Serge Guelton [Thu, 21 Mar 2019 14:47:40 +0000 (14:47 +0000)]
Workaround items/iteritems difference between Python2 and Python3
Differential Revision: https://reviews.llvm.org/D59588
llvm-svn: 356673
Serge Guelton [Thu, 21 Mar 2019 14:44:58 +0000 (14:44 +0000)]
Workaround Python's map difference between Python2/3
When the output of map is not used, using a list comprehension or an explicit
call to list looks awkward.
Differential Revision: https://reviews.llvm.org/D59587
llvm-svn: 356672
Serge Guelton [Thu, 21 Mar 2019 14:41:34 +0000 (14:41 +0000)]
Portable int/long conversion across Python 2 / Python 3
Differential Revision: https://reviews.llvm.org/D59585
llvm-svn: 356671
Serge Guelton [Thu, 21 Mar 2019 14:39:55 +0000 (14:39 +0000)]
Portable exception value access across Python 2 / Python 3
Differential Revision: https://reviews.llvm.org/D59583
llvm-svn: 356670
David Green [Thu, 21 Mar 2019 14:35:06 +0000 (14:35 +0000)]
Fixup opt-remarks.ll gold plugin test. NFC
Now that rL356594 has added a TailCallElim pass to LTO, the call gets marked as
tail.
llvm-svn: 356669
Florian Hahn [Thu, 21 Mar 2019 14:32:09 +0000 (14:32 +0000)]
[DAGCombiner] Use getTokenFactor in a few more cases.
SDNodes can only have 64k operands and for some inputs (e.g. large
number of stores), we can reach this limit when creating TokenFactor
nodes. This patch is a follow up to D56740 and updates a few more places
that potentially can create TokenFactors with too many operands.
Reviewers: efriedma, craig.topper, aemerson, RKSimon
Reviewed By: RKSimon
Differential Revision: https://reviews.llvm.org/D59156
llvm-svn: 356668
Simon Pilgrim [Thu, 21 Mar 2019 14:07:18 +0000 (14:07 +0000)]
[DAGCombine] SimplifySelectCC - call FoldSetCC with the setcc result type
We were calling FoldSetCC with the compare operand type instead of the result type.
Found by OSS-Fuzz #13838 (https://bugs.chromium.org/p/oss-fuzz/issues/detail?id=13838)
llvm-svn: 356667
Peter Smith [Thu, 21 Mar 2019 14:06:40 +0000 (14:06 +0000)]
[ELF][ARM] Redesign of .ARM.exidx handling to use a SyntheticSection
Instead of creating extra Synthetic .ARM.exidx sections to account for
gaps in the table, create a single .ARM.exidx SyntheticSection that can
derive the contents of the gaps from a sorted list of the executable
InputSections. This has the benefit of moving the ARM specific code for
SyntheticSections in SHF_LINK_ORDER processing and the table merging code
into the ARM specific SyntheticSection. This also makes it easier to create
EXIDX_CANTUNWIND table entries for executable InputSections that don't
have an associated .ARM.exidx section.
Fixes pr40277
Differential Revision: https://reviews.llvm.org/D59216
llvm-svn: 356666
Sanjay Patel [Thu, 21 Mar 2019 13:57:07 +0000 (13:57 +0000)]
[CodeGenPrepare] limit formation of overflow intrinsics (PR41129)
This is probably a bigger limitation than necessary, but since we don't have any evidence yet
that this transform led to real-world perf improvements rather than regressions, I'm making a
quick, blunt fix.
In the motivating x86 example from:
https://bugs.llvm.org/show_bug.cgi?id=41129
...and shown in the regression test, we want to avoid an extra instruction in the dominating
block because that could be costly.
The x86 LSR test diff is reversing the changes from D57789. There's no evidence that 1 version
is any better than the other yet.
Differential Revision: https://reviews.llvm.org/D59602
llvm-svn: 356665
Xing GUO [Thu, 21 Mar 2019 13:42:06 +0000 (13:42 +0000)]
[llvm-readobj] Format codes. NFC.
llvm-svn: 356664
Erich Keane [Thu, 21 Mar 2019 13:30:56 +0000 (13:30 +0000)]
Permit redeclarations of a builtin to specify calling convention.
After https://reviews.llvm.org/rL355317 we noticed that quite a decent
amount of code redeclares builtins (memcpy in particular, I believe
reduced from an MSVC header) with a calling convention specified.
This gets particularly troublesome when the user specifies a new
'default' calling convention on the command line.
