Jonas Paulsson [Wed, 7 Jan 2015 13:38:29 +0000 (13:38 +0000)]
New method SDep::isNormalMemoryOrBarrier() in ScheduleDAGInstrs.cpp.
Used to iterate over previously added memory dependencies in
adjustChainDeps() and iterateChainSucc().
SDep::isCtrl() was previously used in these places, that also gave
anti and output edges. The code may be worse if these are followed,
because MisNeedChainEdge() will conservatively return true since a
non-memory instruction has no memory operands, and a false chain dep
will be added. It is also unnecessary since all memory accesses of
interest will be reached by memory dependencies, and there is a budget
limit for the number of edges traversed.
This problem was found on an out-of-tree target with enabled alias
analysis. No test case for an in-tree target has been found.
Reviewed by Hal Finkel.
llvm-svn: 225351
Jonas Paulsson [Wed, 7 Jan 2015 13:20:57 +0000 (13:20 +0000)]
Fix typos in comment and option help texts.
For -enable-aa-sched-mi and -use-tbaa-in-sched-mi.
llvm-svn: 225350
Daniel Jasper [Wed, 7 Jan 2015 12:19:53 +0000 (12:19 +0000)]
clang-format: Fix unary operator detection.
Before:
** outparam = 1;
After:
**outparam = 1;
llvm-svn: 225349
Charlie Turner [Wed, 7 Jan 2015 11:37:40 +0000 (11:37 +0000)]
[ARM] Add missing Tag_DIV_use tests.
llvm-svn: 225348
Asiri Rathnayake [Wed, 7 Jan 2015 11:22:58 +0000 (11:22 +0000)]
Fix regression in r225266.
The change in r225266 was reviewed under D6722. But the commit r225266 has a
typo, causing some MCHammer failures. This patch fixes it.
Change-Id: I573efcff25003af7478ac02548ebbe929fc7f5fd
llvm-svn: 225347
Chandler Carruth [Wed, 7 Jan 2015 11:14:51 +0000 (11:14 +0000)]
[PM] Give slightly less horrible names to the utility pass templates for
requiring and invalidating specific analyses. Also make their printed
names match their class names. Writing these out as prose really doesn't
make sense to me any more.
llvm-svn: 225346
Craig Topper [Wed, 7 Jan 2015 08:10:38 +0000 (08:10 +0000)]
[X86] Merge a switch statement inside a default case of another switch statement on the same variable. There was no additional code in the default so this should be no functional change.
llvm-svn: 225345
Craig Topper [Wed, 7 Jan 2015 08:10:36 +0000 (08:10 +0000)]
[X86] Don't mark the shift by 1 instructions as isConvertibleToThreeAddress. There is no handling for them.
llvm-svn: 225344
Craig Topper [Wed, 7 Jan 2015 07:47:52 +0000 (07:47 +0000)]
[X86] Remove some unused TYPE enums from the disassembler.
llvm-svn: 225343
Tobias Grosser [Wed, 7 Jan 2015 07:43:34 +0000 (07:43 +0000)]
Use explicit StringRef Type
We previously used a Twine here, but as pointed out by David Blaikie
and Mehdi Amini storing a temporary StringRef in a Twine is not a good
idea, as the StringRef will be freed before the Twine is used leaving
a Twine that points to uninitialized memory. We now make it explicit that
we use a StringRef here.
llvm-svn: 225342
Karthik Bhat [Wed, 7 Jan 2015 06:34:34 +0000 (06:34 +0000)]
Revert r225165 and r225169
Even thouh gcc produces simialr instructions as Owen pointed out the two patterns aren’t equivalent in the case
where the original subtraction could have caused an overflow.
Reverting the same.
llvm-svn: 225341
Nico Weber [Wed, 7 Jan 2015 05:25:05 +0000 (05:25 +0000)]
Fix comment typo.
llvm-svn: 225340
Saleem Abdulrasool [Wed, 7 Jan 2015 04:20:26 +0000 (04:20 +0000)]
PE/COFF: teach ARMNT backend about ADDR32NB for exports
This adds the ability to export symbols from a DLL built for ARMNT. Add this
support first to help work towards adding support for import thunks on Windows
on ARM. In order to generate the exports, add support for
IMAGE_REL_ARM_ADDR32NB relocations.
llvm-svn: 225339
Ahmed Bougacha [Wed, 7 Jan 2015 02:42:01 +0000 (02:42 +0000)]
[ADT][SmallVector] Flip an assert comparison to avoid overflows yielding false-negatives. NFC.
r221973 changed SmallVector::operator[] to use size_t instead of unsigned.
