Previously we do symbolization for all the functions and actually we only need the symbols that's hit by the samples.
This can significantly speed up the time for large size binary.
Optimization for per-inliner will come along with next patch.
Reviewed By: hoy, wenlei
Differential Revision: https://reviews.llvm.org/D110465
This patch is in a series of patches to provide builtins for
compatability with the XL compiler. This patch adds builtins for compare
exponent and test data class operations on floating point values.
Reviewed By: #powerpc, lei
Differential Revision: https://reviews.llvm.org/D109437
Note that SystemZMnemonicSpellCheck is defined in
SystemZGenAsmMatcher.inc, which SystemZAsmParser.cpp includes.
Identified with readability-redundant-declaration.
Rename `lenParams` to `typeparams` to be in sync with fir-dev.
This patch is part of the upstreaming effort from fir-dev branch.
Reviewed By: kiranchandramohan
Differential Revision: https://reviews.llvm.org/D110628
Co-authored-by: Jean Perier <jperier@nvidia.com>
Co-authored-by: Eric Schweitz <eschweitz@nvidia.com>
Commit 354ded67b3 ("tsan: align ThreadState to cache line")
did an incomplete thing. It marked ThreadState as cache line
aligned, but the thread local ThreadState instance is declared
as an aligned char array with hard-coded 64-byte alignment.
On PowerPC cache line size is 128 bytes, so the hard-coded
64-byte alignment is not enough.
Use cache line alignment consistently.
Differential Revision: https://reviews.llvm.org/D110629
Move the "slice" and "composite" handling into separate methods to avoid
if/else hell. Use more LLVM types whenever possible. Replace printf()s
with llvm::Error combined with LLDB logging.
Differential Revision: https://reviews.llvm.org/D110619
As we have to split blocks, we may be left in an invalid loop state
after a WLS is reverted to a DLS. Instead remember the WLS that could
not be fixed and revert them after finishing processing all other loops.
Differential Revision: https://reviews.llvm.org/D110567
No idea how my local machine missed this, but I saw no warning for it,
it seems to have been lost in some level of translating this back for
upstreaming.
This reverts the revert commit f85d8a5bed
with bug fixes.
Original message:
MOVi32imm + ANDWrr ==> ANDWri + ANDWri
MOVi64imm + ANDXrr ==> ANDXri + ANDXri
The mov pseudo instruction could be expanded to multiple mov instructions later.
In this case, try to split the constant operand of mov instruction into two
bitmask immediates. It makes only two AND instructions intead of multiple
mov + and instructions.
Added a peephole optimization pass on MIR level to implement it.
Differential Revision: https://reviews.llvm.org/D109963
Try to improve vectorization of the PHI nodes by trying to vectorize
similar instructions at the size of the widest possible vectors, then
aggregating with compatible type PHIs and trying to vectoriza again and
only if this failed, try smaller sizes of the vector factors for
compatible PHI nodes. This restores performance of several benchmarks
after tuning of the fp/int conversion instructions costs.
Differential Revision: https://reviews.llvm.org/D108740
In rG6a076fa9539e, a problem with updating the old/narrow phi nodes after IV
widening was introduced. If after widening of the IV the transformation is
*not* applied, the narrow phi node was incorrectly modified, which should only
happen if flattening happens. This can be seen in the added test widen-iv2.ll,
which incorrectly had 1 incoming value, but should have its original 2 incoming
values, which is now restored.
Differential Revision: https://reviews.llvm.org/D110234
Require it to be always_inline, to more closely match how _FORITFY_SOURCE
behaves.
This avoids generation of `.inline` suffixed functions - these should always be
inlined.
We need more coverage for commuted and (un)signed preds to
verify that things behave as expected here. Currently, we
do not transform signed preds or non-inbounds geps.
The ARMLowOverheadLoops pass recalculates VPT block masks when it
converts VCMP's inside VPT blocks into VPT's. The function to do so
doesn't seem to handle debug info though, leading to invalid block
creation or asserts at compile time. Make sure the function skips any
debug info between the MVE instructions it inspects.
