Sid Manning [Tue, 9 Oct 2012 02:20:47 +0000 (02:20 +0000)]
* Add relocation reference support for Hexagon and PowerPC.
- Hexagon has many relocation types, this patch covers just
a subset, more will be added.
- PowerPC support was added so that we can verify mixed endian
environments.
* Add some basic test cases
* Update ldd-core to accept -endian flag
llvm-svn: 165477
Chandler Carruth [Tue, 9 Oct 2012 01:58:35 +0000 (01:58 +0000)]
Fix PR14034, an infloop / heap corruption / crash bug in the new SROA.
Thanks to Benjamin for the raw test case. This one took about 50 times
longer to reduce than to fix. =/
llvm-svn: 165476
Nadav Rotem [Tue, 9 Oct 2012 01:56:07 +0000 (01:56 +0000)]
whitespace
llvm-svn: 165475
Sean Silva [Tue, 9 Oct 2012 01:47:16 +0000 (01:47 +0000)]
Move misplaced comment.
llvm-svn: 165474
Argyrios Kyrtzidis [Tue, 9 Oct 2012 01:23:50 +0000 (01:23 +0000)]
Move the logic that searches for overridden methods from libclang to
ASTContext so that it can be widely available.
llvm-svn: 165473
Argyrios Kyrtzidis [Tue, 9 Oct 2012 01:23:45 +0000 (01:23 +0000)]
When storing the C++ overridden methods, store them once for the
canonical method; avoid storing them again for an out-of-line definition.
llvm-svn: 165472
Jason Molenda [Tue, 9 Oct 2012 01:17:11 +0000 (01:17 +0000)]
Add a parameter to Symbols::DownloadObjectAndSymbolFile() to control
whether we try to call an external program to load symbols unconditionally,
or if we check the user's preferences before calling it.
ProcessMachCore now sets CanJIT to false - we can't execute code in a core file.
DynamicLoaderDarwinKernel::OSKextLoadedKextSummary::LoadImageUsingMemoryModule changed
to load the kernel from an on-disk file if at all possible.
Don't load the kext binaries out of memory from the remote systems - their linkedit doesn't
seem to be in a good state and we'll error out down in SymbolVendorMacOSX if we try to use
the in-memory images.
Call Symbols::DownloadObjectAndSymbolFile to get the kext/kernel binary -- the external
program may be able to give us a file path on the local filesystem instead of reading
the binary / dSYM over a network drive every time. Fall back to calling
Target::GetSharedModule() like before if DownloadObjectAndSymbolFile fails.
llvm-svn: 165471
Bill Wendling [Tue, 9 Oct 2012 01:03:48 +0000 (01:03 +0000)]
Remove a couple more 'hasAttrSomewhere' calls.
llvm-svn: 165470
Bill Wendling [Tue, 9 Oct 2012 00:51:40 +0000 (00:51 +0000)]
Fix. Apply the no capture attribute to the correct parameter.
llvm-svn: 165469
Bill Wendling [Tue, 9 Oct 2012 00:47:36 +0000 (00:47 +0000)]
Convert to using the Attributes::Builder class to create attributes.
llvm-svn: 165468
Bill Wendling [Tue, 9 Oct 2012 00:34:19 +0000 (00:34 +0000)]
Use an iterator and proper query method instead of the 'hasAttrSomewhere' method.
llvm-svn: 165467
Bill Wendling [Tue, 9 Oct 2012 00:28:54 +0000 (00:28 +0000)]
Remove more uses of the attribute enums by supplying appropriate query methods for them.
