This is the first in a series of patches that introduce C++20 importable
header units.
These differ from clang header modules in that:
(a) they are identifiable by an internal name
(b) they represent the top level source for a single header - although
that might include or import other headers.
We name importable header units with the path by which they are specified
(although that need not be the absolute path for the file).
So "foo/bar.h" would have a name "foo/bar.h". Header units are made a
separate module type so that we can deal with diagnosing places where they
are permitted but a named module is not.
Differential Revision: https://reviews.llvm.org/D121095
Update `WeakUndeclaredIdentifiers` to hold a collection of weak
aliases per identifier instead of only one.
This also allows the "used" state to be removed from `WeakInfo`
because it is really only there as an alternative to removing
processed map entries, and we can represent that using an empty set
now. The serialization code is updated for the removal of the field.
Additionally, a PCH test is added for the new functionality.
The records are grouped by the "target" identifier, which was already
being used as a key for lookup purposes. We also store only one record
per alias name; combined, this means that diagnostics are grouped by
the "target" and limited to one per alias (which should be acceptable).
Fixes PR28611.
Fixesllvm/llvm-project#28985.
Reviewed By: aaron.ballman, cebowleratibm
Differential Revision: https://reviews.llvm.org/D121927
Co-authored-by: Rachel Craik <rcraik@ca.ibm.com>
Co-authored-by: Jamie Schmeiser <schmeise@ca.ibm.com>
This is a re-commit of 98fd3b3598. The
newly added test was failing on the bots, and I've fixed the test now so
that it doesn't actually invoke the linker.
A previous patch removed the compiler generating offloading entries
for variables that were declared on the device but were internal or
hidden. This allowed us to compile programs but turns any attempt to run
'#pragma omp target update' on one of those variables a silent failure.
This patch adds a check in the semantic analysis for if the user is
attempting the update a variable on the device from the host that is not
externally visible.
Reviewed By: jdoerfert
Differential Revision: https://reviews.llvm.org/D122403
Current clang generates extra set of simd variant function attribute
with extra 'v' encoding.
For example:
_ZGVbN2v__Z5add_1Pf vs _ZGVbN2vv__Z5add_1Pf
The problem is due to declaration of ParamAttrs following:
llvm::SmallVector<ParamAttrTy, 8> ParamAttrs(ParamPositions.size());
where ParamPositions.size() is grown after following assignment:
Pos = ParamPositions[PVD];
So the PVD is not find in ParamPositions.
The problem is ParamPositions need to set for each FD decl. To fix this
Move ParamPositions's init inside while loop for each FD.
Differential Revision: https://reviews.llvm.org/D122338
After landing D121813 the binary size increase introduced by this change can be minimized by using --gc-sections link options. D121813 allows each individual callbacks to be optimized out if not used.
Reviewed By: vitalybuka, MaskRay
Differential Revision: https://reviews.llvm.org/D122407
Fix clang crash and add bitfield support in __builtin_dump_struct.
In clang13.0.x, a struct with three or more members and a bitfield at
the same time will cause a crash. In clang15.x, as long as the struct
has one bitfield, it will cause a crash in clang.
Open issue: https://github.com/llvm/llvm-project/issues/54462
Differential Revision: https://reviews.llvm.org/D122248
CUDA/HIP determines whether a function can be called based on
the device/host attributes of callee and caller. Clang assumes the
caller is CurContext. This is correct in most cases, however, it is
not correct in OpenMP parallel region when CUDA/HIP program
is compiled with -fopenmp. This causes incorrect overloading
resolution and missed diagnostics.
To get the correct caller, clang needs to chase the parent chain
of DeclContext starting from CurContext until a function decl
or a lambda decl is reached. Sema API is adapted to achieve that
and used to determine the caller in hostness check.
Reviewed by: Artem Belevich, Richard Smith
Differential Revision: https://reviews.llvm.org/D121765
This information isn't preserved in the DWARF description of function
types (though probably should be - it's preserved on the function
declarations/definitions themselves through the DW_AT_noreturn attribute
- but we should move or also include that in the subroutine type itself
too - but for now, with it not being there, the DWARF is lossy and
can't be reconstructed)
Most intrinsics, especially "default" ones, will not call back into the
IR module. `nocallback` encodes this nicely. As it was not used before,
this patch also makes use of `nocallback` in the Attributor which
results in many more `norecurse` deductions.
Tablegen part is mechanical, test updates by script.
Differential Revision: https://reviews.llvm.org/D118680
Adds basic parsing/sema/serialization support for the
#pragma omp target parallel loop directive.
Differential Revision: https://reviews.llvm.org/D122359
Since function parameters and return values are passed via param space, we
can force special alignment for values hold in it which will add vectorization
options. This change may be done if the function has private or internal
linkage. Special alignment is forced during 2 phases.
1) Instruction selection lowering. Here we use special alignment for function
prototypes (changing both own return value and parameters alignment), call
lowering (changing both callee's return value and parameters alignment).
2) IR pass nvptx-lower-args. Here we change alignment of byval parameters that
belong to param space (or are casted to it). We only handle cases when all
uses of such parameters are loads from it. For such loads, we can change the
alignment according to special type alignment and the load offset. Then,
load-store-vectorizer IR pass will perform vectorization where alignment
allows it.
Special alignment calculated as maximum from default ABI type alignment and
alignment 16. Alignment 16 is chosen because it's the maximum size of
vectorized ld.param & st.param.
Before specifying such special alignment, we should check if it is a multiple
of the alignment that the type already has. For example, if a value has an
enforced alignment of 64, default ABI alignment of 4 and special alignment
of 16, we should preserve 64.
This patch will be followed by a refactoring patch that removes duplicating
code in handling byval and non-byval arguments.
Differential Revision: https://reviews.llvm.org/D120129
Since function parameters and return values are passed via param space, we
can force special alignment for values hold in it which will add vectorization
options. This change may be done if the function has private or internal
linkage. Special alignment is forced during 2 phases.
1) Instruction selection lowering. Here we use special alignment for function
prototypes (changing both own return value and parameters alignment), call
lowering (changing both callee's return value and parameters alignment).
2) IR pass nvptx-lower-args. Here we change alignment of byval parameters that
belong to param space (or are casted to it). We only handle cases when all
uses of such parameters are loads from it. For such loads, we can change the
alignment according to special type alignment and the load offset. Then,
load-store-vectorizer IR pass will perform vectorization where alignment
allows it.
Special alignment calculated as maximum from default ABI type alignment and
alignment 16. Alignment 16 is chosen because it's the maximum size of
vectorized ld.param & st.param.
Before specifying such special alignment, we should check if it is a multiple
of the alignment that the type already has. For example, if a value has an
enforced alignment of 64, default ABI alignment of 4 and special alignment
of 16, we should preserve 64.
This patch will be followed by a refactoring patch that removes duplicating
code in handling byval and non-byval arguments.
Differential Revision: https://reviews.llvm.org/D121549
In some cases, we need to set alternative toolchain path other than the
default with system (headers, libraries, dynamic linker prefix, ld path,
etc.), e.g., to pick up newer components, but keep sysroot at the same
time (to pick up extra packages).
This change introduces a new option --overlay-platform-toolchain to set
up such alternative toolchain path.
Reviewed By: hubert.reinterpretcast
Differential Revision: https://reviews.llvm.org/D121992
Move the SourceRange from the old ParsedAttributesWithRange into
ParsedAttributesView, so we have source range information available
everywhere we use attributes.
This also removes ParsedAttributesWithRange (replaced by simply using
ParsedAttributes) and ParsedAttributesVieWithRange (replaced by using
ParsedAttributesView).
Differential Revision: https://reviews.llvm.org/D121201
Currently we create offloading entries to register device variables with
the host. When we register a variable we will look up the symbol in the
device image and map the device address to the host address. This is a
problem when the symbol is declared with hidden visibility or internal
linkage. This means the symbol is not accessible externally and we
cannot get its address. We should still allow static variables to be
declared on the device, but ew should not create an offloading entry for
them so they exist independently on the host and device.
Fixes#54309
Reviewed By: jdoerfert
Differential Revision: https://reviews.llvm.org/D122352
Clang fails to diagnose:
```
void test() {
int j = 0;
for (int i = 0; i < 1000; i++)
j++;
return;
}
```
Reason: Missing support for UnaryOperator.
We should not warn with volatile variables... so add check for it.
Reviewed By: efriedma
Differential Revision: https://reviews.llvm.org/D122271
Before actually executing the ExtractAPIAction, clear the
CompilationInstance's input list and replace it with a single
synthesized file that just includes (or imports in ObjC) all the inputs.
Depends on D122141
Differential Revision: https://reviews.llvm.org/D122175
CoroSplit lowers various coroutine intrinsics. It's a CGSCC pass and
CGSCC passes don't run on unreachable functions. Normally GlobalDCE will
come along and delete unreachable functions, but we don't run GlobalDCE
under -O0, so an unreachable function with coroutine intrinsics may
never have CoroSplit run on it.
This patch adds GlobalDCE when coroutines intrinsics are present. It
also now runs all coroutine passes conditional when coroutine intrinsics
are present. This should also solve the -O0 regression reported in
D105877 due to LazyCallGraph construction.
Fixes https://github.com/llvm/llvm-project/issues/54117
Reviewed By: ChuanqiXu
Differential Revision: https://reviews.llvm.org/D122275
Currently, Clang handles some qualifiers correctly for __auto_type, but
it does not handle the restrict or _Atomic qualifiers in the same way
that GCC does. This patch handles those qualifiers so that they attach
to the deduced type the same as const and volatile already do.
This fixes https://github.com/llvm/llvm-project/issues/53652
- Add `StructFieldRecord` and `StructRecord` to store API information
for structs
- Implement `VisitRecordDecl` in `ExtractAPIVisitor`
- Implement Symbol Graph serialization for struct records.
- Add test case for struct records.
Depends on D121873
Differential Revision: https://reviews.llvm.org/D122202
Add support for enum records
- Add `EnumConstantRecord` and `EnumRecord` to store API information for
enums
- Implement `VisitEnumDecl` in `ExtractAPIVisitor`
- Implement serializatin for enum records and `MemberOf` relationship
- Add test case for enum records
- Few other improvements
Depends on D122160
Differential Revision: https://reviews.llvm.org/D121873
Adds `--product-name=` flag to the clang driver. This gets forwarded to
cc1 only when we are performing a ExtractAPI Action. This is used to
populate the `name` field of the module object in the generated SymbolGraph.
