Nick Lewycky [Wed, 4 Jun 2014 07:45:54 +0000 (07:45 +0000)]
Fix a use of uninitialized value. OldCC is set when IsCmpZero || IsSwapped and read when ShouldUpdateCC || IsSwapped, and ShouldUpdateCC is independent. Fixes PR19932, but no test since I wasn't able to get any symptoms to appear, not even with valgrind and the testcase from the PR. It's clear what happened from inspection of the code.
llvm-svn: 210168
Andrew Trick [Wed, 4 Jun 2014 07:06:27 +0000 (07:06 +0000)]
Add a subtarget hook: enablePostMachineScheduler.
As requested by AArch64 subtargets.
Note that this will have no effect until the
AArch64 target actually enables the pass like this:
substitutePass(&PostRASchedulerID, &PostMachineSchedulerID);
As soon as armv7 switches over, PostMachineScheduler will become the
default postRA scheduler, so this won't be necessary any more.
Targets using the old postRA schedule would then do:
substitutePass(&PostMachineSchedulerID, &PostRASchedulerID);
llvm-svn: 210167
Andrew Trick [Wed, 4 Jun 2014 07:06:18 +0000 (07:06 +0000)]
Move GenericScheduler and PostGenericScheduler into a header.
These were not exposed previously because I didn't want out-of-tree
targets to be too dependent on their internals. They can be reused for
a very wide variety of processors with casual scheduling needs without
exposing the classes by instead using hooks defined in
MachineSchedPolicy (we can add more if needed). When targets are more
aggressively tuned or want to provide custom heuristics, they can
define their own MachineSchedStrategy. I tend to think this is better
once you start customizing heuristics because you can copy over only
what you need. I don't think that layering heuristics generally works
well.
However, Arch64 targets now want to reuse the Generic scheduling logic
but also provide extensions. I don't see much harm in exposing the
Generic scheduling classes with a major caveat: these scheduling
strategies may change in the future without validating performance on
less mainstream processors. If you want to be immune from changes,
just define your own MachineSchedStrategy.
llvm-svn: 210166
Justin Bogner [Wed, 4 Jun 2014 06:29:38 +0000 (06:29 +0000)]
docs: Remove documentation for legacy PGO options
Late last year r191835 removed a largely unmaintained legacy PGO
infrastructure, but some of the docs were missed. Since these docs are
for things that don't actually exist anymore, they should be removed.
llvm-svn: 210165
Simon Atanasyan [Wed, 4 Jun 2014 05:51:44 +0000 (05:51 +0000)]
[Object][ELF] Range-based loop simplification.
No functional changes.
llvm-svn: 210164
Todd Fiala [Wed, 4 Jun 2014 05:07:40 +0000 (05:07 +0000)]
Fix a couple minor typos in $qMemoryRegionInfo packet docs.
llvm-svn: 210163
Marshall Clow [Wed, 4 Jun 2014 04:49:49 +0000 (04:49 +0000)]
Add a note about debug mode being non-functional
llvm-svn: 210162
Alp Toker [Wed, 4 Jun 2014 04:11:12 +0000 (04:11 +0000)]
GraphWriter: try gv before xdg-open
Avoid changing behaviour for everyone who's used to the traditional ghostview
UI, especially since it knows how to stay in the foreground unlike xdg-open.
Amendment to r210147.
llvm-svn: 210148
Alp Toker [Wed, 4 Jun 2014 03:57:44 +0000 (03:57 +0000)]
GraphWriter: support the XDG open utility
This runs a suitable viewer on Unix desktop environments specified by
Freedesktop.org (GNOME, KDE, Linux distributions etc.)
llvm-svn: 210147
Alp Toker [Wed, 4 Jun 2014 03:34:26 +0000 (03:34 +0000)]
Remove inclusions of LLVM's private config.h
llvm-svn: 210146
Alp Toker [Wed, 4 Jun 2014 03:28:55 +0000 (03:28 +0000)]
Remove the last remaining llvm/Config/config.h includes
This corrects long-standing misuses of LLVM's internal config.h.
In most cases the public llvm-config.h header was intended and we can now
remove the old hacks thanks to LLVM r210144.
The config.h header is private, won't be installed and should no longer be
included by clang or other modules.
llvm-svn: 210145
Alp Toker [Wed, 4 Jun 2014 03:21:38 +0000 (03:21 +0000)]
config.h: fix layering and don't duplicate definitions
Also correct the llvm-config.h header guard so it doesn't depend on 'CONFIG_H'
which is commonly defined in external projects and caused trouble for
embedders.
