Commit Graph

69044 Commits

Author SHA1 Message Date
Gheorghe-Teodor Bercea 47633db42e Add comma to comment.
llvm-svn: 305294
2017-06-13 15:35:27 +00:00
Krasimir Georgiev 54ef4bafbe [clang-format] Document the StartOfTokenColumn parameter, NFC
llvm-svn: 305293
2017-06-13 14:58:55 +00:00
Diana Picus bec724cbb0 Revert "Revert r301742 which made ExprConstant checking apply to all full-exprs."
This reverts commit r305239 because it broke the buildbots (the
diag-flags.cpp test is failing).

llvm-svn: 305287
2017-06-13 12:50:06 +00:00
Francois Ferrand 2a81ca8d61 clang-format: add option to merge empty function body
Summary:
This option supplements the AllowShortFunctionsOnASingleLine flag, to
merge empty function body at the beginning of the line: e.g. when the
function is not short-enough and breaking braces after function.

  int f()
  {}

Reviewers: krasimir, djasper

Reviewed By: djasper

Subscribers: klimek, cfe-commits

Differential Revision: https://reviews.llvm.org/D33447

llvm-svn: 305272
2017-06-13 07:02:43 +00:00
Vedant Kumar 840c2c758d [docs] Add some ubsan changes to the release notes
llvm-svn: 305269
2017-06-13 02:52:31 +00:00
Nick Lewycky 54992386f3 Revert r301742 which made ExprConstant checking apply to all full-exprs.
This patch also exposed pre-existing bugs in clang, see PR32864 and PR33140#c3 .

llvm-svn: 305239
2017-06-12 21:59:18 +00:00
Richard Trieu 11d566acee [ODRHash] Add diagnostic messages for typedef and type alias.
llvm-svn: 305238
2017-06-12 21:58:22 +00:00
Nick Lewycky 13073a6425 Revert r303316, a change to ExprConstant to evaluate function arguments.
The patch was itself correct but it uncovered other bugs which are going to be difficult to fix, per PR33140.

llvm-svn: 305233
2017-06-12 21:15:44 +00:00
Peter Collingbourne 89061b2224 IR: Replace the "Linker Options" module flag with "llvm.linker.options" named metadata.
The new metadata is easier to manipulate than module flags.

Differential Revision: https://reviews.llvm.org/D31349

llvm-svn: 305227
2017-06-12 20:10:48 +00:00
Reid Kleckner 06a4b2ae1d Correct debug info bit offset calculation for big-endian targets
Summary:
The change "[CodeView] Implement support for bit fields in
Clang" (r274201, https://reviews.llvm.org/rL274201) broke the
calculation of bit offsets for the debug info describing bitfields on
big-endian targets.

Prior to commit r274201 the debug info for bitfields got their offsets
from the ASTRecordLayout in CGDebugInfo::CollectRecordFields(), the
current field offset was then passed on to
CGDebugInfo::CollectRecordNormalField() and used directly in the
DIDerivedType.

Since commit r274201, the bit offset ending up in the DIDerivedType no
longer comes directly from the ASTRecordLayout. Instead
CGDebugInfo::CollectRecordNormalField() calls the new method
CGDebugInfo::createBitFieldType(), which in turn calls
CodeGenTypes::getCGRecordLayout().getBitFieldInfo() to fetch a
CGBitFieldInfo describing the field. The 'Offset' member of
CGBitFieldInfo is then used to calculate the bit offset of the
DIDerivedType. Unfortunately the previous and current method of
calculating the bit offset are only equivalent for little endian
targets, as CGRecordLowering::setBitFieldInfo() reverses the bit
offsets for big endian targets as the last thing it does.

