Sanjay Patel [Mon, 22 Nov 2021 19:33:06 +0000 (14:33 -0500)]
[InstCombine] regenerate test checks; NFC
Avoid phantom (cosmetic value naming) diffs in potential future patches.
Sanjay Patel [Mon, 22 Nov 2021 19:29:01 +0000 (14:29 -0500)]
[InstCombine] avoid 'tmp' usage in test files; NFC
The update script ( utils/update_test_checks.py ) warns against this
because it can conflict with the default FileCheck names given to
anonymous values in the IR.
Sanjay Patel [Sun, 21 Nov 2021 16:23:00 +0000 (11:23 -0500)]
[InstCombine] move/add tests for binops with sext operand; NFC
Ed Maste [Mon, 22 Nov 2021 16:26:55 +0000 (11:26 -0500)]
[Driver] correct typo in FreeBSD 14 test
The test specified amd64-unknown-freebsd40.0 rather than 14.0. Since
40 is greater than 14 the test (for behaviour new in FreeBSD 14) worked
despite the typo.
Fixes:
699d47472c3f
Reviewed by: dim (in D77776)
Mehdi Amini [Mon, 22 Nov 2021 19:23:18 +0000 (19:23 +0000)]
Revert "[mlir][Vector] Add a vblendps-based impl for transpose8x8 (both intrin and inline_asm)"
This reverts commit
a9e236bed835c58be381dadb973a1db0681e4795.
This broke the Windows build:
mlir\include\mlir/Dialect/X86Vector/Transforms.h(28): error C2061: syntax error: identifier 'uint'
Matt Morehouse [Mon, 22 Nov 2021 18:32:28 +0000 (10:32 -0800)]
[HWASan] Remove -Wa,-mrelax-relocations=no flag.
We no longer need it after https://reviews.llvm.org/D113220.
Reviewed By: vitalybuka
Differential Revision: https://reviews.llvm.org/D113221
Vy Nguyen [Fri, 19 Nov 2021 15:56:58 +0000 (10:56 -0500)]
[lld-macho] Don't replace local personality symbol with LazySymbol
Follup-up to D107533, where we replaced local syms with non-local.
It doesn't make sense to replace local symbol with lazy.
Differential Revision: https://reviews.llvm.org/D110040
Philip Reames [Mon, 22 Nov 2021 19:00:53 +0000 (11:00 -0800)]
[docs] Incorprate first round of feedback on D114325
This builds on the text landed in
73d52ee78, and incorporates the points that Renato brought up.
Differential Revision: https://reviews.llvm.org/D114325
Stanislav Mekhanoshin [Thu, 11 Nov 2021 20:08:27 +0000 (12:08 -0800)]
[InstCombine] (~(a | b) & c) | ~(c | (a ^ b)) -> ~((a | b) & (c | (b ^ a)))
Transform
```
(~(a | b) & c) | ~(c | (a ^ b)) -> ~((a | b) & (c | (b ^ a)))
```
And swapped case:
```
(a | ~(b & c)) & ~(a & (b ^ c)) --> ~(a | b) | (a ^ b ^ c)
```
```
----------------------------------------
define i3 @src(i3 %a, i3 %b, i3 %c) {
%0:
%or1 = or i3 %b, %c
%not1 = xor i3 %or1, 7
%and1 = and i3 %a, %not1
%xor1 = xor i3 %b, %c
%or2 = or i3 %xor1, %a
%not2 = xor i3 %or2, 7
%or3 = or i3 %and1, %not2
ret i3 %or3
}
=>
define i3 @tgt(i3 %a, i3 %b, i3 %c) {
%0:
%obc = or i3 %b, %c
%xbc = xor i3 %b, %c
%o = or i3 %a, %xbc
%and = and i3 %obc, %o
%r = xor i3 %and, 7
ret i3 %r
}
Transformation seems to be correct!
```
```
----------------------------------------
define i4 @src(i4 %a, i4 %b, i4 %c) {
%0:
%and1 = and i4 %b, %c
%not1 = xor i4 %and1, 15
%or1 = or i4 %not1, %a
%xor1 = xor i4 %b, %c
%and2 = and i4 %xor1, %a
%not2 = xor i4 %and2, 15
%and3 = and i4 %or1, %not2
ret i4 %and3
}
=>
define i4 @tgt(i4 %a, i4 %b, i4 %c) {
%0:
%xor1 = xor i4 %b, %c
%xor2 = xor i4 %xor1, %a
%or1 = or i4 %a, %b
%not1 = xor i4 %or1, 15
%or2 = or i4 %xor2, %not1
ret i4 %or2
}
Transformation seems to be correct!
```
Differential Revision: https://reviews.llvm.org/D112955
Christopher Di Bella [Mon, 22 Nov 2021 18:44:13 +0000 (18:44 +0000)]
[libcxx][NFC] adds var-const@ as the owner for the uninitialised algos
Christopher Di Bella [Fri, 19 Nov 2021 23:26:35 +0000 (23:26 +0000)]
[libcxx][NFC] adds status entry for ranges algorithms
Differential Revision: https://reviews.llvm.org/D114295
Louis Dionne [Mon, 22 Nov 2021 18:33:39 +0000 (13:33 -0500)]
[libc++][NFC] Reformat comment about D68480 support
Manoj Gupta [Sat, 20 Nov 2021 06:03:34 +0000 (22:03 -0800)]
libfuzzer: Disable broken tests for arm
libfuzzer was recently enabled for Arm32 in D112091.