In future llvm/Config/llvm-config.h will be installed, but not
the private llvm/Config/config.h header.
llvm-svn: 210144
David Blaikie [Wed, 4 Jun 2014 01:30:59 +0000 (01:30 +0000)]
DebugInfo: Partial revert r209984 due to more cases where abstract DbgVariables do not have associated DIEs.
Along with a test case to demonstrate that due to inlining order there
are cases where abstract variable DIEs are not constructed since the
abstract subprogram was built due to a previous inlining that optimized
away those variables. This produces incorrect debug info (the 'missing'
abstract variable causes the inlined instance of that variable to be
emitted with a full description (name, line, file) rather than
referencing the abstract origin), but this commit at least ensures that
it doesn't crash...
llvm-svn: 210143
Nick Kledzik [Wed, 4 Jun 2014 00:34:27 +0000 (00:34 +0000)]
[mach-o] Make anonymous atom out of section content before any symbol
In sections that are broken into atoms at symbols, if the first symbol in the
section is not at the start of the section, then make an anonymous atom for
the section content that is before the first symbol.
llvm-svn: 210142
Hans Wennborg [Wed, 4 Jun 2014 00:18:41 +0000 (00:18 +0000)]
Downgrade "definition of dllimport static field" error to warning for class templates (PR19902)
This allows us to compile the following kind of code, which occurs in MSVC
headers:
template <typename> struct S {
__declspec(dllimport) static int x;
};
template <typename T> int S<T>::x;
The definition works similarly to a dllimport inline function definition and
gets available_externally linkage.
Differential Revision: http://reviews.llvm.org/D3998
llvm-svn: 210141
Greg Fitzgerald [Wed, 4 Jun 2014 00:06:41 +0000 (00:06 +0000)]
Disable unstable test on ARM
llvm-svn: 210140
Nick Kledzik [Tue, 3 Jun 2014 23:36:50 +0000 (23:36 +0000)]
[mach-o] Refactor sections -> atoms to be table driven.
Previously each section kind had its own code to loop over the section and
parse it into atoms. This refactoring has two tables. The first maps sections
to ContentType. The second maps ContentType to information on how to find
the atom boundaries.
A few bugs in test cases were discovered as part of the refactoring.
No change in functionality intended.
llvm-svn: 210138
Richard Smith [Tue, 3 Jun 2014 23:27:44 +0000 (23:27 +0000)]
Add __builtin_operator_new and __builtin_operator_delete, which act like calls
to the normal non-placement ::operator new and ::operator delete, but allow
optimizations like new-expressions and delete-expressions do.
llvm-svn: 210137
Matt Arsenault [Tue, 3 Jun 2014 23:06:13 +0000 (23:06 +0000)]
Fix typos
llvm-svn: 210135
Pete Cooper [Tue, 3 Jun 2014 22:42:10 +0000 (22:42 +0000)]
Calculate dead instructions when a live interval is created.
This gets us closer to being able to remove LiveVariables entirely which is where dead instructions are currently tagged as such.
Reviewed by Jakob Olesen
llvm-svn: 210132
Richard Smith [Tue, 3 Jun 2014 21:58:55 +0000 (21:58 +0000)]
Update DR page for latest revision of CWG issues list.
llvm-svn: 210130
Jonathan Roelofs [Tue, 3 Jun 2014 21:50:11 +0000 (21:50 +0000)]
Add test case for inherited exceptions
Test case written by Dana Jansens.
llvm-svn: 210129
Eric Christopher [Tue, 3 Jun 2014 21:01:39 +0000 (21:01 +0000)]
Revert r209381 as it isn't a local variable. Add a testcase so that
we know next time this happens.
llvm-svn: 210127
Eric Christopher [Tue, 3 Jun 2014 21:01:35 +0000 (21:01 +0000)]
Fixup formatting in the pass.
llvm-svn: 210126
Rafael Espindola [Tue, 3 Jun 2014 20:25:26 +0000 (20:25 +0000)]
Update comments to match reality.
llvm-svn: 210125
Tobias Grosser [Tue, 3 Jun 2014 20:20:41 +0000 (20:20 +0000)]
Fix formatting
llvm-svn: 210124
Rafael Espindola [Tue, 3 Jun 2014 20:07:32 +0000 (20:07 +0000)]
Fix a small bug in the parsing of anonymous globals.