When looking to add a diagnostic for this case, it was noticed that we
had 3 other diagnostics that differed only slightly. This patch ALSO
unifies those under a 'select'. Unfortunately, the order of words in
ONE of these diagnostics was reversed ("'thiscall' calling convention"
vs "calling convention 'thiscall'"), so this patch also standardizes on
the former.
Differential Revision: https://reviews.llvm.org/D59560
Change-Id: I79f99fe7c2301640755ffdd774b46eb44526bb22
llvm-svn: 356663
Paul Hoad [Thu, 21 Mar 2019 13:09:22 +0000 (13:09 +0000)]
[clang-format] Add basic support for formatting C# files
Summary:
This revision adds basic support for formatting C# files with clang-format, I know the barrier to entry is high here so I'm sending this revision in to test the water as to whether this might be something we'd consider landing.
Tracking in Bugzilla as:
https://bugs.llvm.org/show_bug.cgi?id=40850
Justification:
C# code just looks ugly in comparison to the C++ code in our source tree which is clang-formatted.
I've struggled with Visual Studio reformatting to get a clean and consistent style, I want to format our C# code on saving like I do now for C++ and i want it to have the same style as defined in our .clang-format file, so it consistent as it can be with C++. (Braces/Breaking/Spaces/Indent etc..)
Using clang format without this patch leaves the code in a bad state, sometimes when the BreakStringLiterals is set, it fails to compile.
Mostly the C# is similar to Java, except instead of JavaAnnotations I try to reuse the TT_AttributeSquare.
Almost the most valuable portion is to have a new Language in order to partition the configuration for C# within a common .clang-format file, with the auto detection on the .cs extension. But there are other C# specific styles that could be added later if this is accepted. in particular how `{ set;get }` is formatted.
Reviewers: djasper, klimek, krasimir, benhamilton, JonasToth
Reviewed By: klimek
Subscribers: llvm-commits, mgorny, jdoerfert, cfe-commits
Tags: #clang, #clang-tools-extra
Differential Revision: https://reviews.llvm.org/D58404
llvm-svn: 356662
Simon Pilgrim [Thu, 21 Mar 2019 12:41:18 +0000 (12:41 +0000)]
[Thumb] Fix infinite loop in ABS expansion (PR41160)
Don't expand ISD::ABS node if its legal.
llvm-svn: 356661
Tim Renouf [Thu, 21 Mar 2019 12:01:21 +0000 (12:01 +0000)]
[AMDGPU] Support for v3i32/v3f32
Added support for dwordx3 for most load/store types, but not DS, and not
intrinsics yet.
SI (gfx6) does not have dwordx3 instructions, so they are not enabled
there.
Some of this patch is from Matt Arsenault, also of AMD.
Differential Revision: https://reviews.llvm.org/D58902
Change-Id: I913ef54f1433a7149da8d72f4af54dbb13436bd9
llvm-svn: 356659
Simon Pilgrim [Thu, 21 Mar 2019 11:58:22 +0000 (11:58 +0000)]
Fix -Wmisleading-indentation gcc7 warning. NFCI.
llvm-svn: 356658
Oliver Stannard [Thu, 21 Mar 2019 11:30:17 +0000 (11:30 +0000)]
[AArch64] Allow -mattr=tpidr-el[1|2|3]
Added subtarget features for AArch64 to use TPIDR_EL[1|2|3] as the TLS base
register, rather than the default TPIDR_EL0.
Patch by Philip Derrin!
Differential revision: https://reviews.llvm.org/D54685
llvm-svn: 356657
Simon Pilgrim [Thu, 21 Mar 2019 11:18:54 +0000 (11:18 +0000)]
[SelectionDAG] Add scalarization of ABS node (PR41149)
Patch by: @ikulagin (Ivan Kulagin)
Differential Revision: https://reviews.llvm.org/D59577
llvm-svn: 356656
Pavel Labath [Thu, 21 Mar 2019 10:21:55 +0000 (10:21 +0000)]
Fix two more issues with r356652
The first problem was a use-after-free in the tests (detected by asan
bots). The temporary array created for the "create" call is guaranteed
to live only until the end of the statement. The fix there is to store
the test data in a local variable to ensure it has the right lifetime
The second issue is broken BUILD_SHARED_LIBS build, which I fix by
adding the appropriate BinaryFormat dependency to the Object unit tests.
