Commit Graph

4670 Commits

Author SHA1 Message Date
Richard Smith 9bd6b77ac6 Don't print `&` as part of reference template arguments.
In passing, also generalize the mechanism used to allow Decl's printName
functions to override qualified name printing.
2020-04-15 00:07:12 -07:00
Jon Roelofs 38b39c34ab [clang] Add missing FileCheck colons 2020-04-14 12:32:48 -06:00
Reid Kleckner 55efb68c19 [MS] Mark vbase dtors used when marking dtor used
In the MS C++ ABI, the complete destructor variant for a class with
virtual bases is emitted whereever it is needed, instead of directly
alongside the base destructor variant. The complete destructor calls the
base destructor of the current class and the base destructors of each
virtual base. In order for this to work reliably, translation units that
use the destructor of a class also need to mark destructors of virtual
bases of that class used.

Fixes PR38521

Reviewed By: rsmith

Differential Revision: https://reviews.llvm.org/D77081
2020-04-09 14:19:36 -07:00
Erich Keane 30588a7395 Make target features check work with ctor and dtor-
The problem was reported in PR45468, applying target features to an
always_inline constructor/destructor runs afoul of GlobalDecl
construction assert when checking for target-feature compatibility.

The core problem is fixed by using the version of the check that takes a
FunctionDecl rather than the GlobalDecl. However, while writing the
test, I discovered that source locations weren't properly set for this
check on ctors/dtors. This patch also fixes constructors and CALLED destructors.

Unfortunately, it doesn't seem too possible to get a meaningful source
location for a 'cleanup' destructor, so those are still 'frontend' level
errors unfortunately. A fixme was added to the test to cover that
situation.
2020-04-08 13:19:55 -07:00
Raul Tambre 878d96011a [clang][CodeGen] Handle throw expression in conditional operator constant folding
Summary:
We're smart and do constant folding when emitting conditional operators.
Thus we emit the live value as a lvalue. This doesn't work if the live value is a throw expression.
Handle this by emitting the throw and returning the dead value as the lvalue.

Fixes PR28184.

Reviewers: rsmith

Reviewed By: rsmith

Subscribers: cfe-commits

Tags: #clang

Differential Revision: https://reviews.llvm.org/D77502
2020-04-08 12:32:21 -07:00
Amy Huang bcf66084ed [DebugInfo] Fix for adding "returns cxx udt" option to functions in CodeView.
Summary:
This change adds DIFlagNonTrivial to forward declarations of
DICompositeType. It adds the flag to nontrivial types and types with
unknown triviality.

It fixes adding the "CxxReturnUdt" flag to functions inconsistently,
since it is added based on whether the return type is marked NonTrivial, and
that changes if the return type was a forward declaration.

continues the discussion at https://reviews.llvm.org/D75215

Bug: https://bugs.llvm.org/show_bug.cgi?id=44785

Reviewers: rnk, dblaikie, aprantl

Subscribers: cfe-commits

Tags: #clang

Differential Revision: https://reviews.llvm.org/D77436
2020-04-07 09:10:27 -07:00
Florian Hahn 338be9c595 [Clang] Add llvm.loop.unroll.disable to loops with -fno-unroll-loops.
Currently Clang does not respect -fno-unroll-loops during LTO. During
D76916 it was suggested to respect -fno-unroll-loops on a TU basis.

This patch uses the existing llvm.loop.unroll.disable metadata to
disable loop unrolling explicitly for each loop in the TU if
unrolling is disabled. This should ensure that loops from TUs compiled
with -fno-unroll-loops are skipped by the unroller during LTO.

This also means that if a loop from a TU with -fno-unroll-loops
gets inlined into a TU without this option, the loop won't be
unrolled.

Due to the fact that some transforms might drop loop metadata, there
potentially are cases in which we still unroll loops from TUs with
-fno-unroll-loops. I think we should fix those issues rather than
introducing a function attribute to disable loop unrolling during LTO.
Improving the metadata handling will benefit other use cases, like
various loop pragmas, too. And it is an improvement to clang completely
ignoring -fno-unroll-loops during LTO.

