| Commit message (Collapse) | Author | Age | Files | Lines |
... | |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
more reliably.
Clang has two different ways it emits array constants (from InitListExprs and
from APValues), and both had some ability to emit zeroinitializer, but neither
was able to catch all cases where we could use zeroinitializer reliably. In
particular, emitting from an APValue would fail to notice if all the explicit
array elements happened to be zero. In addition, for large arrays where only an
initial portion has an explicit initializer, we would emit the complete
initializer (which could be huge) rather than emitting only the non-zero
portion. With this change, when the element would have a suffix of more than 8
zero elements, we emit the array constant as a packed struct of its initial
portion followed by a zeroinitializer constant for the trailing zero portion.
In passing, I found a bug where SemaInit would sometimes walk the entire array
when checking an initializer that only covers the first few elements; that's
fixed here to unblock testing of the rest.
Differential Revision: https://reviews.llvm.org/D47166
llvm-svn: 333044
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
The clang builtins have the same semantics as the stdlib functions.
The stdlib functions are defined in section 7.20.6.1 of the C standard with:
"If the result cannot be represented, the behavior is undefined."
That lets us mark the negation with 'nsw' because "sub i32 0, INT_MIN" would
be UB/poison.
Differential Revision: https://reviews.llvm.org/D47202
llvm-svn: 333038
|
|
|
|
|
|
| |
CGBuiltin.cpp. Just wrap a select builtin around them in the header file instead.
llvm-svn: 333027
|
|
|
|
|
|
|
|
|
| |
clang."
As well as two follow-on commits r332906, r332911 with a fix for
test clang/test/CodeGen/split-debug-filename.c.
llvm-svn: 333013
|
|
|
|
|
|
| |
Due to regression on arm.
llvm-svn: 332991
|
|
|
|
|
|
|
|
| |
We chose the 'slt' form as canonical in IR with:
rL332819
...so we should generate that form directly for efficiency.
llvm-svn: 332989
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Introduced CreateMemTempWithoutCast and CreateTemporaryAllocaWithoutCast to emit alloca
without casting to default addr space.
ActiveFlag is a temporary variable emitted for clean up. It is defined as AllocaInst* type and there is
a cast to AlllocaInst in SetActiveFlag. An alloca casted to generic pointer causes assertion in
SetActiveFlag.
Since there is only load/store of ActiveFlag, it is safe to use the original alloca, therefore use
CreateMemTempWithoutCast is called.
Differential Revision: https://reviews.llvm.org/D47099
llvm-svn: 332982
|
|
|
|
|
|
|
|
|
| |
This change will help Visual Studio resolve forward references to C++ lambda
routines used by captured variables.
Differential Revision: https://reviews.llvm.org/D45438
llvm-svn: 332975
|
|
|
|
|
|
| |
This reverts commit r332885 as it broke several greendragon buildbots.
llvm-svn: 332973
|
|
|
|
|
|
| |
This reverts commit r332906 as a dependency to revert r332885.
llvm-svn: 332972
|
|
|
|
|
|
| |
Suggested by Gabor Buella.
llvm-svn: 332966
|
|
|
|
|
|
|
|
| |
equivalent.
Patch by Hans Wennborg!
llvm-svn: 332964
|
|
|
|
| |
llvm-svn: 332963
|
|
|
|
|
|
|
| |
Reverted due to buildbot failures.
Seems like isnumber() is some Apple addition to cctype.
llvm-svn: 332957
|
|
|
|
|
|
|
| |
This appears to leak in already on libc++ platforms, but is breaking on
some other targets.
llvm-svn: 332955
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Summary:
This includes initial support for the (hopefully final) updated Objective-C ABI, developed here:
https://github.com/davidchisnall/clang-gnustep-abi-2
It also includes some cleanups and refactoring from older GNU ABIs.
The current version is ELF only, other formats to follow.
Reviewers: rjmccall, DHowett-MSFT
Reviewed By: rjmccall
Subscribers: smeenai, cfe-commits
Differential Revision: https://reviews.llvm.org/D46052
llvm-svn: 332950
|
|
|
|
| |
llvm-svn: 332906
|
|
|
|
|
|
|
|
|
|
|
|
| |
instruction instead.
Because the intrinsics in the headers are implemented as macros, we can't just use a select builtin and pternlog builtin. This would require one of the macro arguments to be used twice. Depending on what was passed to the macro we could expand an expression twice leading to weird behavior. We could maybe declare our local variable in the macro, but that would need to worry about name collisions.
