Davide Italiano [Mon, 2 Mar 2015 06:17:38 +0000 (06:17 +0000)]
Update the list of relocations that need to be implemented.
While at it, point the correct document where the missing TLS
relocation(s) are described.
llvm-svn: 230937
Owen Anderson [Mon, 2 Mar 2015 06:00:02 +0000 (06:00 +0000)]
Teach DataLayout that zero-byte pointer sizes don't make sense.
Previously this would result in assertion failures or simply crashes
at various points in the optimizer when trying to create types of zero
bit width.
llvm-svn: 230936
Owen Anderson [Mon, 2 Mar 2015 05:25:09 +0000 (05:25 +0000)]
Teach the LLParser to fail gracefully when it encounters an invalid label name.
Previous it would either assert in +Asserts, or crash in -Asserts. Found by fuzzing LLParser.
llvm-svn: 230935
Owen Anderson [Mon, 2 Mar 2015 05:25:06 +0000 (05:25 +0000)]
Fix a crash in the LL parser where it failed to validate that the pointer operand of a GEP was valid.
This manifested as an assertion failure in +Asserts builds, and a hard crash in -Asserts builds. Found by fuzzing the LL parser.
llvm-svn: 230934
Zachary Turner [Mon, 2 Mar 2015 04:39:56 +0000 (04:39 +0000)]
[llvm-pdbdump] Many minor fixes and improvements
A short list of some of the improvements:
1) Now supports -all command line argument, which implies many
other command line arguments to simplify usage.
2) Now supports -no-compiler-generated command line argument to
exclude compiler generated types.
3) Prints base class list.
4) -class-definitions implies -types.
5) Proper display of bitfields.
6) Can now distinguish between struct/class/interface/union.
And a few other minor tweaks.
llvm-svn: 230933
Nico Weber [Mon, 2 Mar 2015 04:37:11 +0000 (04:37 +0000)]
Revert r230930, it caused PR22747.
llvm-svn: 230932
Craig Topper [Mon, 2 Mar 2015 03:33:11 +0000 (03:33 +0000)]
[X86] There are only 8 mask registers. Fail disassembly if instruction tries to reference more.
llvm-svn: 230931
Adrian Prantl [Mon, 2 Mar 2015 02:38:18 +0000 (02:38 +0000)]
Refactor DebugLocDWARFExpression so it doesn't require access to the
TargetRegisterInfo. DebugLocEntry now holds a buffer with the raw bytes
of the pre-calculated DWARF expression.
Ought to be NFC, but it does slightly alter the output format of the
textual assembly.
llvm-svn: 230930
NAKAMURA Takumi [Mon, 2 Mar 2015 01:14:03 +0000 (01:14 +0000)]
Revert r230921, "Revert some changes that were made to fix PR20680.", for now.
It caused a failure on clang/test/Misc/backend-optimization-failure.cpp .
llvm-svn: 230929
Nico Weber [Mon, 2 Mar 2015 01:12:28 +0000 (01:12 +0000)]
Fix ObjCInterfaceDecl::getCategoryMethod() and give it a caller. No behavior change.
llvm-svn: 230928
Benjamin Kramer [Mon, 2 Mar 2015 01:08:07 +0000 (01:08 +0000)]
Add another missing header that used to be included transitively.
llvm-svn: 230927
NAKAMURA Takumi [Mon, 2 Mar 2015 01:04:34 +0000 (01:04 +0000)]
llvm/examples: Add missing include according to r230907.
llvm-svn: 230926
Benjamin Kramer [Mon, 2 Mar 2015 00:48:06 +0000 (00:48 +0000)]
Add missing includes for make_unique, lld edition.
llvm-svn: 230925
Craig Topper [Mon, 2 Mar 2015 00:22:29 +0000 (00:22 +0000)]
[X86] Fix diassembler crash on AVX512 cmpps/cmppd with immediate that doesn't fit in 5-bits. Fixes PR22743.
llvm-svn: 230924
Sanjoy Das [Mon, 2 Mar 2015 00:17:18 +0000 (00:17 +0000)]
[AArch64] fix an invalid-iterator-use bug.
