The inlined llvm.coro.id should contain the function it refers to.
The modifed test would caused the compiler crash under O2. See
issue52912 for example.
This diff adds support to printing a Value when it is null. We encounter this situation when debugging the PDL bytcode execution (where a null Value is perfectly valid). Currently, the AsmPrinter crashes (with an assert in a cast) when it encounters such Value.
We follow the same format used in other printed entities (e.g., null attribute).
Reviewed By: mehdi_amini, bondhugula
Differential Revision: https://reviews.llvm.org/D116084
Presently the result type verification checks if the type is used by a `pdl::OperationOp` inside the matcher. This is unnecessarily restrictive; the type could come from a `pdl::OperandOp or `pdl::OperandsOp` and still be inferrable.
Reviewed By: rriddle, Mogball
Differential Revision: https://reviews.llvm.org/D116083
This diff adds an integration test to multi-root PDL matching. It consists of two subtests:
1) A 1-layer perceptron with split forward / backward operations.
2) A 2-layer perceptron with fused forward / backward operations.
These tests use a collection of hand-written patterns and TensorFlow operations to be matched. The first test has a DAG / SSA dominant resulting match; the second does not and is therefore stored in a graph region.
This diff also includes two bug fixes:
1) Mark the pdl_interp dialect as a dependent in the TestPDLByteCodePass. This is needed, because we create ops from that dialect as a part of the PDL-to-PDLInterp lowering.
2) Fix of the starting index in the liveness range for the ForEach operations (bug exposed by the integration test).
Reviewed By: Mogball
Differential Revision: https://reviews.llvm.org/D116082
The tree merging of pattern predicates places the predicates in an unordered set. When the predicates are sorted, they are taken in the set order, not the insertion order. This results in nondeterministic behavior.
One solution to this problem would be to use `SetVector`. However, the value `SetVector` does not provide a `find` function for fast O(1) lookups and stores the predicates twice -- once in the set and once in the vector, which is undesirable, because we store patternToAnswer in each predicate. A simpler solution is to store the tie breaking ID (which follows the insertion order), and use this ID to break any ties when comparing predicates.
Reviewed By: Mogball
Differential Revision: https://reviews.llvm.org/D116081
When the original version of multi-root patterns was reviewed, several improvements were made to the pdl_interp operations during the review process. Specifically, the "get users of a value at the specified operand index" was split up into "get users" and "compare the users' operands with that value". The iterative execution was also cleaned up to `pdl_interp.foreach`. However, the positions in the pdl-to-pdl_interp lowering were not similarly refactored. This introduced several problems, including hard-to-detect bugs in the lowering and duplicate evaluation of `pdl_interp.get_users`.
This diff cleans up the positions. The "upward" `OperationPosition` was split-out into `UsersPosition` and `ForEachPosition`, and the operand comparison was replaced with a simple predicate. In the process, I fixed three bugs:
1. When multiple roots were had the same connector (i.e., a node that they shared with a subtree at the previously visited root), we would generate a single foreach loop rather than one foreach loop for each such root. The reason for this is that such connectors shared the position. The solution for this is to add root index as an id to the newly introduced `ForEachPosition`.
2. Previously, we would use `pdl_interp.get_operands` indiscriminately, whether or not the operand was variadic. We now correctly detect variadic operands and insert `pdl_interp.get_operand` when needed.
3. In certain corner cases, we would trigger the "connector has not been traversed yet" assertion. This was caused by not inserting the values during the upward traversal correctly. This has now been fixed.
Reviewed By: Mogball
Differential Revision: https://reviews.llvm.org/D116080
The result value of a llvm.invoke operation is currently not mapped to the corresponding llvm::Value* when exporting to LLVM IR. This leads to any later operations using the result to crash as it receives a nullptr.
Differential Revision: https://reviews.llvm.org/D116564
* Give I[1] and I[-1] a name:
- Easier to understand
- Easier to debug (since you don't go through operator[] everytime)
* TheLine->First != TheLine->Last follows since last is a l brace and
first isn't.