A few tests apparently do not work with arm32 so disable them.
The list of tests was obtained from
https://lab.llvm.org/buildbot/#/builders/190/builds/513
Reviewed By: morehouse
Differential Revision: https://reviews.llvm.org/D114312
LLVM GN Syncbot [Mon, 22 Nov 2021 18:24:45 +0000 (18:24 +0000)]
[gn build] Port
d8380ad977e9
LLVM GN Syncbot [Mon, 22 Nov 2021 18:24:44 +0000 (18:24 +0000)]
[gn build] Port
344cef6695e9
Arthur O'Dwyer [Thu, 29 Jul 2021 02:04:18 +0000 (22:04 -0400)]
[libc++] [P1614] Implement [cmp.alg]'s std::{strong,weak,partial}_order.
This does not include `std::compare_*_fallback`; those are coming later.
There's still an open question of how to implement std::strong_order
for `long double`, which has 80 value bits and 48 padding bits on x86-64,
and which is presumably *not* IEEE 754-compliant on PPC64 and so on.
So that part is left unimplemented.
Differential Revision: https://reviews.llvm.org/D110738
Arthur O'Dwyer [Fri, 19 Nov 2021 19:58:51 +0000 (14:58 -0500)]
[libc++] Granularize the <random> header. NFCI.
Actually there's one functional change here, which is that users can
no longer depend on <random> to include all of C++20 <concepts>. That
inclusion is so new that we believe nobody should be depending on it
yet, even in the presence of Hyrum's Law. We keep the includes of <vector>,
<algorithm>, etc., so as not to break pre-C++20 Hyrum's Law users.
Differential Revision: https://reviews.llvm.org/D114281
Dmitry Vyukov [Mon, 22 Nov 2021 18:03:48 +0000 (19:03 +0100)]
Revert "tsan: new runtime (v3)"
Summary:
This reverts commit
1784fe0532a69ead17793bced060a9bf9d232027.
Broke some bots:
https://lab.llvm.org/buildbot#builders/57/builds/12365
http://green.lab.llvm.org/green/job/clang-stage1-RA/25658/
Reviewers: vitalybuka, melver
Subscribers:
Peter Klausler [Thu, 18 Nov 2021 19:48:42 +0000 (11:48 -0800)]
[flang] Add -fno-automatic, refine IsSaved()
This legacy option (available in other Fortran compilers with various
spellings) implies the SAVE attribute for local variables on subprograms
that are not explicitly RECURSIVE. The SAVE attribute essentially implies
static rather than stack storage. This was the default setting in Fortran
until surprisingly recently, so explicit SAVE statements & attributes
could be and often were omitted from older codes. Note that initialized
objects already have an implied SAVE attribute, and objects in COMMON
effectively do too, as data overlays are extinct; and since objects that are
expected to survive from one invocation of a procedure to the next in static
storage should probably be explicit initialized in the first place, so the
use cases for this option are somewhat rare, and all of them could be
handled with explicit SAVE statements or attributes.
This implicit SAVE attribute must not apply to automatic (in the Fortran sense)
local objects, whose sizes cannot be known at compilation time. To get the
semantics of IsSaved() right, the IsAutomatic() predicate was moved into
Evaluate/tools.cpp to allow for dynamic linking of the compiler. The
redundant predicate IsAutomatic() was noticed, removed, and its uses replaced.
GNU Fortran's spelling of the option (-fno-automatic) was added to
the clang-based driver and used for basic sanity testing.
Differential Revision: https://reviews.llvm.org/D114209
Florian Mayer [Sat, 20 Nov 2021 02:20:43 +0000 (18:20 -0800)]
[hwasan] fix arguments to symbolizer.
new versions do not accept -inlining of -functions (tested with 11 and
13).
Reviewed By: hctim
Differential Revision: https://reviews.llvm.org/D114303
Lei Zhang [Mon, 22 Nov 2021 17:28:39 +0000 (12:28 -0500)]
[mlir][vector] Fix TransferOpReduceRank for 0-D tensors
We cannot unconditionally generate memref.load ops for such cases;
need to check the source's type.
Reviewed By: nicolasvasilache
Differential Revision: https://reviews.llvm.org/D114376
Alex Zinenko [Mon, 22 Nov 2021 12:19:36 +0000 (13:19 +0100)]
[mlir] support recursive types in type conversion infra
MLIR supports recursive types but they could not be handled by the conversion
infrastructure directly as it would result in infinite recursion in
`convertType` for elemental types. Support this case by keeping the "call
stack" of nested type conversions in the TypeConverter class and by passing it
as an optional argument to the individual conversion callback. The callback can
then check if a specific type is present on the stack more than once to detect
and handle the recursive case.
This approach is preferred to the alternative approach of having a separate
callback dedicated to handling only the recursive case as the latter was
observed to introduce ~3% time overhead on a 50MB IR file even if it did not
contain recursive types.
This approach is also preferred to keeping a local stack in type converters
that need to handle recursive types as that would compose poorly in case of
out-of-tree or cross-project extensions.
