Richard Smith [Mon, 21 Jul 2014 05:27:31 +0000 (05:27 +0000)]
Add missing initialization found due to a valgrind false positive.
This field is never inspected in the object state initialized by this
constructor; however, initializing it seems reasonable, since it has
a meaningful value.
Richard Smith [Mon, 21 Jul 2014 04:10:40 +0000 (04:10 +0000)]
[modules] Fix some of the confusion when computing the override set for a macro
introduced by finalization. This is still not entirely correct; more fixes to
follow.
[OPENMP] Added several test cases for clauses 'ordered' and 'nowait': if there are more than one 'nowait' or 'ordered' clause an error message is expected.
[PowerPC] Optimize passing certain aggregates by value
In addition to enabling ELFv2 homogeneous aggregate handling,
LLVM support to pass array types directly also enables a performance
enhancement. We can now pass (non-homogeneous) aggregates that fit
fully in registers as direct integer arrays, using an element type
to encode the alignment requirement (that would otherwise go to the
"byval align" field).
This is preferable since "byval" forces the back-end to write the
aggregate out to the stack, even if it could be passed fully in
registers. This is particularly annoying on ELFv2, if there is
no parameter save area available, since we then need to allocate
space on the callee's stack just to hold those aggregates.
Note that to implement this optimization, this patch does not attempt
to fully anticipate register allocation rules as (defined in the
ABI and) implemented in the back-end. Instead, the patch is simply
passing *any* aggregate passed by value using the array mechanism
if its size is up to 64 bytes. This means that some of those will
end up being passed in stack slots anyway, but the generated code
shouldn't be any worse either. (*Large* aggregates remain passed
using "byval" to enable optimized copying via memcpy etc.)
This patch implements clang support for the PowerPC ELFv2 ABI.
Together with a series of companion patches in LLVM, this makes
clang/LLVM fully usable on powerpc64le-linux.
Most of the ELFv2 ABI changes are fully implemented on the LLVM side.
On the clang side, we only need to implement some changes in how
aggregate types are passed by value. Specifically, we need to:
- pass (and return) "homogeneous" floating-point or vector aggregates in
FPRs and VRs (this is similar to the ARM homogeneous aggregate ABI)
- return aggregates of up to 16 bytes in one or two GPRs
The second piece is trivial to implement in any case. To implement
the first piece, this patch makes use of infrastructure recently
enabled in the LLVM PowerPC back-end to support passing array types
directly, where the array element type encodes properties needed to
handle homogeneous aggregates correctly.
Specifically, the array element type encodes:
- whether the parameter should be passed in FPRs, VRs, or just
GPRs/stack slots (for float / vector / integer element types,
respectively)
- what the alignment requirements of the parameter are when passed in
GPRs/stack slots (8 for float / 16 for vector / the element type
size for integer element types) -- this corresponds to the
"byval align" field
With this support in place, the clang part simply needs to *detect*
whether an aggregate type implements a float / vector homogeneous
aggregate as defined by the ELFv2 ABI, and if so, pass/return it
as array type using the appropriate float / vector element type.
Hal Finkel [Sat, 19 Jul 2014 02:01:03 +0000 (02:01 +0000)]
TypePrinter should not ignore IndexTypeCVRQualifiers on constant-sized arrays
C99 array parameters can have index-type CVR qualifiers, and the TypePrinter
should print them when present (and we were not for constant-sized arrays).
Otherwise, we'd drop the restrict in:
Hal Finkel [Sat, 19 Jul 2014 01:41:07 +0000 (01:41 +0000)]
Use the dereferenceable attribute on C99 array parameters with static
In C99, an array parameter declarator might have the form:
direct-declarator '[' 'static' type-qual-list[opt] assign-expr ']'
where the static keyword indicates that the caller will always provide a
pointer to the beginning of an array with at least the number of elements
specified by the assignment expression. For constant sizes, we can use the
new dereferenceable attribute to pass this information to the optimizer. For
VLAs, we don't know the size, but (for addrspace(0)) do know that the pointer
must be nonnull (and so we can use the nonnull attribute).
Richard Smith [Sat, 19 Jul 2014 01:39:17 +0000 (01:39 +0000)]
PR20356: Fix all Sema warnings with mismatched ext_/warn_ versus
ExtWarn/Warnings. Mostly the name of the warning was changed to match the
semantics, but in the PR20356 cases, the warning was about valid code, so the
diagnostic was changed from ExtWarn to Warning instead.
David Blaikie [Sat, 19 Jul 2014 01:06:45 +0000 (01:06 +0000)]
Remove uses of the redundant ".reset(nullptr)" of unique_ptr, in favor of ".reset()"
It's also possible to just write "= nullptr", but there's some question
of whether that's as readable, so I leave it up to authors to pick which
they prefer for now. If we want to discuss standardizing on one or the
other, we can do that at some point in the future.