Differential Revision: https://reviews.llvm.org/D122141
-fsplit-machine-functions is an optimization in codegen phase. when -flto is use, clang generate IR bitcode in .o files, and linker will call into these codegen optimization passes. Current clang driver doesn't pass this option to linker when both -fsplit-machine-functions and -flto are used, so the optimization is silently ignored. My fix generates linker option -plugin-opt=-split-machine-functions for this case. It allows the linker to pass "split-machine-functions" to code generator to turn on that optimization. It works for both gold and lld.
Reviewed By: hoy, wenlei
Differential Revision: https://reviews.llvm.org/D121969
In D92191, a bunch of test cases were added to check `clang-scan-deps` works in `clang-cl` mode as well.
We don't need to duplicate all test cases, though. Testing the few special cases we have in `clang-scan-deps` for `clang-cl` should be good enough:
1. Deducing output path (and therefore target name in our make output).
2. Ignoring `-Xclang` arguments in step 1.
3. Deducing resource directory by invoking the compiler executuable.
This test de-duplicates the extra clang-cl test cases.
Reviewed By: dexonsmith, saudi
Differential Revision: https://reviews.llvm.org/D121812
This patch gets rid of the ridiculous relative path we use to invoke the `module-deps-to-rsp.py` script and creates proper lit substitution, cleaning up the tests.
Reviewed By: dexonsmith
Differential Revision: https://reviews.llvm.org/D121525
* Check for warnings instead of using -Werror, to avoid masking the
type of diagnostic emitted
* use different -verify labels instead of using conditional
compilation of diagnostic checks
Reviewed By: aaron.ballman
Differential Revision: https://reviews.llvm.org/D122265
This patch extends the support for C/C++ operators for SVE
types to allow one of the arguments to be a scalar, in which
case a vector splat is performed.
Differential Revision: https://reviews.llvm.org/D121829
`HeaderSearch` currently assumes `LookupFileCache` is eventually populated in `LookupFile`. However, that's not always the case with `-fms-compatibility` and its early returns.
This patch adds a defensive check that the iterator pulled out of the cache is actually valid before using it.
(This bug was introduced in D119721. Before that, the cache was initialized to `0` - essentially the `search_dir_begin()` iterator.)
Reviewed By: dexonsmith, erichkeane
Differential Revision: https://reviews.llvm.org/D122237
This change fixes a crash in RangedConstraintManager.cpp:assumeSym due to an
unhandled BO_Div case.
clang: <root>clang/lib/StaticAnalyzer/Core/RangedConstraintManager.cpp:51:
virtual clang::ento::ProgramStateRef
clang::ento::RangedConstraintManager::assumeSym(clang::ento::ProgramStateRef,
clang::ento::SymbolRef, bool):
Assertion `BinaryOperator::isComparisonOp(Op)' failed.
Reviewed By: NoQ
Differential Revision: https://reviews.llvm.org/D122277
This reverts commit 56d46b36fc.
The LIT test SemaCXX/attr-trivial-abi.cpp is failing with 32bit build on
Windows. All the lines with the ifdef WIN32 are asserting but they are
not expected to. It looks like the LIT test was not tested on a 32bit
build of the compiler.
The -mbranch-protection definition in Options.td was not given a Group,
so this was causing clang to emit a -Wunused-command-line-argument
warning when this flag was passed to the linker driver. This was a
problem, because some build systems, like cmake, automatically pass the
C flags to the linker. Therefore, any program that was compiled with
-Werror and -mbranch-protection would fail to link with the error:
argument unused during compilation: '-mbranch-protection=standard' [-Werror,-Wunused-command-line-argument]
Reviewed By: vhscampos
Differential Revision: https://reviews.llvm.org/D121983
GCC supports power-of-2 size structures for the arguments. Clang supports fewer than GCC. But Clang always crashes for the unsupported cases.
This patch adds sema checks to do the diagnosts to solve these crashes.
Reviewed By: jyu2
Differential Revision: https://reviews.llvm.org/D107141
- The name SymbolGraph is inappropriate and confusing for the new library
for clang-extract-api. Refactor and rename things to make it clear that
ExtractAPI is the core functionality and SymbolGraph is one serializer
for the API information.
- Add documentation comments to ExtractAPI classes and methods to improve
readability and clearness of the ExtractAPI work.
Differential Revision: https://reviews.llvm.org/D122160
Allow goto, labelled statements as well as `static`, `thread_local`, and
non-literal variables in `constexpr` functions.
As specified. for all of the above (except labelled statements) constant
evaluation of the construct still fails.
For `constexpr` bodies, the proposal is implemented with diagnostics as
a language extension in older language modes. For determination of
whether a lambda body satisfies the requirements for a constexpr
function, the proposal is implemented only in C++2b mode to retain the
semantics of older modes for programs conforming to them.
Reviewed By: aaron.ballman, hubert.reinterpretcast, erichkeane
Differential Revision: https://reviews.llvm.org/D111400
These diagnostics were added to a diagnostic group, but that diagnostic
group was not under -Wgnu. I've now split them into their own
diagnostic group that is added both to the original group (so user's
currently opting in or out of these should not see a change) and under
the -Wgnu group so that -Wno-gnu can be used to disable all GNU
extension diagnostics. This fixes Issue 54444.
intrinsics.
Those operations are updated under a tail agnostic policy, but they
could have mask agnostic or undisturbed.
Reviewed By: rogfer01
Differential Revision: https://reviews.llvm.org/D120228
@mgehre-amd pointed out the following post-commit review feedback on
the changes in 8cba72177dcd8de5d37177dbaf2347e5c1f0f1e8:
As an example, the paper says 3wb /* Yields an _BitInt(3); two value
bits, one sign bit */.
So I would expect that 0xFwb gives _BitInt(5); four value bits, one
sign bit, but with this implementation I get _BitInt(2).
This is because ResultVal as 4 bits, and getMinSignedBits() inteprets
it as negative and thus says that 1 bit is enough to represent -1.
This corrects the behavior for calculating the bit-width and adds some
test coverage.
Usages of makeNull need to be deprecated in favor of makeNullWithWidth
for architectures where the pointer size should not be assumed. This can
occur when pointer sizes can be of different sizes, depending on address
space for example. See https://reviews.llvm.org/D118050 as an example.
This was uncovered initially in a downstream compiler project, and
tested through those systems tests.
steakhal performed systems testing across a large set of open source
projects.
Co-authored-by: steakhal
Resolves: https://github.com/llvm/llvm-project/issues/53664
Reviewed By: NoQ, steakhal
Differential Revision: https://reviews.llvm.org/D119601
This error was found when analyzing MySQL with CTU enabled.
When there are space characters in the lookup name, the current
delimiter searching strategy will make the file path wrongly parsed.
And when two lookup names have the same prefix before their first space
characters, a 'multiple definitions' error will be wrongly reported.
e.g. The lookup names for the two lambda exprs in the test case are
`c:@S@G@F@G#@Sa@F@operator int (*)(char)#1` and
`c:@S@G@F@G#@Sa@F@operator bool (*)(char)#1` respectively. And their
prefixes are both `c:@S@G@F@G#@Sa@F@operator` when using the first space
character as the delimiter.
Solving the problem by adding a length for the lookup name, making the
index items in the format of `<USR-Length>:<USR File> <Path>`.
---
In the test case of this patch, we found that it will trigger a "triple
mismatch" warning when using `clang -cc1` to analyze the source file
with CTU using the on-demand-parsing strategy in Darwin systems. And
this problem is also encountered in D75665, which is the patch
introducing the on-demand parsing strategy.
We temporarily bypass this problem by using the loading-ast-file
strategy.
Refer to the [discourse topic](https://discourse.llvm.org/t/60762) for
more details.
Differential Revision: https://reviews.llvm.org/D102669
This test assumes that the driver will set -object_path_lto linker
flag and it requests the platform linker with -fuse-ld=. When lld is
used as the host linker, the host linker version is unset and so
Clang won't set -object_path_lto since that flag is set conditionally
only when linker version is at least 116. We set the linker version
explicitly to make sure that -object_path_lto is set. That approach
is already used elsewhere in the test.
Differential Revision: https://reviews.llvm.org/D122110
For the following code,
void test() {
volatile int j = 0;
for (int i = 0; i < 1000; i++)
j += 1;
return;
}
If compiled with
clang -g -Wall -Werror -S -emit-llvm test.c
we will see the following error:
test.c:2:6: error: variable 'j' set but not used [-Werror,-Wunused-but-set-variable]
volatile int j = 0;
^
This is not quite right since 'j' is indeed used due to '+=' operator.
gcc doesn't emit error either in this case.
Also if we change 'j += 1' to 'j++', the warning will disappear
with latest clang.
Note that clang will issue the warning if the volatile declaration
involves only simple assignment (var = ...).
To fix the issue, in function MaybeDecrementCount(), if the
operator is a compound assignment (i.e., +=, -=, etc.) and the
variable is volatile, the count for RefsMinusAssignments will be
decremented, similar to 'j++' case.
Differential Revision: https://reviews.llvm.org/D121715
clang -extract-api should accept multiple headers and forward them to a
single CC1 instance. This change introduces a new ExtractAPIJobAction.
Currently API Extraction is done during the Precompile phase as this is
the current phase that matches the requirements the most. Adding a new
phase would need to change some logic in how phases are scheduled. If
the headers scheduled for API extraction are of different types the
driver emits a diagnostic.
Differential Revision: https://reviews.llvm.org/D121936
Full-expressions are Sema-generated implicit nodes that cover
constant-expressions and expressions-with-cleanup for temporaries.
Ignore those as part of implicit-ignore, and also remove too-aggressive
IgnoreImplicit (which includes nested ImplicitCastExprs, for example)
on unpacked sub-expressions.
Add some unittests to demonstrate that RecursiveASTVisitor sees through
ConstantExpr nodes correctly.
Adjust cxx2a-consteval test to cover diagnostics for nested consteval
expressions that were previously missed.
Fixes bug #53044.
Tests must not include headers from the host system.
It looks like the include wasn't needed for anything, so just remove it.
This makes check-clang work in a `git bash` launched from a cmd.exe
that isn't an MSVC shell (that is, %INCLUDE% isn't set).