No functionality change intended.
llvm-svn: 165466
Bill Wendling [Tue, 9 Oct 2012 00:01:21 +0000 (00:01 +0000)]
Convert to using the Attributes::Builder interface.
llvm-svn: 165465
Eric Christopher [Mon, 8 Oct 2012 23:54:10 +0000 (23:54 +0000)]
Fix typo in docs.
llvm-svn: 165464
Eric Christopher [Mon, 8 Oct 2012 23:53:45 +0000 (23:53 +0000)]
Fix up comment to be more clear.
llvm-svn: 165463
Eli Friedman [Mon, 8 Oct 2012 23:52:38 +0000 (23:52 +0000)]
Make sure we allow "#pragma options align=mac68k" in function-local contexts. <rdar://problem/
12453134>
llvm-svn: 165462
Bill Wendling [Mon, 8 Oct 2012 23:51:19 +0000 (23:51 +0000)]
Update comment.
llvm-svn: 165461
Greg Clayton [Mon, 8 Oct 2012 23:50:19 +0000 (23:50 +0000)]
Fix a build warning and a dangerous possible crasher.
llvm-svn: 165460
Ted Kremenek [Mon, 8 Oct 2012 23:45:09 +0000 (23:45 +0000)]
Add regression test to check if -Wenum-conversion is a subgroup of -Wconversion.
llvm-svn: 165459
Bill Wendling [Mon, 8 Oct 2012 23:27:46 +0000 (23:27 +0000)]
Use the Attributes::Builder to build the attributes in the parser.
llvm-svn: 165458
Nadav Rotem [Mon, 8 Oct 2012 23:14:28 +0000 (23:14 +0000)]
Use DataTypes.h
llvm-svn: 165457
Argyrios Kyrtzidis [Mon, 8 Oct 2012 23:08:41 +0000 (23:08 +0000)]
In VarDecl::getSourceRange() make sure to check that the source location
of the initializer is valid before using it.
Fixes rdar://
12455002&
12449015 where local variables of objc objects in ARC mode
were not annotated because of the ImplicitValueInitExpr initializer having invalid
source range, resulting in the SourceRange of the VarDecl having invalid end location.
llvm-svn: 165456
Nadav Rotem [Mon, 8 Oct 2012 23:06:34 +0000 (23:06 +0000)]
Refactor the AddrMode class out of TLI to its own header file.
This class is used by LSR and a number of places in the codegen.
This is the first step in de-coupling LSR from TLI, and creating
a new interface in between them.
llvm-svn: 165455
Greg Clayton [Mon, 8 Oct 2012 22:48:57 +0000 (22:48 +0000)]
Make DebugMapModule destructor virtual since it inherits from Module.
llvm-svn: 165441
Chandler Carruth [Mon, 8 Oct 2012 22:45:55 +0000 (22:45 +0000)]
Actually add the -Wenum-conversion group to -Wconversion as r165361
seems to have intended.
llvm-svn: 165440
Chandler Carruth [Mon, 8 Oct 2012 22:45:54 +0000 (22:45 +0000)]
Sort the subgroups of the diagnostic group for -Wconversion. No
functionality changed.
llvm-svn: 165439
Greg Clayton [Mon, 8 Oct 2012 22:41:53 +0000 (22:41 +0000)]
Added a new "module" log channel which covers module creation, deletion, and common module list actions.
Also added a new option for "log enable" which is "--stack" which will print out a stack backtrace for each log line.
This was used to track down the leaking module issue I fixed last week.
llvm-svn: 165438
Greg Clayton [Mon, 8 Oct 2012 22:39:38 +0000 (22:39 +0000)]
Checking in fixes that I used to track down a leaking module. The heap module can now search the vm regions with the --vm-regions options to any of the heap functions. This is currently slow and often will time out when run on a large program since our user expression timeout is set to 500000 usec. We need to add an API to LLDB where we can specify the timeout for an expression.
llvm-svn: 165437
Bill Wendling [Mon, 8 Oct 2012 22:20:14 +0000 (22:20 +0000)]
Convert the LLVM parser over to using the new Attributes::Builder to build its
attributes objects.
llvm-svn: 165436
Nick Lewycky [Mon, 8 Oct 2012 22:12:48 +0000 (22:12 +0000)]
Give CaptureTracker::shouldExplore a base implementation. Most users want to do
the same thing. No functionality change.
llvm-svn: 165435
Jakob Stoklund Olesen [Mon, 8 Oct 2012 22:06:44 +0000 (22:06 +0000)]
Don't crash on extra evil irreducible control flow.