Summary:
In AArch64PromoteConstant::appendAndTransferDominatedUses,
`InsertPts[NewPt]` invalidates IPI. Therefore, `InsertPts[NewPt] =
std::move(IPI->second)` is not legal.
This was caught by running `make check` with
http://reviews.llvm.org/D7931.
Reviewers: t.p.northover, grosbach, bkramer
Reviewed By: bkramer
Subscribers: aemerson, llvm-commits
Differential Revision: http://reviews.llvm.org/D7988
llvm-svn: 230923
Benjamin Kramer [Sun, 1 Mar 2015 23:56:19 +0000 (23:56 +0000)]
Fix a really bad typo in my last commit.
llvm-svn: 230922
Sanjoy Das [Sun, 1 Mar 2015 23:36:26 +0000 (23:36 +0000)]
Revert some changes that were made to fix PR20680.
Summary:
As far as I can tell, the real bug causing the issue was fixed in
r230533. SCEVExpander should mark an increment operation as nuw or nsw
only if it can *prove* that the operation does not overflow. There
shouldn't be any situation where we have to do something different
because of no-wrap flags generated by SCEVExpander.
Revert "IndVarSimplify: Allow LFTR to fire more often"
This reverts commit
1ade0f0faa98877b688e0b9da58e876052c1e04e (SVN: 222213).
Revert "IndVarSimplify: Don't let LFTR compare against a poison value"
This reverts commit
c0f2b8b528d8a37b0a1522aae90af649d6357eb5 (SVN: 217102).
Reviewers: majnemer, atrick, spatel
Differential Revision: http://reviews.llvm.org/D7979
llvm-svn: 230921
Benjamin Kramer [Sun, 1 Mar 2015 23:35:20 +0000 (23:35 +0000)]
ArrayRef: Put back std::equal for operator== with a check for the empty ArrayRefs
This has the nice property of compiling down to memcmp when feasible. An empty
ArrayRef can have a nullptr in its Data field. I didn't find anything in the
standard speaking against std::equal(nullptr, nullptr, nullptr) begin valid but
MSVC asserts. The way libstdc++ lowers std::equal down to memcmp also makes
invoking std::equal with a nullptr undefined behavior so checking is the only
way to be safe.
The extra check doesn't cost us perf either because we're essentially peeling
the loop header away from the rotated loop.
llvm-svn: 230920
Vince Harron [Sun, 1 Mar 2015 23:21:29 +0000 (23:21 +0000)]
Linux - debugging of local processes via lldb-gdbserver
Instead of lldb calling the ptrace APIs directly, it launches an llgs
instance and communicates with it via the remote debug protocol.
This has two advantages.
There is more code shared between the remote debugging code path
and the local debugging code path. If a feature works in remote, it
probably works locally and vice-versa.
It makes us more architecturally similar to OSX (which also does
local debugging via a connection to debugserver).
This path is called LLGS local. We think that this configuration is
now at parity with (or better than) local linux debugging.
It is currently invoked if you have an environment variable defined
"PLATFORM_LINUX_FORCE_LLGS_LOCAL"
We would like to switch to LLGS local as the default path and only
use the non-LLGS path if someone has an environment variable defined
"PLATFORM_LINUX_DISABLE_LLGS_LOCAL"
Later, if all goes well, we would like to remove non-LLGS local
debugging support to simplify the codebase and avoid confusion.
llvm-svn: 230919
Benjamin Kramer [Sun, 1 Mar 2015 22:34:04 +0000 (22:34 +0000)]
Another missing include for MSVC.
llvm-svn: 230918
Peter Collingbourne [Sun, 1 Mar 2015 22:07:04 +0000 (22:07 +0000)]
DebugInfo: Give externally defined types a size and alignment where
possible. Fixes PR22736.
llvm-svn: 230914
Benjamin Kramer [Sun, 1 Mar 2015 21:49:21 +0000 (21:49 +0000)]
std::function is part of <functional>, not <utility>
llvm-svn: 230913
Benjamin Kramer [Sun, 1 Mar 2015 21:47:46 +0000 (21:47 +0000)]
Add another missing include for MSVC.
llvm-svn: 230912
Benjamin Kramer [Sun, 1 Mar 2015 21:47:40 +0000 (21:47 +0000)]
X86: Replace variadic function with init list. NFC.
llvm-svn: 230911
Benjamin Kramer [Sun, 1 Mar 2015 21:36:40 +0000 (21:36 +0000)]
Add missing include.
llvm-svn: 230910
Benjamin Kramer [Sun, 1 Mar 2015 21:28:53 +0000 (21:28 +0000)]
Add missing includes. make_unique proliferated everywhere.
llvm-svn: 230909
Arnaud A. de Grandmaison [Sun, 1 Mar 2015 21:22:50 +0000 (21:22 +0000)]
[PBQP] Address post-commit style comment for r230904. NFC.
