In the future the compiler will analyze whether the OpenMP
runtime needs to be (fully) initialized and avoid that overhead
if possible. The functions already take an argument to transfer
that information to the runtime, so pass in the default value 1.
(This is needed for binary compatibility with libomptarget-nvptx
currently being upstreamed.)
Differential Revision: https://reviews.llvm.org/D40354
llvm-svn: 318836
The support for relax relocations is dependent on the linker and
different toolchains within the same compiler can be using different
linkers some of which may or may not support relax relocations.
Give toolchains the option to control whether they want to use relax
relocations in addition to the existing (global) build system option.
Differential Revision: https://reviews.llvm.org/D39831
llvm-svn: 318816
This clang patch changes the __tgt_* API function signatures in preparation for the new map interface.
Changes are: Device IDs 32bits --> 64bits, Flags 32bits --> 64bits
Differential revision: https://reviews.llvm.org/D40281
llvm-svn: 318789
This is an instrumentation flag that's similar to
-finstrument-functions, but it only inserts calls on function entry, the
calls are inserted post-inlining, and they don't take any arugments.
This is intended for users who want to instrument function entry with
minimal overhead.
(-pg would be another alternative, but forces frame pointer emission and
affects link flags, so is probably best left alone to be used for
generating gcov data.)
Differential revision: https://reviews.llvm.org/D40276
llvm-svn: 318785
Summary:
This patch is part of the development effort to add support in the current OpenMP GPU offloading implementation for implicitly sharing variables between a target region executed by the team master thread and the worker threads within that team.
This patch is the first of three required for successfully performing the implicit sharing of master thread variables with the worker threads within a team. The remaining two patches are:
- Patch D38978 to the LLVM NVPTX backend which ensures the lowering of shared variables to an device memory which allows the sharing of references;
- Patch (coming soon) is a patch to libomptarget runtime library which ensures that a list of references to shared variables is properly maintained.
A simple code snippet which illustrates an implicit data sharing situation is as follows:
```
#pragma omp target
{
// master thread only
int v;
#pragma omp parallel
{
// worker threads
// use v
}
}
```
Variable v is implicitly shared from the team master thread which executes the code in between the target and parallel directives. The worker threads must operate on the latest version of v, including any updates performed by the master.
The code generated in this patch relies on the LLVM NVPTX patch (mentioned above) which prevents v from being lowered in the thread local memory of the master thread thus making the reference to this variable un-shareable with the workers. This ensures that the code generated by this patch is correct.
Since the parallel region is outlined the passing of arguments to the outlined regions must preserve the original order of arguments. The runtime therefore maintains a list of references to shared variables thus ensuring their passing in the correct order. The passing of arguments to the outlined parallel function is performed in a separate function which the data sharing infrastructure constructs in this patch. The function is inlined when optimizations are enabled.
Reviewers: hfinkel, carlo.bertolli, arpith-jacob, Hahnfeld, ABataev, caomhin
Reviewed By: ABataev
Subscribers: cfe-commits, jholewinski
Differential Revision: https://reviews.llvm.org/D38976
llvm-svn: 318773
This was previously done in some places, but for example not for
bundling so that single object compilation with -c failed. In
addition cubin was used for all file types during unbundling which
is incorrect for assembly files that are passed to ptxas.
Tighten up the tests so that we can't regress in that area.
Differential Revision: https://reviews.llvm.org/D40250
llvm-svn: 318763
move _WIN64 and _WIN32 defines to lib/Basic/Targets/OSTargets.h
move WIN32, WIN64 and __MINGW64__ to addMinGWDefines
fixes __MINGW64__ not being defined for aarch64
adds WIN32 definition for x64
Reviewers: mstorsjo
Differential Revision: https://reviews.llvm.org/D40285
llvm-svn: 318755
CFG wass built in non-deterministic order due to the fact that indirect
goto labels' declarations (LabelDecl's) are stored in the llvm::SmallSet
container. LabelDecl's are pointers, whose order is not deterministic,
and llvm::SmallSet sorts them by their non-deterministic addresses after
"small" container is exceeded. This leads to non-deterministic processing
of the elements of the container.
The fix is to use llvm::SmallSetVector that was designed to have
deterministic iteration order.
Patch by Ilya Palachev!
Differential Revision: https://reviews.llvm.org/D40073
llvm-svn: 318754
CFG wass built in non-deterministic order due to the fact that indirect
goto labels' declarations (LabelDecl's) are stored in the llvm::SmallSet
container. LabelDecl's are pointers, whose order is not deterministic,
and llvm::SmallSet sorts them by their non-deterministic addresses after
"small" container is exceeded. This leads to non-deterministic processing
of the elements of the container.
The fix is to use llvm::SmallSetVector that was designed to have
deterministic iteration order.
Patch by Ilya Palachev!
Differential Revision: https://reviews.llvm.org/D40073
llvm-svn: 318750
This implements [dcl.modules.export] from the C++ Modules TS, which lets a module re-export another module with the "export import" syntax.
Differential Revision: https://reviews.llvm.org/D40270
llvm-svn: 318744
Summary:
This raises __STDCPP_DEFAULT_NEW_ALIGNMENT__ from 8 to 16 on Win64.
This matches platforms that follow the usual `2 * sizeof(void*)`
alignment requirement for malloc. We might want to consider making that
the default rather than relying on long double alignment.
Fixes PR35356
Reviewers: STL_MSFT, rsmith
Subscribers: cfe-commits
Differential Revision: https://reviews.llvm.org/D40277
llvm-svn: 318723
This is still breaking greendragon.
At this point I give up until someone can fix the greendragon
bots, and I will probably abandon this effort in favor of using
a private github repository.
llvm-svn: 318722
This diff extends StackAddrEscapeChecker
to catch stack addresses leaks via block captures
if the block is executed asynchronously or
returned from a function.
Differential revision: https://reviews.llvm.org/D39438
llvm-svn: 318705
This was reverted due to the tests being run twice on some
build bots. Each run had a slightly different configuration
due to the way in which it was being invoked. This fixes
the problem (albeit in a somewhat hacky way). Hopefully in
the future we can get rid of the workflow of running
debuginfo-tests as part of clang, and then this hack can
go away.
llvm-svn: 318697
https://reviews.llvm.org/D40187
This patch implements code gen for 'teams distribute parallel for' on the host, including all its clauses and related regression tests.
llvm-svn: 318692
As reported in llvm bugzilla 32377.
Here’s a patch to add preinclude of stdc-predef.h.
