Manuel Klimek [Fri, 2 Nov 2012 01:31:03 +0000 (01:31 +0000)]
Insert interception point onStartOfTranslationUnit.
Often users of the ASTMatchers want to add tasks that are done once per
translation unit, for example, cleaning up caches. Combined with the
interception point for the end of source file one can add to the factory
creation, this covers the cases we've seen users need.
Richard Smith [Thu, 1 Nov 2012 22:13:39 +0000 (22:13 +0000)]
Remove divison-by-zero checks from -ftrapv. These checks were incompatible with
g++'s -ftrapv, failed to call the -ftrapv overflow handler, and are still
available under -fcatch-undefined-behavior.
objective-C block meta-data. This patch completes meta-data
generation for captured block variables in arc mode. This includes
inlined version of the meta-data when it can be done. It also includes
severat tests. This is wip. // rdar://12184410.
Make the FilenameRange of the InclusionDirective callback more accurate,
preserve the macro location of the range end if the filename came from a macro.
Richard Smith [Thu, 1 Nov 2012 07:22:08 +0000 (07:22 +0000)]
-fcatch-undefined-behavior: Start checking loads and stores for null pointers.
We want the diagnostic, and if the load is optimized away, we still want to
trap it. Stop checking non-default address spaces; that doesn't work in
general.
Richard Smith [Thu, 1 Nov 2012 03:48:49 +0000 (03:48 +0000)]
Remove CompilerInvocation::toArgs and clang -cc1test mode. These were untested
and apparently unused (and since they are untested, they're presumably also
broken).
The stat cache became essentially useless ever since we started
validating all file entries in the PCH.
But the motivating reason for removing it now is that it also affected
correctness in this situation:
-You have a header without include guards (using "#pragma once" or #import)
-When creating the PCH:
-The same header is referenced in an #include with different filename cases.
-In the PCH, of course, we record only one file entry for the header file
-But we cache in the PCH file the stat info for both filename cases
-Then the source files are updated and the header file is updated in a way that
its size and modification time are the same but its inode changes
-When using the PCH:
-We validate the headers, we check that header file and we create a file entry with its current inode
-There's another #include with a filename with different case than the previously created file entry
-In order to get its stat info we go through the cached stat info of the PCH and we receive the old inode
-because of the different inodes, we think they are different files so we go ahead and include its contents.
Removing the stat cache will potentially break clients that are attempting to use the stat cache
as a way of avoiding having the actual input files available. If that use case is important, patches are welcome
to bring it back in a way that will actually work correctly (i.e., emit a PCH that is self-contained, coping with
literal strings, line/column computations, etc.).
Chad Rosier [Wed, 31 Oct 2012 18:31:33 +0000 (18:31 +0000)]
[driver] Completely rework how superfluous options are stripped out of the crash
diagnostics script.
This addresses the FIXME pertaining to quoted arguments. We also delineate
between those flags that have an argument (e.g., -D macro, -MF file) and
those that do not (e.g., -M, -MM, -MG). Finally, we add the -dwarf-debug-flags
to the list of flags to be removed.
rdar://12329974
Bill Schmidt [Wed, 31 Oct 2012 17:31:46 +0000 (17:31 +0000)]
This patch removes the XFAIL for powerpc for
test/CodeGenCXX/member-alignment.cpp. The test succeeds for
powerpc64-unknown-linux-gnu. If other flavors of powerpc are
shown by buildbots to still be broken, we can adjust the test
at that time.
Jordan Rose [Wed, 31 Oct 2012 16:44:55 +0000 (16:44 +0000)]
[analyzer] Let ConstraintManager subclasses provide a more efficient checkNull.
Previously, every call to a ConstraintManager's isNull would do a full
assumeDual to test feasibility. Now, ConstraintManagers can override
checkNull if they have a cheaper way to do the same thing.
RangeConstraintManager can do this in less than half the work.
Simon Atanasyan [Wed, 31 Oct 2012 12:01:53 +0000 (12:01 +0000)]
Extend -Bprefix functionality and make it closer to gcc. If the "prefix"
is not a directory, Driver::GetProgramPath() routine does not try to append
the program name as a "path component" to it. It just joins the "prefix" with
the program name and checks the resulting path existence.
Chad Rosier [Tue, 30 Oct 2012 21:42:09 +0000 (21:42 +0000)]
[driver] Older versions of ld expect '-L<dir>' not '-L <dir>'. In Xcode4 and
later, '-L <dir>' is allowed, but rewrite these in the driver as '-L<dir>' to
maintain backward compatibility. The same is true for the -I option.
rdar://12366753
Anna Zaks [Tue, 30 Oct 2012 04:17:40 +0000 (04:17 +0000)]
[analyzer] Fix a bug in REGISTER_MAP_WITH_PROGRAMSTATE
The ImmutableMap should not be the key into the GDM map as there could
be several entries with the same map type. Thanks, Jordan.
This complicates the usage of the macro a bit. When we want to retrieve
the whole map, we need to use another name. Currently, I set it to be
Name ## Ty as in "type of the map we are storing in the ProgramState".