[MS-ABI] Assign SEH handler indices to __try blocks
Assigns indices to try blocks. These indices will used in constructing
tables that the mscrt function __except_handler3 reads during SEH.
Testing will occur once we actually emit the tables, in a subsequent
patch.
Hal Finkel [Fri, 18 Jul 2014 23:19:20 +0000 (23:19 +0000)]
TypePrinter should not omit the static keyword in array parameter declarators
In C99, an array parameter declarator might have the form: direct-declarator
'[' 'static' type-qual-list[opt] assign-expr ']'
and when the size of the array is a constant, don't omit the static keyword
when printing the type. Also, in the VLA case, put a space after the static
keyword (some assignment expression must follow it).
Objective-C. Patch to warn if the result of calling a property getter
is unused (this is match behavior when property-dot syntax is used to
use same getter). rdar://17514245
Patch by Anders Carlsson with minor refactoring by me.
Richard Smith [Fri, 18 Jul 2014 22:13:40 +0000 (22:13 +0000)]
Reinstate r213348, reverted in r213395, with an additional bug fix and more
thorough tests.
Original commit message:
[modules] Fix macro hiding bug exposed if:
* A submodule of module A is imported into module B
* Another submodule of module A that is not imported into B exports a macro
* Some submodule of module B also exports a definition of the macro, and
happens to be the first submodule of B that imports module A.
In this case, we would incorrectly determine that A's macro redefines B's
macro, and so we don't need to re-export B's macro at all.
This happens with the 'assert' macro in an LLVM self-host. =(
David Majnemer [Fri, 18 Jul 2014 19:53:17 +0000 (19:53 +0000)]
CodeGen: Properly null-check typeid expressions
Summary:
Thoroughly check for a pointer dereference which yields a glvalue. Look
through casts, comma operators, conditional operators, paren
expressions, etc.
Recommit: Handle diagnostic warnings in Frontend diagnostic handler.
Clang uses a diagnostic handler to grab diagnostic messages so it can print them
with the line of source code they refer to. This patch extends this to handle
optimization failures that were added to llvm to produce a warning when
loop vectorization is explicitly specified (using a pragma clang loop directive)
but fails.
Update renames warning flag name to avoid indicating the flag's severity and
adds a test.
This is breaking the system modules on Darwin, because something that
was defined and re-exported no longer is. Might be this patch, or might
just be a really poor interaction with an existing visibility bug.
Make sure globals created by UBSan are not instrumented by ASan.
Summary:
This change adds description of globals created by UBSan
instrumentation (UBSan handlers, type descriptors, filenames) to
llvm.asan.globals metadata, effectively "blacklisting" them. This can
dramatically decrease the data section in binaries built with UBSan+ASan,
as UBSan tends to create a lot of handlers, and ASan instrumentation
increases the global size to at least 64 bytes.
For i386, gcc 4.9 under Debian (at least) updated its paths to i586-linux-gnu
(i486-linux-gnu previously). Adding this triple to the list of search.
Also impacts clang 3.4.2
Reported on the Debian bug tracking system here:
http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=754963
http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=755183
Hal Finkel [Fri, 18 Jul 2014 15:52:10 +0000 (15:52 +0000)]
Mark C++ reference parameters as dereferenceable
Because references must be initialized using some evaluated expression, they
must point to something, and a callee can assume the reference parameter is
dereferenceable. Taking advantage of a new attribute just added to LLVM, mark
them as such.
Because dereferenceability in addrspace(0) implies nonnull in the backend, we
don't need both attributes. However, we need to know the size of the object to
use the dereferenceable attribute, so for incomplete types we still emit only
nonnull.
Arthur Marble [Fri, 18 Jul 2014 11:38:58 +0000 (11:38 +0000)]
Moved 25 flags to clang_ignored_gcc_optimization_f_Group. Will display a warning
to the user if they try to pass those optimizations. The revision for this patch
is here: http://reviews.llvm.org/D4474. This patch will fix many errors in the
rebuild of Debian with clang. Here is a link to the page for unknown arguments:
http://clang.debian.net/status.php?version=3.4.2&key=UNKNOWN_ARG
Oliver Stannard [Fri, 18 Jul 2014 09:09:31 +0000 (09:09 +0000)]
[ARM] Fix AAPCS regression caused by r211898
r211898 introduced a regression where a large struct, which would
normally be passed ByVal, was causing padding to be inserted to
prevent the backend from using some GPRs, in order to follow the
AAPCS. However, the type of the argument was not being set correctly,
so the backend cannot align 8-byte aligned struct types on the stack.
