- fe9388c Finish making AliasAnalysis aware of the fact that most atomic intrinsics only dereference their arguments, and enhance by Owen Anderson · 16 years ago
- 7bcaefa Fixes a case where we generate an incorrect mask for pshfhw in the presence by Mon P Wang · 16 years ago
- 556b20a While folding vallue comparison terminators ignore dbg intrinsics. by Devang Patel · 16 years ago
- 65085cf Ignore dbg intrinsics while hoisting common code in the two blocks up into the branch block. by Devang Patel · 16 years ago
- 383d7ed by Devang Patel · 16 years ago
- 1aa7056 teach "convert from scalar" to handle loads of fca's. by Chris Lattner · 16 years ago
- 9b872db make scalar conversion handle stores of first class by Chris Lattner · 16 years ago
- 1a3257b Make SROA produce a vector only when the alloca is actually by Chris Lattner · 16 years ago
- 67e3ba3 this produces an undefined result, just check that the alloca is gone by Chris Lattner · 16 years ago
- 95c5f05 Fix PR3411. When replacing values, nodes are analyzed by Duncan Sands · 16 years ago
- 388df62 APInt'fy SimplifyDemandedVectorElts so it can analyze vectors with more than 64 elements. by Evan Cheng · 16 years ago
- 55a683d add another case of undefined behavior without crashing, PR3466. by Chris Lattner · 16 years ago
- 2eafb8b Revert r63600. It didn't fix the bug, it just moved it a bit. by Nick Lewycky · 16 years ago
- dac5c4b Update the callgraph when replacing InvokeInst with CallInst when inlining. by Nick Lewycky · 16 years ago
- ba120aa fix a bitcode reader bug where it can't handle extractelement correctly: by Chris Lattner · 16 years ago
- 3d730f7 Teach ConvertUsesToScalar to handle memset, allowing it to handle by Chris Lattner · 16 years ago
- 50bd558 Delete these two tests. They are specific to x86-64, and there's no by Dan Gohman · 16 years ago
- 7809ecd rearrange how SRoA handles promotion of allocas to vectors. by Chris Lattner · 16 years ago
- d91a61a this test produces an undefined value, we don't care by Chris Lattner · 16 years ago
- 4d4c3da It fails on Linux. XFAIL that machine. by Bill Wendling · 16 years ago
- 1fa3b4b This is passing for us. Should it have been reenabled? by Bill Wendling · 16 years ago
- 7a01d0e Add explicit -march=x86 to these tests so that they don't by Dan Gohman · 16 years ago
- 2ee39de Fix another test to not use -mcpu=yonah with 64-bit code. by Dan Gohman · 16 years ago
- e7f5be7 Yonah does not support x86-64. Change the -mcpu value to one that does. by Dan Gohman · 16 years ago
- 6748f04 Run dsymutil on darwin, when it is expected, before running gdb test. by Devang Patel · 16 years ago
- 3f9495a xfail this for now, will fix shortly. by Chris Lattner · 16 years ago
- 006336d update test by Chris Lattner · 16 years ago
- 996d7a9 Fix a bug which caused us to miscompile a couple of Ada by Chris Lattner · 16 years ago
- d5863dd Do not add redundant arguments in a method definition DIE. by Devang Patel · 16 years ago
- d61051f Make this test case smaller. by Devang Patel · 16 years ago
- 781f2fb This passes on x86-32 linux at least. by Duncan Sands · 16 years ago
- 15d2598 Make the XFAIL line actually match x86-32 targets. by Duncan Sands · 16 years ago
- 67ad9db Teach LowerBRCOND to recognize (xor (setcc x), 1). The xor inverts the condition. It's normally transformed by the dag combiner, unless the condition is set by a arithmetic op with overflow. by Evan Cheng · 16 years ago
- 91dad87 Fix PR3372 by Chris Lattner · 16 years ago
- d42bd99 reduce testcase. by Chris Lattner · 16 years ago
- 0c3c021 add 2 more testcases for -mattr=-sse (r63495). by Torok Edwin · 16 years ago
- 3f142c3 Implement -mno-sse: if SSE is disabled on x86-64, don't store XMM on stack for by Torok Edwin · 16 years ago
- b0d5cdd Fix PR3453 and probably a bunch of other potential by Duncan Sands · 16 years ago
- 4333f49 Reinstate this optimization to fold icmp of xor when possible. Don't try to by Nick Lewycky · 16 years ago
- 39c27ed Fix PR3452 (an infinite loop bootstrapping) by disabling the recent by Chris Lattner · 16 years ago
- 92abc62 Fix PR3401: when using large integers, the type by Duncan Sands · 16 years ago
- d1b5e3f now that all the pieces are in place, teach instcombine's by Chris Lattner · 16 years ago
- 1e19d60 make sure to set Changed=true when instcombine hacks on the code, by Chris Lattner · 16 years ago
- 95be699 Used "-enable-unsafe-fp-math" to allow this transformation - (a * b -c) = c - a *b. by Mon P Wang · 16 years ago
- a7b6cff If unsafe FP optimization is not set, don't allow -(A-B) => B-A because by Mon P Wang · 16 years ago
- 2e0d5f8 Simplify and generalize the SROA "convert to scalar" transformation to by Chris Lattner · 16 years ago
- dd9db66 by Devang Patel · 16 years ago
- 1ff06b2 This is case is to uncover the bug in IntrinsicLowering.cpp, by Zhou Sheng · 16 years ago
- 5585db0 Enable target tripple. by Devang Patel · 16 years ago
- de9e4ce Linux and other target's encoding for DW_AT_declaration may not match. by Devang Patel · 16 years ago
