Reid Kleckner [Thu, 16 Nov 2017 19:09:36 +0000 (19:09 +0000)]
[MS] Apply adjustments after storing 'this'
Summary:
The MS ABI convention is that the 'this' pointer on entry is the address
of the vfptr that was used to make the virtual method call. In other
words, the pointer on entry always points to the base subobject that
introduced the virtual method. Consider this hierarchy:
struct A { virtual void f() = 0; };
struct B { virtual void g() = 0; };
struct C : A, B {
void f() override;
void g() override;
};
On entry to C::g, [ER]CX will contain the address of C's B subobject,
and C::g will have to subtract sizeof(A) to recover a pointer to C.
Before this change, we applied this adjustment in the prologue and
stored the new value into the "this" local variable alloca used for
debug info. However, MSVC does not do this, presumably because it is
often profitable to fold the adjustment into later field accesses. This
creates a problem, because the debugger expects the variable to be
unadjusted. Unfortunately, CodeView doesn't have anything like DWARF
expressions for computing variables that aren't in the program anymore,
so we have to declare 'this' to be the unadjusted value if we want the
debugger to see the right value.
This has the side benefit that, in optimized builds, the 'this' pointer
will usually be available on function entry because it doesn't require
any adjustment.
Zachary Turner [Thu, 16 Nov 2017 18:26:20 +0000 (18:26 +0000)]
Resubmit "Refactor debuginfo-tests"
This was reverted due to some failures on specific darwin buildbots,
the issue being that the new lit configuration was not setting the
SDKROOT environment variable. We've tested a fix locally and confirmed
that it works, so this patch resubmits everything with the fix
applied.
Ilya Biryukov [Thu, 16 Nov 2017 16:25:01 +0000 (16:25 +0000)]
Allow to store precompiled preambles in memory.
Summary:
These preambles are built by ASTUnit and clangd. Previously, preambles
were always stored on disk.
In-memory preambles are routed back to the compiler as virtual files in
a custom VFS.
Interface of ASTUnit does not allow to use in-memory preambles, as
ASTUnit::CodeComplete receives FileManager as a parameter, so we can't
change VFS used by the compiler inside the CodeComplete method.
A follow-up commit will update clangd in clang-tools-extra to use
in-memory preambles.
Erich Keane [Wed, 15 Nov 2017 22:25:39 +0000 (22:25 +0000)]
Split x86 "Processor" info into its own def file. [NFC]
A first step toward removing the repetition of
features/CPU info in the x86 target info, this
patch pulls all the processor information out into
its own .def file.
Hans Wennborg [Wed, 15 Nov 2017 17:11:53 +0000 (17:11 +0000)]
BuiltinOperatorOverloadBuilder: Don't consider types that are unavailable on the target (PR35174)
In the PR, Clang ended up in a situation where it tried to mangle the
__float128 type, which isn't supported when targetingt MSVC, because
Clang instantiated a variable template with that type when searching for
a conversion to use in an arithmetic expression.
David Blaikie [Wed, 15 Nov 2017 16:52:12 +0000 (16:52 +0000)]
ASTMatchers.h: Fix ODR violations by avoiding internal linkage variables in headers
Internal linkage variables ODR referenced from inline functions create
ODR violations (the same inline function ends up having different
definitions in each TU, since it references different variables - rather
than one definition).
This also happens to break modular code generation - so this is the last
fix to allow clang to compile with modular code generation.
Alexey Bader [Wed, 15 Nov 2017 11:38:17 +0000 (11:38 +0000)]
[OpenCL] Fix code generation of function-scope constant samplers.
Summary:
Constant samplers are handled as static variables and clang's code generation
library, which leads to llvm::unreachable. We bypass emitting sampler variable
as static since it's translated to a function call later.
Richard Smith [Wed, 15 Nov 2017 01:33:46 +0000 (01:33 +0000)]
[modules] Fix crash in complex class merging scenario.
When we merge together class definitions, we can end up with the canonical
declaration of a field not being the one that was lexically within the
canonical definition of the class. Additionally, when we merge class
definitions via update records (eg, for a template specialization whose
declaration is instantiated in one module and whose definition is instantiated
in multiple others), we can end up with the list of lexical contents for the
class not including a particular declaration of a field whose lexical parent is
that class definition. In the worst case, we have a field whose canonical
declaration's lexical parent has no fields, and in that case this attempt to
number the fields by walking the fields in the declaration of the class that
contained one of the canonical fields will fail.
Instead, when numbering fields in a class, do the obvious thing: walk the
fields in the definition.
I'm still trying to reduce a testcase; the setup that leads to the above
scenario seems to be quite fragile.
