Richard Smith [Fri, 1 Aug 2014 22:17:28 +0000 (22:17 +0000)]
Actually fix problem with modules buildbot this time.
llvm-svn: 214579
NAKAMURA Takumi [Fri, 1 Aug 2014 22:15:38 +0000 (22:15 +0000)]
llvm/test/CodeGen/Mips/cconv/arguments-varargs.ll: Add explicit -mtriple=(mips|mipsel)-linux on 4 lines.
llvm-svn: 214578
Ben Langmuir [Fri, 1 Aug 2014 22:12:21 +0000 (22:12 +0000)]
Add -fbuild-session-file as an alternative to -fbuild-session-timestamp
Build systems tend to traffic in files and modification times, so having
them touch a file at the beginning of the build can be easier than
having them update the compile command they use every time they build.
llvm-svn: 214577
Adrian Prantl [Fri, 1 Aug 2014 22:11:58 +0000 (22:11 +0000)]
Debug info: Infrastructure to support debug locations for fragmented
variables (for example, by-value struct arguments passed in registers, or
large integer values split across several smaller registers).
On the IR level, this adds a new type of complex address operation OpPiece
to DIVariable that describes size and offset of a variable fragment.
On the DWARF emitter level, all pieces describing the same variable are
collected, sorted and emitted as DWARF expressions using the DW_OP_piece
and DW_OP_bit_piece operators.
http://reviews.llvm.org/D3373
rdar://problem/
15928306
What this patch doesn't do / Future work:
- This patch only adds the backend machinery to make this work, patches
that change SROA and SelectionDAG's type legalizer to actually create
such debug info will follow. (http://reviews.llvm.org/D2680)
- Making the DIVariable complex expressions into an argument of dbg.value
will reduce the memory footprint of the debug metadata.
- The sorting/uniquing of pieces should be moved into DebugLocEntry,
to facilitate the merging of multi-piece entries.
llvm-svn: 214576
Greg Clayton [Fri, 1 Aug 2014 22:10:13 +0000 (22:10 +0000)]
Don't hand compute mydir, do it using "TestBase.compute_mydir(__file__)".
llvm-svn: 214575
Chandler Carruth [Fri, 1 Aug 2014 22:09:43 +0000 (22:09 +0000)]
[SDAG] MorphNodeTo recursively deletes dead operands of the old
fromulation of the node, which isn't really the desired behavior from
within the combiner or legalizer, but is necessary within ISel. I've
added a hopefully helpful comment and fixed the only two places where
this took place.
Yet another step toward the combiner and legalizer not needing to use
update listeners with virtual calls to manage the worklists behind
legalization and combining.
llvm-svn: 214574
Jonathan Roelofs [Fri, 1 Aug 2014 22:02:21 +0000 (22:02 +0000)]
[unwinder] Tell the assembler that functions are functions so that the linker knows how to link them.
This fixes an interworking problem when the unwinder/libcxxabi is built for
Thumb. When unw_getcontext is not marked as a function, 'bl' is used for the
branch instead of 'bx'.
llvm-svn: 214573
Tom Stellard [Fri, 1 Aug 2014 21:55:50 +0000 (21:55 +0000)]
Revert "R600: Move code for generating REGISTER_LOAD into R600ISelLowering.cpp"
This reverts commit r214566.
I did not mean to commit this yet.
llvm-svn: 214572
Reid Kleckner [Fri, 1 Aug 2014 21:54:37 +0000 (21:54 +0000)]
MS inline asm: Hide symbol to attempt to fix test failure on darwin
If the symbol comes from an external DSO, it apparently requires
indirection through a register.
llvm-svn: 214571
Duncan P. N. Exon Smith [Fri, 1 Aug 2014 21:51:52 +0000 (21:51 +0000)]
BitcodeReader: Change mechanics of BlockAddress forward references, NFC
Now that we can reliably handle forward references to `BlockAddress`
(r214563), change the mechanics to simplify predicting use-list order.
Previously, we created dummy `GlobalVariable`s to represent block
addresses. After every function was materialized, we'd go through any
forward references to its blocks and RAUW them with a proper
`BlockAddress` constant. This causes some (potentially a lot of)
unnecessary use-list churn, since any constant expression that it's a
part of will need to be rematerialized as well.
Instead, pre-construct a `BasicBlock` immediately -- without attaching
it to its (empty) `Function` -- and use that to construct a
`BlockAddress`. This constant will not have to be regenerated. When
the function body is parsed, hook this pre-constructed basic block up
in the right place using `BasicBlock::insertInto()`.