Before that, on 64bit platforms, when a large index (say -1) was passed,
truncating it to unsigned avoided an overflow when computing 'begin() + idx',
and failed the range checking assertion, as expected.
With r221973, idx isn't truncated, so the addition wraps to
'(char*)begin() - 1', and doesn't fire anymore when it should have done so.
This commit changes the comparison to instead compute 'end() - begin()'
(i.e., 'size()'), which avoids potentially overflowing additions, and
correctly triggers the assertion when values such as -1 are passed.
Note that the problem already existed before that revision, on platforms
where sizeof(size_t) == sizeof(unsigned).
llvm-svn: 225338
Kostya Serebryany [Wed, 7 Jan 2015 02:37:52 +0000 (02:37 +0000)]
[asan] add flag quarantine_size_mb, deprecate quarantine_size
llvm-svn: 225337
Alexey Samsonov [Wed, 7 Jan 2015 02:12:41 +0000 (02:12 +0000)]
[TSan] Fix Go build.
llvm-svn: 225336
Duncan P. N. Exon Smith [Wed, 7 Jan 2015 02:10:42 +0000 (02:10 +0000)]
IR: Remove MDNode::getWhenValsUnresolved()
Remove dead code. Use `MDNode::get()` instead.
llvm-svn: 225335
Duncan P. N. Exon Smith [Wed, 7 Jan 2015 02:09:51 +0000 (02:09 +0000)]
Remove invalid TODO
We can't drop support for RAUW entirely in `MDNode`s, since it's
required for graph construction. This comment was from before I'd done
the math on that (out-of-tree), and never should have been committed.
llvm-svn: 225334
Chandler Carruth [Wed, 7 Jan 2015 01:58:35 +0000 (01:58 +0000)]
[PM] Fix a pretty nasty bug where the new pass manager would invalidate
passes too many time.
I think this is actually the issue that someone raised with me at the
developer's meeting and in an email, but that we never really got to the
bottom of. Having all the testing utilities made it much easier to dig
down and uncover the core issue.
When a pass manager is running many passes over a single function, we
need it to invalidate the analyses between each run so that they can be
re-computed as needed. We also need to track the intersection of
preserved higher-level analyses across all the passes that we run (for
example, if there is one module analysis which all the function analyses
preserve, we want to track that and propagate it). Unfortunately, this
interacted poorly with any enclosing pass adaptor between two IR units.
It would see the intersection of preserved analyses, and need to
invalidate any other analyses, but some of the un-preserved analyses
might have already been invalidated *and recomputed*! We would fail to
propagate the fact that the analysis had already been invalidated.
The solution to this struck me as really strange at first, but the more
I thought about it, the more natural it seemed. After a nice discussion
with Duncan about it on IRC, it seemed even nicer. The idea is that
invalidating an analysis *causes* it to be preserved! Preserving the
lack of result is trivial. If it is recomputed, great. Until something
*else* invalidates it again, we're good.
The consequence of this is that the invalidate methods on the analysis
manager which operate over many passes now consume their
PreservedAnalyses object, update it to "preserve" every analysis pass to
which it delivers an invalidation (regardless of whether the pass
chooses to be removed, or handles the invalidation itself by updating
itself). Then we return this augmented set from the invalidate routine,
letting the pass manager take the result and use the intersection of
*that* across each pass run to compute the final preserved set. This
accounts for all the places where the early invalidation of an analysis
has already "preserved" it for a future run.
I've beefed up the testing and adjusted the assertions to show that we
no longer repeatedly invalidate or compute the analyses across nested
pass managers.
llvm-svn: 225333
Oleksiy Vyalov [Wed, 7 Jan 2015 01:28:37 +0000 (01:28 +0000)]
Make DynamicLoaderPOSIXDYLD::DidAttach to deduce a target executable by pid if no executable hasn't been assigned to a target so far.
http://reviews.llvm.org/D6740
llvm-svn: 225332
Tom Stellard [Wed, 7 Jan 2015 01:17:37 +0000 (01:17 +0000)]
R600/SI: Add check for amdgcn triple forgotten in r225276.
llvm-svn: 225331
Rui Ueyama [Wed, 7 Jan 2015 01:13:08 +0000 (01:13 +0000)]
[ELF] Remove {ELF,}GNULinkerScript.
Instead of representing a linker script file as an "InputElement",
parse and evaluate scripts in the driver as we see them.
Linker scripts are not regular input files (regular file is one of
object, archive, or shared library file). They are more like
extended command line options. Linker script handling was needlessly
complicated because of that inappropriate abstraction (besides
excessive class hierarchy -- there is no such thing like ELF linker
script but we had two classes there for some reason.)