Reviewed By: rriddle
Differential Revision: https://reviews.llvm.org/D113579
Louis Dionne [Fri, 19 Nov 2021 21:57:36 +0000 (16:57 -0500)]
[runtimes] Do not force -stdlib=libc++ on Apple platforms
It is pointless to specify -stdlib=libc++ on Apple platforms since
that is the default anyway.
Differential Revision: https://reviews.llvm.org/D114283
Peter Klausler [Thu, 18 Nov 2021 17:34:43 +0000 (09:34 -0800)]
[flang] Predefine unit 0 connected to stderr
This is a near-universal language extension; external unit 0
is preconnected to the standard error output.
Differential Revision: https://reviews.llvm.org/D114298
Louis Dionne [Fri, 19 Nov 2021 21:29:15 +0000 (16:29 -0500)]
[libc++] Remove uses of printf in some test support headers
In the test suite, we generally don't use printf or other reporting
utilities. It's not that it wouldn't be useful, it's just that some
platforms don't support IO.
Instead, we try to keep test cases small and self-contained so that
we can reasonably easily reproduce failures locally and debug them.
This patch removes printf in some of the last places in the test suite
that used it. The only remaining places are in a deque test and in the
filesystem tests. The filesystem tests are arguably fine to keep using
IO, since we're testing <filesystem>. The deque test will be handled
separately.
Differential Revision: https://reviews.llvm.org/D114282
Nikolas Klauser [Mon, 22 Nov 2021 02:20:15 +0000 (03:20 +0100)]
[libc++] [NFC] Mark P0858R0 as implemented
P0858R0 seems to be implemented. D93830 sets `__cpp_lib_string_view` and D80452 sets `__cpp_lib_array_constexpr`.
Reviewed By: Quuxplusone, Mordante, ldionne, #libc
Spies: libcxx-commits
Differential Revision: https://reviews.llvm.org/D114344
Nico Weber [Sun, 21 Nov 2021 03:38:54 +0000 (22:38 -0500)]
[asm] Merge EmitMSInlineAsmStr() and EmitGCCInlineAsmStr()
This basically reverts
1778831a3d1, which split them.
Since they were split 9 years ago, EmitGCCInlineAsmStr() grew a bunch of
features that usually weren't added to EmitMSInlineAsmStr(), and
that was usually a mistake. D71677, D113932, D114167 are all examples
of where things were backported to EmitMSInlineAsmStr().
The names were also not great. EmitMSInlineAsmStr() used to be called for `asm
inteldialect`, which clang produces for Microsoft-style __asm { ... } blocks as
well for GCC-style __asm__ / asm statements with -masm=intel. On the other hand,
EmitGCCInlineAsmStr() used to be called for `asm`, whic clang produces for
GCC-style __asm__ / asm statements with -masm=att (the default).
It's also less code (23 insertions, 188 deletions).
No behavior change.
Differential Revision: https://reviews.llvm.org/D114330
Nico Weber [Sun, 21 Nov 2021 03:04:26 +0000 (22:04 -0500)]
[asm] Allow labels as operands in intel asm syntax
This makes a line in llvm/test/CodeGen/X86/asm-block-labels.ll pass
with `asm inteldialect` too.
I don't know if this is something one can hit in practice with inline
asm. The test is from 2007 (
4646aa3e337aa) but in 2009 blockaddr was
introduced and e.g. `__asm__ __volatile__("brl %0" :: "X"(&&foo) : "memory");`
compiles to
call void asm sideeffect "brl $0", "X,..."(i8* blockaddress(@func, %1))
nowadays (thanks to jrtc27 for that example!).
(
6c4d255bf3d64 switched clang to blockaddress on an opt-in basis,
e4801f7844bb added docs for it,
31b132c0b781 added IR support.)
I half-heartedly tried to build clang 2.8 locally, but it didn't
just build. And 2.8 didn't have a prebuilt clang binary yet.
The motivation is to make EmitGCCInlineAsmStr() and EmitMSInlineAsmStr()
more alike, and maybe we should delete this code form EmitGCCInlineAsmStr()
instead. But since it's just 3 lines and it's reachable from LLVM IR,
let's do the safer thing for now.
Differential Revision: https://reviews.llvm.org/D114329
Mark de Wever [Fri, 19 Nov 2021 15:43:27 +0000 (16:43 +0100)]
[libc++] Improve CMake include directory search.
This patch has been tested in D70631, but it should be reviewed
separately.
Reviewed By: ldionne, #libc
Differential Revision: https://reviews.llvm.org/D114248
Gabor Marton [Thu, 11 Nov 2021 13:43:03 +0000 (14:43 +0100)]
[Analyzer][Core] Simplify IntSym in SValBuilder
Make the SimpleSValBuilder capable to simplify existing IntSym
expressions based on a newly added constraint on the sub-expression.
Differential Revision: https://reviews.llvm.org/D113754
Kazu Hirata [Mon, 22 Nov 2021 16:21:09 +0000 (08:21 -0800)]
Use std::string::substr (NFC)
Kazu Hirata [Mon, 22 Nov 2021 16:21:07 +0000 (08:21 -0800)]
[Target] Use range-based for loops (NFC)
Alexey Bataev [Mon, 22 Nov 2021 15:41:07 +0000 (07:41 -0800)]
[SLP][NFC]Add a test that reveals the problem in the emission of
vector int division with undefs.