Both before and after this change, the IR is temporarily in an invalid
state that gets resolved when `materializeForwardReferencedFunctions()`
gets called.
This is a prep commit that's part of PR5680, but the only functionality
change is the reduction of churn in the constant pool.
llvm-svn: 214570
Tom Stellard [Fri, 1 Aug 2014 21:51:05 +0000 (21:51 +0000)]
R600/SI: Remove leftover debugging code
llvm-svn: 214569
Jan Vesely [Fri, 1 Aug 2014 21:50:59 +0000 (21:50 +0000)]
relational: Add islessequal(floatN) builtin
v2: remove the initial undef
Signed-off-by: Jan Vesely <jan.vesely@rutgers.edu>
Reviewed-by: Aaron Watry <awatry@gmail.com>
llvm-svn: 214568
Jan Vesely [Fri, 1 Aug 2014 21:50:55 +0000 (21:50 +0000)]
relational: Add isless(floatN) builtin
v2: remove the initial undef
Signed-off-by: Jan Vesely <jan.vesely@rutgers.edu>
Reviewed-by: Aaron Watry <awatry@gmail.com>
llvm-svn: 214567
Tom Stellard [Fri, 1 Aug 2014 21:50:47 +0000 (21:50 +0000)]
R600: Move code for generating REGISTER_LOAD into R600ISelLowering.cpp
SI doesn't use REGISTER_LOAD anymore, but it was still hitting this code
path for 8-bit and 16-bit private loads.
llvm-svn: 214566
Reid Kleckner [Fri, 1 Aug 2014 21:40:53 +0000 (21:40 +0000)]
docs: Strongly recommend setting rpath when using a local GCC toolchain
Users keep emailing us about the difficulties of getting LD_LIBRARY_PATH
into their environment, which should be completely unecessary. Try to
strengthen the rpath recommentation by putting in an example cmake
invocation.
Speaking of which, we might want to make CMake the recommended build
system in GettingStarted.html.
llvm-svn: 214565
Alexey Samsonov [Fri, 1 Aug 2014 21:35:28 +0000 (21:35 +0000)]
[Sanitizer] Introduce SanitizerMetadata class.
It is responsible for generating metadata consumed by sanitizer instrumentation
passes in the backend. Move several methods from CodeGenModule to SanitizerMetadata.
For now the class is stateless, but soon it won't be the case.
Instead of creating globals providing source-level information to ASan, we will create
metadata nodes/strings which will be turned into actual global variables in the
backend (if needed).
No functionality change.
llvm-svn: 214564
Duncan P. N. Exon Smith [Fri, 1 Aug 2014 21:22:04 +0000 (21:22 +0000)]
IR: Add BasicBlock::insertInto()
Although unlinked `BasicBlock`s can be created, there's currently no way
to insert them into `Function`s after the fact. In particular,
`moveAfter()` and `moveBefore()` require that the basic block is already
linked.
Extract the logic for initially linking a `BasicBlock` out of the
constructor and into a member function that can be used for lazy
insertion.
- Asserts that the basic block is currently unlinked.
- Matches the logic of the constructor.
- Changed the constructor to use it since the logic matches.
This is needed in a follow-up commit for PR5680.
llvm-svn: 214563
Joerg Sonnenberger [Fri, 1 Aug 2014 21:20:02 +0000 (21:20 +0000)]
Let's assume PowerPC has no SSE.
llvm-svn: 214562
Peter Collingbourne [Fri, 1 Aug 2014 21:18:18 +0000 (21:18 +0000)]
[dfsan] Correctly handle loads and stores of zero size.
llvm-svn: 214561
Eric Christopher [Fri, 1 Aug 2014 21:18:01 +0000 (21:18 +0000)]
Add a non-const subtarget returning function to the target machine
so that we can use it to get the old-style JIT out of the subtarget.
This code should be removed when the old-style JIT is removed
(imminently).
llvm-svn: 214560
Duncan P. N. Exon Smith [Fri, 1 Aug 2014 21:11:34 +0000 (21:11 +0000)]
BitcodeReader: Fix some BlockAddress forward reference corner cases
`BlockAddress`es are interesting in that they can reference basic blocks
from *outside* the block's function. Since basic blocks are not global
values, this presents particular challenges for lazy parsing.
One corner case was found in PR11677 and fixed in r147425. In that
case, a global variable references a block address. It's necessary to
load the relevant function to resolve the forward reference before doing
anything with the module.