LinkerScript was one of a few remaining InputElement subclasses
that can be expanded to multiple files. With this patch, we are one
step closer to retire InputElement.
http://reviews.llvm.org/D6648
llvm-svn: 225330
David Majnemer [Wed, 7 Jan 2015 00:39:50 +0000 (00:39 +0000)]
Analysis: Reformulate WillNotOverflowUnsignedAdd for reusability
WillNotOverflowUnsignedAdd's smarts will live in ValueTracking as
computeOverflowForUnsignedAdd. It now returns a tri-state result:
never overflows, always overflows and sometimes overflows.
llvm-svn: 225329
David Majnemer [Wed, 7 Jan 2015 00:39:42 +0000 (00:39 +0000)]
InstCombine: Just a small tidy-up
llvm-svn: 225328
Alexey Samsonov [Wed, 7 Jan 2015 00:38:00 +0000 (00:38 +0000)]
[Sanitizer] Change the runtime flag representation.
This mirrors r225239 to all the rest sanitizers:
ASan, DFSan, LSan, MSan, TSan, UBSan.
Now the runtime flag type, name, default value and
description is located in the single place in the
.inc file.
llvm-svn: 225327
Tobias Grosser [Wed, 7 Jan 2015 00:30:01 +0000 (00:30 +0000)]
Do not use a const Twine here
This has caused segfaults when using Polly in the context of Julia, that are
not reproducible on my machine in 'make check-polly'.
llvm-svn: 225326
Hal Finkel [Wed, 7 Jan 2015 00:15:29 +0000 (00:15 +0000)]
[PowerPC] Transform a README.txt entry into a FIXME
Remove the README.txt entry regarding register allocation of CR logical ops,
and replace it with a FIXME in PPCInstrInfo.td. The text in the README.txt was
not really accurate, and thanks goes to Pat Haugen (and Bill Schmidt) from IBM
for clarifying what was intended and highlighting the relevant text in the ISA
specification.
llvm-svn: 225325
Nico Weber [Tue, 6 Jan 2015 23:54:59 +0000 (23:54 +0000)]
Fix grammar-o in comment.
llvm-svn: 225324
Kostya Serebryany [Tue, 6 Jan 2015 23:53:32 +0000 (23:53 +0000)]
[asan] add a flag soft_rss_limit_mb
llvm-svn: 225323
Rafael Espindola [Tue, 6 Jan 2015 23:53:13 +0000 (23:53 +0000)]
Suppress clang warnings in a codegen test.
This makes the output of FileCheck way easier to read since this test hits
many warnings.
llvm-svn: 225322
Duncan P. N. Exon Smith [Tue, 6 Jan 2015 23:52:35 +0000 (23:52 +0000)]
cmake: Fix 'examples' target after r225319
Add the missing `DEPENDS` keyword. r225319 did almost the right thing
(I didn't notice the problem with it because `Kaleidoscope-Ch8` wasn't
building at all).
llvm-svn: 225321
Duncan P. N. Exon Smith [Tue, 6 Jan 2015 23:48:22 +0000 (23:48 +0000)]
Kaleidoscope: Value => Metadata
llvm-svn: 225320
Duncan P. N. Exon Smith [Tue, 6 Jan 2015 23:42:49 +0000 (23:42 +0000)]
cmake: Add 'examples' target
llvm-svn: 225319
Duncan P. N. Exon Smith [Tue, 6 Jan 2015 23:39:37 +0000 (23:39 +0000)]
cmake: Add Kaleidoscope target
llvm-svn: 225318
Vince Harron [Tue, 6 Jan 2015 23:38:24 +0000 (23:38 +0000)]
Adds UriParser::Parse and unit tests
This can be used to parse URIs passed to 'platform connect'
Differential Revision: http://reviews.llvm.org/D6858
llvm-svn: 225317
Greg Clayton [Tue, 6 Jan 2015 23:33:34 +0000 (23:33 +0000)]
Fix needed for the new terminal test I previously checked in. It was crashing due to a NULL dereference.
llvm-svn: 225316
Eric Christopher [Tue, 6 Jan 2015 23:23:24 +0000 (23:23 +0000)]
Add a subdirectory in CMake for Chapter 8.
llvm-svn: 225315
Matt Arsenault [Tue, 6 Jan 2015 23:14:57 +0000 (23:14 +0000)]
Add __builtin_amdgpu_class
llvm-svn: 225314
Rui Ueyama [Tue, 6 Jan 2015 23:06:49 +0000 (23:06 +0000)]
Convert CoreInputGraph.
This is a part of InputGraph cleanup to represent input files as a flat
list of Files (and some meta-nodes for group etc.)