Differential Revision: https://reviews.llvm.org/D110564
After significant problems in our downstream with the previous
implementation, the SYCL standard has opted to make using macros/etc to
change kernel-naming-lambdas in any way UB (even passively). As a
result, we are able to just emit the itanium mangling.
However, this DOES require a little work in the CXXABI, as the microsoft
and itanium mangler use different numbering schemes for lambdas. This
patch adds a pair of mangling contexts that use the normal 'itanium'
mangling strategy to fill in the "DeviceManglingNumber" used previously
by CUDA.
Differential Revision: https://reviews.llvm.org/D110281
When TargetLibraryInfoImpl::isValidProtoForLibFunc is checking
function signatures to detect lib calls it may check that a parameter
or return value matches with the "size_t" type. For this to work it
has to derive the IR type matching with "size_t". Depending on if
a DataLayout is provided or not, this has been done in two different
way. Either a more strict check being based on IntPtrType (which is
given by the DataLayout) or a more relaxed check assuming that any
integer type matches with "size_t".
Given that the stricter approach exist it seems like we do not want
to trigger rewrites etc if we aren't sure that a function calls
actually match with the library function. Therefore it was questioned
why we actually have the more relaxed approach when not being able
to derive an IR type for "size_t". This patch will take a more
defensive approach, requiring that a DataLayout is passed to
isValidProtoForLibFunc.
Differential Revision: https://reviews.llvm.org/D110584
This patch is for fixing potential insertElement-related bugs like D93818.
```
V = UndefValue::get(VecTy);
for(...)
V = Builder.CreateInsertElementy(V, Elt, Idx);
=>
V = PoisonValue::get(VecTy);
for(...)
V = Builder.CreateInsertElementy(V, Elt, Idx);
```
Like above, this patch changes the placeholder V to poison.
The patch will be separated into several commits.
Reviewed By: aqjune
Differential Revision: https://reviews.llvm.org/D110311
The instruction extractelement/extractvalue are not required to
be scheduled since they only depend on the source vector/aggregate (with
constant indices), smae applies to the parent basic block checks.
Improves compile time and saves scheduling budget.
Differential Revision: https://reviews.llvm.org/D108703
C++17 permits using 'typename' or 'class' for a template template
parameter, but the error message in the parser only refers to 'class'.
This patch, in C++17 or newer modes, adds "or 'template'" to the
diagnostic.
The two module retrieval methods (qXfer:libraries-svr4 and manual list
traversal) differ in how the handle the
manually-added-but-not-yet-loaded modules. The svr4 path will remove it,
while the manual one will keep in the list.
It's likely the two paths need ought to be synchronized, but right now,
this distinction is not relevant for the test.
Recommit https://reviews.llvm.org/D109813 and
https://reviews.llvm.org/D109814.
This implements the second and final entry point for GET_COMMAND_ARGUMENT,
handling the VALUE, STATUS and ERRMSG parameters.
It has a small fix in that we're now using memcpy instead of strncpy
(which was a bad idea to begin with, since we're not actually interested
in a string copy).
mold says it is compatible with GNU ld and gold linkers:
```
$ mold -v
mold 0.9.5 (compatible with GNU ld and GNU gold)
```
And thus it currently gets detected as Gold.
With the following diff, CMake now correctly reports the linker name, and mold keeps being identified as Gold internally for now.
Reviewed By: ldionne, MaskRay
Differential Revision: https://reviews.llvm.org/D110035
We only had that ability for regular debugger launches. This meant that
it was not possible to use the normal dlopen patterns in attach tests.
This fixes that.
We added some support for this mode back in 2015, but the feature was
never productionized. It is completely untested, and there are known
major structural lldb issues that need to be resolved before this
feature can really be supported.
It also complicates making further changes to stop reply packet
handling, which is what I am about to do.
Differential Revision: https://reviews.llvm.org/D110553
This reverts commit 0446f1299f and
df6302311f.
There's a warning on flang-aarch64-latest-gcc related to strncpy using
the result of strlen as a bound. I'll recommit with a fix.
Add support for NOLINTBEGIN ... NOLINTEND comments to suppress
clang-tidy warnings over multiple lines. All lines between the "begin"
and "end" markers are suppressed.
