Daniel Dunbar [Fri, 19 Oct 2012 20:10:10 +0000 (20:10 +0000)]
IRgen: Initialize TargetLoweringInfo with a triple.
- We create two TargetLoweringInfo instances for different pass managers, and
they weren't consistent (the one for codegen didn't have the right info). I'm
not sure this mattered anywhere in practice.
Jordan Rose [Fri, 19 Oct 2012 16:05:26 +0000 (16:05 +0000)]
Allow objc_requires_super to be used to check class methods as well.
Also, unify ObjCShouldCallSuperDealloc and ObjCShouldCallSuperFinalize.
The two have identical behavior and will never be active at the same time.
There's one last simplification now, which is that if we see a call to
[super foo] and we are currently in a method named 'foo', we will
/unconditionally/ clear the ObjCShouldCallSuper flag, rather than check
first to see if we're in a method where calling super is required. There's
no reason to pay the extra lookup price here.
Andy Gibbs [Fri, 19 Oct 2012 12:49:32 +0000 (12:49 +0000)]
Change VerifyDiagnosticConsumer so that it *must* contain at least one "expected-*" directive. As a result, for test-cases that are not expected to generate any diagnostics, an additional directive "expected-no-diagnostics" has been implemented which can then be included in such test-cases. This new directive may not be used in conjunction with any other "expected-*" directive.
This change was initially proposed as a solution to the problem highlighted by check-in r164677, i.e. that -verify will not cause a test-case failure where the compile command does not actually reference the file.
Andy Gibbs [Fri, 19 Oct 2012 12:44:48 +0000 (12:44 +0000)]
Prior to adding the new "expected-no-diagnostics" directive to VerifyDiagnosticConsumer, make the necessary adjustment to 580 test-cases which will henceforth require this new directive.
Andy Gibbs [Fri, 19 Oct 2012 12:36:49 +0000 (12:36 +0000)]
Fix directive parsing in VerifyDiagnosticConsumer so that it ensures that "expected" is at the start of the word and will no longer accept typos such as "junkexpected-*" as a valid "expected-*" directive. A very few test-cases had to be amended to adhere to the new rule.
Richard Smith [Fri, 19 Oct 2012 06:32:17 +0000 (06:32 +0000)]
PR14124: When performing template instantiation of a qualified-id outside of a
class, diagnose if the qualified-id instantiates to a non-static class member.
Rafael Espindola [Fri, 19 Oct 2012 05:04:37 +0000 (05:04 +0000)]
Fix handling of the regparm attribute in the presence of classes with copy
constructors.
When I first moved regparm support to TargetInfo.cpp I tried to isolate it
in classifyArgumentTypeWithReg, but it is actually a lot easier to flip the
code around and check for regparm at the end of the decision tree.
Without this refactoring classifyArgumentTypeWithReg would have to duplicate
the logic about when to use non-byval indirect arguments.
Craig Topper [Fri, 19 Oct 2012 04:40:38 +0000 (04:40 +0000)]
Teach getColumnNumber to use the line cache to get the start of the line if its on the same line as the last call to getLineNumber. Prevents needing to scan backwards for the new line. Fixes PR14106.
Douglas Gregor [Fri, 19 Oct 2012 00:38:02 +0000 (00:38 +0000)]
Move the set of files to be validated in an AST file into the control
block, so the input files are validated early on, before we've
committed to loading the AST file. This (accidentally) fixed a but
wherein the main file used to generate the AST file would *not* be
validated by the existing validation logic.
At the moment, this leads to some duplication of filenames between the
source manager block and input-file blocks, as well as validation
logic. This will be handled via an upcoming patch.
Eric Christopher [Thu, 18 Oct 2012 21:52:18 +0000 (21:52 +0000)]
Add a new option for and disable column number information as there
are no known current users of column info. Robustify and fix up
a few tests in the process. Reduces the size of debug information
by a small amount.
[doc parsing] use getParamName to access parameter
for current(rewritten) comment and getParamNameAsWritten
to access param name coming with \param marker.
Eli Friedman [Thu, 18 Oct 2012 20:14:08 +0000 (20:14 +0000)]
Fix Objective-C implicit property synthesis for C++ classes so we use valid
source locations in places where it is necessary for diagnostics. By itself,
this causes assertions, so while I'm here, also fix property synthesis
for properties of C++ class type so we use so we properly set up a scope
and mark variable declarations.
Chad Rosier [Thu, 18 Oct 2012 19:39:37 +0000 (19:39 +0000)]
[ms-inline asm] Have the LookupInlineAsmIdentifier() callback function return a
*NamedDecl. In turn, build the expressions after we're finished parsing the
asm. This avoids a crasher if the lookup fails.
Douglas Gregor [Thu, 18 Oct 2012 18:27:37 +0000 (18:27 +0000)]
Collapse the "version control revision/tag" AST file record into the
metadata record, which already had other version information. Clean up
the block info block along the way.
Chad Rosier [Thu, 18 Oct 2012 15:49:40 +0000 (15:49 +0000)]
[ms-inline asm] Move most of the AsmParsing logic in clang back into the MC
layer. Use the new ParseMSInlineAsm() API and add an implementation of the
MCAsmParserSemaCallback interface.