The gcc documentation says “On GNU/Linux, <stdc-predef.h> is pre-included.”
See https://gcc.gnu.org/gcc-4.8/porting_to.html;
The preinclude is inhibited with –ffreestanding.
Basically I fixed the failing test cases by adding –ffreestanding which inhibits
this behavior.
I fixed all the failing tests, including some in extra/test, there's a separate
patch for that which is linked here
Patch By: mibintc
Differential Revision: https://reviews.llvm.org/D34158
llvm-svn: 318669
This caused warnings also when the if or else comes from macros. There was an
attempt to fix this in r318556, but that introduced new problems and was
reverted. Reverting this too until the whole issue is sorted.
> This looks like it was just an oversight.
>
> Fixes http://llvm.org/pr35319
>
> git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@318456 91177308-0d34-0410-b5e6-96231b3b80d8
llvm-svn: 318667
It seems this somehow made -Wempty-body fire in some macro cases where
it didn't before, e.g.
../../third_party/ffmpeg/libavcodec/bitstream.c(169,5): error: if statement has empty body [-Werror,-Wempty-body]
ff_dlog(NULL, "new table index=%d size=%d\n", table_index, table_size);
^
../../third_party/ffmpeg\libavutil/internal.h(276,80): note: expanded from macro 'ff_dlog'
# define ff_dlog(ctx, ...) do { if (0) av_log(ctx, AV_LOG_DEBUG, __VA_ARGS__); } while (0)
^
../../third_party/ffmpeg/libavcodec/bitstream.c(169,5): note: put the
semicolon on a separate line to silence this warning
Reverting until this can be figured out.
> Do not show it when `if` or `else` come from macros.
> E.g.,
>
> #define USED(A) if (A); else
> #define SOME_IF(A) if (A)
>
> void test() {
> // No warnings are shown in those cases now.
> USED(0);
> SOME_IF(0);
> }
>
> Patch by Ilya Biryukov!
>
> Differential Revision: https://reviews.llvm.org/D40185
llvm-svn: 318665
The Unified Arm Assembler Language is designed so that the majority of
assembler files can be assembled for both Arm and Thumb with the choice
made as a compilation option.
The way this is done in gcc is to pass -mthumb to the assembler with either
-Wa,-mthumb or -Xassembler -mthumb. This change adds support for these
options to clang. There is no assembler equivalent of -mno-thumb, -marm or
-mno-arm so we don't need to recognize these.
Ideally we would do all of the processing in
CollectArgsForIntegratedAssembler(). Unfortunately we need to change the
triple and at that point it is too late. Instead we look for the option
earlier in ComputeLLVMTriple().
Fixes PR34519
Differential Revision: https://reviews.llvm.org/D40127
llvm-svn: 318647
set -pie as default for musl linux targets
add detection of alpine linux
append appropriate compile flags for alpine
Reviewers: rnk
Differential Revision: https://reviews.llvm.org/D39588
llvm-svn: 318608
Some target devices (e.g. Nvidia GPUs) don't support dynamic stack
allocation and hence no VLAs. Print errors with description instead
of failing in the backend or generating code that doesn't work.
This patch handles explicit uses of VLAs (local variable in target
or declare target region) or implicitly generated (private) VLAs
for reductions on VLAs or on array sections with non-constant size.
Differential Revision: https://reviews.llvm.org/D39505
llvm-svn: 318601
Do not show it when `if` or `else` come from macros.
E.g.,
#define USED(A) if (A); else
#define SOME_IF(A) if (A)
void test() {
// No warnings are shown in those cases now.
USED(0);
SOME_IF(0);
}
Patch by Ilya Biryukov!
Differential Revision: https://reviews.llvm.org/D40185
llvm-svn: 318556
The object is provided by the objc runtime and is never visible in the
module itself, but even so, the address point we compute points into it,
and "+16" is guaranteed not to overflow.
This matches the c++ vtable IRGen.
Note that I'm not entirely convinced the 'i8*' type is correct here: at
the IR level, we're accessing memory that's outside the global object.
But we don't control the allocation, so it's not obviously wrong either.
But either way, this is only in a global initializer, so I don't think
it's going to be mucked with. Filed PR35352 to discuss that.
llvm-svn: 318545
Since SVN r318510, the MinGW/ARM configuration defaults to
dwarf exception handling.
Differential Revision: https://reviews.llvm.org/D39533
llvm-svn: 318511
When mixing PCH and Implicit Modules, missing a header search path
can lead to the implicit built PCM to complaint about not finding its
matching module map.
Instead of adding more magic to implicit modules engine, add a note to
help the user add the appropriate path.
rdar://problem/33388847
llvm-svn: 318503
This is still broken because it causes certain tests to be
run twice with slightly different configurations, which is
wrong in some cases.
You can observe this by running:
ninja -nv check-all | grep debuginfo-tests
And seeing that it passes clang/test and clang/test/debuginfo-tests
to lit, which causes it to run debuginfo-tests twice. The fix is
going to involve either:
a) figuring out that we're running in this "deprecated" configuration,
and then deleting the clang/test/debuginfo-tests path, which should
cause it to behave identically to before, or:
b) make lit smart enough that it doesn't descend into a sub-suite if
that sub-suite already has a lit.cfg file.
llvm-svn: 318486
explicitly instantiated, still emit it with each use.
We don't emit a definition of the member with an explicit instantiation
definition (and indeed it appears that we're not allowed to, since an explicit
instantiation definition does not constitute an odr-use and only odr-use
permits definition for defaulted special members). So we still need to emit a
weak definition with each use.
This also makes defaulted-in-class declarations behave more like
implicitly-declared special members, which matches their design intent.
And it matches the way this problem was solved in GCC.
llvm-svn: 318474
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.
Reviewers: hans
Subscribers: aprantl, cfe-commits
Differential Revision: https://reviews.llvm.org/D40109
llvm-svn: 318440
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.
llvm-svn: 318435
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.
Differential revision: https://reviews.llvm.org/D39579
llvm-svn: 318309
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.
Reviewers: yaxunl, Anastasia
Reviewed By: yaxunl, Anastasia
Subscribers: cfe-commits
Differential Revision: https://reviews.llvm.org/D34342
llvm-svn: 318290
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).
rdar://17068282
Differential Revision: https://reviews.llvm.org/D39446
llvm-svn: 318229
On e.g. PPC the return value and argument were marked 'signext'. This
makes the test expectations a bit more flexible.
Follow-up to r318199.
llvm-svn: 318214
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.