We cannot achieve that goal in one gigantic patch, so I split the task
into small steps as shown below.
(Recap the progress so far: Currently InputGraph contains a list of
InputElements. Each InputElement contain one File (that used to have
multiple Files, but I eliminated that use case in r223867). Files are
currently instantiated in Driver::link(), but I already made a change
to separate file parsing from object instantiation (r224102), so we
can safely instantiate Files when we need them, instead of wrapping
a file with the wrapper class (FileNode class). InputGraph used to
act like a generator class by interpreting groups by itself, but it's
now just a container of a list of InputElements (r223867).)
1. Instantiate Files in the driver and wrap them with WrapperNode.
WrapperNode is a temporary class that allows us to instantiate Files
in the driver while keep using the current InputGraph data structure.
This patch demonstrates how this step 1 looks like, using Core driver
as an example.
2. Do the same thing for the other drivers.
When step 2 is done, an InputGraph consists of GroupEnd objects or
WrapperNodes each of which contains one File. Other types of
FileNode subclasses are removed.
3. Replace InputGraph with std::vector<std::unique_ptr<InputElement>>.
InputGraph is already just a container of list of InputElements,
so this step removes that needless class.
4. Remove WrapperNode.
We need some code cleanup between each step, because many classes
do a bit odd things (e.g. InputGraph::getGroupSize()). I'll straight
things up as I need to.
llvm-svn: 225313
Hal Finkel [Tue, 6 Jan 2015 23:06:41 +0000 (23:06 +0000)]
[PowerPC] Add support for -mcmpb
In r225106, support for the CMPB instruction was added to the PowerPC backend.
This adds the associated GCC-compatible feature flag.
llvm-svn: 225312
Lang Hames [Tue, 6 Jan 2015 23:04:36 +0000 (23:04 +0000)]
Revert r224935 "Refactor duplicated code. No intended functionality change."
This is affecting the behavior of some ObjC++ / AArch64 test cases on Darwin.
Reverting to get the bots green while I track down the source of the changed
behavior.
llvm-svn: 225311
Matt Arsenault [Tue, 6 Jan 2015 23:00:46 +0000 (23:00 +0000)]
R600/SI: Add combine for isinfinite pattern
llvm-svn: 225310
Matt Arsenault [Tue, 6 Jan 2015 23:00:44 +0000 (23:00 +0000)]
Add isNegative helper to ConstantFPSDNode
llvm-svn: 225309
Matt Arsenault [Tue, 6 Jan 2015 23:00:43 +0000 (23:00 +0000)]
Add isInfinity helper to ConstantFPSDNode
llvm-svn: 225308
Matt Arsenault [Tue, 6 Jan 2015 23:00:41 +0000 (23:00 +0000)]
R600/SI: Pattern match isinf to v_cmp_class instructions
llvm-svn: 225307
Matt Arsenault [Tue, 6 Jan 2015 23:00:39 +0000 (23:00 +0000)]
R600/SI: Add basic DAG combines for fp_class
llvm-svn: 225306
Matt Arsenault [Tue, 6 Jan 2015 23:00:37 +0000 (23:00 +0000)]
R600/SI: Add class intrinsic
llvm-svn: 225305
Matt Arsenault [Tue, 6 Jan 2015 23:00:33 +0000 (23:00 +0000)]
Fix using wrong intrinsic in test
This is a leftover from renaming the intrinsic.
It's surprising the unknown llvm. intrinsic wasn't rejected.
llvm-svn: 225304
Rafael Espindola [Tue, 6 Jan 2015 22:55:40 +0000 (22:55 +0000)]
Update for .ll syntax change.
llvm-svn: 225303
Rafael Espindola [Tue, 6 Jan 2015 22:55:16 +0000 (22:55 +0000)]
Change the .ll syntax for comdats and add a syntactic sugar.
In order to make comdats always explicit in the IR, we decided to make
the syntax a bit more compact for the case of a GlobalObject in a
comdat with the same name.
Just dropping the $name causes problems for
@foo = globabl i32 0, comdat
$bar = comdat ...
and
declare void @foo() comdat
$bar = comdat ...
So the syntax is changed to
@g1 = globabl i32 0, comdat($c1)
@g2 = globabl i32 0, comdat
and
declare void @foo() comdat($c1)
declare void @foo() comdat
llvm-svn: 225302
Hal Finkel [Tue, 6 Jan 2015 22:31:02 +0000 (22:31 +0000)]
[PowerPC] Reuse a load operand in int->fp conversions
int->fp conversions on PPC must be done through memory loads and stores. On a
modern core, this process begins by storing the int value to memory, then
loading it using a (sometimes special) FP load instruction. Unfortunately, we
would do this even when the value to be converted was itself a load, and we can
just use that same memory location instead of copying it to another first.