Thanks David !
llvm-svn: 230908
Benjamin Kramer [Sun, 1 Mar 2015 21:05:05 +0000 (21:05 +0000)]
ArrayRef: Remove the equals helper with many arguments.
With initializer lists there is a really neat idiomatic way to write
this, 'ArrayRef.equals({1, 2, 3, 4, 5})'. Remove the equal method which
always had a hard limit on the number of arguments. I considered
rewriting it with variadic templates but that's not really a good fit
for a function with homogeneous arguments.
'ArrayRef == {1, 2, 3, 4, 5}' would've been even more awesome, but C++11
doesn't allow init lists with binary operators.
llvm-svn: 230907
Chaoren Lin [Sun, 1 Mar 2015 20:48:18 +0000 (20:48 +0000)]
Advertise 32 bit support for PlatformRemoteGDBServer on 64 bit systems.
Reviewers: vharron, clayborg, zturner
Subscribers: lldb-commits
Differential Revision: http://reviews.llvm.org/D7970
llvm-svn: 230906
Rui Ueyama [Sun, 1 Mar 2015 20:48:14 +0000 (20:48 +0000)]
Revert "PECOFF: Don't parse files in .drectve asynchronously."
This reverts commit r228955. Previously files appear in a .drectve
section are parsed synchronously to avoid threading issues. I believe
it's now safe to do that asynchronously.
llvm-svn: 230905
Arnaud A. de Grandmaison [Sun, 1 Mar 2015 20:39:34 +0000 (20:39 +0000)]
[PBQP] Do not add an edge between nodes with totally disjoint allowed registers
Such edges are zero matrix, and they bring no additional info to the
allocation problem, apart from contributing to nodes' degree. Removing
those edges is expected to improve allocation time.
Tune the spill cost comparison, as this gives better average performances
now that the nodes' degrees has changed.
llvm-svn: 230904
Daniel Jasper [Sun, 1 Mar 2015 18:55:26 +0000 (18:55 +0000)]
clang-format: Always align */& in multi-var DeclStmts.
Seems like the most consistent thing to do and in multi-var DeclStmts,
it is especially important to point out that the */& bind to the
identifier.
llvm-svn: 230903
Johannes Doerfert [Sun, 1 Mar 2015 18:51:51 +0000 (18:51 +0000)]
[Refactor] Include explicitly what is used
llvm-svn: 230902
Johannes Doerfert [Sun, 1 Mar 2015 18:45:58 +0000 (18:45 +0000)]
[Refactor] Include explicitly what is used
llvm-svn: 230901
Johannes Doerfert [Sun, 1 Mar 2015 18:44:57 +0000 (18:44 +0000)]
[Refactor] Remove Scop * member to simplify JSON class
llvm-svn: 230900
Johannes Doerfert [Sun, 1 Mar 2015 18:43:50 +0000 (18:43 +0000)]
[Refactor] Simplify ScopPass interface
llvm-svn: 230899
Johannes Doerfert [Sun, 1 Mar 2015 18:42:08 +0000 (18:42 +0000)]
[Refactor] Use virtual and override appropriately
+ Add override for overwritten methods.
+ Remove virtual for methods we do not want to be overwritten.
llvm-svn: 230898
Johannes Doerfert [Sun, 1 Mar 2015 18:40:25 +0000 (18:40 +0000)]
[Refactor] Add a Scop & as argument to printScop
This is the first step in the interface simplification.
llvm-svn: 230897
Johannes Doerfert [Sun, 1 Mar 2015 18:38:35 +0000 (18:38 +0000)]
[Cleanup] Remove unused passes
llvm-svn: 230896
Benjamin Kramer [Sun, 1 Mar 2015 18:10:07 +0000 (18:10 +0000)]
Make VTs and UnicodeCharSet ctors constexpr if the compiler supports it.
