Philip Reames [Fri, 2 Apr 2021 16:12:24 +0000 (09:12 -0700)]
[rs4gc] Strip nofree and nosync attributes when lowering from abstract model
The safepoints being inserted exists to free memory, or coordinate with another thread to do so. Thus, we must strip any inferred attributes and reinfer them after the lowering.
I'm not aware of any active miscompiles caused by this, but since I'm working on strengthening inference of both and leveraging them in the optimization decisions, I figured a bit of future proofing was warranted.
Philip Reames [Fri, 2 Apr 2021 15:56:49 +0000 (08:56 -0700)]
[rs4gc] add tests for existing code stripping attributes from function signatures
Aaron Ballman [Fri, 2 Apr 2021 15:33:04 +0000 (11:33 -0400)]
Remove attribute handling code for simple attributes; NFC
Attributes that set the SimpleHandler flag in Attr.td don't need to be
explicitly handled in SemaDeclAttr.cpp.
peter klausler [Fri, 2 Apr 2021 15:26:39 +0000 (08:26 -0700)]
[flang] Fix MSVC build breakage
A recent patch exposed an assumption that "long double" is (at least)
an 80-bit floating-point type, which of course it is not in MSVC.
Also get it right for non-x87 floating-point.
Brendon Cahoon [Tue, 30 Mar 2021 15:19:29 +0000 (11:19 -0400)]
[GlobalISel] Allow different types for G_SBFX and G_UBFX operands
Change the definition of G_SBFX and G_UBFX so that the lsb and width
can have different types than the src and dst operands.
Differential Revision: https://reviews.llvm.org/D99739
Nikita Popov [Fri, 2 Apr 2021 14:41:58 +0000 (16:41 +0200)]
[LVI] Use range metadata on intrinsics
If we don't know how to handle an intrinsic, we should still
make use of normal call range metadata.
Nikita Popov [Fri, 2 Apr 2021 14:38:38 +0000 (16:38 +0200)]
[CVP] Add test for !range on intrinsic (NFC)
Alexey Bataev [Fri, 2 Apr 2021 14:37:40 +0000 (07:37 -0700)]
[SLP]Added a test for min/max reductions with the key store inside, NFC.
Alexey Bataev [Thu, 1 Apr 2021 18:02:34 +0000 (11:02 -0700)]
[SLP]Fix a bug in min/max reduction, number of condition uses.
The ultimate reduction node may have multiple uses, but if the ultimate
reduction is min/max reduction and based on SelectInstruction, the
condition of this select instruction must have only single use.
Differential Revision: https://reviews.llvm.org/D99753
Nico Weber [Fri, 2 Apr 2021 13:54:22 +0000 (09:54 -0400)]
Revert "[X86][SSE] isHorizontalBinOp - use getTargetShuffleInputs helper"
This reverts commit
500969f1d0b1d92d7c4ccfb6bf8807de96b7e4a0.
Makes clang assert compiling avx2 code, see
https://bugs.chromium.org/p/chromium/issues/detail?id=1195353#c4
for a standalone repro.
Paul C. Anagnostopoulos [Tue, 30 Mar 2021 16:37:13 +0000 (12:37 -0400)]
[TableGen] [Docs] Add lldb-tblgen to command guide; add 4 guide stubs
Differential Revision: https://reviews.llvm.org/D99605
Nico Weber [Fri, 2 Apr 2021 13:19:31 +0000 (09:19 -0400)]
Restore
8954fd436c7 after
c06a8f9caa51c
Else, just-built clang can't build programs that include libc++ headers
on macOS if you build via the 'all' target.
Jun Ma [Fri, 2 Apr 2021 02:10:35 +0000 (10:10 +0800)]
[NFC][SVE] update sve-intrinsics-int-arith.ll under update_llc_test_checks.py
Jun Ma [Thu, 1 Apr 2021 11:44:59 +0000 (19:44 +0800)]
[AArch64][SVE] Lowering sve.dot to DOT node
Differential Revision: https://reviews.llvm.org/D99699
Jun Ma [Wed, 31 Mar 2021 09:52:22 +0000 (17:52 +0800)]
[NFC][SVE] Use SVE_4_Op_Imm_Pat for sve_intx_dot_by_indexed_elem
Lei Zhang [Fri, 2 Apr 2021 11:50:45 +0000 (07:50 -0400)]
[mlir][spirv] Add utilities for push constant value
This commit add utility functions for creating push constant
storage variable and loading values from it.
Along the way, performs some clean up:
* Deleted `setABIAttrs`, which is just a 4-liner function
with one user.
* Moved `SPIRVConverstionTarget` into `mlir` namespace,
to be consistent with `SPIRVTypeConverter` and
`LLVMConversionTarget`.
Reviewed By: mravishankar
Differential Revision: https://reviews.llvm.org/D99725
Jeroen Dobbelaere [Fri, 2 Apr 2021 11:45:11 +0000 (13:45 +0200)]
[InstCombine] Fix out-of-bounds ashr(shl) optimization
This fixes a crash found by the oss fuzzer and reported by @fhahn.
The suggestion of @RKSimon seems to be the correct fix here. (See D91343).
The oss fuzz report can be found here: https://bugs.chromium.org/p/oss-fuzz/issues/detail?id=32759
Reviewed By: lebedev.ri
Differential Revision: https://reviews.llvm.org/D99792
Fraser Cormack [Wed, 31 Mar 2021 11:59:22 +0000 (12:59 +0100)]
[RISCV] Test llvm.experimental.vector.insert intrinsics on RV32
RV32 is able to use the llvm.experimental.vector.insert intrinsics too.
This patch ensures they're tested.
Reviewed By: khchen, asb
Differential Revision: https://reviews.llvm.org/D99655
Muhammad Omair Javaid [Fri, 2 Apr 2021 10:54:27 +0000 (15:54 +0500)]
[LLDB] Skip TestLoadUsingLazyBind.py on arm/linux
Simon Pilgrim [Fri, 2 Apr 2021 08:43:51 +0000 (09:43 +0100)]
[X86][SSE] isHorizontalBinOp - use getTargetShuffleInputs helper
Use the getTargetShuffleInputs helper for all shuffle decoding
LLVM GN Syncbot [Fri, 2 Apr 2021 10:22:54 +0000 (10:22 +0000)]
[gn build] Port
0f7bbbc481e2
Sander de Smalen [Wed, 17 Mar 2021 21:35:09 +0000 (21:35 +0000)]
Always emit error for wrong interfaces to scalable vectors, unless cmdline flag is passed.
In order to bring up scalable vector support in LLVM incrementally,
we introduced behaviour to emit a warning, instead of an error, when
asking the wrong question of a scalable vector, like asking for the
fixed number of elements.
This patch puts that behaviour under a flag. The default behaviour is
that the compiler will always error, which means that all LLVM unit
tests and regression tests will now fail when a code-path is taken that
still uses the wrong interface.
The behaviour to demote an error to a warning can be individually enabled
for tools that want to support experimental use of scalable vectors.
This patch enables that behaviour when driving compilation from Clang.
This means that for users who want to try out scalable-vector support,
fixed-width codegen support, or build user-code with scalable vector
intrinsics, Clang will not crash and burn when the compiler encounters
such a case.
This allows us to do away with the following pattern in many of the SVE tests:
RUN: .... 2>%t
RUN: cat %t | FileCheck --check-prefix=WARN
WARN-NOT: warning: ...
The behaviour to emit warnings is only temporary and we expect this flag
to be removed in the future when scalable vector support is more stable.
This patch also has fixes the following tests:
unittests:
ScalableVectorMVTsTest.SizeQueries
SelectionDAGAddressAnalysisTest.unknownSizeFrameObjects
AArch64SelectionDAGTest.computeKnownBitsSVE_ZERO_EXTEND_VECTOR_INREG
regression tests:
Transforms/InstCombine/vscale_gep.ll
Reviewed By: paulwalker-arm, ctetreau
Differential Revision: https://reviews.llvm.org/D98856
Florian Hahn [Fri, 2 Apr 2021 09:40:12 +0000 (10:40 +0100)]
[SLP] Better estimate cost of no-op extracts on target vectors.