llvm-svn: 356655
Alex Bradbury [Thu, 21 Mar 2019 09:53:19 +0000 (09:53 +0000)]
[RISCV][NFC] Remove old CHECK lines from linker-relaxation.s test
The RELOC: check lines are no longer used.
llvm-svn: 356654
Pavel Labath [Thu, 21 Mar 2019 09:44:07 +0000 (09:44 +0000)]
Attempt to fix modules build for r356652
The commit added a new .def file. This adds it to the list of textual
headers.
llvm-svn: 356653
Pavel Labath [Thu, 21 Mar 2019 09:18:59 +0000 (09:18 +0000)]
[Object] Add basic minidump support
Summary:
This patch adds basic support for reading minidump files. It contains
the definitions of various important minidump data structures (header,
stream directory), and of one minidump stream (SystemInfo). The ability
to read other streams will be added in follow-up patches. However, all
streams can be read even now as raw data, which means lldb's minidump
support (where this code is taken from) can be immediately rebased on
top of this patch as soon as it lands.
As we don't have any support for generating minidump files (yet), this
tests the code via unit tests with some small handcrafted binaries in
the form of c char arrays.
Reviewers: Bigcheese, jhenderson, zturner
Subscribers: srhines, dschuff, mgorny, fedor.sergeev, lemo, clayborg, JDevlieghere, aprantl, lldb-commits, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D59291
llvm-svn: 356652
Fangrui Song [Thu, 21 Mar 2019 08:47:40 +0000 (08:47 +0000)]
[BasicAA] Use DenseMap::try_emplace after D59151. NFC
llvm-svn: 356651
Zinovy Nis [Thu, 21 Mar 2019 08:32:07 +0000 (08:32 +0000)]
Reland r356547 after fixing the YAML module missing issue.
[clang-tidy] Parallelize clang-tidy-diff.py
This patch has 2 rationales:
- large patches lead to long command lines and often cause max command line length restrictions imposed by OS;
- clang-tidy runs on modified files are independent and can be done in parallel, the same as done for run-clang-tidy.
Differential Revision: https://reviews.llvm.org/D5766
llvm-svn: 356649
Mikael Holmen [Thu, 21 Mar 2019 07:54:44 +0000 (07:54 +0000)]
Silence warning about unused variable in builds without asserts [NFC]
llvm-svn: 356648
Serge Guelton [Thu, 21 Mar 2019 07:19:09 +0000 (07:19 +0000)]
Use list comprehension instead of map/filter to prepare Python2/3 compat
Differential Revision: https://reviews.llvm.org/D59579
llvm-svn: 356647
Craig Topper [Thu, 21 Mar 2019 05:54:37 +0000 (05:54 +0000)]
[ScalarizeMaskedMemIntrinsics] Reverse some if conditions to reduce indentations to remove curly braces.
Pre-commit for D59180
llvm-svn: 356646
Craig Topper [Thu, 21 Mar 2019 05:33:10 +0000 (05:33 +0000)]
[InstCombine] Add test case for PR41164. NFC
llvm-svn: 356645
Alina Sbirlea [Thu, 21 Mar 2019 05:02:05 +0000 (05:02 +0000)]
[BasicAA] Reduce no of map seaches [NFCI].
Summary:
This is a refactoring patch.
- Reduce the number of map searches by reusing the iterator.
- Add asserts to check that the entry is in the cache, as this is something BasicAA relies on to avoid infinite recursion.
Reviewers: chandlerc, aschwaighofer
Subscribers: sanjoy, jlebar, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D59151
llvm-svn: 356644
Jonas Devlieghere [Thu, 21 Mar 2019 04:08:31 +0000 (04:08 +0000)]
[Reproducers] Properly handle QEnvironment packets
On Linux, a QEnvironment packet is sent for every environment variable.
This breaks replay when the number of environment variables is different
then during capture. The solution is to always reply with OK.
llvm-svn: 356643
Philip Reames [Thu, 21 Mar 2019 03:23:40 +0000 (03:23 +0000)]
[instcombine] Add some todos, and arrange code for readibility
llvm-svn: 356642
George Burgess IV [Thu, 21 Mar 2019 03:11:34 +0000 (03:11 +0000)]
[MSSA] Delete move ctor; remove dynamic never-moved verification
Code archaeology in D59315 revealed that MSSA should never be moved.
Rather than trying to check dynamically that this hasn't happened in the
verify() functions of Walkers, it's likely best to just delete its move
constructor.
Since all these verify() functions did is check that MSSA hasn't moved,
this allows us to remove these verify functions.
