Age | Commit message (Collapse) | Author | Files | Lines |
|
Appeases bots.
|
|
Before this patch all of the 'target', 'target_version' and
'target_clones' attributes were sharing a common mangling logic across
different targets. However we would like to differenciate this logic,
therefore I have moved the default path to ABIInfo and provided
overrides for AArch64. This way we can resolve feature aliases without
affecting the name mangling. The PR #80540 demonstrates a motivating
case.
|
|
This patch takes in json files as input to determine that header files
to process, and in which order, to pass along for CC1 invocations. This
patch also includes an ASTVisitor to collect simple global variables.
|
|
In Sony's document processing, the first "paragraph" of the description
is duplicated into a Brief Description section. Add paragraph breaks to make
those brief descriptions less verbose.
In fmaintrin.h we were including the \code blocks, which are
inappropriate for a brief description. While I was in there, change
\code to \code{.operation} which the Intel document processing wants.
|
|
The F.isFinite() check here was introduced back when the first
floating point implementation was pushed, but I don't think it's
necessary and it fixes the attached test case.
|
|
Just don't do anything and let later operations handle the diagnostics.
|
|
We can call getBase() for pointers where Base != Offset as well,
for example when we've added a constant to the Offset.
|
|
(#82476)"
This reverts commit ffe7049b543adb9739261d28a60d4a47a00aa2e0.
This commit breaks on e.g. arm:
Example:
https://lab.llvm.org/buildbot/#/builders/245/builds/21177/steps/5/logs/FAIL__Clang__stream_c
```
******************** TEST 'Clang :: Analysis/stream.c' FAILED ********************
Exit Code: 1
Command Output (stderr):
--
RUN: at line 1: /home/tcwg-buildbot/worker/clang-armv8-quick/stage1/bin/clang -cc1 -internal-isystem /home/tcwg-buildbot/worker/clang-armv8-quick/stage1/lib/clang/19/include -nostdsysteminc -analyze -analyzer-constraints=range -setup-static-analyzer -analyzer-checker=core,alpha.unix.Stream,debug.ExprInspection -verify /home/tcwg-buildbot/worker/clang-armv8-quick/llvm/clang/test/Analysis/stream.c
+ /home/tcwg-buildbot/worker/clang-armv8-quick/stage1/bin/clang -cc1 -internal-isystem /home/tcwg-buildbot/worker/clang-armv8-quick/stage1/lib/clang/19/include -nostdsysteminc -analyze -analyzer-constraints=range -setup-static-analyzer -analyzer-checker=core,alpha.unix.Stream,debug.ExprInspection -verify /home/tcwg-buildbot/worker/clang-armv8-quick/llvm/clang/test/Analysis/stream.c
error: 'expected-warning' diagnostics expected but not seen:
File /home/tcwg-buildbot/worker/clang-armv8-quick/llvm/clang/test/Analysis/stream.c Line 147: Stream pointer might be NULL
File /home/tcwg-buildbot/worker/clang-armv8-quick/llvm/clang/test/Analysis/stream.c Line 153: Stream pointer might be NULL
error: 'expected-warning' diagnostics seen but not expected:
File /home/tcwg-buildbot/worker/clang-armv8-quick/llvm/clang/test/Analysis/stream.c Line 148: Stream pointer might be NULL [alpha.unix.Stream]
File /home/tcwg-buildbot/worker/clang-armv8-quick/llvm/clang/test/Analysis/stream.c Line 154: Stream pointer might be NULL [alpha.unix.Stream]
4 errors generated.
--
********************
```
|
|
llvm-project/clang/lib/CodeGen/CGCall.cpp:3226:24:
error: unused variable 'StructSize' [-Werror,-Wunused-variable]
llvm::TypeSize StructSize = CGM.getDataLayout().getTypeAllocSize(STy);
^
llvm-project/clang/lib/CodeGen/CGCall.cpp:3227:24:
error: unused variable 'PtrElementSize' [-Werror,-Wunused-variable]
llvm::TypeSize PtrElementSize =
^
llvm-project/clang/lib/CodeGen/CGCall.cpp:5313:24:
error: unused variable 'SrcTypeSize' [-Werror,-Wunused-variable]
llvm::TypeSize SrcTypeSize =
^
llvm-project/clang/lib/CodeGen/CGCall.cpp:5315:24:
error: unused variable 'DstTypeSize' [-Werror,-Wunused-variable]
llvm::TypeSize DstTypeSize = CGM.getDataLayout().getTypeAllocSize(STy);
^
4 errors generated.
|
|
|
|
docs (#81437)
When parsing the ASTMatchers.h file, a matcher could specify an argument
that is a matcher using the not needed namespace
`(clang::)ast_matchers::`.