The motivation for this patch is to better estimate the cost of
extracelement instructions in cases were they are going to be free,
because the source vector can be used directly.
A simple example is
%v1.lane.0 = extractelement <2 x double> %v.1, i32 0
%v1.lane.1 = extractelement <2 x double> %v.1, i32 1
%a.lane.0 = fmul double %v1.lane.0, %x
%a.lane.1 = fmul double %v1.lane.1, %y
Currently we only consider the extracts free, if there are no other
users.
In this particular case, on AArch64 which can fit <2 x double> in a
vector register, the extracts should be free, independently of other
users, because the source vector of the extracts will be in a vector
register directly, so it should be free to use the vector directly.
The SLP vectorized version of noop_extracts_9_lanes is 30%-50% faster on
certain AArch64 CPUs.
It looks like this does not impact any code in
SPEC2000/SPEC2006/MultiSource both on X86 and AArch64 with -O3 -flto.
This originally regressed after D80773, so if there's a better
alternative to explore, I'd be more than happy to do that.
Reviewed By: ABataev
Differential Revision: https://reviews.llvm.org/D99719
Fraser Cormack [Thu, 1 Apr 2021 14:17:32 +0000 (15:17 +0100)]
[RISCV] Optimize more redundant VSETVLIs
D99717 introduced some test cases which showed that the output of one
vsetvli into another would not be picked up by the RISCVCleanupVSETVLI
pass. This patch teaches the optimization about such a pattern. The
pattern is quite common when using the RVV vsetvli intrinsic to pass the
VL onto other intrinsics.
The second test case introduced by D99717 is left unoptimized by this
patch. It is a rarer case and will require us to rewire any uses of the
redundant vset[i]vli's output to the previous one's.
Reviewed By: craig.topper
Differential Revision: https://reviews.llvm.org/D99730
Fraser Cormack [Thu, 1 Apr 2021 10:24:42 +0000 (11:24 +0100)]
[RISCV] Add some tests showing vsetvli cleanup opportunities
Reviewed By: craig.topper
Differential Revision: https://reviews.llvm.org/D99717
Marek Kurdej [Fri, 2 Apr 2021 08:31:18 +0000 (10:31 +0200)]
[libc++] Fix build on macOS older than 10.15.
* This was introduced in D99515 that added -Wundef flag. CI run on macOS 10.15 and this problem wasn't caught before.
Evgeniy Brevnov [Tue, 2 Mar 2021 08:14:32 +0000 (15:14 +0700)]
[NARY-REASSOCIATE] Support reassociation of min/max
Support reassociation for min/max. With that we should be able to transform min(min(a, b), c) -> min(min(a, c), b) if min(a, c) is already available.
Reviewed By: mkazantsev, lebedev.ri
Differential Revision: https://reviews.llvm.org/D88287
Roman Lebedev [Fri, 2 Apr 2021 07:40:12 +0000 (10:40 +0300)]
[PassManager] Run additional LICM before LoopRotate
Loop rotation often has to perform code duplication
from header into preheader, which introduces PHI nodes.
>>! In D99204, @thopre wrote:
>
> With loop peeling, it is important that unnecessary PHIs be avoided or
> it will leads to spurious peeling. One source of such PHIs is loop
> rotation which creates PHIs for invariant loads. Those PHIs are
> particularly problematic since loop peeling is now run as part of simple
> loop unrolling before GVN is run, and are thus a source of spurious
> peeling.
>
> Note that while some of the load can be hoisted and eventually
> eliminated by instruction combine, this is not always possible due to
> alignment issue. In particular, the motivating example [1] was a load
> inside a class instance which cannot be hoisted because the `this'
> pointer has an alignment of 1.
>
> [1] http://lists.llvm.org/pipermail/llvm-dev/attachments/
20210312/
4ce73c47/attachment.cpp
Now, we could enhance LoopRotate to avoid duplicating code when not needed,
but instead hoist loop-invariant code, but isn't that a code duplication? (*sic*)
We have LICM, and in fact we already run it right after LoopRotation.
We could try to move it to before LoopRotation,
that is basically free from compile-time perspective:
https://llvm-compile-time-tracker.com/compare.php?from=
6c93eb4477d88af046b915bc955c03693b2cbb58&to=
a4bee6d07732b1184c436da489040b912f0dc271&stat=instructions
But, looking at stats, i think it isn't great that we would no longer do LICM after LoopRotation, in particular:
| statistic name | LoopRotate-LICM | LICM-LoopRotate | Δ | % | abs(%) |
| asm-printer.EmittedInsts | 9015930 | 9015799 | -131 | 0.00% | 0.00% |
| indvars.NumElimCmp | 3536 | 3544 | 8 | 0.23% | 0.23% |
| indvars.NumElimExt | 36725 | 36580 | -145 | -0.39% | 0.39% |
| indvars.NumElimIV | 1197 | 1187 | -10 | -0.84% | 0.84% |
| indvars.NumElimIdentity | 143 | 136 | -7 | -4.90% | 4.90% |
| indvars.NumElimRem | 4 | 5 | 1 | 25.00% | 25.00% |
| indvars.NumLFTR | 29842 | 29890 | 48 | 0.16% | 0.16% |
| indvars.NumReplaced | 2293 | 2227 | -66 | -2.88% | 2.88% |
| indvars.NumSimplifiedSDiv | 6 | 8 | 2 | 33.33% | 33.33% |
| indvars.NumWidened | 26438 | 26329 | -109 | -0.41% | 0.41% |
| instcount.TotalBlocks | 1178338 | 1173840 | -4498 | -0.38% | 0.38% |
| instcount.TotalFuncs | 111825 | 111829 | 4 | 0.00% | 0.00% |
| instcount.TotalInsts | 9905442 | 9896139 | -9303 | -0.09% | 0.09% |
| lcssa.NumLCSSA | 425871 | 423961 | -1910 | -0.45% | 0.45% |
| licm.NumHoisted | 378357 | 378753 | 396 | 0.10% | 0.10% |
| licm.NumMovedCalls | 2193 | 2208 | 15 | 0.68% | 0.68% |
| licm.NumMovedLoads | 35899 | 31821 | -4078 | -11.36% | 11.36% |
| licm.NumPromoted | 11178 | 11154 | -24 | -0.21% | 0.21% |
| licm.NumSunk | 13359 | 13587 | 228 | 1.71% | 1.71% |
| loop-delete.NumDeleted | 8547 | 8402 | -145 | -1.70% | 1.70% |
| loop-instsimplify.NumSimplified | 12876 | 11890 | -986 | -7.66% | 7.66% |
| loop-peel.NumPeeled | 1008 | 925 | -83 | -8.23% | 8.23% |
| loop-rotate.NumNotRotatedDueToHeaderSize | 368 | 365 | -3 | -0.82% | 0.82% |
| loop-rotate.NumRotated | 42015 | 42003 | -12 | -0.03% | 0.03% |
| loop-simplifycfg.NumLoopBlocksDeleted | 240 | 242 | 2 | 0.83% | 0.83% |
| loop-simplifycfg.NumLoopExitsDeleted | 497 | 20 | -477 | -95.98% | 95.98% |
| loop-simplifycfg.NumTerminatorsFolded | 618 | 336 | -282 | -45.63% | 45.63% |
| loop-unroll.NumCompletelyUnrolled | 11028 | 11032 | 4 | 0.04% | 0.04% |
| loop-unroll.NumUnrolled | 12608 | 12529 | -79 | -0.63% | 0.63% |
| mem2reg.NumDeadAlloca | 10222 | 10221 | -1 | -0.01% | 0.01% |
| mem2reg.NumPHIInsert | 192110 | 192106 | -4 | 0.00% | 0.00% |
| mem2reg.NumSingleStore | 637650 | 637643 | -7 | 0.00% | 0.00% |
| scalar-evolution.NumBruteForceTripCountsComputed | 814 | 812 | -2 | -0.25% | 0.25% |
| scalar-evolution.NumTripCountsComputed | 283108 | 282934 | -174 | -0.06% | 0.06% |
| scalar-evolution.NumTripCountsNotComputed | 106712 | 106718 | 6 | 0.01% | 0.01% |
| simple-loop-unswitch.NumBranches | 5178 | 4752 | -426 | -8.23% | 8.23% |
| simple-loop-unswitch.NumCostMultiplierSkipped | 914 | 503 | -411 | -44.97% | 44.97% |
| simple-loop-unswitch.NumSwitches | 20 | 18 | -2 | -10.00% | 10.00% |
| simple-loop-unswitch.NumTrivial | 183 | 95 | -88 | -48.09% | 48.09% |
... but that actually regresses LICM (-12% `licm.NumMovedLoads`),
loop-simplifycfg (`NumLoopExitsDeleted`, `NumTerminatorsFolded`),
simple-loop-unswitch (`NumTrivial`).