Previously, OpenMP variant declarations for a function declaration that included
the 'cpu_dispatch', 'cpu_specific', or 'target' attributes was diagnosed, but
one with the 'target_clones' attribute was not. Now fixed.
Reviewed By: erichkeane, jdoerfert
Differential Revision: https://reviews.llvm.org/D121963
This change extends the existing diagnostic tests for OpenMP variant
declarations to cover diagnostics for declarations that include
multiversion function attributes. The new tests demonstrate a missing
check for the 'target_clones' attribute.
Reviewed By: erichkeane, jdoerfert
Differential Revision: https://reviews.llvm.org/D121962
Previously, an attempt to declare an overload of a multiversion function
in C was not properly diagnosed. In some cases, diagnostics were simply
missing. In other cases the following assertion failure occured...
```
Assertion `(Previous.empty() || llvm::any_of(Previous, [](const NamedDecl *ND) { return ND->hasAttr(); })) && "Non-redecls shouldn't happen without overloadable present"' failed.
```
... or the following diagnostic was spuriously issued.
```
error: at most one overload for a given name may lack the 'overloadable' attribute
```
The diagnostics issued in some cases could be improved. When the function
type of a redeclaration does not match the prior declaration, it would be
preferable to diagnose the type mismatch before diagnosing mismatched
attributes. Diagnostics are also missing for some cases.
Reviewed By: erichkeane
Differential Revision: https://reviews.llvm.org/D121959
This change adds test cases to validate diagnostics for overloaded sets
that contain declarations of multiversion functions. Many of the added test
cases exercise declarations that are intended to be valid. Others are
intended to be valid if and when restrictions on multiversion functions
being declared with the overloadable attribute are lifted.
Several of the new test cases currently trigger the following assertion
failure in SemaDecl.cpp; the relevant test is therefore marked as an
expected failure pending a fix.
```
Assertion `(Previous.empty() || llvm::any_of(Previous, [](const NamedDecl *ND) { return ND->hasAttr(); })) && "Non-redecls shouldn't happen without overloadable present"' failed.
```
Reviewed By: erichkeane
Differential Revision: https://reviews.llvm.org/D121954
The EmitLoadOfPointer() call already specified the right pointer
type, but it did not match the Address we're loading from, so we
need to insert a bitcast first.
%S refers to the directory of %s, not to the cwd. This is mostly
handled correctly, but update_cc_test_checks.py used the wrong
path for non-FileCheck RUN lines.
Reapplying this with a fix for an update_cc_test_checks test that
was based on cwd semantics.
Partially fixes PR24883.
The patch sets Reference bit while instantiating a typedef if it
previously was found referenced.
Reviewed By: thakis
Differential Revision: https://reviews.llvm.org/D114382
Rather than using a dummy void pointer type, we should specify the
correct private type and perform the bitcast beforehand rather than
afterwards. This way, the Address will have correct alignment
information.
This is a follow up to 565603cc94,
which made macOS the default target OS for `-arch arm64` when
running on an Apple Silicon Mac. Now it'll be the default when
running on an Intel Mac too.
clang/test/Driver/apple-arm64-arch.c was a bit odd before: it was added
for the above commit, but tested the inverse behaviour and XFAIL'ed on
Apple Silicon. This inverts it to the (new) behaviour (that's now
correct regardless) and removes the XFAIL.
Radar-Id: rdar://90500294
https://reviews.llvm.org/D23944 implemented the #pragma intrinsic from
MSVC. This causes the statement #pragma intrinsic(cpuid) to fail [0]
on Clang because cpuid is currently implemented in intrin.h instead
of a Clang builtin. Reimplementing cpuid (as well as it's releated
function, cpuidex) should resolve this.
[0]: https://crbug.com/1279344
Differential revision: https://reviews.llvm.org/D121653
The ppc64be bot emits the dtor metadata first for some reason. We should
investigate this or make the _cc_ update script able to use variables
instead of fixed numbers (e.g., !1). The IR update script does that
already.
__builtin_memcpy_inline doesn't use the usual builtin argument validation code,
so it crashed when receiving wrong number of argument. Add the missing validation
check.
Open issue: https://github.com/llvm/llvm-project/issues/52949
Reviewed By: gchatelet
Differential Revision: https://reviews.llvm.org/D121965
Committed by gchatelet on behalf of "Roy Jacobson <roi.jacobson1@gmail.com>"
Rather than specifying a dummy type in EmitLoadOfPointer() and
then casting it to the correct one, we should instead specify the
correct type and cast beforehand. Otherwise the computed alignment
will be incorrect.
This adds support for multiple attributes in `#pragma clang attribute push`, for example:
```
```
or
```
```
Related attributes can now be applied with a single pragma, which makes it harder for developers to make an accidental error later when editing the code.
rdar://78269653
Differential Revision: https://reviews.llvm.org/D121283
PowerPC is lacking tests checking `_Atomic` alignment in cfe. Adding these tests since we're going to make change to align with gcc on Linux.
Reviewed By: hubert.reinterpretcast, jsji
Differential Revision: https://reviews.llvm.org/D121441
Reimplements MisExpect diagnostics from D66324 to reconstruct its
original checking methodology only using MD_prof branch_weights
metadata.
New checks rely on 2 invariants:
1) For frontend instrumentation, MD_prof branch_weights will always be
populated before llvm.expect intrinsics are lowered.
2) for IR and sample profiling, llvm.expect intrinsics will always be
lowered before branch_weights are populated from the IR profiles.
These invariants allow the checking to assume how the existing branch
weights are populated depending on the profiling method used, and emit
the correct diagnostics. If these invariants are ever invalidated, the
MisExpect related checks would need to be updated, potentially by
re-introducing MD_misexpect metadata, and ensuring it always will be
transformed the same way as branch_weights in other optimization passes.
Frontend based profiling is now enabled without using LLVM Args, by
introducing a new CodeGen option, and checking if the -Wmisexpect flag
has been passed on the command line.
Differential Revision: https://reviews.llvm.org/D115907
Summary:
Specifically, for trap handling, for targets that do not support getDoorbellID,
we load the queue_ptr from the implicit kernarg, and move queue_ptr to s[0:1].
To get aperture bases when targets do not have aperture registers, we load
private_base or shared_base directly from the implicit kernarg. In clang, we use
implicitarg_ptr + offsets to implement __builtin_amdgcn_workgroup_size_{xyz}.
Reviewers: arsenm, sameerds, yaxunl
Differential Revision: https://reviews.llvm.org/D120265
FLT_EVAL_METHOD tells the user the precision at which, temporary results
are evaluated but when fast-math is enabled, the numeric values are not
guaranteed to match the source semantics, so the eval-method is
meaningless.
For example, the expression `x + y + z` has as source semantics `(x + y)
+ z`. FLT_EVAL_METHOD is telling the user at which precision `(x + y)`
is evaluated. With fast-math enable the compiler can choose to
evaluate the expression as `(y + z) + x`.
The correct behavior is to set the FLT_EVAL_METHOD to `-1` to tell the
user that the precision of the intermediate values is unknow. This
patch is doing that.
Differential Revision: https://reviews.llvm.org/D121122
Currently we allow half types in vectors if the scalar Zfh extension
is enabled. This behavior is not inline with the vector spec. For f32
and f64 types, the Zve32f, Zve64f, Zve64d, and V explicitly control
the availablity of floating point types in vectors.
In order to make our compiler compliant, we either need to remove all support
for half in vectors or we need an extension to control it.
Draft spec here https://github.com/riscv/riscv-v-spec/pull/780
Reviewed By: kito-cheng
Differential Revision: https://reviews.llvm.org/D121345
If we are equality comparing an FP literal with a value cast from a type
where the literal can't be represented, that's known true or false and
probably a programmer error.
Fixes issue #54222.
https://github.com/llvm/llvm-project/issues/54222
Note - I added the optimizer change with:
9397bdc67e
...and as discussed in the post-commit comments, that transform might be
too dangerous without this warning in place, so it was reverted to allow
this change first.
Differential Revision: https://reviews.llvm.org/D121306
Poison trivial class members one-by-one in the reverse order of their
construction, instead of all-at-once at the very end.
For example, in the following code access to `x` from `~B` will
produce an undefined value.
struct A {
struct B b;
int x;
};
Reviewed By: kda
Differential Revision: https://reviews.llvm.org/D119600
-fsanitize-memory-use-after-dtor detects memory access after a
subobject is destroyed but its memory is not yet deallocated.
This is done by poisoning each object memory near the end of its destructor.
Subobjects (members and base classes) do this in their respective
destructors, and the parent class does the same for its members with
trivial destructors.
Inexplicably, base classes with trivial destructors are not handled at
all. This change fixes this oversight by adding the base class poisoning logic
to the parent class destructor.
Reviewed By: vitalybuka
Differential Revision: https://reviews.llvm.org/D119300
This is a preliminary patch ahead of D119792 (I'll rebase that one on top of this).
This shows what Clang's _current_ behaviour is for calculating NRVO in various
common cases. Then, in D119792 (and future patches), I'll be able to demostrate
exactly how LLVM IR for each of these cases changes.
Reviewed By: Quuxplusone
Differential Revision: https://reviews.llvm.org/D119927
The clang/SymbolGraph/global_record.c test case explicitly diffs the
clang version in use, which causes failures. Fix the issue by normalize
the `generator` field before checking the output.
Add facilities for extract-api:
- Structs/classes to hold collected API information: `APIRecord`, `API`
- Structs/classes for API information:
- `AvailabilityInfo`: aggregated availbility information
- `DeclarationFragments`: declaration fragments
- `DeclarationFragmentsBuilder`: helper class to build declaration
fragments for various types/declarations
- `FunctionSignature`: function signature
- Serialization: `Serializer`
- Add output file for `ExtractAPIAction`
- Refactor `clang::RawComment::getFormattedText` to provide an
additional `getFormattedLines` for a more detailed view of comment lines
used for the SymbolGraph format
Add support for global records (global variables and functions)
- Add `GlobalRecord` based on `APIRecord` to store global records'
information
- Implement `VisitVarDecl` and `VisitFunctionDecl` in `ExtractAPIVisitor` to
collect information
- Implement serialization for global records
- Add test case for global records
Differential Revision: https://reviews.llvm.org/D119479
In AMD GPU device code the globals are in AS(1). Before, we crashed if
the global was a structure. Now we simply cast away the AS before we
generate the code to initialize the global.
