| Commit message (Expand) | Author | Age | Files | Lines |
* | [InstCombine] Fold Select with binary op | David Bolvansky | 2018-07-30 | 1 | -0/+33 |
* | [InstCombine] canonicalize abs pattern | Chen Zheng | 2018-07-27 | 1 | -20/+50 |
* | [InstCombine] add more SPFofSPF folding | Chen Zheng | 2018-07-16 | 1 | -0/+5 |
* | [InstCombine] Correct the cmp operand type used when canonicalizing abs/nabs | John Brawn | 2018-06-05 | 1 | -1/+1 |
* | [InstCombine] narrow select to match condition operands' size | Sanjay Patel | 2018-05-31 | 1 | -8/+11 |
* | [InstCombine] choose 1 form of abs and nabs as canonical | Sanjay Patel | 2018-05-20 | 1 | -0/+51 |
* | [InstCombine] Qualify a select pattern based transform to restrct to only min... | Craig Topper | 2018-05-18 | 1 | -1/+1 |
* | [InstCombine] refine select-of-constants to bitwise ops | Sanjay Patel | 2018-05-03 | 1 | -57/+34 |
* | [InstCombine] clean up foldSelectICmpAnd(); NFC | Sanjay Patel | 2018-04-25 | 1 | -46/+39 |
* | [InstCombine]: foldSelectICmpAndAnd(): and is commutative | Roman Lebedev | 2018-04-13 | 1 | -24/+20 |
* | [InstCombine] Get rid of select of bittest (PR36950 / PR17564) | Roman Lebedev | 2018-04-07 | 1 | -0/+49 |
* | [PatternMatch] allow undef elements when matching vector FP +0.0 | Sanjay Patel | 2018-03-25 | 1 | -4/+4 |
* | [InstCombine] canonicalize fcmp+select to fabs | Sanjay Patel | 2018-03-19 | 1 | -1/+31 |
* | [InstCombine] Replace calls to getNumUses with hasNUses or hasNUsesOrMore | Craig Topper | 2018-03-12 | 1 | -2/+2 |
* | [InstCombine] simplify min/max canonicalization; NFCI | Sanjay Patel | 2018-03-06 | 1 | -10/+5 |
* | [ValueTracking] move helpers for SelectPatterns from InstCombine to ValueTrac... | Sanjay Patel | 2018-03-06 | 1 | -51/+11 |
* | [InstCombine] Don't fold select(C, Z, binop(select(C, X, Y), W)) -> select(C,... | Craig Topper | 2018-02-14 | 1 | -2/+17 |
* | [InstCombine] add unsigned saturation subtraction canonicalizations | Sanjay Patel | 2018-02-05 | 1 | -1/+56 |
* | [InstCombine] Make foldSelectOpOp able to handle two-operand getelementptr | John Brawn | 2018-01-19 | 1 | -7/+19 |
* | [InstCombine] fold min/max tree with common operand (PR35717) | Sanjay Patel | 2018-01-08 | 1 | -0/+60 |
* | [InstCombine] relax use constraint for min/max (~a, ~b) --> ~min/max(a, b) | Sanjay Patel | 2018-01-06 | 1 | -2/+2 |
* | [InstCombine] add folds for min(~a, b) --> ~max(a, b) | Sanjay Patel | 2018-01-05 | 1 | -22/+12 |
* | [InstCombine] Simplify binops that are only used by a select and are fed by a... | Craig Topper | 2017-11-15 | 1 | -0/+40 |
* | [InstCombine] Simplify selects that test cmpxchg instructions | Matthew Simpson | 2017-10-31 | 1 | -0/+76 |
* | [Transforms] Fix some Clang-tidy modernize and Include What You Use warnings;... | Eugene Zelenko | 2017-10-24 | 1 | -6/+26 |
* | [InstCombine] Move foldSelectICmpAnd helper function earlier in the file to e... | Craig Topper | 2017-09-05 | 1 | -105/+105 |