What if we instead have LICM both before and after LoopRotate?
| statistic name | LoopRotate-LICM | LICM-LoopRotate-LICM | Δ | % | abs(%) |
| asm-printer.EmittedInsts | 9015930 | 9014474 | -1456 | -0.02% | 0.02% |
| indvars.NumElimCmp | 3536 | 3546 | 10 | 0.28% | 0.28% |
| indvars.NumElimExt | 36725 | 36681 | -44 | -0.12% | 0.12% |
| indvars.NumElimIV | 1197 | 1185 | -12 | -1.00% | 1.00% |
| indvars.NumElimIdentity | 143 | 146 | 3 | 2.10% | 2.10% |
| indvars.NumElimRem | 4 | 5 | 1 | 25.00% | 25.00% |
| indvars.NumLFTR | 29842 | 29899 | 57 | 0.19% | 0.19% |
| indvars.NumReplaced | 2293 | 2299 | 6 | 0.26% | 0.26% |
| indvars.NumSimplifiedSDiv | 6 | 8 | 2 | 33.33% | 33.33% |
| indvars.NumWidened | 26438 | 26404 | -34 | -0.13% | 0.13% |
| instcount.TotalBlocks | 1178338 | 1173652 | -4686 | -0.40% | 0.40% |
| instcount.TotalFuncs | 111825 | 111829 | 4 | 0.00% | 0.00% |
| instcount.TotalInsts | 9905442 | 9895452 | -9990 | -0.10% | 0.10% |
| lcssa.NumLCSSA | 425871 | 425373 | -498 | -0.12% | 0.12% |
| licm.NumHoisted | 378357 | 383352 | 4995 | 1.32% | 1.32% |
| licm.NumMovedCalls | 2193 | 2204 | 11 | 0.50% | 0.50% |
| licm.NumMovedLoads | 35899 | 35755 | -144 | -0.40% | 0.40% |
| licm.NumPromoted | 11178 | 11163 | -15 | -0.13% | 0.13% |
| licm.NumSunk | 13359 | 14321 | 962 | 7.20% | 7.20% |
| loop-delete.NumDeleted | 8547 | 8538 | -9 | -0.11% | 0.11% |
| loop-instsimplify.NumSimplified | 12876 | 12041 | -835 | -6.48% | 6.48% |
| loop-peel.NumPeeled | 1008 | 924 | -84 | -8.33% | 8.33% |
| loop-rotate.NumNotRotatedDueToHeaderSize | 368 | 365 | -3 | -0.82% | 0.82% |
| loop-rotate.NumRotated | 42015 | 42005 | -10 | -0.02% | 0.02% |
| loop-simplifycfg.NumLoopBlocksDeleted | 240 | 241 | 1 | 0.42% | 0.42% |
| loop-simplifycfg.NumTerminatorsFolded | 618 | 619 | 1 | 0.16% | 0.16% |
| loop-unroll.NumCompletelyUnrolled | 11028 | 11029 | 1 | 0.01% | 0.01% |
| loop-unroll.NumUnrolled | 12608 | 12525 | -83 | -0.66% | 0.66% |
| mem2reg.NumPHIInsert | 192110 | 192073 | -37 | -0.02% | 0.02% |
| mem2reg.NumSingleStore | 637650 | 637652 | 2 | 0.00% | 0.00% |
| scalar-evolution.NumTripCountsComputed | 283108 | 282998 | -110 | -0.04% | 0.04% |
| scalar-evolution.NumTripCountsNotComputed | 106712 | 106691 | -21 | -0.02% | 0.02% |
| simple-loop-unswitch.NumBranches | 5178 | 5185 | 7 | 0.14% | 0.14% |
| simple-loop-unswitch.NumCostMultiplierSkipped | 914 | 925 | 11 | 1.20% | 1.20% |
| simple-loop-unswitch.NumTrivial | 183 | 179 | -4 | -2.19% | 2.19% |
| simple-loop-unswitch.NumBranches | 5178 | 4752 | -426 | -8.23% | 8.23% |
| simple-loop-unswitch.NumCostMultiplierSkipped | 914 | 503 | -411 | -44.97% | 44.97% |
| simple-loop-unswitch.NumSwitches | 20 | 18 | -2 | -10.00% | 10.00% |
| simple-loop-unswitch.NumTrivial | 183 | 95 | -88 | -48.09% | 48.09% |
I.e. we end up with less instructions, less peeling, more LICM activity,
also note how none of those 4 regressions are here. Namely:
| statistic name | LICM-LoopRotate | LICM-LoopRotate-LICM | Δ | % | abs(%) |
| asm-printer.EmittedInsts | 9015799 | 9014474 | -1325 | -0.01% | 0.01% |
| indvars.NumElimCmp | 3544 | 3546 | 2 | 0.06% | 0.06% |
| indvars.NumElimExt | 36580 | 36681 | 101 | 0.28% | 0.28% |
| indvars.NumElimIV | 1187 | 1185 | -2 | -0.17% | 0.17% |
| indvars.NumElimIdentity | 136 | 146 | 10 | 7.35% | 7.35% |
| indvars.NumLFTR | 29890 | 29899 | 9 | 0.03% | 0.03% |
| indvars.NumReplaced | 2227 | 2299 | 72 | 3.23% | 3.23% |
| indvars.NumWidened | 26329 | 26404 | 75 | 0.28% | 0.28% |
| instcount.TotalBlocks | 1173840 | 1173652 | -188 | -0.02% | 0.02% |
| instcount.TotalInsts | 9896139 | 9895452 | -687 | -0.01% | 0.01% |
| lcssa.NumLCSSA | 423961 | 425373 | 1412 | 0.33% | 0.33% |
| licm.NumHoisted | 378753 | 383352 | 4599 | 1.21% | 1.21% |
| licm.NumMovedCalls | 2208 | 2204 | -4 | -0.18% | 0.18% |
| licm.NumMovedLoads | 31821 | 35755 | 3934 | 12.36% | 12.36% |
| licm.NumPromoted | 11154 | 11163 | 9 | 0.08% | 0.08% |
| licm.NumSunk | 13587 | 14321 | 734 | 5.40% | 5.40% |
| loop-delete.NumDeleted | 8402 | 8538 | 136 | 1.62% | 1.62% |
| loop-instsimplify.NumSimplified | 11890 | 12041 | 151 | 1.27% | 1.27% |
| loop-peel.NumPeeled | 925 | 924 | -1 | -0.11% | 0.11% |
| loop-rotate.NumRotated | 42003 | 42005 | 2 | 0.00% | 0.00% |
| loop-simplifycfg.NumLoopBlocksDeleted | 242 | 241 | -1 | -0.41% | 0.41% |
| loop-simplifycfg.NumLoopExitsDeleted | 20 | 497 | 477 | 2385.00% | 2385.00% |
| loop-simplifycfg.NumTerminatorsFolded | 336 | 619 | 283 | 84.23% | 84.23% |
| loop-unroll.NumCompletelyUnrolled | 11032 | 11029 | -3 | -0.03% | 0.03% |
| loop-unroll.NumUnrolled | 12529 | 12525 | -4 | -0.03% | 0.03% |
| mem2reg.NumDeadAlloca | 10221 | 10222 | 1 | 0.01% | 0.01% |
| mem2reg.NumPHIInsert | 192106 | 192073 | -33 | -0.02% | 0.02% |
| mem2reg.NumSingleStore | 637643 | 637652 | 9 | 0.00% | 0.00% |
| scalar-evolution.NumBruteForceTripCountsComputed | 812 | 814 | 2 | 0.25% | 0.25% |
| scalar-evolution.NumTripCountsComputed | 282934 | 282998 | 64 | 0.02% | 0.02% |
| scalar-evolution.NumTripCountsNotComputed | 106718 | 106691 | -27 | -0.03% | 0.03% |
| simple-loop-unswitch.NumBranches | 4752 | 5185 | 433 | 9.11% | 9.11% |
| simple-loop-unswitch.NumCostMultiplierSkipped | 503 | 925 | 422 | 83.90% | 83.90% |
| simple-loop-unswitch.NumSwitches | 18 | 20 | 2 | 11.11% | 11.11% |
| simple-loop-unswitch.NumTrivial | 95 | 179 | 84 | 88.42% | 88.42% |
{
F15983613} {
F15983615} {
F15983616}
(this is vanilla llvm testsuite + rawspeed + darktable)
As an example of the code where early LICM only is bad, see:
https://godbolt.org/z/GzEbacs4K
This does have an observable compile-time regression of +~0.5% geomean
https://llvm-compile-time-tracker.com/compare.php?from=
7c5222e4d1a3a14f029e5f614c9aefd0fa505f1e&to=
5d81826c3411982ca26e46b9d0aff34c80577664&stat=instructions
but i think that's basically nothing, and there's potential that it might
be avoidable in the future by fixing clang to produce alignment information
on function arguments, thus making the second run unneeded.