Change the argument parsing in dump_ast_matchers.py to remove those
namespaces such that when parameters with these namespaces slip through,
the namespaces will be not be shown in the matchers reference, like it
is done with the `internal` namespace.
Additionally, remove the not needed namespaces from arguments in
ASTMatchers.h.
|
|
Model `getc` and `putc` as equivalent to `fgetc` and `fputc` respectively.
Model `vfscanf` and `vfprintf` as `fscanf` and `fprintf`, except that
`vfscanf` can not invalidate the parameters due to the indirection via a
`va_list`. Nevertheless, we can still track EOF and errors as for `fscanf`.
|
|
|
|
|
|
We used to run them, but not check if they failed. If they do,
the expression is invalid, even if we already have a result.
I do have a suspicion that we need to manually call destroyLocals()
in more places (everywhere basically?), but I'll wait with that
until I have a reproducer at hand.
|
|
trivial function. (#82291)
|
|
Emitting the basic block address map with
`-fbasic-block-sections=labels` is allowed for AArch64 ELF since
7eaf94fefa1250fc8a46982cea8ce99abacae11f. Allow doing so with
`-fbasic-block-address-map`.
|
|
5." (#83214)
Reverts llvm/llvm-project#82840
|
|
This reverts commit 55783bd0f9cfc30aa93c718919dab5419d86a2c6.
Due to regressions in hipCUB.
hipCUB/hipcub/include/hipcub/backend/rocprim/device/device_spmv.hpp:142:33: error: call to 'min' is ambiguous
https://github.com/ROCm/hipCUB/blob/develop/hipcub/include/hipcub/backend/rocprim/device/device_spmv.hpp#L142
The ambuguity seems due to missing min(int, unsigned int).
Previously, there is only min(int, int). After the change,
there are min(int, int) and min(unsigned int, unsigned int),
therefore there is ambiguity.
|
|
Updates the extension feature `define-target-os-macros` to support the
recently-added XROS target (TARGET_OS_VISION).
|
|
At the moment, clang generates what I believe are incorrect !tbaa.struct
fields for named bitfields. At the moment, the base type size is used
for named bifields (e.g. sizeof(int)) instead of the bifield width per
field. This results in overalpping fields in !tbaa.struct metadata.
This causes incorrect results when extracting individual copied fields
from !tbaa.struct as in added in dc85719d5.
This patch fixes that by skipping by combining adjacent bitfields
in fields with correct sizes.
Fixes https://github.com/llvm/llvm-project/issues/82586
|
|
When analysing whether we should handle a binary expression as an
overloaded operator call or a builtin operator, we were calling
`checkPlaceholderForOverload()`, which takes care of any placeholders
that are not overload sets—which would usually make sense since those
need to be handled as part of overload resolution.
Unfortunately, we were also doing that for `.*`, which is not
overloadable, and then proceeding to create a builtin operator anyway,
which would crash if the RHS happened to be an unresolved overload set
(due hitting an assertion in `CreateBuiltinBinOp()`—specifically, in one
of its callees—in the `.*` case that makes sure its arguments aren’t
placeholders).
This pr instead makes it so we check for *all* placeholders early if the
operator is `.*`.
It’s worth noting that,
1. In the `.*` case, we now additionally also check for *any*
placeholders (not just non-overload-sets) in the LHS; this shouldn’t
make a difference, however—at least I couldn’t think of a way to trigger
the assertion with an overload set as the LHS of `.*`; it is worth
noting that the assertion in question would also complain if the LHS
happened to be of placeholder type, though.
