Commit Graph

3382 Commits

Author SHA1 Message Date
David Majnemer 9f77e90dcc Update testcase to match r236539
llvm-svn: 236542
2015-05-05 20:34:29 +00:00
David Majnemer ab1309252c [MS ABI] Fix a crash in vptr path calculation
I discovered a case where the old algorithm would crash.  Instead of
trying to patch the algorithm, rewrite it.  The new algorithm operates
in three phases:
1. Find all paths to the subobject with the vptr.
2. Remove paths which are subsets of other paths.
3. Select the best path where 'best' is defined as introducing the most
   covariant overriders.  If two paths introduce different overriders,
   raise a diagnostic.

llvm-svn: 236444
2015-05-04 18:47:54 +00:00
David Majnemer cfcac9f64a [MS ABI] Add testcase which was missing from r236354
This test verifies that we can detect when the inheritance paths are
ambiguous due to covariant thunks, ensuring we don't regress PR16759.

llvm-svn: 236381
2015-05-02 05:12:53 +00:00
David Majnemer ead97576a9 [MS ABI] NV bases may indirectly contain covariant thunks from V Bases
A class might contain multiple ways of getting to a vbase, some of which
are virtual and other non-virtual.  It may be the case that a
non-virtual base contains an override of a method in a vbase.  This
means that we must carefully pick between a set of nvbases to determine
which is the best.

As a consequence, the findPathForVPtr algorithm is considerably simpler.

llvm-svn: 236353
2015-05-01 21:35:41 +00:00
Richard Smith e831690957 PR22354: Add a test that we emit copy-ctors for captures in generic lambdas.
The bug was fixed by r235921.

llvm-svn: 236254
2015-04-30 20:51:03 +00:00
David Majnemer d950f15ee5 [MS ABI] Correctly make paths through covariant virtual bases
There can be multiple virtual bases which are on the path to a vfptr
when one vbase virtually inherits from another.  We should prefer the
most derived virtual base which covariantly overrides a method in the
vfptr class;  if we do not lengthen the path this way, we will end up
with too few vftable entries.

This fixes PR21073.

llvm-svn: 236239
2015-04-30 17:15:48 +00:00
Richard Smith 13afe197a0 Remove XFAIL now that this test passes (fixed by r236184).
llvm-svn: 236194
2015-04-30 01:15:19 +00:00
NAKAMURA Takumi 83a758a19a Update clang/test/CodeGenCXX/copy-constructor-synthesis-2.cpp to pass for targeting i686.
r236155 missed the suffix in "@llvm.memcpy.p0i8.p0i8.i32".

llvm-svn: 236184
2015-04-29 23:55:46 +00:00
Rick Foos 2688a49981 XFAIL Hexagon until more codegen in place.
Summary:
Hexagon is being updated, but there is not enough to pass these tests.
These sections are now on top of Colin's list.

Test Plan: Ran changes on hexagon-build-03.

Reviewers: colinl, rfoos

Reviewed By: rfoos

Subscribers: cfe-commits

Differential Revision: http://reviews.llvm.org/D9356

llvm-svn: 236173
2015-04-29 22:54:40 +00:00
Richard Smith 0ae6b2b839 Stop assuming a 64-bit target here.
llvm-svn: 236155
2015-04-29 20:49:22 +00:00
Richard Smith 419bd09415 PR23373: A defaulted union copy constructor that is not trivial must still be
emitted as a memcpy.

llvm-svn: 236142
2015-04-29 19:26:57 +00:00
Adrian Prantl 00820abf02 Debug Info: Represent local anonymous unions as anonymous unions
and as artificial local variables in the debug info.

This is a follow-up to r236059. We can't get rid of the local variables
entirely because the gdb buildbot depends on them, but we can mark them
as artificial while still emitting the correct debug info. As I learned
from review comments other compilers also follow this model.

A paired commit in LLVM temporarily relaxes the debug info verifier to
not check the integrity of DW_OP_bit_pieces of artificial variables.

rdar://problem/20730771

llvm-svn: 236125
2015-04-29 16:52:31 +00:00
Duncan P. N. Exon Smith 9dd4e4e63a DebugInfo: Metadata constructs now start with DI*
LLVM r236120 renamed debug info IR constructs to use a `DI` prefix, now
that the `DIDescriptor` hierarchy has been gone for about a week.  This
commit was generated using the rename-md-di-nodes.sh upgrade script
attached to PR23080, followed by running clang-format-diff.py on the
`lib/` portion of the patch.

llvm-svn: 236121
2015-04-29 16:40:08 +00:00
Adrian Prantl 0d82089233 Revert "Debug Info: Represent local anonymous unions as anonymous unions"
This reverts commit r236059 as it breaks the gdb buildbot.

llvm-svn: 236110
2015-04-29 15:05:50 +00:00
Adrian Prantl 0a4ddc180b Debug Info: Represent local anonymous unions as anonymous unions
in the debug info. This patch deletes a hack that emits the members
of local anonymous unions as local variables.

Besides being morally wrong, the existing representation using local
variables breaks internal assumptions about the local variables' storage
size.

Compiling

```
   void fn1() {
     union {
       int i;
       char c;
     };
     i = c;
   }

```

with -g -O3 -verify will cause the verifier to fail after SROA splits
the 32-bit storage for the "local variable" c into two pieces because the
second piece is clearly outside the 8-bit range that is expected for a
variable of type char. Given the choice I'd rather fix the debug
representation than weaken the verifier.

Debuggers generally already know how to deal with anonymous unions when
they are members of C++ record types, but they may have problems finding
the local anonymous struct members in the expression evaluator.

rdar://problem/20730771

llvm-svn: 236059
2015-04-28 23:01:24 +00:00
Richard Smith 6d69e8c00b Fix assertion failure if a lambda array-capture is followed by a this capture.
llvm-svn: 236043
2015-04-28 21:41:14 +00:00
John McCall f3e86a7a55 ms_struct does not imply the MS base-layout ABI; separate these
conditions in the IRGen struct layout code.

rdar://20636558

llvm-svn: 235949
2015-04-28 00:17:18 +00:00
Reid Kleckner 4cb2dbde5b [MS ABI] Use 'continue' instead of 'return false' where intended
This was a bug in r218285 that prevented us from seeing subsequent
virtual bases in the class hierarchy, leading to crashes later.

Also add some comments to this function, now that we better understand
what it's trying to do.

Fixes PR21062 and PR21064.

llvm-svn: 235899
2015-04-27 17:19:49 +00:00
John McCall 9fc700e76d Correctly handle zero-sized but non-empty base classes in IRGen.
Fixes rdar://20621065.

A more elegant fix would preclude this case by defining the
rules such that zero-size classes are always formally empty.
I believe the only extensions which create zero-size classes
right now are flexible arrays and zero-length arrays; it's
not abstractly unreasonable to say that those don't count
as members for the purposes of emptiness, just as zero-width
bitfields don't count.  But that's an ABI-affecting change
and requires further discussion; in the meantime, let's not
assert / miscompile.

llvm-svn: 235815
2015-04-26 04:43:26 +00:00
David Majnemer e154456d4a [MS ABI] Fix the preferred alignment of member pointers
Member pointers in the MS ABI have different alignment depending on
whether they were created on the stack or live in a record.

llvm-svn: 235681
2015-04-24 01:25:05 +00:00
David Majnemer 9595b7da1a [MS ABI] Add support for mangling VLA types
Treat a VLA type like an incomplete array type.

llvm-svn: 235575
2015-04-23 07:42:08 +00:00
David Majnemer 96183f6b06 [MS ABI] Treat ConstantArrayType like IncompleteArrayType in args
Type backreferences for arguments use the DecayedType's original type.
Because of this, arguments with the same canonical type with the same
mangling would not backreference each other if one was a
ConstantArrayType while the other was an IncompleteArrayType.  Solve
this by canonicalizing the ConstantArrayType to a suitable
IncompleteArrayType.

This fixes PR23325.

llvm-svn: 235572
2015-04-23 05:21:20 +00:00
Reid Kleckner 7d4bc9c0fb [WinEH] Don't emit an exceptional cleanup for llvm.eh.endcatch
These extra endcatch markers aren't helping identify regions to outline,
so let's get rid of them.  LLVM outlines (more or less) from begincatch
to endcatch.  Any unwind edge from an enclosed invoke is a transition to
a new exception handler, which has it's own outlining markers.

llvm-svn: 235562
2015-04-22 23:39:15 +00:00
David Majnemer dc012fa266 Revert "Revert r234581, it might have caused a few miscompiles in Chromium."
This reverts commit r234700.  It turns out that the lifetime markers
were not the cause of Chromium failing but a bug which was uncovered by
optimizations exposed by the markers.

llvm-svn: 235553
2015-04-22 21:38:15 +00:00
Reid Kleckner 787dc438c9 Set normal LLVM function attributes on global initializer functions
Otherwise -fno-omit-frame-pointer and other flags like it aren't
applied.

Basic idea taken from Gao's patch, thanks!

Differential Revision: http://reviews.llvm.org/D9203

llvm-svn: 235537
2015-04-22 19:37:32 +00:00
Hans Wennborg 287231cb7a Don't dllimport/export class members with internal linkage (PR23308)
For example, a function taking a parameter with internal linkage will
itself have internal linkage since it cannot be called outside the
translation unit.

Differential Revision: http://reviews.llvm.org/D9182

llvm-svn: 235471
2015-04-22 04:05:17 +00:00
David Majnemer 2ccba83401 [MS ABI] Use the right types for filter and finally blocks
The type for abnormal_termination can't be an i1, it an i8.
Filter functions return 'LONG', not 'int'.

llvm-svn: 235161
2015-04-17 06:57:25 +00:00
David Blaikie d6c88ece21 [opaque pointer types] Explicit non-pointer type for call expressions
(migration for recent LLVM change to textual IR for calls)

llvm-svn: 235147
2015-04-16 23:25:00 +00:00
Renato Golin c640ff63e2 Revert "[CodeGen] Fix crash with duplicated mangled name."
This reverts commit r234767, as it was breaking all ARM buildbots for two days and the
assert is not in the code, making it difficult to spot the error, which would keep the
bots red for a few more days. New errors were silently introduced because of this bug,
and we don't want this to escalate.

llvm-svn: 234983
2015-04-15 08:44:40 +00:00
Reid Kleckner e07140eb3f Move the logic to avoid double global emission from Sema to CodeGen
Reverts the code changes from r234675 but keeps the test case.

We were already maintaining a DenseMap of globals with dynamic
initializers anyway.