By inspection, I found (and have fixed here) two other cases:
- An instruction from one function references a block address from
another function, and only the first function is lazily loaded.
I fixed this the same way as PR11677: by eagerly loading the
referenced function.
- A function whose block address is taken is dematerialized, leaving
invalid references to it.
I fixed this by refusing to dematerialize functions whose block
addresses are taken (if you have to load it, you can't unload it).
llvm-svn: 214559
Duncan P. N. Exon Smith [Fri, 1 Aug 2014 21:06:59 +0000 (21:06 +0000)]
Try to fix configure+make after r214556
llvm-svn: 214558
Duncan P. N. Exon Smith [Fri, 1 Aug 2014 21:01:04 +0000 (21:01 +0000)]
Rewrite BitReaderTest, NFC
Rewrite the single unit test in `BitReaderTest` so that it's easier to
add more tests.
- Parse from an assembly string rather than using API.
- Use more helper functions.
- Use a separate context for the module on the other side.
Aside from relying on the assembly parser, there's no functionality
change intended.
llvm-svn: 214556
Richard Smith [Fri, 1 Aug 2014 20:39:36 +0000 (20:39 +0000)]
Re-commit r214547 with tests fixed. Hopefully all the bots will be happy now.
Original message:
Fix iterator invalidation issues that are breaking my modules buildbot's
bootstrap.
llvm-svn: 214555
Reid Kleckner [Fri, 1 Aug 2014 20:23:29 +0000 (20:23 +0000)]
MS inline asm: Tests for r214550
These tests seem like an exception to the rule against assembly emitting
tests in clang. I made an LLVM side change that can only be tested by
setting up the inline assembly machinery that is only implemented by
Clang.
llvm-svn: 214552
Reid Kleckner [Fri, 1 Aug 2014 20:23:03 +0000 (20:23 +0000)]
MS inline asm: Add tests for LLVM r214468
This used to assert.
llvm-svn: 214551
Reid Kleckner [Fri, 1 Aug 2014 20:21:24 +0000 (20:21 +0000)]
MS inline asm: Use memory constraints for functions instead of registers
This is consistent with how we parse them in a standalone .s file, and
inline assembly shouldn't differ.
This fixes errors about requiring more registers than available in
cases like this:
void f();
void __declspec(naked) g() {
__asm pusha
__asm call f
__asm popa
__asm ret
}
There are no registers available to pass the address of 'f' into the asm
blob. The asm should now directly call 'f'.
Tests will land in Clang shortly.
llvm-svn: 214550
Richard Smith [Fri, 1 Aug 2014 20:09:39 +0000 (20:09 +0000)]
Revert r214547 due to test breakage.
llvm-svn: 214549
Justin Bogner [Fri, 1 Aug 2014 19:59:48 +0000 (19:59 +0000)]
llvm-profdata: Replace redundant tests with more targeted ones
llvm-svn: 214548
Richard Smith [Fri, 1 Aug 2014 19:59:14 +0000 (19:59 +0000)]
Fix iterator invalidation issues that are breaking my modules buildbot's bootstrap.
llvm-svn: 214547
Chandler Carruth [Fri, 1 Aug 2014 19:49:59 +0000 (19:49 +0000)]
[SDAG] Begin simplifying the way in which the legalizer deletes nodes.
This lifts the (very few) places the legalizer would delete dead nodes
into the outer loop around the legalizer. This is significantly simpler
because it doesn't require the legalizer itself to manage the iterator
validity, and it doesn't require the legalizer to be a DAG update
listener in order to remove things from the legalized set. It also makes
the interface much less contrived for the case of the legalizer running
inside the last phase of DAG combining.
I'm working on centralizing the deletion of nodes during both legalizing
and combining as much as possible. My hope is to remove the need for DAG
update listeners from the combiner next, which would remove a costly
virtual dispatch chain on every deletion. This in turn should allow us
to more aggressively delete DAG nodes during combining which will in
turn allow us to combine more aggressively by exposing the actual nodes
which have single users to the combine phases.
llvm-svn: 214546
Juergen Ributzka [Fri, 1 Aug 2014 19:40:16 +0000 (19:40 +0000)]
[FastISel][AArch64] Fold offset into the memory operation.
