For the block smaller than a page size, one block is unlikely to
introduce more unused pages (at most 2 if it acrosses the page boundary
and both touched pages are unused). So it's better to apply the
threshold to reduce the time of scanning groups that can't release any
new pages.
Currently, WebAssembly/WASI target does not provide direct support for
code coverage.
This patch set fixes several issues to unlock the feature. The main
changes are:
1. Port `compiler-rt/lib/profile` to WebAssembly/WASI.
2. Adjust profile metadata sections for Wasm object file format.
- [CodeGen] Emit `__llvm_covmap` and `__llvm_covfun` as custom sections
instead of data segments.
- [lld] Align the interval space of custom sections at link time.
- [llvm-cov] Copy misaligned custom section data if the start address is
not aligned.
- [llvm-cov] Read `__llvm_prf_names` from data segments
3. [clang] Link with profile runtime libraries if requested
See each commit message for more details and rationale.
This is part of the effort to add code coverage support in Wasm target
of Swift toolchain.
Refer: 7.3.1 from [ISO
SPEC](https://www.open-std.org/jtc1/sc22/wg14/www/docs/n3220.pdf)
I have added complex variants of F16 and F128 in libc doc but have
omitted support for them since we will have to first investigate how
their support matrix for clang and gcc looks like, and then add header
guards for them accordingly. Planning to add them in follow up PRs once
this gets landed.
Check invoked tool with `starts_with`.
Addresses the issue where `perf2bolt` invoked using a distro symlink
`perf2bolt-16` fails to run in perf2bolt mode and runs in llvm-bolt mode
instead.
The issue is mentioned in https://vondra.me/posts/playing-with-bolt-and-postgres/
Test Plan:
```
ln -sf perf2bolt perf2bolt-20
perf2bolt-20 clang -p perf.data -o fdata.clang -w yaml.clang
...
PERF2BOLT: wrote 188593 objects and 0 memory objects to fdata.clang
```
Reviewers: ayermolo, rafaelauler, dcci, maksfb
Reviewed By: maksfb
Pull Request: https://github.com/llvm/llvm-project/pull/111072
Currently, our unwinder assumes that the functions are continuous (or at
least, that there are no functions which are "in the middle" of other
functions). Neither of these assumptions is true for functions optimized
by tools like propeller and (probably) bolt.
While there are many things that go wrong for these functions, the
biggest damage is caused by the unwind plan caching code, which
currently takes the maximalist extent of the function and assumes that
the unwind plan we get for that is going to be valid for all code inside
that range. If a part of the function has been moved into a "cold"
section, then the range of the function can be many megabytes, meaning
that any function within that range will probably fail to unwind.
We end up with this maximalist range because the unwinder asks for the
Function object for its range. This is only one of the strategies for
determining the range, but it is the first one -- and also the most
incorrect one. The second choice would is asking the eh_frame section
for the range of the function, and this one returns something reasonable
here (the address range of the current function fragment) -- which it
does because each fragment gets its own eh_frame entry (it has to,
because they have to be continuous).
With this in mind, this patch moves the eh_frame (and debug_frame) to
the front of the queue. I think that preferring this range makes sense
because eh_frame is one of the unwind plans that we return, and some
others (augmented eh_frame) are based on it. In theory this could break
some functions, where the debug info and eh_frame disagree on the extent
of the function (and eh_frame is the one who's wrong), but I don't know
of any such scenarios.
Closes#107985.
LanguageExtensions.rst states that `__builtin_shufflevector` and
`__builtin_convertvector` can be evaluated as constants, but this is not
reflected in Butiltins.td. This patch aligns these two.
The worker clause specifies iterations of the loop/ that are executed in
parallel by distributing the iterations among the multiple works within
a single gang.
The sema rules for this type are simply that it cannot be combined with
a `kernel` construct with a `num_workers` clause, child `loop` clauses
cannot contain a `gang` or `worker` clause, and that the argument is oly
allowed when associated with a `kernel`.
Bug fix: `@autoreleasepool`, `@synchronized`, and `@finally` were not
being noticed and treated as function effect violations.
---------
Co-authored-by: Doug Wyatt <dwyatt@apple.com>
This paper adds 'i' and 'j' as suffixes for forming a _Complex constant.
This feature has been supported in Clang since at least Clang 3.0, so
only test coverage is needed.
It does remove -Wgnu-imaginary-constant in C mode (still used in C++
mode) because the feature is now a C2y feature rather than a GNU one.
This is part of the effort to support for enabling plugins on windows by
adding better support for building llvm and clang as a DLL. These macros
are similar to the ones i added in #96630, but are for clang.
Added explicit symbol visibility macros definitions that are defined in
a new header and will be used to for shared library builds of clang to
export
symbols.
Updated clang cmake to define a macro to enable the symbol visibility
macros and explicitly disable them for clang tools that always use
static linking.
---------
Co-authored-by: Aaron Ballman <aaron@aaronballman.com>
The underlying issue was caused by a file included in two different
places which resulted in duplicate definition errors when linking
individual shared libraries. This was fixed in c3201ddaeac02a2c86a38b
[#109874].
Added .setMIFlag(MachineInstr::FrameSetup) to all BuildMI calls in
HexagonFrameLowering::insertAllocframe. This change ensures that the
test sugared-constants.ll passes upstream by correctly marking
instructions as part of the frame setup.
This patch fixes the sm90 cluster test by:
* Fixing a typo in LowerGpuOpsToNVVMOps where one of the ClusterDim Op
conversion pattern should actually be for the
ClusterDimBlocks Op. This addresses the compilation error for this test.
* The grid-size should be (4,4,1) instead of (2,2,1). This passes the
scf-if check against the threshold of 3 below and actually
generates the required prints from the GPU.
Signed-off-by: Durgadoss R <durgadossr@nvidia.com>
Windows doesn't have a built-in system log. Previously we got away with
writing to stdout and stderr because it was used only sporadically. As
we're trying to make the system log more useful on the other platforms,
the increased use become a concern. Make it a NOOP until someone figures
out a reasonable alternative.
The part of the code for parsing Verilog module instantiations
dereferenced a pointer without checking for null pointer. The pointer
may be null if the input is not complete and a line starts with a comma.
Reverse the generic foldVSelectToSignBitSplatMask fold on AVX512 targets where we can use the SETCC result directly in predicated moves/instructions.
Fixes#109272
For test program like this variable array is mentioned in both shared
clause and map clause. For OMP TARGET compound directives like this
where we have OMP TARGET TEAMS, map clause applies to TARGET directive
and SHARED clause applies to TEAMS directive. So both SHARED and MAP
clauses can co-exist.
> program test
> implicit none
> integer :: array(10,10),i,j
> !$omp target teams shared(array) map(tofrom:array)
> do i=1,10
> !$omp parallel do
> do j=1,10
> array(j,i)=i+j
> end do
> end do
> !$omp end target teams
> print *, array
> end program test
>
>
Before this PR we were checking for exclusivity for all target
directives set which is now relaxed to exclusivity check not being
applied to compound directives which can accept SHARED clause.
Reuse llvm::isTriviallyVectorizable in llvm::isNotCrossLaneOperation, in
order to get it to handle more intrinsics.
Alive2 proofs for changed tests: https://alive2.llvm.org/ce/z/XSV_GT