Fixes the test case from PR23234.

llvm-svn: 234961
2015-04-15 01:08:06 +00:00
Reid Kleckner ebaf28d13d Reland r234613 (and follow-ups 234614, 234616, 234618)
The frameescape intrinsic cannot be inlined, so I fixed the inliner in
r234937. This should address PR23216.

llvm-svn: 234942
2015-04-14 20:59:00 +00:00
Nico Weber ad108337cf Revert r234613 (and follow-ups 234614, 234616, 234618), it caused PR23216.
llvm-svn: 234789
2015-04-13 20:04:22 +00:00
Nico Weber f2a39a7b4e Revert r234786, it contained a bunch of stuff I did not mean to commit.
llvm-svn: 234787
2015-04-13 20:03:03 +00:00
Nico Weber b31abb05fb Revert r234613 (and follow-ups 234614, 234616, 234618), it caused PR23216.
llvm-svn: 234786
2015-04-13 20:01:20 +00:00
Argyrios Kyrtzidis 6f1fadf989 [CodeGen] Fix crash with duplicated mangled name.
Patch by Yunzhong Gao!

llvm-svn: 234767
2015-04-13 17:40:46 +00:00
Nico Weber 1c565c31b1 Revert r234581, it might have caused a few miscompiles in Chromium.
If the revert helps, I'll get a repro this Monday.  Else I'll put the change
back in.

llvm-svn: 234700
2015-04-11 23:51:38 +00:00
NAKAMURA Takumi 7b24b0573d clang/test/CodeGenCXX/static-data-member-single-emission.cpp: Tweak for MS mangler.
llvm-svn: 234687
2015-04-11 14:57:11 +00:00
Reid Kleckner 620d0ce55c Only notify consumers about static data members of class templates once
llvm-svn: 234675
2015-04-11 01:25:36 +00:00
Reid Kleckner 5522d815eb [SEH] Add test for inline functions using SEH helpers
Such helpers should always be comdat with the parent function and have
internal linkage.

llvm-svn: 234618
2015-04-10 18:07:38 +00:00
Arnaud A. de Grandmaison 047a686d53 Remove threshold for inserting lifetime markers for named temporaries
Now that TailRecursionElimination has been fixed with r222354, the
threshold on size for lifetime marker insertion can be removed. This
only affects named temporary though, as the patch for unnamed temporaries
is still in progress.

My previous commit (r222993) was not handling debuginfo correctly, but
this could only be seen with some asan tests. Basically, lifetime markers
are just instrumentation for the compiler's usage and should not affect
debug information; however, the cleanup infrastructure was assuming it
contained only destructors, i.e. actual code to be executed, and was
setting the breakpoint for the end of the function to the closing '}', and
not the return statement, in order to show some destructors have been
called when leaving the function. This is wrong when the cleanups are only
lifetime markers, and this is now fixed.

llvm-svn: 234581
2015-04-10 10:13:52 +00:00
Benjamin Kramer 85be49d26d Update test case for r234543.
llvm-svn: 234544
2015-04-09 22:54:53 +00:00
Benjamin Kramer f3e67de85a [CodeGen] Do a more principled fix for PR231653, always use the inner type.
We were still using the MaterializeTemporaryExpr's type to check if the
transform is legal. Always use the inner Expr type.

llvm-svn: 234543
2015-04-09 22:50:07 +00:00
Benjamin Kramer b2b81439a3 [CodeGen] When promoting a reference temporary to a global use the inner type to fold it.
The MaterializeTemporaryExpr can have a different type than the inner
expression, miscompiling the constant. PR23165.

llvm-svn: 234499
2015-04-09 16:09:29 +00:00
Reid Kleckner e5b064205a [WinEH] Don't wrap cleanups in terminate actions
_CxxFrameHandler3 calls terminate if a cleanup action throws, regardless
of what bits you put in the xdata tables. There's no need to model this
in the IR, since we just have to take it out later.

llvm-svn: 234448
2015-04-08 22:48:50 +00:00
Reid Kleckner 31a1bb0c14 Reland "[SEH] Implement filter capturing in CodeGen"
The test should be fixed. It was failing in NDEBUG builds due to a
missing '*' character in a regex. In asserts builds, the pattern matched
a single digit value, which became a double digit value in NDEBUG
builds. Go figure.

This reverts commit r234261.

llvm-svn: 234447
2015-04-08 22:23:48 +00:00
Reid Kleckner 0f764e57fc Revert "Mark instantiated function decls as inline specified if any pattern is"
It breaks down on this test case:
  void foo();
  template <typename T> class C {
    friend void foo();
  };
  inline void foo() {}
  C<int> c;

We shouldn't be marking the instantiation of the friend decl of foo as
inline-specified. It may be possible to fix this by determining if the
full definition is part of the current template, but it seems better to
rever tot green until we come up with a full solution.

This reverts commit r233817, as well as follow-ups r233820 and r233821.

llvm-svn: 234355
2015-04-07 20:46:51 +00:00
Daniel Jasper 303c3ac925 Revert "[SEH] Implement filter capturing in CodeGen"
Test fails:
http://lab.llvm.org:8080/green/job/clang-stage2-configure-Rlto_check/3182/

llvm-svn: 234306
2015-04-07 10:07:47 +00:00
Reid Kleckner 67cf035a99 [WinEH] Don't create an alloca for unnamed catch parameters
The catch object parameter to llvm.eh.begincatch is optional, and can be
null. We can save some ourselves the stack space, copy ctor, and dtor
calls if we pass null.

llvm-svn: 234264
2015-04-07 00:09:59 +00:00
Reid Kleckner 0ada50f17f [SEH] Implement filter capturing in CodeGen
While capturing filters aren't very common, we'd like to outline
__finally blocks in the frontend to simplify -O0 EH preparation and
reduce code size. Finally blocks are usually have captures, and this is
the first step towards that.

Currently we don't support capturing 'this' or VLAs.

Reviewers: majnemer

Differential Revision: http://reviews.llvm.org/D8825

llvm-svn: 234261
2015-04-06 23:51:44 +00:00
David Majnemer f205f5324b [MS ABI] A pointer-to-function cannot be caught as a pointer-to-void
Don't assume that all pointers are convertible to void pointer.
Instead correctly respect [conv.ptr]p2; only allow pointer types with an
object pointee type to be caught as pointer-to-void.

llvm-svn: 234090
2015-04-04 05:37:48 +00:00
Manman Ren 2738278b7f [i386 ABI] expand small C like structs in C++, just like how we handle small
C structs.

This comes up when we have a function that takes a struct and is defined in a
C++ file and used in a C file.

Before this commit, we will generate byval for C++ and will expand the struct
for C, thus causing difference at IR level. We will use bitcast of function type
at the callsite, which causes the inliner to not inline the function.

This commit changes how we handle small C like structs at IR level, but at
backend, we should generate the same argument passing before and after the
commit.

Note that the condition for expanding is still over conservative. We should be
able to expand type that is spelled with “class” and types that are not C-like.
But this commit fixes the inconsistent argument passing between C/C++.

Reviewed by John.

rdar://20121030

llvm-svn: 234033
2015-04-03 18:10:29 +00:00
Peter Collingbourne 1a7488afaa Implement CFI type checks for non-virtual calls.
This uses the same class metadata currently used for virtual call and
cast checks.

The new flag is -fsanitize=cfi-nvcall. For consistency, the -fsanitize=cfi-vptr
flag has been renamed -fsanitize=cfi-vcall.

Differential Revision: http://reviews.llvm.org/D8756

llvm-svn: 233874
2015-04-02 00:23:30 +00:00
Reid Kleckner d2f6c726e7 Re-land "MS ABI: lambda call operators are instance methods and should use thiscall"
Update the test cases to pass when lambda call operators use thiscall.

Update the lambda-to-block conversion operator to use the default free
function calling convention instead of the call operator's convention.

This reverts commit r233082 and re-instates r233023.

llvm-svn: 233835
2015-04-01 20:22:13 +00:00
Reid Kleckner 2dc063fa16 Fix inlinehint.cpp with an explicit triple to avoid x86_thiscallcc problems
llvm-svn: 233821
2015-04-01 16:48:54 +00:00
Reid Kleckner e20ce1fcbf Remove comdats from inlinehint.cpp to pass test on Darwin
llvm-svn: 233820
2015-04-01 16:45:17 +00:00
Reid Kleckner 0a33e615f3 Mark instantiated function decls as inline specified if any pattern is
A function template pattern can be declared without the 'inline'
specifier and defined later with the 'inline' specifier. However, during
instantiation, we were only looking at the canonical decl to see if we
should mark the instantiated decl as inline specified. Since the
instantiated decl actually represents many pattern declarations, put the
inline specifier on the instantiation decl if any of the pattern decls
have it.

llvm-svn: 233817
2015-04-01 16:23:44 +00:00
David Majnemer 1f192e26fd [MS ABI] Disregard restrictive exception specifications
MSVC treats all non-empty exception specifications the same way: all
exceptions are permitted.  The .xdata tables provide a way to
efficiently lower exception specifications *but* this probably has to be
implemented as a catch-all/rethrow mechanism instead of the Itanium way.

This fixes PR23092.

llvm-svn: 233787
2015-04-01 04:45:52 +00:00
David Majnemer 37b417f1ed [MS ABI] Rework .xdata HandlerType emission
Utilizing IMAGEREL relocations for synthetic IR constructs isn't
valuable, just clutter.  While we are here, simplify HandlerType names
by making the numeric value for the 'adjective' part of the mangled name
instead of appending '.const', etc.  The old scheme made for very long
global names and leads to wordy things like '.std_bad_alloc'

llvm-svn: 233503
2015-03-29 21:55:10 +00:00
Benjamin Kramer 52bd867399 Remove test for PR12917 for now. I'm relatively sure this is ill-formed per [dcl.fct.default]p3.
llvm-svn: 233379
2015-03-27 14:12:44 +00:00
Aaron Ballman 368701a0c8 Reverting r233023 -- it caused test failures on Windows with MSVC x86.
http://bb.pgr.jp/builders/ninja-clang-i686-msc18-R/builds/572

llvm-svn: 233082
2015-03-24 13:51:13 +00:00
Reid Kleckner a7b275d3b2 MS ABI: lambda call operators are instance methods and should use thiscall
Fixes an issue reported by Daniel Berenyi on cfe-dev.

llvm-svn: 233023
2015-03-23 21:16:49 +00:00
Reid Kleckner 7ffc3fbb2f C++14: Disable sized deallocation by default due to ABI breakage
There are no widely deployed standard libraries providing sized
deallocation functions, so we have to punt and ask the user if they want
us to use sized deallocation. In the future, when such libraries are
deployed, we can teach the driver to detect them and enable this
feature.

N3536 claimed that a weak thunk from sized to unsized deallocation could
be emitted to avoid breaking backwards compatibility with standard
libraries not providing sized deallocation. However, this approach and
other variations don't work in practice.

With the weak function approach, the thunk has to have default
visibility in order to ensure that it is overridden by other DSOs
providing sized deallocation. Weak, default visibility symbols are
particularly expensive on MachO, so John McCall was considering
disabling this feature by default on Darwin. It also changes behavior
ELF linking behavior, causing certain otherwise unreferenced object
files from an archive to be pulled into the link.

Our second approach was to use an extern_weak function declaration and
do an inline conditional branch at the deletion call site. This doesn't
work because extern_weak only works on MachO if you have some archive
providing the default value of the extern_weak symbol. Arranging to
provide such an archive has the same challenges as providing the symbol
in the standard library. Not to mention that extern_weak doesn't really
work on COFF.

Reviewers: rsmith, rjmccall

Differential Revision: http://reviews.llvm.org/D8467

llvm-svn: 232788
2015-03-20 00:31:07 +00:00
David Majnemer 73768709f2 Don't crash-on-valid when an inline function is friend of class template
We assumed that the most recent declaration of an inline function would
also be inline.  However, a more recent declaration can come from a
friend declaration in a class template that is instantiated at the
definition of the function.

llvm-svn: 232786
2015-03-20 00:02:27 +00:00
David Majnemer a7f8c46439 MS ABI: Implement the MSVC 2015 scheme for scope disambiguation
consider C++ that looks like:
  inline int &f(bool b) {
    if (b) {
      static int i;
      return i;
    }
    static int i;
    return i;
  }

Both 'i' variables must have distinct (and stable) names for linkage
purposes.  The MSVC 2013 ABI would number the variables using a count of
the number of scopes that have been created.  However, the final 'i'
returns to a scope that has already been created leading to a mangling
collision.