If that direction looks good, we can use a similar approach to also
respect -fno-vectorize during LTO, at least for LoopVectorize.

In the future, this might also allow us to remove the UnrollLoops option
LLVM's PassManagerBuilder.

Reviewers: Meinersbur, hfinkel, dexonsmith, tejohnson

Reviewed By: Meinersbur, tejohnson

Differential Revision: https://reviews.llvm.org/D77058
2020-04-07 14:01:55 +01:00
Amy Huang 11a04a64aa [DebugInfo] Change to constructor homing debug info mode: skip literal types
Summary:
In constructor type homing mode sometimes complete debug info for constexpr
types was missing, because there was not a constructor emitted. This change
makes constructor type homing ignore constexpr types.

Reviewers: rnk, dblaikie

Subscribers: aprantl, cfe-commits

Tags: #clang

Differential Revision: https://reviews.llvm.org/D77432
2020-04-06 09:52:53 -07:00
David Blaikie e9644e6f4f DebugInfo: Fix default template parameter computation for dependent non-type template parameters
This addresses the immediate bug, though in theory we could still
produce a default parameter for the DWARF in this test case - but other
cases will be definitely unachievable (you could have a default
parameter that cannot be evaluated - so long as the user overrode it
with another value rather than relying on that default)
2020-04-05 16:31:30 -07:00
Lucas Prates e6cb4b659a [Clang][CodeGen] Fixing mismatch between memory layout and const expressions for oversized bitfields
Summary:
The construction of constants for structs/unions was conflicting the
expected memory layout for over-sized bit-fields. When building the
necessary bits for those fields, clang was ignoring the size information
computed for the struct/union memory layout and using the original data
from the AST's FieldDecl information. This caused an issue in big-endian
targets, where the field's contant was incorrectly misplaced due to
endian calculations.

This patch aims to separate the constant value from the necessary
padding bits, using the proper size information for each one of them.
With this, the layout of constants for over-sized bit-fields matches the
ABI requirements.

Reviewers: rsmith, eli.friedman, efriedma

Reviewed By: efriedma

Subscribers: efriedma, cfe-commits

Tags: #clang

Differential Revision: https://reviews.llvm.org/D77048
2020-04-02 11:55:20 +01:00
David Blaikie db92719c1d DebugInfo: Defaulted non-type template parameters of bool type
Caused an assertion due to mismatched bit widths - this seems like the
right API to use for a possibly width-varying equality test. Though
certainly open to some post-commit review feedback if there's a more
suitable way to do this comparison/test.
2020-04-01 13:21:13 -07:00
Sam McCall 159a9f7e76 [AST] Print a<b<c>> without extra spaces in C++11 or later.
Summary: It's not 1998 anymore.

Reviewers: kadircet

Subscribers: jkorous, arphaman, usaxena95, cfe-commits

Tags: #clang

Differential Revision: https://reviews.llvm.org/D76801
2020-03-26 09:53:54 +01:00
Eli Friedman 3f1defa6e2 [clang codegen] Clean up handling of vectors with trivial-auto-var-init.
The code was pretending to be doing something useful with vectors, but
really it was doing nothing: the element type of a vector is always a
scalar type, so constWithPadding would always just return the input constant.

Split off from D75661 so it can be reviewed separately.

While I'm here, also add testcase to show missing vector handling.

Differential Revision: https://reviews.llvm.org/D76528
2020-03-24 14:34:40 -07:00
Erik Pilkington de98cf92e3 [CodeGen] Add an alignment attribute to all sret parameters
This fixes a miscompile when the parameter is actually underaligned.
rdar://58316406

Differential revision: https://reviews.llvm.org/D74183
2020-03-24 15:31:57 -04:00
Reid Kleckner ce5173c0e1 Use FinishThunk to finish musttail thunks
FinishThunk, and the invariant of setting and then unsetting
CurCodeDecl, was added in 7f416cc426 (2015). The invariant didn't
exist when I added this musttail codepath in ab2090d107 (2014).
Recently in 28328c3771, I started using this codepath on non-Windows
platforms, and users reported problems during release testing (PR44987).