Zarko Todorovski [Mon, 22 Nov 2021 14:39:21 +0000 (09:39 -0500)]
[NFC][llvm][Hexagon] Inclusive Terms remove uses of sanity in Hexagon taget
Most changes are rewording comments but there are some assertions that I rephrased.
Reviewed By: kparzysz
Differential Revision: https://reviews.llvm.org/D114132
Hsiangkai Wang [Tue, 16 Nov 2021 08:01:37 +0000 (16:01 +0800)]
[RISCV] Reverse the order of loading/storing callee-saved registers.
Currently, we restore the return address register as the last restoring
instruction in the epilog. The next instruction is `ret` usually. It is
a use of return address register. In some microarchitectures, there is
load-to-use data hazard. To avoid the load-to-use data hazard, we could
separate the load instruction from its use as far as possible. In this
patch, we reverse the order of restoring callee-saved registers to
increase the distance of `load ra` and `ret` in the epilog.
Differential Revision: https://reviews.llvm.org/D113967
Dmitry Vyukov [Tue, 27 Apr 2021 11:55:41 +0000 (13:55 +0200)]
tsan: new runtime (v3)
This change switches tsan to the new runtime which features:
- 2x smaller shadow memory (2x of app memory)
- faster fully vectorized race detection
- small fixed-size vector clocks (512b)
- fast vectorized vector clock operations
- unlimited number of alive threads/goroutimes
Depends on D112602.
Reviewed By: melver
Differential Revision: https://reviews.llvm.org/D112603
Dmitry Vyukov [Mon, 22 Nov 2021 07:22:01 +0000 (08:22 +0100)]
tsan: disable instrumentation in runtime callbacks in tests
All runtime callbacks must be non-instrumented with the new tsan runtime
(it's now more picky with respect to recursion into runtime).
Disable instrumentation in Darwin tests as we do in all other tests now.
Differential Revision: https://reviews.llvm.org/D114348
Nikita Popov [Mon, 22 Nov 2021 14:46:46 +0000 (15:46 +0100)]
Revert "[SCEV] Fix and validate ValueExprMap/ExprValueMap consistency"
This reverts commit
d633db8f9dd4a361e60a9030c82adc490d5797e3.
Causes bootstrap assertion failures:
https://lab.llvm.org/buildbot/#/builders/168/builds/3459/steps/9/logs/stdio
Guillaume Chatelet [Mon, 22 Nov 2021 14:31:56 +0000 (14:31 +0000)]
[libc] add memmove basic building blocks
Differential Revision: https://reviews.llvm.org/D113321
Arjun P [Mon, 22 Nov 2021 14:22:54 +0000 (19:52 +0530)]
[MLIR] PresburgerSetTest: fix comment and add a test case
Nikita Popov [Sat, 30 Oct 2021 19:40:14 +0000 (21:40 +0200)]
[SCEV] Fix and validate ValueExprMap/ExprValueMap consistency
This adds validation for consistency of ValueExprMap and
ExprValueMap, and fixes identified issues:
* Addrec construction directly wrote to ValueExprMap in a few places,
without updating ExprValueMap. Add a helper to ensures they stay
consistent. The adjustment in forgetSymbolicName() explicitly
drops the old value from the map, so that we don't rely on it
being overwritten.
* forgetMemoizedResultsImpl() was dropping the SCEV from
ExprValueMap, but not dropping the corresponding entries from
ValueExprMap.
Differential Revision: https://reviews.llvm.org/D113349
Pavel Labath [Thu, 18 Nov 2021 20:27:27 +0000 (21:27 +0100)]
[lldb] Fix [some] leaks in python bindings
Using an lldb_private object in the bindings involves three steps
- wrapping the object in it's lldb::SB variant
- using swig to convert/wrap that to a PyObject
- wrapping *that* in a lldb_private::python::PythonObject
Our SBTypeToSWIGWrapper was only handling the middle part. This doesn't
just result in increased boilerplate in the callers, but is also a
functionality problem, as it's very hard to get the lifetime of of all
of these objects right. Most of the callers are creating the SB object
(step 1) on the stack, which means that we end up with dangling python
objects after the function terminates. Most of the time this isn't a
problem, because the python code does not need to persist the objects.
However, there are legitimate cases where they can do it (and even if
the use case is not completely legitimate, crashing is not the best
response to that).
For this reason, some of our code creates the SB object on the heap, but
it has another problem -- it never gets cleaned up.
This patch begins to add a new function (ToSWIGWrapper), which does all
of the three steps, while properly taking care of ownership. In the
first step, I have converted most of the leaky code (except for
SBStructuredData, which needs a bit more work).
Differential Revision: https://reviews.llvm.org/D114259
Pavel Labath [Thu, 18 Nov 2021 12:52:44 +0000 (13:52 +0100)]
[lldb/test] Make it possible to run the mock gdb server on a single thread
This is a preparatory commit to enable mocking of qemu startup. That
will involve running the mock server in a separate process, so there's
no need for multithreading.
Initialization is moved from the start function into the constructor
(which can then take an actual socket instead of a class), and the run
method is made public.