To avoid that just generate IR directly in CGBuiltin.cpp.
Differential Revision: https://reviews.llvm.org/D47125
llvm-svn: 332891
|
|
|
|
|
|
| |
initialization.
llvm-svn: 332886
|
|
|
|
|
|
|
|
| |
Fixes PR37466.
Differential Revision: https://reviews.llvm.org/D47093
llvm-svn: 332885
|
|
|
|
|
|
|
|
|
|
| |
up to dwo output.
Part of PR37466.
Differential Revision: https://reviews.llvm.org/D47089
llvm-svn: 332881
|
|
|
|
| |
llvm-svn: 332879
|
|
|
|
|
|
|
|
|
|
|
|
| |
1. added restrictions to memory scope, order and volatile parameters
2. added custom processing for these builtins - currently is not used code,
needed to switch off GCCBuiltin link to the builtins (ongoing change to llvm
tree)
3. builtins renamed as requested
Differential Revision: https://reviews.llvm.org/D43281
llvm-svn: 332848
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
If a variable has an initializer, codegen tries to build its value. If
the variable is large in size, building its value requires substantial
resources. It causes strange behavior from user viewpoint: compilation
of huge zero initialized arrays like:
char data_1[2147483648u] = { 0 };
consumes enormous amount of time and memory.
With this change codegen tries to determine if variable initializer is
equivalent to zero initializer. In this case variable value is not
constructed.
This change fixes PR18978.
Differential Revision: https://reviews.llvm.org/D46241
llvm-svn: 332847
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
The first version of the patch (r332228) was flawed because it was
putting structors into C5/D5 comdats very eagerly. This is correct only
if we can ensure the comdat contains all required versions of the
structor (which wasn't the case). This version uses a more nuanced
approach:
- for local structor symbols we use an alias because we don't have to
worry about comdats or other compilation units.
- linkonce symbols are emitted separately, as we cannot guarantee we
will have all symbols we need to form a comdat (they are emitted
lazily, only when referenced).
- available_externally symbols are also emitted separately, as the code
seemed to be worried about emitting an alias in this case.
- other linkage types are not affected by the optimization level. They
either get put into a comdat (weak) or get aliased (external).
Reviewers: rjmccall, aprantl
Subscribers: cfe-commits
Differential Revision: https://reviews.llvm.org/D46685
llvm-svn: 332839
|
|
|
|
|
|
|
|
|
|
| |
When a lambda capture captures a __block in the same statement, the compiler asserts out because isCapturedBy assumes that an Expr can only be a BlockExpr, StmtExpr, or if it's a Stmt then all the statement's children are expressions. That's wrong, we need to visit all sub-statements even if they're not expressions to see if they also capture.
Fix this issue by pulling out the isCapturedBy logic to use RecursiveASTVisitor.
<rdar://problem/39926584>
llvm-svn: 332801
|
|
|
|
|
|
|
|
|
|
|
|
| |
To support linking device code in different source files, it is necessary to
embed fat binary at host linking stage.
This patch emits an external symbol for fat binary in host codegen, then
embed the fat binary by lld through a linker script.
Differential Revision: https://reviews.llvm.org/D46472
llvm-svn: 332724
|
|
|
|
|
|
|
|
|
|
| |
This will eventually replace MCObjectWriter::WriteZeros.
Part of PR37466.
Differential Revision: https://reviews.llvm.org/D47033
llvm-svn: 332675
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
MethodVFTableLocations in MigrosoftVTableContext contains canonicalized
decl. But, it's sometimes asked to lookup for non-canonicalized decl,
and that causes assertion failure, and compilation failure.
Fixes PR37481.
Patch by Taiju Tsuiki!
Differential Revision: https://reviews.llvm.org/D46929
llvm-svn: 332639
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
lifetime.start/end expects pointer argument in alloca address space.
However in C++ a temporary variable is in default address space.
This patch changes API CreateMemTemp and CreateTempAlloca to
get the original alloca instruction and pass it lifetime.start/end.
It only affects targets with non-zero alloca address space.
Differential Revision: https://reviews.llvm.org/D45900
llvm-svn: 332593
|
|
|
|
|
|
|
|
|
|
|
| |
functions.
