Use sint_to_fp instead of select.
Reduce the number of branch instructions and
avoid generating TargetConstantPool for double.

(select cc, 1.0, 0.0) -> (sint_to_fp (zext cc))
https://alive2.llvm.org/ce/z/aoEcd9
https://godbolt.org/z/n543Y9v3e

(select cc, 0.0, 1.0) -> (sint_to_fp (zext (xor cc, 1)))
https://alive2.llvm.org/ce/z/zngvSB

Reviewed By: craig.topper

Differential Revision: https://reviews.llvm.org/D151719

Also add missing marking to the test of related issue 621.

https://cplusplus.github.io/CWG/issues/621.html
https://cplusplus.github.io/CWG/issues/873.html

Reviewed By: #clang-language-wg, shafik

Differential Revision: https://reviews.llvm.org/D151704

Use llvm::raw_ostream::indent().

The revision adds the alias analysis and access group
interfaces to the call operation.

Reviewed By: Dinistro

Differential Revision: https://reviews.llvm.org/D151693

This patch adds the Zvfhmin extension for clang.

Reviewed By: craig.topper, michaelmaitland

Differential Revision: https://reviews.llvm.org/D150253

class VLIWSchedBoundary manages resources such as dynamically allocated memory, it's generally a good practice to either implement a custom copy constructor or disable the default one.

Reviewed By: JamesNagurne

Differential Revision: https://reviews.llvm.org/D151699

This patch makes the variables names for callback functions more consistent. Changes no functionality.

Reviewed By: vitalybuka

Differential Revision: https://reviews.llvm.org/D151605

This should fix the Clang CI.

[RISCV] Merge emitDirectiveOptionArchPlus and emitDirectiveOptionArchMinus into a single interface. NFC

Probably going to do some other refactors after this, but this one
was easy and clearly reduces duplicate code.

Reviewed By: StephenFan

Differential Revision: https://reviews.llvm.org/D151771

Show line numbers to the left of diagnostic code snippets and increase
the numbers of lines shown from 1 to 16.

Differential Revision: https://reviews.llvm.org/D147875

I forgot to request a regression test in review of D151429, so adding one myself.

Revert "Fix -u option in dsymutil, to not emit an extra DW_LNE_set_address if the original line table was empty"

This reverts commit 9bd3ff8b.

Reverting because of test failures:

TEST 'LLVM :: tools/dsymutil/ARM/fat-dylib-update.test' FAILED

Command Output (stderr):
--
+ : 'RUN: at line 2'
+ /b/ml-opt-rel-x86-64-b1/build/bin/dsymutil -oso-prepend-path /b/ml-opt-rel-x86-64-b1/llvm-project/llvm/test/tools/dsymutil/ARM/.. /b/ml-opt-rel-x86-64-b1/llvm-project/llvm/test/tools/dsymutil/ARM/../Inputs/fat-test.arm.dylib -o /b/ml-opt-rel-x86-64-b1/build/test/tools/dsymutil/ARM/Output/fat-dylib-update.test.tmp.dSYM
warning: /b/ml-opt-rel-x86-64-b1/llvm-project/llvm/test/tools/dsymutil/ARM/../Inputs/fat-test.arm.o: timestamp mismatch between object file (2023-05-24 23:42:39.442778779) and debug map (2015-08-05 21:31:26.000000000)
warning: /b/ml-opt-rel-x86-64-b1/llvm-project/llvm/test/tools/dsymutil/ARM/../Inputs/fat-test.arm.o: timestamp mismatch between object file (2023-05-24 23:42:39.442778779) and debug map (2015-08-05 21:31:26.000000000)
error: lipo: No such file or directory

github issue: https://github.com/llvm/llvm-project/issues/62912

Reviewed By: jansvoboda11

Differential Revision: https://reviews.llvm.org/D151429

This patch changes the argument type to `HeaderSearch::LookupFile()` from `const DirectoryEntry *` to `DirectoryEntryRef` in order to remove some calls to the deprecated `DirectoryEntry::getName()`.

Depends on D127660.

Reviewed By: bnbarham, benlangmuir

Differential Revision: https://reviews.llvm.org/D127663

Fix -u option in dsymutil, to not emit an extra DW_LNE_set_address if the original line table was empty

With dsymutil's -u option, only the accelerator tables should be
updated, but with https://reviews.llvm.org/D150554 the -u option will
still re-generate the line table. If the line table was empty, that is,
it was a dummy line table, with no entries in it, dsymutil will always
generate a line table with a DW_LNE_end_sequence, a funky side effect of
this is that when the line table is re-generated, it will always emit a
DW_LNE_set_address first, which will change the line table total size.
This patch addresses this by making sure that if all the line table has
in it is a DW_LNE_end_sequence, it is the same as a dummy entry.

Differential Revision: https://reviews.llvm.org/D151579

Most users of `Module::Header` already assume its `Entry` is populated. Enforce this assumption in the type system and handle the only case where this is not the case by wrapping the whole struct in `std::optional`. Do the same for `Module::DirectoryName`.

Depends on D151584.