There is a slight complication when handling int_to_fp(fp_to_int(x)) pairs,
because the fp_to_int operand has not been lowered when the int_to_fp is being
lowered. We handle this specially by invoking fp_to_int's lowering logic
(partially) and getting the necessary memory location (some trivial refactoring
was done to make this possible).
This is all somewhat ugly, and it would be nice if some later CodeGen stage
could just clean this stuff up, but because doing so would involve modifying
target-specific nodes (or instructions), it is not immediately clear how that
would work.
Also, remove a related entry from the README.txt for which we now generate
reasonable code.
llvm-svn: 225301
Dan Albert [Tue, 6 Jan 2015 22:18:27 +0000 (22:18 +0000)]
Move a test to the new tests directory.
I had written the patch that added this test before the tests were
moved into tests/std, so the test ended up in the wrong directory.
llvm-svn: 225300
Ed Maste [Tue, 6 Jan 2015 22:13:48 +0000 (22:13 +0000)]
Put FreeBSD note types in their own namespace
Note types are inherently OS-specific, but some note type names are
common to both FreeBSD and Linux.
llvm-svn: 225299
Alexey Samsonov [Tue, 6 Jan 2015 20:58:40 +0000 (20:58 +0000)]
[Sanitizer] Improve unit tests in COMPILER_RT_DEBUG=ON mode.
Propagate -DSANITIZER_DEBUG definition to unit tests.
Make sure unit tests depend on compiler-rt headers.
llvm-svn: 225298
Zachary Turner [Tue, 6 Jan 2015 20:56:12 +0000 (20:56 +0000)]
Remove a lock acquisition from ProcessWindows::OnExitProcess.
This was causing a race condition where DoDestroy() would acquire
the lock and then initiate a shutdown and then wait for it to
complete. But part of the shutdown involved acquiring the same
lock from a different thread. So the main thread would timeout
waiting for the shutdown to complete and return too soon.
The end result of this is that SBProcess::Kill() was broken on
Windows.
llvm-svn: 225297
Tom Stellard [Tue, 6 Jan 2015 20:42:12 +0000 (20:42 +0000)]
Use amdgcn triple for SI+ GPUs
llvm-svn: 225296
Tobias Grosser [Tue, 6 Jan 2015 20:40:33 +0000 (20:40 +0000)]
Make registerPollyPasses public
This function is needed for the integration of Polly into Julia.
llvm-svn: 225295
Tom Stellard [Tue, 6 Jan 2015 20:34:47 +0000 (20:34 +0000)]
R600: Handle amdgcn triple
For now there is no difference between amdgcn and r600.
llvm-svn: 225294
Alexey Samsonov [Tue, 6 Jan 2015 20:25:34 +0000 (20:25 +0000)]
Use -DSANITIZER_DEBUG=1 if COMPILER_RT_DEBUG=ON, now that PR22083 is fixed.
llvm-svn: 225293
Mehdi Amini [Tue, 6 Jan 2015 20:05:02 +0000 (20:05 +0000)]
Use a Factory Method for MachineFunctionInfo Creation
The goal is to allows MachineFunctionInfo to override this create
function to customize the creation.
No change intended in existing backend in this patch.
llvm-svn: 225292
Colin LeMahieu [Tue, 6 Jan 2015 20:03:31 +0000 (20:03 +0000)]
[Hexagon] Adding compound jump encodings.
llvm-svn: 225291
Tom Stellard [Tue, 6 Jan 2015 19:52:07 +0000 (19:52 +0000)]
R600/SI: Insert s_waitcnt before s_barrier instructions.
This ensures that all memory operations are complete when all threads
reach the barrier.
llvm-svn: 225290
Tom Stellard [Tue, 6 Jan 2015 19:52:04 +0000 (19:52 +0000)]
R600/SI: Fix dependency calculation for DS writes instructions in SIInsertWaits
In DS write instructions, the address operand comes before the value
operand(s) which is reversed from every other instruction type.
The SIInsertWait assumed that the first use for each instruction
was the value, so for DS write it was protecting the address
operand with s_waitcnt instructions when it should have been
protecting the value operand.
llvm-svn: 225289
Adrian Prantl [Tue, 6 Jan 2015 19:47:27 +0000 (19:47 +0000)]
Revert "Reapply: Teach SROA how to update debug info for fragmented variables."
because of a tsan buildbot failure.
This reverts commit 225272.