The fix is to not insert the padding arguments when the argument is
being passed ByVal.
Kevin Qin [Fri, 18 Jul 2014 07:03:22 +0000 (07:03 +0000)]
[AArch64] Implement Clang CLI interface proposal about "-march".
1. Revert "Add default feature for CPUs on AArch64 target in Clang"
at r210625. Then, all enabled feature will by passed explicitly by
-target-feature in -cc1 option.
2. Get "-mfpu" deprecated.
3. Implement support of "-march". Usage is:
-march=armv8-a+[no]feature
For instance, "-march=armv8-a+neon+crc+nocrypto". Here "armv8-a" is
necessary, and CPU names are not acceptable. Candidate features are
fp, neon, crc and crypto. Where conflicting feature modifiers are
specified, the right-most feature is used.
4. Implement support of "-mtune". Usage is:
-march=CPU_NAME
For instance, "-march=cortex-a57". This option will ONLY get
micro-architectural feature enabled specifying to target CPU,
like "+zcm" and "+zcz" for cyclone. Any architectural features
WON'T be modified.
5. Change usage of "-mcpu" to "-mcpu=CPU_NAME+[no]feature", which is
an alias to "-march={feature of CPU_NAME}+[no]feature" and
"-mtune=CPU_NAME" together. Where this option is used in conjunction
with -march or -mtune, those options take precedence over the
appropriate part of this option.
Richard Smith [Fri, 18 Jul 2014 04:53:37 +0000 (04:53 +0000)]
[modules] Fix macro hiding bug exposed if:
* A submodule of module A is imported into module B
* Another submodule of module A that is not imported into B exports a macro
* Some submodule of module B also exports a definition of the macro, and
happens to be the first submodule of B that imports module A.
In this case, we would incorrectly determine that A's macro redefines B's
macro, and so we don't need to re-export B's macro at all.
This happens with the 'assert' macro in an LLVM self-host. =(
Mark the vtable used when defining implicit copy and move ctors
I don't think other implicit members like copy assignment and move
assignment require this treatment, because they should already be
operating on a constructed object.
Fix parsing certain kinds of strings in the MS section pragmas
We were crashing on the relevant test case inputs. Also, refactor this
code a bit so we can report failure and slurp the pragma tokens without
returning a diagnostic id. This is more consistent with the rest of the
parser and sema code.
If, during the initial parse of a template, we perform aggregate initialization
and form an implicit value initialization for an array type, then when we come
to instantiate the template and redo the initialization step, we would try to
match the implicit value initialization up against an array *element*, not to
the complete array.
Remarkably, we've had this bug since ~the dawn of time, but only noticed it
recently.
David Blaikie [Thu, 17 Jul 2014 20:40:36 +0000 (20:40 +0000)]
unique_ptr-ify ownership of ASTConsumers
(after fixing a bug in MultiplexConsumer I noticed the ownership of the
nested consumers was implemented with raw pointers - so this fixes
that... and follows the source back to its origin pushing unique_ptr
ownership up through there too)
If char/short are shorter than int, do not use U as suffix for
constants. Comparing int against a constant of the given type like
UINT8_MAX will otherwise force a promotion to unsigned int, which is
typically not expected.
clang-cl: Flush stdout after writing the /showIncludes output
Summary:
Before this patch, you could get lines in the output such as:
Note: including file: ../../dist/include/js/Tc:/path/to/foo.cpp(1,1) : error(clang): static_assert failed...
This patch ensures that the stdout output from showIncludes won't be garbled
in the terminal like this, and it also helps applications that use the output
to generate dependency information if they happen to capture both stdout and
stderr.
Test Plan:
Tested locally, it's hard to write an automated test for this as
the behavior depends on the buffering of the ostreams.
[UBSan] Add !nosanitize metadata to the code generated by UBSan.
This is used to mark the instructions emitted by Clang to implement
variety of UBSan checks. Generally, we don't want to instrument these
instructions with another sanitizers (like ASan).
Objective-C. deprecated attribute is not inherited on methods
overriden in interfaces and protocols (this is already the case
for properties). rdar://16068470
Hal Finkel [Thu, 17 Jul 2014 14:49:58 +0000 (14:49 +0000)]
Handle __assume in the VoidExprEvaluator
This is a follow-up to an IRC conversation with Richard last night; __assume
does not evaluate its argument, and so the argument should not contribute to
whether (__assume(e), constant) can be used where a constant is required.
Hal Finkel [Thu, 17 Jul 2014 14:25:55 +0000 (14:25 +0000)]
Add an __assume side-effects warning
In MS-compatibility mode, we support the __assume builtin. The __assume builtin
does not evaluate its arguments, and we should issue a warning if __assume is
provided with an argument with side effects (because these effects will be
discarded).