When the CFG contains a loop with multiple entry blocks, the traces
computed by MachineTraceMetrics don't always have the same nice
properties. Loop back-edges are normally excluded from traces, but
MachineLoopInfo doesn't recognize loops with multiple entry blocks, so
those back-edges may be included.
Avoid asserting when that happens by adding an isEarlierInSameTrace()
function that accurately determines if a dominating block is part of the
same trace AND is above the currrent block in the trace.
llvm-svn: 165434
Bill Wendling [Mon, 8 Oct 2012 21:47:17 +0000 (21:47 +0000)]
Begin the transition to using the AttributesImpl object for the Attributes ivar.
Start using the AttributesImpl object to hold the value of the attributes. All
queries go through the interfaces now.
This has one unfortunate consequence. I needed to move the AttributesImpl.h file
into include/llvm. But this is only temporary! Otherwise, the changes needed to
support this would be too large.
llvm-svn: 165433
Eric Christopher [Mon, 8 Oct 2012 21:41:30 +0000 (21:41 +0000)]
Add names for the accelerator table sections so that they can
be emitted if they're wanted on elf platforms.
llvm-svn: 165432
Eric Christopher [Mon, 8 Oct 2012 21:41:18 +0000 (21:41 +0000)]
Update comment.
llvm-svn: 165431
Chandler Carruth [Mon, 8 Oct 2012 21:31:38 +0000 (21:31 +0000)]
Make Bitrig's clang understand -stdlib= correctly.
With this patch Bitrig can use a different c++ library without pain and
within the normal commandline parameters.
Original patch by David Hill, with lots of fixes and cleanup by me.
llvm-svn: 165430
Bill Wendling [Mon, 8 Oct 2012 21:17:59 +0000 (21:17 +0000)]
Use a special path to place the .o files in.
llvm-svn: 165429
Bill Wendling [Mon, 8 Oct 2012 21:17:45 +0000 (21:17 +0000)]
Use a special path to place the .o files in.
llvm-svn: 165428
Eric Christopher [Mon, 8 Oct 2012 20:48:54 +0000 (20:48 +0000)]
Fixup comment.
llvm-svn: 165427
Eric Christopher [Mon, 8 Oct 2012 20:48:49 +0000 (20:48 +0000)]
Fixup comments.
llvm-svn: 165426
Daniel Jasper [Mon, 8 Oct 2012 20:32:51 +0000 (20:32 +0000)]
CompilationDatabaseTest: Fix another Windows path issue.
llvm-svn: 165425
Daniel Jasper [Mon, 8 Oct 2012 20:08:25 +0000 (20:08 +0000)]
CompilationDatabase: Use //net paths for tests, as they should be
considered absolute on all platforms.
llvm-svn: 165422
Enrico Granata [Mon, 8 Oct 2012 19:06:11 +0000 (19:06 +0000)]
<rdar://problem/
12200505> Fixing a logical error in SBProcess, where the get_process_thread_list function was creating invalid threads_access instances, and hence failing to correctly fill in the list
llvm-svn: 165421
Enrico Granata [Mon, 8 Oct 2012 19:01:10 +0000 (19:01 +0000)]
Silly me! There was a closing ) missing from one of the lines - and Python complained about syntax errors on the next line. It being a Friday afternoon made the rest
llvm-svn: 165420
Adhemerval Zanella [Mon, 8 Oct 2012 18:59:53 +0000 (18:59 +0000)]
PR12716: PPC crashes on vector compare
Vector compare using altivec 'vcmpxxx' instructions have as third argument
a vector register instead of CR one, different from integer and float-point
compares. This leads to a failure in code generation, where 'SelectSETCC'
expects a DAG with a CR register and gets vector register instead.
This patch changes the behavior by just returning a DAG with the
vector compare instruction based on the type. The patch also adds a testcase
for all vector types llvm defines.