Douglas Gregor [Thu, 18 Oct 2012 05:31:06 +0000 (05:31 +0000)]
Start factoring the on-disk records for an AST file into a control
block, which stores information about how the AST file to generated,
from the AST block, which stores the actual serialized AST. The
information in the control block should be enough to determine whether
the AST file is up-to-date and compatible with the current translation
unit, and reading it should not cause any side effects that aren't
easy to undo. That way, we can back out from an attempt to read an
incompatible or out-of-date AST file.
Note that there is still more factoring to do. In particular,
information about the source files used to generate the AST file
(along with their time stamps, sizes, etc.) still resides in the
source manager block.
David Blaikie [Wed, 17 Oct 2012 23:45:39 +0000 (23:45 +0000)]
Fix -Woverloaded-virtual when the using statement refers to a base declaration of a virtual function.
GCC and Clang both do not warn on:
struct a { virtual void func(); };
struct b: a { virtual void func(); void func(int); };
struct c: b { void func(int); using b::func; };
but if the "using" was using a::func GCC would still remain silent where Clang
would warn. This change makes Clang consistent with GCC's existing behavior.
[Doc parsing]: This patch adds <Declaration> tag to
XML comment for declarations which pretty-prints
declaration. I had to XFAIL one test annotate-comments.cpp.
This test is currently unmaintainable as written.
Dmitri G., can you see what we can do about this test.
We should change this test such that adding a new tag does not wreck
havoc to the test.
Jordan Rose [Wed, 17 Oct 2012 19:35:44 +0000 (19:35 +0000)]
[analyzer] When binding to a ParenExpr, bind to its inner expression instead.
This actually looks through several kinds of expression, such as
OpaqueValueExpr and ExprWithCleanups. The idea is that binding and lookup
should be consistent, and so if the environment needs to be modified later,
the code doing the modification will not have to manually look through these
"transparent" expressions to find the real binding to change.
This is necessary for proper updating of struct rvalues as described in
the previous commit.
Jordan Rose [Wed, 17 Oct 2012 19:35:37 +0000 (19:35 +0000)]
[analyzer] Create a temporary region when accessing a struct rvalue.
In C++, rvalues that need to have their address taken (for example, to be
passed to a function by const reference) will be wrapped in a
MaterializeTemporaryExpr, which lets CodeGen know to create a temporary
region to store this value. However, MaterializeTemporaryExprs are /not/
created when a method is called on an rvalue struct, even though the 'this'
pointer needs a valid value. CodeGen works around this by creating a
temporary region anyway; now, so does the analyzer.
The analyzer also does this when accessing a field of a struct rvalue.
This is a little unfortunate, since the rest of the struct will soon be
thrown away, but it does make things consistent with the rest of the
analyzer.
This allows us to bring back the assumption that all known 'this' values
are Locs. This is a revised version of r164828-9, reverted in r164876-7.
John McCall [Wed, 17 Oct 2012 04:53:31 +0000 (04:53 +0000)]
Set a special flag in class metadata when an Objective-C class
has ivars that require destruction, but none that require anything
except zero-initialization. This is common in ARC and (when true
throughout a class hierarchy) permits the elimination of an
unnecessary message-send during allocation.
John McCall [Wed, 17 Oct 2012 02:28:37 +0000 (02:28 +0000)]
At -O0, prefer objc_storeStrong with a null new value to the
combination of a load+objc_release; this is generally better
for tools that try to track why values are retained and
released. Also use objc_storeStrong when copying a block
(again, only at -O0), which requires us to do a preliminary
store of null in order to compensate for objc_storeStrong's
assign semantics.
Douglas Gregor [Tue, 16 Oct 2012 23:40:58 +0000 (23:40 +0000)]
Serialize TargetOptions into an AST file, and make sure that we keep
target options around so they can be accessed at any point (rather
than keeping them transient).
David Blaikie [Tue, 16 Oct 2012 22:56:05 +0000 (22:56 +0000)]
PR13684: Emit vtable entries for deleted functions as __cxa_deleted_function.
This is consistent/interoperable with GCC 4.7 (& __cxa_deleted_function isn't
present in 4.4 - not sure when it got added, but you'll need something with
that function available for this to work).
Derek Schuff [Tue, 16 Oct 2012 22:30:41 +0000 (22:30 +0000)]
Add pnaclcall convention to Native Client targets.
Because PNaCl bitcode must be target-independent, it uses some
different bitcode representations from other targets (e.g. byval and
sret for structures). This means that without additional type
information, it cannot meet some native ABI requirements for some
targets (e.g. passing structures containing unions by value on
x86-64). To allow generation of code which uses the correct native
ABIs, we also support triples such as x86_64-nacl, which uses
target-dependent IR (as opposed to le32-nacl, which uses byval and
sret).
To allow interoperation between the two types of code, this patch adds
a calling convention attribute to be used in code compiled with the
target-dependent triple, which will generate code using the le32-style
bitcode. This calling convention does not need to be explicitly
supported in the backend because it determines bitcode representation
rather than native conventions (the backend just needs to undersand
how to handle byval and sret for the Native Client OS).
This patch implements __attribute__((pnaclcall)) to generate calls in
bitcode according to the le32 bitcode conventions, an attribute which
is accepted by any Native Client target, but issues a warning
otherwise.
projects / clang / log