The issue was already present for users of EH on i686-windows-msvc, so I
added a test for that case as well.

Reviewed By: hans

Differential Revision: https://reviews.llvm.org/D76444
2020-03-20 09:02:21 -07:00
Djordje Todorovic d9b9621009 Reland D73534: [DebugInfo] Enable the debug entry values feature by default
The issue that was causing the build failures was fixed with the D76164.
2020-03-19 13:57:30 +01:00
Yaxun (Sam) Liu e778db6d72 recommit b58f6bb120
Add a test for UsedDeclVisitor

This test is reduced from mlir/lib/Transforms/AffineDataCopyGeneration.cpp
to make sure there is no assertion due to UsedDeclVisitor.
2020-03-19 00:26:38 -04:00
Yaxun (Sam) Liu f528df8e26 Revert "Add a test for UsedDeclVisitor"
This reverts commit b58f6bb120.
2020-03-19 00:15:47 -04:00
Yaxun (Sam) Liu b58f6bb120 Add a test for UsedDeclVisitor
This test is reduced from mlir/lib/Transforms/AffineDataCopyGeneration.cpp
to make sure there is no assertion due to UsedDeclVisitor.
2020-03-19 00:05:10 -04:00
Vedant Kumar 47622efc6f [clang/test] Add test for DIFlagAllCallsDescribed under -ggdb + -gdwarf-4, NFC 2020-03-18 15:24:13 -07:00
Jon Chesterfield cc691f3384 Disable loader-uninitialized tests on Windows 2020-03-17 23:33:12 +00:00
Jon Chesterfield 1d19b15395 Fix arm build broken by D74361 by dropping align from filecheck pattern 2020-03-17 22:15:19 +00:00
Jon Chesterfield c45eaeabb7 [Clang] Undef attribute for global variables
Summary:
[Clang] Attribute to allow defining undef global variables

Initializing global variables is very cheap on hosted implementations. The
C semantics of zero initializing globals work very well there. It is not
necessarily cheap on freestanding implementations. Where there is no loader
available, code must be emitted near the start point to write the appropriate
values into memory.

At present, external variables can be declared in C++ and definitions provided
in assembly (or IR) to achive this effect. This patch provides an attribute in
order to remove this reason for writing assembly for performance sensitive
freestanding implementations.

A close analogue in tree is LDS memory for amdgcn, where the kernel is
responsible for initializing the memory after it starts executing on the gpu.
Uninitalized variables in LDS are observably cheaper than zero initialized.

Patch is loosely based on the cuda __shared__ and opencl __local variable
implementation which also produces undef global variables.

Reviewers: kcc, rjmccall, rsmith, glider, vitalybuka, pcc, eugenis, vlad.tsyrklevich, jdoerfert, gregrodgers, jfb, aaron.ballman

Reviewed By: rjmccall, aaron.ballman

Subscribers: Anastasia, aaron.ballman, davidb, Quuxplusone, dexonsmith, cfe-commits

Tags: #clang

Differential Revision: https://reviews.llvm.org/D74361
2020-03-17 21:22:23 +00:00
Nico Weber f82b32a51e Revert "Reland "[DebugInfo] Enable the debug entry values feature by default""
This reverts commit 5aa5c943f7.
Causes clang to assert, see
https://bugs.chromium.org/p/chromium/issues/detail?id=1061533#c4
for a repro.
2020-03-13 15:37:44 -04:00
Djordje Todorovic 5aa5c943f7 Reland "[DebugInfo] Enable the debug entry values feature by default"
Differential Revision: https://reviews.llvm.org/D73534
2020-03-10 09:15:06 +01:00
Sjoerd Meijer 3d9a0445cc Recommit #2 "[Driver] Default to -fno-common for all targets"
After a first attempt to fix the test-suite failures, my first recommit
caused the same failures again. I had updated CMakeList.txt files of
tests that needed -fcommon, but it turns out that there are also
Makefiles which are used by some bots, so I've updated these Makefiles
now too.