It also included a fix on signed 5-bits predicates printing, where
signed values were not handled correctly as signed (char are unsigned by
default for PowerPC). This generates 'vspltisw' (vector splat)
instruction with SIM out of range.
llvm-svn: 165419
Andrew Trick [Mon, 8 Oct 2012 18:54:00 +0000 (18:54 +0000)]
misched: remove the unused getSpecialAddressLatency hook.
llvm-svn: 165418
Andrew Trick [Mon, 8 Oct 2012 18:53:57 +0000 (18:53 +0000)]
misched: remove forceUnitLatencies. Defaults are handled by the default SchedModel
llvm-svn: 165417
Andrew Trick [Mon, 8 Oct 2012 18:53:53 +0000 (18:53 +0000)]
misched: avoid scheduling an instruction twice.
llvm-svn: 165416
Benjamin Kramer [Mon, 8 Oct 2012 18:49:39 +0000 (18:49 +0000)]
Don't emit double parentheses in __clang_version__.
PR14040.
llvm-svn: 165415
Daniel Jasper [Mon, 8 Oct 2012 18:37:21 +0000 (18:37 +0000)]
Hopefully appease Windows buildbot with Japanese (?) error message.
llvm-svn: 165414
Daniel Jasper [Mon, 8 Oct 2012 18:31:54 +0000 (18:31 +0000)]
FileNameMatchTrie: Use StringRef instead of Twines where possible.
llvm-svn: 165412
Adhemerval Zanella [Mon, 8 Oct 2012 18:25:11 +0000 (18:25 +0000)]
PowerPC: Fix object creation with PPC::MTCRF8 instruction
llvm-svn: 165411
Enrico Granata [Mon, 8 Oct 2012 17:32:55 +0000 (17:32 +0000)]
Retrying to apply Vishal's patch - hopefully this time it won't break Jason's build
llvm-svn: 165410
Adhemerval Zanella [Mon, 8 Oct 2012 17:27:24 +0000 (17:27 +0000)]
Add floating-point to and from integer conversion
This patch add altivec support for v4i32 to v4f32 and for v4f32 to
v4i32 vector rounding conversion.
llvm-svn: 165409
Micah Villmow [Mon, 8 Oct 2012 17:26:19 +0000 (17:26 +0000)]
Move TargetData to DataLayout to fix build breakage caused by LLVM r16540
llvm-svn: 165408
Micah Villmow [Mon, 8 Oct 2012 17:06:25 +0000 (17:06 +0000)]
Fix the ocaml binding breakage from TargetData -> DataLayout changes.
llvm-svn: 165406
Micah Villmow [Mon, 8 Oct 2012 16:40:38 +0000 (16:40 +0000)]
Move TargetData to DataLayout.
llvm-svn: 165404
Micah Villmow [Mon, 8 Oct 2012 16:39:34 +0000 (16:39 +0000)]
Move TargetData to DataLayout.
llvm-svn: 165403
Micah Villmow [Mon, 8 Oct 2012 16:38:25 +0000 (16:38 +0000)]
Move TargetData to DataLayout.
llvm-svn: 165402
Micah Villmow [Mon, 8 Oct 2012 16:37:04 +0000 (16:37 +0000)]
Move TargetData to DataLayout.
llvm-svn: 165401
Micah Villmow [Mon, 8 Oct 2012 16:29:26 +0000 (16:29 +0000)]
Move TargetData to DataLayout.
llvm-svn: 165397
Micah Villmow [Mon, 8 Oct 2012 16:28:57 +0000 (16:28 +0000)]
Move TargetData to DataLayout.
llvm-svn: 165396
Micah Villmow [Mon, 8 Oct 2012 16:25:52 +0000 (16:25 +0000)]
Move TargetData to DataLayout.
llvm-svn: 165395
Daniel Jasper [Mon, 8 Oct 2012 16:08:15 +0000 (16:08 +0000)]
Support symlinks and relative paths in complilation databases.