Depends on D114156.
Differential Revision: https://reviews.llvm.org/D114157
Tobias Gysi [Mon, 22 Nov 2021 13:15:06 +0000 (13:15 +0000)]
[mlir][linalg] Use getAsOpFoldResult in padding (NFC).
After padding, we introduce a ExtractSliceOp to get the final unpadded result. This revision uses getAsOpFoldResult to compute the size of the unpadded result, which guarantees the result type has a partially static shape if some of the sizes of the unpadded result are statically known. At the moment, we rely on canonicalization to cleanup the types after padding.
Depends On D114085
Reviewed By: nicolasvasilache
Differential Revision: https://reviews.llvm.org/D114153
Tobias Gysi [Mon, 22 Nov 2021 12:49:23 +0000 (12:49 +0000)]
[mlir][linalg] Always generate an extract/insert slice pair when tiling output tensors.
Adapt tiling to always generate an extract/insert slice pair for output tensors even if the tensor is not tiled. Having an explicit extract/insert slice pair simplifies followup transformations such as padding and bufferization. In particular, it makes read and written iteration argument slices explicit.
Depends On D114067
Reviewed By: nicolasvasilache
Differential Revision: https://reviews.llvm.org/D114085
Tres Popp [Mon, 22 Nov 2021 09:37:42 +0000 (10:37 +0100)]
Rename MlirExecutionEngine lookup to lookupPacked
The purpose of the change is to make clear whether the user is
retrieving the original function or the wrapper function, in line with
the invoke commands. This new functionality is useful for users that
already have defined their own packed interface, so they do not want the
extra layer of indirection, or for users wanting to the look at the
resulting primary function rather than the wrapper function.
All locations, except the python bindings now have a `lookupPacked`
method that matches the original `lookup` functionality. `lookup`
still exists, but with new semantics.
- `lookup` returns the function with a given name. If `bool f(int,int)`
is compiled, `lookup` will return a reference to `bool(*f)(int,int)`.
- `lookupPacked` returns the packed wrapper of the function with the
given name. If `bool f(int,int)` is compiled, `lookupPacked` will return
`void(*mlir_f)(void**)`.
Differential Revision: https://reviews.llvm.org/D114352
Tobias Gysi [Mon, 22 Nov 2021 12:31:40 +0000 (12:31 +0000)]
[mlir][linalg] Remove tile and fuse test pass (NFC).
Remove the tile and fuse test pass that has been replaced by codegen strategy.
Depends On D114067
Reviewed By: nicolasvasilache
Differential Revision: https://reviews.llvm.org/D114068
Bradley Smith [Thu, 18 Nov 2021 17:03:05 +0000 (17:03 +0000)]
[AArch64][ARM] Add missing SVE/SVE2 features from Cortex-A710
Differential Revision: https://reviews.llvm.org/D114169
Simon Moll [Mon, 22 Nov 2021 11:58:12 +0000 (12:58 +0100)]
[DA][NFC] Update publication - add remarks
Update the reference publication for the SyncDependenceAnalysis and Divergence Analysis. Fix phrasing, formatting. Add comments on reducible loop limitation.
Reviewed By: sameerds
Differential Revision: https://reviews.llvm.org/D114146
Roman Lebedev [Mon, 22 Nov 2021 11:31:25 +0000 (14:31 +0300)]
[X86][TTI] Finish costmodel for AVX512BW's VPMOVM2[BW] / VPMOV[BW]2M instructions
Apparently my methodology was suboptimal, and not only did miss all the +VL tuples,
i also missed some plain tuples. I believe, this adds everything missing.
Indeed, these manual costmodels are just not okay long-term.
Reviewed By: RKSimon
Differential Revision: https://reviews.llvm.org/D114334
Roman Lebedev [Mon, 22 Nov 2021 11:31:18 +0000 (14:31 +0300)]
[X86][TTI] Costmodel for AVX512DQ's VPMOVM2[DQ] / VPMOV[DQ]2M instructions
Much like the VPMOVM2[BW] / VPMOV[BW]2M from AVX512BW,
these either sign-extent the mask register into a vector,
or pack the mask from vector register.
Apparently, we didn't even have MCA tests for these,
added in rG2f364f6f0d3a2420ca78cbd80abb186657180e05,
so i'm just guessing that their perf characteristics
are optimal.
Reviewed By: RKSimon
Differential Revision: https://reviews.llvm.org/D114314
Nicolas Vasilache [Mon, 22 Nov 2021 10:57:33 +0000 (10:57 +0000)]
[mlir] Add InitializeNativeTargetAsmParser to ExecutionEngine.
This is required to allow python to work with lowerings that use inline_asm.
Differential Revision: https://reviews.llvm.org/D114338
Tobias Gysi [Mon, 22 Nov 2021 10:56:34 +0000 (10:56 +0000)]
[mlir][linalg] Add a tile and fuse on tensors pattern.
Add a pattern to apply the new tile and fuse on tensors method. Integrate the pattern into the CodegenStrategy and use the CodegenStrategy to implement the tests.