See the original commit message for more details on this change:
0a9fc9233e
2020-03-09 19:57:03 +00:00
Sjoerd Meijer f35d112efd Revert "Recommit "[Driver] Default to -fno-common for all targets""
This reverts commit 2c36c23f34.

Still problems in the test-suite, which I really thought I had fixed...
2020-03-09 10:37:28 +00:00
Sjoerd Meijer 2c36c23f34 Recommit "[Driver] Default to -fno-common for all targets"
This includes fixes for:
- test-suite: some benchmarks need to be compiled with -fcommon, see D75557.
- compiler-rt: one test needed -fcommon, and another a change, see D75520.
2020-03-09 10:07:37 +00:00
Sjoerd Meijer 4e363563fa Revert "[Driver] Default to -fno-common for all targets"
This reverts commit 0a9fc9233e.

Going to look at the asan failures.

I find the failures in the test suite weird, because they look
like compile time test and I don't understand how that can be
failing, but will have a brief look at that too.
2020-03-03 10:00:36 +00:00
Sjoerd Meijer 0a9fc9233e [Driver] Default to -fno-common for all targets
This makes -fno-common the default for all targets because this has performance
and code-size benefits and is more language conforming for C code.
Additionally, GCC10 also defaults to -fno-common and so we get consistent
behaviour with GCC.

With this change, C code that uses tentative definitions as definitions of a
variable in multiple translation units will trigger multiple-definition linker
errors. Generally, this occurs when the use of the extern keyword is neglected
in the declaration of a variable in a header file. In some cases, no specific
translation unit provides a definition of the variable. The previous behavior
can be restored by specifying -fcommon.

As GCC has switched already, we benefit from applications already being ported
and existing documentation how to do this. For example:
- https://gcc.gnu.org/gcc-10/porting_to.html
- https://wiki.gentoo.org/wiki/Gcc_10_porting_notes/fno_common

Differential revision: https://reviews.llvm.org/D75056
2020-03-03 09:15:07 +00:00
Awanish Pandey 1cb0e01e42 [DebugInfo][DWARF5]: Added support for debuginfo generation for defaulted parameters
This patch adds support for dwarf emission/dumping part of debuginfo
generation for defaulted parameters.

Reviewers: probinson, aprantl, dblaikie

Reviewed By: aprantl, dblaikie

Differential Revision: https://reviews.llvm.org/D73462
2020-03-03 13:09:53 +05:30
Hiroshi Yamauchi f16d2bec40 Devirtualize a call on alloca without waiting for post inline cleanup and next DevirtSCCRepeatedPass iteration.
This aims to fix a missed inlining case.

If there's a virtual call in the callee on an alloca (stack allocated object) in
the caller, and the callee is inlined into the caller, the post-inline cleanup
would devirtualize the virtual call, but if the next iteration of
DevirtSCCRepeatedPass doesn't happen (under the new pass manager), which is
based on a heuristic to determine whether to reiterate, we may miss inlining the
devirtualized call.

This enables inlining in clang/test/CodeGenCXX/member-function-pointer-calls.cpp.

This is a second commit after a revert
https://reviews.llvm.org/rG4569b3a86f8a4b1b8ad28fe2321f936f9d7ffd43 and a fix
https://reviews.llvm.org/rG41e06ae7ba91.

Differential Revision: https://reviews.llvm.org/D69591
2020-02-28 09:43:32 -08:00
Kirill Bobyrev 4569b3a86f
Revert "Devirtualize a call on alloca without waiting for post inline cleanup and next"
This reverts commit 59fb9cde7a.

The patch caused internal miscompilations.
2020-02-27 15:58:39 +01:00
Hiroshi Yamauchi 59fb9cde7a Devirtualize a call on alloca without waiting for post inline cleanup and next
DevirtSCCRepeatedPass iteration.  Needs ReviewPublic

This aims to fix a missed inlining case.

