When computing the smallest and largest types for selecting the maximum
vectorization factor, we currently ignore loads and stores of pointer types if
the memory access is non-consecutive. We do this because such accesses must be
scalarized regardless of vectorization factor, and thus shouldn't be considered
when determining the factor. This patch makes this check less aggressive by
also considering non-consecutive accesses that may be vectorized, such as
interleaved accesses. Because we don't know at the time of the check if an
accesses will certainly be vectorized (this is a cost model decision given a
particular VF), we consider all accesses that can potentially be vectorized.
Differential Revision: https://reviews.llvm.org/D30305
llvm-svn: 296747
These loads cannot be savely hoisted as the condition guarding the
non-affine region cannot be duplicated to also protect the hoisted load
later on. Today they are dropped in ScopInfo. By checking for this early, we
do not even try to model them and possibly can still optimize smaller regions
not containing this specific required-invariant load.
llvm-svn: 296744
If dominator tree is not calculated or is invalidated, set corresponding
pointer in the pass state to nullptr. Such pointer value will indicate
that operations with dominator tree are not allowed. In particular, it
allows to skip verification for such pass state. The dominator tree is
not calculated if the machine dominator pass was skipped, it occures in
the case of entities with linkage available_externally.
The change fixes some test fails observed when expensive checks
are enabled.
Differential Revision: https://reviews.llvm.org/D29280
llvm-svn: 296742
this test was using the VPATH hack to avoid having a copy of the
inferior source code. This makes the test fail if in happens to run
concurrently with a test in the parent folder. Fix that by moving it up
to the parent.
llvm-svn: 296741
MSVC (at least the version I am using) does not want to implicitly
capture a const bool variable. Move it into the lambda, as it is not
used outside anyway.
llvm-svn: 296738
Summary:
This patch makes the namespace comment fixer use the number of unwrapped lines
that a namespace spans to detect it that namespace is short, thus not needing
end comments to be added.
This is needed to ensure clang-format is idempotent. Previously, a short namespace
was detected by the original source code lines. This has the effect of requiring two
runs for this example:
```
namespace { class A; }
```
after first run:
```
namespace {
class A;
}
```
after second run:
```
namespace {
class A;
} // namespace
```
Reviewers: djasper
Reviewed By: djasper
Subscribers: cfe-commits, klimek
Differential Revision: https://reviews.llvm.org/D30528
llvm-svn: 296736
Summary:
Hello everybody,
this is an incremental patch for the NoMalloc-Checker I wrote. It allows to configure the memory-management functions, that are checked,
This might be helpful for a code base with custom functions in use, or non-standard functionality, like posix_memalign.
Reviewers: aaron.ballman, hokein, alexfh
Reviewed By: aaron.ballman, alexfh
Subscribers: sbenza, nemanjai, JDevlieghere
Tags: #clang-tools-extra
Patch by Jonas Toth!
Differential Revision: https://reviews.llvm.org/D28239
llvm-svn: 296734
Summary:
Most classes annotated with _LIBCPP_TYPE_VIS need to have at least some
of their members exported, otherwise we have a lot of link errors when
linking against a libc++ built with hidden visibility. This also makes
_LIBCPP_TYPE_VIS be consistent across platforms, since on Windows it
already exports members.
With this change made, any template methods of a class marked
_LIBCPP_TYPE_VIS will also get default visibility when instantiatied,
which is not desirable for clients of libc++ headers who wish to control
their visibility; this is the same issue as PR30642. Annotate all
problematic methods with an explicit visibility specifier to avoid this.
The problematic methods were found by running bad-visibility-finder [1]
against the libc++ headers after making the _LIBCPP_TYPE_VIS change. The
small methods were marked for inlining; the larger ones hidden.
[1] https://github.com/smeenai/bad-visibility-finder
Reviewers: mclow.lists, EricWF
Subscribers: cfe-commits
Differential Revision: https://reviews.llvm.org/D25208
llvm-svn: 296732
When building libc++ with hidden visibility, we want explicit template
instantiations to export members. This is consistent with existing
Windows behavior, and is necessary for clients to be able to link
against a hidden visibility built libc++ without running into lots of
missing symbols.