Example:
// NOLINTBEGIN(some-check)
<Code with warnings to be suppressed, line 1>
<Code with warnings to be suppressed, line 2>
<Code with warnings to be suppressed, line 3>
// NOLINTEND(some-check)
Follows similar syntax as the NOLINT and NOLINTNEXTLINE comments
that are already implemented, i.e. allows multiple checks to be provided
in parentheses; suppresses all checks if the parentheses are omitted,
etc.
If the comments are misused, e.g. using a NOLINTBEGIN but not
terminating it with a NOLINTEND, a clang-tidy-nolint diagnostic
message pointing to the misuse is generated.
As part of implementing this feature, the following bugs were fixed in
existing code:
IsNOLINTFound(): IsNOLINTFound("NOLINT", Str) returns true when Str is
"NOLINTNEXTLINE". This is because the textual search finds NOLINT as
the stem of NOLINTNEXTLINE.
LineIsMarkedWithNOLINT(): NOLINTNEXTLINEs on the very first line of a
file are ignored. This is due to rsplit('\n\').second returning a blank
string when there are no more newline chars to split on.
ScalarizationResult's destructor makes sure ToFreeze is not ignored if
set. Currently, scalarizeLoadExtract has an early exit if the index is
not safe directly. But when it is SafeWithFreeze, we need to discard the
state first, otherwise we hit the assert in the destructor.
Fixes PR51992.
It is a common practice in glibc header to provide an inline redefinition of an
existing function. It is especially the case for fortified function.
Clang currently has an imperfect approach to the problem, using a combination of
trivially recursive function detection and noinline attribute.
Simplify the logic by suffixing these functions by `.inline` during codegen, so
that they are not recognized as builtin by llvm.
After that patch, clang passes all tests from https://github.com/serge-sans-paille/fortify-test-suite
Differential Revision: https://reviews.llvm.org/D109967
MOVi32imm + ANDWrr ==> ANDWri + ANDWri
MOVi64imm + ANDXrr ==> ANDXri + ANDXri
The mov pseudo instruction could be expanded to multiple mov instructions later.
In this case, try to split the constant operand of mov instruction into two
bitmask immediates. It makes only two AND instructions intead of multiple
mov + and instructions.
Added a peephole optimization pass on MIR level to implement it.
Differential Revision: https://reviews.llvm.org/D109963
Sometimes I see people unsure about which options they can use in specific versions of clang-format because
https://clang.llvm.org/docs/ClangFormatStyleOptions.html points to the latest and greatest versions.
The reality is this says its version 13.0, but actually anything we add now, will not be in 13.0 GA but
instead 14.0 GA (as 13.0 has already been branched).
How about we introduce some nomenclature to the Format.h so that we can mark which options in the
documentation were introduced for which version?
Reviewed By: HazardyKnusperkeks
Differential Revision: https://reviews.llvm.org/D110432
Call RemoveMachineInstrFromMaps before erasing instrs.
repairIntervalsInRange will do this for you after erasing the
instruction, but it's not safe to rely on it because assertions in
SlotIndexes::removeMachineInstrFromMaps refer to fields in the erased
instruction.
This fixes asan buildbot failures caused by D110335.
Add the fir.save_result operation. It is use to save an
array, box, or record function result SSA-value to a memory location
Reviewed By: jeanPerier
Differential Revision: https://reviews.llvm.org/D110407
Co-authored-by: Jean Perier <jperier@nvidia.com>
Co-authored-by: Valentin Clement <clementval@gmail.com>
This reverts commit 8fdac7cb7a.
The issue causing the revert has been fixed a while ago in 60b852092c.
Original message:
Now that SCEVExpander can preserve LCSSA form,
we do not have to worry about LCSSA form when
trying to look through PHIs. SCEVExpander will take
care of inserting LCSSA PHI nodes as required.
This increases precision of the analysis in some cases.
Reviewed By: mkazantsev, bmahjour
Differential Revision: https://reviews.llvm.org/D71539
This patch emits DW_TAG_namelist and DW_TAG_namelist_item for fortran
namelist variables. DICompositeType is extended to support this fortran
feature.
Reviewed By: aprantl
Differential Revision: https://reviews.llvm.org/D108553
wlei