A simple reproducer for this error is the following module:

struct fields {
  unsigned a : 4;
  unsigned b : 4;
} flags = {0x0f, 0x1};

Compiled for Mips, with commit r274200 both the DIDerivedType bit
offsets on the IR-level and the DWARF information on the ELF-level
will have the expected values: the offsets of 'a' and 'b' are 0 and 4
respectively. With r274201 the offsets are switched to 4 and 0. By
noting that the static initialization of 'flags' in both cases is the
same, we can eliminate a change in record layout as the cause of the
change in the debug info. Also compiling this example with gcc,
produces the same record layout and debug info as commit r274200.

In order to restore the previous function we extend
CGDebugInfo::createBitFieldType() to compensate for the reversal done
in CGRecordLowering::setBitFieldInfo().

Patch by Frej Drejhammar!

Reviewers: cfe-commits, majnemer, rnk, aaboud, echristo, aprantl

Reviewed By: rnk, aprantl

Subscribers: aprantl, arichardson, frej

Differential Revision: https://reviews.llvm.org/D32745

llvm-svn: 305224
2017-06-12 19:57:56 +00:00
Alex Lorenz 1345ea2a05 Recommit r305117: [libclang] Merge multiple availability clauses when
getting the platform's availability

Patch by Ronald Wampler!

Differential Revision: https://reviews.llvm.org/D33478

llvm-svn: 305221
2017-06-12 19:06:30 +00:00
Vedant Kumar 6dbf4274a5 [ubsan] Detect invalid unsigned pointer index expression (clang)
Adding an unsigned offset to a base pointer has undefined behavior if
the result of the expression would precede the base. An example from
@regehr:

  int foo(char *p, unsigned offset) {
    return p + offset >= p; // This may be optimized to '1'.
  }

  foo(p, -1); // UB.

This patch extends the pointer overflow check in ubsan to detect invalid
unsigned pointer index expressions. It changes the instrumentation to
only permit non-negative offsets in pointer index expressions when all
of the GEP indices are unsigned.

Testing: check-llvm, check-clang run on a stage2, ubsan-instrumented
build.

Differential Revision: https://reviews.llvm.org/D33910

llvm-svn: 305216
2017-06-12 18:42:51 +00:00
Yaron Keren 374b41a6aa Address David Blaikie comment by replacing grep with FileCheck.
llvm-svn: 305215
2017-06-12 18:29:37 +00:00
Yaron Keren c91c46c85e Add regression test for r305179.
llvm-svn: 305213
2017-06-12 18:05:13 +00:00
Artem Dergachev 11150c009a [analyzer] Fix a crash when an ObjC object is constructed in AllocaRegion.
Memory region allocated by alloca() carries no implicit type information.
Don't crash when resolving the init message for an Objective-C object
that is being constructed in such region.

rdar://problem/32517077

Differential Revision: https://reviews.llvm.org/D33828

llvm-svn: 305211
2017-06-12 17:59:50 +00:00
George Burgess IV b5fe855dfb [Sema] Use the right FoldingSet.
We were doing FindNodeOrInsertPos on SubstTemplateTypeParmPackTypes, so
we should presumably be inserting into SubstTemplateTypeParmPackTypes.

Looks like the FoldingSet API can be tweaked a bit so that we can catch
things like this at compile-time. I'll look into that shortly.

I'm unsure of how to test this; suggestions welcome.

Thanks to Vladimir Voskresensky for bringing this up!

llvm-svn: 305207
2017-06-12 17:44:30 +00:00
Erik Pilkington 21ff345d64 [Sema][C++1z] Ensure binding in dependent range for have non-null type
Fixes PR32172

Differential revision: https://reviews.llvm.org/D34096

llvm-svn: 305195
2017-06-12 16:11:06 +00:00
Daniel Jasper cdc4408bbf Revert r305164/5/7.
cc1as does not currently access the "--" version of this flag. At the
very least this needs to be fixed and proper test cases need to be
added.

Simple reproducer:
clang -Wa,--compress-debug-sections /tmp/test.cc

Result:
error: unknown argument: '--compress-debug-sections'

llvm-svn: 305182
2017-06-12 08:08:18 +00:00
John McCall cb731548fa Don't crash when forming a destructor name on an incomplete type.
Fixes PR25156.