An unfortunate side effect, however, is that any template methods of a
class with an explicit instantiation will get default visibility when
instantiated, unless the methods are explicitly marked inline or hidden
visibility. This is not desirable for clients of libc++ headers who wish
to control their visibility, and led to PR30642.
Annotate all problematic methods with an explicit visibility specifier
to avoid this. The problematic methods were found by running
https://github.com/smeenai/bad-visibility-finder against the libc++
headers after making the _LIBCPP_EXTERN_TEMPLATE_TYPE_VIS change. The
methods were marked with the new _LIBCPP_METHOD_TEMPLATE_IMPLICIT_INSTANTIATION_VIS
macro, which was created for this purpose.
It should be noted that _LIBCPP_EXTERN_TEMPLATE_TYPE_VIS was originally
intended to expand to default visibility, and was changed to expanding
to default type visibility to fix PR30642. The visibility macro
documentation was not updated accordingly, however, so this change makes
the macro consistent with its documentation again, while explicitly
fixing the methods which resulted in that PR.
Differential Revision: https://reviews.llvm.org/D29157
llvm-svn: 296731
D29157 will make explicit template instantiations expand to default
visibility, at which point these method templates will need to be
explicitly marked hidden visibility to avoid leaking into other DSOs.
Unfortunately, because of clang PR32114, they must be marked inline (in
conjunction with `-fvisibility-inlines-hidden`) to actually hide them,
since clang doesn't respect the hidden visibility annotation.
Since this involves an ABI change, mark these methods inline in a
separate change, so that the ABI changes can be reviewed separately and
verified to be safe.
Differential Revision: https://reviews.llvm.org/D30523
llvm-svn: 296729
The test is passing with c++11 and c++14 but not c++1z on this
particular version of the compiler. Try to use lit boolean condition
to satisfy this constaint.
llvm-svn: 296725
This reverts commit r296712. It broke our bot.
It turns out that the test is passing with c++11 and c++14 but
not c++1z on this particular version of the compiler. Since one
job is defaulting to c++1z and the other is testing all config I'm
not sure how to fix this...
llvm-svn: 296724
We were not gcing any section whose name was a C identifier. Both gold
and bfd only keep those if they are used.
To avoid having to create the __start/__stop symbols early or doing
string lookups in resolvedReloc, this patch just looks for undefined
symbols __start/__stop to decide if a section is needed or not.
llvm-svn: 296723
Surprisingly, one of the three interference checks in LiveRegMatrix was
using the main live range instead of the apropriate subregister range
resulting in unnecessarily conservative results.
llvm-svn: 296722
Original commit message:
[ARM] Fix insert point for store rescheduling.
In ARMPreAllocLoadStoreOpt::RescheduleOps, LastOp should be the last
operation which we want to merge. If we break out of the loop because
an operation has the wrong offset, we shouldn't use that operation as
LastOp.
This patch fixes some cases where we would sink stores for no reason.
llvm-svn: 296718
This patch cleans up how libc++abi handles the definitions for new/delete.
It is in preperation for upcoming changes to fix how both libc++ and libc++abi
handle new/delete.
The primary changes in this patch are:
* Move the definitions for bad_array_length and bad_new_array_length
into stdlib_exception.cpp. This way stdlib_new_delete.cpp only
contains new/delete.
* Rename cxa_new_delete.cpp -> stdlib_new_delete.cpp for consistency
with other files.
* Add a FIXME regarding when stdlib_new_delete.cpp is actually compiled
as part of the dylib.
llvm-svn: 296715
This tests is failing in XCode 7.0. But Xcode 7.3 that shipped
an updated clang has this test passing. This is fixing green dragon
which runs this configuration.
llvm-svn: 296712
Summary:
This can be used to optimize large multiplications after legalization.
Depends on D29565
Reviewers: mkuper, spatel, RKSimon, zvi, bkramer, aaboud, craig.topper
Subscribers: llvm-commits
Differential Revision: https://reviews.llvm.org/D29587
llvm-svn: 296711
Until now, we've had to use -global-isel to enable GISel. But using
that on other targets that don't support it will result in an abort, as we
can't build a full pipeline.