* Factor the check for is(tok::l_brace) out.
* Drop else after return.
Differential Revision: https://reviews.llvm.org/D115060
In the post commit discussion of https://reviews.llvm.org/D116484 it was concluded that `--no-as-needed` should not be ignored. `--as-needed` stays ignored as it is already the default behaviour on COFF, which cannot be changed.
In D116472 we created conditionally defined variables for the tools to
unbreak the legacy build where they are in `llvm/tools`.
The runtimes are not tools, so that flexibility doesn't matter. Still,
it might be nice to define (unconditionally) and use the variable for
the runtimes simply to make the code a bit clearer and document what is
going on.
Also, consistently put project dirs at the beginning, not end of `CMAKE_MODULE_PATH`. This ensures they will properly shadow similarly named stuff that happens to be later on the path.
Reviewed By: mstorsjo, #libunwind, #libc, #libc_abi, ldionne
Differential Revision: https://reviews.llvm.org/D116477
This commit resolves GitHub issue #45895 (Bugzilla #46550), to
add or remove empty line between definition blocks including
namespaces, classes, structs, enums and functions.
Reviewed By: MyDeveloperDay, curdeius, HazardyKnusperkeks
Differential Revision: https://reviews.llvm.org/D116314
My team has a vendetta against lines ending with an open parenthesis, thought it might be useful for others too 😊
Reviewed By: HazardyKnusperkeks, curdeius
Differential Revision: https://reviews.llvm.org/D116170
This reverts commit 9bd22595bad36cd19f5e7ae18ccd9f41cba29dc5.
Seeing some bot failures which look plausibly connected. Revert while investigating/waiting for bots to stablize.
e.g. https://lab.llvm.org/buildbot#builders/36/builds/15933
This function returns an upper bound on the number of bits needed
to represent the signed value. Use "Max" to match similar functions
in KnownBits like countMaxActiveBits.
Rename APInt::getMinSignedBits->getSignificantBits. Keeping the old
name around to keep this patch size down. Will do a bulk rename as
follow up.
Rename KnownBits::countMaxSignedBits->countMaxSignificantBits.
Reviewed By: lebedev.ri, RKSimon, spatel
Differential Revision: https://reviews.llvm.org/D116522
This reverts commit fd4808887ee47f3ec8a030e9211169ef4fb094c3.
This patch causes gcc to issue a lot of warnings like:
warning: base class ‘class llvm::MCParsedAsmOperand’ should be
explicitly initialized in the copy constructor [-Wextra]
Implementation is based on the "expected type" as used for
designated-initializers in braced init lists. This means it can deduce the type
in some cases where it's not written:
void foo(Widget);
foo({ /*help here*/ });
Only basic constructor calls are in scope of this patch, excluded are:
- aggregate initialization (no help is offered for aggregates)
- initializer_list initialization (no help is offered for these constructors)
Fixes https://github.com/clangd/clangd/issues/306
Differential Revision: https://reviews.llvm.org/D116317
MOV64ri results in a significantly longer encoding, and use of this
operator is fairly avoidable as we can always check the size of the
immediate we're using.
This is an updated version of D99045.
Co-authored-by: Simonas Kazlauskas <git@kazlauskas.me>
Reviewed By: craig.topper
Differential Revision: https://reviews.llvm.org/D116458
There is some similar looking code in `TokenAnnotator.cpp` but given that I've
never worked on clang-format before I don't know what the purpose of that code
is and how it's related to `UnwrappedLineParser.cpp`.
Either way, it fixes clang-format with `BraceWrapping.AfterEnum=true` and
`AllowShortEnumsOnASingleLine=false` to behave like the documentation says.
Before this patch:
```
enum
{
A,
B
} myEnum;
```
After this patch:
```
enum {
A,
B
} myEnum;
```
According to the unittests which I had to modify this would change the LLVM
style. Please evaluate if you want to change the defaults or if you consider
the current style a bug.
