Changes in ConstantRangeTest.cpp

2e912669 - [ConstantRangeTest] Migrate known bits test to generic infrastructure (NFC)

Mon, 18 Jul 2022 13:18:47 +0000

[ConstantRangeTest] Migrate known bits test to generic infrastructure (NFC)This can't make use of TestBinaryOpExhaustive, but it can make useof the general TestRange approach that collects the precise elementsin a bit vector.This allows us to remove the obsolete "op range gatherer" infrastructure. List of files: /llvm-project-15.0.7/llvm/unittests/IR/ConstantRangeTest.cpp

b57d6138 - [ConstantRangeTest] Move nowrap binop tests to generic infrastructure (NFC)

Mon, 18 Jul 2022 13:13:12 +0000

[ConstantRangeTest] Move nowrap binop tests to generic infrastructure (NFC)Move testing for add/sub with nowrap flags to TestBinaryOpExhaustive,rather than separate homegrown exhaustive testing functions. List of files: /llvm-project-15.0.7/llvm/unittests/IR/ConstantRangeTest.cpp

ba1e04b9 - [ConstantRange] Fix sdiv() with one bit values (PR56333)

Fri, 01 Jul 2022 13:43:27 +0000

[ConstantRange] Fix sdiv() with one bit values (PR56333)Signed one bit values can only be -1 or 0, not positive. The codewas interpreting the 1 as -1 and intersecting with a full rangerather than an empty one.Fixes https://github.com/llvm/llvm-project/issues/56333. List of files: /llvm-project-15.0.7/llvm/unittests/IR/ConstantRangeTest.cpp

d152e50c - [llvm] Don't use Optional::{hasValue,getValue} (NFC)

Sat, 25 Jun 2022 18:24:23 +0000

[llvm] Don't use Optional::{hasValue,getValue} (NFC) List of files: /llvm-project-15.0.7/llvm/unittests/IR/ConstantRangeTest.cpp

9398caf3 - Recommit "[ConstantRange] Improve the implementation of binaryOr"

Fri, 20 May 2022 18:39:21 +0000

Recommit "[ConstantRange] Improve the implementation of binaryOr"This recommits https://reviews.llvm.org/rG6990e7477d24ff585ae86549f5280f0be65422a6as the problematic test has been updated updated inhttps://reviews.llvm.org/rG3bd112c720dc614a59e3f34ebf9b45075037bfa0. List of files: /llvm-project-15.0.7/llvm/unittests/IR/ConstantRangeTest.cpp

54e3bf5f - Revert "[ConstantRange] Improve the implementation of binaryOr"

Fri, 20 May 2022 17:21:24 +0000

Revert "[ConstantRange] Improve the implementation of binaryOr"This reverts commit 6990e7477d24ff585ae86549f5280f0be65422a6.This change was causing the test compiler-rt/test/fuzzer/merge_two_step.test to fail onour internal bot as well as other build bots such as https://lab.llvm.org/buildbot/#/builders/179/builds/3712. List of files: /llvm-project-15.0.7/llvm/unittests/IR/ConstantRangeTest.cpp

6990e747 - [ConstantRange] Improve the implementation of binaryOr

Thu, 19 May 2022 20:25:47 +0000

[ConstantRange] Improve the implementation of binaryOrThis diff adjusts binaryOr to take advantage of the analysisbased on KnownBits.Differential revision: https://reviews.llvm.org/D125933Test plan:1/ ninja check-llvm2/ ninja check-llvm-unit List of files: /llvm-project-15.0.7/llvm/unittests/IR/ConstantRangeTest.cpp

0f4d9f9b - [ConstantRange] Improve the implementation of binaryAnd

Tue, 17 May 2022 21:30:44 +0000

[ConstantRange] Improve the implementation of binaryAndThis diff adjusts binaryAnd to take advantage of the analysisbased on KnownBits.Differential revision: https://reviews.llvm.org/D125603Test plan:1/ ninja check-llvm2/ ninja check-llvm-unit List of files: /llvm-project-15.0.7/llvm/unittests/IR/ConstantRangeTest.cpp

2db4dc7e - [ConstantRange] Implement binaryXor() using known bits

Tue, 17 May 2022 08:02:50 +0000

[ConstantRange] Implement binaryXor() using known bitsThis allows us to compute known high bits. It's not optimal, butbetter than nothing. List of files: /llvm-project-15.0.7/llvm/unittests/IR/ConstantRangeTest.cpp

a694546f - [KnownBits] Add operator==

Mon, 16 May 2022 15:20:19 +0000

[KnownBits] Add operator==Checking whether two KnownBits are the same is somewhat common,mainly in test code.I don't think there is a lot of room for confusion with "determinewhat the KnownBits for an icmp eq would be", as that has adifferent result type (this is what the eq() method implements,which returns Optional).Differential Revision: https://reviews.llvm.org/D125692 List of files: /llvm-project-15.0.7/llvm/unittests/IR/ConstantRangeTest.cpp

8ab819ad - [ConstantRange] Add toKnownBits() method

Mon, 16 May 2022 14:04:11 +0000

[ConstantRange] Add toKnownBits() methodAdd toKnownBits() method to mirror fromKnownBits(). We know thetop bits that are constant between min and max.The return value for an empty range is chosen to be conservative. List of files: /llvm-project-15.0.7/llvm/unittests/IR/ConstantRangeTest.cpp

1c5d636a - [ConstantRangeTest] Add helper to enumerate APInts (NFC)