MSVC 2015 fixes this by giving the second 'i' the name it would have if
it were declared before the 'if'.  However, this results in ABI breakage
because the mangled name, in cases where there was no ambiguity, would
now be different.

We implement the new behavior and only enable it if we are targeting the
MSVC 2015 ABI, otherwise the old behavior will be used.

This fixes PR18131.

llvm-svn: 232766
2015-03-19 21:54:30 +00:00
Reid Kleckner 7270ef5705 Revert "Improvement on sized deallocation from r230160"
This reverts commit r230580.

extern_weak functions don't appear to work on Darwin (PR22951), so we'll
need to come up with a new approach.

llvm-svn: 232731
2015-03-19 17:03:58 +00:00
David Majnemer a03849b2e7 MS ABI: Don't try to emit VF/VB-Tables for extern class templates
There will be an explicit template instantiation in another translation
unit which will provide the definition of the VF/VB-Tables.

This fixes PR22932.

llvm-svn: 232680
2015-03-18 22:04:43 +00:00
David Majnemer 4deaac780c MS ABI: Empty pack expansions had their mangling changed in 2013->2015
We used to support the 2013 mangling and changed it to the more
reasonable 2015 mangling.  Let's make the mangling conditional on what
version of MSVC is targeted.

This fixes PR21888.

llvm-svn: 232609
2015-03-18 03:56:27 +00:00
NAKAMURA Takumi 59aca19708 Fix test in release mode. This reapplies r232456, corresponding to r232579.
llvm-svn: 232582
2015-03-18 01:41:58 +00:00
Nick Lewycky 84146bee6c Fix the LLVM type used when lowering initializer list reference temporaries to global variables. Reapplies r232454 with fix for PR22940.
llvm-svn: 232579
2015-03-18 01:06:24 +00:00
Reid Kleckner 93f661a1da MS ABI: Build C++ default argument exprs for exported template classes
This was an omission from r232229.

llvm-svn: 232554
2015-03-17 21:51:43 +00:00
David Majnemer 5f0dd6162c MS ABI: Emit HandlerMap entries for C++ catch
The HandlerMap describes, to the runtime, what sort of catches surround
the try.  In principle, this structure has to be emitted by the backend
because only it knows the layout of the stack (the runtime needs to know
where on the stack the destination of a copy lives, etc.) but there is
some C++ specific information that the backend can't reason about.

Stick this information in special LLVM globals with the relevant
"const", "volatile", "reference" info mangled into the name.

llvm-svn: 232538
2015-03-17 20:35:05 +00:00
Reid Kleckner bba3cb95cc MS ABI: Delay default constructor closure checking until the outermost class scope ends
Previously, we would error out on this code because the default argument
wasn't parsed until the end of Outer:

  struct __declspec(dllexport) Outer {
    struct __declspec(dllexport) Inner {
      Inner(void *p = 0);
    };
  };

Now we do the checking on the closing brace of Outer instead of Inner.

llvm-svn: 232519
2015-03-17 19:00:50 +00:00
Hans Wennborg f9d865b059 Revert r232454 and r232456: "Fix the LLVM type used when lowering initializer list reference temporaries to global variables."
This caused PR22940.

llvm-svn: 232496
2015-03-17 16:38:58 +00:00
Nick Lewycky 563dfdb111 Fix test in release mode.
llvm-svn: 232456
2015-03-17 02:31:46 +00:00
Nick Lewycky cf191adaf5 Fix the LLVM type used when lowering initializer list reference temporaries to global variables.
llvm-svn: 232454
2015-03-17 02:21:31 +00:00
Duncan P. N. Exon Smith a7fd07f3fc CodeGenCXX: Test that linetables work with variadic virtual thunks
Add a frontend test for PR22929, which was fixed by LLVM r232449.
Besides the crash test, check that the `!dbg` attachment is sane since
its presence was the trigger.

llvm-svn: 232450
2015-03-17 01:19:01 +00:00
David Majnemer ad803d4b76 MS ABI: Don't use qualified pointee types for 'catch' EH TypeDescriptors
Qualifiers are located next to the TypeDescriptor in order to properly
ensure that a pointer type can only be caught by a more qualified catch
handler.  This means that a catch handler of type 'const int *' requires
an RTTI object for 'int *'.  We got this correct for 'throw' but not for
'catch'.

N.B.  We don't currently have the means to store the qualifiers because
LLVM's EH strategy is tailored to the Itanium scheme.  The Itanium ABI
stores qualifiers inside the type descriptor in such a way that the
manner of qualification is stored in addition to the pointee type's
descriptor.  Perhaps the best way of modeling this for the MS ABI is
using an aggregate type to bundle the qualifiers with the descriptor?
This is tricky because we want to make it clear to the optimization
passes which catch handlers invalidate other handlers.

My current thoughts on a design for this is along the lines of:
  { { TypeDescriptor* TD, i32 QualifierFlags }, i32 MiscFlags }

The idea is that the inner most aggregate is all that is needed to
communicate that one catch handler might supercede another.  The
'MiscFlags' field would be used to hold the bitpattern for the notion
that the 'catch' handler does not need to invoke a copy-constructor
because we are catching by reference.

llvm-svn: 232318
2015-03-15 07:10:01 +00:00
David Majnemer 8062eb6bed CodeGen: Correctly initialize bitfields with non-constant initializers
It is possible to construct an initializer for a bitfield which is not
constant.  Instead of emitting code to initialize the field before the
execution of main, clang would crash.

llvm-svn: 232285
2015-03-14 22:24:38 +00:00
David Majnemer bc02d32f4d MS ABI: Mangle virtual member pointer thunks with the correct CC
Virtual member pointers are implemented using a thunk.  We assumed that
the calling convention for this thunk was always __thiscall for 32-bit
targets and __cdecl for 64-bit targets.  However, this is not the case.
Mangle in whichever calling convention is appropriate for this member
function thunk.

llvm-svn: 232254
2015-03-14 06:34:41 +00:00
Peter Collingbourne d2926c91d5 Implement bad cast checks using control flow integrity information.
This scheme checks that pointer and lvalue casts are made to an object of
the correct dynamic type; that is, the dynamic type of the object must be
a derived class of the pointee type of the cast. The checks are currently
only introduced where the class being casted to is a polymorphic class.

Differential Revision: http://reviews.llvm.org/D8312

llvm-svn: 232241
2015-03-14 02:42:25 +00:00
Ahmed Bougacha 5a4aa42a59 Add a bunch of missing "CHECK" colons in tests. NFC.
llvm-svn: 232237
2015-03-14 01:10:19 +00:00
David Majnemer 37fd66e78b MS ABI: Generate default constructor closures
The MS ABI utilizes a compiler generated function called the "vector
constructor iterator" to construct arrays of objects with
non-trivial constructors/destructors.  For this to work, the constructor
must follow a specific calling convention.  A thunk must be created if
the default constructor has default arguments, is variadic or is
otherwise incompatible.  This thunk is called the default constructor
closure.

N.B.  Default constructor closures are only generated if the default
constructor is exported because clang itself does not utilize vector
constructor iterators.  Failing to export the default constructor
closure will result in link/load failure if a translation unit compiled
with MSVC is on the import side.

Differential Revision: http://reviews.llvm.org/D8331

llvm-svn: 232229
2015-03-13 22:36:55 +00:00
David Majnemer ba3e5ecf07 MS ABI: Implement __GetExceptionInfo for std::make_exception_ptr
std::make_exception_ptr calls std::__GetExceptionInfo in order to figure
out how to properly copy the exception object.

Differential Revision: http://reviews.llvm.org/D8280

llvm-svn: 232188
2015-03-13 18:26:17 +00:00
David Blaikie bdf40a62a7 Test case updates for explicit type parameter to the gep operator
llvm-svn: 232187
2015-03-13 18:21:46 +00:00
David Majnemer a1aea9aad4 MS ABI: Allow a nullptr_t exception to be caught by void * catch handler
A nullptr exception object can be caught by any pointer type catch
handler.  However, it is not possible to express this in the exception
info for the MS ABI.  As a middle ground, allow such exception objects
to be caught with pointer-to-void catch handlers.

llvm-svn: 232069
2015-03-12 17:44:49 +00:00
David Majnemer 678732933f Don't overconstrain a FileCheck pattern
llvm-svn: 231971
2015-03-11 21:50:09 +00:00
David Majnemer dfa6d2067c MS ABI: Implement copy-ctor closures, finish implementing throw
This adds support for copy-constructor closures.  These are generated
when the C++ runtime has to call a copy-constructor with a particular
calling convention or with default arguments substituted in to the call.

Because the runtime has no mechanism to call the function with a
different calling convention or know-how to evaluate the default
arguments at run-time, we create a thunk which will do all the
appropriate work and package it in a way the runtime can use.

Differential Revision: http://reviews.llvm.org/D8225

llvm-svn: 231952
2015-03-11 18:36:39 +00:00
David Majnemer f9bde287e8 Sema: Properly track mangling number/name for linkage for using decls
Using declarations which are aliases to struct types have their name
used as the struct type's name for linkage purposes.  Otherwise, make
sure to give an anonymous struct defined inside a using declaration a
mangling number to disambiguate it from other anonymous structs in the
same context.

This fixes PR22809.

llvm-svn: 231909
2015-03-11 06:45:39 +00:00
Daniel Jasper a61e212a43 Prevent test from accessing files.
llvm-svn: 231907
2015-03-11 06:01:01 +00:00
Richard Smith 67375c3bbc Fix test to specify an Itanium triple.
llvm-svn: 231900
2015-03-11 01:09:01 +00:00
Richard Smith ed1cb88c8a For PR22870: produce an error rather than asserting if a designated initializer appears in a signature.
llvm-svn: 231892
2015-03-11 00:12:17 +00:00
David Majnemer 999cbf9d21 MS ABI: Mangle the location of the catchable type into it's name
Because the catchable type has a reference to its name, mangle the
location to ensure that two catchable types with different locations are
distinct.

llvm-svn: 231819
2015-03-10 19:01:51 +00:00
NAKAMURA Takumi f65421ad9f Suppress a couple of tests, clang/test/CodeGen/catch-undef-behavior.c and one, for -Asserts for now. They were introduced in r231711.
llvm-svn: 231717
2015-03-09 22:32:03 +00:00
Alexey Samsonov 21d2dda3d2 [UBSan] Split -fsanitize=shift into -fsanitize=shift-base and -fsanitize=shift-exponent.
This is a recommit of r231150, reverted in r231409. Turns out
that -fsanitize=shift-base check implementation only works if the
shift exponent is valid, otherwise it contains undefined behavior
itself.

Make sure we check that exponent is valid before we proceed to
check the base. Make sure that we actually report invalid values
of base or exponent if -fsanitize=shift-base or
-fsanitize=shift-exponent is specified, respectively.

llvm-svn: 231711
2015-03-09 21:50:19 +00:00
Benjamin Kramer f8b86964ca Reapply r231508 "CodeGen: Emit constant temporaries into read-only globals."
I disabled putting the new global into the same COMDAT as the function for now.
There's a fundamental problem when we inline references to the global but still
have the global in a COMDAT linked to the inlined function. Since this is only
an optimization there may be other versions of the COMDAT around that are
missing the new global and hell breaks loose at link time.