Differential Revision: https://reviews.llvm.org/D121837
Includes verifier changes checking the elementtype, clang codegen
changes to emit the elementtype, and ISel changes using the elementtype.
Basically the same as D120527.
Reviewed By: #opaque-pointers, nikic
Differential Revision: https://reviews.llvm.org/D121847
Fix the instruction names to match the WebAssembly spec:
- `i32x4.trunc_sat_zero_f64x2_{s,u}` => `i32x4.trunc_sat_f64x2_{s,u}_zero`
- `f32x4.demote_zero_f64x2` => `f32x4.demote_f64x2_zero`
Also rename related things like intrinsics, builtins, and test functions to
match.
Reviewed By: aheejin
Differential Revision: https://reviews.llvm.org/D121661
Current ASTContext.getAttributedType() takes attribute kind,
ModifiedType and EquivType as the hash to decide whether an AST node
has been generated or note. But this is not enough for btf_type_tag
as the attribute might have the same ModifiedType and EquivType, but
still have different string associated with attribute.
For example, for a data structure like below,
struct map_value {
int __attribute__((btf_type_tag("tag1"))) __attribute__((btf_type_tag("tag3"))) *a;
int __attribute__((btf_type_tag("tag2"))) __attribute__((btf_type_tag("tag4"))) *b;
};
The current ASTContext.getAttributedType() will produce
an AST similar to below:
struct map_value {
int __attribute__((btf_type_tag("tag1"))) __attribute__((btf_type_tag("tag3"))) *a;
int __attribute__((btf_type_tag("tag1"))) __attribute__((btf_type_tag("tag3"))) *b;
};
and this is incorrect.
It is very difficult to use the current AttributedType as it is hard to
get the tag information. To fix the problem, this patch introduced
BTFTagAttributedType which is similar to AttributedType
in many ways but with an additional BTFTypeTagAttr. The tag itself can
be retrieved with BTFTypeTagAttr.
With the new BTFTagAttributed type, the debuginfo code can be greatly
simplified compared to previous TypeLoc based approach.
Differential Revision: https://reviews.llvm.org/D120296
Add the rest of intrinsics to clang except intrinsics using vector mask
registers.
Reviewed By: simoll
Differential Revision: https://reviews.llvm.org/D121586
CT_Dependent
When compile following code without -std=c++17, clang will abort by
llvm_unreachable:
class A {
public:
static const char X;
};
const char A::X = 0;
template<typename U> void func() noexcept(U::X);
template<class... B, char x>
void foo(void(B...) noexcept(x)) {}
void bar()
{
foo(func<A>);
}
So, my solution is to let EST_Uninstantiated in
FunctionProtoType::canThrow return CT_Dependent
Differential Revision: https://reviews.llvm.org/D121498
The code for traversing precompiled dependencies is somewhat complicated and contains a dangling iterator bug.
This patch simplifies the code and fixes the bug.
Reviewed By: dexonsmith
Differential Revision: https://reviews.llvm.org/D121533
When pruning header search paths (to reduce the number of modules we need to build explicitly), we can't prune the search paths used in (transitive) dependencies of a module. Otherwise, we could end up with either of the following dependency graphs:
```
X:<hash1> -> Y:<hash2>
X:<hash1> -> Y:<hash3>
```
depending on the search paths of the translation unit we discovered `X` and `Y` from.
This patch fixes that.
Depends on D121295.
Reviewed By: dexonsmith
Differential Revision: https://reviews.llvm.org/D121303
To reduce the number of modules we build in explicit builds (which use strict context hash), we prune unused header search paths. This essentially merges parts of the dependency graph.
Determining whether a search path was used to discover a module (through implicit module maps) proved to be somewhat complicated. Initial support landed in D102923, while D113676 attempts to fix some bugs.
However, now that we don't use implicit module maps in explicit builds (since D120465), we don't need to consider such search paths as used anymore. Modules are no longer discovered through the header search mechanism, so we can drop such search paths (provided they are not needed for other reasons).
This patch removes whatever support for detecting such usage we had, since it's buggy and not required anymore.
Depends on D120465.
Reviewed By: dexonsmith
Differential Revision: https://reviews.llvm.org/D121295
This is the `ext_vector_type` alternative to D81083.
This patch extends Clang to allow 'bool' as a valid vector element type
(attribute ext_vector_type) in C/C++.
This is intended as the canonical type for SIMD masks and facilitates
clean vector intrinsic declarations. Vectors of i1 are supported on IR
level and below down to many SIMD ISAs, such as AVX512, ARM SVE (fixed
vector length) and the VE target (NEC SX-Aurora TSUBASA).
The RFC on cfe-dev: https://lists.llvm.org/pipermail/cfe-dev/2020-May/065434.html
Reviewed By: erichkeane
Differential Revision: https://reviews.llvm.org/D88905
This reverts commit 049f4e4eab.
The problem was a stray dependency in CLANG_TEST_DEPS which caused cmake
to fail if clang-pseudo wasn't built. This is now removed.
This should make clearer that:
- it's not part of clang proper
- there's no expectation to update it along with clang (beyond green tests)
- clang should not depend on it
This is intended to be expose a library, so unlike other tools has a split
between include/ and lib/.
The main renames are:
clang/lib/Tooling/Syntax/Pseudo/* => clang-tools-extra/pseudo/lib/*
clang/include/clang/Tooling/Syntax/Pseudo/* => clang-tools-extra/pseudo/include/clang-pseudo/*
clang/tools/clang/pseudo/* => clang-tools-extra/pseudo/tool/*
clang/test/Syntax/* => clang-tools-extra/pseudo/test/*
clang/unittests/Tooling/Syntax/Pseudo/* => clang-tools-extra/pseudo/unittests/*
#include "clang/Tooling/Syntax/Pseudo/*" => #include "clang-pseudo/*"
namespace clang::syntax::pseudo => namespace clang::pseudo
check-clang => check-clang-pseudo
clangToolingSyntaxPseudo => clangPseudo
The clang-pseudo and ClangPseudoTests binaries are not renamed.
See discussion around:
https://discourse.llvm.org/t/rfc-a-c-pseudo-parser-for-tooling/59217/50
Differential Revision: https://reviews.llvm.org/D121233
This fixes a bug that happens when using -fdebug-prefix-map to remap an
absolute path to a relative path. Since the path was absolute before
remapping, it is safe to assume that concatenating the remapped working
directory would be wrong.
This was originally submitted as https://reviews.llvm.org/D113718, but
reverted because when testing with dwarf 5 enabled, the tests were too
strict.
Differential Revision: https://reviews.llvm.org/D121663
function candidates
See: https://github.com/llvm/llvm-project/issues/54379
I tried to see if I can reuse ResolveAddressOfOverloadedFunction for
explicit function instantiation and so I managed to hit this ICE.
Bug was the diagnostic required an argument (%0) and specific code path
didn't pass an argument.
Differential Revision: https://reviews.llvm.org/D121646
GCC supports:
- `namespace <gnu attributes> identifier`
- `namespace identifier <gnu attributes>`
But clang supports only `namespace identifier <gnu attributes>` and diagnostics for `namespace <gnu attributes> identifier` case looks unclear:
Code:
```
namespace __attribute__((visibility("hidden"))) A
{
}
```
Diags:
```
test.cpp:1:49: error: expected identifier or '{'
namespace __attribute__((visibility("hidden"))) A
^
test.cpp:1:49: error: C++ requires a type specifier for all declarations
test.cpp:3:2: error: expected ';' after top level declarator
}
```
This patch adds support for `namespace <gnu attributes> identifier` and also forbids gnu attributes for nested namespaces (this already done for C++ attributes).
Reviewed By: aaron.ballman
Differential Revision: https://reviews.llvm.org/D121245
This patch follows the same reasoning as D114481. The PCH reader looks for `__clangast` section in the precompiled module file, which is not present in the file on AIX and not supported in XCOFF yet.
Reviewed By: daltenty
Differential Revision: https://reviews.llvm.org/D121709
On unix systems this logic would not separate the file and directory of
the DIFile unless they shared more components at the start than just the
root path character. The logic to do this was unix specific so it didn't
work on Windows. Now we check if the entire root_path is the same as
what you were going to set as the Dir and use the full filepath in that
case.
Differential Revision: https://reviews.llvm.org/D111579
This patch adds the offload kind to the embedded section name in
preparation for offloading to different kinda like CUDA or HIP.
Depends on D120288
Reviewed By: jdoerfert
Differential Revision: https://reviews.llvm.org/D120271
Currently we use the `-fembed-offload-object` option to embed a binary
file into the host as a named section. This is currently only used as a
codegen action, meaning we only handle this option correctly when the
input is a bitcode file. This patch adds the same handling to embed an
offloading object after we complete code generation. This allows us to
embed the object correctly if the input file is source or bitcode.
Reviewed By: jdoerfert
Differential Revision: https://reviews.llvm.org/D120270
Motivation:
```
int test(int x, int y) {
int r = 0;
[[clang::always_inline]] r += foo(x, y); // force compiler to inline this function here
return r;
}
```
In 2018, @kuhar proposed "Introduce per-callsite inline intrinsics" in https://reviews.llvm.org/D51200 to solve this motivation case (and many others).
This patch solves this problem with call site attribute. "noinline" statement attribute already landed in D119061. Also, some LLVM Inliner fixes landed so call site attribute is stronger than function attribute.
Reviewed By: aaron.ballman
Differential Revision: https://reviews.llvm.org/D120717
Includes verifier changes checking the elementtype, clang codegen
changes to emit the elementtype, and ISel changes using the elementtype.
Reviewed By: #opaque-pointers, nikic
Differential Revision: https://reviews.llvm.org/D120527
RequireAnalysis<GlobalsAA> doesn't actually recompute GlobalsAA.
GlobalsAA isn't invalidated (unless specifically invalidated) because
it's self-updating via ValueHandles, but can be imprecise during the
self-updates.
Rather than invalidating GlobalsAA, which would invalidate AAManager and
any analyses that use AAManager, create a new pass that recomputes
GlobalsAA.
Fixes#53131.