Vedant Kumar [Tue, 14 Nov 2017 23:56:53 +0000 (23:56 +0000)]
[PGO] Detect more structural changes with the stable hash
Lifting from Bob Wilson's notes: The hash value that we compute and
store in PGO profile data to detect out-of-date profiles does not
include enough information. This means that many significant changes to
the source will not cause compiler warnings about the profile being out
of date, and worse, we may continue to use the outdated profile data to
make bad optimization decisions. There is some tension here because
some source changes won't affect PGO and we don't want to invalidate the
profile unnecessarily.
This patch adds a new hashing scheme which is more sensitive to loop
nesting, conditions, and out-of-order control flow. Here are examples
which show snippets which get the same hash under the current scheme,
and different hashes under the new scheme:
Loop Nesting Example
--------------------
// Snippet 1
while (foo()) {
while (bar()) {}
}
// Snippet 2
while (foo()) {}
while (bar()) {}
Condition Example
-----------------
// Snippet 1
if (foo())
bar();
baz();
// Snippet 2
if (foo())
bar();
else
baz();
Out-of-order Control Flow Example
---------------------------------
// Snippet 1
while (foo()) {
if (bar()) {}
baz();
}
// Snippet 2
while (foo()) {
if (bar())
continue;
baz();
}
In each of these cases, it's useful to differentiate between the
snippets because swapping their profiles gives bad optimization hints.
The new hashing scheme considers some logical operators in an effort to
detect more changes in conditions. This isn't a perfect scheme. E.g, it
does not produce the same hash for these equivalent snippets:
// Snippet 1
bool c = !a || b;
if (d && e) {}
// Snippet 2
bool f = d && e;
bool c = !a || b;
if (f) {}
This would require an expensive data flow analysis. Short of that, the
new hashing scheme looks reasonably complete, based on a scan over the
statements we place counters on.
Profiles which use the old version of the PGO hash remain valid and can
be used without issue (there are tests in tree which check this).
Hans Wennborg [Tue, 14 Nov 2017 21:13:27 +0000 (21:13 +0000)]
Switch -mcount and -finstrument-functions to emit EnterExitInstrumenter attributes
This updates -mcount to use the new attribute names (LLVM r318195), and
switches over -finstrument-functions to also use these attributes rather
than inserting instrumentation in the frontend.
It also adds a new flag, -finstrument-functions-after-inlining, which
makes the cygprofile instrumentation get inserted after inlining rather
than before.
Dave Lee [Tue, 14 Nov 2017 14:17:26 +0000 (14:17 +0000)]
Make isDefinition matcher support ObjCMethodDecl
Summary:
Allow the `isDefinition()` matcher to apply to `ObjCMethodDecl` nodes, in
addition to those it already supports. For whatever reason, `ObjCMethodDecl`
does not inherit from `FunctionDecl` and so this is specialization is necessary.
Erik Verbruggen [Tue, 14 Nov 2017 09:34:39 +0000 (09:34 +0000)]
[libclang] Allow crash recovery with LIBCLANG_NOTHREADS
Enabled crash recovery for some libclang operations on a calling thread even
when LIBCLANG_NOTHREAD is specified.
Previously it would only run under crash recovery if LIBCLANG_NOTHREAD is not
set. Moved handling of LIBCLANG_NOTHREAD env variable into RunSafely from its
call sites.
Create more orthogonal pieces. The restructuring made it easy to try out
several alternatives to D33589, and while none of the alternatives
turned out to be the right solution, the underlying simplification of
the structure is helpful.
Craig Topper [Tue, 14 Nov 2017 08:48:28 +0000 (08:48 +0000)]
[NewPassManager] Pass the -fdebug-pass-manager flag setting into the Analysis managers to match what we do in opt
Summary: Currently the -fdebug-pass-manager flag for clang doesn't enable the debug logging in the analysis managers. This is different than what the switch does when passed to opt.
Chandler Carruth [Tue, 14 Nov 2017 01:59:18 +0000 (01:59 +0000)]
[PM] Wire up support for the bounds checking sanitizer with the new PM.
Not much interesting here. Mostly wiring things together.
One thing worth noting is that the approach is substantially different
from the old PM. Here, the -O0 case works fundamentally differently in
that we just directly build the pipeline without any callbacks or other
cruft. In some ways, this is nice and clean. However, I don't like that
it causes the sanitizers to be enabled with different changes at
different times. =/ Suggestions for a better way to do this are welcome.
Zachary Turner [Mon, 13 Nov 2017 23:33:29 +0000 (23:33 +0000)]
Revert "Update test_debuginfo.pl script to point to new tree location."
This reverts the aforementioned patch and 2 subsequent follow-ups,
as some buildbots are still failing 2 tests because of it.