Differential Revision: https://reviews.llvm.org/D39331
llvm-svn: 318199
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.
Differential Revision: https://reviews.llvm.org/D39085
llvm-svn: 318131
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.
llvm-svn: 318112
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.
Differential Revision: https://reviews.llvm.org/D39641
llvm-svn: 318093
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.
rdar://problem/33603303
Differential Revision: https://reviews.llvm.org/D39711
llvm-svn: 318054
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).
llvm-svn: 318042
Change Header files of the intrinsics for lowering test and testn intrinsics to IR code.
Removed test and testn builtins from clang
Differential Revision: https://reviews.llvm.org/D38737
llvm-svn: 318035
This patch, together with a matching llvm patch (https://reviews.llvm.org/D38671), implements the lowering of X86 shuffle i/f intrinsics to IR.
Differential Revision: https://reviews.llvm.org/D38672
Change-Id: I9b3c2f2b34323bd9ccb21d0c1832f848b88ec047
llvm-svn: 318025
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.
Reviewers: rjmccall, hfinkel, eric_niebler
Reviewed By: eric_niebler
Subscribers: EricWF, cfe-commits
Differential Revision: https://reviews.llvm.org/D39768
llvm-svn: 317981
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.
llvm-svn: 317957
Summary:
The LTO Config field wasn't being set when invoking a ThinLTO backend
via clang (i.e. for distributed builds).
Reviewers: danielcdh
Subscribers: mehdi_amini, inglorion, eraman, cfe-commits
Differential Revision: https://reviews.llvm.org/D39923
llvm-svn: 317951
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.
Differential Revision: https://reviews.llvm.org/D39605
llvm-svn: 317925
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).
Reviewers: ABataev
Reviewed By: ABataev
Subscribers: cfe-commits
Differential Revision: https://reviews.llvm.org/D39859
llvm-svn: 317891
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.
llvm-svn: 317873
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.
rdar://problem/32223263
Reviewers: thakis, dexonsmith, tra
Reviewed By: tra
Subscribers: jlebar, cfe-commits
Differential Revision: https://reviews.llvm.org/D39502
llvm-svn: 317860
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.
Differential Revision: https://reviews.llvm.org/D39862
llvm-svn: 317839
The 'decl' role is more canonical than the 'ref'. This helps us establish the
'specialization-of' relation just by looking at decls or defs.
rdar://31884960
llvm-svn: 317832
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.
Differential Revision: https://reviews.llvm.org/D39707
llvm-svn: 317820
Summary:
This is an alternative to JSONCompilationDatabase for simple projects that
don't use a build system such as CMake.
(You can also drop one in ~, to make your tools use e.g. C++11 by default)
There's no facility for varying flags per-source-file or per-machine.
Possibly this could be accommodated backwards-compatibly using cpp, but even if
not the simplicity seems worthwhile for the cases that are addressed.
Tested with clangd, works great! (requires clangd restart)
Reviewers: klimek
Subscribers: ilya-biryukov, cfe-commits
Differential Revision: https://reviews.llvm.org/D39799
llvm-svn: 317777
There are some limitations with emitting regions in macro expansions
because we don't gather file IDs within the expansions. Fix the check
that prevents us from emitting deferred regions in expansions to make an
exception for headers, which is something we can handle.
rdar://35373009
llvm-svn: 317760
The area immediately after a terminated region in the function top-level
should have the same count as the label it precedes.
This solves another problem with wrapped segments. Consider:
1| a:
2| return 0;
3| b:
4| return 1;
Without a gap area starting after the first return, the wrapped segment
from line 2 would make it look like line 3 is executed, when it's not.
rdar://35373009
llvm-svn: 317759
The area immediately after the closing right-paren of an if condition
should have a count equal to the 'then' block's count. Use a gap region
to set this count, so that region highlighting for the 'then' block
remains precise.
This solves a problem we have with wrapped segments. Consider:
1| if (false)
2| foo();
Without a gap area starting after the condition, the wrapped segment
from line 1 would make it look like line 2 is executed, when it's not.
rdar://35373009
llvm-svn: 317758
Block literals can have a type with attributes in its signature, e.g.
ns_returns_retained. The code that inspected the type loc of the block when
declaring its parameters didn't account for this fact, and only looked through
paren type loc. This commit ensures that getAsAdjusted is used instead of
IgnoreParens to find the block's FunctionProtoTypeLoc. This ensures that
block parameters are declared correctly in the block and avoids the
'undeclared identifier' error.
rdar://35416160
llvm-svn: 317736
Objective-C NSString has a class method stringWithUTF8String that creates a new
NSString from a C string. Objective-C box expression @(...) can be used to
create an NSString instead of invoking the stringWithUTF8String method directly
(The compiler lowers it down to the invocation though). This commit ensures that
the type of @(string-value) gets the same nullability attributes as the return
type of stringWithUTF8String to ensure that the diagnostics are consistent
between the two.
rdar://33847186
Differential Revision: https://reviews.llvm.org/D39762
llvm-svn: 317727
This un-breaks builds on other platforms. Otherwise, they fail due to warnings like:
warning: unable to find a Visual Studio installation; try running Clang from a developer command prompt [-Wmsvc-not-found]
llvm-svn: 317716
Summary:
The -coverage option is not a CoreOption, so it is not available to clang-cl.
This patch adds the CoreOption flag to "-coverage" to allow it to be used with clang-cl.
Reviewers: rnk
Reviewed By: rnk
Subscribers: cfe-commits
Differential Revision: https://reviews.llvm.org/D38221
llvm-svn: 317709
Do not crash when trying to compute x && y or x || y where x and y are
of a vector type.
For now we do not seem to properly model operations with vectors. In particular,
operations && and || on a pair of vectors are not short-circuit, unlike regular
logical operators, so even our CFG is incorrect.
Avoid the crash, add respective FIXME tests for later.
Differential Revision: https://reviews.llvm.org/D39682
rdar://problem/34317663
llvm-svn: 317700
Clang's completion output is non-deterministic, causing test failures
with turned on LLVM_REVERSE_ITERATION.
The workaround is to use CHECK-DAGs for now, will remove them when
PR35244 gets fixed.
llvm-svn: 317687
Summary:
Adjusted PrintingPolicy inside code completion to avoid printing some
redundant name qualifiers.
Before this change, typedefs that were written unqualified in source
code were printed with qualifiers in completion. For example, in the
following code
struct foo {
typedef int type;
type method();
};
completion item for `method` had return type of `foo::type`, even
though the original code used `type` without qualifiers.