Additionally, we want to experiment with enabling GISel by default for
some targets: we can't just enable GISel by default, even among those
target that do have some support, because the level of support varies.
This first step adds an override for the target to explicitly define its
level of support. For AArch64, do that using
a new command-line option (I know..):
-aarch64-enable-global-isel-at-O=<N>
Where N is the opt-level below which GISel should be used.
Default that to -1, so that we still don't enable GISel anywhere.
We're not there yet!
While there, remove a couple LLVM_UNLIKELYs. Building the pipeline is
such a cold path that in practice that shouldn't matter at all.
llvm-svn: 296710
In ARMPreAllocLoadStoreOpt::RescheduleOps, LastOp should be the last
operation which we want to merge. If we break out of the loop because
an operation has the wrong offset, we shouldn't use that operation as
LastOp.
This patch fixes some cases where we would sink stores for no reason.
Differential Revision: https://reviews.llvm.org/D30124
llvm-svn: 296708
Summary:
This patch allows us to move away from using __thread on darwin,
which is requiring for building lsan for darwin on ios version 7
and on iossim i386.
Reviewers: kubamracek, kcc
Subscribers: llvm-commits
Differential Revision: https://reviews.llvm.org/D29994
llvm-svn: 296707
Summary:
The current size is flaky, as revealed by checking
the stack size attr after setting it.
Reviewers: kubamracek, rnk
Subscribers: llvm-commits
Differential Revision: https://reviews.llvm.org/D30267
llvm-svn: 296706
These tests are failing in XCode 8.0, 8.1, and 8.2, but not in Xcode
8.3. Annoyingly the version numbering for clang does not follow Xcode
and is bumped to 8.1 only in Xcode 8.3. So Xfailing apple-clang-8.0
should catch all cases here.
llvm-svn: 296704
This patch adds an option named --thinlto-cache-dir, which specifies the
path to a directory in which to cache native object files for ThinLTO
incremental builds.
Differential Revision: https://reviews.llvm.org/D30509
llvm-svn: 296702
This code starts from the high end of the sorted vector of offsets, and
works backwards: it tries to find contiguous offsets, process them, then
pops them from the end of the vector. Most of the code agrees with this
order of processing, but one loop doesn't: it instead processes elements
from the low end of the vector (which are nodes with unrelated offsets).
Fix that loop to process the correct elements.
This has a few implications. One, we don't incorrectly return early when
processing multiple groups of offsets in the same block (which allows
rescheduling prera-ldst-insertpt.mir). Two, we pick the correct insert
point for loads, so they're correctly sorted (which affects the
scheduling of vldm-liveness.ll). I think it might also impact some of
the heuristics slightly.
Differential Revision: https://reviews.llvm.org/D30368
llvm-svn: 296701
That class had three member functions, and all of them are just reader
methods that did not depend on class members, so they can be just non-
member functions.
Probably we should reorganize the functions themselves because their
return types doesn't make much sense to me, but for now I just moved
these functions out of the class.
llvm-svn: 296700
This is part of the ongoing attempt to improve select codegen for all targets and select
canonicalization in IR (see D24480 for more background). The transform is a subset of what
is done in InstCombine's FoldOpIntoSelect().
I first noticed a regression in the x86 avx512-insert-extract.ll tests with a patch that
hopes to convert more selects to basic math ops. This appears to be a general missing DAG
transform though, so I added tests for all standard binops in rL296621
(PowerPC was chosen semi-randomly; it has scripted FileCheck support, but so do ARM and x86).
The poor output for "sel_constants_shl_constant" is tracked with:
https://bugs.llvm.org/show_bug.cgi?id=32105
Differential Revision: https://reviews.llvm.org/D30502
llvm-svn: 296699
Now that terminators can be EH pads, this code needs to iterate over the
immediate dominators of the EH pad to find a valid insertion point.
Fix for PR32107
Patch by Robert Olliff!
Differential Revision: https://reviews.llvm.org/D30511
llvm-svn: 296698
Matthew Simpson