break; case Instruction::ShuffleVector: @@ -1253,13 +1255,13 @@ static void WriteInstruction(const Instruction &I, unsigned InstID, case Instruction::ExtractElement: Code = bitc::FUNC_CODE_INST_EXTRACTELT; PushValueAndType(I.getOperand(0), InstID, Vals, VE); - pushValue(I.getOperand(1), InstID, Vals, VE); + PushValueAndType(I.getOperand(1), InstID, Vals, VE); break; case Instruction::InsertElement: Code = bitc::FUNC_CODE_INST_INSERTELT; PushValueAndType(I.getOperand(0), InstID, Vals, VE); pushValue(I.getOperand(1), InstID, Vals, VE); - pushValue(I.getOperand(2), InstID, Vals, VE); + PushValueAndType(I.getOperand(2), InstID, Vals, VE); break; case Instruction::ShuffleVector: Code = bitc::FUNC_CODE_INST_SHUFFLEVEC; -- cgit v1.1