2. There is another case in which we also don’t perform overload
resolution—namely `=` if the LHS is not of class or enumeration type
after handling non-overload-set placeholders—as in the `.*` case, but
similarly to 1., I first couldn’t think of a way of getting this case to
crash, and secondly, `CreateBuiltinBinOp()` doesn’t seem to care about
placeholders in the LHS or RHS in the `=` case (from what I can tell,
it, or rather one of its callees, only checks that the LHS is not a
pseudo-object type, but those will have already been handled by the call
to `checkPlaceholderForOverload()` by the time we get to this function),
so I don’t think this case suffers from the same problem.
This fixes #53815.
---------
Co-authored-by: Aaron Ballman <aaron@aaronballman.com>
|
|
ISO/IEC TR 18037 defines %r, %R, %k, and %K for fixed point format
specifiers. -Wformat should not warn on these when they are provided.
|
|
fallback archive format. (#82888)
|
|
Unnamed bitfields need to be ignored here.
|
|
Evaluate all the semantic expressions.
|
|
Same fix we had for global ones: leave a pointer on the stack.
|
|
CUDA defines min/max functions for host in global namespace.
HIP header needs to define them too to be compatible.
Currently only min/max(int, int) is defined. This causes
wrong result for arguments that are out of range for int.
This patch defines host min/max functions to be compatible
with CUDA.
Also allows users to define
`__HIP_NO_HOST_MIN_MAX_IN_GLOBAL_NAMESPACE__` to disable
host max/min in global namespace.
min/max functions with mixed signed/unsigned integer
parameters are not defined unless
`__HIP_DEFINE_MIXED_HOST_MIN_MAX__` is defined.
Fixes: SWDEV-446564
|
|
The old name clashes with the Init opcode.
|
|
When -gsplit-dwarf is passed in clang emmmits -ggnu-pubnames which
results in
.debug_gnu_pubnames/..debug_gnu_pubtypes being generated.
This is used by GDB, but not by LLDB.
Changed so that these sections are not emitted for LLDB tuning.
|
|
(#81418)
…C mode
Factored logic from `CheckImplicitConversion` into new methods
`Expr::getEnumConstantDecl` and `Expr::getEnumCoercedType` for use in
`checkEnumArithmeticConversions`.
Fix #29217
|
|
Noticed while implementing #82536 that this test was also missing the
call the FileCheck.
|
|
To align TableGen consecutive definitions.
|
|
|
|
I'm not aware of any remaining overrides of this function.
While I'm here, change an outdated comment in DataflowAnalysis.h that
still
referred to `merge()`. I've made the comment more general, referring
simply to
`ValueModel`, as we shouldn't really be repeating the documentation of
that
class here anyway.
|
|
This include 2 fixes:
1. Disallow 'f' for softfloat.
2. Allow 'r' for softfloat.
Currently, 'f' is accpeted by clang, then LLVM meets an internal error.
'r' is rejected by LLVM by: couldn't allocate input reg for constraint
'r'.
Fixes: #64241, #63632
---------
Co-authored-by: Fangrui Song <i@maskray.me>
|
|
|
|
This fixes a few breakages introduced during upstreaming.
|
|
Extra tests with unnamed bitfields for
https://github.com/llvm/llvm-project/pull/82922.
|
|
Clang was incorrectly finding the start of the exponent in a fixed point
hex literal. It would unconditionally find the first `e/E/p/P` in a
constant regardless of if it were hex or not and parser the remaining
digits as an APInt. In a debug build, this would be caught by an
assertion, but in a release build, the assertion is removed and we'd end
up in an infinite loop.
Fixes #83050
|
|
This patch inserts 1-byte counters instead of an 8-byte counters into
llvm profiles for source-based code coverage. The origial idea was
proposed as block-cov for PGO, and this patch repurposes that idea for
coverage: https://groups.google.com/g/llvm-dev/c/r03Z6JoN7d4
The current 8-byte counters mechanism add counters to minimal regions,
and infer the counters in the remaining regions via adding or
subtracting counters. For example, it infers the counter in the if.else
region by subtracting the counters between if.entry and if.then regions
in an if statement. Whenever there is a control-flow merge, it adds the
counters from all the incoming regions. However, we are not going to be
able to infer counters by subtracting two execution counts when using
single-byte counters. Therefore, this patch conservatively inserts
additional counters for the cases where we need to add or subtract
counters.