I hope the chromium build doesn't break this time :)

llvm-svn: 231564
2015-03-07 13:37:13 +00:00
Hans Wennborg cd8f011157 Revert r231508 "CodeGen: Emit constant temporaries into read-only globals."
This broke the Chromium build. Links were failing with messages like:

obj/dbus/libdbus_test_support.a(obj/dbus/dbus_test_support.mock_object_proxy.o):../../dbus/mock_object_proxy.cc:function dbus::MockObjectProxy::Detach(): warning: relocation refers to discarded section
/usr/local/google/work/chromium/src/third_party/binutils/Linux_x64/Release/bin/ld.gold: error: treating warnings as errors

llvm-svn: 231541
2015-03-07 00:46:19 +00:00
David Majnemer b6207883e3 Unbreak build bots
llvm-svn: 231536
2015-03-06 23:56:30 +00:00
David Majnemer 322fe4188f MS ABI: Stick throw-related data into the .xdata section
This is a little nicer as it keeps the contents of .xdata away from
normal .rdata; we expect .xdata to be far colder than .rdata.

llvm-svn: 231534
2015-03-06 23:45:23 +00:00
David Majnemer d3d7669ced MS ABI: Correctly generate throw-info for pointer to const qual types
We didn't create type info based on the unqualified pointee type,
causing RTTI mismatches.

llvm-svn: 231533
2015-03-06 23:45:20 +00:00
Benjamin Kramer 3d8aa5c77d CodeGen: Emit constant temporaries into read-only globals.
Instead of creating a copy on the stack just stash them in a private
constant global. This saves both the copying overhead and the stack
space, and gives the optimizer more room to constant fold.

This tries to make array temporaries more similar to regular arrays,
they can't use the same logic because a temporary has no VarDecl to be
bound to so we roll our own version here.

The original use case for this optimization was code like
  for (int i : {1, 2, 3, 4, 5, 6, 7, 8, 10})
    foo(i);
where without this patch (assuming that the loop is not unrolled) we
would alloca an array on the stack, copy the 10 values over and
iterate on that. With this patch we put the array in .text use it
directly. Apart from that case this helps on virtually any passing of
a constant std::initializer_list as a function argument.

Differential Revision: http://reviews.llvm.org/D8034

llvm-svn: 231508
2015-03-06 20:00:03 +00:00
David Majnemer e7a818fec8 MS ABI: Insert copy-constructors into the CatchableType
Find all unambiguous public classes of the exception object's class type
and reference all of their copy constructors.  Yes, this is not
conforming but it is necessary in order to implement their ABI.  This is
because the copy constructor is actually referenced by the metadata
describing which catch handlers are eligible to handle the exception
object.

N.B.  This doesn't yet handle the copy constructor closure case yet,
that work is ongoing.

Differential Revision: http://reviews.llvm.org/D8101

llvm-svn: 231499
2015-03-06 18:53:55 +00:00
Nico Weber 55905145e7 Don't crash on non-public referenced dtors in toplevel classes.
Fixes PR22793, a bug that caused self-hosting to fail after the innocuous
r231254. See the bug for details.

llvm-svn: 231451
2015-03-06 06:01:06 +00:00
Fariborz Jahanian 28d94b1df2 [PATCH] Patch to fix the AST for vector splat from any
arithmetic type to a vector so that the arithmatic type
matches the vector element type. Without which it crashes
in Code Gen. rdar://20000762

llvm-svn: 231419
2015-03-05 23:06:09 +00:00
Alexey Samsonov 48a9db034a Revert "[UBSan] Split -fsanitize=shift into -fsanitize=shift-base and -fsanitize=shift-exponent."
It's not that easy. If we're only checking -fsanitize=shift-base we
still need to verify that exponent has sane value, otherwise
UBSan-inserted checks for base will contain undefined behavior
themselves.

llvm-svn: 231409
2015-03-05 21:57:35 +00:00
David Majnemer 7c23707174 MS ABI: Implement support for throwing a C++ exception
Throwing a C++ exception, under the MS ABI, is implemented using three
components:
- ThrowInfo structure which contains information like CV qualifiers,
  what destructor to call and a pointer to the CatchableTypeArray.
- In a significant departure from the Itanium ABI, copying by-value
  occurs in the runtime and not at the catch site.  This means we need
  to enumerate all possible types that this exception could be caught as
  and encode the necessary information to convert from the exception
  object's type to the catch handler's type.  This includes complicated
  derived to base conversions and the execution of copy-constructors.

N.B. This implementation doesn't support the execution of a
copy-constructor from within the runtime for now.  Adding support for
that functionality is quite difficult due to things like default
argument expressions which may evaluate arbitrary code hiding in the
copy-constructor's parameters.

Differential Revision: http://reviews.llvm.org/D8066

llvm-svn: 231328
2015-03-05 00:46:22 +00:00
Rick Foos e9c019a7a6 Temporary XFAILs for Hexagon
Summary: Temporary XFAIL's until patches done.

Reviewers: echristo, adasgupt, colinl

Reviewed By: colinl

Subscribers: llvm-commits

Differential Revision: http://reviews.llvm.org/D8044

llvm-svn: 231318
2015-03-04 23:40:38 +00:00
Alexey Samsonov 783b8174ad [UBSan] Split -fsanitize=shift into -fsanitize=shift-base and -fsanitize=shift-exponent.
-fsanitize=shift is now a group that includes both these checks, so
exisiting users should not be affected.

This change introduces two new UBSan kinds that sanitize only left-hand
side and right-hand side of shift operation. In practice, invalid
exponent value (negative or too large) tends to cause more portability
problems, including inconsistencies between different compilers, crashes
and inadequeate results on non-x86 architectures etc. That is,
-fsanitize=shift-exponent failures should generally be addressed first.

As a bonus, this change simplifies CodeGen implementation for emitting left
shift (separate checks for base and exponent are now merged by the
existing generic logic in EmitCheck()), and LLVM IR for these checks
(the number of basic blocks is reduced).

llvm-svn: 231150
2015-03-03 22:15:35 +00:00
Reid Kleckner fff8e7f6ba Split catch IRgen into ItaniumCXXABI and MicrosoftCXXABI
Use llvm.eh.begincatch for Microsoft-style catches.

This moves lots of CGException code into ItaniumCXXABI. Sorry for the
blame pain.

llvm-svn: 231105
2015-03-03 19:21:04 +00:00
Reid Kleckner f7e1e75828 Disable the right RUN line
llvm-svn: 231098
2015-03-03 18:36:38 +00:00
Filipe Cabecinhas a96989dc8d Disabled the other test from r231086 (like in r231087) since it also had problems
llvm-svn: 231096
2015-03-03 18:32:38 +00:00
Reid Kleckner 1859025df8 Disable a Clang test until the begincatch change lands
llvm-svn: 231087
2015-03-03 17:51:26 +00:00
Duncan P. N. Exon Smith f04be1fb3a DebugInfo: Move new hierarchy into place (clang)
Update testcases for LLVM change in r231082 to use the new debug info
hierarchy.

llvm-svn: 231083
2015-03-03 17:25:55 +00:00
Benjamin Kramer 83b1bf3a27 CodeGen: Fix passing of classes with only one AVX vector member in AVX registers
isSingleElementStruct was a bit too tight in its definition of struct
so we got a mismatch between classify() and the actual code generation.
To make matters worse the code in GetByteVectorType still defaulted to
<2 x double> if it encountered a type it didn't know, making this a
silent miscompilation (PR22753).

Completely remove the "preferred type" stuff from GetByteVectorType and
make it fail an assertion if someone tries to use it with a type not
suitable for a vector register.

llvm-svn: 230971
2015-03-02 16:09:24 +00:00
Peter Collingbourne d251b0a60e DebugInfo: Give externally defined types a size and alignment where
possible. Fixes PR22736.

llvm-svn: 230914
2015-03-01 22:07:04 +00:00
Richard Smith 91c18de755 Rework our handling of key functions. We used to track a complete list of all
dynamic classes in the translation unit and check whether each one's key
function is defined when we got to the end of the TU (and when we got to the
end of each module). This is really terrible for modules performance, since it
causes unnecessary deserialization of every dynamic class in every compilation.

We now use a much simpler (and, in a modules build, vastly more efficient)
system: when we see an out-of-line definition of a virtual function, we check
whether that function was in fact its class's key function. (If so, we need to
emit the vtable.)

llvm-svn: 230830
2015-02-28 01:01:56 +00:00
Saleem Abdulrasool cd187f033e DebugInfo: hoist definition into global context when needed
When generating debug info for a static inline member which is initialized for
the DLLExport storage class, hoist the definition into a non-composite type
context. Otherwise, we would trigger an assertion when generating the DIE for
the associated global value as the debug context has a type association. This
addresses PR22669.

Thanks to David Blakie for help in coming up with a solution to this!

llvm-svn: 230816
2015-02-28 00:13:13 +00:00
David Blaikie a953f2825b Update Clang tests to handle explicitly typed load changes in LLVM.
llvm-svn: 230795
2015-02-27 21:19:58 +00:00
David Blaikie 218b783192 Update Clang tests to handle explicitly typed gep changes in LLVM.
llvm-svn: 230783
2015-02-27 19:18:17 +00:00
Larisse Voufo 697b32a0cb Improvement on sized deallocation from r230160:
Do not declare sized deallocation functions dependently on whether it is found in global scope. Instead, enforce the branching in emitted code by (1) declaring the functions extern_weak and (2) emitting sized delete expressions as a branching between both forms delete.

llvm-svn: 230580
2015-02-25 23:48:43 +00:00
Peter Collingbourne 69b004d987 UBSan: Use the correct function prologue for x32.
llvm-svn: 230571
2015-02-25 23:18:42 +00:00
David Majnemer dbdab4037e MS ABI: Turn throw into std::terminate for now, make try/catch "work"
This lets us compile programs which make use of exceptional constructs
statically without executing any of them dynamically.

llvm-svn: 230568
2015-02-25 23:01:21 +00:00
Reid Kleckner 8b6d034cab Reland r230446, "MS ABI: Try to respect external AST source record layouts"
It broke test/PCH/headersearch.cpp because it was using -Wpadding, which
only works for Itanium layout. Before this commit, we would use Itanium
record layout when using PCH, which is crazy. Now that the test uses an
explicit Itanium triple, we can reland.

llvm-svn: 230525
2015-02-25 19:17:45 +00:00
NAKAMURA Takumi a2acc360ed Revert r230446, "MS ABI: Try to respect external AST source record layouts"
It fails on Clang::PCH/headersearch.cpp for targeting msvc.

llvm-svn: 230474
2015-02-25 10:32:13 +00:00
Reid Kleckner 3990db79c5 MS ABI: Try to respect external AST source record layouts
Covered by existing tests in test/CodeGen/override-layout.c and
test/CodeGenCXX/override-layout.cpp. Seriously, they found real bugs in
my code. :)

llvm-svn: 230446
2015-02-25 02:16:09 +00:00
Larisse Voufo a65298af65 Relax the requirement on sized deallocation a bit: Default on unsized delete if sized delete is not provided in global scope, and -fdefine-sized-deallocation option is disabled.
llvm-svn: 230160
2015-02-22 06:36:53 +00:00
Peter Collingbourne a4ccff3281 Implement Control Flow Integrity for virtual calls.
This patch introduces the -fsanitize=cfi-vptr flag, which enables a control
flow integrity scheme that checks that virtual calls take place using a vptr of
the correct dynamic type. More details in the new docs/ControlFlowIntegrity.rst
file.