After this change, the completion item has return type `type`, as
originally written in the source code.
Note that this change does not suppress qualifiers written by the
user. For example, in the following code
typedef int type;
struct foo {
typedef int type;
::type method(foo::type);
};
completion item for `method` has return type of `::type` and
parameter type of `foo::type`, as originally written in the source
code.
Reviewers: arphaman, bkramer, klimek
Reviewed By: arphaman
Subscribers: mgorny, eraman, cfe-commits
Differential Revision: https://reviews.llvm.org/D38538
llvm-svn: 317677
Summary:
This just seems to have been an oversight. We already supported the f64
atomic add with an explicit scope (e.g. "cta"), but not the scopeless
version.
Reviewers: tra
Subscribers: jholewinski, sanjoy, cfe-commits, llvm-commits, hiraditya
Differential Revision: https://reviews.llvm.org/D39638
llvm-svn: 317623
The files are already large, and we may need to add even more RUNs to
distinguish differences based on OS, environment, or other platform things.
llvm-svn: 317583
Do not crash when trying to define and call a non-standard
strcpy(unsigned char *, unsigned char *) during analysis.
At the same time, do not try to actually evaluate the call.
Differential Revision: https://reviews.llvm.org/D39422
llvm-svn: 317565
The analyzer's BodyFarm models dispatch_once() by comparing the passed-in
predicate against a known 'done' value. If the predicate does not have that
value, the model updates the predicate to have that value and executes the
passed in block.
Unfortunately, the current model uses the wrong 'done' value: 1 instead of ~0.
This interferes with libdispatch's static inline function _dispatch_once(),
which enables a fast path if the block has already been executed. That function
uses __builtin_assume() to tell the compiler that the done flag is set to ~0 on
exit. When r302880 added modeling of __builtin_assume(), this caused the
analyzer to assume 1 == ~0. This in turn caused the analyzer to never explore any code after a call to dispatch_once().
This patch regains the missing coverage by updating BodyFarm to use the correct
'done' value.
rdar://problem/34413048
Differential Revision: https://reviews.llvm.org/D39691
llvm-svn: 317516
Summary:
This change allows binutils to be used for linking with MSVC. Currently, when
using an MSVC target and `-fuse-ld=bfd`, the driver produces an invalid linker
invocation.
Reviewers: rnk, compnerd
Reviewed By: compnerd
Subscribers: smeenai, cfe-commits
Differential Revision: https://reviews.llvm.org/D39509
llvm-svn: 317511
Basically a regression after r316268.
However the diagnostic is correct, but the test coverage is bad.
So just like rL316500, introduce yet more tests,
and adjust the release notes.
See https://bugs.llvm.org/show_bug.cgi?id=35200
llvm-svn: 317421
GNU frontends don't have options like /MT, /MD
This fixes a few link error regressions with libc++ and libc++abi
Reviewers: rnk, mstorsjo, compnerd
Differential Revision: https://reviews.llvm.org/D33620
llvm-svn: 317398
This is a re-apply of rL313082 which was reverted in rL313088
In rL289668 the ability to specify the default linker at compile time
was added but because the MinGW driver used custom detection we could
not take advantage of this new CMAKE flag CLANG_DEFAULT_LINKER.
rL289668 added no test cases and the mingw driver was either overlooked
or purposefully skipped because it has some custom linker tests
Removing them here because they are covered by the generic case.
Reviewers: rnk
Differntial Revision: https://reviews.llvm.org/D37727
llvm-svn: 317397
Add a mix of postive and negative tests to check that wrong Decls won't be
flagged in the diagnostic. Split the check everything test and moved the
pieces closer to where the related tests are.
llvm-svn: 317394
when needed
This commit implements the semicolon insertion logic into the extract
refactoring. The following rules are used:
- extracting expression: add terminating ';' to the extracted function.
- extracting statements that don't require terminating ';' (e.g. switch): add
terminating ';' to the callee.
- extracting statements with ';': move (if possible) the original ';' from the
callee and add terminating ';'.
- otherwise, add ';' to both places.
Differential Revision: https://reviews.llvm.org/D39441
llvm-svn: 317343
Summary:
This change adds Scudo as a possible Sanitizer option via -fsanitize=.
This allows for easier static & shared linking of the Scudo library, it allows
us to enforce PIE (otherwise the security of the allocator is moot), and check
for incompatible Sanitizers combo.
In its current form, Scudo is not compatible with any other Sanitizer, but the
plan is to make it work in conjunction with UBsan (-fsanitize=scudo,undefined),
which will require additional work outside of the scope of this change.
Reviewers: eugenis, kcc, alekseyshl
Reviewed By: eugenis, alekseyshl
Subscribers: llvm-commits, srhines
Differential Revision: https://reviews.llvm.org/D39334
llvm-svn: 317337
When a preamble ends in a conditional preprocessor block that is being
skipped, the preprocessor needs to continue skipping that block when
the preamble is used.
This fixes PR34570.
llvm-svn: 317308
Since they'll likely (not always - if the address is taken, etc) be
inlined away, even at -O0, separately provided weak definitions are
likely to be unused so skip all of that.
llvm-svn: 317279
Consistent with various workarounds in the backwards compatible modules
that allow static functions in headers to exist, be deduplicated to some
degree, and not generally fail right out of the gate... do the same with
modular codegen as there are enough cases (including in libstdc++ and in
LLVM itself - though I cleaned up the easy ones) that it's worth
supporting as a migration/backcompat step.
Simply create a separate, internal linkage function in each object that
needs it. If an available_externally/modularized function references a
static function, but the modularized function is eventually dropped and
not inlined, the static function will be dropped as unreferenced.
llvm-svn: 317274
Added support for regcall as default calling convention. Also added code to
exclude main when applying default calling conventions.
Patch-By: eandrews
Differential Revision: https://reviews.llvm.org/D39210
llvm-svn: 317268
This just makes const-ness of the builtins match const-ness of their lib function siblings.
We're deferring fixing some of these that are obviously wrong to follow-up patches.
Hopefully, the bugs are visible in the new test file (added at rL317220).
As the description in Builtins.def says: "e = const, but only when -fmath-errno=0".
This is step 2 of N to fix builtins and math calls as discussed in D39204.
Differential Revision: https://reviews.llvm.org/D39481
llvm-svn: 317265
If the thread id is requested in windows mode within funclets, we may
generate incorrect function call that could lead to broken codegen.
llvm-svn: 317208
We can generate constant sized arrays whenever the array section has constant
length, even if the base expression itself is a VLA.