Fold simple offsets into the memory operation:
add x0, x0, #8
ldr x0, [x0]
-->
ldr x0, [x0, #8]
Fixes <rdar://problem/
17887945>.
llvm-svn: 214545
Viktor Kutuzov [Fri, 1 Aug 2014 19:37:05 +0000 (19:37 +0000)]
Add FreeBSD support to the address sanitizer's deep_stack_uaf.cc test case
Differential Revision: http://reviews.llvm.org/D4668
llvm-svn: 214544
Viktor Kutuzov [Fri, 1 Aug 2014 19:33:14 +0000 (19:33 +0000)]
Fix the waitid sanitizer interceptor to work on FreeBSD in 32-bit mode
Differential Revision: http://reviews.llvm.org/D4670
llvm-svn: 214543
Rafael Espindola [Fri, 1 Aug 2014 19:28:15 +0000 (19:28 +0000)]
Include Archive.h
MSVC was complaining about Archive being an incomplete type.
llvm-svn: 214542
Viktor Kutuzov [Fri, 1 Aug 2014 19:23:15 +0000 (19:23 +0000)]
Fix building with in-tree libc++abi on FreeBSD
Differential Revision: http://reviews.llvm.org/D4743
llvm-svn: 214541
Rafael Espindola [Fri, 1 Aug 2014 18:49:24 +0000 (18:49 +0000)]
Move virtual method out of line.
Should fix the MSVC build.
llvm-svn: 214539
Philip Reames [Fri, 1 Aug 2014 18:47:09 +0000 (18:47 +0000)]
Add support for StackMap section for ELF/Linux systems
This patch adds code to emits the StackMap section on ELF systems. This section is required to support llvm.experimental.stackmap and llvm.experimental.patchpoint intrinsics.
Reviewers: ributzka, echristo
Differential Revision: http://reviews.llvm.org/D4574
llvm-svn: 214538
Juergen Ributzka [Fri, 1 Aug 2014 18:39:24 +0000 (18:39 +0000)]
[FastISel][AArch64] Add branch weights.
Add branch weights to branch instructions, so that the following passes can
optimize based on it (i.e. basic block ordering).
Fixes <rdar://problem/
17887137>.
llvm-svn: 214537
Greg Clayton [Fri, 1 Aug 2014 18:32:07 +0000 (18:32 +0000)]
Convert over to using pipes for the initial LLDB commands in the Driver.
This will avoid having to use a temp file and adding temp file related calls to the public API.
llvm-svn: 214536
Rafael Espindola [Fri, 1 Aug 2014 18:31:17 +0000 (18:31 +0000)]
Use object::Archive::create instead of new object::Archive.
Also fix the error handling. No testcaes, issue found by inspection.
Thanks to David Blaikie for the suggestion.
llvm-svn: 214535
Philip Reames [Fri, 1 Aug 2014 18:26:27 +0000 (18:26 +0000)]
Explicitly report runtime stack realignment in StackMap section
This change adds code to explicitly mark a function which requires runtime stack realignment as not having a fixed frame size in the StackMap section. As it happens, this is not actually a functional change. The size that would be reported without the check is also "-1", but as far as I can tell, that's an accident. The code change makes this explicit.
Note: There's a separate bug in handling of stackmaps and patchpoints in functions which need dynamic frame realignment. The current code assumes that offsets can be calculated from RBP, but realigned frames must use RSP. (There's a variable gap between RBP and the spill slots.) This change set does not address that issue.
Reviewers: atrick, ributzka
Differential Revision: http://reviews.llvm.org/D4572
llvm-svn: 214534
Rafael Espindola [Fri, 1 Aug 2014 18:09:32 +0000 (18:09 +0000)]
Replace comment about ownership with std::unique_ptr.
llvm-svn: 214533
Ehsan Akhgari [Fri, 1 Aug 2014 18:04:53 +0000 (18:04 +0000)]
[ASan] Support the asan_loadN/asan_storeN functions in the DLL thunk library
Summary:
This is required for linking DLLs with large functions exceeding
san-instrumentation-with-call-threshold. One such function is
vp9_fdct16x16_sse2 in libvpx.
Reviewers: timurrrr
Subscribers: llvm-commits
Differential Revision: http://reviews.llvm.org/D4758
llvm-svn: 214532
Juergen Ributzka [Fri, 1 Aug 2014 18:04:14 +0000 (18:04 +0000)]
[FastISel][ARM] Do not emit stores for undef arguments.