If there's a virtual call in the callee on an alloca (stack allocated object) in
the caller, and the callee is inlined into the caller, the post-inline cleanup
would devirtualize the virtual call, but if the next iteration of
DevirtSCCRepeatedPass doesn't happen (under the new pass manager), which is
based on a heuristic to determine whether to reiterate, we may miss inlining the
devirtualized call.

This enables inlining in clang/test/CodeGenCXX/member-function-pointer-calls.cpp.
2020-02-26 09:51:24 -08:00
Roman Lebedev 3dd5a298bf
[clang] Annotating C++'s `operator new` with more attributes
Summary:
Right now we annotate C++'s `operator new` with `noalias` attribute,
which very much is healthy for optimizations.

However as per [[ http://eel.is/c++draft/basic.stc.dynamic.allocation | `[basic.stc.dynamic.allocation]` ]],
there are more promises on global `operator new`, namely:
* non-`std::nothrow_t` `operator new` *never* returns `nullptr`
* If `std::align_val_t align` parameter is taken, the pointer will also be `align`-aligned
* ~~global `operator new`-returned pointer is `__STDCPP_DEFAULT_NEW_ALIGNMENT__`-aligned ~~ It's more caveated than that.

Supplying this information may not cause immediate landslide effects
on any specific benchmarks, but it for sure will be healthy for optimizer
in the sense that the IR will better reflect the guarantees provided in the source code.

The caveat is `-fno-assume-sane-operator-new`, which currently prevents emitting `noalias`
attribute, and is automatically passed by Sanitizers ([[ https://bugs.llvm.org/show_bug.cgi?id=16386 | PR16386 ]]) - should it also cover these attributes?
The problem is that the flag is back-end-specific, as seen in `test/Modules/explicit-build-flags.cpp`.
But while it is okay to add `noalias` metadata in backend, we really should be adding at least
the alignment metadata to the AST, since that allows us to perform sema checks on it.

Reviewers: erichkeane, rjmccall, jdoerfert, eugenis, rsmith

Reviewed By: rsmith

Subscribers: xbolva00, jrtc27, atanasyan, nlopes, cfe-commits

Tags: #llvm, #clang

Differential Revision: https://reviews.llvm.org/D73380
2020-02-26 01:37:17 +03:00
Djordje Todorovic 2f215cf36a Revert "Reland "[DebugInfo] Enable the debug entry values feature by default""
This reverts commit rGfaff707db82d.
A failure found on an ARM 2-stage buildbot.
The investigation is needed.
2020-02-20 14:41:39 +01:00
Reid Kleckner 0edb212925 [MS] Mark vectorcall FP and vector args inreg
This has no effect on how LLVM passes the arguments, but it prevents
rewriteWithInAlloca from thinking that these parameters should be part
of the inalloca pack.

Follow-up to D72114

Reviewed By: erichkeane

Differential Revision: https://reviews.llvm.org/D74452
2020-02-19 16:37:50 -08:00
Djordje Todorovic faff707db8 Reland "[DebugInfo] Enable the debug entry values feature by default"
Differential Revision: https://reviews.llvm.org/D73534
2020-02-19 11:12:26 +01:00
Djordje Todorovic 2bf44d11cb Revert "Reland "[DebugInfo] Enable the debug entry values feature by default""
This reverts commit rGa82d3e8a6e67.
2020-02-18 16:38:11 +01:00
Djordje Todorovic a82d3e8a6e Reland "[DebugInfo] Enable the debug entry values feature by default"
This patch enables the debug entry values feature.

  - Remove the (CC1) experimental -femit-debug-entry-values option
  - Enable it for x86, arm and aarch64 targets
  - Resolve the test failures
  - Leave the llc experimental option for targets that do not
    support the CallSiteInfo yet

Differential Revision: https://reviews.llvm.org/D73534
2020-02-18 14:41:08 +01:00
Erik Pilkington e26c24b849 Revert "[IRGen] Emit lifetime intrinsics around temporary aggregate argument allocas"
This reverts commit fafc6e4fdf.