Review: http://llvm-reviews.chandlerc.com/D30
llvm-svn: 165392
Alexey Samsonov [Mon, 8 Oct 2012 13:11:18 +0000 (13:11 +0000)]
[ASan] use llvm-symbolizer for ASan lit-style tests on Mac
llvm-svn: 165391
James Molloy [Mon, 8 Oct 2012 13:06:30 +0000 (13:06 +0000)]
Some regression tests which are testing the old jit and are exercising functionality which is both known to be broken and not expected to be fixed in the old jit. To remove these from the regression test output, I've marked them XFAIL (for lit tests) and ifdef'd them out (unit tests). These modifications remove the last long-standing regression test failures from the buildbots (though updating the triple to reflect new ubuntu configuration has temporarily caused some new failures). Tested on x86-64 and ARM Linux.
Patch by David Tweed!
llvm-svn: 165390
Alexey Samsonov [Mon, 8 Oct 2012 10:17:44 +0000 (10:17 +0000)]
llvm-symbolizer for Mac: add support for reading DWARF from separate object file in resource directory generated by dsymutil. Fix minor symbol table lookup issues.
llvm-svn: 165389
Tobias Grosser [Mon, 8 Oct 2012 08:56:52 +0000 (08:56 +0000)]
Rename TargetData -> DataLayout
Contributed by: Sameer Sahasrabuddhe <sameer.sahasrabuddhe@amd.com>
llvm-svn: 165387
Peter Collingbourne [Mon, 8 Oct 2012 03:39:21 +0000 (03:39 +0000)]
Implement any() builtin. Patch by Tom Stellard!
llvm-svn: 165386
Peter Collingbourne [Mon, 8 Oct 2012 03:39:05 +0000 (03:39 +0000)]
Add native_powr builtin. Patch by Tom Stellard!
llvm-svn: 165385
David Blaikie [Mon, 8 Oct 2012 01:19:49 +0000 (01:19 +0000)]
Use a single note diagnostic for all the precedent/parentheses warnings.
llvm-svn: 165384
David Blaikie [Mon, 8 Oct 2012 01:11:04 +0000 (01:11 +0000)]
StringRef-ify Binary/UnaryOperator::getOpcodeStr
llvm-svn: 165383
Craig Topper [Sun, 7 Oct 2012 23:03:22 +0000 (23:03 +0000)]
Remove unused MachineInstr constructors that don't take a DebugLoc argument.
llvm-svn: 165382
Craig Topper [Sun, 7 Oct 2012 20:31:05 +0000 (20:31 +0000)]
Fix indentation. Remove 'else' after return. No functional change.
llvm-svn: 165381
Sean Silva [Sun, 7 Oct 2012 18:49:28 +0000 (18:49 +0000)]
Silence Sphinx warnings.
Found the fix on this page:
http://permalink.gmane.org/gmane.comp.python.sphinx.devel/112
llvm-svn: 165380
Tobias Grosser [Sun, 7 Oct 2012 17:43:23 +0000 (17:43 +0000)]
ScopLib: Support negated access functions.
Scoplib only supports access functions, but not the more generic
access relations. This commit now also supports access functions
that where not directly expresses as A[sub] with sub = i + 5b,
but with A[sub] with -sub = -i + (-5b).
Test case to come.
Contributed by: Dustin Feld <d3.feld@gmail.com>
llvm-svn: 165379
Tobias Grosser [Sun, 7 Oct 2012 17:43:18 +0000 (17:43 +0000)]
Update the 'News' section on the Polly page
llvm-svn: 165378
Benjamin Kramer [Sun, 7 Oct 2012 15:34:27 +0000 (15:34 +0000)]
X86: fcmov doesn't handle all possible EFLAGS, fall back to a branch for the others.
Otherwise it will try to use SSE patterns and fail horribly if sse is disabled.
Fixes PR14035.
llvm-svn: 165377
Dmitry Vyukov [Sun, 7 Oct 2012 14:21:24 +0000 (14:21 +0000)]
tsan: better message about found races for Go ThreadSanitizer
llvm-svn: 165376
Benjamin Kramer [Sun, 7 Oct 2012 11:51:45 +0000 (11:51 +0000)]
Python bindings: fix typo.
llvm-svn: 165375
Benjamin Kramer [Sun, 7 Oct 2012 11:46:37 +0000 (11:46 +0000)]
Python bindings: Add missing availability kind.