I can readd the verification checks if someone's super concerned about
us trying to `memcpy` MemorySSA or something somewhere, but I imagine we
have other problems if we're trying anything like that...
llvm-svn: 356641
Eric Fiselier [Thu, 21 Mar 2019 02:13:30 +0000 (02:13 +0000)]
Add dylib-has-no-filesystem when filesystem is disabled
llvm-svn: 356640
Eric Fiselier [Thu, 21 Mar 2019 02:07:55 +0000 (02:07 +0000)]
Mark debug death tests as unsupported on Windows
llvm-svn: 356639
Jonas Devlieghere [Thu, 21 Mar 2019 01:57:33 +0000 (01:57 +0000)]
[Reproducers] Log inconsistencies during replay (NFC)
Make debugging of the GDB remote packet aspect of reproducers easier by
logging both requests and replies. This enables some sanity checking
during replay.
llvm-svn: 356638
Eric Fiselier [Thu, 21 Mar 2019 01:48:15 +0000 (01:48 +0000)]
Remove type visibility specifiers from new chrono types.
_LIBCPP_TYPE_VIS is only really needed on types with a vtable.
And on Windows it doesn't work with types that have only inline methods.
This patch removes the unneeded attributes.
llvm-svn: 356637
George Burgess IV [Thu, 21 Mar 2019 01:01:53 +0000 (01:01 +0000)]
creduce-clang-crash.py: preprocess file + reduce commandline
This CL causes our creduce-clang-crash.py util to:
- try to preprocess the file before reducing
- try to remove some command line arguments
- now require a llvm bin directory, since the generated crash script
doesn't have an absolute path for clang
It also marks it as executable, since I forgot to do that in the last
commit. :)
Patch by Amy Huang!
Differential Revision: https://reviews.llvm.org/D59440
llvm-svn: 356636
Billy Robert O'Neal III [Thu, 21 Mar 2019 00:24:43 +0000 (00:24 +0000)]
[libcxx] [test] Revert r356632 add (void) casts to operator new calls, to suppress warnings generated by [[nodiscard]]."
llvm-svn: 356635
Artem Dergachev [Thu, 21 Mar 2019 00:15:07 +0000 (00:15 +0000)]
[CFG] [analyzer] pr41142: C++17: Skip transparent InitListExprs in constructors.
When searching for construction contexts, i.e. figuring out which statements
define the object that is constructed by each construct-expression, ignore
transparent init-list expressions because they don't add anything to the
context. This allows the Static Analyzer to model construction, destruction,
materialization, lifetime extension correctly in more cases. Also fixes
a crash caused by incorrectly evaluating initial values of variables
initialized with such expressions.
Differential Revision: https://reviews.llvm.org/D59573
llvm-svn: 356634
Eric Fiselier [Thu, 21 Mar 2019 00:04:31 +0000 (00:04 +0000)]
Allow disabling of filesystem library.
Summary: Filesystem doesn't work on Windows, so we need a mechanism to turn it off for the time being.
Reviewers: ldionne, serge-sans-paille, EricWF
Reviewed By: EricWF
Subscribers: mstorsjo, mgorny, christof, jdoerfert, libcxx-commits
Differential Revision: https://reviews.llvm.org/D59619
llvm-svn: 356633
Billy Robert O'Neal III [Wed, 20 Mar 2019 23:58:46 +0000 (23:58 +0000)]
[libcxx] [test] Add (void) casts to operator new calls, to suppress warnings generated by [[nodiscard]].
This allows these tests to pass when compiled by MSVC++.
llvm-svn: 356632
Craig Topper [Wed, 20 Mar 2019 23:35:49 +0000 (23:35 +0000)]
[X86] Add CMPXCHG8B feature flag. Set it for all CPUs except i386/i486 including 'generic'. Disable use of CMPXCHG8B when this flag isn't set.
CMPXCHG8B was introduced on i586/pentium generation.
If its not enabled, limit the atomic width to 32 bits so the AtomicExpandPass will expand to lib calls. Unclear if we should be using a different limit for other configs. The default is 1024 and experimentation shows that using an i256 atomic will cause a crash in SelectionDAG.
Differential Revision: https://reviews.llvm.org/D59576
llvm-svn: 356631
Douglas Yung [Wed, 20 Mar 2019 23:21:43 +0000 (23:21 +0000)]
Revert "Reland r356547 after fixing the tests for Linux."
This reverts commit
538fb72226cf6dff95af83f7777e12b8dbd061ea (r356565).