Jordan Rose [Tue, 30 Oct 2012 01:21:35 +0000 (01:21 +0000)]
[analyzer] New checker for missing super calls in UIViewController subclasses.
This is a syntactic checker aimed at helping iOS programmers correctly
subclass and override the methods of UIViewController. While this should
eventually be covered by the 'objc_requires_super' attribute, this
checker can be used with the existing iOS SDKs without any header changes.
This new checker is currently named 'alpha.osx.cocoa.MissingSuperCall'.
Patch by Julian Mayer!
Rafael Espindola [Tue, 30 Oct 2012 00:38:13 +0000 (00:38 +0000)]
getOriginalSourceFileName and getOriginalSourceFile can return a StringRef.
MaybeAddSystemRootToFilename doesn't need to return anything, it modifies
its argument.
[PCH] The diagnostic state points can refer to previously created
diagnostic states; make sure the ASTReader sets the diagnostic state
properly instead of always recreating it.
Rafael Espindola [Tue, 30 Oct 2012 00:13:16 +0000 (00:13 +0000)]
In the past "production" clang builds would not be used for c++, and
we had the -ccc-clang-cxx and -ccc-no-clang-cxx options to force them
on or off for testing.
Clang c++ support is now production quality and these options are dead.
Richard Smith [Mon, 29 Oct 2012 18:26:47 +0000 (18:26 +0000)]
Partially roll back r166898; it exposed a bug in the standard.
The problem is as follows: C++11 has contexts which are not
potentially-evaluated, and yet in which we are required or encouraged to
perform constant evaluation. In such contexts, we are not permitted to
implicitly define special member functions for literal types, therefore
we cannot evalaute those constant expressions.
Punt on this in one more context for now by skipping checking constexpr
variable initializers if they occur in dependent contexts.
Jordan Rose [Mon, 29 Oct 2012 17:46:47 +0000 (17:46 +0000)]
-Warc-repeated-use-of-weak: allow single reads in loops from local variables.
Previously, the warning would erroneously fire on this:
for (Test *a in someArray)
use(a.weakProp);
...because it looks like the same property is being accessed over and over.
However, clearly this is not the case. We now ignore loops like this for
local variables, but continue to warn if the base object is a parameter,
global variable, or instance variable, on the assumption that these are
not repeatedly usually assigned to within loops.
Additionally, do-while loops where the condition is 'false' are not really
loops at all; usually they're just used for semicolon-swallowing macros or
using "break" like "goto".
Jordan Rose [Mon, 29 Oct 2012 17:31:59 +0000 (17:31 +0000)]
[analyzer] New option to not suppress null return paths if an argument is null.
Our one basic suppression heuristic is to assume that functions do not
usually return NULL. However, when one of the arguments is NULL it is
suddenly much more likely that NULL is a valid return value. In this case,
we don't suppress the report here, but we do attach /another/ visitor to
go find out if this NULL argument also comes from an inlined function's
error path.
This new behavior, controlled by the 'avoid-suppressing-null-argument-paths'
analyzer-config option, is turned off by default. Turning it on produced
two false positives and no new true positives when running over LLVM/Clang.
This is one of the possible refinements to our suppression heuristics.
<rdar://problem/12350829>
Jordan Rose [Mon, 29 Oct 2012 17:31:53 +0000 (17:31 +0000)]
[analyzer] Use the CallEnter node to get a value for tracked null arguments.
Additionally, don't collect PostStore nodes -- they are often used in
path diagnostics.
Previously, we tried to track null arguments in the same way as any other
null values, but in many cases the necessary nodes had already been
collected (a memory optimization in ExplodedGraph). Now, we fall back to
using the value of the argument at the time of the call, which may not
always match the actual contents of the region, but often will.
This is a precursor to improving our suppression heuristic.
<rdar://problem/12350829>
Bill Schmidt [Mon, 29 Oct 2012 14:59:24 +0000 (14:59 +0000)]
This patch adds alignment information for long double to the 64-bit PowerPC
ELF subtarget.
The existing description string is moved from PPC64TargetInfo to its
DarwinTargetInfo subclass, to avoid any changes to the Darwin ABI.
PPC64TargetInfo now has two possible description strings: one for FreeBSD,
which requires 8-byte alignment, and a default string that requires
16-byte alignment.
I've added a test for PPC64 Linux to verify the 16-byte alignment. If
somebody wants to add a separate test for FreeBSD, that would be great.
Note that there is a companion patch to update the alignment information
in LLVM, which I am committing now as well.
Richard Smith [Sun, 28 Oct 2012 07:39:29 +0000 (07:39 +0000)]
Revert functional part of r166896 and just suppress -Wunneeded-internal-declaration for reference types for now. This needs more work; the cases we currently miss are a bit random.
Richard Smith [Sun, 28 Oct 2012 06:18:02 +0000 (06:18 +0000)]
When determining whether to try evaluating the initializer of a variable, check
whether the initializer is value-dependent rather than whether we are in a
dependent context. This allows us to detect some errors sooner, and fixes a
crash-on-invalid if a dependent type leaks out to a non-dependent context in
error recovery.