Patch by Leo Liu, test case by me.
llvm-svn: 165374
Bill Wendling [Sun, 7 Oct 2012 08:55:05 +0000 (08:55 +0000)]
Move more methods out-of-line. This is in preparation for changing the internal
contents of the Attributes class over to an AttributesImpl.
llvm-svn: 165373
Bill Wendling [Sun, 7 Oct 2012 07:10:13 +0000 (07:10 +0000)]
Sphinxify the GettinStarted documentation.
llvm-svn: 165372
Bill Wendling [Sun, 7 Oct 2012 04:56:08 +0000 (04:56 +0000)]
Sphinxify the ExtendingLLVM documentation.
llvm-svn: 165371
Rafael Espindola [Sun, 7 Oct 2012 04:44:33 +0000 (04:44 +0000)]
Use getArch instead of getArchName + string compare.
llvm-svn: 165370
Bill Wendling [Sun, 7 Oct 2012 04:34:10 +0000 (04:34 +0000)]
Sphinxify the compiler writer info documentation.
llvm-svn: 165369
Rafael Espindola [Sun, 7 Oct 2012 03:23:40 +0000 (03:23 +0000)]
Use getArch instead of getArchName.
The darwin change should be a nop since Triple::getArchTypeForDarwinArchName
doesn't know about amd64.
If things like amd64-mingw32 are to be rejected, we should print a error
earlier on instead of silently using the wrong abi.
Remove old comment that looks out of place, this is "in clang".
llvm-svn: 165368
Bob Wilson [Sun, 7 Oct 2012 01:11:19 +0000 (01:11 +0000)]
Make sure always-inline functions get inlined. <rdar://problem/
12423986>
Without this change, when the estimated cost for inlining a function with
an "alwaysinline" attribute was lower than the inlining threshold, the
getInlineCost function was returning that estimated cost rather than the
special InlineCost::AlwaysInlineCost value. That is fine in the normal
inlining case, but it can fail when the inliner considers the opportunity
cost of inlining into an internal or linkonce-odr function. It may decide
not to inline the always-inline function in that case. The fix here is just
to make getInlineCost always return the special value for always-inline
functions. I ran into this building clang with libc++. Tablegen failed to
link because of an always-inline function that was not inlined. I have been
unable to reduce the testcase down to a reasonable size.
llvm-svn: 165367
Rafael Espindola [Sun, 7 Oct 2012 00:56:09 +0000 (00:56 +0000)]
Document MapVector.
llvm-svn: 165366
Dmitri Gribenko [Sat, 6 Oct 2012 17:23:59 +0000 (17:23 +0000)]
List of potential checkers: smart pointer checker: actually, unique_ptr<T[]>
will do the right thing for new[] allocated memory. Thanks David!
llvm-svn: 165365
Dmitri Gribenko [Sat, 6 Oct 2012 17:14:39 +0000 (17:14 +0000)]
Clang static analyzer docs: fix a couple of HTML markup issues and non-UTF-8
characters.
llvm-svn: 165364
Dmitri Gribenko [Sat, 6 Oct 2012 16:59:15 +0000 (16:59 +0000)]
List of potential checkers: more C++11 details for the smart pointer checker.
llvm-svn: 165363
Benjamin Kramer [Sat, 6 Oct 2012 14:42:22 +0000 (14:42 +0000)]
Expose __builtin_bswap16.