This is still breaking a build bot:
http://lab.llvm.org:8011/builders/llvm-clang-lld-x86_64-scei-ps4-ubuntu-fast/builds/45557
llvm-svn: 356630
Michael Trent [Wed, 20 Mar 2019 23:21:16 +0000 (23:21 +0000)]
Fix Mach-O bind and rebase validation errors in libObject
Summary:
llvm-objdump (via libObject) validates DYLD_INFO rebase and bind
entries against the basic structure found in the Mach-O file before
evaluating the contents of those entries. Certain malformed Mach-Os can
defeat the validation check and force llvm-objdump (libObject) to crash.
The previous logic verified a rebase or bind started in a valid Mach-O
section, but did not verify that the section wholely contained the
fixup. It also generally allows rebases or binds to start immediately
after a valid section even if that range is not itself part of a valid
section. Finally, bind and rebase opcodes that indicate more than one
fixup (apply N times...) are not completely validated: only the first
and final fixups are checked.
The previous logic also rejected certain binaries as false positives.
Some bind and rebase opcodes can modify the state machine such that the
next bind or rebase will fail. libObject will reject these opcodes as
invalid in order to be helpful and print an error message associated
with the instruction that caused the problem, even though the binary is
not actually illegal until it consumes the invalid state in the state
machine. In other words, libObject may reject a Mach-O binary that
Apple's dynamic linker may consider legal. The original version of
macho-rebase-add-addr-uleb-too-big is an example of such a binary.
I have replaced the existing checkSegAndOffset and checkCountAndSkip
functions with a single function, checkSegAndOffsets, which validates
all of the fixups realized by a DYLD_INFO opcode. checkSegAndOffsets
verifies that a Mach-O section fully contains each fixup. Every fixup
realized by an opcode is validated, and some (but not all!)
inconsistencies in the state machine are allowed until a fixup is
realized. This means that libObject may fail on an opcode that realizes
a fixup, not on the opcode that introduced the arithmetic error.
Existing test cases have been modified to reflect the changes in error
messages returned by libObject. What's more, the test case for
macho-rebase-add-addr-uleb-too-big has been modified so that it actually
triggers the error condition; the new code in libObject considers the
original test binary "legal".
rdar://
47797757
Reviewers: lhames, pete, ab
Reviewed By: pete
Subscribers: rupprecht, jdoerfert, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D59574
llvm-svn: 356629
Jennifer Yu [Wed, 20 Mar 2019 23:05:18 +0000 (23:05 +0000)]
Remove extra white spaces
llvm-svn: 356628
Stephane Moore [Wed, 20 Mar 2019 23:05:00 +0000 (23:05 +0000)]
[clang-tidy] Disable google-runtime-int in Objective-C++ 🔓
Summary:
In contrast to Google C++, Objective-C often uses built-in integer types
other than `int`. In fact, the Objective-C runtime itself defines the
types NSInteger¹ and NSUInteger² which are variant types depending on
the target architecture. The Objective-C style guide indicates that
usage of system types with variant sizes is appropriate when handling
values provided by system interfaces³. Objective-C++ is commonly the
result of conversion from Objective-C to Objective-C++ for the purpose
of integrating C++ functionality. The opposite of Objective-C++ being
used to expose Objective-C functionality to C++ is less common,
potentially because Objective-C has a signficantly more uneven presence
on different platforms compared to C++. This generally predisposes
Objective-C++ to commonly being more Objective-C than C++. Forcing
Objective-C++ developers to perform conversions between variant system types
and fixed size integer types depending on target architecture when
Objective-C++ commonly uses variant system types from Objective-C is
likely to lead to more bugs and overhead than benefit. For that reason,
this change proposes to disable google-runtime-int in Objective-C++.
[1] https://developer.apple.com/documentation/objectivec/nsinteger?language=objc
[2] https://developer.apple.com/documentation/objectivec/nsuinteger?language=objc
[3] "Types long, NSInteger, NSUInteger, and CGFloat vary in size between
32- and 64-bit builds. Use of these types is appropriate when handling
values exposed by system interfaces, but they should be avoided for most
other computations."
https://github.com/google/styleguide/blob/gh-pages/objcguide.md#types-with-inconsistent-sizes
Subscribers: xazax.hun, jdoerfert, cfe-commits
Tags: #clang
Differential Revision: https://reviews.llvm.org/D59336
llvm-svn: 356627
Jason Molenda [Wed, 20 Mar 2019 22:59:16 +0000 (22:59 +0000)]
Change the logging on ptrace(PT_KILL) in MachProcess::Kill to log
if LOG_PROCESS is enabled or if there was an error making that call.