Differential Revision: https://reviews.llvm.org/D99249
Vitaly Buka [Fri, 2 Apr 2021 07:58:09 +0000 (00:58 -0700)]
[NFC][scudo] Inline some functions into ScudoPrimaryTest
Vitaly Buka [Fri, 2 Apr 2021 07:17:45 +0000 (00:17 -0700)]
[NFC][scudo] Convert ScudoPrimaryTest into TYPED_TEST
Martin Storsjö [Wed, 31 Mar 2021 06:17:00 +0000 (09:17 +0300)]
[libcxx] [test] Fix invocable tests on Windows
MSVC had a bug regarding preferring intergral conversions over
floating conversions. This is fixed in MSVC 19.28 and newer. Clang in
MSVC mode so far only mimics the old, buggy behaviour, but will
hopefully soon be fixed to comply with the new behaviour too
(see https://reviews.llvm.org/D99663).
Make the negative test to use a distinctly different type,
leaving checks for compiler specific bugs out of the libcxx test.
Differential Revision: https://reviews.llvm.org/D99641
Martin Storsjö [Thu, 1 Apr 2021 20:53:36 +0000 (23:53 +0300)]
[libcxx] [test] Make the condvar wait_for tests a bit more understandable. NFC.
This was requested in the review of D99175; rename the "runs"
variable to clarify what it means wrt the test, and move updating of
it to the main function to clarify its behaviour wrt the two runs
further.
Differential Revision: https://reviews.llvm.org/D99768
oToToT [Thu, 1 Apr 2021 23:18:51 +0000 (07:18 +0800)]
[clang][ItaniumMangle] Check SizeExpr for DependentSizedArrayType
(PR49478)
As ArrayType::ArrayType mentioned in clang/lib/AST/Type.cpp, a
DependentSizedArrayType might not have size expression because it it
used as the type of a dependent array of unknown bound with a dependent
braced initializer.
Thus, I add a check when mangling array of that type.
This should fix https://bugs.llvm.org/show_bug.cgi?id=49478
Reviewed By: Richard Smith - zygoloid
Differential Revision: https://reviews.llvm.org/D99407
Alex Zinenko [Thu, 1 Apr 2021 12:19:56 +0000 (14:19 +0200)]
[mlir] add memref dialect as dependent of lower-affine pass
The lower-affine pass also processes affine load and store operations
that get converted to load and store operations now available in the
memref dialect. Since it produces operations from the memref dialect,
this dialect should be registered as dependent for this pass. It is rare
but possible to have code that doesn't have memref operations in the
input and calls this pass.
Reviewed By: mehdi_amini
Differential Revision: https://reviews.llvm.org/D99720
Marek Kurdej [Fri, 2 Apr 2021 06:57:42 +0000 (08:57 +0200)]
[clang-cl] [Sema] Do not prefer integral conversion over floating-to-integral for MS compatibility 19.28 and higher.
As of MSVC 19.28 (2019 Update 8), integral conversion is no longer preferred over floating-to-integral, and so MSVC is more standard conformant and will generate a compiler error on ambiguous call.
Cf. https://godbolt.org/z/E8xsdqKsb.
Initially found during the review of D99641.
Reviewed By: rnk
Differential Revision: https://reviews.llvm.org/D99663
Aaron Green [Fri, 2 Apr 2021 06:20:35 +0000 (23:20 -0700)]
Tweak SimpleFastHash
This change adds a SimpleFastHash64 variant of SimpleFastHash which allows call sites to specify a starting value and get a 64 bit hash in return. This allows a hash to be "resumed" with more data.
A later patch needs this to be able to hash a sequence of module-relative values one at a time, rather than just a region a memory.
Reviewed By: morehouse
Differential Revision: https://reviews.llvm.org/D94510
Christopher Di Bella [Wed, 31 Mar 2021 21:42:37 +0000 (21:42 +0000)]
[libcxx] adds concepts `std::totally_ordered` and `std::totally_ordered_with`
Implements parts of:
- P0898R3 Standard Library Concepts
- P1754 Rename concepts to standard_case for C++20, while we still can
Reviewed By: Mordante
Differential Revision: https://reviews.llvm.org/D98983
Wenlei He [Fri, 2 Apr 2021 05:04:40 +0000 (22:04 -0700)]
[CSSPGO] Skip dangling probe value when computing profile summary
Recently we switched to use InvalidProbeCount = UINT64_MAX (instead of 0) to represent dangling probe, but UINT64_MAX is not excluded when computing profile summary. This caused profile summary to produce incorrect hot/cold threshold. The change fixed it by excluding UINT64_MAX from summary builder.
Differential Revision: https://reviews.llvm.org/D99788
Craig Topper [Fri, 2 Apr 2021 03:48:18 +0000 (20:48 -0700)]
[RISCV] Add missing nxvXf64 intrinsics tests cases for floating-point compare for RV32.
Samuel [Fri, 2 Apr 2021 03:38:39 +0000 (20:38 -0700)]
[llvm-reduce] Add header guards and fix clang-tidy warnings
Add header guards and fix other clang-tidy warnings in .h files.
Also align misaligned header docs
Reviewed By: aeubanks
Differential Revision: https://reviews.llvm.org/D99634
Craig Topper [Fri, 2 Apr 2021 03:20:15 +0000 (20:20 -0700)]
[RISCV] Add more nxvi64 vector intrinsic tests for RV32. NFC
This confirms we handle most instrutions gracefully. We do
currently fail for vslide1up and vslide1down though.
Yang Fan [Fri, 2 Apr 2021 03:25:20 +0000 (11:25 +0800)]
[lld][MachO] Fix -Wsign-compare warning (NFC)
GCC warning:
```
/llvm-project/lld/MachO/InputFiles.cpp:484:24: warning: comparison of integer expressions of different signedness: ‘int64_t’ {aka ‘long int’} and ‘uint64_t’ {aka ‘long unsigned int’} [-Wsign-compare]
484 | return value < subsectionEntry.offset;
| ~~~~~~^~~~~~~~~~~~~~~~~~~~~~~~
```
Juneyoung Lee [Fri, 2 Apr 2021 02:53:45 +0000 (11:53 +0900)]
[AssumeBundles] offset should be added to correctly calculate align
This is a patch to fix the bug in alignment calculation (see https://reviews.llvm.org/D90529#2619492).