Differential Revision: https://reviews.llvm.org/D39504
llvm-svn: 317207
for instantiating its definition.
We model the 'inline'ness as being instantiated with the static data member in
order to track whether the declaration has become a definition yet.
llvm-svn: 317147
Given a choice between a constructor call and a conversion function in C++17,
we prefer the constructor for direct-initialization and the conversion function
for copy-initialization, matching the behavior in C++14 and before. The
guaranteed copy elision rules were not intended to change the meaning of such
code (other than by removing unnecessary copy constructor calls).
This tweak will be raised with CWG.
llvm-svn: 317066
The cloning happens before all metadata nodes are resolved. Prevent the value
mapper from running into unresolved or temporary MD nodes.
Differential Revision: https://reviews.llvm.org/D39396
llvm-svn: 317047
Summary:
This change allows generalizing pointers in type signatures used for
cfi-icall by enabling the -fsanitize-cfi-icall-generalize-pointers flag.
This works by 1) emitting an additional generalized type signature
metadata node for functions and 2) llvm.type.test()ing for the
generalized type for translation units with the flag specified.
This flag is incompatible with -fsanitize-cfi-cross-dso because it would
require emitting twice as many type hashes which would increase artifact
size.
Reviewers: pcc, eugenis
Reviewed By: pcc
Subscribers: kcc
Differential Revision: https://reviews.llvm.org/D39358
llvm-svn: 317044
The right shift operator was not seen as a valid operator in a fold expression, which is PR32563.
Patch by Nicolas Lesser ("Blitz Rakete")!
llvm-svn: 317032
The LLVM sqrt intrinsic definition changed with:
D28797
...so we don't have to use any relaxed FP settings other than errno handling.
This patch sidesteps a question raised in PR27435:
https://bugs.llvm.org/show_bug.cgi?id=27435
Is a programmer using __builtin_sqrt() invoking the compiler's intrinsic definition of sqrt or the mathlib definition of sqrt?
But we have an answer now: the builtin should match the behavior of the libm function including errno handling.
Differential Revision: https://reviews.llvm.org/D39204
llvm-svn: 317031
They might have different visibility, and thus discarding all but one of them
can result in rejecting valid code. Also fix name lookup to cope with multiple
using-directives being found that denote the same namespace, where some are not
visible -- don't cache an "already visited" state for a using-directive that we
didn't visit because it was hidden.
llvm-svn: 316965
The analyzer did not return an UndefVal in case a negative value was left
shifted. I also altered the UndefResultChecker to emit a clear warning in this
case.
Differential Revision: https://reviews.llvm.org/D39423
llvm-svn: 316924
For non-zero alloca addr space, alloca is usually casted to default addr
space immediately.
For non-vla, alloca is inserted at AllocaInsertPt, therefore the addr
space cast should also be insterted at AllocaInsertPt. However,
for vla, alloca is inserted at the current insertion point of IRBuilder,
therefore the addr space cast should also inserted at the current
insertion point of IRBuilder.
Currently clang always insert addr space cast at AllocaInsertPt, which
causes invalid IR.
This patch fixes that.
Differential Revision: https://reviews.llvm.org/D39374
llvm-svn: 316909
Now when a template is instantiated more times and there is a bug found in the
instantiations the issue hash will be different for each instantiation even if
every other property of the bug (path, message, location) is the same.
This patch aims to resolve this issue. Note that explicit specializations still
generate different hashes but that is intended.
Differential Revision: https://reviews.llvm.org/D38728
llvm-svn: 316900
Extend ExprInspection checker to make it possible to dump the issue hash of
arbitrary expressions. This change makes it possible to make issue hash related
tests more concise and also makes debugging issue hash related problems easier.
Differential Revision: https://reviews.llvm.org/D38844
llvm-svn: 316899
AAPCS and AAPCS64 mandate that `wchar_t` with `-fno-short-wchar` is an
`unsigned int` rather than a `signed int`. Ensure that the driver does
not flip the signedness of `wchar_t` for those targets.
Add additional tests to ensure that this does not regress.
llvm-svn: 316858
Added new enum in order to differentiate the warning messages on "misusing" into
3 categories: function calls, moving an object, copying an object. (At the
moment the checker gives the same message in case of copying and moving.)
Additional test cases added as well.
Differential Revision: https://reviews.llvm.org/D38674
llvm-svn: 316852
An earlier solution from Artem r315301 solves the reset problem, however, the
reports should be handled the same way in case of method calls. We should not
just report the base class of the object where the method was defined but the
whole object.
Fixed false positive which came from not removing the subobjects in case of a
state-resetting function. (Just replaced the State->remove(...) call to
removeFromState(..) which was defined exactly for that purpose.)
Some minor typos fixed in this patch as well which did not worth a whole new
patch in my opinion, so included them here.
Differential Revision: https://reviews.llvm.org/D31538
llvm-svn: 316850
OpaqueValueExpr in a GNU binary conditional expression.
It's not meaningful for a non-materialized temporary object to be used as a
common subexpression of multiple expressions.
llvm-svn: 316836
The loop unrolling feature aims to track the maximum possible steps a loop can
make. In order to implement this, it investigates the initial value of the
counter variable and the bound number. (It has to be known.)
These numbers are used as llvm::APInts, however, it was not checked if their
bitwidths are the same which lead to some crashes.
This revision solves this problem by extending the "shorter" one (to the length
of the "longer" one).
For the detailed bug report, see: https://bugs.llvm.org/show_bug.cgi?id=34943
Differential Revision: https://reviews.llvm.org/D38922
llvm-svn: 316830
The existing coverage for the Darwin targets wasn't enough to catch all
the variations. Improve the coverage a bit further and fix a few cases
for Darwin targets.
llvm-svn: 316826
Centralise the definitions of these compiler vended types to aid
inspection to ensure that they are defined similarly. The one case that
stands out is the Darwin case where the types do not match up. This
fixes the API conformance for APCS-GNU as well.
llvm-svn: 316810
Summary:
Clang typically warns that in the following class hierarchy, 'A' is
inaccessible because there is no series of casts that the user can
write to access it unambiguously:
struct A { };
struct B : A { };
struct C : A, B { };
MSVC allows the user to convert from C* to A*, though, and we've
encountered this issue in the latest Windows SDK headers.
This patch allows this conversion when -fms-compatibility is set and
adds a warning for it under -Wmicrosoft-inaccessible-base.