It was able to parse
hidden dllexport global i32 42
but not
dllexport global i32 42
llvm-svn: 210121
Rafael Espindola [Tue, 3 Jun 2014 20:00:20 +0000 (20:00 +0000)]
Simplify the parsing of unnamed globals. No functionality changed.
llvm-svn: 210120
Rafael Espindola [Tue, 3 Jun 2014 19:38:37 +0000 (19:38 +0000)]
Update comment. This should have been part of r210062.
llvm-svn: 210119
Alp Toker [Tue, 3 Jun 2014 19:06:49 +0000 (19:06 +0000)]
Remove some redundant doc comments
llvm-svn: 210118
Sebastian Pop [Tue, 3 Jun 2014 18:16:31 +0000 (18:16 +0000)]
record delinearization result and reuse it in polyhedral translation
Without this patch, the testcase would fail on the delinearization of the second
array:
; void foo(long n, long m, long o, double A[n][m][o]) {
; for (long i = 0; i < n; i++)
; for (long j = 0; j < m; j++)
; for (long k = 0; k < o; k++) {
; A[i+3][j-4][k+7] = 1.0;
; A[i][0][k] = 2.0;
; }
; }
; CHECK: [n, m, o] -> { Stmt_for_body6[i0, i1, i2] -> MemRef_A[3 + i0, -4 + i1, 7 + i2] };
; CHECK: [n, m, o] -> { Stmt_for_body6[i0, i1, i2] -> MemRef_A[i0, 0, i2] };
Here is the output of FileCheck on the testcase without this patch:
; CHECK: [n, m, o] -> { Stmt_for_body6[i0, i1, i2] -> MemRef_A[i0, 0, i2] };
^
<stdin>:26:2: note: possible intended match here
[n, m, o] -> { Stmt_for_body6[i0, i1, i2] -> MemRef_A[o0] };
^
It is possible to find a good delinearization for A[i][0][k] only in the context
of the delinearization of both array accesses.
There are two ways to delinearize together all array subscripts touching the
same base address: either duplicate the code from scop detection to first gather
all array references and then run the delinearization; or as implemented in this
patch, use the same delinearization info that we computed during scop detection.
llvm-svn: 210117
Todd Fiala [Tue, 3 Jun 2014 18:09:56 +0000 (18:09 +0000)]
Added gdb-remote memory read ($m) test.
Added set-memory:{content} and get-memory-address-hex: commands
to the test exe for gdb-remote. Added a test that sets the content
via the inferior command line, then reads it back via gdb-remote
with $m.
Passing on debugserver. Marked as fail on llgs. Implementing
in the llgs branch next.
llvm-svn: 210116
Alp Toker [Tue, 3 Jun 2014 17:23:34 +0000 (17:23 +0000)]
Fix leak from r210059
Also revert r210096 which temporarily disabled the test while this was being
investigated.
llvm-svn: 210115
Tilmann Scheller [Tue, 3 Jun 2014 16:33:13 +0000 (16:33 +0000)]
[AArch64] Fix typo in load/store optimizer.
llvm-svn: 210114
Tilmann Scheller [Tue, 3 Jun 2014 16:03:00 +0000 (16:03 +0000)]
[AArch64] Add regression tests for the load/store optimizer which cover post-index update folding with sub rather than add.
The tests check that the following transform happens:
(ldr|str) X, [x20]
...
sub x20, x20, #16
->
(ldr|str) X, [x20], #-16
with X being either w0, x0, s0, d0 or q0.
llvm-svn: 210113
Jeroen Ketema [Tue, 3 Jun 2014 15:43:57 +0000 (15:43 +0000)]
Fix build broken by LLVM commit r209103
Reviewed-by: Aaron Watry <awatry@gmail.com>
llvm-svn: 210111
Rafael Espindola [Tue, 3 Jun 2014 15:36:31 +0000 (15:36 +0000)]
Remove dead code.
New code should probably use shared_ptr anyway.
llvm-svn: 210110
Evgeniy Stepanov [Tue, 3 Jun 2014 15:27:15 +0000 (15:27 +0000)]
[sancov] Delay opening dump file until the first module constructor.
llvm-svn: 210109
Evgeniy Stepanov [Tue, 3 Jun 2014 15:25:43 +0000 (15:25 +0000)]
[sancov] Handle spaces in module name.
llvm-svn: 210108
Rafael Espindola [Tue, 3 Jun 2014 15:06:22 +0000 (15:06 +0000)]
Implement one operator== with another.