<rdar://problem/
49036508>
llvm-svn: 356626
Jason Molenda [Wed, 20 Mar 2019 22:57:33 +0000 (22:57 +0000)]
Update DWARF files.
llvm-svn: 356625
Nico Weber [Wed, 20 Mar 2019 22:55:03 +0000 (22:55 +0000)]
libc++/win: Make once_flag have the same size as a pointer.
`unsigned long` is 32-bit on 32-bit systems and 64-bit on 64-bit systems
on LP64 systems -- which most Unix systems are, but Windows isn't.
Windows is LLP64, which means unsigned long is 32-bit even on 64-bit
systems.
pplwin.h contains
static_assert(alignof(void *) == alignof(::std::once_flag), ...)
which fails due to this problem.
Instead of unsigned long, use uintptr_t, which consistently is 32-bit
on 32-bit systems and 64-bit on 64-bit systems.
No functional change except on 64-bit Windows.
Differential Revision: https://reviews.llvm.org/D59607
llvm-svn: 356624
Jordan Rupprecht [Wed, 20 Mar 2019 22:51:56 +0000 (22:51 +0000)]
Revert "[clangd] Print arguments in template specializations"
This reverts commit
44a63f6a150dec72dea43730d2a89d292e58bd6f. It segfaults on an internal test case (will follow up off thread).
llvm-svn: 356623
Thomas Lively [Wed, 20 Mar 2019 22:34:34 +0000 (22:34 +0000)]
[WebAssembly][NFC] Fix formatting error from rL356610
llvm-svn: 356622
Tim Renouf [Wed, 20 Mar 2019 22:02:09 +0000 (22:02 +0000)]
[AMDGPU] Do not generate spurious PAL metadata
My previous fix rL356591 "[AMDGPU] Added MsgPack format PAL metadata"
accidentally caused a spurious PAL metadata .note record to be emitted
for any AMDGPU output. That caused failures in the lld test
amdgpu-relocs.s. Fixed.
Differential Revision: https://reviews.llvm.org/D59613
Change-Id: Ie04a2aaae890dcd490f22c89edf9913a77ce070e
llvm-svn: 356621
Nikita Popov [Wed, 20 Mar 2019 21:42:17 +0000 (21:42 +0000)]
[InstCombine] Add additional sub nsw inference tests; NFC
nsw can be determined based on known bits here, but currently
isn't.
llvm-svn: 356620
Stanislav Mekhanoshin [Wed, 20 Mar 2019 21:42:05 +0000 (21:42 +0000)]
Allow machine dce to remove uses in the same instruction
Machine DCE cannot remove a dead definition if there are non-dbg uses.
A use however can be in the same instruction:
dead %0 = INST %0
Such instructions sometimes created by Detect dead lanes pass.
Allow this instruction to be deleted despite the use if the only use
belongs to the same instruction.
Differential Revision: https://reviews.llvm.org/D59565
llvm-svn: 356619
Craig Topper [Wed, 20 Mar 2019 21:30:20 +0000 (21:30 +0000)]
[X86] Call lowerShuffleAsBitMask for 512-bit vectors in lowerShuffleAsBlend.
This patch enables the use of lowerShuffleAsBitMask for 512-bit blends before
falling back to move immedate, GPR to k-register, and masked op.
I had to make some changes to support v8i64 when i64 is not a legal type. And to
support floating point types.
This trades a load for the move immediate and GPR move which is higher latency.
But its probably better for register pressure not having to hop through other
register classes. The load+and should play better with LICM and
rematerialization I think.
Differential Revision: https://reviews.llvm.org/D59479
llvm-svn: 356618
Michael Liao [Wed, 20 Mar 2019 21:22:27 +0000 (21:22 +0000)]
[AMDGPU] Fix dependency on `BinaryFormat`
Summary: - The linking is broken when this library is built as shared one.
Subscribers: arsenm, kzhuravl, jvesely, wdng, nhaehnle, yaxunl, dstuttard, tpr, t-tye, hiraditya, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D59610
llvm-svn: 356617
Louis Dionne [Wed, 20 Mar 2019 21:18:14 +0000 (21:18 +0000)]
[libc++] Mark <filesystem> as unavailable on Apple platforms using pragmas
Summary:
Also add the corresponding XFAILs to tests that require filesystem.
The approach taken to mark <filesystem> as unavailable in this patch
is to mark all the header as unavailable using #pragma clang attribute.