Consider this code:
```
call void @llvm.assume(i1 true) ["align"(i32* %a, i32 32, i32 28)]
%arrayidx = getelementptr inbounds i32, i32* %a, i64 -1
; aligment of %arrayidx?
```
The llvm.assume guarantees that `%a - 28` is 32-bytes aligned, meaning that `%a` is 32k + 28 for some k.
Therefore `a - 4` cannot be 32-bytes aligned but the existing code was calculating the pointer as 32-bytes aligned.
The reason why this happened is as follows.
`DiffSCEV` stores `%arrayidx - %a` which is -4.
`OffSCEV` stores the offset value of “align”, which is 28.
`DiffSCEV` + `OffSCEV` = 24 should be used for `a - 4`'s offset from 32k, but `DiffSCEV` - `OffSCEV` = 32 was being used instead.
Reviewed By: Tyker
Differential Revision: https://reviews.llvm.org/D98759
Petr Hosek [Fri, 2 Apr 2021 03:30:49 +0000 (20:30 -0700)]
[CMake] Use append instead of set with the list
This addresses an issue introduced by D99706.
Vitaly Buka [Fri, 2 Apr 2021 01:44:55 +0000 (18:44 -0700)]
[NFC][scudo] Move some shared stuff into ScudoCombinedTest
Yang Fan [Fri, 2 Apr 2021 03:03:08 +0000 (11:03 +0800)]
[lld] Add missing header guard (NFC)
Jonas Devlieghere [Fri, 2 Apr 2021 02:57:46 +0000 (19:57 -0700)]
[lldb] Account for objc_debug_class_getNameRaw returning NULL
On macOS Catalina, calling objc_debug_class_getNameRaw on some of the
ISA pointers returns NULL, causing us to crash and unwind before reading
all the Objective-C classes. This does not happen on macOS Big Sur.
Account for that possibility and skip the class when that happens.
Jason Molenda [Fri, 2 Apr 2021 01:59:36 +0000 (18:59 -0700)]
Handle all standalone combinations of LC_NOTEs w/ & w/o addr & uuid
Fill out ProcessMachCore::DoLoadCore to handle LC_NOTE hints with
a UUID or with a UUID+address, and load the binary at the specified
offset correctly. Add tests for all four combinations. Change
DynamicLoaderStatic to not re-set a Section's load address in the
Target if it's already been specified.
Differential Revision: https://reviews.llvm.org/D99571
rdar://
51490545
Yang Fan [Fri, 2 Apr 2021 01:29:18 +0000 (09:29 +0800)]
[X86] Fix -Wunused-function warning (NFC)
GCC warning:
```
/llvm-project/llvm/lib/Target/X86/X86ISelLowering.cpp:9212:13: warning: ‘bool isHorizOp(unsigned int)’ defined but not used [-Wunused-function]
9212 | static bool isHorizOp(unsigned Opcode) {
| ^~~~~~~~~
```
Vitaly Buka [Fri, 2 Apr 2021 01:33:49 +0000 (18:33 -0700)]
[NFC][scudo] Move globals into related test
Chen Zheng [Thu, 1 Apr 2021 04:50:49 +0000 (00:50 -0400)]
[debug-info][XCOFF] set `-gno-column-info` by default for DBX
For DBX, it does not handle column info well. Set -gno-column-info
by default for DBX.
Reviewed By: dblaikie
Differential Revision: https://reviews.llvm.org/D99703
Aart Bik [Thu, 1 Apr 2021 23:23:17 +0000 (16:23 -0700)]
[mlir][sparse] support for very narrow index and pointer types
Rationale:
Small indices and values, when allowed by the required range of the
input tensors, can reduce the memory footprint of sparse tensors
even more. Note, however, that we must be careful zero extending
the values (since sparse tensors never use negatives for indexing),
but LLVM treats the index type as signed in most memory operations
(like the scatter and gather). This CL dots all the i's in this regard.
Reviewed By: bixia
Differential Revision: https://reviews.llvm.org/D99777
Tony [Fri, 26 Mar 2021 02:01:18 +0000 (02:01 +0000)]
[NFC][AMDGPU] Add product names for gfx908 and gfx10 processors
Reviewed By: msearles
Differential Revision: https://reviews.llvm.org/D99781
Philip Reames [Fri, 2 Apr 2021 00:45:53 +0000 (17:45 -0700)]
[indvars[ Fix pr49802 by checking for SCEVCouldNotCompute
The code is assuming that having an exact exit count for the loop implies that exit counts for every exit are known. This used to be true, but when we added handling for dead exits we broke this invariant. The new invariant is that an exact loop count implies that any exits non trivially dead have exit counts.
We could have fixed this by either a) explicitly checking for a dead exit, or b) just testing for SCEVCouldNotCompute. I chose the second as it was simpler.
(Debugging this took longer than it should have since I'd mistyped the original assert and it wasn't checking what it was meant to...)
p.s. Sorry for the lack of test case. Getting things into a state to actually hit this is difficult and fragile. The original repro involves loop-deletion leaving SCEV in a slightly inprecise state which lets us bypass other transforms in IndVarSimplify on the way to this one. All of my attempts to separate it into a standalone test failed.
Alexander Shaposhnikov [Fri, 2 Apr 2021 00:48:09 +0000 (17:48 -0700)]
[lld][MachO] Make emitEndFunStab independent from .subsections_via_symbols
This diff addresses FIXME in SyntheticSections.cpp and removes
the dependency of emitEndFunStab on .subsections_via_symbols.
Test plan: make check-lld-macho
Differential revision: https://reviews.llvm.org/D99054
Vitaly Buka [Fri, 2 Apr 2021 00:31:52 +0000 (17:31 -0700)]
[NFC][scudo] Use TYPED_TEST to split large test
Daniel Rodríguez Troitiño [Fri, 2 Apr 2021 00:19:09 +0000 (17:19 -0700)]
[TextAPI] Add support for arm64_32
Add a new architecture definition for arm64_32. The change should allow
the new architecture arm64_32 to be recognized in several pieces of
code, TextAPI parsing one of them. llvm-lipo will also recognize the
architecture and will allow lipoing files with this architecture without
failing.
Includes a small test that the architecture is recognized by llvm-nm.
Reviewed By: cishida
Differential Revision: https://reviews.llvm.org/D99673
Daniel Rodríguez Troitiño [Fri, 2 Apr 2021 00:15:56 +0000 (17:15 -0700)]
[builtins] Build for arm64_32 for watchOS (Darwin)
Trying to build the builtins code fails because `arm64_32_SOURCES` is
missing. Setting it to the same list used for `aarch64_SOURCES` solves
that problem and allow the builtins to compile for that architecture.
Additionally, arm64_32 is added as a possible architecture for watchos
platforms.
Reviewed By: compnerd
Differential Revision: https://reviews.llvm.org/D99690
Craig Topper [Thu, 1 Apr 2021 22:58:46 +0000 (15:58 -0700)]
[RISCV] Add nxvXi64 test cases to the RV32 Zvamo intrinsic test files. NFC
peter klausler [Thu, 1 Apr 2021 23:27:15 +0000 (16:27 -0700)]
[flang] Disable some new unit tests (non-portable results)
Due to architectural variation on the C++ functions std::ceil, std::floor,
and std::trunc, diable some new Fortran unit tests for now that depending
on specifical results for IEEE floating-point edge cases of infinities
and NaNs.
Thomas Preud'homme [Thu, 1 Apr 2021 23:41:56 +0000 (00:41 +0100)]
[MIPS, test] Fix use of undef FileCheck var
LLVM test CodeGen/Mips/sr1.ll tries to check for the absence of a
sequence of instructions with several CHECK-NOT with one of those
directives using a variable defined in another. However CHECK-NOT are
checked independently so that is using a variable defined in a pattern
that should not occur in the input.
This commit removes the definition and uses of variable to check each
line independently, making the check stronger than the current one.