There are static variables of this around that we really want to go
into a read-only segment. Sadly compilers are not smart enough to figure
that out without constexpr.
llvm-svn: 230895
Benjamin Kramer [Sun, 1 Mar 2015 18:09:56 +0000 (18:09 +0000)]
Make some non-constant static variables non-static or fully const.
Otherwise we have to emit thread-safe initialization for them. NFC.
llvm-svn: 230894
Benjamin Kramer [Sun, 1 Mar 2015 18:09:50 +0000 (18:09 +0000)]
Sema: Remove stray static
This is a real bug if the code path is ever used with different pointer
sizes in the same process.
llvm-svn: 230893
Elena Demikhovsky [Sun, 1 Mar 2015 08:23:41 +0000 (08:23 +0000)]
Reverted 230471 - gather scatter handling in table gen.
llvm-svn: 230892
Elena Demikhovsky [Sun, 1 Mar 2015 07:44:04 +0000 (07:44 +0000)]
AVX-512: Added mask and rounding mode for scalar arithmetics
Added more tests for scalar instructions to destinguish between AVX and AVX-512 forms.
llvm-svn: 230891
Zachary Turner [Sun, 1 Mar 2015 06:59:57 +0000 (06:59 +0000)]
[llvm-pdbdump] Fix GCC build.
llvm-svn: 230890
Zachary Turner [Sun, 1 Mar 2015 06:51:29 +0000 (06:51 +0000)]
[llvm-pdbdump] Clean up method signatures.
llvm-svn: 230889
Zachary Turner [Sun, 1 Mar 2015 06:49:49 +0000 (06:49 +0000)]
[llvm-pdbdump] Add regex-based filtering.
llvm-svn: 230888
Chaoren Lin [Sun, 1 Mar 2015 04:31:16 +0000 (04:31 +0000)]
Fix attaching to 32 bit inferior with 64 bit llgs.
Summary: Executable module should use inferior architecture instead of host architecture.
Reviewers: ovyalov, vharron, clayborg
Subscribers: lldb-commits
Differential Revision: http://reviews.llvm.org/D7977
llvm-svn: 230887
Rui Ueyama [Sun, 1 Mar 2015 04:25:04 +0000 (04:25 +0000)]
Revert "PECOFF: Temporarily add a lock to un-break buildbot."
This reverts commit r230086. I added a lock to guard FileCOFF::doParse(),
which killed parallel file parsing. Now the buildbots got back to green,
I believe the threading issue was resolved, so it's time to remove the
guard to see if it works with the buildbots.
llvm-svn: 230886
NAKAMURA Takumi [Sun, 1 Mar 2015 04:16:28 +0000 (04:16 +0000)]
Revert r230655, "gold-plugin: "Upgrade" debug info and handle its warnings."
It emits *millions of warnings* during selfhosting LTO build, to choke the buildbot with gigbytes of log.
llvm-svn: 230885
Sanjay Patel [Sun, 1 Mar 2015 00:09:35 +0000 (00:09 +0000)]
avoid infinite looping when folding vector multiplies of constants (PR22698)
We were missing a check for the following fold in DAGCombiner:
// fold (fmul (fmul x, c1), c2) -> (fmul x, (fmul c1, c2))
If 'x' is also a constant, then we shouldn't do anything. Otherwise, we could end up swapping the operands back and forth forever.
This should fix:
http://llvm.org/bugs/show_bug.cgi?id=22698
Differential Revision: http://reviews.llvm.org/D7917
llvm-svn: 230884
Sanjay Patel [Sun, 1 Mar 2015 00:02:03 +0000 (00:02 +0000)]
fixed to test only the feature, not the feature and a CPU
llvm-svn: 230883
Duncan P. N. Exon Smith [Sat, 28 Feb 2015 23:57:16 +0000 (23:57 +0000)]
DebugInfo: Convert DW_OP_piece => DW_OP_bit_piece
r228631 stopped using `DW_OP_piece` inside `DIExpression`s in the IR,
but it apparently missed updating these testcases. Caught by verifier
checks for `MDExpression` while working on moving the new hierarchy into
place.
llvm-svn: 230882
Sanjay Patel [Sat, 28 Feb 2015 23:55:24 +0000 (23:55 +0000)]
make the tested feature (SSE2) explicit
llvm-svn: 230881
Duncan P. N. Exon Smith [Sat, 28 Feb 2015 23:52:24 +0000 (23:52 +0000)]
DebugInfo: Fix invalid file reference in CodeGen/X86/unknown-location.ll
There are two types of files in the old (current) debug info schema.