Differential Revision: https://reviews.llvm.org/D121167
WG14 adopted N2775 (http://www.open-std.org/jtc1/sc22/wg14/www/docs/n2775.pdf)
at our Feb 2022 meeting. This paper adds a literal suffix for
bit-precise types that automatically sizes the bit-precise type to be
the smallest possible legal _BitInt type that can represent the literal
value. The suffix chosen is wb (for a signed bit-precise type) which
can be combined with the u suffix (for an unsigned bit-precise type).
The preprocessor continues to operate as-if all integer types were
intmax_t/uintmax_t, including bit-precise integer types. It is a
constraint violation if the bit-precise literal is too large to fit
within that type in the context of the preprocessor (when still using
a pp-number preprocessing token), but it is not a constraint violation
in other circumstances. This allows you to make bit-precise integer
literals that are wider than what the preprocessor currently supports
in order to initialize variables, etc.
Due to various implementation constraints, despite the programmer
choosing a 'processor' cpu_dispatch/cpu_specific needs to use the
'feature' list of a processor to identify it. This results in the
identified processor in source-code not being propogated to the
optimizer, and thus, not able to be tuned for.
This patch changes to use the actual cpu as written for tune-cpu so that
opt can make decisions based on the cpu-as-spelled, which should better
match the behavior expected by the programmer.
Note that the 'valid' list of processors for x86 is in
llvm/include/llvm/Support/X86TargetParser.def. At the moment, this list
contains only Intel processors, but other vendors may wish to add their
own entries as 'alias'es (or with different feature lists!).
If this is not done, there is two potential performance issues with the
patch, but I believe them to be worth it in light of the improvements to
behavior and performance.
1- In the event that the user spelled "ProcessorB", but we only have the
features available to test for "ProcessorA" (where A is B minus
features),
AND there is an optimization opportunity for "B" that negatively affects
"A", the optimizer will likely choose to do so.
2- In the event that the user spelled VendorI's processor, and the
feature
list allows it to run on VendorA's processor of similar features, AND
there
is an optimization opportunity for VendorIs that negatively affects
"A"s,
the optimizer will likely choose to do so. This can be fixed by adding
an
alias to X86TargetParser.def.
Differential Revision: https://reviews.llvm.org/D121410
This change teaches the Sema logic for `__builtin_memcpy_inline` to implicitly convert arrays passed as arguments to pointers, similarly to regular `memcpy`.
This code will no longer cause a compiler crash:
```
void f(char *p) {
char s[1] = {0};
__builtin_memcpy_inline(p, s, 1);
}
```
rdar://88147527
Differential Revision: https://reviews.llvm.org/D121475
Instead of outputting the test directory into the JSON result file, parsing it with `FileCheck` and then potentially stripping it, simply use `FileCheck`'s `-D` option.
Note that we use `%/t` instead of `%t` in order to normalize to forward slashes on Windows, which matches what we do with `sed 's:\\\\\?:/:g'`.
Differential Revision: https://reviews.llvm.org/D121516
With explicit modules build, the '-fmodules-cache-path=' argument is unused.
This patch removes the argument to avoid warnings or errors (with '-Werror') stemming from that.
Depends on D118915.
Reviewed By: dexonsmith
Differential Revision: https://reviews.llvm.org/D120474
The `clang-scan-deps` tool currently generates `-fmodule-file=` command-line arguments for the whole transitive closure of modular dependencies. This is not necessary, we only need to provide the direct dependencies on the command line. Information about transitive dependencies is stored within the `.pcm` files of direct dependencies. This makes the command lines shorter, but should be a NFC otherwise (unless there are bugs in the loading mechanism for explicit modules).
Depends on D120465.
Reviewed By: Bigcheese
Differential Revision: https://reviews.llvm.org/D118915
Since D113473, we don't report any module map files via `-fmodule-map-file=` in explicit builds. The ultimate goal here is to make sure Clang doesn't open/read/parse/evaluate unnecessary module maps.
However, implicit module maps still end up reading all reachable module maps. This patch disables implicit module maps in explicit builds.
Unfortunately, we still need to report some module map files that aren't encoded in PCM files of dependencies: module maps that are necessary to correctly evaluate includes in modules marked as `[no_undeclared_includes]`.
Depends on D120464.
Reviewed By: dexonsmith
Differential Revision: https://reviews.llvm.org/D120465
This clarifies that this is an LLVM specific variable and avoids
potential conflicts with other projects.
Differential Revision: https://reviews.llvm.org/D119918
This patch implements support for the +, -, *, / and % operators on sizeless SVE
types. Support for these operators on svbool_t is excluded.
Differential Revision: https://reviews.llvm.org/D120323
Giving an int parameter to SVE intrinsics svptrue and svcnt caused Clang
to crash on compilation. Changing their parameter types to void instead of
omitting args results in a diagnostic error message instead.
Differential Revision: https://reviews.llvm.org/D121294
The llvm pre-merge test got timeout due to large test files, this commit
split up the files that have over 10k lines under clang/test/CodeGen/RISCV
into even smaller ones.
Differential Revision: https://reviews.llvm.org/D121431
As far as I can tell, MSVC allows the relevant conversions for all
pointer types. Found compiling a Windows SDK header.
I've verified that the updated errors in MicrosoftExtensions.cpp match
the ones that MSVC actually emits, except for the one with a FIXME. (Not
sure why this wasn't done for the patch that added the tests.)
To make up for the missing error, add a warning that triggers on
conversions that drop the __unaligned qualfier.
Differential Revision: https://reviews.llvm.org/D120936
35ca7d9ddf broke 471c4f8299 for -arch flags that don't map 1:1
to the triple arch. This has been broken for the many years since.
It hasn't mattered much since then, mostly because few people use it,
but also because it works for x86_64/i386, armv7/armv7s
don't differ much, arm64 is its own arch, and arm64/arm64_32 have
different arches (and it's a rare combination anyway).
But arm64/arm64e exposes this issue again.
Patch by: Justin Bogner <mail@justinbogner.com>
with some added tests.
Fix a typo about -fno-gpu-sanitize handling and disable warnings when
-fno-gpu-sanitize is specified.
Reviewed by: Artem Belevich
Differential Revision: https://reviews.llvm.org/D121302
The clang compiler prepends the HIP header include paths to the search
list using -internal-isystem when building for the HIP language. This
prevents warnings related to things like reserved identifiers when
including the HIP headers even when ROCm is installed in a non-system
directory, such as /opt/rocm.
However, when HIP is installed in /usr, then the prepended include
path would be /usr/include. That is a problem, because the C standard
library headers are stored in /usr/include and the C++ standard
library headers must come before the C library headers in the search
path list (because the C++ standard library headers use #include_next
to include the C standard library headers).
While the HIP wrapper headers _do_ need to be earlier in the search
than the C++ headers, those headers get their own subdirectory and
their own explicit -internal-isystem argument. This include path is for
<hip/hip_runtime_api.h> and <hip/hip_runtime.h>, which do not require a
particular search ordering with respect to the C or C++ headers. Thus,
HIP include path is added after other system include paths.
With contribution from Cordell Bloor.
Reviewed by: Artem Belevich
Differential Revision: https://reviews.llvm.org/D120132
Regresses:
typedef struct {
static void f() {
}
} a_t;
Causing this to error instead of warn, because the linkage is computed
earlier/too early perhaps. I'll send out a review to see if there's some
other path forward or if this is an acceptable regression, etc.
This reverts commit 275c56226d.
The crypto extension have several shorthand extensions that don't consist of any extra instructions.
Take `zk` for example, while the extension would imply `zkn, zkr, zkt`. The 3 extensions should also
combine back into `zk` to maintain the canonical order in isa strings.
This patch addresses the above.
Reviewed By: VincentWu
Differential Revision: https://reviews.llvm.org/D119530
We were not creating an evaluation context for the TU scope, so we
never popped an evaluation context for it. Popping the evaluation
context triggers a number of diagnostics, including warnings about
immediate invocations that we were previously missing.
Note: I think we have an additional issue that we should solve, but not
as part of this patch. I don't think Clang is properly modeling static
initialization as happening before constant expression evaluation. I
think structure members members are zero initialized per
http://eel.is/c++draft/basic.start.static#1,
https://eel.is/c++draft/basic.start.static#2.sentence-2, and
http://eel.is/c++draft/dcl.init#general-6.2 and the new test case
actually should be accepted. However, it's also worth noting that other
compilers behave the way this patch makes Clang behave:
https://godbolt.org/z/T7noqhdPr
The existing module symbol mangling scheme turns out to be
undemangleable. It is also desirable to switch to the
strong-ownership model as the hoped-for C++17 compatibility turns out
to be fragile, and we also now have a better way of controlling that.
The issue is captured on the ABI list at:
https://github.com/itanium-cxx-abi/cxx-abi/issues/134
A document describing the issues and new mangling is at:
https://drive.google.com/file/d/1qQjqptzOFT_lfXH8L6-iD9nCRi34wjft/view
This patch is the code-generation part. I have a demangler too, but
that patch is based on some to-be-landed refactoring of the demangler.
The old mangling is unceremoniously dropped. No backwards
compatibility, no deprectated old-mangling flag. It was always
labelled experimental. (Old and new manglings cannot be confused.)
Reviewed By: ChuanqiXu
Differential Revision: https://reviews.llvm.org/D118352
Otherwise we will fail to generate the definition of a defaulted destructor,
if the only reference was in a templated temporary.
rdar://89366678
Differential Revision: https://reviews.llvm.org/D120426
A load via pointer cast to constant will return true from
pointsToConstantMemory which is not necessarily so.
Fixes: SWDEV-326463
Differential Revision: https://reviews.llvm.org/D121172
This relands commit 7313474319.
It failed on Windows/Mac because `-fjmc` is only checked for ELF targets.
Check the flag unconditionally instead and issue a warning for non-ELF targets.
The paramemter of hint clause in OpenMP critical hint should be
non-negative. The omp_lock_hint_none is 0 in omp.h.
Reviewed By: Alexey Bataev
Differential Revision: https://reviews.llvm.org/D121101
The motivation is to enable the MSVC-style JMC instrumentation usable by a ELF-based
debugger. Since there is no prior experience implementing JMC feature for ELF-based
debugger, it might be better to just reuse existing MSVC-style JMC instrumentation.
For debuggers that support both ELF&COFF (like lldb), the JMC implementation might
be shared between ELF&COFF. If this is found to inadequate, it is pretty low-cost
switching to alternatives.