Reviewed By: dsanders
Differential Revision: https://reviews.llvm.org/D99776
Daniel Sanders [Thu, 1 Apr 2021 23:47:43 +0000 (16:47 -0700)]
Revert "[globalisel][unittests] Rename setUp() to avoid potential mix up with SetUp() from gtest"
Forgot to apply commit message changes from phabricator
This reverts commit
3a016e31ecef7eeb876b540c928a25a7c5d2e07a.
Daniel Sanders [Wed, 31 Mar 2021 23:06:46 +0000 (16:06 -0700)]
[globalisel][unittests] Rename setUp() to avoid potential mix up with SetUp() from gtest
Also, make it structurally required so it can't be forgotten and re-introduce
the bug that led to the rotten green tests.
Differential Revision: https://reviews.llvm.org/D99692
Thomas Preud'homme [Thu, 1 Apr 2021 23:25:37 +0000 (00:25 +0100)]
[OpenMP, test] Fix use of undef VAR_PRIV FileCheck var
Remove the CHECK-NOT directive referring to as-of-yet undefined VAR_PRIV
variable since the pattern of the following CHECK-NOT in the same
CHECK-NOT block covers a superset of the case caught by the first
CHECK-NOT.
Reviewed By: ABataev
Differential Revision: https://reviews.llvm.org/D99775
Thomas Preud'homme [Thu, 1 Apr 2021 21:31:06 +0000 (22:31 +0100)]
[OpenMP, test] Fix use of undef DECL FileCheck var
OpenMP test target_data_use_device_ptr_if_codegen contains a CHECK-NOT
directive using an undefined DECL FileCheck variable. It seems copied
from target_data_use_device_ptr_codegen where there's a CHECK for a load
that defined the variable. Since there is no corresponding load in this
testcase, the simplest is to simply forbid any store and get rid of the
variable altogether.
Reviewed By: ABataev
Differential Revision: https://reviews.llvm.org/D99771
Thomas Preud'homme [Thu, 1 Apr 2021 16:12:14 +0000 (17:12 +0100)]
[OpenMP, test] Fix uses of undef S*VAR FileCheck var
Fix the many cases of use of undefined SIVAR/SVAR/SFVAR in OpenMP
*private_codegen tests, due to a missing BLOCK directive to capture the
IR variable when it is declared. It also fixes a few typo in its use.
Reviewed By: MaskRay
Differential Revision: https://reviews.llvm.org/D99770
Mehdi Amini [Thu, 1 Apr 2021 23:34:03 +0000 (23:34 +0000)]
Setup OpBuilder to support detached block in loopUnrollByFactor (NFC)
Setting the builder from a block is looking up for a parent operation
to get a context, instead by setting up the builder with an explicit
context we can support invoking this helper in absence of a parent
operation.
peter klausler [Thu, 1 Apr 2021 22:59:56 +0000 (15:59 -0700)]
[flang] Fix unit test failure on POWER
A new unit test for the Fortran runtime needs to allow for some
architectural variation on Infinity and NaN edge cases of NINT().
cchen [Thu, 1 Apr 2021 23:07:12 +0000 (18:07 -0500)]
[OpenMP51] Accept `primary` as proc bind affinity policy in Clang
Reviewed By: ABataev
Differential Revision: https://reviews.llvm.org/D99622
peter klausler [Thu, 1 Apr 2021 19:59:59 +0000 (12:59 -0700)]
[flang] Implement numeric intrinsic functions in runtime
Adds APIs, implementations, and unit tests for AINT, ANINT,
CEILING, EXPONENT, FLOOR, FRACTION, MOD, MODULO, NEAREST, NINT,
RRSPACING, SCALE, SET_EXPONENT, & SPACING.
Differential Revision: https://reviews.llvm.org/D99764
Tom Stellard [Thu, 1 Apr 2021 04:35:04 +0000 (21:35 -0700)]
llvm-shlib: Create object libraries for each component and link against them
This makes it possible to build libLLVM.so without first creating a
static library for each component. In the case where only libLLVM.so is
built (i.e. ninja LLVM) this eliminates 150 linker jobs.
Reviewed By: stellaraccident
Differential Revision: https://reviews.llvm.org/D95727
Philip Reames [Thu, 1 Apr 2021 21:45:14 +0000 (14:45 -0700)]
[funcattrs] Respect nofree attribute on callsites (not just callee)
Fangrui Song [Thu, 1 Apr 2021 21:37:34 +0000 (14:37 -0700)]
[Driver] -nostdinc -nostdinc++: don't warn for -Wunused-command-line-argument
Craig Topper [Thu, 1 Apr 2021 21:07:04 +0000 (14:07 -0700)]
[RISCV] Add isel patterns to handle vrsub intrinsic with 2 vector operands.
This occurs when we type legalize an i64 scalar input on RV32. We
need to manually splat, which requires a vector input. Rather
than special case this in lowering just pattern match it.
Philip Reames [Thu, 1 Apr 2021 20:58:13 +0000 (13:58 -0700)]
[tests] Add tests for forthcoming funcattrs nosync inference improvement
These are basically all the attributor tests for the same attribute with some minor cleanup for readability and autogened.
Jian Cai [Thu, 1 Apr 2021 19:22:50 +0000 (12:22 -0700)]
Reland "Add support to -Wa,--version in clang""
This relands commit
3cc3c0f8352ec33ca2f2636f94cb1d85fc57ac16 with fixed
test cases, which was reverted by commit
bf2479c347c8ca88fefdb144d8bae0a7a4231e2a.
Louis Dionne [Tue, 2 Mar 2021 21:18:37 +0000 (16:18 -0500)]
[libc++][NFC] Increase readability of typeinfo comparison of ARM64
We wasted a good deal of time trying to figure out whether our implementation
was correct. In the end, it was, but it wasn't so easy to determine. This
patch dumbs down the implementation and improves the documentation to make
it easier to validate.
See https://lists.llvm.org/pipermail/libcxx-dev/2020-December/001060.html.
Differential Revision: https://reviews.llvm.org/D97802
Vitaly Buka [Thu, 1 Apr 2021 19:40:28 +0000 (12:40 -0700)]
[scudo][NFC] Make tests runs with --gtest_repeat=2
Reviewed By: cryptoad
Differential Revision: https://reviews.llvm.org/D99766
Vitaly Buka [Thu, 1 Apr 2021 19:44:31 +0000 (12:44 -0700)]
[Scudo] Fix SizeClassAllocatorLocalCache::drain
It leaved few blocks in PerClassArray[0].
Reviewed By: cryptoad
Differential Revision: https://reviews.llvm.org/D99763
David Green [Thu, 1 Apr 2021 20:21:40 +0000 (21:21 +0100)]
[ARM] Allow v6m runtime loop unrolling
This removes the restriction that only Thumb2 targets enable runtime
loop unrolling, allowing it for Thumb1 only cores as well. The existing
T2 heuristics are used (for the time being) to control when and how
unrolling is performed.
Differential Revision: https://reviews.llvm.org/D99588
Vitaly Buka [Thu, 1 Apr 2021 04:46:53 +0000 (21:46 -0700)]
[NFC][scudo] Simplify UseQuarantine initialization
peter klausler [Thu, 1 Apr 2021 19:54:22 +0000 (12:54 -0700)]
[flang] Fix arm clang build
The new source file flang/runtime/complex-reduction.c contains
a portability work-around that implicitly assumed that a recent
version of clang would be used; this patch changes the code and
should be portable to older clangs and any other C compilers that
don't support the standard CMPLXF/CMPLX/CMPLXL macros.
Joseph Huber [Wed, 31 Mar 2021 20:00:38 +0000 (16:00 -0400)]
[OpenMP] Pass mapping names to add components in a user defined mapper
Summary:
Currently the mapping names are not passed to the mapper components that set up
the array region. This means array mappings will not have their names availible
in the runtime. This patch fixes this by passing the argument name to the region
correctly. This means that the mapped variable's name will be the declared
mapper that placed it on the device.