Reviewed By: curdeius, HazardyKnusperkeks
Differential Revision: https://reviews.llvm.org/D106349
In that revision, I make LLD match Clang in deprecating `llvm-config`.
This patch isn't to worthwhile on its own --- there isn't a sense in
which the new order is "better" in isolation --- but by putting the
steps that LLD also neeeds to do first, I make the diff between LLD and
Clang's top-level `CMakeLists.txt` very legible.
Longer term I hope:
1. We can remove calling `llvm-config` altogether, and just go strait to
finding the CMake config file. This is what Flang does, at least.
2. Hopefully the diffable part is smaller then --- i.e. there is less
duplicated boilerplate.
3. Any duplicate boilerplate that remains can be factored out.
I didn't both trying to factor anything out in e.g. the top level common
CMake Utility modules because this deprecated-but-not-removed state is a
merely transitional.
Reviewed By: beanz
Differential Revision: https://reviews.llvm.org/D116548
As Visual Studio's CMake support is getting better and better the line
between IDE generator and non-IDE generators is blurring. Visual Studio
2019 and later have a very useful UI that can handle all of the various
targets we create, but if they are unsorted it is wildly unwieldy.
This change sorts the lit testsuite targets and per-component install
targets into folders, which are not generated for IDE generators but
are generated by default under Visual Studio's CMake + Ninja
integration.
If we have an exit which is controlled by a loop invariant condition and which dominates the latch, we know only the copy in the first unrolled iteration can be taken. All other copies are dead.
The change itself is pretty straight forward, but let me add two points of context:
* I'd have expected other transform passes to catch this after unrolling, but I'm seeing multiple examples where we get to the end of O2/O3 without simplifying.
* I'd like to do a stronger change which did CSE during unroll and accounted for invariant expressions (as defined by SCEV instead of trivial ones from LoopInfo), but that doesn't fit cleanly into the current code structure.
Differential Revision: https://reviews.llvm.org/D116496
There is an ongoing CI outage with our Linux nodes, so I temporarily
set up a couple of nodes. These nodes will be much slower than the
usual ones and there's only a few of them, so I am temporarily disabling
most of our CI to keep things working.
Segmentation fault in ompt_tsan_dependences function due to an unchecked NULL pointer dereference is as follows:
```
ThreadSanitizer:DEADLYSIGNAL
==140865==ERROR: ThreadSanitizer: SEGV on unknown address 0x000000000050 (pc 0x7f217c2d3652 bp 0x7ffe8cfc7e00 sp 0x7ffe8cfc7d90 T140865)
==140865==The signal is caused by a READ memory access.
==140865==Hint: address points to the zero page.
/usr/bin/addr2line: DWARF error: could not find variable specification at offset 1012a
/usr/bin/addr2line: DWARF error: could not find variable specification at offset 133b5
/usr/bin/addr2line: DWARF error: could not find variable specification at offset 1371a
/usr/bin/addr2line: DWARF error: could not find variable specification at offset 13a58
#0 ompt_tsan_dependences(ompt_data_t*, ompt_dependence_t const*, int) /ptmp/bhararit/llvm-project/openmp/tools/archer/ompt-tsan.cpp:1004 (libarcher.so+0x15652)
#1 __kmpc_doacross_post /ptmp/bhararit/llvm-project/openmp/runtime/src/kmp_csupport.cpp:4280 (libomp.so+0x74d98)
#2 .omp_outlined. for_ordered_01.c:? (for_ordered_01.exe+0x5186cb)
#3 __kmp_invoke_microtask /ptmp/bhararit/llvm-project/openmp/runtime/src/z_Linux_asm.S:1166 (libomp.so+0x14e592)
#4 __kmp_invoke_task_func /ptmp/bhararit/llvm-project/openmp/runtime/src/kmp_runtime.cpp:7556 (libomp.so+0x909ad)
#5 __kmp_fork_call /ptmp/bhararit/llvm-project/openmp/runtime/src/kmp_runtime.cpp:2284 (libomp.so+0x8461a)
#6 __kmpc_fork_call /ptmp/bhararit/llvm-project/openmp/runtime/src/kmp_csupport.cpp:308 (libomp.so+0x6db55)
#7 main ??:? (for_ordered_01.exe+0x51828f)
#8 __libc_start_main ??:? (libc.so.6+0x24349)
#9 _start /home/abuild/rpmbuild/BUILD/glibc-2.26/csu/../sysdeps/x86_64/start.S:120 (for_ordered_01.exe+0x4214e9)
ThreadSanitizer can not provide additional info.
