Fix PR3667
[oota-llvm.git] / test / Transforms / InstCombine /
2009-02-25 Chris LattnerFix PR3667
2009-02-24 Dan GohmanAdd a testcase for the problem fixed in r65289.
2009-02-24 Dan GohmanFix a ValueTracking rule: RHS means operand 1, not...
2009-02-21 Nick LewyckyDon't sign extend the char when expanding char -> int...
2009-02-21 Chris Lattnerrename a function to indicate that it checks for profit...
2009-02-18 Dan GohmanChange the argument type in this test to something...
2009-02-17 Chris Lattnercommit a tweaked version of Daniel's patch for PR3599...
2009-02-16 Dan GohmanFix EnforceKnownAlignment so that it doesn't ever reduc...
2009-02-16 Dan GohmanChange these tests to use regular loads instead of...
2009-02-12 Nate BegemanAdd suppport for ConstantExprs of shufflevectors whose...
2009-02-07 Mon P WangInstrcombine should not change load(cast p) to cast...
2009-02-04 Duncan SandsAllow the inverse transform x86_fp80 -> i80 (also
2009-02-04 Duncan SandsFix PR3468: a crash when constant folding a bitcast of
2009-02-03 Evan ChengAPInt'fy SimplifyDemandedVectorElts so it can analyze...
2009-02-02 Chris Lattnerreduce testcase.
2009-01-31 Nick LewyckyReinstate this optimization to fold icmp of xor when...
2009-01-31 Chris LattnerFix PR3452 (an infinite loop bootstrapping) by disablin...
2009-01-31 Chris Lattnernow that all the pieces are in place, teach instcombine's
2009-01-31 Chris Lattnermake sure to set Changed=true when instcombine hacks...
2009-01-26 Mon P WangFixed optimization of combining two shuffles where...
2009-01-24 Torok Edwintestcase for PR3381.
2009-01-24 Chris LattnerMake InstCombineStoreToCast handle aggregates more...
2009-01-21 Dale JohannesenDo not use host floating point types when emitting
2009-01-21 Dale JohannesenDisable on x86_64 until I figure out what's wrong.
2009-01-21 Dale JohannesenMake special cases (0 inf nan) work for frem.
2009-01-20 Dale JohannesenCalls to fmod, it turns out, are constant-folded by
2009-01-19 Dale JohannesenMove & restructure test per review.
2009-01-16 Chris LattnerFix PR3335 by not turning a store to one address space...
2009-01-16 Evan ChengClean up previous cast optimization a bit. Also make...
2009-01-15 Evan Cheng- Teach CanEvaluateInDifferentType of this xform: sext...
2009-01-13 Dan GohmanMake instcombine ensure that all allocas are explicitly...
2009-01-09 Chris LattnerImplement rdar://6480391, extending of equality icmp...
2009-01-09 Chris LattnerFix part 3/2 of PR3290, making instcombine zap (gep...
2009-01-08 Chris LattnerValueTracker can't assume that an alloca with no specif...
2009-01-05 Chris Lattnermake m_ConstantInt(int64_t) safely match ConstantInt...
2009-01-04 Bill WendlingXFAIL this test. The xform was removed.
2009-01-01 Bill WendlingAdd transformation:
2008-12-18 Nick LewyckyMake all the vector elements positive in an srem of...
2008-12-01 Bill WendlingUse m_Specific() instead of double matching.
2008-12-01 Chris Lattnersimplify these patterns using m_Specific. No need...
2008-12-01 Chris LattnerTeach inst combine to merge GEPs through PHIs. This...
2008-12-01 Bill WendlingImplement ((A|B)&1)|(B&-2) -> (A&1) | B transformation...
2008-11-30 Bill WendlingAdd instruction combining for ((A&~B)|(~A&B)) -> A...
2008-11-30 Bill WendlingImplement (A&((~A)|B)) -> A&B transformation in the...
2008-11-30 Bill WendlinggetSExtValue() doesn't work for ConstantInts with bitwi...
2008-11-30 Bill WendlingStrengthen check for div inst-combining.
2008-11-30 Bill WendlingInstcombine was illegally transforming -X/C into X...
2008-11-27 Nick LewyckyChris prefers icmp/select over udiv!
2008-11-27 Nick LewyckyAdd a couple of missed optimizations on integer vectors...
2008-11-21 Nick LewyckyOptimize (x/y)*y into x-(x%y) in general. Div and rem...