Reviewers: rsmith
Subscribers: cfe-commits
Differential Revision: https://reviews.llvm.org/D39389
llvm-svn: 316807
Craig noticed that CodeGen wasn't properly ignoring the
values sent to the target attribute. This patch ignores
them.
This patch also sets the 'default' for this checking to
'supported', since only X86 has implemented the support
for checking valid CPU names and Feature Names.
One test was changed to i686, since it uses a lakemont,
which would otherwise be prohibited in x86_64.
Differential Revision: https://reviews.llvm.org/D39357
llvm-svn: 316783
deduction for invalid functions
The fabricated template parameters cause an assertion because their depth
is invalid.
rdar://34109988
Differential Revision: https://reviews.llvm.org/D37341
llvm-svn: 316778
Also, for OS unknown targets like wasm, don't include
'unknown' in the library path. This is a fix for rL316719.
Differential Revision: https://reviews.llvm.org/D39354
llvm-svn: 316777
Fixes an assertion failure when ivar is a struct containing incomplete
array. Also completes support for direct flexible array members.
rdar://problem/21054495
Reviewers: rjmccall, theraven
Reviewed By: rjmccall
Subscribers: cfe-commits
Differential Revision: https://reviews.llvm.org/D38774
llvm-svn: 316723
clang currently uses .init_array instead of .ctors on Linux if it detects gcc
4.7+. Make it so that it also uses .init_array if no gcc installation is found
at all – if there's no old gcc, there's nothing we need to be compatible with.
icecc for example runs clang in a very small chroot, so before this change
clang would use .ctors if run under icecc. And lld currently silently mislinks
inputs with .ctors sections, so before this clang + icecc + lld would produce
broken binaries. (But this seems like a good change independent of that lld
bug.)
https://reviews.llvm.org/D39317
llvm-svn: 316713
I think the only reason they are different is because we don't set tune_i686 for -march=i686 to match GCC. But GCC 4.9.0 seems to have changed this behavior and they do set it now. So I think they can aliases now.
Differential Revision: https://reviews.llvm.org/D39349
llvm-svn: 316712
Instead of only setting a non-zero debug location on the return
instruction in *_helper_block functions, set a proper location on all
instructions within these functions. Pick the start location of the
block literal expr for maximum clarity.
The debugger does not step into *_helper_block functions during normal
single-stepping because we mark their parameters as artificial. This is
what we want (the functions are implicitly generated and uninteresting
to most users). The stepping behavior is unchanged by this patch.
rdar://32907581
Differential Revision: https://reviews.llvm.org/D39310
llvm-svn: 316704
The exisiting code goes out of its way to put block parameters into an
alloca only at -O0, and then describes the funciton argument with a
dbg.declare, which is undocumented in the LLVM-CFE contract and does
not actually behave as intended after LLVM r642022.
This patch just generates the alloca unconditionally, the mem2reg pass
will eliminate it at -O1 and up anyway and points the dbg.declare to
the alloca as intended (which mem2reg will then correctly rewrite into
a dbg.value).
This reapplies r316684 with some dead code removed.
rdar://problem/35043980
Differential Revision: https://reviews.llvm.org/D39305
llvm-svn: 316689
The exisiting code goes out of its way to put block parameters into an
alloca only at -O0, and then describes the funciton argument with a
dbg.declare, which is undocumented in the LLVM-CFE contract and does
not actually behave as intended after LLVM r642022.
This patch just generates the alloca unconditionally, the mem2reg pass
will eliminate it at -O1 and up anyway and points the dbg.declare to
the alloca as intended (which mem2reg will then correctly rewrite into
a dbg.value).
rdar://problem/35043980
Differential Revision: https://reviews.llvm.org/D39305
llvm-svn: 316684
Summary:
The warning was initially introduced in D32914 by @thakis,
and the concerns were raised there, and later in rL302247
and PR33771.
I do believe that it makes sense to relax the diagnostic
e.g. in this case, when the expression originates from the
system header, which can not be modified. This prevents
adoption for the diagnostic for codebases which use pthreads
(`PTHREAD_MUTEX_INITIALIZER`), gtest, etc.
As @malcolm.parsons suggests, it *may* make sense to also
not warn for the template types, but it is not obvious to
me how to do that in here.
Though, it still makes sense to complain about `NULL` macro.
While there, add more tests.
Reviewers: dblaikie, thakis, rsmith, rjmccall, aaron.ballman
Reviewed By: thakis
Subscribers: Rakete1111, hans, cfe-commits, thakis, malcolm.parsons
Tags: #clang
Differential Revision: https://reviews.llvm.org/D38954
llvm-svn: 316662
https://bugs.llvm.org/show_bug.cgi?id=33904
Happens when static function is accessed via the class variable. That leads to incorrect overloads number because the variable is considered as the first argument.
struct Bar {
static void foo(); static void foo(int);
};
int main() {
Bar b;
b.foo(/*complete here*/); // did not work before
Bar::foo(/*complete here*/); // worked fine
}
Patch by Ivan Donchevskii!
Differential Revision: https://reviews.llvm.org/D36390
llvm-svn: 316646
Summary:
CUDA 9's minimum sm is sm_30.
Ideally we should also make sm_30 the default when compiling with CUDA
9, but that seems harder than it should be.
Subscribers: sanjoy
Differential Revision: https://reviews.llvm.org/D39109
llvm-svn: 316611
Summary:
Also enable -no-pie on Gnu toolchain (previously available on Darwin only).
Non-PIE executables won't even start on recent Android, and DT_RPATH is ignored by the loader.
Reviewers: srhines, danalbert
Subscribers: cfe-commits
Differential Revision: https://reviews.llvm.org/D38430
llvm-svn: 316606
Darwin uses char * for the variadic list type (va_list). We use the PPC
SVR4 ABI for PPC, which uses a structure type for the va_list. When
constructing the GEP, we would fail due to the incorrect handling for
the va_list. Correct this to use the right type.
llvm-svn: 316599
As indicated by Table 1-1 in Intel Architecture Instruction Set Extensions and Future Features Programming Reference from October 2017.
llvm-svn: 316593
Ensure that we check the ivar containing decl for the DLL storage
attribute rather than the ivar itself as the dll storage is associated
to the interface decl not the ivar decl.
llvm-svn: 316545
determining whether a RecordDecl is CFError.
CFErrorRef used to be declared with "objc_bridge(NSError)" but is now
declared with "objc_bridge_mutable(NSError)". Look for either when
checking whether a RecordDecl is CFError.
rdar://problem/35034779
llvm-svn: 316531
Fastcall doesn't support variadic function calls, so
setting the default calling convention to Fastcall would
result in incorrect code being emitted for these conditions.