Patch by Don Hinton!

llvm-svn: 305169
2017-06-11 20:33:00 +00:00
Saleem Abdulrasool 617034da8f test: attempt to repair build bots
Split the no-ias tests and give them a target to ensure that they go
down the GNU toolchain path.  Adjust the no compression support tests.

llvm-svn: 305167
2017-06-11 18:55:17 +00:00
Saleem Abdulrasool 7289ba9165 Driver: add support for `-gz` and `-gz=`
These options control the behaviour of the compression of debug info
sections on ELF targets.  Our behaviour slightly diverges from the
behaviour of GCC.  `-gz` maps to the `-compress-debug-sections` rather
than `-compress-debug-sections=zlib` or
`-compress-debug-sections=zlib-gnu`.  This small divergence allows us to
be compatible across versions of binutils (=zlib support was introduced
in 2.26, while earlier versions only support =zlib-gnu).  This also
allows users to not have to worry about the version of the assembler
they may be using if they are not using the IAS.  Previously, users
would have had to go through the internal option
`-compress-debug-sectionss` and pass that through to the assembler,
which is no longer needed.

llvm-svn: 305165
2017-06-11 17:49:23 +00:00
Saleem Abdulrasool d3ba0ac2a5 Driver: pass along [-]-[no]compress-debug-sections unfiltered
Rather than validating the flags, pass them through without any
validation.  Arguments passed via -Wa or -Xassembler are passed directly
to the assembler without validation.  The validation was previously
required since we did not provide proper driver level support for
controlling the debug compression on ELF targets.  A subsequent change
will add support for the `-gz` and `-gz=` flags which provide proper
driver level control of the ELF compressed debug sections.

llvm-svn: 305164
2017-06-11 17:49:17 +00:00
Roman Lebedev 5806d9f205 Revert "[clang] Implement -Wcast-qual for C++"
Breaks -Werror builders.

llvm-svn: 305148
2017-06-10 17:49:23 +00:00
Roman Lebedev b0120740c4 [clang] Implement -Wcast-qual for C++
Summary:
This way, the behavior of that warning flag
more closely resembles that of GCC.

Do note that there is at least one false-negative (see FIXME in tests).

Fixes PR4802.

Testing:
```
ninja check-clang-sema check-clang-semacxx
```

Reviewers: dblaikie, majnemer, rnk

Reviewed By: dblaikie, rnk

Subscribers: cfe-commits, alexfh, rnk

Differential Revision: https://reviews.llvm.org/D33102

llvm-svn: 305147
2017-06-10 17:19:19 +00:00
Richard Trieu be5cb93088 Revert r305110 to fix buildbot
llvm-svn: 305130
2017-06-09 23:03:40 +00:00
Erich Keane 3cf69bce69 Support operator keywords used in Windows SDK(fix ubsan)
UBSan found an issue with a nullptr being assigned to a reference.
This was because a following function went back and checked the 
identifier in the CPPOperatorName case.  This patch corrects that
location with the original logic as well.

llvm-svn: 305128
2017-06-09 22:50:02 +00:00
Richard Smith 01e4a7f29b 27037: Use correct CVR qualifier on an upcast on method pointer call
Patch by Taiju Tsuiki!

Differential Revision: https://reviews.llvm.org/D33875

llvm-svn: 305126
2017-06-09 22:25:28 +00:00
Alexander Shaposhnikov f3e877017b [clang] Cleanup fixit.c
This diff removes temporary file t2 in fixit.c and updates the test command accordingly.
NFC.