2008-11-16 Chris LattnerHandle the case where there is no "not". It is possibl...
2008-11-16 Chris Lattnermake this actually test what it is trying to.
2008-11-09 Bill WendlingIf the LHS of the FCMP is coming from a UIToFP instruct...
2008-11-02 Nick LewyckyFix demanded bits analysis with srem by negative number...
2008-11-02 Dan GohmanFix this recently moved code to use the correct type...
2008-10-30 Dan GohmanCanonicalize sext(i1) to i1?-1:0, and update various...
2008-10-28 Dan Gohman(A & sext(C)) | (B & ~sext(C) -> C ? A : B
2008-10-24 Nick LewyckyDon't try to create a mask when we don't need one....
2008-10-15 Dan GohmanTeach instcombine's visitLoad to scan back several...
2008-10-14 Evan ChengCombine (fcmp cc0 x, y) | (fcmp cc1 x, y) into a single...
2008-10-14 Evan Cheng- Somehow I forgot about one / une.
2008-10-14 Evan ChengOptimize anding of two fcmp into a single fcmp if the...
2008-10-11 Chris LattnerFix PR2697 by rewriting the '(X / pos) op neg' logic...
2008-10-05 Chris Lattnerrewrite bswap matching to be more general, allowing...
2008-09-30 Nick LewyckyFix misoptimization of: xor i1 (icmp eq (X, C1), icmp...
2008-09-16 Dan GohmanImprove instcombine's handling of integer min and max...
2008-09-11 Dan GohmanOn 64-bit targets, change 32-bit getelementptr indices...
2008-09-11 Dan GohmanFix a vectorshuffle instcombine bug introduced by r55995.
2008-09-10 Dan GohmanFix an icmp+sdiv optimization to check for and handle...
2008-09-09 Dan GohmanMake SimplifyDemandedVectorElts simplify vectors with...
2008-09-03 Nick LewyckyDon't crash when trying to constant fold a vector with...
2008-08-21 Nick LewyckyRevert r54876 r54877 r54906 and r54907. Evan found...
2008-08-17 Nick LewyckyConsider the case where xor by -1 and xor by 128 have...
2008-08-17 Nick LewyckyXor'ing both sides of icmp by sign-bit is equivalent...
2008-08-16 Owen AndersonRemove GCSE and LoadVN from the testsuite.
2008-08-13 Dan GohmanFix a bogus srem rule - a negative value srem'd by...
2008-08-11 Chris LattnerImplement support for simplifying vector comparisons...
2008-08-06 Dan GohmanFix a shufflevector instcombine that was emitting inval...
2008-08-06 Chris Lattneroptimize a common idiom generated by clang for bitfield...
2008-08-06 Chris LattnerZap sitofp/fptoui pairs. In all cases when the sign...
2008-08-06 Nick LewyckyReinstate this optimization, but without the miscompile...
2008-08-06 Bill WendlingJust grep for through the LL code instead of the ASM...
2008-08-05 Bill WendlingAdd default architecture.
2008-08-05 Bill WendlingTestcase for PR2629.
2008-08-05 Bill WendlingRevert r53282. This was causing a miscompile on Linux...
2008-07-17 Chris LattnerFix PR2553
2008-07-16 Matthijs KooijmanAdd a few cases to instcombine's extractvalue testcase.
2008-07-16 Evan ChengFix PR2296. Do not transform x86_sse2_storel_dq into...
2008-07-14 Chris LattnerFix PR2506 by being a bit more careful about reverse...
2008-07-12 Nick LewyckyEnhance analysis of srem.
2008-07-11 Nick LewyckyAdd another optimization from PR2330. Also catch some...
2008-07-11 Chris LattnerFix folding of icmp's of i1 where the comparison is...
2008-07-11 Chris LattnerFix a bogus optimization: folding (slt (zext i1 A to...
2008-07-11 Chris Lattnermake this condition more precise.
2008-07-10 Nick LewyckyFix overzealous optimization. Thanks to Duncan Sands...
2008-07-09 Nick LewyckyFold (a < 8) && (b < 8) into (a|b) < 8 for unsigned...
2008-07-09 Nick LewyckyFold ((1 << a) & 1) to (a == 0).
2008-07-09 Chris LattnerFix a broken test. Neither load is eliminable without...
2008-07-09 Nick LewyckyReduce x - y to -y when we know the 'x' part will get...
2008-07-08 Chris Lattnernew testcase for PR2496
next