This is similar in spirit to the warnings issued by other compilers (Intel
Diagnostic 2261, MS Compiler Warning C4557).
Using a std::string instead of a StringRef because the Default case synthesizes a temporary std::string from a Twine. Assigning that into a StringRef causes the StringRef to refer to a temporary, and bad things happen.
Upstream an MS inline assembly test from Mozilla's inline assembly code
Summary:
I'm planning on upstreaming some test cases for the inline assembly
usage in the Mozilla code base. A lot of these test cases test the
recent fixes to this code.
Daniel Sanders [Thu, 17 Jul 2014 09:46:40 +0000 (09:46 +0000)]
[mips] Pass the ABI to the integrated assembler and add tests the existing arguments.
Summary:
With this patch (and a corresponding LLVM patch), assembling an empty file with
GCC and Clang -fintegrated-as produce near identical objects. The remaining
differences are:
* GCC/GAS produce objects have a .pdr section
* GCC/GAS produce objects have a .gnu.attributes section
Other differences are insignificant such as precise file offsets and the order
of strings in the string table.
Richard Smith [Thu, 17 Jul 2014 05:12:35 +0000 (05:12 +0000)]
Track the difference between
-- a constructor list initialization that unpacked an initializer list into
constructor arguments and
-- a list initialization that created as std::initializer_list and passed it
as the first argument to a constructor
in the AST. Use this flag while instantiating templates to provide the right
semantics for the resulting initialization.
David Blaikie [Wed, 16 Jul 2014 23:52:46 +0000 (23:52 +0000)]
DebugInfo: Forward HandleTagDeclRequiredDefinition through MultiplexConsumer to fix debug info emission in the presence of plugins.
When plugins are used the Multiplex(AST)Consumer is employed to dispatch
to both the plugin ASTConsumers and the IRGen ASTConsumer. It wasn't
dispatching a critical call for debug info, resulting in plugin users
having a negative debugging experience.
While I'm here, forward a bunch of other missing calls through the
consumer that seem like they should be there.
To test this, use the example plugin (requires plugins and examples) and
split the test case up so that the plugin testing can be done under that
requirement while the non-plugin testing will execute even in builds
that don't include plugin support or examples.
David Blaikie [Wed, 16 Jul 2014 23:25:44 +0000 (23:25 +0000)]
DebugInfo: Ensure the ASTConsumer "HandleTagDeclRequireDefinition" callback path is tested.
By having the two variables 'a' and 'b' in this test in a namespace, the
type was required to be complete before any debug info was ever emitted
(the entire namespace is parsed before the variables were emitted), this
meant that the codepath in which a declaration is emitted, then later on
the type is required to be complete and the debug info must be upgraded
to a definition was not used.
Moving the variables outside a namespace fixes this test coverage bug.
(interestingly, code coverage didn't help here -
HandleTagDeclRequireDefinition is fully covered because it's called even
in cases where the type hasn't been emitted for debug info at all
(further down in CGDebugInfo this no-ops) - so CC wouldn't've helped
catch this test coverage problem)
Hal Finkel [Wed, 16 Jul 2014 22:44:54 +0000 (22:44 +0000)]
Add basic (noop) CodeGen support for __assume
Clang supports __assume, at least at the semantic level, when MS extensions are
enabled. Unfortunately, trying to actually compile code using __assume would
result in this error:
error: cannot compile this builtin function yet
__assume is an optimizer hint, and can be ignored at the IR level. Until LLVM
supports assumptions at the IR level, a noop lowering is valid, and that is
what is done here.
Richard Smith [Wed, 16 Jul 2014 21:33:43 +0000 (21:33 +0000)]
When list-initializing an object of class type, if we pick an initializer list
constructor (and pass it an implicitly-generated std::initializer_list object),
be sure to mark the resulting construction as list-initialization. This fixes
an assert in template instantiation where we previously thought we'd got direct
non-list initialization without any parentheses.
Arthur Marble [Wed, 16 Jul 2014 21:16:16 +0000 (21:16 +0000)]
Typically linker options are protected with -Xlinker or -Wl,
however certain sloppy Makefiles pass -z options directly to
the compiler. This patch enables clang to recognize these
options (because -z is not used by clang itself).
Specifying the diagnostic argument through the attribute table generator instead of having to enter it manually as part of the attribute subject list. This only affects attributes appertaining to ObjC interfaces and protocols.
No new tests required as this is covered by existing tests.
Removing a FIXME from the attribute parsing code by now passing along the scope and syntax information for attributes with custom parsing. It turns out not to matter too much because the FIXME wasn't quite true -- none of these attributes have a C++11 spelling. However, it's still a good change (for instance, we may add an attribute with a type arg in the future for which this code now behaves properly).