Test plan:
make check-all

Differential revision: https://reviews.llvm.org/D34066

llvm-svn: 305124
2017-06-09 22:20:52 +00:00
Alex Lorenz 35cc699d7a Revert r305117
It caused `Index/availability.c` test failure on Linux

llvm-svn: 305122
2017-06-09 22:06:36 +00:00
Richard Smith 76c51dc5b2 Attempt to fix shared library build: RewriteFrontend depends on Serialization after r305116.
llvm-svn: 305121
2017-06-09 22:02:33 +00:00
Vassil Vassilev 497a99523a [modules] D29951: Load lazily the template specialization in multi-module setups.
Currently, we load all template specialization if we have more than one module
attached and we touch anything around the template definition.

This patch registers the template specializations as lazily-loadable entities.
In some TUs it reduces the amount of deserializations by 1%.

llvm-svn: 305120
2017-06-09 21:54:18 +00:00
Vassil Vassilev 41cafcd49c [modules] Fix that global delete operator get's assigned to a submodule.
n the current local-submodule-visibility mode, as soon as we discover a virtual
destructor, we declare on demand a global delete operator. However, this causes
that this delete operator is owned by the submodule which contains said virtual
destructor. This means that other modules no longer can see the global delete
operator which is hidden inside another submodule and fail to compile.

This patch unhides those global allocation function once they're created to
prevent this issue.

Patch by Raphael Isemann (D33366)!

llvm-svn: 305118
2017-06-09 21:36:28 +00:00
Alex Lorenz 2e34be23a2 [libclang] Merge multiple availability clauses when getting the platform's
availability

Patch by Ronald Wampler!

Differential Revision: https://reviews.llvm.org/D33478

llvm-svn: 305117
2017-06-09 21:29:45 +00:00
Richard Smith 86a3ef5b03 Add -frewrite-imports flag.
If specified, when preprocessing, the contents of imported .pcm files will be
included in preprocessed output. The resulting preprocessed file can then be
compiled standalone without the module sources or .pcm files.

llvm-svn: 305116
2017-06-09 21:24:02 +00:00
Richard Trieu 4ab80f2e9e [ODRHash] Add support for TemplateArgument types.
Recommit r304592 that was reverted in r304618.  r305104 should have fixed the
issue.

llvm-svn: 305110
2017-06-09 21:00:10 +00:00
Richard Trieu f21b803876 [ODRHash] Skip inline namespaces when hashing.
Speculatively try to fix the underlying issue from r304592, of underlying types
being confused when inline namespaces are used.

llvm-svn: 305104
2017-06-09 20:11:51 +00:00
Benjamin Kramer 29e23f3ccf Bringt back -triple so the test passes on non-x86.
llvm-svn: 305103
2017-06-09 19:47:36 +00:00
Richard Smith 5d2ed48987 Add #pragma clang module build/endbuild pragmas for performing a module build
as part of a compilation.

This is intended for two purposes:

1) Writing self-contained test cases for modules: we can now write a single
source file test that builds some number of module files on the side and
imports them.

2) Debugging / test case reduction. A single-source testcase is much more
amenable to reduction, compared to a VFS tarball or .pcm files.

llvm-svn: 305101
2017-06-09 19:22:32 +00:00
Benjamin Kramer 96a0b3fc6c [ASTMatchers] Fix use after free.
Found by asan.

llvm-svn: 305094
2017-06-09 17:55:42 +00:00
Vassil Vassilev 0e1a4ca42e Repair 2010-05-31-palignr.c test
This test was silently failing since a long time because it failed to include
stdlib.h (as it's running in a freestanding environment). However, because we
 used just not clang_cc1 instead of the verify mode, this regression was never
 noticed and the test was just always passing.

This adds -ffreestanding to the invocation, so that tmmintrin.h doesn't
indirectly include mm_malloc.h, which in turns includes the unavailable stdlib.h.
We also run now in the -verify mode to prevent that we silently regress again.

I've also updated the test to no longer check the return value of _mm_alignr_epi8
as this is also causing it to fail (and it's not really the job of this test to
test this).