Implementation:
- The '-fjmc' is already a driver and cc1 flag. Wire it up for ELF in the driver.
- Refactor the JMC instrumentation pass a little bit.
- The ELF handling is different from MSVC in two places:
* the flag section name is ".just.my.code" instead of ".msvcjmc"
* the way default function is provided: MSVC uses /alternatename; ELF uses weak function.
Based on D118428.
Reviewed By: rnk
Differential Revision: https://reviews.llvm.org/D119910
Since D106876, PCM files don't report module maps as input files unless they contributed to the compilation.
Reporting only module maps of (transitively) imported modules is not enough, though. For modules marked with `[no_undeclared_includes]`, other module maps affect the compilation by introducing anti-dependencies.
This patch makes sure such module maps are being reported as input files.
Depends on D120463.
Reviewed By: dexonsmith
Differential Revision: https://reviews.llvm.org/D120464
This patch simplifies a test that checks only used module map files are reported as input files in PCM files.
Instead of using opaque `diff`, this patch uses `clang -module-file-info` and `FileCheck` to verify this.
Reviewed By: dexonsmith
Differential Revision: https://reviews.llvm.org/D120463
Currently in Clang, we have two types of builtins for fnmsub operation:
one for float/double vector, they'll be transformed into IR operations;
one for float/double scalar, they'll generate corresponding intrinsics.
But for the vector version of builtin, the 3 op chain may be recognized
as expensive by some passes (like early cse). We need some way to keep
the fnmsub form until code generation.
This patch introduces ppc.fnmsub.* intrinsic to unify four fnmsub
intrinsics.
Reviewed By: shchenz
Differential Revision: https://reviews.llvm.org/D116015
The OpenMPIRBuilder has a bug. Specifically, suppose you have two nested openmp parallel regions (writing with MLIR for ease)
```
omp.parallel {
%a = ...
omp.parallel {
use(%a)
}
}
```
As OpenMP only permits pointer-like inputs, the builder will wrap all of the inputs into a stack allocation, and then pass this
allocation to the inner parallel. For example, we would want to get something like the following:
```
omp.parallel {
%a = ...
%tmp = alloc
store %tmp[] = %a
kmpc_fork(outlined, %tmp)
}
```
However, in practice, this is not what currently occurs in the context of nested parallel regions. Specifically to the OpenMPIRBuilder,
the entirety of the function (at the LLVM level) is currently inlined with blocks marking the corresponding start and end of each
region.
```
entry:
...
parallel1:
%a = ...
...
parallel2:
use(%a)
...
endparallel2:
...
endparallel1:
...
```
When the allocation is inserted, it presently inserted into the parent of the entire function (e.g. entry) rather than the parent
allocation scope to the function being outlined. If we were outlining parallel2, the corresponding alloca location would be parallel1.
This causes a variety of bugs, including https://github.com/llvm/llvm-project/issues/54165 as one example.
This PR allows the stack allocation to be created at the correct allocation block, and thus remedies such issues.
Reviewed By: jdoerfert
Differential Revision: https://reviews.llvm.org/D121061
We previously had logic to disable pthreads, set the ThreadModel to Single, and
disable thread-safe statics when the atomics target features is disabled, since
that means that the resulting program will not be used in a threaded context.
Similarly check for the presence of the bulk-memory feature, since that is also
necessary to produce multithreaded programs.
Differential Revision: https://reviews.llvm.org/D121014
When both CUDA or HIP programs and C++ programs are passed
to clang driver without -c, C++ programs are treated as CUDA
or HIP program, which is incorrect.
This is because action builder sets the offloading kind of input
job actions to the linking action to be the union of offloading
kind of the input job actions, i.e. if there is one HIP or CUDA
input to the linker, then all the input to the linker is marked
as HIP or CUDA.
To fix this issue, the offload action builder tracks the originating
input argument of each host action, which allows it to determine
the active offload kind of each host action. Then the offload
kind of each input action to the linker can be determined
individually.
Reviewed by: Artem Belevich
Differential Revision: https://reviews.llvm.org/D120911
When both HIP and C++ programs are input files to clang
with -c, clang treats C++ programs as HIP programs,
which is incorrect.
This is due to action builder does not set correct
offloading kind for job actions for C++ programs.
Reviewed by: Artem Belevich
Differential Revision: https://reviews.llvm.org/D120910
This commit divides the large test files(over 30k lines) under clang/test/CodeGen/RISCV including:
rvv-intrinsics/vloxseg.c
rvv-intrinsics/vluxseg.c
rvv-intrinsics-overloaded/vloxseg.c
rvv-intrinsics-overloaded/vluxseg.c
into "non-masked" version and "masked" version which can reduce the test cases by 50% in a single file.
Differential Revision: https://reviews.llvm.org/D120967
This test file has grown to the point where it takes a huge amount of
time to run. At the moment, this test seems to consistently time out
when running in the pre-commit checks in Phabricator with a 10 minute
timeout. For example see
https://reviews.llvm.org/harbormaster/unit/view/2832724/
While splitting up the test file is not ideal, it is even more
undesirable to have huge test files that time out in common settings.
This patch splits up the test file roughly in the middle.
Reviewed By: dmgreen
Differential Revision: https://reviews.llvm.org/D120876
This test file has grown to the point where it takes a huge amount of
time to run. At the moment, this test seems to consistently time out
when running in the pre-commit checks in Phabricator with a 10 minute
timeout. For example see
https://reviews.llvm.org/harbormaster/unit/view/2832723/
While splitting up the test file is not ideal, it is even more
undesirable to have huge test files that time out in common settings.
This patch splits up the test file roughly in the middle.
Reviewed By: dmgreen
Differential Revision: https://reviews.llvm.org/D120875
The baremetal-sysroot test fails when the toolchain is configured with
DEFAULT_SYSROOT. So, to emulate not having passed one at all, let's
pass an empty sysroot instead.
https://reviews.llvm.org/D119144
Patch by Carlo Cabrera <carlo.antonio.cabrera@gmail.com>
Few weeks back I was experimenting with reading the uninitialized values from src , which is actually a bug but the CSA seems to give up at that point . I was curious about that and I pinged @steakhal on the discord and according to him this seems to be a genuine issue and needs to be fix. So I goes with fixing this bug and thanks to @steakhal who help me creating this patch. This feature seems to break some tests but this was the genuine problem and the broken tests also needs to fix in certain manner. I add a test but yeah we need more tests,I'll try to add more tests.Thanks
Reviewed By: steakhal, NoQ
Differential Revision: https://reviews.llvm.org/D120489
The standard requires[0] member function constraints to be checked when
explicitly instantiating classes. This patch adds this constraints
check.
This issue is tracked as #46029 [1].
Note that there's an related open CWG issue (2421[2]) about what to do when
multiple candidates have satisfied constraints. This is particularly an
issue because mangling doesn't contain function constraints, and so the
following code still ICEs with definition with same mangled name
'_ZN1BIiE1fEv' as another definition:
template<class T>
struct B {
int f() requires std::same_as<T, int> {
return 0;
}
int f() requires (std::same_as<T, int> &&
!std::same_as<T, char>) {
return 1;
}
};
template struct B<int>;
Also note that the constraints checking while instantiating *functions*
is still not implemented. I started looking at it but It's a bit more
complicated. I believe in such a case we have to consider the partial
constraints order and potentially choose the best candidate out of the
set of multiple valid ones.
[0]: https://eel.is/c++draft/temp.explicit#10
[1]: https://github.com/llvm/llvm-project/issues/46029
[2]: https://cplusplus.github.io/CWG/issues/2421.html
Differential Revision: https://reviews.llvm.org/D120255
The problem with leak bug reports is that the most interesting event in the code
is likely the one that did not happen -- lack of ownership change and lack of
deallocation, which is often present within the same function that the analyzer
inlined anyway, but not on the path of execution on which the bug occured. We
struggle to understand that a function was responsible for freeing the memory,
but failed.
D105819 added a new visitor to improve memory leak bug reports. In addition to
inspecting the ExplodedNodes of the bug pat, the visitor tries to guess whether
the function was supposed to free memory, but failed to. Initially (in D108753),
this was done by checking whether a CXXDeleteExpr is present in the function. If
so, we assume that the function was at least party responsible, and prevent the
analyzer from pruning bug report notes in it. This patch improves this heuristic
by recognizing all deallocator functions that MallocChecker itself recognizes,
by reusing MallocChecker::isFreeingCall.
Differential Revision: https://reviews.llvm.org/D118880
Previously, we didin't build a DeclRefExpr which refers to an invalid declaration.
In this patch, we handle this case by building an empty RecoveryExpr,
which will preserve more broken code (AST parent nodes that contain the
RecoveryExpr is preserved in the AST).
Differential Revision: https://reviews.llvm.org/D120812
Clang is crashing on the following statement
char var[9];
__asm__ ("" : "=r" (var) : "0" (var));
This is similar to existing test: crbug_999160_regtest
The issue happens when EmitAsmStmt is trying to convert input to match
output type length. However, that is not guaranteed to be successful all the
time and if the statement itself is invalid like having an array type in
the example, we should give a regular error message here instead of
using assert().
Reviewed By: MaskRay
Differential Revision: https://reviews.llvm.org/D120596
`hip-openmp-compatible` flag treats hip and hipv4 offload kinds
as compatible with openmp offload kind while extracting code objects
from a heterogenous archive library. Vice versa is also considered
compatible if hip code was compiled with -fgpu-rdc.
This flag only relaxes compatibility criteria on `OffloadKind`,
rest of the components like `Triple` and `GPUArhc` still needs to
be compatible.
Reviewed By: yaxunl
Differential Revision: https://reviews.llvm.org/D120697
C++20 non-type template parameter prints `MyType<{{116, 104, 105, 115}}>` when the code is as simple as `MyType<"this">`. This patch prints `MyType<{"this"}>`, with one layer of braces preserved for the intermediate structural type to trigger CTAD.
`StringLiteral` handles this case, but `StringLiteral` inside `APValue` code looks like a circular dependency. The proposed patch implements a cheap strategy to emit string literals in diagnostic messages only when they are readable and fall back to integer sequences.
Reviewed By: aaron.ballman
Differential Revision: https://reviews.llvm.org/D115031
NOTE: this is a follow-up commit with the missing clang-side changes.
This patch adds builtins and intrinsics for the f16 and f16x2 variants of the ex2
instruction.