It also introduces the -fsanitize=cfi flag, which is currently a synonym for
-fsanitize=cfi-vptr, but will eventually cover all CFI checks implemented
in Clang.

Differential Revision: http://reviews.llvm.org/D7424

llvm-svn: 230055
2015-02-20 20:30:56 +00:00
Larisse Voufo 89088232b2 Add -fno-sized-deallocation option for completeness of fix in r229241 in documentation in r229818.
llvm-svn: 229950
2015-02-20 02:07:22 +00:00
Hans Wennborg 97cbed422c Don't dllexport inline methods when targeting MinGW.
MinGW neither imports nor exports such methods. The import bit was
committed earlier, in r221154, and this takes care of the export part.

This also partially fixes PR22591.

llvm-svn: 229922
2015-02-19 22:39:24 +00:00
David Majnemer 67a8ec6258 Itanium ABI: Pack expansions change the arity of expressions to unknown
llvm-svn: 229918
2015-02-19 21:41:48 +00:00
Reid Kleckner 4343922dde Avoid using a COMDAT for sized delete on MachO
llvm-svn: 229915
2015-02-19 21:13:45 +00:00
Reid Kleckner 66abf2f92f Put the implicit weak sized deallocation funciton in C++14 in a comdat
Fixes PR22635.

llvm-svn: 229913
2015-02-19 21:01:34 +00:00
David Majnemer f63bcaa2c5 CodeGen: Weak reference temporaries belong in a COMDAT
llvm-svn: 229902
2015-02-19 19:35:18 +00:00
David Majnemer c9a9c7a673 CodeGen: static constexpr data members should have a linkonce_odr init
Classes can be defined in multiple translation units.  This means that
the static constexpr data members should have identical initializers in
all translation units.  Implement this by giving the reference temporary
linkonce_odr linkage.

llvm-svn: 229900
2015-02-19 19:25:17 +00:00
Larisse Voufo 62d8aa5caf Fix a test case.
llvm-svn: 229812
2015-02-19 03:03:23 +00:00
David Majnemer b8014dd7c0 Itanium ABI: Properly qualify the destructor-name
We didn't have enough qualificaiton before the scope specifier and we
had too much qualification in the destructor name itself.

llvm-svn: 229809
2015-02-19 02:16:16 +00:00
David Majnemer eed9c8bb44 Itanium ABI: Restore disabled tests which are correctly mangled
llvm-svn: 229725
2015-02-18 19:08:12 +00:00
David Majnemer f8c02e6bfb Itanium ABI: Properly mangle extern "C" template arguments
extern "C" declarations should be considered like global declarations
for mangling purposes.

Differential Revision: http://reviews.llvm.org/D7718

llvm-svn: 229724
2015-02-18 19:08:11 +00:00
Zoran Jovanovic 26a1216a74 Change representation of member function pointers for MIPS targets
Differential Revision: http://reviews.llvm.org/D7148

llvm-svn: 229680
2015-02-18 15:21:35 +00:00
David Majnemer 7ff7eb706a Itanium ABI: Mangle <mangled-name> according to the ABI
We attempted to be compatible with GCC's buggy mangling for templates
with a declaration for a template argument.

However, we weren't completely successful in copying their bug in cases
like:
  char foo;
  template <char &C> decltype(C) f() { return foo; };
  template char &f<foo>();

Instead, just follow the ABI specification.  This fixes PR22621.

llvm-svn: 229644
2015-02-18 07:47:09 +00:00
David Majnemer a88b359fdc Itanium ABI: Improve our mangling of <destructor-name>
Our mangling of <destructor-name> wasn't quite right: we'd introduce
mangling substitutions where one shouldn't be possible.  We also didn't
correctly handle the case where the destroyed type was not dependent but
still a TemplateSpecializationType.

N.B. There isn't a mangling for a template-template parameter showing up
as the destroyed type.  We do the 'obvious' thing and mangle the index
of the parameter.

llvm-svn: 229615
2015-02-18 02:28:01 +00:00
Larisse Voufo e990a3f60c Rename flags and options to match current naming: from -fdef-sized-delete to -fdefine-sized-deallocation, and from DefaultSizedDelete to DefineSizedDeallocation.
llvm-svn: 229597
2015-02-18 01:04:10 +00:00
Larisse Voufo 0c691d029a More on fixing sized deallocation implementation logic: Fix PR21754.
llvm-svn: 229291
2015-02-15 08:16:37 +00:00
David Blaikie 4e0ccd1a48 Update test to be ready for typeless pointer changes
llvm-svn: 229270
2015-02-15 04:12:16 +00:00
David Blaikie a23916f1cd Update test to make it robust to migration to typeless pointers
llvm-svn: 229269
2015-02-15 04:12:14 +00:00
David Majnemer 1dabfdcf6f ItaniumMangle: Correctly mangle <base-unresolved-name>
We had two bugs:
- We were missing the "on" prefix for unresolved operators.
- We didn't handle the mangling of destructors at all.

This fixes PR22584.

llvm-svn: 229255
2015-02-14 13:23:54 +00:00
Larisse Voufo 5526f4f094 Revise the implementation logic of sized deallocation: Do not automatically generate weak definitions of the sized operator delete (in terms of unsized operator delete). Instead, provide the funcitonality via a new compiler flag, -fdef-sized-delete.
The current implementation causes link-time ODR violations when the delete symbols are exported into the dynamic table.

llvm-svn: 229241
2015-02-14 05:42:57 +00:00
Richard Smith 902a0238aa DR1748: the reserved placement allocation functions have undefined behavior if
they're given a null pointer as an argument, so we do not need to emit null
checks on their results.

llvm-svn: 229213
2015-02-14 01:52:20 +00:00
Reid Kleckner a593000f01 Add the 'noinline' attribute to call sites within __try bodies
LLVM doesn't support non-call exceptions, so inlining makes it harder to
catch such asynchronous exceptions.

llvm-svn: 228876
2015-02-11 21:40:48 +00:00
Reid Kleckner 4c209c7e1f Add a comdat to __clang_call_terminate
llvm-svn: 228863
2015-02-11 18:50:13 +00:00
Reid Kleckner e7b3f7c70d Emit landing pads for SEH even if nounwind is present
Disabling exceptions applies nounwind to lots of functions. SEH catches
asynch exceptions, so emit the landing pad anyway.

llvm-svn: 228769
2015-02-11 00:00:21 +00:00
David Blaikie 20937be183 DebugInfo: Suppress the location of instructions in complex default arguments.
llvm-svn: 228589
2015-02-09 18:55:57 +00:00
David Blaikie 2221aba85b DebugInfo: Suppress the location of instructions in aggregate default arguments.
Matches the existing code for scalar default arguments. Complex default
arguments probably need the same handling too (test/fix to that coming
next).

llvm-svn: 228588
2015-02-09 18:47:14 +00:00
Reid Kleckner deeddeced3 Re-land r228258 and make clang-cl's /EHs- disable -fexceptions again
After r228258, Clang started emitting C++ EH IR that LLVM wasn't ready
to deal with, even when exceptions were disabled with /EHs-. This time,
make /EHs- turn off -fexceptions while still emitting exceptional
constructs in functions using __try.  Since Sema rejects C++ exception
handling constructs before CodeGen, landingpads should only appear in
such functions as the result of a __try.

llvm-svn: 228329
2015-02-05 18:56:03 +00:00
Richard Smith ed83ebd77e PR22465: when performing list-initialization for a class type C, if we see an
initializer of the form {x}, where x is of type C or a type derived from C,
perform *non-list* initialization of the entity from x, but create a
CXXConstructExpr that knows that we used list-initialization syntax.

Plus some fixes to ensure we mangle correctly in this and related cases.

llvm-svn: 228276
2015-02-05 07:02:11 +00:00
Richard Smith 520449d55e Various fixes to mangling of list-initialization.
llvm-svn: 228274
2015-02-05 06:15:50 +00:00
Nico Weber 94b2368c24 Revert r228258.
It caused a chromium base unittest that tests throwing and catching SEH
exceptions to fail (http://crbug.com/455488) and I suspect it might also
be the cause of the chromium clang win 64-bit shared release builder timing
out during compiles.  So revert to see if that's true.

llvm-svn: 228262
2015-02-05 02:08:50 +00:00
NAKAMURA Takumi b1980e63d7 Tweak clang/test/CodeGenCXX/debug-info-line.cpp for -Asserts.
CHECK: call {{.*}}src

matched

  %call = invoke dereferenceable(4) i32* @_Z3srcv()

I don't think it was intentional.

llvm-svn: 228245
2015-02-05 01:02:34 +00:00
David Blaikie 4d52443c0e DebugInfo: Attribute cleanup code to the end of the scope, not the end of the function.
Now if you break on a dtor and go 'up' in your debugger (or you get an
asan failure in a dtor) during an exception unwind, you'll have more
context. Instead of all dtors appearing to be called from the '}' of the
function, they'll be attributed to the end of the scope of the variable,
the same as the non-exceptional dtor call.

This doesn't /quite/ remove all uses of CurEHLocation (which might be
nice to remove, for a few reasons) - it's still used to choose the
location for some other work in the landing pad. It'd be nice to
attribute that code to the same location as the exception calls within
the block and to remove CurEHLocation.

llvm-svn: 228181
2015-02-04 19:47:54 +00:00
NAKAMURA Takumi 2e4c20109a Tweak clang/test/CodeGenCXX/copy-constructor-elim.cpp to satisfy i686-linux.
llvm-svn: 228146
2015-02-04 14:51:30 +00:00
David Majnemer df1e1aee4c Fix ARM buildbot
llvm-svn: 228111
2015-02-04 03:30:00 +00:00
Larisse Voufo 5ecfede65c Fix typo in test case.
llvm-svn: 228110
2015-02-04 03:16:46 +00:00
Larisse Voufo eebe962eda Fix typo in test case.
llvm-svn: 228108
2015-02-04 02:46:33 +00:00
Larisse Voufo 404e142ed7 Generalize r228066 to give all implicit global allocation functions default visibility.
llvm-svn: 228107
2015-02-04 02:34:32 +00:00
NAKAMURA Takumi 493dae89e8 clang/test/CodeGenCXX/cxx14-implicit-sized-delete.cpp: Unbreak build to add explicit triple x86_64-unknown.
It seems incompatible to i686.

llvm-svn: 228100
2015-02-04 01:00:12 +00:00
NAKAMURA Takumi 44d96be561 clang/test/CodeGenCXX/cxx14-implicit-sized-delete.cpp: Fix for -Asserts.
llvm-svn: 228099
2015-02-04 01:00:04 +00:00
NAKAMURA Takumi 46333480f2 Add LF to EOF.
llvm-svn: 228098
2015-02-04 00:59:59 +00:00
David Majnemer bc18d6ab05 Unbreak the MIPS buildbot
llvm-svn: 228088
2015-02-04 00:26:13 +00:00
Larisse Voufo fce61d3fc8 PR22419: Give implicit sized deallocation functions default visibility
llvm-svn: 228066
2015-02-03 23:31:50 +00:00
David Majnemer 64b2a2628f Unbreak hexagon bot
llvm-svn: 228065
2015-02-03 23:21:21 +00:00
David Majnemer fd1e739a44 CodeGen: Copy-ctorm must obey the destination's alignment requirement
We would synthesize memcpy intrinsics when emitting calls to trivial C++
constructors but we wouldn't take into account the alignment of the
destination.

llvm-svn: 228061
2015-02-03 23:04:06 +00:00
Reid Kleckner 11ca834bef SEH: Track users of __try so we can pick a per-func EH personality
There are four major kinds of declarations that cause code generation:
- FunctionDecl (includes CXXMethodDecl etc)
- ObjCMethodDecl
- BlockDecl
- CapturedDecl

This patch tracks __try usage on FunctionDecls and diagnoses __try usage
in other decls. If someone wants to use __try from ObjC, they can use it
from a free function, since the ObjC code will need an ObjC-style EH
personality.