Thanks for David Blaikie for the suggestion.
llvm-svn: 210107
Kostya Serebryany [Tue, 3 Jun 2014 15:06:13 +0000 (15:06 +0000)]
[asan] enable detect_odr_violation=2; https://code.google.com/p/address-sanitizer/wiki/OneDefinitionRuleViolation . This feature has been working quite well so far, found a few bugs and zero false positives. The main problem is that there could be lots of true positives and users will have to disable this checking until the bugs are fixed.
llvm-svn: 210106
Colin Riley [Tue, 3 Jun 2014 14:37:35 +0000 (14:37 +0000)]
Windows fix: Disable editline for MSVC. Since r208369 there have been issues, probably related to the editline wrapper. For now, it's more stable and usable disabled.
llvm-svn: 210105
Colin Riley [Tue, 3 Jun 2014 14:33:41 +0000 (14:33 +0000)]
Windows fix: Condition::Wait returned failure when it actually succeeded (SleepConditionVariableCS returns non-zero for success)
llvm-svn: 210104
Evgeniy Stepanov [Tue, 3 Jun 2014 14:16:00 +0000 (14:16 +0000)]
[asan] Fix coverage instrumentation with -asan-globals=0.
llvm-svn: 210103
Tim Northover [Tue, 3 Jun 2014 13:54:53 +0000 (13:54 +0000)]
AArch64: mark small types (i1, i8, i16) as promoted
This means the output of LowerFormalArguments returns a lowered
SDValue with the correct type (expected in SelectionDAGBuilder).
Without this, an assertion under a DEBUG macro triggers when those
types are passed on the stack.
llvm-svn: 210102
Rafael Espindola [Tue, 3 Jun 2014 13:27:48 +0000 (13:27 +0000)]
Use IntrusiveRefCntPtr's == and != nullptr operators.
This is a partial revert of r210075.
llvm-svn: 210101
Rafael Espindola [Tue, 3 Jun 2014 13:26:18 +0000 (13:26 +0000)]
Add operator== and operator!= to compare with nullptr.
llvm-svn: 210100
Colin Riley [Tue, 3 Jun 2014 13:01:18 +0000 (13:01 +0000)]
Fix windows build: SBTypeEnumMember.cpp added to the separate windows liblldb.dll project.
llvm-svn: 210099
Evgeniy Stepanov [Tue, 3 Jun 2014 12:15:43 +0000 (12:15 +0000)]
[sancov] Fix map update logic on Android.
dlopen()/dlclose() are not interceptable on Android,
so we update .sancov.map in module constructor callbacks.
llvm-svn: 210098
Daniel Jasper [Tue, 3 Jun 2014 12:02:45 +0000 (12:02 +0000)]
clang-format: Refactor indentation behavior for multiple nested blocks.
This fixes a few oddities when formatting multiple nested JavaScript
blocks, e.g.:
Before:
promise.then(
function success() {
doFoo();
doBar();
},
[], function error() {
doFoo();
doBaz();
});
promise.then([],
function success() {
doFoo();
doBar();
},
function error() {
doFoo();
doBaz();
});
After:
promise.then(
function success() {
doFoo();
doBar();
},
[],
function error() {
doFoo();
doBaz();
});
promise.then([],
function success() {
doFoo();
doBar();
},
function error() {
doFoo();
doBaz();
});
llvm-svn: 210097
Kostya Serebryany [Tue, 3 Jun 2014 11:45:37 +0000 (11:45 +0000)]
temporary disable part of a test because it causes clang to leak memory (want to have the sanitizer bot green)
llvm-svn: 210096
Alexander Musman [Tue, 3 Jun 2014 10:16:47 +0000 (10:16 +0000)]
[OPENMP] Loop canonical form analysis (Sema)
This patch implements semantic analysis to make sure that the loop is in OpenMP canonical form.
This is the form required for 'omp simd', 'omp for' and other loop pragmas.
Differential revision: http://reviews.llvm.org/D3778
llvm-svn: 210095
Simon Atanasyan [Tue, 3 Jun 2014 09:59:42 +0000 (09:59 +0000)]
[Mips] Make la25-stub.test self contained.