!0 = !{!"some/filename", !"/path/to/dir"}
!1 = !{!"0x29", !0} ; [ DW_TAG_file_type ]
!1 has a wrapper class called `DIFile` which inherits from `DIScope` and
is referenced in 'scope' fields.
!0 is called a "file node", and debug info nodes with a 'file' field
point at one of these directly -- although they're built in `DIBuilder`
by sending in a `DIFile` and reaching into it.
In the new hierarchy, I unified these nodes as `MDFile` (which `DIFile`
is a lightweight wrapper for) in r230057. Moving the new hierarchy into
place (and upgrading testcases) caused CodeGen/X86/unknown-location.ll
to start failing -- apparently "0x29" was previously showing up in the
linetable as a filename, causing:
.loc 2 4 3
(where 2 points at filename "0x29") instead of:
.loc 1 4 3
(where 1 points at the actual filename).
Change the testcase to use the old schema correctly.
llvm-svn: 230880
Duncan P. N. Exon Smith [Sat, 28 Feb 2015 23:48:02 +0000 (23:48 +0000)]
DebugInfo: Use TempMDNode in DIDescriptor::replaceAllUsesWith()
Start using `TempMDNode` in `DIDescriptor::replaceAllUsesWith()`
(effectively `std::unique_ptr<MDNode, MDNode::deleteTemporary>`).
Besides making ownership more explicit, this prepares for when
`DIDescriptor` refers to nodes that are *not* `MDTuple`. The old logic
for "replacing" a node with itself used `MDNode::get()` to return a new
(uniqued) `MDTuple`, while the new logic just defers to
`MDNode::replaceWithUniqued()` (which also typically saves an allocation
and RAUW traffic by mutating the temporary in place).
llvm-svn: 230879
Sanjay Patel [Sat, 28 Feb 2015 23:47:09 +0000 (23:47 +0000)]
fixed to test only the feature, not the feature and a CPU
llvm-svn: 230878
Duncan P. N. Exon Smith [Sat, 28 Feb 2015 23:21:38 +0000 (23:21 +0000)]
Optimize metadata node fields for CHECK-ability
While gaining practical experience hand-updating CHECK lines (for moving
the new debug info hierarchy into place), I learnt a few things about
CHECK-ability of the specialized node assembly output.
- The first part of a `CHECK:` is to identify the "right" node (this
is especially true if you intend to use the new `CHECK-SAME`
feature, since the first CHECK needs to identify the node correctly
before you can split the line).
- If there's a `tag:`, it should go first.
- If there's a `name:`, it should go next (followed by the
`linkageName:`, if any).
- If there's a `scope:`, it should follow after that.
- When a node type supports multiple DW_TAGs, but one is implied by
its name and is overwhelmingly more common, the `tag:` field is
terribly uninteresting unless it's different.
- `MDBasicType` is almost always `DW_TAG_base_type`.
- `MDTemplateValueParameter` is almost always
`DW_TAG_template_value_parameter`.
- Printing `name: ""` doesn't improve CHECK-ability, and there are far
more nodes than I realized that are commonly nameless.
- There are a few other fields that similarly aren't very interesting
when they're empty.
This commit updates the `AsmWriter` as suggested above (and makes
necessary changes in `LLParser` for round-tripping).
llvm-svn: 230877
Sanjay Patel [Sat, 28 Feb 2015 22:25:06 +0000 (22:25 +0000)]
fix typo; NFC
llvm-svn: 230876
Duncan P. N. Exon Smith [Sat, 28 Feb 2015 22:20:16 +0000 (22:20 +0000)]
AsmWriter: Escape string fields in metadata
Properly escape string fields in metadata. I've added a spot-check with
direct coverage for `MDFile::getFilename()`, but we'll get more coverage
once the hierarchy is moved into place (since this comes up in various
checked-in testcases).