This is a followup patch for r214366, which added the same behavior to the
AArch64 and X86 FastISel code. This fix reproduces the already existing
behavior of SelectionDAG in FastISel.
llvm-svn: 214531
Rafael Espindola [Fri, 1 Aug 2014 18:04:14 +0000 (18:04 +0000)]
Use range loop.
llvm-svn: 214530
Hans Wennborg [Fri, 1 Aug 2014 17:38:53 +0000 (17:38 +0000)]
Update build instructions for the clang-format VS plugin
llvm-svn: 214529
Renato Golin [Fri, 1 Aug 2014 17:27:31 +0000 (17:27 +0000)]
Add missing breaks to AArch64InstrInfo::isGPRCopy
llvm-svn: 214528
Ed Maste [Fri, 1 Aug 2014 17:04:31 +0000 (17:04 +0000)]
Add decorator for FreeBSD test failure - malloc() breakpoint issue
llvm.org/pr20476
llvm-svn: 214527
Aaron Ballman [Fri, 1 Aug 2014 17:02:34 +0000 (17:02 +0000)]
Dropping some else-after-returns. No functional changes intended.
llvm-svn: 214526
Matt Arsenault [Fri, 1 Aug 2014 17:00:29 +0000 (17:00 +0000)]
R600: Cleanup test
Remove -CHECKs, use multiple prefixes, name values,
also test the @llvm.fabs version
llvm-svn: 214525
Matt Arsenault [Fri, 1 Aug 2014 17:00:27 +0000 (17:00 +0000)]
Make getNamedOperandIdx readonly
llvm-svn: 214524
Matt Arsenault [Fri, 1 Aug 2014 17:00:26 +0000 (17:00 +0000)]
R600/SI: Don't display GDS bit for read2
This isn't displayed for any other instructions anymore,
and isn't ever used.
llvm-svn: 214523
Ed Maste [Fri, 1 Aug 2014 16:57:46 +0000 (16:57 +0000)]
Fix test execution: the cmake keyword is CACHE, not CACHED
llvm-svn: 214522
Chad Rosier [Fri, 1 Aug 2014 15:30:41 +0000 (15:30 +0000)]
[AArch64] Fix test from r214518 in an attempt to appease buildbots.
llvm-svn: 214521
Eli Bendersky [Fri, 1 Aug 2014 15:01:10 +0000 (15:01 +0000)]
Add IR Mangler for more stable mangling.
llvm-svn: 214520
Rafael Espindola [Fri, 1 Aug 2014 14:57:05 +0000 (14:57 +0000)]
Remove lto_codegen_set_attr.
It was never exported, so no functionality change.
llvm-svn: 214519
Chad Rosier [Fri, 1 Aug 2014 14:48:56 +0000 (14:48 +0000)]
[AArch64] Generate tbz/tbnz when comparing against zero.
The tbz/tbnz checks the sign bit to convert
op w1, w1, w10
cmp w1, #0
b.lt .LBB0_0
to
op w1, w1, w10
tbnz w1, #31, .LBB0_0
Differential Revision: http://reviews.llvm.org/D4440
llvm-svn: 214518
Ulrich Weigand [Fri, 1 Aug 2014 14:35:58 +0000 (14:35 +0000)]
[PowerPC] PR20280 - Slots for byval parameters are not immutable
Found by inspection while looking at PR20280: code would mark slots
in the parameter save area where a byval parameter is passed as
"immutable". This is not correct since code is allowed to modify
byval parameters in place in the parameter save area.
llvm-svn: 214517
Rafael Espindola [Fri, 1 Aug 2014 14:31:55 +0000 (14:31 +0000)]
Remove some calls to std::move.
Instead of moving out the data in a ErrorOr<std::unique_ptr<Foo>>, get
a reference to it.
Thanks to David Blaikie for the suggestion.
llvm-svn: 214516
Rafael Espindola [Fri, 1 Aug 2014 14:16:40 +0000 (14:16 +0000)]
[pr20127] Check for leading \1 in the Twine version of getNameWithPrefix.
No functionality change, but will simplify an upcoming patch that uses the
Twine version.
llvm-svn: 214515
Rafael Espindola [Fri, 1 Aug 2014 14:11:14 +0000 (14:11 +0000)]
Simplify the code a bit with std::unique_ptr.
llvm-svn: 214514
Aaron Ballman [Fri, 1 Aug 2014 13:49:00 +0000 (13:49 +0000)]
The GNU-style aligned attribute has an optional expression, but the generated pretty printing logic was unaware of this. Fixed the pretty printing logic, and added a test to ensure it no longer asserts.