Eventually we will want to look through CapturedDecls and track SEH
usage on the parent FunctionDecl, if present.

llvm-svn: 228058
2015-02-03 22:52:35 +00:00
David Blaikie 14177b748a DebugInfo: Ensure calls to functions with default arguments which themselves have default arguments, still have locations.
To handle default arguments in C++ in the debug info, we disable code
updating the debug location during the emission of default arguments.

This code was buggy in the case of default arguments which, themselves,
have default arguments - the inner default argument would re-enable
debug info when it was finished, but before the outer default argument
was finished.

This was already a bug, but got worse (because a crasher instead of just
a quality bug) with the recent improvements to debug info line quality
because... The ApplyDebugLocation scoped device would find the debug
info disabled and not save any debug location. But then in
~ApplyDebugLocation it would find the debug info had been enabled and
would then apply the no-location. Then the outer function call would be
emitted without any location. That's bad.

Arguably we could /also/ fix the ApplyDebugLocation to assert on this
situation (where debug info was disabled in the ctor and enabled in the
dtor, or the other way around) but this is at least the necessary fix
regardless.

(also, I imagine this disabling behavior might need to be in-place for
CGExprComplex and CGExprAgg too, maybe... ?)

And I seem to recall seeing some weird default arg stepping behavior
recently which might be related to this too... I'll have to look into
it.

llvm-svn: 228053
2015-02-03 22:37:17 +00:00
David Majnemer 2ba2b26632 MS ABI: Add more documentation and tests for novtable
llvm-svn: 227838
2015-02-02 19:05:46 +00:00
David Blaikie 298720d324 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.

llvm-svn: 227356
2015-01-28 19:50:09 +00:00
Nico Weber b3a9978dc8 Don't let virtual calls and dynamic casts call Sema::MarkVTableUsed().
clang currently calls MarkVTableUsed() for classes that get their virtual
methods called or that participate in a dynamic_cast. This is unnecessary,
since CodeGen only emits vtables when it generates constructor, destructor, and
vtt code. (*)

Note that Sema::MarkVTableUsed() doesn't cause the emission of a vtable.
Its main user-visible effect is that it instantiates virtual member functions
of template classes, to make sure that if codegen decides to write a vtable
all the entries in the vtable are defined.

While this shouldn't change the behavior of codegen (other than being faster),
it does make clang more permissive: virtual methods of templates (in particular
destructors) end up being instantiated less often. In particular, classes that
have members that are smart pointers to incomplete types will now get their
implicit virtual destructor instantiated less frequently. For example, this
used to not compile but does now compile:

    template <typename T> struct OwnPtr {
      ~OwnPtr() { static_assert((sizeof(T) > 0), "TypeMustBeComplete"); }
    };
    class ScriptLoader;
    struct Base { virtual ~Base(); };
    struct Sub : public Base {
      virtual void someFun() const {}
      OwnPtr<ScriptLoader> m_loader;
    };
    void f(Sub *s) { s->someFun(); }

The more permissive behavior matches both gcc (where this is not often
observable, since in practice most things with virtual methods have a key
function, and Sema::DefineUsedVTables() skips vtables for classes with key
functions) and cl (which is my motivation for this change) – this fixes
PR20337.  See this issue and the review thread for some discussions about
optimizations.

This is similar to r213109 in spirit. r225761 was a prerequisite for this
change.

Various tests relied on "a->f()" marking a's vtable as used (in the sema
sense), switch these to just construct a on the stack. This forces
instantiation of the implicit constructor, which will mark the vtable as used.

(*) The exception is -fapple-kext mode: In this mode, qualified calls to
virtual functions (`a->Base::f()`) still go through the vtable, and since the
vtable pointer off this doesn't point to Base's vtable, this needs to reference
Base's vtable directly. To keep this working, keep referencing the vtable for
virtual calls in apple kext mode.

llvm-svn: 227073
2015-01-26 06:23:36 +00:00
Nico Weber 220b3ad598 Test that member functions of constexpr constructed templates are instantiated.
They are referenced from the vtable. (This worked fine, but I couldn't find
an existing test for this.  Maybe I didn't look hard enough.)

llvm-svn: 227072
2015-01-26 06:08:07 +00:00
Nico Weber 4a18449c2c Remove -verify from a codegen test that didn't have any expected-foo lines.
Makes the error output of the test more readable when it fails. Also allows
removing a "not" from the run line.

llvm-svn: 227071
2015-01-26 05:47:24 +00:00
Nico Weber fe0053eb42 Reword comment.
llvm-svn: 227069
2015-01-26 03:03:49 +00:00
Nico Weber 99e05d02cb Test that qualified virtual calls mark vtables referenced in apple kext mode.
I broke this locally while working on PR20337 and no test caught that.  Now
there's coverage for this, and a comment explaining why this is needed.

llvm-svn: 227068
2015-01-26 02:59:07 +00:00
Nico Weber a24b299c00 Rename four test files from .C to .cpp.
lit.cfg has never supported running .C files, so these tests were never
executed by check-clang.  Rename them to .cpp so that they run as part of
the test suite, and minorly tweak two of them that look like they were broken
when checked in to actually pass.

llvm-svn: 227029
2015-01-25 02:04:36 +00:00
David Blaikie 9f7ae2c948 DebugInfo: Attribute calls to overloaded operators with the operator, not the start of the whole expression
llvm-svn: 227028
2015-01-25 01:25:37 +00:00
David Blaikie 9b47966615 DebugInfo: Use the preferred location rather than the start location for expression line info
This causes things like assignment to refer to the '=' rather than the
LHS when attributing the store instruction, for example.

There were essentially 3 options for this:

* The beginning of an expression (this was the behavior prior to this
  commit). This meant that stepping through subexpressions would bounce
  around from subexpressions back to the start of the outer expression,
  etc. (eg: x + y + z would go x, y, x, z, x (the repeated 'x's would be
  where the actual addition occurred)).

* The end of an expression. This seems to be what GCC does /mostly/, and
  certainly this for function calls. This has the advantage that
  progress is always 'forwards' (never jumping backwards - except for
  independent subexpressions if they're evaluated in interesting orders,
  etc). "x + y + z" would go "x y z" with the additions occurring at y
  and z after the respective loads.
  The problem with this is that the user would still have to think
  fairly hard about precedence to realize which subexpression is being
  evaluated or which operator overload is being called in, say, an asan
  backtrace.

* The preferred location or 'exprloc'. In this case you get sort of what
  you'd expect, though it's a bit confusing in its own way due to going
  'backwards'. In this case the locations would be: "x y + z +" in
  lovely postfix arithmetic order. But this does mean that if the op+
  were an operator overload, say, and in a backtrace, the backtrace will
  point to the exact '+' that's being called, not to the end of one of
  its operands.

(actually the operator overload case doesn't work yet for other reasons,
but that's being fixed - but this at least gets scalar/complex
assignments and other plain operators right)

llvm-svn: 227027
2015-01-25 01:19:10 +00:00
David Blaikie 2d321fb79d DebugInfo: Correct the line location of geps on array accesses
llvm-svn: 227023
2015-01-24 23:35:17 +00:00
David Blaikie a627202e19 Disable warnings in an IRGen test to make test failures less noisy
llvm-svn: 226954
2015-01-23 22:47:05 +00:00
Rafael Espindola 6565e92f3f [pr22293] Don't crash during codegen of a recursive destructor.
In ItaniumCXXABI::EmitCXXDestructors we first emit the base destructor
and then try to emit the complete one as an alias.

If in the base ends up calling the complete destructor, the GD for the
complete will be in the list of deferred decl by the time we replace
it with an alias and delete the original GV.

llvm-svn: 226896
2015-01-23 05:26:38 +00:00
Rafael Espindola e5df59ff78 Emit DeferredDeclsToEmit in a DFS order.
Currently we emit DeferredDeclsToEmit in reverse order. This patch changes that.

The advantages of the change are that

* The output order is a bit closer to the source order. The change to
test/CodeGenCXX/pod-member-memcpys.cpp is a good example.

* If we decide to deffer more, it will not cause as large changes in the
estcases as it would without this patch.

llvm-svn: 226751
2015-01-22 00:24:57 +00:00
Rafael Espindola 701d2dd4b7 Use CHECK-LABEL when possible. NFC.
llvm-svn: 226743
2015-01-21 23:33:55 +00:00
David Blaikie 07c6bfd502 DebugInfo: Remove distinct-call-inlining test case as this is being fixed in LLVM.
This test will start failing shortly once this bug is fixed in LLVM. At
that point this behavior is no longer required in Clang and will be
removed. In the interim, remove this test just to avoid the race between
the LLVM and Clang commits.

After the LLVM commit, I'll cleanup the workaround behavior in Clang.

llvm-svn: 226735
2015-01-21 22:57:22 +00:00
Rafael Espindola 2144377536 Use a CHECK-LABEL. NFC
llvm-svn: 226709
2015-01-21 22:38:24 +00:00
Reid Kleckner 9da9448339 Add the "thunk" attribute to MS ABI virtual member pointers
This attribute implies indicates that the function musttail calls
another function and returns whatever it returns. The return type of the
thunk is meaningless, as the thunk can dynamically call different
functions with different return types. So long as the callers bitcast
the thunk with the correct type, behavior is well defined.

This attribute was necessary to fix PR20944, where the indirect call
combiner noticed that the thunk returned void and replaced the results
of the indirect call instruction with undef.