Marking each declaration using the attribute is more intrusive and
does not provide a lot of value right now because pretty much everything
in <filesystem> requires dylib support, often transitively.
This is an alternative to https://reviews.llvm.org/D59093.
A similar (but partial) patch was already applied in r356558.
Reviewers: mclow.lists, EricWF, serge-sans-paille
Subscribers: christof, jkorous, dexonsmith, libcxx-commits
Differential Revision: https://reviews.llvm.org/D59224
llvm-svn: 356616
Paul Hoad [Wed, 20 Mar 2019 21:02:12 +0000 (21:02 +0000)]
[clang-format][NFC] fix release notes build issue
build issue from r356613
llvm-svn: 356615
Jordan Rupprecht [Wed, 20 Mar 2019 21:01:56 +0000 (21:01 +0000)]
[clang][OpenMP] Fix build when using libgomp
Summary: rL356570 introduced a test which only passes with the default openmp library, libomp, and fails with other openmp libraries, such as libgomp. Explicitly choose libomp.
Reviewers: lebedev.ri
Subscribers: guansong, jdoerfert, cfe-commits
Tags: #clang
Differential Revision: https://reviews.llvm.org/D59609
llvm-svn: 356614
Paul Hoad [Wed, 20 Mar 2019 20:49:43 +0000 (20:49 +0000)]
[clang-format] BeforeHash added to IndentPPDirectives
Summary:
The option BeforeHash added to IndentPPDirectives.
Fixes Bug 36019. https://bugs.llvm.org/show_bug.cgi?id=36019
Reviewers: djasper, klimek, krasimir, sammccall, mprobst, Nicola, MyDeveloperDay
Reviewed By: klimek, MyDeveloperDay
Subscribers: kadircet, MyDeveloperDay, mnussbaum, geleji, ufna, cfe-commits
Patch by to-mix.
Differential Revision: https://reviews.llvm.org/D52150
llvm-svn: 356613
Zachary Turner [Wed, 20 Mar 2019 20:49:25 +0000 (20:49 +0000)]
Introduce DWARFContext.
LLVM's DWARF parsing library has a class called DWARFContext which holds
all of the various DWARF data sections and lots of other information.
LLDB's on the other hand stores all of this directly in SymbolFileDWARF
/ SymbolFileDWARFDwo and passes this interface around through the
parsing library. Obviously this is incompatible with a world where the
low level interface does not depend on the high level interface, so we
need to move towards a model similar to LLVM's - i.e. all of the context
needed for low level parsing should be in a single class, and that class
gets passed around.
This patch is a small incremental step towards achieving this. The
interface and internals deviate from LLVM's for technical reasons, but
the high level idea is the same. The goal is, eventually, to remove all
occurrences of SymbolFileDWARF from the low level parsing code.
For now I've chosen a very simple section - the .debug_aranges section
to move into DWARFContext while leaving everything else unchanged. In
the short term this is a bit confusing because now the information you
need might come from either of 2 different locations. But it's a huge
refactor to do this all at once and runs a much higher risk of breaking
things. So I think it would be wise to do this in very small pieces.
TL;DR - No functional change
Differential Revision: https://reviews.llvm.org/D59562
llvm-svn: 356612
Matt Arsenault [Wed, 20 Mar 2019 20:41:34 +0000 (20:41 +0000)]
AMDGPU: Don't look for constant in insert/extract_vector_elt regbankselect
The constantness shouldn't change the register bank choice. We also
don't need to restrict this to only indexing VGPRs, since it's
possible to index SGPRs (but SelectionDAG made using this
difficult). Allow directly indexing SGPRs when appropriate.
llvm-svn: 356611
Thomas Lively [Wed, 20 Mar 2019 20:26:45 +0000 (20:26 +0000)]
[WebAssembly] Target features section
Summary:
Implements a new target features section in assembly and object files
that records what features are used, required, and disallowed in
WebAssembly objects. The linker uses this information to ensure that
all objects participating in a link are feature-compatible and records
the set of used features in the output binary for use by optimizers
and other tools later in the toolchain.
The "atomics" feature is always required or disallowed to prevent
linking code with stripped atomics into multithreaded binaries. Other
features are marked used if they are enabled globally or on any
function in a module.
Future CLs will add linker flags for ignoring feature compatibility
checks and for specifying the set of allowed features, implement using
the presence of the "atomics" feature to control the type of memory
and segments in the linked binary, and add front-end flags for
relaxing the linkage policy for atomics.