Depends On D114012
Reviewed By: nicolasvasilache
Differential Revision: https://reviews.llvm.org/D114067
Diego Caballero [Mon, 22 Nov 2021 10:18:45 +0000 (10:18 +0000)]
[LV] Drop integer poison-generating flags from instructions that need predication
This patch fixes PR52111. The problem is that LV propagates poison-generating flags (`nuw`/`nsw`, `exact`
and `inbounds`) in instructions that contribute to the address computation of widen loads/stores that are
guarded by a condition. It may happen that when the code is vectorized and the control flow within the loop
is linearized, these flags may lead to generating a poison value that is effectively used as the base address
of the widen load/store. The fix drops all the integer poison-generating flags from instructions that
contribute to the address computation of a widen load/store whose original instruction was in a basic block
that needed predication and is not predicated after vectorization.
Reviewed By: fhahn, spatel, nlopes
Differential Revision: https://reviews.llvm.org/D111846
Nicolas Vasilache [Mon, 22 Nov 2021 08:52:40 +0000 (08:52 +0000)]
[mlir] Fix unintentional mutation by VectorType/RankedTensorType::Builder dropDim
Differential Revision: https://reviews.llvm.org/D113933
Tobias Gysi [Mon, 22 Nov 2021 10:17:53 +0000 (10:17 +0000)]
[mlir][linalg] Fix tile and fuse for outermost reduction.
Tile and fuse failed if the outermost tile loop is a reduction dimension. Add the necessary check to handle outermost reductions and introduce a test case to verify the change.
Reviewed By: nicolasvasilache
Differential Revision: https://reviews.llvm.org/D114012
Nicolas Vasilache [Mon, 22 Nov 2021 10:22:37 +0000 (10:22 +0000)]
[mlir][Vector] Add a vblendps-based impl for transpose8x8 (both intrin and inline_asm)
This revision follows up on the conversation titled:
```[llvm-dev] Understanding and controlling some of the AVX shuffle emission paths```
The revision adds a vblendps-based implementation for transpose8x8 and further distinguishes between and intrinsics and an inline_asm implementation.
This results in roughly 20% fewer cycles as reported by llvm-mca:
After this revision (intrinsic version, resolves to virtually identical assembly as per the llvm-dev discussion, no vblendps instruction is emitted):
```
Iterations: 100
Instructions: 5900
Total Cycles: 2415
Total uOps: 7300
Dispatch Width: 6
uOps Per Cycle: 3.02
IPC: 2.44
Block RThroughput: 24.0
Cycles with backend pressure increase [ 89.90% ]
Throughput Bottlenecks:
Resource Pressure [ 89.65% ]
- SKXPort1 [ 0.04% ]
- SKXPort2 [ 12.42% ]
- SKXPort3 [ 12.42% ]
- SKXPort5 [ 89.52% ]
Data Dependencies: [ 37.06% ]
- Register Dependencies [ 37.06% ]
- Memory Dependencies [ 0.00% ]
```
After this revision (inline_asm version, vblendps instructions are indeed emitted):
```
Iterations: 100
Instructions: 6300
Total Cycles: 2015
Total uOps: 7700
Dispatch Width: 6
uOps Per Cycle: 3.82
IPC: 3.13
Block RThroughput: 20.0
Cycles with backend pressure increase [ 83.47% ]
Throughput Bottlenecks:
Resource Pressure [ 83.18% ]
- SKXPort0 [ 14.49% ]
- SKXPort1 [ 14.54% ]
- SKXPort2 [ 19.70% ]
- SKXPort3 [ 19.70% ]
- SKXPort5 [ 83.03% ]
- SKXPort6 [ 14.49% ]
Data Dependencies: [ 39.75% ]
- Register Dependencies [ 39.75% ]
- Memory Dependencies [ 0.00% ]
```
An accessible copy of the conversation is available [here](https://gist.github.com/nicolasvasilache/
68c7f34012584b0e00f335bcb374ede0).
Reviewed By: ftynse, dcaballe
Differential Revision: https://reviews.llvm.org/D114335
Sjoerd Meijer [Thu, 18 Nov 2021 14:08:37 +0000 (14:08 +0000)]
[BPI] Look-up tables for non-loop branches. NFC.
This adds and uses look-up tables for non-loop branch probabilities, which have
have probabilities directly encoded into the tables for the different condition
codes. Compared to having this logic inlined in different functions, as it used
to be the case, I think this is compacter and thus also easier to check/cross
reference. This also adds a test for pointer heuristics that was missing.
Differential Revision: https://reviews.llvm.org/D114009
Arjun P [Sun, 21 Nov 2021 19:55:25 +0000 (01:25 +0530)]
[MLIR][NFC] Simplex: remove repeated words in comment
Diego Caballero [Mon, 22 Nov 2021 10:12:25 +0000 (10:12 +0000)]
[LV] Pre-commit test for D111846
Reviewed By: fhahn
Differential Revision: https://reviews.llvm.org/D112054
Guillaume Chatelet [Mon, 22 Nov 2021 10:12:46 +0000 (10:12 +0000)]
[libc] Remove unused variable
Manuel Klimek [Mon, 22 Nov 2021 08:07:57 +0000 (09:07 +0100)]
Fix various problems found by fuzzing.
1. IndexTokenSource::getNextToken cannot return nullptr; some code was
still written assuming it can; make getNextToken more resilient against
incorrect input and fix its call-sites.