Added a FIXME to the code about eliding the parenthesis when pretty printing such a construct.
llvm-svn: 214513
Daniel Sanders [Fri, 1 Aug 2014 13:26:28 +0000 (13:26 +0000)]
Revert r214497: [mips] Defer va_arg expansion to the backend.
It appears that the backend does not handle all cases that were handled by clang.
In particular, it does not handle structs as used in
SingleSource/UnitTests/2003-05-07-VarArgs.
llvm-svn: 214512
Aaron Ballman [Fri, 1 Aug 2014 13:20:09 +0000 (13:20 +0000)]
Replacing some more complex logic with a helper function call to ObjCMethod::getReturnTypeSourceRange. No functional changes intended.
llvm-svn: 214511
Robert Lytton [Fri, 1 Aug 2014 13:11:46 +0000 (13:11 +0000)]
Xcore target: Fix CXXStdlibIncludeArgs to check 'nostdinc++' flag too.
Summary:
There are no tests as it is dependant upon the environment variables
XCC_C_INCLUDE_PATH & XCC_CPLUS_INCLUDE_PATH being set.
Differential Revision: http://reviews.llvm.org/D4621
llvm-svn: 214510
Tim Northover [Fri, 1 Aug 2014 13:07:19 +0000 (13:07 +0000)]
llvm-objdump: implement printing for MachO __compact_unwind info.
llvm-svn: 214509
Daniel Jasper [Fri, 1 Aug 2014 13:03:05 +0000 (13:03 +0000)]
clang-format: Add a space in ObjC protocols.
Before:
@interface Foo (HackStuff)<MyProtocol>
After:
@interface Foo (HackStuff) <MyProtocol>
llvm-svn: 214508
Aaron Ballman [Fri, 1 Aug 2014 12:58:11 +0000 (12:58 +0000)]
Improving diagnostic source ranges for the nonnull attribute. Now it highlights the attribute and the faulty nonpointer type when possible.
llvm-svn: 214507
Eli Bendersky [Fri, 1 Aug 2014 12:55:44 +0000 (12:55 +0000)]
Fix a memory leak - dispose of the CXString after printing it in mangling.
Thanks to kcc@ for noticing.
llvm-svn: 214506
James Molloy [Fri, 1 Aug 2014 12:42:11 +0000 (12:42 +0000)]
Allow only disassembling of M-class MSR masks that the assembler knows how to assemble back.
Note: The current code in DecodeMSRMask() rejects the unpredictable A/R MSR mask '0000' with Fail. The code in the patch follows this style and rejects unpredictable M-class MSR masks also with Fail (instead of SoftFail). If SoftFail is preferred in this case then additional changes to ARMInstPrinter (to print non-symbolic masks) and ARMAsmParser (to parse non-symbolic masks) will be needed.
Patch by Petr Pavlu!
llvm-svn: 214505
Aaron Ballman [Fri, 1 Aug 2014 12:41:37 +0000 (12:41 +0000)]
Initializing the loop hint option token again.
llvm-svn: 214504
Aaron Ballman [Fri, 1 Aug 2014 12:34:58 +0000 (12:34 +0000)]
Improve some const-correctness to remove a -Wcast-qual warning. No functional changes intended.
llvm-svn: 214503
Aaron Ballman [Fri, 1 Aug 2014 12:20:20 +0000 (12:20 +0000)]
Initializing the loop hint option token.
llvm-svn: 214502
Sylvestre Ledru [Fri, 1 Aug 2014 12:19:15 +0000 (12:19 +0000)]
Update the code with clang changes r214450 (FunctionProtoType::ExtProtoInfo update)
llvm-svn: 214501
Tilmann Scheller [Fri, 1 Aug 2014 12:08:04 +0000 (12:08 +0000)]
[ARM] Make the assembler reject unpredictable pre/post-indexed ARM LDRB/LDRSB instructions.
The ARM ARM prohibits LDRB/LDRSB instructions with writeback into the destination register. With this commit this constraint is now enforced and we stop assembling LDRH/LDRSH instructions with unpredictable behavior.
llvm-svn: 214500
Tilmann Scheller [Fri, 1 Aug 2014 11:33:47 +0000 (11:33 +0000)]
[ARM] Make the assembler reject unpredictable pre/post-indexed ARM LDRH/LDRSH instructions.