Fix should be coming soon.
llvm-svn: 225288
Dan Albert [Tue, 6 Jan 2015 19:32:30 +0000 (19:32 +0000)]
Make a test UNSUPPORTED if libcpp-has-no-threads.
llvm-svn: 225287
Dan Albert [Tue, 6 Jan 2015 19:23:25 +0000 (19:23 +0000)]
Appease MSAN buildbots.
This is just a compile time test, but we have MSAN buildbots that will
fail since `exp` was uninitialized.
llvm-svn: 225286
Marshall Clow [Tue, 6 Jan 2015 19:20:49 +0000 (19:20 +0000)]
Fix PR 22106; make std::swap work for multi-dimensional arrays. Thanks to Peter Griess for the report and suggested fix
llvm-svn: 225285
Greg Clayton [Tue, 6 Jan 2015 19:17:58 +0000 (19:17 +0000)]
Added a test case for launching a process in a separate terminal window to ensure we don't regress on this.
A recent POSIX host thread issue where HostThreadPosix::Join() wasn't returning the thread result was responsible for this regression, yet we had no test case covering this so it wasn't discovered.
llvm-svn: 225284
Colin LeMahieu [Tue, 6 Jan 2015 19:03:20 +0000 (19:03 +0000)]
[Hexagon] Adding encoding for misc v4 instructions: boundscheck, tlbmatch, dcfetch.
llvm-svn: 225283
Sanjoy Das [Tue, 6 Jan 2015 19:02:56 +0000 (19:02 +0000)]
This patch teaches IndVarSimplify to add nuw and nsw to certain kinds
of operations that provably don't overflow. For example, we can prove
%civ.inc below does not sign-overflow. With this change,
IndVarSimplify changes %civ.inc to an add nsw.
define i32 @foo(i32* %array, i32* %length_ptr, i32 %init) {
entry:
%length = load i32* %length_ptr, !range !0
%len.sub.1 = sub i32 %length, 1
%upper = icmp slt i32 %init, %len.sub.1
br i1 %upper, label %loop, label %exit
loop:
%civ = phi i32 [ %init, %entry ], [ %civ.inc, %latch ]
%civ.inc = add i32 %civ, 1
%cmp = icmp slt i32 %civ.inc, %length
br i1 %cmp, label %latch, label %break
latch:
store i32 0, i32* %array
%check = icmp slt i32 %civ.inc, %len.sub.1
br i1 %check, label %loop, label %break
break:
ret i32 %civ.inc
exit:
ret i32 42
}
Differential Revision: http://reviews.llvm.org/D6748
llvm-svn: 225282
Kostya Serebryany [Tue, 6 Jan 2015 18:48:49 +0000 (18:48 +0000)]
[ubsan] disable coverage-levels.cc on darwin until we can investigate what's wrong there.
llvm-svn: 225281
Dan Albert [Tue, 6 Jan 2015 18:39:37 +0000 (18:39 +0000)]
Obey [atomics.types.operations.req]/21 for GCC.
Summary:
Excerpt from [atomics.types.operations.req]/21:
> When only one memory_order argument is supplied, the value of
> success is order, and the value of failure is order except that a
> value of memory_order_acq_rel shall be replaced by the value
> memory_order_acquire and a value of memory_order_release shall be
> replaced by the value memory_order_relaxed.
Clean up some copy pasta while I'm here (someone added a return
statement to a void function).
Reviewers: EricWF, jroelofs, mclow.lists
Reviewed By: mclow.lists
Subscribers: cfe-commits
Differential Revision: http://reviews.llvm.org/D6632
llvm-svn: 225280
Colin LeMahieu [Tue, 6 Jan 2015 18:38:26 +0000 (18:38 +0000)]
[Hexagon] Adding encoding information for absolute address loads.
llvm-svn: 225279
Mehdi Amini [Tue, 6 Jan 2015 18:20:04 +0000 (18:20 +0000)]
SelectionDAGBuilder: move constant initialization out of loop
No semantic change intended.
Reviewers: resistor
Differential Revision: http://reviews.llvm.org/D6834
llvm-svn: 225278
Tom Stellard [Tue, 6 Jan 2015 18:00:21 +0000 (18:00 +0000)]
R600/SI: Add a stub GCNTargetMachine
This is equivalent to the AMDGPUTargetMachine now, but it is the
starting point for separating R600 and GCN functionality into separate
targets.