I've replicated the `if` logic using the `ShouldSkipEmpty` flag
(although a follow-up commit is going to change how often this flag is
specified); no NFCI other than escaping the string fields.
llvm-svn: 230875
Hans Wennborg [Sat, 28 Feb 2015 22:18:11 +0000 (22:18 +0000)]
Revert r229678 "tsan: fix signal handling during stop-the-world"
This was causing Chromium's browser_tests to flakily segfault during
leak check.
llvm-svn: 230874
Duncan P. N. Exon Smith [Sat, 28 Feb 2015 22:16:56 +0000 (22:16 +0000)]
AsmWriter: Extract writeStringField(), NFCI
Extract logic for escaping a string field in the new debug info
hierarchy from `GenericDebugNode`. A follow-up commit will use it far
more widely (hence the dead code for `ShouldSkipEmpty`).
llvm-svn: 230873
Duncan P. N. Exon Smith [Sat, 28 Feb 2015 21:58:10 +0000 (21:58 +0000)]
Fix buildbot issues for MDScope::getFile() after r230871
I hope this extra cast will make everyone happy...
llvm-svn: 230872
Duncan P. N. Exon Smith [Sat, 28 Feb 2015 21:47:02 +0000 (21:47 +0000)]
IR: Specialize MDScope::getFile() for MDFile
Fix `MDScope::getFile()` so that it correctly returns a valid `MDFile`
even when it's an instance of `MDFile`. This logic is necessary because
of r230057. I'm working on moving the new hierarchy into place
out-of-tree (on track to commit Monday morning, BTW), and this was
exposed by a few failing tests.
llvm-svn: 230871
Duncan P. N. Exon Smith [Sat, 28 Feb 2015 21:38:32 +0000 (21:38 +0000)]
Fix line endings on Transforms/Inline/inline_dbg_declare.ll
llvm-svn: 230870
Saleem Abdulrasool [Sat, 28 Feb 2015 20:30:11 +0000 (20:30 +0000)]
build: remove the stub linux SDK
The SDK is insufficient to really build much. The builtins can be built
standalone now which is what the stub SDK was meant to permit. Remove the
unnecessary files.
llvm-svn: 230869
Zachary Turner [Sat, 28 Feb 2015 20:23:18 +0000 (20:23 +0000)]
[llvm-pdbdump] Better error handling.
Previously it was impossible to distinguish between "There is
no PDB implementation for this platform" and "I tried to load
the PDB, but couldn't find the file", making it hard to figure
out if you built llvm-pdbdump incorrectly or if you just mistyped
a file name.
This patch adds proper error handling so that we can know exactly
what went wrong.
llvm-svn: 230868
Saleem Abdulrasool [Sat, 28 Feb 2015 20:18:39 +0000 (20:18 +0000)]
cmath: account for MSVCRT 12.0 changes
MSVCRT 12.0 introduces better compatibility for C99. This includes a number of
math routines that were previously undefined. Use the crtversion.h header to
detect the version of MSVCRT being targeted and avoid re-declaring the
variables.
Since copysign has been introduced in MSVCRT, importing the definition via using
makes it difficult to provide overloads (due to minor differences between
throw () and noexcept. Avoid defining the overloads on newer MSVCRT
targets.
llvm-svn: 230867
Benjamin Kramer [Sat, 28 Feb 2015 20:15:07 +0000 (20:15 +0000)]
IndexedMap: Default to SmallVector<T, 0>
This looks ridiculous but SmallVector's realloc tricks really help with
large vectors of PODs, such as our virtreg IndexedMap.
llvm-svn: 230866
Benjamin Kramer [Sat, 28 Feb 2015 20:15:00 +0000 (20:15 +0000)]
DwarfAccelTable: We know how many hashes we have in the output, just reserve the precise number
llvm-svn: 230865
Benjamin Kramer [Sat, 28 Feb 2015 20:14:38 +0000 (20:14 +0000)]
StackColoring: Move set instead of copying. NFC.
llvm-svn: 230864
Benjamin Kramer [Sat, 28 Feb 2015 20:14:27 +0000 (20:14 +0000)]
LiveRange: Replace a creative vector erase loop with std::remove_if.