- d234e59 Add DW_AT_declaration for class methods. by Devang Patel · 16 years ago
- af399a6 XFAIL this test. It only worked before because of a bug in the spill point selection code. Not deleting because by Owen Anderson · 16 years ago
- d5a4802 Local register allocator shouldn't assume only the entry and landing pad basic blocks have live-ins. by Evan Cheng · 16 years ago
- 197e88f In the case of an extractelement on an insertelement value, by Dan Gohman · 16 years ago
- b691b70 Add a always_inline test case. by Evan Cheng · 16 years ago
- 6db3612 Add a test case for Chris lvalue alignment fixes. by Evan Cheng · 16 years ago
- 5a3c6a8 Exit with nice warnings when register allocator run out of registers. by Evan Cheng · 16 years ago
- e5af2d3 Make x86's BT instruction matching more thorough, and add some by Dan Gohman · 16 years ago
- e91a000 Fixed lowering of v816 shuffles. by Mon P Wang · 16 years ago
- 7540561 Make test platform agnostic. by Bill Wendling · 16 years ago
- ef42a3b Give this test an explicit target, to make it host-independent. by Dan Gohman · 16 years ago
- f923574 by Devang Patel · 16 years ago
- 6e733d3 Fix some issues with volatility, move "CanConvertToScalar" check by Chris Lattner · 16 years ago
- 6c8e35f strengthen this test. by Chris Lattner · 16 years ago
- 1632782 The memory alignment requirement on some of the mov{h|l}p{d|s} patterns are 16-byte. That is overly strict. These instructions read / write f64 memory locations without alignment requirement. by Evan Cheng · 16 years ago
- f4646d9 Added sse test patterns for r62979 and r63193. by Mon P Wang · 16 years ago
- 19d3e82 Add three new option properties. by Mikhail Glushenkov · 16 years ago
- ac7de20 Add testcase for r63142. by Bill Wendling · 16 years ago
- 6a86bd7 Implement multiple with overflow by 2 with an add instruction. by Evan Cheng · 16 years ago
- ccc497f Forgot this test case. by Evan Cheng · 16 years ago
- 49237f0 Add a FrontendC testcase for the x86-64 Red Zone feature, by Dan Gohman · 16 years ago
- ad165be by Devang Patel · 16 years ago
- 4e1fae5 Add a regression test for x86-64 red zone usage. by Dan Gohman · 16 years ago
- 8bca3b7 Testcase for 6522054. by Dale Johannesen · 16 years ago
- 5bb11b8 Fix PR3393, which amounts to a bug in the expensive by Duncan Sands · 16 years ago
- 4699934 At Nick Lewycky's request, rename this test with a more informative name. by Dan Gohman · 16 years ago
- 5b2e589 Enhance logic in X86DAGToDAGISel::PreprocessForRMW which move load inside callseq_start to allow it to be folded into a call. It was not considering the cases where a token factor is between the load and the callseq_start. by Evan Cheng · 16 years ago
- fe6d2cd Fixed optimization of combining two shuffles where the first shuffle inputs by Mon P Wang · 16 years ago
- c9c8b2a CellSPU: by Scott Michel · 16 years ago
- 18f0231 Handle single-entry phi nodes gracefully in condprop. by Chris Lattner · 16 years ago
- 3796a26 Fix PR3408 by making a non-obvious assumption very obvious, and by Chris Lattner · 16 years ago
- 51a0437 Map address space 256 to gs; similar mappings could be supported for the by Nate Begeman · 16 years ago
- c8ad2d2 revert this patch for now, because Codegen does still want to generate SSE code, by Torok Edwin · 16 years ago
- a10b879 testcase for llvm-gcc part of PR3402. by Torok Edwin · 16 years ago
- 7f1d525 If user explicitly asks not to use SSE, don't force it. This fixes LLVM part of PR3402. by Torok Edwin · 16 years ago
- 67e1f49 Actually run the test in this directory. by Nick Lewycky · 16 years ago
- 9d49791 The function that does nothing but call malloc is noalias return. by Nick Lewycky · 16 years ago
- 1902a12 Private linkage support for PPC / Darwin. by Evan Cheng · 16 years ago
- d498c8f Teach 2addr pass to be do more commuting. If both uses of a two-address instruction are killed, but the first operand has a use before and after the def, commute if the second operand does not suffer from the same issue. by Evan Cheng · 16 years ago
- 54e853a Add a PR comment to this test. by Dan Gohman · 16 years ago
- 08ffee5 testcase for PR3381. Also it was an empty struct, not a void after all. by Torok Edwin · 16 years ago
- 3914f72 Make InstCombineStoreToCast handle aggregates more aggressively, by Chris Lattner · 16 years ago
- e591b20 Fix test case. Use valid file name and directory in global variable's debug info entry. by Devang Patel · 16 years ago
- 95900f2 fix two more cases where we could let the NLPDI cache get unsorted. by Chris Lattner · 16 years ago
- 3e2351f Update test to reflect command line option name change. by Evan Cheng · 16 years ago
- 760f86f Don't create ISD::FNEG nodes after legalize if they aren't legal. by Dan Gohman · 16 years ago
- 5e3c013 Do not use buggy llvm-gcc to generate testcases. by Devang Patel · 16 years ago
- 10dfea8 Remove no-longer relevant comment. Pointed out by Gabor. by Duncan Sands · 16 years ago
- 4be3a22 This passes on linux. by Duncan Sands · 16 years ago
- 1c2ad9f fix a testcase. by Chris Lattner · 16 years ago