Over-the-shoulder reviewed by David Majnemer.

llvm-svn: 226707
2015-01-21 22:18:17 +00:00
Rafael Espindola 17c57ecf4a Used CHECK-DAG since the order is not important.
llvm-svn: 226677
2015-01-21 17:12:04 +00:00
Rafael Espindola 469e2ee83a Used CHECK-DAG since the order is not important.
llvm-svn: 226675
2015-01-21 16:56:43 +00:00
Rafael Espindola 11bb53bc77 Don't assume variable name.
Should fix the test in -Asserts builds.

llvm-svn: 226668
2015-01-21 16:18:42 +00:00
Rafael Espindola e20c966cf4 Make the test a bit stricter. NFC.
llvm-svn: 226667
2015-01-21 16:13:57 +00:00
Rafael Espindola 3a0e228cca Make the test a bit stricter. NFC.
llvm-svn: 226666
2015-01-21 16:03:26 +00:00
David Majnemer 8c9cdb6573 MS ABI: Virtual member pointer thunks should be in COMDAT groups
They can be emitted by multiple translation units and thus belong in a
COMDAT group.

llvm-svn: 226630
2015-01-21 01:21:31 +00:00
David Majnemer 3072fc885e MS ABI: Let guard variables be present in COMDATs
A guard variable in a COMDAT'd function should also be in a COMDAT.

llvm-svn: 226629
2015-01-21 01:04:30 +00:00
David Majnemer 740d59ec49 CodeGen: Compiler generated __declspec(uuid) objects should be COMDAT'd
llvm-svn: 226628
2015-01-21 01:04:28 +00:00
Adrian Prantl 7c6f944cdf Migrate all uses of DIVariable's FlagIndirectVariable to use a DIExpression
with a DW_OP_deref instead.

llvm-svn: 226474
2015-01-19 17:51:58 +00:00
Rafael Espindola f9b1730d41 Add comdat to thunks.
llvm-svn: 226465
2015-01-19 14:02:14 +00:00
Rafael Espindola 1be4c4546c Make this test a bit stricter. NFC.
llvm-svn: 226463
2015-01-19 13:59:31 +00:00
David Blaikie c6593075bf DebugInfo: Attribute complex expressions to the source location of the expression
Just as r225956 did for scalar expressions (CGExprScalar::Visit), do the
same for complex expressions.

llvm-svn: 226390
2015-01-18 01:57:54 +00:00
David Blaikie 01fb5fb128 DebugInfo: Attribute aggregate expressions to the source location of the expression
Just as r225956 did for scalar expressions (CGExprScalar::Visit), do the
same for aggregate expressions.

llvm-svn: 226388
2015-01-18 01:48:19 +00:00
David Blaikie 1c39bbb741 Refactor test so it's not lazily emitted on a global, simplifying ordering when more test cases are added
llvm-svn: 226384
2015-01-18 00:37:04 +00:00
David Blaikie a81d410d4f DebugInfo: Correct the debug location of non-static data member initializers
This was causing some trouble for otherwise dead code removed in r225085
(reverted in r225361). The location being set for function arguments was
leaking out to the call which wasn't setting its own location (so a
quality bug turned into a crasher with r225085). Fix this so r225085 can
be recommitted.

llvm-svn: 226382
2015-01-18 00:12:58 +00:00
Rafael Espindola d3e0469a84 Make sure all weak destructors go in a comdat in the ms abi.
Destructors have a special treatment in getFunctionLinkage. Instead of
duplicating the logic, check the resulting linkage.

llvm-svn: 226361
2015-01-17 01:47:39 +00:00
David Blaikie 71fb83ea7a Isolate test for PR22096 to clang.
Emitting inlinable calls without debug locations (in functions with
debug info, to functions with debug info) is problematic for debug info
when inlining occurs. Test specifically that we don't do that in this
case - thus the test isn't simply "don't crash", it's "include debug
location for this call" (granted it's the wrong location - fix for that
is coming)

llvm-svn: 226337
2015-01-16 22:55:09 +00:00
Rafael Espindola 4af2cdb732 Also put vtables in a comdat when rtti is disabled.
llvm-svn: 226325
2015-01-16 21:41:44 +00:00
Rafael Espindola d9b26d563a Add comdat to string literal variables on COFF.
llvm-svn: 226317
2015-01-16 20:32:35 +00:00
Rafael Espindola 654542a5ed Add comdats to the RTTI variables in the microsoft abi.
llvm-svn: 226303
2015-01-16 19:23:42 +00:00
Tim Northover 7ad87af277 AArch64: implement AAPCS layout rules for bit-fields.
llvm-svn: 226294
2015-01-16 18:44:04 +00:00
Rafael Espindola 77abc3a7da Add comdats to dynamic init functions in the microsoft abi.
llvm-svn: 226286
2015-01-16 16:04:45 +00:00
Rafael Espindola a8eb9931ad Make this test a bit stricter. NFC.
llvm-svn: 226285
2015-01-16 16:02:03 +00:00
Rafael Espindola 694cb5d9b7 Add comdats to constructs and destructor in the microsoft abi.
llvm-svn: 226280
2015-01-16 15:37:11 +00:00
Rafael Espindola cb92c19fc4 Use a trivial comdat for C++ tables.
This produces comdats for vtables, typeinfo, typeinfo names, and vtts.

When combined with llvm not producing implicit comdats, not doing this would
cause code bloat on ELF and link errors on COFF.

llvm-svn: 226227
2015-01-15 23:18:01 +00:00
Rafael Espindola dbee8a7a7a Use a trivial comdat for inline ctor/dtor when not using C5/D5.
When combined with llvm not producing implicit comdats, not doing this would
cause code bloat on ELF and link errors on COFF.

llvm-svn: 226211
2015-01-15 21:36:08 +00:00
Hans Wennborg fd76d91366 Warn about dllexported explicit class template instantiation declarations (PR22035)
Clang would previously become confused and crash here.

It does not make a lot of sense to export these, so warning seems appropriate.

MSVC will export some member functions for this kind of specializations, whereas
MinGW ignores the dllexport-edness. The latter behaviour seems better.

Differential Revision: http://reviews.llvm.org/D6984

llvm-svn: 226208
2015-01-15 21:18:30 +00:00
David Majnemer f8637360e2 AST: Ensure implicit records have default visibility
Types composed with certain implicit record types would have their RTTI
marked as hidden because the implicit record type didn't have any
visibility.

This manifests itself as triggering false positives from tools like
clang's -fsantize=function feature.  The RTTI for a function type's
return type wouldn't match if the return type was an implicit record
type.

Patch by Stephan Bergmann!

llvm-svn: 226148
2015-01-15 08:41:25 +00:00
Duncan P. N. Exon Smith 8d3ef611ce IR: Move MDLocation into place (clang testcases)
Update testcases to match LLVM change in r226048.

llvm-svn: 226049
2015-01-14 22:28:03 +00:00
David Blaikie 66e4197f07 Reapply r225000 (reverted in r225555): DebugInfo: Generalize debug info location handling (and follow-up commits).
Several pieces of code were relying on implicit debug location setting
which usually lead to incorrect line information anyway. So I've fixed
those (in r225955 and r225845) separately which should pave the way for
this commit to be cleanly reapplied.

The reason these implicit dependencies resulted in crashes with this
patch is that the debug location would no longer implicitly leak from
one place to another, but be set back to invalid. Once a call with
no/invalid location was emitted, if that call was ever inlined it could
produce invalid debugloc chains and assert during LLVM's codegen.

There may be further cases of such bugs in this patch - they're hard to
flush out with regression testing, so I'll keep an eye out for reports
and investigate/fix them ASAP if they come up.

Original commit message:

Reapply "DebugInfo: Generalize debug info location handling"

Originally committed in r224385 and reverted in r224441 due to concerns
this change might've introduced a crash. Turns out this change fixes the
crash introduced by one of my earlier more specific location handling
changes (those specific fixes are reverted by this patch, in favor of
the more general solution).

Recommitted in r224941 and reverted in r224970 after it caused a crash
when building compiler-rt. Looks to be due to this change zeroing out
the debug location when emitting default arguments (which were meant to
inherit their outer expression's location) thus creating call
instructions without locations - these create problems for inlining and
must not be created. That is fixed and tested in this version of the
change.

Original commit message:

This is a more scalable (fixed in mostly one place, rather than many
places that will need constant improvement/maintenance) solution to
several commits I've made recently to increase source fidelity for
subexpressions.

This resetting had to be done at the DebugLoc level (not the
SourceLocation level) to preserve scoping information (if the resetting
was done with CGDebugInfo::EmitLocation, it would've caused the tail end
of an expression's codegen to end up in a potentially different scope
than the start, even though it was at the same source location). The
drawback to this is that it might leave CGDebugInfo out of sync. Ideally
CGDebugInfo shouldn't have a duplicate sense of the current
SourceLocation, but for now it seems it does... - I don't think I'm
going to tackle removing that just now.

I expect this'll probably cause some more buildbot fallout & I'll
investigate that as it comes up.

Also these sort of improvements might be starting to show a weakness/bug
in LLVM's line table handling: we don't correctly emit is_stmt for
statements, we just put it on every line table entry. This means one
statement split over multiple lines appears as multiple 'statements' and
two statements on one line (without column info) are treated as one
statement.

I don't think we have any IR representation of statements that would
help us distinguish these cases and identify the beginning of each
statement - so that might be something we need to add (possibly to the
lexical scope chain - a scope for each statement). This does cause some
problems for GDB and possibly other DWARF consumers.

llvm-svn: 225956
2015-01-14 07:38:27 +00:00
David Blaikie 47d28e07c1 DebugInof: Correct the location of exception cleanups in global ctors/dtors and ObjC methods
Without setting the CurEHLocation these cleanups would be attributed to
whatever the last active debug line location was (the 'fn' call in the
included test cases). By setting CurEHLocation correctly the line
information is improved/corrected.

This quality bug turned into a crasher with r225000 when, instead of
allowing the last location to persist, it would be zero'd out. This
could lead to a function call (such as the dtor) being made without a
debug location - if that call was subsequently inlined (and the caller
and callee had debug info, just not the call instruction) the inliner
would violate important constraints about the debug location chains by
not updating the inlined instructions to chain up to the callee
locations.

So, by fixing this bug, I am addressing the assertion failures
introduced by r225000 and should be able to recommit that patch with
impunity...

llvm-svn: 225955
2015-01-14 07:10:46 +00:00
Alexey Samsonov 8845952b54 Reimplement -fsanitize-recover family of flags.
Introduce the following -fsanitize-recover flags:
  - -fsanitize-recover=<list>: Enable recovery for selected checks or
      group of checks. It is forbidden to explicitly list unrecoverable
      sanitizers here (that is, "address", "unreachable", "return").
  - -fno-sanitize-recover=<list>: Disable recovery for selected checks or
     group of checks.
  - -f(no-)?sanitize-recover is now a synonym for
    -f(no-)?sanitize-recover=undefined,integer and will soon be deprecated.

These flags are parsed left to right, and mask of "recoverable"
sanitizer is updated accordingly, much like what we do for -fsanitize= flags.
-fsanitize= and -fsanitize-recover= flag families are independent.

CodeGen change: If there is a single UBSan handler function, responsible
for implementing multiple checks, which have different recoverable setting,
then we emit two handler calls instead of one:
the first one for the set of "unrecoverable" checks, another one - for
set of "recoverable" checks. If all checks implemented by a handler have the
same recoverability setting, then the generated code will be the same.

llvm-svn: 225719
2015-01-12 22:39:12 +00:00
Rafael Espindola 0d4fb98504 [patch][pr19848] Produce explicit comdats in clang.
The llvm IR until recently had no support for comdats. This was a problem when
targeting C++ on ELF/COFF as just using weak linkage would cause quite a bit of
dead bits to remain on the executable (unless -ffunction-sections,
-fdata-sections and --gc-sections were used).

To fix the problem, llvm's codegen will just assume that any weak or linkonce
that is not in an explicit comdat should be output in one with the same name as
the global.

This unfortunately breaks cases like pr19848 where a weak symbol is not
xpected to be part of any comdat.

Now that we have explicit comdats in the IR, we can finally get both cases
right.