I didn't see this so far because it scans backwards, but that doesn't
make it any less quadratic. NFC.
llvm-svn: 230863
Nick Lewycky [Sat, 28 Feb 2015 19:43:20 +0000 (19:43 +0000)]
Revert r230812. Do not break builds for no reason.
llvm-svn: 230862
Mehdi Amini [Sat, 28 Feb 2015 19:34:54 +0000 (19:34 +0000)]
Fixup for recent -fast-isel-abort change: code didn't match description
Level 1 should abort for all instructions but call/terminators/args.
Instead it was aborting only if the level was > 2
From: Mehdi Amini <mehdi.amini@apple.com>
llvm-svn: 230861
Craig Topper [Sat, 28 Feb 2015 19:33:17 +0000 (19:33 +0000)]
[X86] Remove the blendpd/blendps/pblendw/pblendd intrinsics. They can represented by shuffle_vector instructions.
llvm-svn: 230860
Zachary Turner [Sat, 28 Feb 2015 19:08:27 +0000 (19:08 +0000)]
[raw_ostream] When printing color on Windows, use correct bg color.
When using SetConsoleTextAttribute() to set the foreground or
background color, if you don't explicitly set both colors, then
a default value of black will be chosen for whichever you don't
specify a value for.
This is annoying when you have a non default console background
color, for example, and you try to set the foreground color.
This patch gets the existing fg/bg color and when you set one
attribute, sets the opposite attribute to its existing color
prior to comitting the update.
Reviewed by: Aaron Ballman
Differential Revision: http://reviews.llvm.org/D7967
llvm-svn: 230859
Alexei Starovoitov [Sat, 28 Feb 2015 18:03:04 +0000 (18:03 +0000)]
bpf: fix build
complete the plumbing of passing TargetRegisterInfo through
computeRegisterProperties started by r230583
llvm-svn: 230858
Johannes Doerfert [Sat, 28 Feb 2015 17:10:06 +0000 (17:10 +0000)]
Update obsolete comment
llvm-svn: 230857
Benjamin Kramer [Sat, 28 Feb 2015 16:47:27 +0000 (16:47 +0000)]
TRE: Just erase dead BBs and tweak the iteration loop not to increment the deleted BB iterator.
Leaving empty blocks around just opens up a can of bugs like PR22704. Deleting
them early also slightly simplifies code.
Thanks to Sanjay for the IR test case.
llvm-svn: 230856
Yaron Keren [Sat, 28 Feb 2015 15:54:04 +0000 (15:54 +0000)]
Silence variable set but not used warning in CodeGenRegisters.cpp, NFC.
llvm-svn: 230854
Yaron Keren [Sat, 28 Feb 2015 15:29:17 +0000 (15:29 +0000)]
Silence three more variable set but not used warnings, NFC.
llvm-svn: 230853
Johannes Doerfert [Sat, 28 Feb 2015 14:48:15 +0000 (14:48 +0000)]
Format non-affine subregions news item
llvm-svn: 230852
Johannes Doerfert [Sat, 28 Feb 2015 14:46:26 +0000 (14:46 +0000)]
Add non-affine subregions to the news
llvm-svn: 230851
Peter Zotov [Sat, 28 Feb 2015 13:48:23 +0000 (13:48 +0000)]
[OCaml] Generate documentation again with autoconf buildsystem.
Patch by Evangelos Foutras:
r220899 started using ocamlfind to build the OCaml bindings but
docs/Makefile still contains references to the OCAMLDOC macro which
is no longer being defined. The result is that OCaml documentation
isn't generated/installed.
llvm-svn: 230850
Benjamin Kramer [Sat, 28 Feb 2015 13:20:15 +0000 (13:20 +0000)]
Convert push_back loops into append calls.
No functionality change intended.
llvm-svn: 230849
Yaron Keren [Sat, 28 Feb 2015 13:11:24 +0000 (13:11 +0000)]
Silence variable set but not used warning, NFC.
llvm-svn: 230848
Kuba Brecka [Sat, 28 Feb 2015 12:25:10 +0000 (12:25 +0000)]
Fix OS X build failure with Command Line Tools and without full Xcode installation
On OS X, if you don't have a full Xcode installation, but just the Command Line Tools package, xcrun and xcodebuild don't return a valid SDK root path. In these cases, let's use "/" as the SDK root (which is where the headers and libraries are installed).