No functional changes.
llvm-svn: 210094
Rui Ueyama [Tue, 3 Jun 2014 08:40:49 +0000 (08:40 +0000)]
Improve error message.
Previously the parser always printed out an error message followed
by "Invalid file type" even if the file type is correct.
llvm-svn: 210093
Richard Smith [Tue, 3 Jun 2014 08:40:27 +0000 (08:40 +0000)]
Remove incorrect assertion.
llvm-svn: 210092
Richard Smith [Tue, 3 Jun 2014 08:26:00 +0000 (08:26 +0000)]
Implement DR990 and DR1070. Aggregate initialization initializes uninitialized
elements from {}, rather than value-initializing them. This permits calling an
initializer-list constructor or constructing a std::initializer_list object.
(It would also permit initializing a const reference or rvalue reference if
that weren't explicitly prohibited by other rules.)
llvm-svn: 210091
Richard Smith [Tue, 3 Jun 2014 08:24:28 +0000 (08:24 +0000)]
Teach AST dumper to dump the array filler in an initializer list.
llvm-svn: 210090
Rui Ueyama [Tue, 3 Jun 2014 08:12:33 +0000 (08:12 +0000)]
Rename FileToMutable -> SimpleFileWrapper.
FileToMutable is what this class does, but this class (or, to be precise,
an instance of this class) is a wrapper of the other SimpleFile. It's odd
that the class was named like a function.
llvm-svn: 210089
Rui Ueyama [Tue, 3 Jun 2014 07:53:37 +0000 (07:53 +0000)]
Fix build breakage.
llvm-svn: 210087
Rui Ueyama [Tue, 3 Jun 2014 07:52:42 +0000 (07:52 +0000)]
Inline short member function.
llvm-svn: 210086
Rui Ueyama [Tue, 3 Jun 2014 07:43:15 +0000 (07:43 +0000)]
Use auto for obvious types.
llvm-svn: 210085
Rui Ueyama [Tue, 3 Jun 2014 07:39:32 +0000 (07:39 +0000)]
No need to compute valBit until mask bit is 1.
llvm-svn: 210084
Richard Smith [Tue, 3 Jun 2014 07:28:54 +0000 (07:28 +0000)]
PR11410: Extend diagnostic to cover all cases of aggregate initialization, not
just the extremely specific case of a trailing array element that couldn't be
initialized because the default constructor for the element type is deleted.
Also reword the diagnostic to better match our other context diagnostics and to
prepare for the implementation of core issue 1070.
llvm-svn: 210083
Rui Ueyama [Tue, 3 Jun 2014 07:27:49 +0000 (07:27 +0000)]
Use range-based for loop.
llvm-svn: 210082
Rui Ueyama [Tue, 3 Jun 2014 07:24:46 +0000 (07:24 +0000)]
Remove unusual use of using.
llvm-svn: 210081
Nikola Smiljanic [Tue, 3 Jun 2014 07:06:20 +0000 (07:06 +0000)]
List DR532 as implemented on the defect report status page by renaming the namespace.
llvm-svn: 210080
Richard Smith [Tue, 3 Jun 2014 06:58:52 +0000 (06:58 +0000)]
When emitting a multidimensional array new, emit the initializers for the
trailing elements as a single loop, rather than sometimes emitting a nest of
several loops. This fixes a bug where CodeGen would sometimes try to emit an
expression with the wrong type for the element being initialized. Plus various
other minor cleanups to the IR produced for array new initialization.
llvm-svn: 210079
Rafael Espindola [Tue, 3 Jun 2014 05:26:12 +0000 (05:26 +0000)]
Use an enum class.
llvm-svn: 210078
Rafael Espindola [Tue, 3 Jun 2014 05:12:33 +0000 (05:12 +0000)]
Use an enum class.