Reviewed By: jdoerfert
Differential Revision: https://reviews.llvm.org/D99681
Craig Topper [Thu, 1 Apr 2021 18:47:11 +0000 (11:47 -0700)]
[RISCV] Use softPromoteHalf legalization for fp16 without Zfh rather than PromoteFloat.
The default legalization strategy is PromoteFloat which keeps
half in single precision format through multiple floating point
operations. Conversion to/from float is done at loads, stores,
bitcasts, and other places that care about the exact size being 16
bits.
This patches switches to the alternative method softPromoteHalf.
This aims to keep the type in 16-bit format between every operation.
So we promote to float and immediately round for any arithmetic
operation. This should be closer to the IR semantics since we
are rounding after each operation and not accumulating extra
precision across multiple operations. X86 is the only other
target that enables this today. See https://reviews.llvm.org/D73749
I had to update getRegisterTypeForCallingConv to force f16 to
use f32 when the F extension is enabled. This way we can still
pass it in the lower bits of an FPR for ilp32f and lp64f ABIs.
The softPromoteHalf would otherwise always give i16 as the
argument type.
Reviewed By: asb, frasercrmck
Differential Revision: https://reviews.llvm.org/D99148
Anastasia Stulova [Thu, 1 Apr 2021 19:31:00 +0000 (20:31 +0100)]
[OpenCL][Docs] Update links to the C++ for OpenCL documentation
Philip Reames [Thu, 1 Apr 2021 19:17:01 +0000 (12:17 -0700)]
Update a test missed in 6ef4505
Philip Reames [Thu, 1 Apr 2021 19:01:29 +0000 (12:01 -0700)]
[Attributor] Cleanup detection of non-relaxed atomics in nosync inference
The code was checking for cases which are disallowed by the verifier. Delete dead code and adjust style.
Philip Reames [Thu, 1 Apr 2021 18:48:19 +0000 (11:48 -0700)]
[Attributor] Cleanup intrinsic handling in nosync inference [mostly NFC]
Mostly stylistic adjustment, but the old code didn't handle the memcpy.inline intrinsic. By using the matcher class, we now do.
Martin Storsjö [Tue, 23 Mar 2021 11:09:26 +0000 (13:09 +0200)]
[libcxx] [test] Make the condvar wait_for tests less brittle
These seem to fail occasionally (they are marked as possibly requiring
a retry).
When doing a condvar wait_for(), it can wake up before the timeout
as a spurious wakeup. In these cases, the wait_for() method returns that
the timeout wasn't hit, and the test reruns another wait_for().
On Windows, it seems like the wait_for() operation often can end up
returning slightly before the intended deadline - when intending to
wait for 250 milliseconds, it can return after e.g. 235 milliseconds.
In these cases, the wait_for() doesn't indicate a timeout.
Previously, the test then reran a new wait_for() for a full 250
milliseconds each time. So for N consecutive wakeups slightly too early,
we'd wait for (N+1)*250 milliseconds. Now it only reruns wait_for() for
the remaining intended wait duration.
Differential Revision: https://reviews.llvm.org/D99175
Philip Reames [Thu, 1 Apr 2021 18:34:43 +0000 (11:34 -0700)]
[funcattrs] Infer nosync from readnone and non-convergent
This implements the most basic possible nosync inference. The choice of inference rule is taken from the comments in attributor and the discussion on the review of the change which introduced the nosync attribute (
0626367202c).
This is deliberately minimal. As noted in code comments, I do plan to add a more robust inference which actually scans the function IR directly, but a) I need to do some refactoring of the attributor code to use common interfaces, and b) I wanted to get something in. I also wanted to minimize the "interesting" analysis discussion since that's time intensive.
Context: This combines with existing nofree attribute inference to help prove dereferenceability in the ongoing deref-at-point semantics work.
Differential Revision: https://reviews.llvm.org/D99749
Greg Clayton [Wed, 17 Mar 2021 05:10:29 +0000 (22:10 -0700)]
Fix "image lookup --address" Summary results for inline functions.
Inline callstacks were being incorrectly displayed in the results of "image lookup --address". The deepest frame wasn't displaying the line table line entry, it was always showing the inline information's call file and line on the previous frame. This is now fixed and has tests to make sure it doesn't regress.
Differential Revision: https://reviews.llvm.org/D98761
Philip Reames [Thu, 1 Apr 2021 18:32:22 +0000 (11:32 -0700)]
Infer dereferenceability from malloc and friends
Hookup TLI when inferring object size from allocation calls. This allows the analysis to prove dereferenceability for known allocation functions (such as malloc/new/etc) in addition to those marked explicitly with the allocsize attribute.
This is a follow up to 0129cd5 now that the bug fixed by
e2c6621e6 is resolved.
As noted in the test, this relies on being able to prove that there is no free between allocation and context (e.g. hoist location). At the moment, this is handled conservatively. I'm working strengthening out ability to reason about no-free regions separately.
Differential Revision: https://reviews.llvm.org/D99737
Martin Storsjö [Tue, 30 Mar 2021 09:40:30 +0000 (12:40 +0300)]
[ARM] Remove an unused parameter in ARMWinCOFFObjectWriter. NFC.
This writer only ever operates on 32 bit arm code.
Differential Revision: https://reviews.llvm.org/D99575
Martin Storsjö [Tue, 23 Mar 2021 12:00:12 +0000 (14:00 +0200)]
[libcxx] [test] Remove XFAIL LIBCXX-WINDOWS-FIXME from time.clock.file/now.pass.cpp
This doesn't fail when _LIBCPP_HAS_NO_INT128 is defined consistently
in both CMAKE_CXX_FLAGS and LIBCXX_TEST_COMPILER_FLAGS; the XFAIL was
added based on early CI testruns where that flag was missing in
LIBCXX_TEST_COMPILER_FLAGS.
Differential Revision: https://reviews.llvm.org/D99705
Philip Reames [Thu, 1 Apr 2021 18:21:34 +0000 (11:21 -0700)]
Extract isVolatile helper on Instruction [NFCI]
We have this logic duplicated in several cases, none of which were exhaustive. Consolidate it in one place.
I don't believe this actually impacts behavior of the callers. I think they all filter their inputs such that their partial implementations were correct. If not, this might be fixing a cornercase bug.
peter klausler [Wed, 31 Mar 2021 16:14:08 +0000 (09:14 -0700)]
[flang] Implement reductions in the runtime
Add runtime APIs, implementations, and tests for ALL, ANY, COUNT,
MAXLOC, MAXVAL, MINLOC, MINVAL, PRODUCT, and SUM reduction
transformantional intrinsic functions for all relevant argument
and result types and kinds, both without DIM= arguments
(total reductions) and with (partial reductions).
Complex-valued reductions have their APIs in C so that
C's _Complex types can be used for their results.
Some infrastructure work was also necessary or noticed:
* Usage of "long double" in the compiler was cleaned up a
bit, and host dependences on x86 / MSVC have been isolated
in a new Common/long-double header.
* Character comparison has been exposed via an extern template
so that reductions could use it.
* Mappings from Fortran type category/kind to host C++ types
and vice versa have been isolated into runtime/cpp-type.h and
then used throughout the runtime as appropriate.
* The portable 128-bit integer package in Common/uint128.h
was generalized to support signed comparisons.
* Bugs in descriptor indexing code were fixed.
Differential Revision: https://reviews.llvm.org/D99666
Raphael Isemann [Thu, 1 Apr 2021 17:50:08 +0000 (19:50 +0200)]
[lldb] Prevent that LLDB randomly crashes in CommandLineParser::addOption by initializing LLVM's command line parser
Since quite a while Apple's LLDB fork (that contains the Swift debugging
support) is randomly crashing in `CommandLineParser::addOption` with an error
such as `CommandLine Error: Option 'h' registered more than once!`
The backtrace of the crashing thread is shown below. There are also usually many
other threads also performing similar clang::FrontendActions which are all
trying to generate (usually outdated) Clang modules which are used by Swift for
various reasons.