2. Change various asserts that can happen due to user provided input to
conditionals in the code.
Salman Javed [Mon, 22 Nov 2021 09:49:49 +0000 (22:49 +1300)]
Add missing clang-tidy args in index.rst (NFC)
The RST docs have gone out of sync with the command-line args that the
clang-tidy program actually supports.
Kirill Bobyrev [Mon, 22 Nov 2021 09:44:21 +0000 (10:44 +0100)]
[clangd] IncludeCleaner: Mark possible expr resolutions as used
Fixes: https://github.com/clangd/clangd/issues/934
Reviewed By: sammccall
Differential Revision: https://reviews.llvm.org/D114287
David Green [Mon, 22 Nov 2021 08:11:35 +0000 (08:11 +0000)]
[AArch64] Sink splat shuffles to lane index intrinsics
This teaches AArch64TargetLowering::shouldSinkOperands to sink splat
shuffles to certain neon intrinsics, so that they can make use of the
lane variants of the instructions that are available.
Differential Revision: https://reviews.llvm.org/D112994
Salman Javed [Mon, 22 Nov 2021 08:06:08 +0000 (21:06 +1300)]
Fix nits in clang-tidy's documentation (NFC)
Add commas, articles, and conjunctions where missing.
Chuanqi Xu [Mon, 22 Nov 2021 07:53:51 +0000 (15:53 +0800)]
[C++20] [Coroutines] Warn for deprecated form 'for co_await'
The form 'for co_await' is part of CoroutineTS instead of C++20.
So if we detected the use of 'for co_await' in C++20, we should emit
a warning at least.
Dmitry Vyukov [Fri, 19 Nov 2021 15:51:30 +0000 (16:51 +0100)]
tsan: add another fork test
Add a fork test that models what happens on Mac
where fork calls malloc/free inside of our atfork
callbacks.
Reviewed By: vitalybuka, yln
Differential Revision: https://reviews.llvm.org/D114250
Igor Kudrin [Mon, 22 Nov 2021 07:19:07 +0000 (14:19 +0700)]
[ELF][NFC] Do not pass region name to expandMemoryRegion()
The name can be easily got on-site.
Differential Revision: https://reviews.llvm.org/D114228
wangpc [Mon, 22 Nov 2021 06:01:37 +0000 (14:01 +0800)]
[RISCV] Generate pseudo instruction li
Add an alias of `addi [x], zero, imm` to generate pseudo
instruction li, which makes assembly mush more readable.
For existed tests, users can update them by running script
`llvm/utils/update_llc_test_checks.py`.
Reviewed By: asb
Differential Revision: https://reviews.llvm.org/D112692
Kazu Hirata [Mon, 22 Nov 2021 03:24:17 +0000 (19:24 -0800)]
[llvm] Use make_early_inc_range (NFC)
Kazu Hirata [Mon, 22 Nov 2021 03:24:15 +0000 (19:24 -0800)]
[llvm] Use range-based for loops (NFC)
Roland McGrath [Mon, 22 Nov 2021 02:14:30 +0000 (18:14 -0800)]
NFC: clang-format lib/Transforms/Instrumentation/InstrProfiling.cpp
Differential Revision: https://reviews.llvm.org/D114343
Joe Loser [Sun, 21 Nov 2021 00:13:18 +0000 (19:13 -0500)]
[libc++][NFC] Sort includes in __ranges/concepts.h
Differential Revision: https://reviews.llvm.org/D114328
LLVM GN Syncbot [Mon, 22 Nov 2021 00:29:19 +0000 (00:29 +0000)]
[gn build] Port
1dc62f2653f8
Nikolas Klauser [Sun, 21 Nov 2021 23:22:55 +0000 (00:22 +0100)]
[libc++] Implement P1272R4 (std::byteswap)
Implement P1274R4
Reviewed By: Quuxplusone, Mordante, #libc
Spies: jloser, lebedev.ri, mgorny, libcxx-commits, arichardson
Differential Revision: https://reviews.llvm.org/D114074
Jacques Pienaar [Sun, 21 Nov 2021 23:06:08 +0000 (15:06 -0800)]
[mlir] Fix unused function warning (NFC)
Delete function no longer needed as all derived classes override
printer.
Jacques Pienaar [Sun, 21 Nov 2021 22:41:11 +0000 (14:41 -0800)]
[mlir] Move trait to InferTypeOpInterface
Step towards removing the hard coded behavior for this trait and to instead use common interface.
Differential Revision: https://reviews.llvm.org/D114208
Kazu Hirata [Sun, 21 Nov 2021 18:36:20 +0000 (10:36 -0800)]
[CodeGen] Use llvm::is_contained (NFC)
Kazu Hirata [Sun, 21 Nov 2021 18:36:18 +0000 (10:36 -0800)]
[llvm] Use range-based for loops (NFC)
Simon Pilgrim [Sun, 21 Nov 2021 18:33:05 +0000 (18:33 +0000)]
[ARM] Regenerate sxt_rot.ll tests
Simon Pilgrim [Sun, 21 Nov 2021 18:32:10 +0000 (18:32 +0000)]
[Thumb2] Regenerate ext + rot tests
Simon Pilgrim [Sun, 21 Nov 2021 18:30:58 +0000 (18:30 +0000)]
[PowerPC] Regenerate rlwinm2.ll test
Philip Reames [Sun, 21 Nov 2021 16:00:34 +0000 (08:00 -0800)]
Add a best practice section on how to configure a fast builder
This is based on conversations with a couple of folks currently running buildbots. There's a couple pieces which didn't make it in, but this tries to cover the common themes.