The ARM ARM prohibits LDRH/LDRSH instructions with writeback into the source register. With this commit this constraint is now enforced and we stop assembling LDRH/LDRSH instructions with unpredictable behavior.
llvm-svn: 214499
Tilmann Scheller [Fri, 1 Aug 2014 11:08:51 +0000 (11:08 +0000)]
[ARM] Make the assembler reject unpredictable pre/post-indexed ARM LDR instructions.
The ARM ARM prohibits LDR instructions with writeback into the destination register. With this commit this constraint is now enforced and we stop assembling LDR instructions with unpredictable behavior.
llvm-svn: 214498
Daniel Sanders [Fri, 1 Aug 2014 10:29:21 +0000 (10:29 +0000)]
[mips] Defer va_arg expansion to the backend.
Summary:
This patch causes clang to emit va_arg instructions to the backend instead of
expanding them into an implementation itself. The backend already implements
va_arg since this is necessary for NaCl so this patch is removing redundant
code.
Together with the llvm patch (D4556) that accounts for the effect of endianness
on the expansion of va_arg, this fixes PR19612.
Depends on D4556
Reviewers: sstankovic, dsanders
Reviewed By: dsanders
Subscribers: rnk, cfe-commits
Differential Revision: http://reviews.llvm.org/D4742
llvm-svn: 214497
Erik Eckstein [Fri, 1 Aug 2014 09:47:38 +0000 (09:47 +0000)]
SLPVectorizer: fix build problem in Release configuration
llvm-svn: 214496
Simon Atanasyan [Fri, 1 Aug 2014 09:47:21 +0000 (09:47 +0000)]
[Mips] Replace assembler code by YAML to make the 'interpreter.test' test
target independent.
llvm-svn: 214495
Erik Eckstein [Fri, 1 Aug 2014 09:20:42 +0000 (09:20 +0000)]
SLPVectorizer: improved scheduling algorithm.
llvm-svn: 214494
Daniel Sanders [Fri, 1 Aug 2014 09:17:39 +0000 (09:17 +0000)]
[mips][PR19612] Fix va_arg for big-endian mode.
Summary:
Big-endian mode was not correctly adjusting the offset for types smaller
than an ABI slot.
Fixes PR19612
Reviewers: dsanders
Reviewed By: dsanders
Subscribers: sstankovic, llvm-commits
Differential Revision: http://reviews.llvm.org/D4556
llvm-svn: 214493
Johannes Doerfert [Fri, 1 Aug 2014 08:44:49 +0000 (08:44 +0000)]
[Format] Remove blank line in Dependences.h
llvm-svn: 214492
Johannes Doerfert [Fri, 1 Aug 2014 08:20:26 +0000 (08:20 +0000)]
[Fix] Annotate the IslAst with broken reductions (Missing files)
+ test cases of r214489.
llvm-svn: 214491
Johannes Doerfert [Fri, 1 Aug 2014 08:18:39 +0000 (08:18 +0000)]
[Typo] Added dot at the end of a sentence.
llvm-svn: 214490
Johannes Doerfert [Fri, 1 Aug 2014 08:17:19 +0000 (08:17 +0000)]
Annotate the IslAst with broken reductions
+ Split all reduction dependences and map them to the causing memory accesses.
+ Print the types & base addresses of broken reductions for each "reduction
parallel" marked loop (OpenMP style).
+ 3 test cases to show how reductions are now represented in the isl ast.
The mapping "(ast) loops -> broken reductions" is also needed to find the
memory accesses we need to privatize in a loop.
llvm-svn: 214489
Johannes Doerfert [Fri, 1 Aug 2014 08:14:28 +0000 (08:14 +0000)]
Change the semantics of is*Parallel
The functions isParallel, isInnermostParallel and IsOutermostParallel in
IslAstInfo will now return true even in the presence of broken reductions.
To compensate for this change the negated result of isReductionParallel can
be used.
llvm-svn: 214488
Erik Eckstein [Fri, 1 Aug 2014 08:14:28 +0000 (08:14 +0000)]
SLP Vectorizer: added statistics counter
llvm-svn: 214487
Johannes Doerfert [Fri, 1 Aug 2014 08:13:25 +0000 (08:13 +0000)]
Change the printing of reduction types
We use the C operator representation when applicable.
+ Update all the test cases accordingly.
llvm-svn: 214486
Erik Eckstein [Fri, 1 Aug 2014 08:05:55 +0000 (08:05 +0000)]
SLP Vectorizer: improve canonicalize tree operands of commutitive binary operands.
