Nico Weber [Tue, 3 Feb 2015 17:06:08 +0000 (17:06 +0000)]
Implement jump scope SEHmantic analysis.
Thou shall not jump into SEH blocks. Jumping out of SEH __try and __excepts
is A-ok. Jumping out of __finally blocks is B-ok (msvc doesn't error about it,
but warns that it has undefined behavior).
I've checked that clang's behavior with this patch matches msvc's behavior.
We don't have the warning on jumping out of a __finally yet, see the FIXME
in the test. clang also currently crashes on codegen for a jump out of a
__finally block, see PR22414 comment 7.
I also added a few tests for the interaction of indirect jumps and SEH blocks.
MSVC doesn't support indirect jumps, so there's no way to know if clang behave
the same way as msvc here. clang's behavior with this patch does make sense
to me, but maybe it could be argued that it should be more permissive (see
FIXME in the indirect jump tests -- shout if you have an opinion on this).
Daniel Jasper [Tue, 3 Feb 2015 09:45:52 +0000 (09:45 +0000)]
Add some overloads so that floating point literals can be AST matched properly.
I am not entirely sure whether the implemented sematics are ideal. In
particular, should floatLiteral(equals(0.5)) match "0.5f" and should
floatLiteral(equals(0.5f)) match "0.5". With the overloads in this
patch, the answer to both questions is yes, but I am happy to change
that.
Richard Smith [Tue, 3 Feb 2015 03:32:14 +0000 (03:32 +0000)]
[modules] Be sure to load the lexical definition of a class template
specialization from an update record exactly once, even if we needed to fake up
the definition.
Daniel Jasper [Mon, 2 Feb 2015 23:04:00 +0000 (23:04 +0000)]
Add some overloads so that floating point literals can be AST matched properly.
I am not entirely sure whether the implemented sematics are ideal. In
particular, should floatLiteral(equals(0.5)) match "0.5f" and should
floatLiteral(equals(0.5f)) match "0.5". With the overloads in this
patch, the answer to both questions is yes, but I am happy to change
that.
-save-temps=cwd is equivalent to -save-temps
-save-temps=obj saves temporary file in the same directory as output
This helps to avoid clobbering of temp files in case of parallel
compilation with -save-temps of the files that have the same name
but located in different directories.
Reid Kleckner [Mon, 2 Feb 2015 22:15:31 +0000 (22:15 +0000)]
SEH: Diagnose use of C++ EH and SEH in the same function
This check does not apply when Borland extensions are enabled, as they
have a checked in test case indicating that mixed usage of SEH and C++
is supported.
When generating llvm.used, we may need an addrspacecast instead of a bitcast.
Summary:
This is especially important for targets that use multiple address spaces,
and commonly place global variables in address spaces other than zero.
David Majnemer [Mon, 2 Feb 2015 10:22:20 +0000 (10:22 +0000)]
MS ABI: Implement support for 'novtable'
It is common for COM interface classes to be marked as 'novtable' to
tell the compiler that constructors and destructors should not reference
virtual function tables.
Nico Weber [Mon, 2 Feb 2015 05:33:50 +0000 (05:33 +0000)]
Follow-up to r217302: Don't crash on ~A::A in a postfix expr suffix followed by '<'.
This used to crash, complaining "ObjectType and scope specifier cannot coexist":
struct A { } b = b.~A::A <int>;
The only other caller of ParseOptionalCXXScopeSpecifier() that passes in a
non-empty ObjectType clears the ObjectType of the scope specifier comes back
non-empty (see the tok::period case in Parser::ParsePostfixExpressionSuffix()),
so do that here too.
[multiversion] Update Clang for the API change in LLVM r227731.
This moves all of the PassManager <-> Target communication to use the
new pass manager's TargetIRAnalysis even with the old pass manager. See
the LLVM commit for some of why things are moving in this direction, but
the short version is that this will enable us to create per-function
TargetTransformInfo objects that have correct subtarget information for
that function.
Craig Topper [Sun, 1 Feb 2015 07:35:40 +0000 (07:35 +0000)]
[X86] Rename _mm512_valign_epi64/32 intrinsics to _mm512_alignr_epi64/32 to match Intel docs. Make immediate argument to them an ICE. Fix mask size for the alignd version.
Create a new TargetCodeGenInfo for Windows on ARM to permit annotating the
functions with stack-probe-size (for /Gs and -mstack-probe-support) for
generating the stack probe necessary for Windows targets. This will be used by
the backend when lowering the frame to generate the stack probe appropriately.
Nico Weber [Fri, 30 Jan 2015 16:53:11 +0000 (16:53 +0000)]
Follow-up to r217302 and r227555: Don't crash on inline ~A::A() if A is an int.
Even with r227555, this still crashed:
struct S {
int A;
~A::A() {}
};
That's because ParseOptionalCXXScopeSpecifier()'s call to
ActOnCXXNestedNameSpecifier() doesn't mark the scope spec as invalid if sema
thought it's a good idea to fixit-correct "::" to ":". For the diagnostic
improvement done in r217302, we never want :: to be interpreted as :, so fix
this by setting ColonSacred to false temporarily.
Fraser Cormack [Fri, 30 Jan 2015 10:51:46 +0000 (10:51 +0000)]
Fix OpenCL 1.2 double as an optional core feature behaviour
In OpenCL 1.2, using double no longer requires using the pragma cl_khr_fp64,
instead a kernel is allowed to use double, but must first have queried
clGetDeviceInfo's CL_DEVICE_DOUBLE_FP_CONFIG.