Fri, 12 Nov 2021 17:17:27 +0000

[ConstantRangeTest] Add helper to enumerate APInts (NFC)While ForeachNumInConstantRange(ConstantRange::getFull(Bits))works, it's somewhat roundabout, and I keep looking for thisfunction. List of files: /llvm-project-15.0.7/llvm/unittests/IR/ConstantRangeTest.cpp

2060895c - [ConstantRange] Add exact union/intersect (NFC)

Sun, 07 Nov 2021 20:14:21 +0000

[ConstantRange] Add exact union/intersect (NFC)For some optimizations on comparisons it's necessary that theunion/intersect is exact and not a superset. Add methods thatreturn Optional only if the result is exact.For the sake of simplicity this is implemented by comparingthe subset and superset approximations for now, but it should bepossible to do this more directly, as unionWith() and intersectWith()already distinguish the cases where the result is imprecise for thepreferred range type functionality. List of files: /llvm-project-15.0.7/llvm/unittests/IR/ConstantRangeTest.cpp

cf71a5ea - [ConstantRange] Support zero size in isSizeLargerThan()

Sun, 07 Nov 2021 20:21:41 +0000

[ConstantRange] Support zero size in isSizeLargerThan()From an API perspective, it does not make a lot of sense that 0is not a valid argument to this function. Add the exact check neededto support it. List of files: /llvm-project-15.0.7/llvm/unittests/IR/ConstantRangeTest.cpp

9f0194be - [ConstantRange] Add getEquivalentICmp() variant with offset (NFCI)

Sat, 06 Nov 2021 20:56:53 +0000

[ConstantRange] Add getEquivalentICmp() variant with offset (NFCI)Add a variant of getEquivalentICmp() that produces an optionaloffset. This allows us to create an equivalent icmp for all ranges.Use this in the with.overflow folding code, which was doing thisadjustment separately -- this clarifies that the fold will indeedalways apply. List of files: /llvm-project-15.0.7/llvm/unittests/IR/ConstantRangeTest.cpp

3348b841 - Make enum iteration with seq safe by default

Thu, 04 Nov 2021 00:47:57 +0000

Make enum iteration with seq safe by defaultBy default `llvm::seq` would happily iterate over enums, which may be unsafe if the enum values are not continuous. This patch disable enum iteration with `llvm::seq` and `llvm::seq_inclusive` and adds two new functions: `enum_seq` and `enum_seq_inclusive`.To make sure enum iteration is safe, we require users to declare their enum types as iterable by specializing `enum_iteration_traits`. Because it's not always possible to add these traits next to enum definition (e.g., for enums defined in external libraries), we provide an escape hatch to allow iteration on per-callsite basis by passing `force_iteration_on_noniterable_enum`.The main benefit of this approach is that these global declarations via traits can appear just next to enum definitions, making easy to spot when enums are miss-labeled, e.g., after introducing new enum values, whereas `force_iteration_on_noniterable_enum` should stand out and be easy to grep for.This emerged from a discussion with gchatelet@ about reusing llvm's `Sequence.h` in lieu of https://github.com/GPUOpen-Drivers/llpc/blob/dev/lgc/interface/lgc/EnumIterator.h.Reviewed By: dblaikie, gchatelet, aaron.ballmanDifferential Revision: https://reviews.llvm.org/D107378 List of files: /llvm-project-15.0.7/llvm/unittests/IR/ConstantRangeTest.cpp

03a4f1f3 - [ConstantRange] Sign-flipping of signedness-invariant comparisons

Fri, 06 Nov 2020 11:06:00 +0000

[ConstantRange] Sign-flipping of signedness-invariant comparisonsFor certain combination of LHS and RHS constant ranges,the signedness of the relational comparison predicate is irrelevant.This implements complete and precise model for all predicates,as confirmed by the brute-force tests. I'm not sure if there aresome more cases that we can handle here.In a follow-up, CVP will make use of this.Reviewed By: nikicDifferential Revision: https://reviews.llvm.org/D90924 List of files: /llvm-project-15.0.7/llvm/unittests/IR/ConstantRangeTest.cpp

25043c82 - [NFCI] Introduce `ICmpInst::compare()` and use it where appropriate

Sat, 30 Oct 2021 14:36:23 +0000

[NFCI] Introduce `ICmpInst::compare()` and use it where appropriateAs noted in https://reviews.llvm.org/D90924#inline-1076197apparently this is a pretty common pattern,let's not repeat it yet again, but have it in a common place.There may be some more places where it could be used,but these are the most obvious ones. List of files: /llvm-project-15.0.7/llvm/unittests/IR/ConstantRangeTest.cpp

274b2439 - [ConstantRange] Add fast signed multiply

Sat, 16 Oct 2021 19:02:04 +0000

[ConstantRange] Add fast signed multiplyThe multiply() implementation is very slow -- it performs sixmultiplications in double the bitwidth, which means that it willtypically work on allocated APInts and bypass fast-pathimplementations. Add an additional implementation that doesn'ttry to produce anything better than a full range if overflow ispossible. At least for the BasicAA use-case, we really don't careabout more precise modeling of overflow behavior. The currentuse of multiply() is fine while the implementation is limited toa single index, but extending it to the multiple-index case makesthe compile-time impact untenable. List of files: /llvm-project-15.0.7/llvm/unittests/IR/ConstantRangeTest.cpp

587493b4 - [ConstantRange] Compute precise shl range for single elements

Fri, 15 Oct 2021 19:25:39 +0000

[ConstantRange] Compute precise shl range for single elementsFor the common case where the shift amount is constant (a singleelement range) we can easily compute a precise range (up tounsigned envelope), so do that. List of files: /llvm-project-15.0.7/llvm/unittests/IR/ConstantRangeTest.cpp