Investigation is ongoing into the cause of the failures.
Sanjay Patel [Mon, 13 Nov 2017 22:11:49 +0000 (22:11 +0000)]
[CodeGen] fix const-ness of cbrt and fma
cbrt() is always constant because it can't overflow or underflow. Therefore, it can't set errno.
fma() is not always constant because it can overflow or underflow. Therefore, it can set errno.
But we know that it never sets errno on GNU / MSVC, so make it constant in those environments.
Devin Coughlin [Mon, 13 Nov 2017 17:35:29 +0000 (17:35 +0000)]
[analyzer] ObjCGenerics: Don't warn on cast conversions involving explicit cast
The ObjCGenerics checker warns on a cast when there is no subtyping relationship
between the tracked type of the value and the destination type of the cast. It
does this even if the cast was explicitly written. This means the user can't
write an explicit cast to silence the diagnostic.
This commit treats explicit casts involving generic types as an indication from
the programmer that the Objective-C type system is not rich enough to express
the needed invariant. On explicit casts, the checker now removes any existing
information inferred about the type arguments. Further, it no longer assumes
the casted-to specialized type because the invariant the programmer specifies
in the cast may only hold at a particular program point and not later ones. This
prevents a suppressing cast from requiring a cascade of casts down the
line.
Teresa Johnson [Mon, 13 Nov 2017 15:38:33 +0000 (15:38 +0000)]
[ThinLTO] Handle -fdebug-pass-manager for backend invocations via clang
Recommit of r317951 and r317951 along with what I believe should fix
the remaining buildbot failures - the target triple should be specified
for both the ThinLTO pre-thinlink compile and backend (post-thinlink)
compile to ensure it is consistent.
Original description:
The LTO Config field wasn't being set when invoking a ThinLTO backend
via clang (i.e. for distributed builds).
Michal Gorny [Sat, 11 Nov 2017 20:01:41 +0000 (20:01 +0000)]
[python] [tests] Fix test_linkage for unique external linkage
Starting with r314037, anonymous namespaces no longer give
unique-external linkage to variables. However, this linkage can still be
achieved by using a type which is not exterally visible,
e.g. through being declared in an anonymous namespace but used outside
it. Fix the test to take advantage of that.
Faisal Vali [Sat, 11 Nov 2017 18:02:29 +0000 (18:02 +0000)]
Adjust r316292 - remove the anonymous union for sharing a bitfield in FunctionDecl.
The anonymous union did NOT save us storage, but instead behaved as if we added an additional integer data member to FunctionDecl.
For additional context, the anonymous union renders the bit fields as non-adjacent and prevents them from sharing the same 'memory location' (i.e. bit-storage) by requiring the anonymous union object to be appropriately aligned.
This was confirmed through discussion with Richard Smith in Albuquerque (ISO C++ Meeting)
Gor Nishanov [Sat, 11 Nov 2017 17:00:43 +0000 (17:00 +0000)]
[coroutines] Promote cleanup.dest.slot allocas to registers to avoid storing it in the coroutine frame
Summary:
We don't want to store cleanup dest slot saved into the coroutine frame (as some of the cleanup code may
access them after coroutine frame destroyed).
This is an alternative to https://reviews.llvm.org/D37093
It is possible to do this for all functions, but, cursory check showed that in -O0, we get slightly longer function (by 1-3 instructions), thus, we are only limiting cleanup.dest.slot elimination to coroutines.
Jake Ehrlich [Sat, 11 Nov 2017 01:15:41 +0000 (01:15 +0000)]
Add CLANG_DEFAULT_OBJCOPY to allow Clang to use llvm-objcopy for dwarf fission
llvm-objcopy is getting to where it can be used in non-trivial ways
(such as for dwarf fission in clang). It now supports dwarf fission but
this feature hasn't been thoroughly tested yet. This change allows
people to optionally build clang to use llvm-objcopy rather than GNU
objcopy. By default GNU objcopy is still used so nothing should change.
Richard Trieu [Sat, 11 Nov 2017 00:54:25 +0000 (00:54 +0000)]
Handle lambda captures of variable length arrays in profiling and printing.
From http://reviews.llvm.org/D4368 these cases were thought to not be reachable
and the checks removed before the rest of the code was committed in r216649.
However, these cases are reachable and the checks are added back.
Zachary Turner [Fri, 10 Nov 2017 20:57:57 +0000 (20:57 +0000)]
[debuginfo-tests] Make debuginfo-tests work in a standard configuration.
Previously, debuginfo-tests was expected to be checked out into
clang/test and then the tests would automatically run as part of
check-clang. This is not a standard workflow for handling
external projects, and it brings with it some serious drawbacks
such as the inability to depend on things other than clang, which
we will need going forward.