These two variants were added in PTX7.0, and are supported by sm_75 and above.
Note that this isn't wired with the exp2 llvm intrinsic because the ex2
instruction is only available in its approx variant.
Running ptxas on the assembly generated by the test f16-ex2.ll works as
expected.
Differential Revision: https://reviews.llvm.org/D119157
Currently adding attribute no_sanitize("bounds") isn't disabling
-fsanitize=local-bounds (also enabled in -fsanitize=bounds). The Clang
frontend handles fsanitize=array-bounds which can already be disabled by
no_sanitize("bounds"). However, instrumentation added by the
BoundsChecking pass in the middle-end cannot be disabled by the
attribute.
The fix is very similar to D102772 that added the ability to selectively
disable sanitizer pass on certain functions.
In this patch, if no_sanitize("bounds") is provided, an additional
function attribute (NoSanitizeBounds) is attached to IR to let the
BoundsChecking pass know we want to disable local-bounds checking. In
order to support this feature, the IR is extended (similar to D102772)
to make Clang able to preserve the information and let BoundsChecking
pass know bounds checking is disabled for certain function.
Reviewed By: melver
Differential Revision: https://reviews.llvm.org/D119816
This new flag enables `__has_feature(cxx_unstable)` that would replace libc++ macros for individual unstable/experimental features, e.g. `_LIBCPP_HAS_NO_INCOMPLETE_RANGES` or `_LIBCPP_HAS_NO_INCOMPLETE_FORMAT`.
This would make it easier and more convenient to opt-in into all libc++ unstable features at once.
Differential Revision: https://reviews.llvm.org/D120160
NVVM IR specification defines them with i32 return type:
declare i32 @llvm.nvvm.match.any.sync.i64(i32 %membermask, i64 %value)
declare {i32, i1} @llvm.nvvm.match.all.sync.i64(i32 %membermask, i64 %value)
...
The i32 return value is a 32-bit mask where bit position in mask corresponds
to thread’s laneid.
as well as PTX ISA:
9.7.12.8. Parallel Synchronization and Communication Instructions: match.sync
match.any.sync.type d, a, membermask;
match.all.sync.type d[|p], a, membermask;
...
Destination d is a 32-bit mask where bit position in mask corresponds
to thread’s laneid.
Additionally, ptxas doesn't accept intructions, produced by NVPTX backend.
After this patch, it compiles with no issues.
Reviewed By: tra
Differential Revision: https://reviews.llvm.org/D120499
Implementation partitions bring two extra cases where we have
visibility of module-private data.
1) When we import a module implementation partition.
2) When a partition implementation imports the primary module intertace.
We maintain a record of direct imports into the current module since
partition decls from direct imports (but not trasitive ones) are visible.
The rules on decl-reachability are much more relaxed (with the standard
giving permission for an implementation to load dependent modules and for
the decls there to be reachable, but not visible).
Differential Revision: https://reviews.llvm.org/D118599
Add a checker to maintain the system-defined value 'errno'.
The value is supposed to be set in the future by existing or
new checkers that evaluate errno-modifying function calls.
Reviewed By: NoQ, steakhal
Differential Revision: https://reviews.llvm.org/D120310
Given a dependent `T` (maybe an undeduced `auto`),
Before:
new T(z) --> new T((z)) # changes meaning with more args
new T{z} --> new T{z}
T(z) --> T(z)
T{z} --> T({z}) # forbidden if T is auto
After:
new T(z) --> new T(z)
new T{z} --> new T{z}
T(z) --> T(z)
T{z} --> T{z}
Depends on D113393
Reviewed By: aaron.ballman
Differential Revision: https://reviews.llvm.org/D120608
Add applyStaticChunkedWorkshareLoop method implementing static schedule when chunk-size is specified. Unlike a static schedule without chunk-size (where chunk-size is chosen by the runtime such that each thread receives one chunk), we need two nested loops: one for looping over the iterations of a chunk, and a second for looping over all chunks assigned to the threads.
This patch includes the following related changes:
* Adapt applyWorkshareLoop to triage between the schedule types, now possible since all schedules have been implemented. The default schedule is assumed to be non-chunked static, as without OpenMPIRBuilder.
* Remove the chunk parameter from applyStaticWorkshareLoop, it is ignored by the runtime. Change the value for the value passed to the init function to 0, as without OpenMPIRBuilder.
* Refactor CanonicalLoopInfo::setTripCount and CanonicalLoopInfo::mapIndVar as used by both, applyStaticWorkshareLoop and applyStaticChunkedWorkshareLoop.
* Enable Clang to use the OpenMPIRBuilder in the presence of the schedule clause.
Differential Revision: https://reviews.llvm.org/D114413
Motivation:
```
int foo(int x, int y) { // any compiler will happily inline this function
return x / y;
}
int test(int x, int y) {
int r = 0;
[[clang::noinline]] r += foo(x, y); // for some reason we don't want any inlining here
return r;
}
```
In 2018, @kuhar proposed "Introduce per-callsite inline intrinsics" in https://reviews.llvm.org/D51200 to solve this motivation case (and many others).
This patch solves this problem with call site attribute. The implementation is "smaller" wrt approach which uses new intrinsics and thanks to https://reviews.llvm.org/D79121 (Add nomerge statement attribute to clang), we have got some basic infrastructure to deal with attrs on statements with call expressions.
GCC devs are more inclined to call attribute solution as well, as builtins are problematic for them - https://gcc.gnu.org/bugzilla/show_bug.cgi?id=104187. But they have no patch proposal yet so.. We have free hands here.
If this approach makes sense, next future steps would be support for call site attributes for always_inline / flatten.
Reviewed By: aaron.ballman, kuhar
Differential Revision: https://reviews.llvm.org/D119061
Otherwise, the driver will insert e.g. -lgcc_s when
CLANG_DEFAULT_UNWINDLIB=libgcc is set during the clang build.
Differential Revision: https://reviews.llvm.org/D120644
See post-commit discussion on https://reviews.llvm.org/D120305.
This change breaks the clang-ppc64le-rhel buildbot, though
there is suspicion that it's an issue with the bot. The change
also had a larger than expected impact on compile-time and
code-size.
This reverts commit 3c4ed02698
and some followup changes.
We cannot export partition implementation CMIs, but we can export the content
of partition interface CMIs.
Differential Revision: https://reviews.llvm.org/D118588
The purpose of this change is to fix the following codegen bug:
```
// main.c
__attribute__((cpu_specific(generic)))
int *foo(void) { static int z; return &z;}
int main() { return *foo() = 5; }
// other.c
__attribute__((cpu_dispatch(generic))) int *foo(void);
// run:
clang main.c other.c -o main; ./main
```
This will segfault prior to the change, and return the correct
exit code 5 after the change.
The underlying cause is that when a translation unit contains
a cpu_specific function without the corresponding cpu_dispatch
the generated code binds the reference to foo() against a
GlobalIFunc whose resolver is undefined. This is invalid: the
resolver must be defined in the same translation unit as the
ifunc, but historically the LLVM bitcode verifier did not check
that. The generated code then binds against the resolver rather
than the ifunc, so it ends up calling the resolver rather than
the resolvee. In the example above it treats its return value as
an int *, therefore trying to write to program text.
The root issue at the representation level is that GlobalIFunc,
like GlobalAlias, does not support a "declaration" state. The
object which provides the correct semantics in these cases
is a Function declaration, but unlike Functions, changing a
declaration to a definition in the GlobalIFunc case constitutes
a change of the object type, as opposed to simply emitting code
into a Function.
I think this limitation is unlikely to change, so I implemented
the fix by returning a function declaration rather than an ifunc
when encountering cpu_specific, and upgrading it to an ifunc
when emitting cpu_dispatch.
This uses `takeName` + `replaceAllUsesWith` in similar vein to
other places where the correct IR object type cannot be known
locally/up-front, like in `CodeGenModule::EmitAliasDefinition`.
Previous discussion in: https://reviews.llvm.org/D112349
Signed-off-by: Itay Bookstein <ibookstein@gmail.com>
Reviewed By: erichkeane
Differential Revision: https://reviews.llvm.org/D120266
* For `-###`, `-o %t.o` is unnecessary if we don't specifically test the
output filename.
* --target= is the canonical spelling. -target is a legacy spelling which
unfortunately cannot be removed because there are too many uses.
* -no-canonical-prefixes uses the dereferenced absolute path for the cc1
command. For most tests "-cc1" is sufficient to identify the command line, no
need to specifically test the "clang" command, and -no-canonical-prefixes can
removed.
* --unwindlib= is the preferred spelling. -u is a short option taking a value,
which means a -uwindlib= typo cannot be detected.
I recommend that you take a look at linux-cross.cpp. Testing include paths and
library paths in one RUN line is sometimes more readable than having separate
include/library tests.
Having separate RUN lines for misc features like -fdata-sections
(`aix-data-sections.c`) is wasteful. It may be better testing multiple
options in a single RUN command.
Reviewed By: jsji
Differential Revision: https://reviews.llvm.org/D119309
This fixes a bug that happens when using -fdebug-prefix-map to remap
an absolute path to a relative path. Since the path was absolute
before remapping, it is safe to assume that concatenating the remapped
working directory would be wrong.
Differential Revision: https://reviews.llvm.org/D113718
Changed the we handle llvm::Constants in sizes arrays. ConstExprs and
GlobalValues cannot be used as initializers, need to put them at the
runtime, otherwise there wight be the compilation errors.
Differential Revision: https://reviews.llvm.org/D105297
Value-dependent ConstantExprs are not meant to be evaluated.
There is an assert in Expr::EvaluateAsConstantExpr that ensures this condition.
But before this patch the method was called without prior check.
Fixes https://github.com/llvm/llvm-project/issues/52768
Reviewed By: erichkeane
Differential Revision: https://reviews.llvm.org/D119375
Exactly what it says on the tin! We had a nasty crash with the following incovation:
$ clang --analyze -Xclang -analyzer-constraints=z3 test.c
fatal error: error in backend: LLVM was not compiled with Z3 support, rebuild with -DLLVM_ENABLE_Z3_SOLVER=ON
... <stack trace> ...
Differential Revision: https://reviews.llvm.org/D120325
Add a checker to maintain the system-defined value 'errno'.
The value is supposed to be set in the future by existing or
new checkers that evaluate errno-modifying function calls.