SUMMARY: ThreadSanitizer: SEGV /ptmp/bhararit/llvm-project/openmp/tools/archer/ompt-tsan.cpp:1004 in ompt_tsan_dependences(ompt_data_t*, ompt_dependence_t const*, int)
==140865==ABORTING
```
To reproduce the error, use the following openmp code snippet:
```
/* initialise testMatrixInt Matrix, cols, r and c */
#pragma omp parallel private(r,c) shared(testMatrixInt)
{
#pragma omp for ordered(2)
for (r=1; r < rows; r++) {
for (c=1; c < cols; c++) {
#pragma omp ordered depend(sink:r-1, c+1) depend(sink:r-1,c-1)
testMatrixInt[r][c] = (testMatrixInt[r-1][c] + testMatrixInt[r-1][c-1]) % cols ;
#pragma omp ordered depend (source)
}
}
}
```
Compilation:
```
clang -g -stdlib=libc++ -fsanitize=thread -fopenmp -larcher test_case.c
```
It seems like the changes introduced by the commit https://reviews.llvm.org/D114005 causes this particular SEGV while using Archer.
Reviewed By: protze.joachim
Differential Revision: https://reviews.llvm.org/D115328
There are some limitations here, so this is behind a flag for now (in addition
to the config setting for the overall feature).
- symbols without exactly one associated header aren't handled right
- no macro support
- referencing std::size_t usually doesn't leave any trace in the AST that the
alias in std was used, so we associate with stddef.h instead of cstddef.
(An AST issue not specific to stdlib, but much worse there)
Differential Revision: https://reviews.llvm.org/D114077
This mechanism is used almost exclusively to enable extra warnings in clang-tidy
using ExtraArgs=-Wfoo, Checks="clang-diagnostic-foo".
Its presence is a strong signal that these flags are useful.
We choose not to actually emit them as clang-tidy diagnostics, but under their
"main" name - this ensures we show the same diagnostic in a consistent way.
We don't add the ExtraArgs to the compile command in general, but rather just
handle the -W<group> flags, which is the common case and avoids unexpected
side-effects.
And we only do this for the main file parse, when producing diagnostics.
Differential Revision: https://reviews.llvm.org/D116147
- This partially reverts d677a7cb056b17145a50ec8ca2ab6d5f4c494749 to
pacify GCC warnings like
```
base class should be explicitly initialized in the copy constructor
```
- Shall we keep turning on option `IgnoreBaseInCopyConstructors` when
enabling `readability-redundant-member-init` check?
This patch introduces support for targetting the Armv9.3-A architecture,
which should map to the existing Armv8.8-A extensions.
Differential Revision: https://reviews.llvm.org/D116159
After https://reviews.llvm.org/D115821 it became possible to create
`tensor<elem_type>` with a single `tensor.from_elements` operation without
collapsing tensor shape from `tensor<1xelem_type>` to `tensor<elem_type>`
Differential Revision: https://reviews.llvm.org/D115891
Provide signature while typing template arguments: Foo< ^here >
Here the parameters are e.g. "typename x", and the result type is e.g.
"struct" (class template) or "int" (variable template) or "bool (std::string)"
(function template).
Multiple overloads are possible when a template name is used for several
overloaded function templates.
Fixes https://github.com/clangd/clangd/issues/299
Differential Revision: https://reviews.llvm.org/D116352