Should fix ppc stage2 failure: http://lab.llvm.org:8011/builders/clang-ppc64be-linux-multistage/builds/23546

Conflicts:
	clang/lib/CodeGen/CGCall.cpp
2020-02-12 12:26:46 -08:00
Djordje Todorovic 97ed706a96 Revert "[DebugInfo] Enable the debug entry values feature by default"
This reverts commit rG9f6ff07f8a39.

Found a test failure on clang-with-thin-lto-ubuntu buildbot.
2020-02-12 11:59:04 +01:00
Djordje Todorovic 9f6ff07f8a [DebugInfo] Enable the debug entry values feature by default
This patch enables the debug entry values feature.

  - Remove the (CC1) experimental -femit-debug-entry-values option
  - Enable it for x86, arm and aarch64 targets
  - Resolve the test failures
  - Leave the llc experimental option for targets that do not
    support the CallSiteInfo yet

Differential Revision: https://reviews.llvm.org/D73534
2020-02-12 10:25:14 +01:00
Reid Kleckner 2c6a3896ab Re-land "[MS] Overhaul how clang passes overaligned args on x86_32"
This brings back 2af74e27ed and reverts
eaabaf7e04.

The changes were correct, the code that was broken contained an ODR
violation that assumed that these types are passed equivalently:
  struct alignas(uint64_t) Wrapper { uint64_t P };
  void f(uint64_t p);
  void f(Wrapper p);

MSVC does not pass them the same way, and so clang-cl should not pass
them the same way either.
2020-02-11 16:49:28 -08:00
Richard Smith 9ce6dc9872 CWG1423: don't permit implicit conversion of nullptr_t to bool.
The C++ rules briefly allowed this, but the rule changed nearly 10 years
ago and we never updated our implementation to match. However, we've
warned on this by default for a long time, and no other compiler accepts
(even as an extension).
2020-02-11 06:52:45 -08:00
Michael Liao a067891389 [clang][codegen] Fix another lifetime emission on alloca on non-default address space.
- Lifetime intrinsics expect the pointer directly from alloca. Need
  extra handling for targets with alloca on non-default (or non-zero)
  address space.
2020-02-10 00:15:56 -05:00
Erik Pilkington fafc6e4fdf [IRGen] Emit lifetime intrinsics around temporary aggregate argument allocas
These temporaries are only used in the callee, and their memory can be reused
after the call is complete.

rdar://58552124

Differential revision: https://reviews.llvm.org/D74094
2020-02-07 14:39:31 -08:00
Michael Liao 2926917f43 [clang] Fix linkage of nested lambdas.
patch from Philippe Daouadi <blastrock@free.fr>

This is an attempt to fix
[PR#44368](https://bugs.llvm.org/show_bug.cgi?id=44368)

This effectively reverts [D1783](https://reviews.llvm.org/D1783). It
doesn't break the current tests and fixes the test that this commit
adds.

We now decide of a lambda linkage only depending on the visibility of
its parent context.

Differential Revision: https://reviews.llvm.org/D73701
2020-02-07 13:24:21 -05:00
Richard Smith 96c899449b C++ DR2026: static storage duration variables are not zeroed before
constant initialization.

Removing this zeroing regressed our code generation in a few cases, also
fixed here. We now compute whether a variable has constant destruction
even if it doesn't have a constant initializer, by trying to destroy a
default-initialized value, and skip emitting a trivial default
constructor for a variable even if it has non-trivial (but perhaps
constant) destruction.
2020-02-06 16:37:22 -08:00
shafik 428583dd22 [DebugInfo] Fix debug-info generation for block invocations so that we set the LinkageName
Currently when generating debug-info for a BlockDecl we are setting the Name to the mangled name and not setting the LinkageName.
This means we see the mangled name for block invcations ends up in DW_AT_Name and not in DW_AT_linkage_name.

This patch fixes this case so that we also set the LinkageName as well.

Differential Revision: https://reviews.llvm.org/D73282
2020-02-05 11:07:30 -08:00