This first patch just makes clang give explicit comdats to GlobalValues where
t is allowed to.

A followup patch to llvm will then stop implicitly producing comdats.

llvm-svn: 225705
2015-01-12 22:13:53 +00:00
Richard Smith dbc8249224 Fix temporary lifetime extension from an initializer using braced "functional"
cast notation T{...} when T is a reference type.

llvm-svn: 225571
2015-01-10 01:28:13 +00:00
Hans Wennborg 43a0f99b10 Don't emit implicit template instantiations eagerly (PR21718)
Their linkage can change if they are later explicitly instantiated. We would
previously emit such functions eagerly (as opposed to lazily on first use) if
they have a 'dllexport' or 'used' attribute, and fail an assert when hitting the
explicit instantiation.

This is achieved by replacing the old CodeGenModule::MayDeferGeneration() method
with two new ones: MustBeEmitted() and MayBeEmittedEagerly().

Differential Revision: http://reviews.llvm.org/D6674

llvm-svn: 225570
2015-01-10 01:19:48 +00:00
David Blaikie 9cdf582f72 Remove unnecessary/incorrect XFAIL after the revert of 225000
llvm-svn: 225561
2015-01-09 23:38:45 +00:00
David Blaikie f353d3ecd0 Revert "DebugInfo: Generalize debug info location handling" and related commits
This reverts commit r225000, r225021, r225083, r225086, r225090.

The root change (r225000) still has several issues where it's caused
calls to be emitted without debug locations. This causes assertion
failures if/when those calls are inlined.

I'll work up some test cases and fixes before recommitting this.

llvm-svn: 225555
2015-01-09 23:00:28 +00:00
Duncan P. N. Exon Smith 4bbe428cc5 IR: Add 'distinct' MDNodes to bitcode and assembly (clang)
Update testcases for LLVM change in r225474 to make `MDNode`s explicitly
distinct (when they aren't uniqued).

Part of PR22111.

llvm-svn: 225475
2015-01-08 22:39:28 +00:00
Nico Weber c1d42ad99a Slightly simplify the test from r225361.
Shorter and doesn't need -O2 -- but still suboptimal as it's still doing
-emit-obj.  dblaikie says he'll improve this when he'll reland his change
with a fix.

llvm-svn: 225364
2015-01-07 18:47:51 +00:00
Nico Weber eac50037fb Revert r225085, it caused PR22096.
PR22096 has several test cases that assert that look fairly different. I'm
adding one of those as an automated test, but when relanding the other cases
should probably be checked as well.

llvm-svn: 225361
2015-01-07 18:23:08 +00:00
Adrian Prantl 0772dbd125 Debug info: pass in the correct size for a pointer to a member function.
This corrects a bug I introduced in r224781.

llvm-svn: 225359
2015-01-07 17:49:30 +00:00
Rafael Espindola c7bd85424c Suppress clang warnings in a codegen test.
This makes the output of FileCheck way easier to read since this test hits
many warnings.

llvm-svn: 225322
2015-01-06 23:53:13 +00:00
Rafael Espindola c0f4a30ee7 Update for .ll syntax change.
llvm-svn: 225303
2015-01-06 22:55:40 +00:00
Rafael Espindola 58d4c5a9c0 Make this test a bit stricter.
The first function is named __cxx_global_var_init, which is a substring of
the following functions @__cxx_global_var_init(1,2,3,etc).

llvm-svn: 225191
2015-01-05 18:48:18 +00:00
Rafael Espindola 6f39d8525c Disable warnings in this test.
This makes it a lot easier to read the output from FileCheck when it fails.

llvm-svn: 225190
2015-01-05 18:19:35 +00:00
David Blaikie ba90b04b7b DebugInfo: Fix cases where location failed to be updated after r225000
The optimization (that appears to have been here since the earliest
implementation (r50848) & has become more complicated over the years) to
avoid recreating the debugloc if it would be the same was out of date
because ApplyDebugLocation was not re-updating the CurLoc/PrevLoc. This
optimization doesn't look terribly beneficial/necessary, so I'm removing
it - if it turns up in benchmarks, I'm happy to reconsider/reimplement
this with justification, but for now it just seems to add
complexity/problems.

llvm-svn: 225083
2015-01-02 19:06:25 +00:00
David Majnemer 602cfe7d62 CodeGen: Don't crash when a lambda uses a local constexpr variable
The DeclRefExpr might be for a variable initialized by a constant
expression which hasn't been ODR used.
Emit the initializer for the variable instead of trying to capture the
variable itself.

This fixes PR22071.

llvm-svn: 225060
2015-01-01 09:49:44 +00:00
David Blaikie 508d29d5b7 XFAIL test on win32 due to missing __complex support
llvm-svn: 225051
2014-12-31 22:30:31 +00:00
David Blaikie f936984e33 Handle PPC64 return type (signext i32 rather than plain i32) in test case
llvm-svn: 225021
2014-12-31 00:06:08 +00:00
David Blaikie 84fe79cfc3 Reapply "DebugInfo: Generalize debug info location handling"
Originally committed in r224385 and reverted in r224441 due to concerns
this change might've introduced a crash. Turns out this change fixes the
crash introduced by one of my earlier more specific location handling
changes (those specific fixes are reverted by this patch, in favor of
the more general solution).

Recommitted in r224941 and reverted in r224970 after it caused a crash
when building compiler-rt. Looks to be due to this change zeroing out
the debug location when emitting default arguments (which were meant to
inherit their outer expression's location) thus creating call
instructions without locations - these create problems for inlining and
must not be created. That is fixed and tested in this version of the
change.

Original commit message:

This is a more scalable (fixed in mostly one place, rather than many
places that will need constant improvement/maintenance) solution to
several commits I've made recently to increase source fidelity for
subexpressions.

This resetting had to be done at the DebugLoc level (not the
SourceLocation level) to preserve scoping information (if the resetting
was done with CGDebugInfo::EmitLocation, it would've caused the tail end
of an expression's codegen to end up in a potentially different scope
than the start, even though it was at the same source location). The
drawback to this is that it might leave CGDebugInfo out of sync. Ideally
CGDebugInfo shouldn't have a duplicate sense of the current
SourceLocation, but for now it seems it does... - I don't think I'm
going to tackle removing that just now.

I expect this'll probably cause some more buildbot fallout & I'll
investigate that as it comes up.

Also these sort of improvements might be starting to show a weakness/bug
in LLVM's line table handling: we don't correctly emit is_stmt for
statements, we just put it on every line table entry. This means one
statement split over multiple lines appears as multiple 'statements' and
two statements on one line (without column info) are treated as one
statement.

I don't think we have any IR representation of statements that would
help us distinguish these cases and identify the beginning of each
statement - so that might be something we need to add (possibly to the
lexical scope chain - a scope for each statement). This does cause some
problems for GDB and possibly other DWARF consumers.

llvm-svn: 225000
2014-12-30 19:39:33 +00:00
David Blaikie 608a24501c Revert "DebugInfo: Generalize debug info location handling"
Asserting when building compiler-rt when using a GCC host compiler.
Reverting while I investigate.

This reverts commit r224941.

llvm-svn: 224970
2014-12-29 23:49:00 +00:00
David Blaikie 4d5c7288cc PR22051: Missing debug location on calls in dtor thunks in Windows.
llvm-svn: 224963
2014-12-29 22:53:52 +00:00
David Blaikie 3945d1bd99 Reapply "DebugInfo: Generalize debug info location handling"
Originally committed in r224385 and reverted in r224441 due to concerns
this change might've introduced a crash. Turns out this change fixes the
crash introduced by one of my earlier more specific location handling
changes (those specific fixes are reverted by this patch, in favor of
the more general solution).

Original commit message:

This is a more scalable (fixed in mostly one place, rather than many
places that will need constant improvement/maintenance) solution to
several commits I've made recently to increase source fidelity for
subexpressions.

This resetting had to be done at the DebugLoc level (not the
SourceLocation level) to preserve scoping information (if the resetting
was done with CGDebugInfo::EmitLocation, it would've caused the tail end
of an expression's codegen to end up in a potentially different scope
than the start, even though it was at the same source location). The
drawback to this is that it might leave CGDebugInfo out of sync. Ideally
CGDebugInfo shouldn't have a duplicate sense of the current
SourceLocation, but for now it seems it does... - I don't think I'm
going to tackle removing that just now.

I expect this'll probably cause some more buildbot fallout & I'll
investigate that as it comes up.

Also these sort of improvements might be starting to show a weakness/bug
in LLVM's line table handling: we don't correctly emit is_stmt for
statements, we just put it on every line table entry. This means one
statement split over multiple lines appears as multiple 'statements' and
two statements on one line (without column info) are treated as one
statement.

I don't think we have any IR representation of statements that would
help us distinguish these cases and identify the beginning of each
statement - so that might be something we need to add (possibly to the
lexical scope chain - a scope for each statement). This does cause some
problems for GDB and possibly other DWARF consumers.

llvm-svn: 224941
2014-12-29 18:18:45 +00:00
David Majnemer 3087a2b949 Sema: Permit an atomic type to be initialized by the same atomic type
We forgot a conversion step when initializing an atomic type with an
rvalue of the same type.

This fixes PR22043.

llvm-svn: 224902
2014-12-28 21:47:31 +00:00
Adrian Prantl ee24e14bdc Debug Info: Pass the pointer size into createMemberPointerType().
Paired commit with LLVM.

llvm-svn: 224781
2014-12-23 19:11:54 +00:00
Nico Weber 08ef80f4b8 Rename test.cc files to test.cpp.
The lit.cfg files only add .cpp to suffixes, so these tests used to never run,
oops.  (Also tweak to of these tests in minor ways to make the actually pass.)

llvm-svn: 224718
2014-12-22 18:13:07 +00:00
Hans Wennborg 0c0a8c868c Tweak the assert in ModuleBuilder from r224533 (PR21989)
Turns out there will be left-over deferred inline methods if there have
been errors, because in that case HandleTopLevelDecl bails out early.

llvm-svn: 224649
2014-12-19 23:35:11 +00:00
Alexey Bataev 7cb1789011 Fix for PR21915: assert on multidimensional VLA in function arguments.
Fixed assertion on type checking for arguments and parameters on function call if arguments are pointers to VLA
Differential Revision: http://reviews.llvm.org/D6655

llvm-svn: 224504
2014-12-18 06:54:53 +00:00
Eric Christopher 560cc4fb44 Make sure that arm-linux-gnu is still the apcs-gnu ABI when we
use clang -cc1 matching the front end and backend. Fix up a couple
of tests that were testing aapcs for arm-linux-gnu.

The test that removes the aapcs abi calling convention removes
them because the default triple matches what the backend uses
for the calling convention there and so it doesn't need to be
explicitly stated - see the code in TargetInfo.cpp.

llvm-svn: 224491
2014-12-18 02:08:55 +00:00
Duncan P. N. Exon Smith c8ee63e2d0 CGDebugInfo: Use DIBuilder API for self-referencing DICompositeTypes
Use new `DIBuilder` API from LLVM r224482 to mutate `DICompositeType`s,
rather than changing them directly.  This allows `DIBuilder` to track
otherwise orphaned cycles when `CollectContainingType()` creates a
self-reference.

Fixes PR21941.

llvm-svn: 224483
2014-12-18 00:48:56 +00:00