GCC has always supported this on PowerPC and 4.8 supports it on all platforms,
so it's a good idea to expose it in clang too. LLVM supports this on all targets.
llvm-svn: 165362
Ted Kremenek [Sat, 6 Oct 2012 05:25:43 +0000 (05:25 +0000)]
Place warn_impcast_different_enum_types in a new warning group, -Wenum-conversion, that is a subgroup of -Wconversion.
llvm-svn: 165361
Ted Kremenek [Sat, 6 Oct 2012 05:11:15 +0000 (05:11 +0000)]
Add link to potential future checkers.
llvm-svn: 165360
Ted Kremenek [Sat, 6 Oct 2012 05:09:43 +0000 (05:09 +0000)]
Add list of potential checkers. Provided by Anton Yartsev!
llvm-svn: 165359
Jordan Rose [Sat, 6 Oct 2012 02:06:17 +0000 (02:06 +0000)]
[analyzer] Tweak test to run the retain-count checker even on non-Darwin.
This should fix the bots.
llvm-svn: 165358
Jason Molenda [Sat, 6 Oct 2012 02:02:26 +0000 (02:02 +0000)]
In DynamicLoaderDarwinKernel::OSKextLoadedKextSummary::LoadImageUsingMemoryModule(),
if we have a kernel binary, set the target's architecture to match.
Include the target's architecture in the ModuleSpec when we're searching for the
kext binaries on the local system -- otherwise we won't get a specific slice of
a fat file picked out for us and we won't use the returned Module correctly.
Remove the redundant attempt to find a file on the local filesystem from this method.
In ProcessGDBRemote::CheckForKernel(), if we have a kernel binary in memory, mark
the canJIT as false. There is no jitting code in kernel debug sessions.
llvm-svn: 165357
Jason Molenda [Sat, 6 Oct 2012 01:46:12 +0000 (01:46 +0000)]
Revert Vishal's patch that Enrico commited, at least for the weekend. With it applied,
starting lldb I get
% ./lldb -x
Traceback (most recent call last):
File "<string>", line 1, in <module>
File "/private/tmp/build/Debug/LLDB.framework/Versions/A/Resources/Python/lldb/__init__.py", line 9008
raise TypeError("No array item of type %s" % str(type(key)))
^
SyntaxError: invalid syntax
Traceback (most recent call last):
File "<string>", line 1, in <module>
NameError: name 'run_one_line' is not defined
Traceback (most recent call last):
File "<string>", line 1, in <module>
NameError: name 'run_one_line' is not defined
Traceback (most recent call last):
File "<string>", line 1, in <module>
NameError: name 'run_one_line' is not defined
(lldb)
I did a clean build and still got the problem so I'm backing this out until Enrico can
look at it.
llvm-svn: 165356
Jordan Rose [Sat, 6 Oct 2012 01:19:36 +0000 (01:19 +0000)]
ParentMap: Restore the ability to update an existing map.
The Clang ASTs are a DAG, not a pure tree. However, ParentMap has to
choose a single parent for each object. In the main (only?) cases in
which the AST forms a DAG, it protects from multiple traversal by using
OpaqueValueExprs. Previously, ParentMap would just unconditionally look
through all OpaqueValueExprs when building its map.
In order to make this behavior better for the analyzer's diagnostics,
ParentMap was changed to not set a statement's parent if there already
was one in the map. However, ParentMap is supposed to allow updating
existing mappings by calling addStmt once again. This change makes the
"transparency" of OpaqueValueExprs explicit, and disables it when it
is not desired, rather than checking the current contents of the map.
This new code seems like a big change, but it should actually have
essentially the same performance as before. Only OpaqueValueExprs and
their users (PseudoObjectExpr and BinaryConditionalOperator) will
have any different behavior.
There should be no user-visible functionality change, though a test
has been added for the current behavior of BinaryConditionalOperator
source locations and accompanying Xcode arrows (which are not so great...).
llvm-svn: 165355
Jordan Rose [Sat, 6 Oct 2012 01:19:30 +0000 (01:19 +0000)]
[analyzer] Handle implicit statements used for end-of-path nodes' source locs.
Some implicit statements, such as the implicit 'self' inserted for "free"
Objective-C ivar access, have invalid source locations. If one of these
statements is the location where an issue is reported, we'll now look at
the enclosing statements for a valid source location.
<rdar://problem/
12446776>
llvm-svn: 165354