This patch adds a 'variadic' test to the default calling conv
test, as well as fixes the behavior of fastcall.
llvm-svn: 316528
rL316268 / D39122 has fixed PR35009, and now when in C,
these three(?) diagnostics properly use the enum's underlying
datatype.
While it was fixed, the test coverage was clearly insufficient,
because the -Wsign-compare change didn't show up in any of the
tests, until it was reported in the post-commit mail for rL316268.
So add the test for the -Wsign-compare diagnostic for enum
for C code, and while there, document this in the release notes.
The fix itself was obviously correct, so unless we want to silence
this new diagnosed case, i deem this commit to be NFC.
llvm-svn: 316500
Builder save/restores insertion pointer when emitting addr space cast
for alloca, but does not save/restore debug loc, which causes verifier
failure for certain call instructions.
This patch fixes that.
Differential Revision: https://reviews.llvm.org/D39069
llvm-svn: 316484
This commit adds an initial, skeleton outline of the "extract function"
refactoring. The extracted function doesn't capture variables / rewrite code
yet, it just basically does a simple copy-paste.
The following initiation rules are specified:
- extraction can only be done for executable code in a function/method/block.
This means that you can't extract a global variable initialize into a function
right now.
- simple literals and references are not extractable.
This commit also adds support for full source ranges to clang-refactor's test
mode.
Differential Revision: https://reviews.llvm.org/D38982
llvm-svn: 316465
With enabled CINDEXTEST_CODE_COMPLETE_PATTERNS env option (which enables
IncludeCodePatterns in completion options) code completion after colon
currently suggests access modifiers with 2 completion chunks which is
incorrect.
Example:
class A : <Cursor>B
{
}
Currently we get 'NotImplemented:{TypedText public}{Colon :} (40)'
but the correct line is just 'NotImplemented:{TypedText public} (40)'
The fix introduces more specific scope that occurs between ':' and '{'
It allows us to determine when we don't need to add ':' as a second
chunk to the public/protected/private access modifiers.
Patch by Ivan Donchevskii!
Differential Revision: https://reviews.llvm.org/D38618
llvm-svn: 316436
When -mtune is used on AArch64 the -target-cpu is passed the value of the
cpu given to -mtune. As well as setting micro-architectural features of the
-mtune cpu, this will also add the architectural features such as support
for instructions. This can result in the backend using instructions that
are supported in the -mtune cpu but not supported in the target
architecture. For example use of the v8.1-a LSE extensions with -march=v8.
This change removes the setting of -target-cpu for -mtune, the -mcpu must
be used to set -target-cpu. This has the effect of removing all non-hard
coded benefits of mtune but it does produce correct output when -mtune cpu
with a later architecture than v8 is used.
Fixes PR34625
Differential Revision: https://reviews.llvm.org/D39179
llvm-svn: 316424
Lit tries to inject the shared library paths, but no action is taken
when platform.system() is not recognized, results in an environment
variable with an empty name, which is illegal.
The patch fixes this mechanism for FreeBSD and NetBSD, and gives an
warning on other platforms, so that the latecomers don't have to spend
time on debugging lit.
Thanks Zhihao Yuan for the patch!
Differential Revision: https://reviews.llvm.org/D39162
llvm-svn: 316411
C++14 [dcl.constexpr]p4 states that in the body of a constexpr
constructor,
> every non-variant non-static data member and base class sub-object
shall be initialized
However, [class.bit]p2 notes that
> Unnamed bit-fields are not members and cannot be initialized.
Therefore, we should make sure to filter them out of the check that
all fields are initialized.
Fixing this makes the constant evaluator a bit smarter, and
specifically allows constexpr constructors to avoid tripping
-Wglobal-constructors when the type contains unnamed bitfields.
Reviewed at https://reviews.llvm.org/D39035.
llvm-svn: 316408
In getLValueElement Base may represent the address of a label
(as in the newly-added test case), in this case it's not a loc::MemRegionVal
and Base.castAs<loc::MemRegionVal>() triggers an assert, this diff makes
getLValueElement return UnknownVal instead.
Differential revision: https://reviews.llvm.org/D39174
llvm-svn: 316399
Allow Obj-C ivars with incomplete array type but only as the last ivar.
Also add a requirement for ivars that contain a flexible array member to
be at the end of class too. It is possible to add in a subclass another
ivar at the end but we'll emit a warning in this case. Also we'll emit a
warning if a variable sized ivar is declared in class extension or in
implementation because subclasses won't know they should avoid adding
new ivars.
In ARC incomplete array objects are treated as __unsafe_unretained so
require them to be marked as such.
Prohibit synthesizing ivars with flexible array members because order of
synthesized ivars is not obvious and tricky to control. Spelling out
ivar explicitly gives control to developers and helps to avoid surprises
with unexpected ivar ordering.
For C and C++ changed diagnostic to tell explicitly a field is not the
last one and point to the next field. It is not as useful as in Obj-C
but it is an improvement and it is consistent with Obj-C. For C for
unions emit more specific err_flexible_array_union instead of generic
err_field_incomplete.
rdar://problem/21054495
Reviewers: rjmccall, theraven
Reviewed By: rjmccall
Subscribers: cfe-commits
Differential Revision: https://reviews.llvm.org/D38773
llvm-svn: 316381
Passing a flavor to LLD requires command line argument, but if these
are being passed through a response file, this will fail because LLD
needs to know which driver to use before processing the response file.
Use ld.lld directly instead to avoid this issue.
Differential Revision: https://reviews.llvm.org/D39176
llvm-svn: 316379
In some cases the compiler can deduce the length of an array section
as constants. With this information, VLAs can be avoided in place of
a constant sized array or even a scalar value if the length is 1.
Example:
int a[4], b[2];
pragma omp parallel reduction(+: a[1:2], b[1:1])
{ }
For chained array sections, this optimization is restricted to cases
where all array sections except the last have a constant length 1.
This trivially guarantees that there are no holes in the memory region
that needs to be privatized.
Example:
int c[3][4];
pragma omp parallel reduction(+: c[1:1][1:2])
{ }
This relands commit r316229 that I reverted in r316235 because it
failed on some bots. During investigation I found that this was because
Clang and GCC evaluate the two arguments to emplace_back() in
ReductionCodeGen::emitSharedLValue() in a different order, hence
leading to a different order of generated instructions in the final
LLVM IR. Fix this by passing in the arguments from temporary variables
that are evaluated in a defined order.