Patch by Raphael Isemann (D34022)

llvm-svn: 305089
2017-06-09 16:42:26 +00:00
Erich Keane 33c3d8a916 support operator keywords used in Windows SDK
to support operator keywords used in Windows SDK, alter token type when 
seen in system headers

Hello, I submitted D33505 to address this problem, but the 
proposal was rejected as too big a hammer.
This change will allow clang to parse the WindowsSDK header <query.h> 
which uses the operator name "or" as a field name. Treat cpp operator 
keywords as ordinary identifiers inside the Microsoft headers, but 
treat them as usual in the user's program.

Original Submitter: Melanie Blower (mibintc)

Differential Revision: https://reviews.llvm.org/D33782

llvm-svn: 305087
2017-06-09 16:29:35 +00:00
Alexey Bataev 24f7101876 [DebugInfo] Fix comment, NFC.
llvm-svn: 305076
2017-06-09 13:55:08 +00:00
Alexey Bataev 56223237b0 [DebugInfo] Add kind of ImplicitParamDecl for emission of FlagObjectPointer.
Summary:
If the first parameter of the function is the ImplicitParamDecl, codegen
automatically marks it as an implicit argument with `this` or `self`
pointer. Added internal kind of the ImplicitParamDecl to separate
'this', 'self', 'vtt' and other implicit parameters from other kind of
parameters.

Reviewers: rjmccall, aaron.ballman

Subscribers: cfe-commits

Differential Revision: https://reviews.llvm.org/D33735

llvm-svn: 305075
2017-06-09 13:40:18 +00:00
Nikolai Bozhenov de57795cc8 Revert of r305066 "Reapply Frontend support for Nios2 target"
llvm-svn: 305068
2017-06-09 12:09:29 +00:00
Nikolai Bozhenov b2de17c734 Reapply "Frontend support for Nios2 target"
Summary:
- Implements TargetInfo class for Nios2 target.
- Enables handling of -march and -mcpu options for Nios2 target.
- Definition of Nios2 builtin functions.

Reviewed By: craig.topper

Differential Revision: https://reviews.llvm.org/D33356

Author: belickim <mateusz.belicki@intel.com>
llvm-svn: 305066
2017-06-09 10:56:18 +00:00
Erik Verbruggen efe6fa501c Speed up preamble loading
Cache filename - SourceLocation pairs to speed up preamble loading and
global completion. This is especially relevant for windows, where
preamble loading takes a while.

Patch by Ivan Donchevskii!

Differential Revision: http://reviews.llvm.org/D33493

llvm-svn: 305061
2017-06-09 08:29:58 +00:00
Richard Smith d9259c2f5e Remove 'Filename' parameter from BeginSourceFileAction.
No-one was using this, and it's not meaningful in general -- FrontendActions
can be run on inputs that don't have a corresponding source file. The current
frontend input can be obtained by asking the FrontendAction if any future
action actually needs it.

llvm-svn: 305045
2017-06-09 01:36:10 +00:00
Argyrios Kyrtzidis 735e92c29a [libclang] Introduce a new parsing option 'CXTranslationUnit_SingleFileParse' that puts preprocessor in a mode for parsing a single file only.
This is useful for parsing a single file, as a fast/inaccurate 'mode' that can still provide declarations from the file, like the classes and their methods.

llvm-svn: 305044
2017-06-09 01:20:48 +00:00
Saleem Abdulrasool 54448909bf Represent debug information compression type fully
This is tied with the LLVM side of the change to expose the debug
information compression types to clang.  We now track the compression
type as an enumeration rather than a boolean.  We still use the same
value (GNU) that we did previously.  This is in preparation to support
passing down the compression type and switch it based on the command
line.

llvm-svn: 305039
2017-06-09 00:40:30 +00:00
Kostya Serebryany 2c2fb8896b [sanitizer-coverage] one more flavor of coverage: -fsanitize-coverage=inline-8bit-counters. Experimental so far, not documenting yet. Reapplying revisions 304630, 304631, 304632, 304673, see PR33308
llvm-svn: 305026
2017-06-08 22:58:19 +00:00