This reverts r214338 (except the test file) and replaces it with a more general algorithm.
llvm-svn: 214485
Eric Fiselier [Fri, 1 Aug 2014 06:30:18 +0000 (06:30 +0000)]
Adding ABI information to linux test results
llvm-svn: 214484
Eric Fiselier [Fri, 1 Aug 2014 06:27:40 +0000 (06:27 +0000)]
Update information about compiler used during linux tests and reformat run information.
llvm-svn: 214483
Sylvestre Ledru [Fri, 1 Aug 2014 06:16:03 +0000 (06:16 +0000)]
Revert of 214418:
"Create a default symver on Linux like ELF OSes."
Fails the build under Debian with ld.gold:
/usr/bin/ld.gold: --default-symver: unknown option
llvm-svn: 214482
Hal Finkel [Fri, 1 Aug 2014 05:20:41 +0000 (05:20 +0000)]
[PowerPC] Generate unaligned vector loads using intrinsics instead of regular loads
Altivec vector loads on PowerPC have an interesting property: They always load
from an aligned address (by rounding down the address actually provided if
necessary). In order to generate an actual unaligned load, you can generate two
load instructions, one with the original address, one offset by one vector
length, and use a special permutation to extract the bytes desired.
When this was originally implemented, I generated these two loads using regular
ISD::LOAD nodes, now marked as aligned. Unfortunately, there is a problem with
this:
The alignment of a load does not contribute to its identity, and SDNodes
are uniqued. So, imagine that we have some unaligned load, L1, that is not
aligned. The routine will create two loads, L1(aligned) and (L1+16)(aligned).
Further imagine that there had already existed a load (L1+16)(unaligned) with
the same chain operand as the load L1. When (L1+16)(aligned) is created as part
of the lowering of L1, this load *is* also the (L1+16)(unaligned) node, just
now marked as aligned (because the new alignment overwrites the old). But the
original users of (L1+16)(unaligned) now get the data intended for the
permutation yielding the data for L1, and (L1+16)(unaligned) no longer exists
to get its own permutation-based expansion. This was PR19991.
A second potential problem has to do with the MMOs on these loads, which can be
used by AA during instruction scheduling to break chain-based dependencies. If
the new "aligned" loads get the MMO from the original unaligned load, this does
not represent the fact that it will load data from below the original address.
Normally, this would not matter, but this load might be combined with another
load pair for a previous vector, and then the dependency on the otherwise-
ignored lower bytes can matter.
To fix both problems, instead of generating the necessary loads using regular
ISD::LOAD instructions, ppc_altivec_lvx intrinsics are used instead. These are
provided with MMOs with a conservative address range.
Unfortunately, I no longer have a failing test case (since PR19991 was
reported, other changes in CodeGen have forced this bug back into hiding it
again). Nevertheless, this should fix the underlying problem.
llvm-svn: 214481
Matthew Gardiner [Fri, 1 Aug 2014 05:12:23 +0000 (05:12 +0000)]
Change the encoding of the Triple string exchanged across GDB-RSP
and update documentation to suit, as suggested by Jason Molenda and
discussed in:
http://lists.cs.uiuc.edu/pipermail/lldb-commits/Week-of-Mon-
20140721/011978.html
Differential Revision: http://reviews.llvm.org/D4704
llvm-svn: 214480
Suyog Sarda [Fri, 1 Aug 2014 05:07:20 +0000 (05:07 +0000)]
This patch implements transform for pattern "(A & ~B) ^ (~A) -> ~(A & B)".
Differential Revision: http://reviews.llvm.org/D4653
llvm-svn: 214479
Suyog Sarda [Fri, 1 Aug 2014 04:59:26 +0000 (04:59 +0000)]
This patch implements transform for pattern "(A | B) & ((~A) ^ B) -> (A & B)".
Differential Revision: http://reviews.llvm.org/D4628
llvm-svn: 214478
Suyog Sarda [Fri, 1 Aug 2014 04:50:31 +0000 (04:50 +0000)]
This patch implements transform for pattern "( A & (~B)) | (A ^ B) -> (A ^ B)"
Differential Revision: http://reviews.llvm.org/D4652
llvm-svn: 214477
Suyog Sarda [Fri, 1 Aug 2014 04:41:43 +0000 (04:41 +0000)]
This patch implements transform for pattern "(A & B) | ((~A) ^ B) -> (~A ^ B)".
Patch Credit to Ankit Jain !
Differential Revision: http://reviews.llvm.org/D4655
llvm-svn: 214476