The goal of this patch is to migrate towards a more standard
workflow. To ease the transition for build bot maintainers,
this patch tries not to break the existing workflow, but instead
simply deprecate it to give maintainers a chance to update
the build infrastructure.
Michal Gorny [Fri, 10 Nov 2017 16:44:12 +0000 (16:44 +0000)]
[python] [tests] Rewrite to use standard unittest module
Rewrite the tests from using plain 'assert' mixed with some nosetests
methods to the standard unittest module layout. Improve the code
to use the most canonical assertion methods whenever possible.
This has a few major advantages:
- the code uses standard methods now, resulting in a reduced number
of WTFs whenever someone with basic Python knowledge gets to read it,
- completely unnecessary dependency on nosetests is removed since
the standard library supplies all that is necessary for the tests
to run,
- the tests can be run via any test runner, including the one built-in
in Python,
- the failure output for most of the tests is improved from 'assertion
x == y failed' to actually telling the values.
Alexey Bataev [Fri, 10 Nov 2017 15:39:50 +0000 (15:39 +0000)]
[OpenMP] diagnose assign to firstprivate const, patch by Joel E. Denny
Summary:
[OpenMP] diagnose assign to firstprivate const
Clang does not diagnose assignments to const variables declared
firstprivate. Furthermore, codegen is broken such that, at run time,
such assignments simply have no effect. For example, the following
prints 0 not 1:
int main() {
const int i = 0;
#pragma omp parallel firstprivate(i)
{ i=1; printf("%d\n", i); }
return 0;
}
This commit makes these assignments a compile error, which is
consistent with other OpenMP compilers I've tried (pgcc 17.4-0, gcc
6.3.0).
Summary:
This change implements the changes required in both clang and
compiler-rt to allow building XRay-instrumented binaries in Darwin. For
now we limit this to x86_64. We also start building the XRay runtime
library in compiler-rt for osx.
A caveat to this is that we don't have the tests set up and running
yet, which we'll do in a set of follow-on changes.
This patch uses the monorepo layout for the coordinated change across
multiple projects.
Craig Topper [Fri, 10 Nov 2017 05:20:32 +0000 (05:20 +0000)]
[X86] Reduce the number of FMA builtins needed by the frontend by adding negates to operands of the fmadd and fmaddsub builtins.
The backend should be able to combine the negates to create fmsub, fnmadd, and fnmsub. faddsub converting to fsubadd still needs work I think, but should be very doable.
This matches what we already do for the masked builtins.
This only covers the packed builtins. Scalar builtins will be done after FMA4 is fixed.
Steven Wu [Fri, 10 Nov 2017 01:32:47 +0000 (01:32 +0000)]
[Driver] Make clang/cc conforms to UNIX standard
Summary:
This is basically reverting r261774 with a tweak for clang-cl. UNIX
standard states:
When c99 encounters a compilation error that causes an object file not
to be created, it shall write a diagnostic to standard error and
continue to compile other source code operands, but it shall not perform
the link phase and it shall return a non-zero exit status
The same goes for c89 or cc. And they are all alias or shims pointing to
clang on Darwin.
The original commit was intended for CUDA so the error message doesn't
get emit twice for both host and device. It seems that the clang driver
has been changed to model the CUDA dependency differently. Now the
driver behaves the same without this commit.
[analyzer] do not crash when trying to convert an APSInt to an unexpected type
This is the issue breaking the postgresql bot, purely by chance exposed
through taint checker, somehow appearing after
https://reviews.llvm.org/D38358 got committed.
The backstory is that the taint checker requests SVal for the value of
the pointer, and analyzer has a "fast path" in the getter to return a
constant when we know that the value is constant.
Unfortunately, the getter requires a cast to get signedness correctly,
and for the pointer `void *` the cast crashes.
This is more of a band-aid patch, as I am not sure what could be done
here "correctly", but it should be applied in any case to avoid the
crash.
Patches the solver to assume that bitwise OR of an unsigned value with a
constant always produces a value larger-or-equal than the constant, and
bitwise AND with a constant always produces a value less-or-equal than
the constant.
This patch is especially useful in the context of using bitwise
arithmetic for error code encoding: the analyzer would be able to state
that the error code produced using a bitwise OR is non-zero.
Ben Hamilton [Thu, 9 Nov 2017 16:01:16 +0000 (16:01 +0000)]
[VirtualFileSystem] InMemoryFileSystem::addFile(): Type and Perms
Summary:
This implements a FIXME in InMemoryFileSystem::addFile(), allowing
clients to specify User, Group, Type, and/or Perms when creating a
file in an in-memory filesystem.
projects / clang / log