If the combined construct is specified in the declare target function
and the device code is emitted, the compiler crashes because of the
incorrectly chosen captured stmt. We should choose the innermost
captured statement, not the outermost.
llvm-svn: 332477
|
|
|
|
|
|
|
|
| |
If the orphaned directive is executed in SPMD mode, we need to emit the
check for the SPMD mode and run the orphaned parallel directive in
sequential mode.
llvm-svn: 332467
|
|
|
|
| |
llvm-svn: 332400
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
- Define a function (canPassInRegisters) that determines whether a
record can be passed in registers based on language rules and
target-specific ABI rules.
- Set flag RecordDecl::ParamDestroyedInCallee to true in MSVC mode and
remove ASTContext::isParamDestroyedInCallee, which is no longer needed.
- Use the same type (unsigned) for RecordDecl's bit-field members.
For more background, see the following discussions that took place on
cfe-commits.
http://lists.llvm.org/pipermail/cfe-commits/Week-of-Mon-20180326/223498.html
http://lists.llvm.org/pipermail/cfe-commits/Week-of-Mon-20180402/223688.html
http://lists.llvm.org/pipermail/cfe-commits/Week-of-Mon-20180409/224754.html
http://lists.llvm.org/pipermail/cfe-commits/Week-of-Mon-20180423/226494.html
http://lists.llvm.org/pipermail/cfe-commits/Week-of-Mon-20180507/227647.html
llvm-svn: 332397
|
|
|
|
|
|
|
|
| |
In generic data-sharing mode we do not need to globalize
variables/parameters of reference/pointer types. They already are placed
in the global memory.
llvm-svn: 332380
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
The DEBUG() macro is very generic so it might clash with other projects.
The renaming was done as follows:
- git grep -l 'DEBUG' | xargs sed -i 's/\bDEBUG\s\?(/LLVM_DEBUG(/g'
- git diff -U0 master | ../clang/tools/clang-format/clang-format-diff.py -i -p1 -style LLVM
Explicitly avoided changing the strings in the clang-format tests.
Differential Revision: https://reviews.llvm.org/D44975
llvm-svn: 332350
|
|
|
|
|
|
|
|
| |
Initial support for passing the virtual base pointer offset to CodeViewDebug.
https://reviews.llvm.org/D46271
llvm-svn: 332296
|
|
|
|
|
|
|
|
|
| |
Some targets have constant address space (e.g. amdgcn). For them string literal should be
emitted in constant address space then casted to default address space.
Differential Revision: https://reviews.llvm.org/D46643
llvm-svn: 332279
|
|
|
|
|
|
|
|
|
| |
It breaks the sanitizer build
<http://lab.llvm.org:8011/builders/sanitizer-x86_64-linux-autoconf/builds/23739>
This reverts commit r332228.
llvm-svn: 332232
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Summary:
Removing the full structor and replacing all usages with the base one
can degrade debug quality as it will leave the debugger unable to locate
the full object structor. This is apparent when evaluating an expression
in the debugger which requires constructing an object of class which has
had this optimization applied to it. When compiling the expression, we
pretend that the class and its methods have been defined in another
compilation unit, so the expression compiler assumes the structor
definition must be available. This didn't use to be the case for
structors with internal linkage. Less aggressive optimizations like
emitting the full structor as an alias remain in place, as they do not
cause the structor symbol to disappear completely.
This improves debug quality on non-darwin platforms (darwin does not
have -mconstructor-aliases on by default, so it is spared these
problems) and enable us to remove some workarounds from LLDB which attempt to
mitigate this issue.
Reviewers: rjmccall, aprantl
Subscribers: cfe-commits
Differential Revision: https://reviews.llvm.org/D46685
llvm-svn: 332228
|
|
|
|
|
|
|
|
|
| |
These intrinsics work exactly as all other atomic_fetch_* intrinsics and allow to create *atomicrmw* with ordering.
Updated the clang-extensions document.
Differential Revision: https://reviews.llvm.org/D46386
llvm-svn: 332193
|
|
|
|
|
|
|
|
|
| |
distribute simd directives.
Directives `target simd` and `target teams distribute simd` must be
executed in non-SPMD mode.
llvm-svn: 332129
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Summary:
The Itanium ABI requires that the type info for pointer-to-incomplete types to have internal linkage, so that it doesn't interfere with the type info once completed. Currently it also marks the type info name as internal as well. However, this causes a bug with the STL implementations, which use the type info name pointer to perform ordering and hashing of type infos.