Differential Revision: https://reviews.llvm.org/D114325
Arjun P [Sun, 21 Nov 2021 13:53:15 +0000 (19:23 +0530)]
[MLIR][NFC] Simplex::restoreRow: improve documentation
Simon Pilgrim [Sun, 21 Nov 2021 12:01:44 +0000 (12:01 +0000)]
[ARM][ParallelDSP] Regenerate complex_dot_prod.ll test
David Green [Sun, 21 Nov 2021 11:46:34 +0000 (11:46 +0000)]
[AArch64] Extra testing for sinking splats to various instructions. NFC
Fangrui Song [Sun, 21 Nov 2021 06:18:09 +0000 (22:18 -0800)]
[ELF] Move getOutputSectionName from Writer.cpp to LinkerScript.cpp. NFC
and internalize it.
Kazu Hirata [Sun, 21 Nov 2021 02:42:10 +0000 (18:42 -0800)]
[llvm] Use range-based for loops (NFC)
Phoebe Wang [Sun, 21 Nov 2021 01:12:46 +0000 (09:12 +0800)]
[X86][FP16] Relax the pattern condition for VZEXT_MOVL to match more cases
Fixes pr52560
Reviewed By: RKSimon
Differential Revision: https://reviews.llvm.org/D114313
Joe Loser [Sun, 21 Nov 2021 00:15:00 +0000 (19:15 -0500)]
[libc++][NFC] Fix typo in ranges::iterator_t synopsis
The `iterator_t` alias template is on `T` not a `R` like the other
neighboring alias templates. Fix the typo.
Arthur O'Dwyer [Thu, 14 Oct 2021 20:49:58 +0000 (16:49 -0400)]
[libc++] [doc] Mark some spaceship-related LWG issues as "Complete."
LWG3330 has been "Completed" since D99309, which was in the 13.x timeframe.
Reviewed as part of D110738.
Roman Lebedev [Sat, 20 Nov 2021 22:11:05 +0000 (01:11 +0300)]
[NFC][X86][Costmodel] Actually test +prefer-256-bit in replication-shuffle-related tests :(
While -prefer-256-bit indeed becomes complete with D114314,
the real-world (the one with +prefer-256-bit) coverage is lacking.
Hilarious.
Nikita Popov [Sat, 20 Nov 2021 22:17:41 +0000 (23:17 +0100)]
[DSE] Drop hasAnalyzableMemoryWrite() (NFCI)
The functionality of hasAnalyzableMemoryWrite() is effectively
subsumed by getLocForWriteEx(), which will return None if the
instruction is not analyzable. The implementations don't match
exactly (e.g. getLocForWriteEx() does not limit non-calls to
stores), but in conjunction with the isRemovable() check, it ends
up being the same.
Felix Berger [Fri, 19 Nov 2021 01:33:22 +0000 (20:33 -0500)]
[clang-tidy] performance-unnecessary-copy-initialization: Correctly match the type name of the thisPointertype.
The matching did not work correctly for pointer and reference types.
Differential Revision: https://reviews.llvm.org/D114212
Reviewed-by: courbet
Nikita Popov [Sat, 20 Nov 2021 20:06:08 +0000 (21:06 +0100)]
[LVI] Drop requirement that modulus is constant
If we're looking only at the lower bound, the actual modulus
doesn't matter. This is a leftover from when I wanted to consider
the upper bound as well, where the modulus does matter.
Nikita Popov [Sat, 20 Nov 2021 18:03:45 +0000 (19:03 +0100)]
[LVI] Support urem in implied conditions
If (X urem M) >= C we know that X >= C. Make use of this fact
when computing the implied condition range.
In some cases we could also establish an upper bound, but that's
both tricker and not interesting in practice.
Alive: https://alive2.llvm.org/ce/z/R5ZGSW
Nikita Popov [Sat, 20 Nov 2021 19:48:56 +0000 (20:48 +0100)]
[CVP] Add tests for implied conditions using urem (NFC)
Florian Hahn [Sat, 20 Nov 2021 17:59:47 +0000 (17:59 +0000)]
[VPlan] Wrap vector loop blocks in region.
A first step towards modeling preheader and exit blocks in VPlan as well.
Keeping the vector loop in a region allows for changing the VF as we
traverse region boundaries.
Reviewed By: Ayal
Differential Revision: https://reviews.llvm.org/D113182
Sanjay Patel [Sat, 20 Nov 2021 15:55:41 +0000 (10:55 -0500)]
[InstCombine] add folds for binop with sexted bool and constant operands
This is a generalization/extension of the existing and/or
folds noted with TODO comments. Those have a one-use
constraint that is not necessary.
Potential follow-ups are noted by the TODO comments in
the new function. We can also call this function from
other binop visit* functions, but we need to add tests
first.
This solves:
https://llvm.org/PR52543
https://alive2.llvm.org/ce/z/NWuCR5