Might also fix the windows build.
llvm-svn: 210077
Rafael Espindola [Tue, 3 Jun 2014 05:05:15 +0000 (05:05 +0000)]
Remove the last unspecified_bool_type from llvm.
llvm-svn: 210076
Rafael Espindola [Tue, 3 Jun 2014 05:04:52 +0000 (05:04 +0000)]
Don't assume an implicit IntrusiveRefCntPtr -> bool operator.
llvm-svn: 210075
Rui Ueyama [Tue, 3 Jun 2014 05:04:07 +0000 (05:04 +0000)]
Reference::target() doesn't and shouldn't return a nullptr.
llvm-svn: 210074
Rafael Espindola [Tue, 3 Jun 2014 04:54:15 +0000 (04:54 +0000)]
Make this operator bool() explicit to match the standard library.
llvm-svn: 210073
Rafael Espindola [Tue, 3 Jun 2014 04:42:24 +0000 (04:42 +0000)]
Make this operator bool() explicit to match the standard library.
llvm-svn: 210072
Rafael Espindola [Tue, 3 Jun 2014 04:41:30 +0000 (04:41 +0000)]
This lambda returns a boolean, update the return statements.
llvm-svn: 210071
Rafael Espindola [Tue, 3 Jun 2014 04:40:55 +0000 (04:40 +0000)]
Don't assume an implicit error_code -> bool operator.
llvm-svn: 210070
Greg Fitzgerald [Tue, 3 Jun 2014 04:29:46 +0000 (04:29 +0000)]
Fix Android build after r210053
llvm-svn: 210069
Nick Lewycky [Tue, 3 Jun 2014 04:25:36 +0000 (04:25 +0000)]
Ignore line numbers on debug intrinsics. Add an assert to ensure that we aren't emitting line number zero, the .gcno format uses this to indicate that the next field is a filename.
llvm-svn: 210068
Jiangning Liu [Tue, 3 Jun 2014 03:25:09 +0000 (03:25 +0000)]
[AArch64] Correctly deal with VPR stack parameter passing.
llvm-svn: 210067
Rui Ueyama [Tue, 3 Jun 2014 03:07:49 +0000 (03:07 +0000)]
Remove group-parent references.
Previously section groups are doubly linked to their children.
That is, an atom representing a group has group-child references
to its group contents, and content atoms also have group-parent
references to the group atom. That relationship was invariant;
if X has a group-child edge to Y, Y must have a group-parent
edge to X.
However we were not using group-parent references at all. The
resolver only needs group-child edges.
This patch simplifies the section group by removing the unused
reverse edge. No functionality change intended.
Differential Revision: http://reviews.llvm.org/D3945
llvm-svn: 210066
Alp Toker [Tue, 3 Jun 2014 03:01:03 +0000 (03:01 +0000)]
Process::GetRandomNumber(): fix insecure RNG
This could have generated non-random output under error conditions in release
builds.
llvm-svn: 210065
Nikola Smiljanic [Tue, 3 Jun 2014 02:56:59 +0000 (02:56 +0000)]
Move DR532 test where it belongs.
llvm-svn: 210064
Rafael Espindola [Tue, 3 Jun 2014 02:42:01 +0000 (02:42 +0000)]
Update for llvm API change.
Aliases in llvm now hold an arbitrary expression.
llvm-svn: 210063
Rafael Espindola [Tue, 3 Jun 2014 02:41:57 +0000 (02:41 +0000)]
Allow alias to point to an arbitrary ConstantExpr.
This patch changes GlobalAlias to point to an arbitrary ConstantExpr and it is
up to MC (or the system assembler) to decide if that expression is valid or not.
This reduces our ability to diagnose invalid uses and how early we can spot
them, but it also lets us do things like
@test5 = alias inttoptr(i32 sub (i32 ptrtoint (i32* @test2 to i32),
i32 ptrtoint (i32* @bar to i32)) to i32*)
An important implication of this patch is that the notion of aliased global
doesn't exist any more. The alias has to encode the information needed to
access it in its metadata (linkage, visibility, type, etc).
Another consequence to notice is that getSection has to return a "const char *".
It could return a NullTerminatedStringRef if there was such a thing, but when
that was proposed the decision was to just uses "const char*" for that.
llvm-svn: 210062
David Majnemer [Tue, 3 Jun 2014 02:40:39 +0000 (02:40 +0000)]
[libc++] Don't return uninitialized data from random_device::operator()
Make sure we appropriately retry calls to read if the return result is
less than what we asked for.
Additionally, check and handle IO errors: EINTR results in the read
operation getting restarted; other errors turn into exceptions.
llvm-svn: 210061
David Majnemer [Tue, 3 Jun 2014 02:21:37 +0000 (02:21 +0000)]
[libc++] random_device fails if open returns zero
random_device::random_device(const string&) wrongly assumes that open
can only validly return a file descriptor greater than zero.