For example:
```
// header.h
struct T;
extern std::type_info const& Info;
// tu_one.cpp
#include "header.h"
std::type_info const& Info = typeid(T*);
// tu_two.cpp
#include "header.h"
struct T {};
int main() {
auto &TI1 = Info;
auto &TI2 = typeid(T*);
assert(TI1 == TI2); // Fails
assert(TI1.hash_code() == TI2.hash_code()); // Fails
}
```
This patch fixes the STL bug by emitting the type info name as linkonce_odr when the type-info is for a pointer-to-incomplete type.
Note that libc++ could fix this without a compiler change, but the quality of fix would be poor. The library would either have to:
(A) Always perform strcmp/string hashes.
(B) Determine if we have a pointer-to-incomplete type, and only do strcmp then. This would require an ABI break for libc++.
Reviewers: rsmith, rjmccall, majnemer, vsapsai
Reviewed By: rjmccall
Subscribers: smeenai, cfe-commits
Differential Revision: https://reviews.llvm.org/D46665
llvm-svn: 332028
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
This commit relands r331904.
Adding a SrcMgr::CharacteristicKind parameter to the InclusionDirective
in PPCallbacks, and updating calls to that function. This will be useful
in https://reviews.llvm.org/D43778 to determine which includes are
system
headers.
Differential Revision: https://reviews.llvm.org/D46614
llvm-svn: 332021
|
|
|
|
|
|
|
|
| |
Added initial support for L2 parallelism in SPMD mode. Note, though,
that the orphaned parallel directives are not currently supported in
SPMD mode.
llvm-svn: 332016
|
|
|
|
|
|
|
|
| |
when current field is not legal integer type.
Differential Revision: https://reviews.llvm.org/D39053
llvm-svn: 331979
|
|
|
|
|
|
|
| |
This reverts commit r331957. It seems to be causing failures
on ppc64le-linux.
llvm-svn: 331963
|
|
|
|
|
|
|
|
|
|
| |
not use a 512-bit intermediate vector.
This is unnecessary for AVX512VL supporting CPUs like SKX. We can just emit a 128-bit masked load/store here no matter what. The backend will widen it to 512-bits on KNL CPUs.
Fixes the frontend portion of PR37386. Need to fix the backend to optimize the new sequences well.
llvm-svn: 331958
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Summary:
The Itanium ABI requires that the type info for pointer-to-incomplete types to have internal linkage, so that it doesn't interfere with the type info once completed. Currently it also marks the type info name as internal as well. However, this causes a bug with the STL implementations, which use the type info name pointer to perform ordering and hashing of type infos.
For example:
```
// header.h
struct T;
extern std::type_info const& Info;
// tu_one.cpp
#include "header.h"
std::type_info const& Info = typeid(T*);
// tu_two.cpp
#include "header.h"
struct T {};
int main() {
auto &TI1 = Info;
auto &TI2 = typeid(T*);
assert(TI1 == TI2); // Fails
assert(TI1.hash_code() == TI2.hash_code()); // Fails
}
```
This patch fixes the STL bug by emitting the type info name as linkonce_odr when the type-info is for a pointer-to-incomplete type.
Note that libc++ could fix this without a compiler change, but the quality of fix would be poor. The library would either have to:
(A) Always perform strcmp/string hashes.
(B) Determine if we have a pointer-to-incomplete type, and only do strcmp then. This would require an ABI break for libc++.
Reviewers: rsmith, rjmccall, majnemer, vsapsai
Reviewed By: rjmccall
Subscribers: smeenai, cfe-commits
Differential Revision: https://reviews.llvm.org/D46665
llvm-svn: 331957
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
match what the middle and backends understand
Previously we emitted something like
rotl(x, n) {
n &= bitwidth-1;
return n != 0 ? ((x << n) | (x >> (bitwidth - n)) : x;
}
We use a select to avoid the undefined behavior on the (bitwidth - n) shift.
The middle and backend don't really recognize this as a rotate and end up emitting a cmov or control flow because of the select.
A better pattern is (x << (n & mask)) | (x << (-n & mask)) where mask is bitwidth - 1.
Fixes the main complaint in PR37387. There's still some work to be done if the user writes that sequence directly on a short or char where type promotion rules can prevent it from being recognized. The builtin is emitting direct IR with unpromoted types so that isn't a problem for it.
Differential Revision: https://reviews.llvm.org/D46656
llvm-svn: 331943
|