Reviewed By: NoQ, steakhal
Differential Revision: https://reviews.llvm.org/D120310
Partition implementations are special, they generate a CMI, but it
does not have an 'export' line, and we cannot export anything from the
it [that is it can only make decls available to other members of the
owning module, not to importers of that].
Add initial testcases for partition handling, derived from the examples in
Section 10 of the C++20 standard, which identifies what should be accepted
and/or rejected.
Differential Revision: https://reviews.llvm.org/D118587
(resubmit https://reviews.llvm.org/D119207 after fixing the test for
some build settings)
This patch converts CUDA pointer kernel arguments with default address
space to CrossWorkGroup address space (__global in OpenCL). This is
because Generic or Function (OpenCL's private) is not supported as
storage class for kernel pointer types.
Differential revision: https://reviews.llvm.org/D120366
GCC's compiled with --enable-version-specific-runtime-libs change the paths where includes and libs are found.
This patch adds support for these cases
Reviewed By: MaskRay
Differential Revision: https://reviews.llvm.org/D118700
Default the option introduced in D113372 to ON to match all(?) major Linux
distros. This matches GCC and improves consistency with Android and linux-musl
which always default to PIE.
Note: CLANG_DEFAULT_PIE_ON_LINUX will be removed in the future.
Reviewed By: thesamesam
Differential Revision: https://reviews.llvm.org/D120305
The `objc_precise_lifetime` attribute is applied to Objective-C pointers to ensure the optimizer does not prematurely release an object under Automatic Reference Counting (ARC). It is a common enough pattern to assign values to these variables but not reference them otherwise, and annotating them with `__unused` is not really correct as they are being used to ensure an object's lifetime.
Differential Revision: https://reviews.llvm.org/D120372
This patch adds -Wno-strict-prototypes to all of the test cases that
use functions without prototypes, but not as the primary concern of the
test. e.g., attributes testing whether they can/cannot be applied to a
function without a prototype, etc.
This is done in preparation for enabling -Wstrict-prototypes by
default.
A significant number of our tests in C accidentally use functions
without prototypes. This patch converts the function signatures to have
a prototype for the situations where the test is not specific to K&R C
declarations. e.g.,
void func();
becomes
void func(void);
This is the final batch of tests being updated to add prototypes,
hopefully.
Changed the we handle llvm::Constants in sizes arrays. ConstExprs and
GlobalValues cannot be used as initializers, need to put them at the
runtime, otherwise there wight be the compilation errors.
Differential Revision: https://reviews.llvm.org/D105297
Summary:
We use a section to embed offloading code into the host for later
linking. This is normally unique to the translation unit as it is thrown
away during linking. However, if the user performs a relocatable link
the sections will be merged and we won't be able to access the files
stored inside. This patch changes the section variables to have external
linkage and a name defined by the section name, so if two sections are
combined during linking we get an error.
This adds a diagnostic when an unqualified call is resolved
to std::move or std::forward.
This follows some C++ committee discussions where some
people where concerns that this might be an usual anti pattern
particularly britle worth warning about - both because move
is a common name and because these functions accept any values.
This warns inconditionnally of whether the current context is in
std:: or not, as implementations probably want to always qualify
these calls too, to avoid triggering adl accidentally.
Differential Revision: https://reviews.llvm.org/D119670
This implements the parsing and recognition of module partition CMIs
and removes the FIXMEs in the parser.
Module partitions are recognised in the base computation of visibility,
however additional amendments to visibility follow in subsequent patches.
Differential Revision: https://reviews.llvm.org/D118586
Now that integer min/max intrinsics have good support in both
InstCombine and other passes, start canonicalizing SPF min/max
to intrinsic min/max.
Once this sticks, we can stop matching SPF min/max in various
places, and can remove hacks we have for preventing infinite loops
and breaking of SPF canonicalization.
Differential Revision: https://reviews.llvm.org/D98152
Introduce -fgpu-default-stream={legacy|per-thread} option to
support per-thread default stream for HIP runtime.
When -fgpu-default-stream=per-thread, HIP kernels are
launched through hipLaunchKernel_spt instead of
hipLaunchKernel. Also HIP_API_PER_THREAD_DEFAULT_STREAM=1
is defined by the preprocessor to enable other per-thread stream
API's.
Reviewed by: Artem Belevich
Differential Revision: https://reviews.llvm.org/D120298
-fdata-sections decides whether global variables go into different sections.
This is orthogonal to whether we place their metadata (`.data` or `asan_globals`) into different sections.
With -fno-data-sections, `-fsanitize-address-globals-dead-stripping` can still:
* deduplicate COMDAT `asan.module_ctor` and `asan.module_dtor`
* (with ld --gc-sections): for a data section (e.g. `.data`), if all global variables defined relative to it are unreferenced, discard them and associated `asan_globals` sections (rare but no need to exclude this case)
Similar to c7b90947bd for PE/COFF.
Reviewed By: #sanitizers, kstoimenov, vitalybuka
Differential Revision: https://reviews.llvm.org/D120394
Currently when we generate OpenMP offloading code we always make
fallback code for the CPU. This is necessary for implementing features
like conditional offloading and ensuring that unhandled pragmas don't
result in missing symbols. However, this is problematic for a few cases.
For offloading tests we can silently fail to the host without realizing
that offloading failed. Additionally, this makes it impossible to
provide interoperabiility to other offloading schemes like HIP or CUDA
because those methods do not provide any such host fallback guaruntee.
this patch adds the `-fopenmp-offload-mandatory` flag to prevent
generating the fallback symbol on the CPU and instead replaces the
function with a dummy global and the failed branch with 'unreachable'.
Reviewed By: ABataev
Differential Revision: https://reviews.llvm.org/D120353
It's customary for these options to have the -fno- form which is sometimes
handy to work around issues. Using the supported driver option is preferred over
the internal cl::opt option `-mllvm -asan-globals-live-support=0`
Reviewed By: kstoimenov, vitalybuka
Differential Revision: https://reviews.llvm.org/D120391
The TokenStream class is the representation of the source code that will
be fed into the GLR parser.
This patch allows a "raw" TokenStream to be built by reading source code.
It also supports scanning a TokenStream to find the directive structure.
Next steps (with placeholders in the code): heuristically choosing a
path through #ifs, preprocessing the code by stripping directives and comments.
These will produce a suitable stream to feed into the parser proper.
Differential Revision: https://reviews.llvm.org/D119162
The dependency scanner already generates canonical -cc1 command lines that can be used to compile discovered modular dependencies.
For translation unit command lines, the scanner only generates additional driver arguments the build system is expected to append to the original command line.
While this works most of the time, there are situations where that's not the case. For example with `-Wunused-command-line-argument`, Clang will complain about the `-fmodules-cache-path=` argument that's not being used in explicit modular builds. Combine that with `-Werror` and the build outright fails.
To prevent such failures, this patch changes the dependency scanner to return the full driver command line to compile the original translation unit. This gives us more opportunities to massage the arguments into something reasonable.
Reviewed By: Bigcheese
Differential Revision: https://reviews.llvm.org/D118986
This patch enables inferring framework modules in explicit builds in all contexts. Until now, inferring framework modules only worked with `-fimplicit-module-maps` due to this block of code:
```
// HeaderSearch::loadFrameworkModule
case LMM_InvalidModuleMap:
// Try to infer a module map from the framework directory.
if (HSOpts->ImplicitModuleMaps)
ModMap.inferFrameworkModule(Dir, IsSystem, /*Parent=*/nullptr);
break;
```
Reviewed By: Bigcheese
Differential Revision: https://reviews.llvm.org/D113880
Until now, subgroup builtins are available with `opencl-c.h` when at
least one of `cl_intel_subgroups`, `cl_khr_subgroups`, or
`__opencl_c_subgroups` is defined. With `-fdeclare-opencl-builtins`,
subgroup builtins are conditionalized on `cl_khr_subgroups` only.
Align `-fdeclare-opencl-builtins` to `opencl-c.h` by introducing the
internal `__opencl_subgroup_builtins` macro.
Differential Revision: https://reviews.llvm.org/D120254
This switches the testcase committed for initial C++20 modules import tracking to
use split-file rather than preprocessor directives.
Differential Revision: https://reviews.llvm.org/D120352
It is useful to be able to visualise the C++20 modules content of a PCM file
both for inspection and for testing. In particular, when adding more module
types to support C++20 Partitions and Header Units, we would like to be able
to confirm that the output PCM has the intended structure.
The existing scheme for dumping data is restricted to the content of the AST
file control block, which does not include structural data beyond imports.
The change here makes use of the AST unit that is set up by BeginSourceFile
to query for the information on the primary and sub-modules. We can then
inspect each of these in turn, accounting for Global, Private, Imported and
Exported modules/fragments and then showing the sub-stucture of the main
module(s).
The disadvantage of this mechanism is that it has no easy method to control
the granularity of the output. Perhaps more detailed inspection would be
better handled by a stand-alone module inspection tool.
Differential Revision: https://reviews.llvm.org/D119823
This patch introduces a dense implementation of the LR parsing table, which is
used by LR parsers.
We build a SLR(1) parsing table from the LR(0) graph.
Statistics of the LR parsing table on the C++ spec grammar:
- number of states: 1449
- number of actions: 83069
- size of the table (bytes): 334928
Differential Revision: https://reviews.llvm.org/D118196
This is the driver part of D91605 <https://reviews.llvm.org/D91605>, a
workaround to allow direct calls to `__tls_get_addr` on Solaris/amd64.
Tested on `amd64-pc-solaris2.11` and `sparcv9-sun-solaris2.11`.
Differential Revision: https://reviews.llvm.org/D119829
Otherwise callers of these functions have to check both the return value
for and the contents of the returned llvm::Optional.
Fixes#53742
Differential Revision: https://reviews.llvm.org/D119525
Summary:
This patch adds checks that were missing in clang for Armv8.5/6/7-A. These include:
* ACLE macro defines for AArch32.
* Handling of crypto and SM4, SHA and AES feature flags on clang's driver.
Reviewers: dmgreen, SjoerdMeijer, tmatheson
Differential Revision: https://reviews.llvm.org/D116153
We need to capture the local variables into a record in task untied
regions but clang does not support record with VLA data members.
Differential Revision: https://reviews.llvm.org/D99436
Iain Sandoe