Fix ARMAttributeParser::CPU_arch_profile so that it doesn't switch on the value
'0' as a legal value of this build attribute.
Change-Id: Ie05a08900a82bb10b78c841b437df747ce3bb38e
llvm-svn: 222743
Summary:
This resolves [[ http://llvm.org/bugs/show_bug.cgi?id=17391 | PR17391 ]].
GCC's sources were used as a guide (couldn't find much information in ARM documentation).
Reviewers: doug.gregor, asl
Reviewed By: asl
Subscribers: asl, aemerson, cfe-commits
Differential Revision: http://reviews.llvm.org/D6339
llvm-svn: 222741
clearly only exactly equal width ptrtoint and inttoptr casts are no-op
casts, it says so right there in the langref. Make the code agree.
Original log from r220277:
Teach the load analysis to allow finding available values which require
inttoptr or ptrtoint cast provided there is datalayout available.
Eventually, the datalayout can just be required but in practice it will
always be there today.
To go with the ability to expose available values requiring a ptrtoint
or inttoptr cast, helpers are added to perform one of these three casts.
These smarts are necessary to finish canonicalizing loads and stores to
the operational type requirements without regressing fundamental
combines.
I've added some test cases. These should actually improve as the load
combining and store combining improves, but they may fundamentally be
highlighting some missing combines for select in addition to exercising
the specific added logic to load analysis.
llvm-svn: 222739
Summary:
First, remove lit configuration that sets ASAN_OPTIONS to detect_leaks=1
because this is already the default when leak detection is supported.
This removes a bit of duplication between various lit.cfg files.
Second, add a new feature 'leak-detection' if we're targetting x86_64
(not i386) on Linux.
Third, change a couple of tests that need leak detection to require the
new 'leak-detection' feature.
Reviewers: kcc, earthdok, samsonov
Reviewed By: samsonov
Subscribers: llvm-commits
Differential Revision: http://reviews.llvm.org/D6396
llvm-svn: 222738
Only the super register flat_scr was marked as reserved,
so in some cases with high register usage it would still
try to allocate the subregisters.
llvm-svn: 222737
Rethrowing exceptions in the MS model is very simple: just call
_CxxThrowException with nullptr for both arguments.
N.B. They chose stdcall as the calling convention for x86 but cdecl for
all other platforms.
llvm-svn: 222733
The pattern matching failed to recognize all instances of "-1", because when
comparing against "-1" we didn't use an APInt of the same bitwidth.
This commit fixes this and also adds inverse versions of the conditon to catch
more cases.
llvm-svn: 222722
This handles cases where we are comparing a masked value against itself.
The analysis could be further improved by making it recursive but such
expense is not currently justified.
llvm-svn: 222716
The attn instruction is not part of the Power ISA, but is documented in the A2
user manual, and is accepted by the GNU assembler for the A2 and the POWER4+.
Reported as part of PR21650.
llvm-svn: 222712
This does not matter on newer cores (where we can use reciprocal estimates in
fast-math mode anyway), but for older cores this allows us to generate better
fast-math code where we have multiple FDIVs with a common divisor.
llvm-svn: 222710
We were matching against the assume intrinsic in every check. Since we know that it must be an assume, this is just wasted work. Somewhat surprisingly, matching an intrinsic id is actually relatively expensive. It devolves to a string construction and comparison in Function::isIntrinsic.
I originally spotted this because it showed up in a performance profile of my compiler. I've since discovered a separate issue which seems to be the actual root cause, but this is minor perf goodness regardless.
I'm likely to follow up with another change to factor out the comparison matching. There's no need to match the compare instruction in every single one of the tests.
Differential Revision: http://reviews.llvm.org/D6312
llvm-svn: 222709
Summary:
This patch adds CMake support for building and testing libc++abi without threads.
1. Add `LIBCXXABI_ENABLE_THREADS` option to CMake.
2. Propagate `LIBCXXABI_ENABLE_THREADS` to lit via lit.site.cfg.in
3. Configure tests for `LIBCXXABI_ENABLE_THREADS=OFF
Currently the test suite does not work when libc++abi is built without threads because that information does not propagate to the test suite.
Reviewers: danalbert, mclow.lists, jroelofs
Reviewed By: jroelofs
Subscribers: cfe-commits
Differential Revision: http://reviews.llvm.org/D6393
llvm-svn: 222702
Summary:
When using LIT the timing output is entirely unused but introduces a dependency on `<chrono>`. When libc++ is built without a montonic clock this causes some of the tests to fail.
This patch factors out all of the timing logic into `support/timer.hpp` and disables it by default. To enable the timing you must define `LIBCXXABI_TIME_TESTS`.
Reviewers: mclow.lists, danalbert, jroelofs
Reviewed By: jroelofs
Subscribers: cfe-commits
Differential Revision: http://reviews.llvm.org/D6391
llvm-svn: 222701
Clarify the wording around !invariant.load to properly reflect the semantics of such loads with respect to control dependence and location lifetime. To the best of my knowledge, the revised wording respects the actual implementation and understanding of issues involved highlighted in the recent 'Optimization hints for "constant" loads' thread on LLVMDev.
In particular, I'm aiming for the following results:
- To clarify that an invariant.load can fault and must respect control dependence. In particular, it is not sound to unconditionally pull an invariant load out of a loop if that loop would potentially never execute.
- To clarify that the invariant nature of a given pointer does not preclude the modification of that location through a pointer which is unrelated to the load operand. In particular, initializing a location and then passing a pointer through an opaque intrinsic which produces a new unrelated pointer, should behave as expected provided that the intrinsic is memory dependent on the initializing store.
- To clarify that storing a value to an invariant location is defined. It can not, for example, be considered unreachable. The value stored can be assumed to be equal to the value of any previous (or following!) invariant load, but the store itself is defined.
I recommend that anyone interested in using !invariant.load, or optimizing for them, read over the discussion in the review thread. A number of motivating examples are discussed.
Differential Revision: http://reviews.llvm.org/D6346
llvm-svn: 222700
Summary:
I finally got around to merging the many, many changes to lit.cfg into
Android's libc++. This patch makes it simpler to actually use a custom
configuration and test format.
First, I've factored out _build, _run, and _clean methods from
_execute_test, since these are the likely parts that will need to be
overridden. This is likely a first step in the work jroelofs has been
doing with improving cross-compiling test execution.
Second, I've added a `configuration_variant` to the config. This
entry, if present, is a string that forms the prefix of the class that
is to be used to configure the test runner. For example, Android sets
`config.configuration_variant = 'Android'`, and this causes an object
of type `AndroidConfiguration` to be constructed.
As an example of how this will be used, see:
https://android-review.googlesource.com/#/c/116022/
Reviewers: jroelofs, mclow.lists, EricWF
Reviewed By: EricWF
Subscribers: cfe-commits
Differential Revision: http://reviews.llvm.org/D6373
llvm-svn: 222698
All supported environments on x86 Windows return structs in EAX:EDX.
This removes code added in r204978 that had to get updated in r222680.
We should now have the same behavior we had before r204978.
llvm-svn: 222697
Summary:
Filter the toplevel matchers by kind.
Decl and Stmt matchers are tied to a specific node kind and trying to
match incompatible nodes is a waste.
Precalculate a filtered list of matchers that have a chance of matching
the node and ignore the rest.
Speeds up our clang-tidy benchmark by ~10%
Reviewers: klimek
Subscribers: klimek, cfe-commits
Differential Revision: http://reviews.llvm.org/D6361
llvm-svn: 222688
Charlie Turner