Page 197, section 6.1.1 of the OpenCL 1.2 specification has a footnote 23
describing this behaviour.
I've also added test cases such that the pragma must be used if targeting
OpenCL 1.0 or 1.1, but is ignored in 1.2 and 2.0.
David Majnemer [Fri, 30 Jan 2015 05:01:23 +0000 (05:01 +0000)]
Sema: Disable template instantiation once a fatal error is raised
Fatal errors disable all further diagnostics. Continuing to permit
template instantiation does nothing but make it take longer for clang to
finish with the TU.
Ulrich Weigand [Thu, 29 Jan 2015 19:08:51 +0000 (19:08 +0000)]
[PowerPC] Work around TLS linker bug
Work around a bug in GNU ld (and gold) linker versions up to 2.25
that may mis-optimize code generated by this version of clang/LLVM
to access general-dynamic or local-dynamic TLS variables.
Bug is fixed here:
https://sourceware.org/ml/binutils/2015-01/msg00318.html
Rafael Espindola [Thu, 29 Jan 2015 17:22:53 +0000 (17:22 +0000)]
This reverts commit r227432, r227438 and r227448.
It should bring the bots back.
Original messagses:
r227448:
Remove unnecessary default.
r227438:
Fix Index/print-type.cpp test following r227432.
r227432:
libclang: Add three functions useful for dealing with anonymous fields:
clang_Cursor_getOffsetOfField
clang_Cursor_isAnonymous
clang_Type_visitFields
Python: Add corresponding methods for dealing with anonymous fields.
Aaron Ballman [Thu, 29 Jan 2015 16:58:53 +0000 (16:58 +0000)]
Reverting r227453, which adds back the fuzzer library. Now excluding the clang-format fuzzer functionality based on LLVM_USE_SANITIZE_COVERAGE being set or unset.
Francois Pichet [Thu, 29 Jan 2015 12:45:29 +0000 (12:45 +0000)]
libclang: Add three functions useful for dealing with anonymous fields:
clang_Cursor_getOffsetOfField
clang_Cursor_isAnonymous
clang_Type_visitFields
Python: Add corresponding methods for dealing with anonymous fields.
David Majnemer [Thu, 29 Jan 2015 09:29:21 +0000 (09:29 +0000)]
MS ABI: Implement proper support for setjmp
On targets which use the MSVCRT, setjmp is a macro which expands to
_setjmp or _setjmpex.
_setjmp and _setjmpex have a secret, hidden argument which is not listed
in the function prototype on X64 and WoA. This hidden argument always
seems to be the frame pointer.
_setjmpex isn't used on X86, _setjmp is magically replaced with a call
to _setjmp3. The second argument is zero for 'normal' setjmp/longjmp
pairs, otherwise it is a count of additional variadic arguments. This
is used when setjmp appears inside of a try or __try.
It is not safe to use a pointer to setjmp because _setjmp, _setjmpex and
_setmp3 are not compatible with setjmp.
For the --be8 flag, check explicitly for pre-v7 / pre-v6m cores.
Those used the old Big Endian support on ARM and don't need flags.
Refactor the logic in a separate common function, which also looks at
-march. Add corresponding logic for the Linux toolchain.
Richard Smith [Wed, 28 Jan 2015 22:06:01 +0000 (22:06 +0000)]
PR22367: Don't forget to create a CXXFunctionalCastExpr around
list-initialization that gets converted to some form other than an
InitListExpr. CXXTemporaryObjectExpr is a special case here, because it
represents a fused CXXFunctionalCastExpr + CXXConstructExpr. That, in
itself, is probably a design error...
Kaelyn Takata [Wed, 28 Jan 2015 21:10:46 +0000 (21:10 +0000)]
Revert a change from r222797 that is no longer needed and can cause
infinite recursion.
Also guard against said infinite recursion by adding an assert that will
trigger if CorrectDelayedTyposInExpr is called before a previous call to
CorrectDelayedTyposInExpr returns (i.e. if the TreeTransform run by
CorrectDelayedTyposInExpr calls a sequence of methods that
end up calling CorrectDelayedTyposInExpr, as the new test case had done
prior to this commit). Fixes PR22292.
Derek Schuff [Wed, 28 Jan 2015 20:24:52 +0000 (20:24 +0000)]
Remove support for pnaclcall attribute
Summary:
It was used for interoperability with PNaCl's calling conventions, but
it's no longer needed.
Also Remove NaCl*ABIInfo which just existed to delegate to either the portable
or native ABIInfo, and remove checkCallingConvention which was now a no-op
override.
David Blaikie [Wed, 28 Jan 2015 19:50:09 +0000 (19:50 +0000)]
DebugInfo: Attribute implicit boolean tests to the expression being tested, not to the outer use of that expression.
This is half a fix for a GDB test suite failure that expects to start at
'a' in the following code:
void func(int a)
if (a
&&
b)
...
But instead, without this change, the comparison was assigned to '&&'
(well, worse actually - because there was a chained 'a && b && c' and it
was assigned to the second '&&' because of a recursive application of
this bug) and then the load folded into the comparison so breaking on
the function started at '&&' instead of 'a'.
The other part of this needs to be fixed in LLVM where it's ignoring the
location of the icmp and instead using the location of the branch
instruction.
The fix to the conditional operator is actually a no-op currently,
because the conditional operator's location coincides with 'a' (the
start of the conditional expression) but should probably be '?' instead.
See the FIXME in the test case that mentions the ARCMigration tool
failures when I tried to make that change.
projects / clang / log