It is recommened that users start using the gcn triple for GCN-based
GPUs, because using the r600 triple for these GPUs will be deprecated in
the future.
llvm-svn: 225277
Tom Stellard [Tue, 6 Jan 2015 18:00:00 +0000 (18:00 +0000)]
Triple: Add amdgcn triple
This will be used for AMD GPUs with the Graphics Core Next architecture,
which are currently using by the r600 triple.
llvm-svn: 225276
Tom Stellard [Tue, 6 Jan 2015 17:59:56 +0000 (17:59 +0000)]
R600/SI: Remove MachineFunction dump from AsmPrinter
The dump was dependent on a feature string, which meant that it couldn't
be disabled or enable on a per compile basis.
llvm-svn: 225275
Andrea Di Biagio [Tue, 6 Jan 2015 17:41:18 +0000 (17:41 +0000)]
[CodeGenPrepare] Improved logic to speculate calls to cttz/ctlz.
This patch improves the logic added at revision 224899 (see review D6728) that
teaches the backend when it is profitable to speculate calls to cttz/ctlz.
The original algorithm conservatively avoided speculating more than one
instruction from a basic block in a control flow grap modelling an if-statement.
In particular, the only allowed instruction (excluding the terminator) was a
call to cttz/ctlz. However, there are cases where we could be less conservative
and still be able to speculate a call to cttz/ctlz.
With this patch, CodeGenPrepare now tries to speculate a cttz/ctlz if the
result is zero extended/truncated in the same basic block, and the zext/trunc
instruction is "free" for the target.
Added new test cases to CodeGen/X86/cttz-ctlz.ll
Differential Revision: http://reviews.llvm.org/D6853
llvm-svn: 225274
Dan Albert [Tue, 6 Jan 2015 17:34:51 +0000 (17:34 +0000)]
[libcxx] Set _LIBCPP_ELAST for mingw.
Reviewers: K-ballo, mclow.lists, EricWF
Reviewed By: EricWF
Subscribers: jfb, jroelofs, majnemer, cfe-commits
Differential Revision: http://reviews.llvm.org/D6558
llvm-svn: 225273
Adrian Prantl [Tue, 6 Jan 2015 17:14:10 +0000 (17:14 +0000)]
Reapply: Teach SROA how to update debug info for fragmented variables.
This also rolls in the changes discussed in http://reviews.llvm.org/D6766.
Defers migrating the debug info for new allocas until after all partitions
are created.
Thanks to Chandler for reviewing!
llvm-svn: 225272
Filipe Cabecinhas [Tue, 6 Jan 2015 17:08:26 +0000 (17:08 +0000)]
Don't loop endlessly for MachO files with 0 ncmds
llvm-svn: 225271
Colin LeMahieu [Tue, 6 Jan 2015 16:52:38 +0000 (16:52 +0000)]
[Hexagon] Fix 225267. GP register is not yet fully implemented. Removing Uses [GP] maintains existing behavior.
llvm-svn: 225270
Adrian Prantl [Tue, 6 Jan 2015 16:50:25 +0000 (16:50 +0000)]
Implement a very basic colored syntax highlighting for llvm-dwarfdump.
The color scheme is the same as the one used by the colorize dwarfdump
script on Darwin.
A new --color option can be used to forcibly turn color on or off.
http://reviews.llvm.org/D6852
llvm-svn: 225269
Hal Finkel [Tue, 6 Jan 2015 16:46:37 +0000 (16:46 +0000)]
[PowerPC] Add a regression test for r225251
In r225251, I removed an old entry from the README.txt file. While there are
several contributing factors (including pieces in Clang's ABI code), upon
further reflection, the backend part deserves a regression test.
llvm-svn: 225268
Colin LeMahieu [Tue, 6 Jan 2015 16:15:15 +0000 (16:15 +0000)]
[Hexagon] Adding dealloc_return encoding and absolute address stores.
llvm-svn: 225267
Asiri Rathnayake [Tue, 6 Jan 2015 15:55:09 +0000 (15:55 +0000)]
[ARM] Cleanup so_imm* tblgen defintions
No functional changes. Support for ARM's modified immediate syntax was added
in r223113 and r223115 (review: D6408). That patch introduced the mod_imm*
tblegen definitions which renders the existing so_imm* definitions redundant.
This patch gets rid of them completely.
Reviewed as: D6722
llvm-svn: 225266
Matt Arsenault [Tue, 6 Jan 2015 15:50:59 +0000 (15:50 +0000)]
Convert fcmp with 0.0 from casted integers to icmp
This is already handled in general when it is known the
conversion can't lose bits with smaller integer types
casted into wider floating point types.
This pattern happens somewhat often in GPU programs that cast
workitem intrinsics to float, which are often compared with 0.
Specifically handle the special case of compares with zero which
should also be known to not lose information. I had a more general
version of this which allows equality compares if the casted float is
exactly representable in the integer, but I'm not 100% confident that
is always correct.