Reviewed By: benlangmuir

Differential Revision: https://reviews.llvm.org/D151586

Currently when clang fails to deduce auto return type of a function,
it does not emit any notes about why it fails. This causes difficulty
for users to fix such errors.

Actually, clang already generates the information for emitting notes
about the failed deduction. There is a TODO for actually emitting
them.

This patch tries to implement the TODO. Basically it passes the
failed template specialization candidate set from the point of
specialization failure back to the point where the deduction starts.

It is not comprehensive but would be a start for further improvement.

Reviewed by: Richard Smith, Matheus Izvekov

Differential Revision: https://reviews.llvm.org/D150212

Fixes: SWDEV-354278

class CodeViewContext manages resources such as dynamically allocated memory, it's generally a good practice to either implement a custom copy constructor or disable the default one.

Reviewed By: skan

Differential Revision: https://reviews.llvm.org/D151695

class SchedBoundary manages resources such as dynamically allocated memory, it's generally a good practice to either implement a custom copy constructor or disable the default one.

Reviewed By: pengfei

Differential Revision: https://reviews.llvm.org/D151686

Breaks Windows.

This reverts commit 8ac08472.

DeduceTemplateArgumentsByTypeMatch() returns null value which is dereferenced without checking since getAsIncompleteArrayType() returns nullptr and we are dereferencing null pointer value for S.Context->getAsIncompleteArrayType(P) when calling getElementType().

This patch adds an assert.

Reviewed By: erichkeane

Differential Revision: https://reviews.llvm.org/D151529

Make unmasked vp_ftrunc, vp_fceil, vp_floor and vp_fround select to unmasked instruction.

Reviewed By: frasercrmck

Differential Revision: https://reviews.llvm.org/D151676

This patch uses castAs instead of getAs which will assert if the type
doesn't match in clang::Type::getRVVEltType(clang::ASTContext const &)

Reviewed By: erichkeane

Differential Revision: https://reviews.llvm.org/D151721

`DataAggregator::recordTrace` serves two purposes:
  - Attaching LBR fallthrough ("trace") information to CFG (`getBranchInfo`),
    which eventually gets emitted as YAML profile.
  - Populating vector of offsets that gets added to `FuncBranchData`, which
    eventually gets emitted as fdata profile.

`recordTrace` is invoked from `getFallthroughsInTrace` which checks its return
status and passes on the collected vector of offsets to `doTrace`.

However, if a malformed trace is passed to `recordTrace` it might partially
attach the profile to CFG and exit with false, not propagating the vector of
offsets to `doTrace`. This leads to a difference between fdata and yaml profile
collected from the same binary and the same perf file.

(Skylake LBR errata might produce such malformed traces where the last entry
is duplicated, resulting in invalid fallthrough path between the last two
entries).

There are two ways to handle this mismatch: c...

With D148542, we ran into the following libc++ build error when using musl libc.

```
.../musl/include/bits/alltypes.h:354:16:
error: definition of type '__mbstate_t' conflicts with typedef of the same name
typedef struct __mbstate_t { unsigned __opaque1, __opaque2; } mbstate_t;
               ^
.../sysroot/usr/include/bits/types/__mbstate_t.h:21:3: note: '__mbstate_t' declared here
} __mbstate_t;
  ^
1 error generated.
```
This is because the mbstate_t definition in musl libc conflicts with the one
from "bits/types/mbstate_t.h", and this patch attempts to fix this build issue
when musl libc is used.

Reviewed By: iana

Differential Revision: https://reviews.llvm.org/D151740

This patch makes the following changes to `SparseTensorDimSliceAttr` methods:
* Mark `isDynamic` constexpr.
* Add new helpers `getStatic` and `getStaticString` to avoid repetition.
* Moved the definitions for `getStatic{Offset,Stride,Size}` and `isCompletelyDynamic` out of the class declaration; because there's no benefit to inlining them.
* Changed `parse` to use `kDynamic` rather than literals.
* Changed `verify` to use the `isDynamic` helper.

Reviewed By: aartbik

Differential Revision: https://reviews.llvm.org/D150919

This was just a thinko. The API StackFrame::GetVariableList takes a
bool for "get_file_globals" which if true will also find file statics
and file globals. But we only were passing that as true if the
ValueType was eValueTypeVariableGlobal, which meant that we never find
file statics. It's okay if we cast too wide a net when we do
GetVariableList as later on we check against the ValueType to filter
globals from statics.

There was a test that had a whole bunch of globals and tested
FindValue on all of them, but had no statics. So I just made one of
the globals a file static, which verifies the fix.

Differential Revision: https://reviews.llvm.org/D151392

Currently compiler assert when passing variable "memspace" in
omp_init_allocator.

omp_allocator_handle_t alloc=omp_init_allocator(memspace,1,traits)

The problem is memspace is not mapping to the target region.  During
the call to emitAllocatorInit, calls to EmitVarDecl for "alloc", then
emit initialization of "alloc" that cause to assert.

If I understant correct, it is not necessary to emit variable
initialization, since "allocator" is private to target region.

To fix this call CGF.EmitAutoVarAlloca(allocator) instead
CGF.EmitVarDecl(allocator).

Differential Revision: https://reviews.llvm.org/D151743