Differential Revision: https://reviews.llvm.org/D39136
llvm-svn: 316362
This is for consistency with lld-link, see https://reviews.llvm.org/D38972
Also give --version a help text so it shows up in --help / /? output (for
both clang-cl and regular clang).
llvm-svn: 316335
In order to identify the copy deduction candidate, I considered two approaches:
- attempt to determine whether an implicit guide is a copy deduction candidate by checking certain properties of its subsituted parameter during overload-resolution.
- using one of the many bits (WillHaveBody) from FunctionDecl (that CXXDeductionGuideDecl inherits from) that are otherwise irrelevant for deduction guides
After some brittle gymnastics w the first strategy, I settled on the second, although to avoid confusion and to give that bit a better name, i turned it into a member of an anonymous union.
Given this identification 'bit', the tweak to overload resolution was a simple reordering of the deduction guide checks (in SemaOverload.cpp::isBetterOverloadCandidate), in-line with Jason Merrill's p0620r0 drafting which made it into the working paper. Concordant with that, I made sure the copy deduction candidate is always added.
References:
See https://bugs.llvm.org/show_bug.cgi?id=34970
See http://wg21.link/p0620r0
llvm-svn: 316292
Summary:
As Mattias Eriksson has reported in PR35009, in C, for enums, the underlying type should
be used when checking for the tautological comparison, unlike C++, where the enumerator
values define the value range. So if not in CPlusPlus mode, use the enum underlying type.
Also, i have discovered a problem (a crash) when evaluating tautological-ness of the following comparison:
```
enum A { A_a = 0 };
if (a < 0) // expected-warning {{comparison of unsigned enum expression < 0 is always false}}
return 0;
```
This affects both the C and C++, but after the first fix, only C++ code was affected.
That was also fixed, while preserving (i think?) the proper diagnostic output.
And while there, attempt to enhance the test coverage.
Yes, some tests got moved around, sorry about that :)
Fixes PR35009
Reviewers: aaron.ballman, rsmith, rjmccall
Reviewed By: aaron.ballman
Subscribers: Rakete1111, efriedma, materi, cfe-commits
Tags: #clang
Differential Revision: https://reviews.llvm.org/D39122
llvm-svn: 316268
constant expressions.
We permit array-to-pointer decay on such arrays, but disallow pointer
arithmetic (since we do not know whether it will have defined behavior).
This is based on r311970 and r301822 (the former by me and the latter by Robert
Haberlach). Between then and now, two things have changed: we have committee
feedback indicating that this is indeed the right direction, and the code
broken by this change has been fixed.
This is necessary in C++17 to continue accepting certain forms of non-type
template argument involving arrays of unknown bound.
llvm-svn: 316245
In some cases the compiler can deduce the length of an array section
as constants. With this information, VLAs can be avoided in place of
a constant sized array or even a scalar value if the length is 1.
Example:
int a[4], b[2];
pragma omp parallel reduction(+: a[1:2], b[1:1])
{ }
For chained array sections, this optimization is restricted to cases
where all array sections except the last have a constant length 1.
This trivially guarantees that there are no holes in the memory region
that needs to be privatized.
Example:
int c[3][4];
pragma omp parallel reduction(+: c[1:1][1:2])
{ }
Differential Revision: https://reviews.llvm.org/D39136
llvm-svn: 316229
Summary:
Change clang-refactor default behavior to print the new code after refactoring
(instead of editing the source files), which would make it easier to use
and debug the refactoring action.
Reviewers: arphaman, ioeric
Reviewed By: arphaman
Subscribers: cfe-commits
Differential Revision: https://reviews.llvm.org/D39092
llvm-svn: 316212
NetBSD uses `long int` for `intptr_t` on ARM. This was changed in SVN
r316046, referenced against other compilers. However, NetBSD's
reference was incorrect as the current clang behaviour is more
up-to-date. Restore the original behaviour for that target.
llvm-svn: 316204
GCC tries to shorten system headers in depfiles using its real path
(resolving components like ".." and following symlinks). Mimic this
feature to ensure that the Ninja build tool detects the correct
dependencies when a symlink changes directory levels, see
https://github.com/ninja-build/ninja/issues/1330
An option to disable this feature is added in case "these changed header
paths may conflict with some compilation environments", see
https://gcc.gnu.org/ml/gcc-patches/2012-09/msg00287.html
Note that the original feature request for GCC
(https://gcc.gnu.org/bugzilla/show_bug.cgi?id=52974) also included paths
preprocessed output (-E) and diagnostics. That is not implemented now
since I am not sure if it breaks something else.
Differential Revision: https://reviews.llvm.org/D37954
llvm-svn: 316193
In function GetIntrinsic, not all types are covered. Types double and long long are missed, type long is wrongly treated same as int, it should be same as long long. These problems cause compiler crashes when compiling code in PR31161. This patch fixed the problem.
Differential Revision: https://reviews.llvm.org/D38820
llvm-svn: 316179
instantiation declarations if they are usable from constant expressions.
We are permitted to instantiate in these cases, and required to do so in order
to have an initializer available for use within constant evaluation.
llvm-svn: 316136
This patch has the following changes
A new flag "-mhvx-length={64B|128B}" is introduced to specify the length of the vector.
Previously we have used "-mhvx-double" for 128 Bytes. This adds the target-feature "+hvx-length{64|128}b"
The "-mhvx" flag must be provided on command line to enable HVX for Hexagon. If no -mhvx-length flag
is specified, a default length is picked from the arch mentioned in this priority order from either -mhvx=vxx
or -mcpu. For v60 and v62 the default length is 64 Byte. For unknown versions, the length is 128 Byte. The
-mhvx flag adds the target-feature "+hvxv{hvx_version}"
The 64 Byte mode is soon going to be deprecated. A warning is emitted if 64 Byte is enabled. A warning is
still emitted for the default 64 Byte as well. This warning can be suppressed with a -Wno flag.
The "-mhvx-double" and "-mno-hvx-double" flags are deprecated. A warning is emitted if the driver sees
them on commandline. "-mhvx-double" is an alias to "-mhvx-length=128B"
The compilation will error out if -mhvx-length is specified with out an -mhvx/-mhvx= flag
The macro HVX_LENGTH is defined and is set to the length of the vector.
Eg: #define HVX_LENGTH 64
The macro HVX_ARCH is defined and is set to the version of the HVX.
Eg: #define HVX_ARCH 62
Differential Revision: https://reviews.llvm.org/D38852
llvm-svn: 316102
Alexey Bataev