Also fold cases that aren't integers to true / false.
llvm-svn: 225265
Tobias Grosser [Tue, 6 Jan 2015 12:57:44 +0000 (12:57 +0000)]
www: Draft changelog for the upcoming release.
llvm-svn: 225264
Tobias Grosser [Tue, 6 Jan 2015 12:31:31 +0000 (12:31 +0000)]
Fix html
llvm-svn: 225263
Tobias Grosser [Tue, 6 Jan 2015 12:30:26 +0000 (12:30 +0000)]
We currently do not create these result for polly
We should reconsider this after having switched to imath (instead of gmp)
as the default isl backend, as this would allow us to keep a copy of isl
in the polly svn and to consequently make it easier to distribute Polly.
llvm-svn: 225262
Jay Foad [Tue, 6 Jan 2015 10:01:15 +0000 (10:01 +0000)]
[asan] Bump the self-imposed stack limit from 128k to 256k. PowerPC64
Linux has 64k pages, so the old limit was only two pages. With ASLR the
initial sp might be right at the start of the second page, so the stack
will immediately grow down into the first page; and if you use all pages
of a limited stack then asan hits a kernel bug to do with how stack
guard pages are reported in /proc/self/maps:
http://lkml.iu.edu//hypermail/linux/kernel/1501.0/01025.html
We should still fix the underlying problems, but in the mean time this
patch makes the test work with 64k pages as well as it does with 4k
pages.
llvm-svn: 225261
NAKAMURA Takumi [Tue, 6 Jan 2015 09:44:44 +0000 (09:44 +0000)]
[CMake] Silence stderr on "COMMAND ${CMAKE_C_COMPILER} -Wl,--version". It was noisy during configuraion.
llvm-svn: 225260
NAKAMURA Takumi [Tue, 6 Jan 2015 09:44:29 +0000 (09:44 +0000)]
Reformat.
llvm-svn: 225259
Chandler Carruth [Tue, 6 Jan 2015 09:10:47 +0000 (09:10 +0000)]
[PM] Hide a function we only use in an assert behind NDEBUG.
llvm-svn: 225258
Chandler Carruth [Tue, 6 Jan 2015 09:06:35 +0000 (09:06 +0000)]
[PM] Introduce a utility pass that preserves no analyses.
Use this to test that path of invalidation. This test actually shows
redundant invalidation here that is really bad. I'm going to work on
fixing that next, but wanted to commit the test harness now that its all
working.
llvm-svn: 225257
Craig Topper [Tue, 6 Jan 2015 08:59:30 +0000 (08:59 +0000)]
[X86] Add OpSize32 to XBEGIN_4. Add XBEGIN_2 with OpSize16.
Requires new AsmParserOperand types that detect 16-bit and 32/64-bit mode so that we choose the right instruction based on default sizing without predicates. This is necessary since predicates mess up the disassembler table building.
llvm-svn: 225256
David Majnemer [Tue, 6 Jan 2015 08:41:31 +0000 (08:41 +0000)]
InstCombine: Bitcast call arguments from/to pointer/integer type
Try harder to get rid of bitcast'd calls by ptrtoint/inttoptr'ing
arguments and return values when DataLayout says it is safe to do so.
llvm-svn: 225254
Chandler Carruth [Tue, 6 Jan 2015 08:37:58 +0000 (08:37 +0000)]
[PM] Simplify how we parse the outer layer of the pass pipeline text and
remove an extra, redundant pass manager wrapping every run.
I had kept seeing these when manually testing, but it was getting really
annoying and was going to cause problems with overly eager invalidation.
The root cause was an overly complex and unnecessary pile of code for
parsing the outer layer of the pass pipeline. We can instead delegate
most of this to the recursive pipeline parsing.
I've added some somewhat more basic and precise tests to catch this.
llvm-svn: 225253
Craig Topper [Tue, 6 Jan 2015 07:35:50 +0000 (07:35 +0000)]
[X86] Make isel select the 2-byte register form of INC/DEC even in non-64-bit mode. Convert to the 1-byte form in non-64-bit mode as part of MCInst lowering.
Overall this seems simpler. It reduces duplication of patterns between both modes and it simplifies the memory folding/unfolding tables as they don't need to create fake instructions just to keep track of 64-bitness.
llvm-svn: 225252
Hal Finkel [Tue, 6 Jan 2015 07:23:13 +0000 (07:23 +0000)]
[PowerPC] Remove old README.txt entry regarding struct passing
Because of how Clang represents structs as arrays (at least on non-Darwin
platforms), and what SROA does, etc. this is no longer a problem.
llvm-svn: 225251