This results in random_device believing that it didn't successfully open
the device causing it to throw in it's constructor, this ends up leaking
a file descriptor.
The fix is simple, don't error on file descriptors which are zero.
llvm-svn: 210060
Alp Toker [Tue, 3 Jun 2014 02:14:20 +0000 (02:14 +0000)]
Fix -emit-codegen-only to not generate binaries
llvm-svn: 210059
Alp Toker [Tue, 3 Jun 2014 02:13:57 +0000 (02:13 +0000)]
Eliminate redundant MangleBuffer class
The only remaining user didn't actually use the non-dynamic storage facility
this class provides.
The std::string is transitional and likely to be StringRefized shortly.
llvm-svn: 210058
Rui Ueyama [Tue, 3 Jun 2014 01:59:02 +0000 (01:59 +0000)]
Make dead-striping to handle reverse edges.
Layout-before edges are no longer used for layout, but they are
still there for dead-stripping. If we would just remove them
from code, LLD would wrongly remove live atoms that were
referenced by layout-befores.
This patch fixes the issue. Before dead-stripping, it scans all
atoms to construct a reverse map for layout-after edges. Dead-
stripping pass uses the map to traverse the graph.
Differential Revision: http://reviews.llvm.org/D3986
llvm-svn: 210057
Marshall Clow [Mon, 2 Jun 2014 23:37:13 +0000 (23:37 +0000)]
First cut at a post c++14 status page
llvm-svn: 210056
Hans Wennborg [Mon, 2 Jun 2014 23:13:03 +0000 (23:13 +0000)]
Itanium ABI: Update getAddrOfVTable to set the DLL storage class for vtables
This corresponds to the same change for the MS ABI in r209908.
Differential Revision: http://reviews.llvm.org/D3993
llvm-svn: 210054
Greg Fitzgerald [Mon, 2 Jun 2014 23:11:24 +0000 (23:11 +0000)]
Cleanup Android build and provide better diagnostics
No longer need to set ANDROID if COMPILER_RT_TEST_TARGET_TRIPLE is
arm-linux-androideabi.
No need to set ANDROID_COMMON_FLAGS. These flags are already in
CMAKE_CXX_FLAGS which are used in try_compile().
llvm-svn: 210053
Rafael Espindola [Mon, 2 Jun 2014 22:01:04 +0000 (22:01 +0000)]
Add back commit r210029.
The code was actually correct. Sorry for the confusion. I have expanded the
comment saying why the analysis is valid to avoid me misunderstaning it
again in the future.
llvm-svn: 210052
Greg Clayton [Mon, 2 Jun 2014 21:58:30 +0000 (21:58 +0000)]
Small cleanups for the new enum fixes:
- Fix Xcode project to have source files for SBTypeEnumMember.h/SBTypeEnumMember.cpp in the right place
- Rename a member variable to inluce "_sp" suffix since it is a shared pointer
- Cleanup initialization code for TypeEnumMemberImpl to not warn about out of order initialization
llvm-svn: 210051
Will Schmidt [Mon, 2 Jun 2014 21:47:14 +0000 (21:47 +0000)]
allow optional signext attribute
Allow the tests to succeed with tne signext (or other) attribute is present. The attributes
show up for Power, but not for x86*, so need to be appropriately wildcarded.
llvm-svn: 210050
Rafael Espindola [Mon, 2 Jun 2014 21:23:54 +0000 (21:23 +0000)]
Convert test to FileCheck.
llvm-svn: 210049
Rafael Espindola [Mon, 2 Jun 2014 21:12:19 +0000 (21:12 +0000)]
Revert "Add the nsw flag when we detect that an add will not signed overflow."
This reverts commit r210029.
It was not correctly handling cases where LHS and RHS had multiple but different
sign bits.
llvm-svn: 210048
Alexey Samsonov [Mon, 2 Jun 2014 21:05:54 +0000 (21:05 +0000)]
Delete apparently unused method
llvm-svn: 210047
Todd Fiala [Mon, 2 Jun 2014 20:55:29 +0000 (20:55 +0000)]
Add support for inspecting enum members.
Change by Russell Harmon.
Xcode project updates (and all errors therein)
by Todd Fiala.
llvm-svn: 210046
Eric Christopher [Mon, 2 Jun 2014 20:51:49 +0000 (20:51 +0000)]
InitLibcallNames can take a Triple instead of a TargetMachine.
llvm-svn: 210045