* | [InstCombine] In foldSelectIntoOp, avoid creating a Constant before we know f... | Craig Topper | 2017-09-05 | 1 | -17/+18 |
* | [InstCombine][InstSimplify] Teach decomposeBitTestICmp to look through trunca... | Craig Topper | 2017-09-01 | 1 | -1/+1 |
* | [InstCombine] remove unnecessary vector select fold; NFCI | Sanjay Patel | 2017-08-30 | 1 | -4/+0 |
* | [InstCombine] Teach foldSelectICmpAndOr to handle vector splats | Craig Topper | 2017-08-29 | 1 | -6/+8 |
* | [InstCombine] Teach select01 helper of foldSelectIntoOp to handle vector splats | Craig Topper | 2017-08-28 | 1 | -7/+6 |
* | [InstCombine] Teach foldSelectICmpAnd to recognize a (icmp slt X, 0) and (icm... | Craig Topper | 2017-08-21 | 1 | -19/+49 |
* | [InstCombine] Make folding (X >s -1) ? C1 : C2 --> ((X >>s 31) & (C2 - C1)) +... | Craig Topper | 2017-08-16 | 1 | -17/+22 |
* | [InstCombine] Cast to BinaryOperator earlier in foldSelectIntoOp to simplify ... | Craig Topper | 2017-08-08 | 1 | -14/+10 |
* | [InstCombine] Support vector splats in foldSelectICmpAnd. | Craig Topper | 2017-08-05 | 1 | -15/+23 |
* | [InstCombine] In foldSelectICmpAnd, if we need to to truncate from the 'and' ... | Craig Topper | 2017-08-05 | 1 | -8/+8 |
* | [InstCombine] Use ConstantInt::getFalse to reduce some code. NFC | Craig Topper | 2017-08-04 | 1 | -2/+1 |
* | [InstCombine] Canonicalize clamp of float types to minmax in fast mode. | Nikolai Bozhenov | 2017-08-04 | 1 | -7/+17 |
* | [InstCombine] Move the call to foldSelectICmpAnd into foldSelectInstWithICmp.... | Craig Topper | 2017-08-04 | 1 | -80/+80 |
* | [Value Tracking] Default argument to true and rename accordingly. NFC. | Chad Rosier | 2017-08-01 | 1 | -2/+2 |
* | Disable loop unswitching for some patterns containing equality comparison wit... | Wei Mi | 2017-07-25 | 1 | -0/+16 |
* | [IR] Add Type::isIntOrIntVectorTy(unsigned) similar to the existing isInteger... | Craig Topper | 2017-07-09 | 1 | -2/+2 |
* | [InstCombine] Speculatively implement a fix for what might be the root cause ... | Craig Topper | 2017-07-09 | 1 | -1/+2 |
* | [InstCombine] Make InstCombine's IRBuilder be passed by reference everywhere | Craig Topper | 2017-07-07 | 1 | -70/+69 |
* | [Constants] If we already have a ConstantInt*, prefer to use isZero/isOne/isM... | Craig Topper | 2017-07-06 | 1 | -5/+5 |
* | [InstCombine] Don't create extra ConstantInt objects in foldSelectICmpAnd. NFCI | Craig Topper | 2017-07-06 | 1 | -19/+17 |
* | Revert of r306525: "Canonicalize clamp of float types to minmax" | Nikolai Bozhenov | 2017-06-30 | 1 | -10/+3 |
* | [InstCombine] Canonicalize clamp of float types to minmax in fast mode. | Nikolai Bozhenov | 2017-06-28 | 1 | -3/+10 |
* | [InstCombine] canonicalize icmp predicate feeding select | Sanjay Patel | 2017-06-27 | 1 | -0/+17 |
* | [InstCombine] Teach foldSelectICmpAndOr to recognize (select (icmp slt (trunc... | Craig Topper | 2017-06-22 | 1 | -11/+38 |