Reviewed at http://reviews.llvm.org/D7641
llvm-svn: 230847
Benjamin Kramer [Sat, 28 Feb 2015 12:04:00 +0000 (12:04 +0000)]
ArrayRefize memory operand folding. NFC.
llvm-svn: 230846
Benjamin Kramer [Sat, 28 Feb 2015 10:11:12 +0000 (10:11 +0000)]
Replace std::copy with a back inserter with vector append where feasible
All of the cases were just appending from random access iterators to a
vector. Using insert/append can grow the vector to the perfect size
directly and moves the growing out of the loop. No intended functionalty
change.
llvm-svn: 230845
Eric Christopher [Sat, 28 Feb 2015 10:04:18 +0000 (10:04 +0000)]
Remove option.ll as part of the Forward Control Flow Integrity
removal.
llvm-svn: 230844
Richard Smith [Sat, 28 Feb 2015 09:58:41 +0000 (09:58 +0000)]
Add change accidentally missed from r230840.
llvm-svn: 230843
Kuba Brecka [Sat, 28 Feb 2015 09:39:05 +0000 (09:39 +0000)]
Symbolizer refactoring: SymbolizerTool and better interface
Reviewed at: http://reviews.llvm.org/D7936
llvm-svn: 230842
Filipe Cabecinhas [Sat, 28 Feb 2015 07:35:07 +0000 (07:35 +0000)]
Add PS4's autolink syntax to autolink.m's CHECK lines.
llvm-svn: 230841
Richard Smith [Sat, 28 Feb 2015 05:57:02 +0000 (05:57 +0000)]
[modules] Avoid adding a redecl chain to the 'pending out of date' list as the
very first step in updating it.
llvm-svn: 230840
Richard Smith [Sat, 28 Feb 2015 03:09:52 +0000 (03:09 +0000)]
Give better diagnostics when -fmodule-file= finds a bad file: if the file is
found indirectly, explain how we got there, and distinguish between 'file not
found' and 'file found but invalid'.
llvm-svn: 230839
Petar Jovanovic [Sat, 28 Feb 2015 02:32:55 +0000 (02:32 +0000)]
[MIPS] Add missing header for syscall
The syscall function itself is declared in <unistd.h>
Compiling compiler-rt with -Werror-implicit-function-declaration will
cause a build failure. This change fixes it.
Differential Revision: http://reviews.llvm.org/D7949
llvm-svn: 230838
Saleem Abdulrasool [Sat, 28 Feb 2015 02:23:54 +0000 (02:23 +0000)]
Unwind: make it build on Darwin again
The support for the LSB eh_frame_hdr extension was overzealous in trying to
include headers. Be more careful to permit building on Darwin.
llvm-svn: 230837
Philip Reames [Sat, 28 Feb 2015 01:57:44 +0000 (01:57 +0000)]
[RewriteStatepointsForGC] Reduce indentation via early continue [NFC]
llvm-svn: 230836
Philip Reames [Sat, 28 Feb 2015 01:52:09 +0000 (01:52 +0000)]
[RewriteStatepointsForGC] Fix another order of iteration bug
It turns out the naming of inserted phis and selects is sensative to the order in which two sets are iterated. We need to nail this down to avoid non-deterministic output and possible test failures.
The modified test is the one I first noticed something odd in. The change is making it more strict to report the error. With the test change, but without the code change, the test fails roughly 1 in 5. With the code change, I've run ~30 runs without error.
Long term, the right fix here is to adjust the naming scheme. I'm checking in this hack to avoid any possible non-determinism in the tests over the weekend. HJust because I only noticed one case doesn't mean it's actually the only case. I hope to get to the right change Monday.
std->llvm data structure changes bugfix change #3
llvm-svn: 230835
Richard Smith [Sat, 28 Feb 2015 01:45:19 +0000 (01:45 +0000)]
[modules] Deduplicate when merging lists of template specializations from imported modules.
llvm-svn: 230834