```
[ 6] LLDB`CommandLineParser::addOption(llvm::cl::Option*, llvm::cl::SubCommand*) + 856
[ 7] LLDB`CommandLineParser::addOption(llvm::cl::Option*, llvm::cl::SubCommand*) + 733
[ 8] LLDB`CommandLineParser::addOption(llvm::cl::Option*, bool) + 184
[ 9] LLDB`llvm::cl::ParseCommandLineOptions(...) [inlined] ::CommandLineParser::ParseCommandLineOptions(... + 1279
[ 9] LLDB`llvm::cl::ParseCommandLineOptions(...) + 497
[ 10] LLDB`setCommandLineOpts(clang::CodeGenOptions const&) + 416
[ 11] LLDB`EmitAssemblyHelper::EmitAssemblyWithNewPassManager(...) + 98
[ 12] LLDB`clang::EmitBackendOutput(...) + 4580
[ 13] LLDB`PCHContainerGenerator::HandleTranslationUnit(clang::ASTContext&) + 871
[ 14] LLDB`clang::MultiplexConsumer::HandleTranslationUnit(clang::ASTContext&) + 43
[ 15] LLDB`clang::ParseAST(clang::Sema&, bool, bool) + 579
[ 16] LLDB`clang::FrontendAction::Execute() + 74
[ 17] LLDB`clang::CompilerInstance::ExecuteAction(clang::FrontendAction&) + 1808
```
The underlying reason for the crash is that the CommandLine code in LLVM isn't
thread-safe and will never be thread-safe with its current architecture. The way
LLVM's CommandLine logic works is that all parts of the LLVM can provide command
line arguments by defining `cl::opt` global variables and their constructors
(which are invoked during static initialisation) register the variable in LLVM's
CommandLineParser (which is also just a global variable). At some later point
after static initialization we actually try to parse command line arguments and
we ask the CommandLineParser to parse our `argv`. The CommandLineParser then
lazily constructs it's internal parsing state in a non-thread-safe way (this is
where the crash happens), parses the provided command line and then goes back to
the respective `cl::opt` global variables and sets their values according to the
parse result.
As all of this is based on global state, this whole mechanism isn't thread-safe
so the only time to ever use it is when we know we only have one active thread
dealing with LLVM logic. That's why nearly all callers of
`llvm::cl::ParseCommandLineOptions` are at the top of the `main` function of the
some LLVM-based tool. One of the few exceptions to this rule is in the
`setCommandLineOpts` function in `BackendUtil.cpp` which is in our backtrace:
```
static void setCommandLineOpts(const CodeGenOptions &CodeGenOpts) {
SmallVector<const char *, 16> BackendArgs;
BackendArgs.push_back("clang"); // Fake program name.
if (!CodeGenOpts.DebugPass.empty()) {
BackendArgs.push_back("-debug-pass");
BackendArgs.push_back(CodeGenOpts.DebugPass.c_str());
}
if (!CodeGenOpts.LimitFloatPrecision.empty()) {
BackendArgs.push_back("-limit-float-precision");
BackendArgs.push_back(CodeGenOpts.LimitFloatPrecision.c_str());
}
BackendArgs.push_back(nullptr);
llvm::cl::ParseCommandLineOptions(BackendArgs.size() - 1,
BackendArgs.data());
}
```
This is trying to set `cl::opt` variables in the LLVM backend to their right
value as the passed via CodeGenOptions by invoking the CommandLine parser. As
this is just in some generic Clang CodeGen code (where we allow having multiple
threads) this is code is clearly wrong. If we're unlucky it either overwrites
the value of the global variables or it causes the CommandLine parser to crash.
So the next question is why is this only crashing in LLDB? The main reason seems
to be that easiest way to crash this code is to concurrently enter the initial
CommandLineParser construction where it tries to collect all the registered
`cl::opt` options and checks for sanity:
```
// If it's a DefaultOption, check to make sure it isn't already there.
if (O->isDefaultOption() &&
SC->OptionsMap.find(O->ArgStr) != SC->OptionsMap.end())
return;
// Add argument to the argument map!
if (!SC->OptionsMap.insert(std::make_pair(O->ArgStr, O)).second) {
errs() << ProgramName << ": CommandLine Error: Option '" << O->ArgStr
<< "' registered more than once!\n";
HadErrors = true;
}
```
The `OptionsMap` here is global variable and if we end up in this code with two
threads at once then two threads at the same time can register an option (such
as 'h') when they pass the first `if` and then we fail with the sanity check in
the second `if`.
After this sanity check and initial setup code the only remaining work is just
parsing the provided CommandLine which isn't thread-safe but at least doesn't
crash in all my attempts at breaking it (as it's usually just reading from the
already generated parser state but not further modifying it). The exception to
this is probably that once people actually specify the options in the code
snippet above we might run into some new interesting ways to crash everything.
To go back to why it's only affecting LLDB: Nearly all LLVM tools I could find
(even if they are using threads) seem to call the CommandLine parser at the
start so they all execute the initial parser setup at a point where there is
only one thread. So once the code above is executed they are mostly safe from
the sanity check crashes. We even have some shady code for the gtest `main` in
`TestMain.cpp` which is why this also doesn't affect unit tests.
The only exception to this rule is ... *drum roll* ... LLDB! it's not using that
CommandLine library for parsing options so it also never ends up calling it in
`main`. So when we end up in the `FrontendAction` code from the backtrace we are
already very deep in some LLDB logic and usually already have several threads.
In a situation where Swift decides to compile a large amount of Clang modules in
parallel we then end up entering this code via several threads. If several
threads reach this code at the same time we end up in the situation where the
sanity-checking code of CommandLine crashes. I have a very reliable way of
demonstrating the whole thing in D99650 (just run the unit test several times,
it usually crashes after 3-4 attempts).
We have several ways to fix this:
1. Make the whole CommandLine mechanism in LLVM thread-safe.
2. Get rid of `setCommandLineOpts` in `BackendUtil.cpp` and other callers of the
command line parsing in generic Clang code.
3. Initialise the CommandLine library in a safe point in LLDB.
Option 1 is just a lot of work and I'm not even sure where to start. The whole
mechanism is based on global variables and global state and this seems like a
humongous task.
Option 2 is probably the best thing we can do in the near future. There are only
two callers of the command line parser in generic Clang code. The one in
`BackendUtils.cpp` looks like it can be replaced with some reasonable
refactoring (as it only deals with two specific options). There is another one
in `ExecuteCompilerInvocation` which deals with forwarding the generic `-mllvm`
options to the backend which seems like it will just end up requiring us to do
Option 1.
Option 3 is what this patch is doing. We just parse some dummy command line
invocation in a point of the LLDB execution where we only have one thread that
is dealing with LLVM/Clang stuff. This way we are at least prevent the frequent
crashes for users as parsing the dummy command line invocation will set up the
initial parser state safely.
Fixes rdar://
70989856
Reviewed By: mib, JDevlieghere
Differential Revision: https://reviews.llvm.org/D99652
Jim Ingham [Wed, 31 Mar 2021 23:59:52 +0000 (16:59 -0700)]
Add support for fetching signed values from tagged pointers.
The ObjC runtime offers both signed & unsigned tagged pointer value
accessors to tagged pointer providers, but lldb's tagged pointer
code only implemented the unsigned one. This patch adds an
emulation of the signed one.
The motivation for doing this is that NSNumbers use the signed
accessor (they are always signed) and we need to follow that in our
summary provider or we will get incorrect values for negative
NSNumbers.
The data-formatter-objc test file had NSNumber examples (along with lots of other
goodies) but the NSNumber values weren't tested. So I also added
checks for those values to the test.
I also did a quick audit of the other types in that main.m file, and
it looks like pretty much all the other values are either intermediates
or are tested.
Differential Revision: https://reviews.llvm.org/D99694
Alexey Bataev [Thu, 1 Apr 2021 17:56:25 +0000 (10:56 -0700)]
[SLP]Test for min/max reductions bug, NFC.
Aden Grue [Thu, 1 Apr 2021 17:55:50 +0000 (10:55 -0700)]
Add a pattern to combine composed subview ops
Differential Revision: https://reviews.llvm.org/D99229