RFC:
https://discourse.llvm.org/t/rfc-single-byte-counters-for-source-based-code-coverage/75685
|
|
There are two issues here. first `ICK_Floating_Integral` were always
defaulting to `CK_FloatingToIntegral` for vectors regardless of
direction of cast. Check was scalar only so added a vec float check to
the conditional.
Second issue was float to int casts were resolving to
ICK_Integral_Promotion when they need to be resolving to
CK_FloatingToIntegral. This was fixed by changing the ordering of
conversion checks.
This fixes #82826
|
|
Fixes #82058.
|
|
incomplete context (#83024)
When compiling the following with `-fms-compatibility`:
```
template<typename T> struct C;
// Test lookup with incomplete lookup context
template<typename T>
auto C<T>::f() -> decltype(x) { }
```
An assert fails because `CXXRecordDecl::hasAnyDependentBases` is called
on an incomplete class. This patch ensures we don't perform unqualified
lookup into dependent base classes when the lookup context is
incomplete.
|
|
Follow-up for #81971 to fix the disabled LIT test and add
LIT tests for lowering of the added math intrinsics.
|
|
BSD, Linux, and z/OS enable `LLVM_ENABLE_PER_TARGET_RUNTIME_DIR` by
default.
When a compiler-rt library is not found, we currently report an
incorrect filename `libclang_rt.XXX-$arch.a`
```
% /tmp/Debug/bin/clang++ a.cc -fsanitize=address -o a
ld.lld: error: cannot open /tmp/Debug/lib/clang/19/lib/linux/libclang_rt.asan-x86_64.a: No such file or directory
clang++: error: linker command failed with exit code 1 (use -v to see invocation)
```
With this change, we will correctly report:
```
% /tmp/Debug/bin/clang++ a.cc -fsanitize=address -o a
ld.lld: error: cannot open /tmp/Debug/lib/clang/19/lib/x86_64-unknown-linux-gnu/libclang_rt.asan.a: No such file or directory
clang++: error: linker command failed with exit code 1 (use -v to see invocation)
```
Link: https://discourse.llvm.org/t/runtime-directory-fallback/76860
|
|
I called this function when investigating the issue
(https://github.com/llvm/llvm-project/issues/78131), and I was surprised
to see the definition is commented out.
I think it makes sense to provide the definition even though the
implementation is not stable.
|
|
The second parameter needs to be a uint64_t and nothing else.
This broke windows builders, see
https://github.com/llvm/llvm-project/commit/264d828ea6399c31c210b67a050fbf084634da6a
|
|
This change implements https://github.com/llvm/llvm-project/issues/70073
HLSL has a dot intrinsic defined here:
https://learn.microsoft.com/en-us/windows/win32/direct3dhlsl/dx-graphics-hlsl-dot
The intrinsic itself is defined as a HLSL_LANG LangBuiltin in
Builtins.td.
This is used to associate all the dot product typdef defined
hlsl_intrinsics.h
with a single intrinsic check in CGBuiltin.cpp & SemaChecking.cpp.
In IntrinsicsDirectX.td we define the llvmIR for the dot product.
A few goals were in mind for this IR. First it should operate on only
vectors. Second the return type should be the vector element type. Third
the second parameter vector should be of the same size as the first
parameter. Finally `a dot b` should be the same as `b dot a`.
In CGBuiltin.cpp hlsl has built on top of existing clang intrinsics via
EmitBuiltinExpr. Dot
product though is language specific intrinsic and so is guarded behind
getLangOpts().HLSL.
The call chain looks like this: EmitBuiltinExpr -> EmitHLSLBuiltinExp
EmitHLSLBuiltinExp dot product intrinsics makes a destinction
between vectors and scalars. This is because HLSL supports dot product
on scalars which simplifies down to multiply.
Sema.h & SemaChecking.cpp saw the addition of
CheckHLSLBuiltinFunctionCall, a language specific semantic validation
that can be expanded for other hlsl specific intrinsics.
Fixes #70073
|
|
No functional changes intended.
Fixes some leaks found by running under asan with `--gtest_repeat=2`.
|