Reviewers: aheejin, sbc100, dschuff
Subscribers: jgravelle-google, hiraditya, sunfish, mgrang, jfb, jdoerfert, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D59173
llvm-svn: 356610
Craig Topper [Wed, 20 Mar 2019 20:25:28 +0000 (20:25 +0000)]
[X86] Add __crc32b/__crc32w/__crc32d/__crc32q intrinsics to match gcc and icc.
gcc has these intrinsics in ia32intrin.h as well. And icc implements them
though they aren't documented in the Intel Intrinsics Guide.
Differential Revision: https://reviews.llvm.org/D59533
llvm-svn: 356609
Michael Liao [Wed, 20 Mar 2019 20:18:56 +0000 (20:18 +0000)]
[AMDGPU] Fix clamp bit DAG operand
Summary:
- Should use `targetconstant` instead of `constant` operand for clamp
bit, which is expected as an immediate operand. Under certain
conditions, such as a common `i1 false` constant is used in other
place and selected before the instruction with clamp bit, register
operand may be added instead of immediate one. Use `targetcosntant` to
enforce that.
Subscribers: arsenm, kzhuravl, jvesely, wdng, nhaehnle, yaxunl, dstuttard, tpr, t-tye, hiraditya, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D59608
llvm-svn: 356608
Alexey Bataev [Wed, 20 Mar 2019 20:14:22 +0000 (20:14 +0000)]
[OPENMP]Improve detection of omp_allocator_handle_t type and predefined
allocators.
It is better to deduce omp_allocator_handle_t type from the predefined
allocators, because omp.h header might not define it explicitly. Plus,
it allows to identify the predefined allocators correctly when trying to
build the allcoator for the global variables.
llvm-svn: 356607
Pete Couperus [Wed, 20 Mar 2019 20:06:21 +0000 (20:06 +0000)]
[ARC] Add ARCOptAddrMode pass to generate postincrement loads/stores.
Build on newly introduced ARC postincrement loads/stores from r356200.
Patch By Denis Antrushin! <denis@synopsys.com>
Differential Revision: https://reviews.llvm.org/D59409
llvm-svn: 356606
Alex Lorenz [Wed, 20 Mar 2019 20:02:00 +0000 (20:02 +0000)]
Fix implicit ios -> watchOS availability version mapping for
versions that have the major number only
rdar://
48018651
llvm-svn: 356605
Evandro Menezes [Wed, 20 Mar 2019 19:57:59 +0000 (19:57 +0000)]
[AArch64] Fix formatting (NFC)
Indent macro instances properly.
llvm-svn: 356604
Konstantin Zhuravlyov [Wed, 20 Mar 2019 19:44:47 +0000 (19:44 +0000)]
AMDHSA: Fix COMPUTE_PGM_RSRC2.USER_SGPR calculation when parsing ISA assembly
It must match https://llvm.org/docs/AMDGPUUsage.html#initial-kernel-execution-state
Differential Revision: https://reviews.llvm.org/D59570
llvm-svn: 356603
Marshall Clow [Wed, 20 Mar 2019 19:43:22 +0000 (19:43 +0000)]
Mark <charconv> tests as unsupported for C++11 and C++14 if you're not testing libc++. Thanks to Louis for the suggestion.
llvm-svn: 356602
Eli Friedman [Wed, 20 Mar 2019 19:40:45 +0000 (19:40 +0000)]
[ARM] Eliminate redundant "mov rN, sp" instructions in Thumb1.
This takes sequences like "mov r4, sp; str r0, [r4]", and optimizes them
to something like "str r0, [sp]".
For regular stack variables, this optimization was already implemented:
we lower loads and stores using frame indexes, which are expanded later.
However, when constructing a call frame for a call with more than four
arguments, the existing optimization doesn't apply. We need to use
stores which are actually relative to the current value of sp, and don't
have an associated frame index.
This patch adds a special case to handle that construct. At the DAG
level, this is an ISD::STORE where the address is a CopyFromReg from SP
(plus a small constant offset).
This applies only to Thumb1: in Thumb2 or ARM mode, a regular store
instruction can access SP directly, so the COPY gets eliminated by
existing code.
The change to ARMDAGToDAGISel::SelectThumbAddrModeSP is a related
cleanup: we shouldn't pretend that it can select anything other than
frame indexes.
Differential Revision: https://reviews.llvm.org/D59568
llvm-svn: 356601
Erik Pilkington [Wed, 20 Mar 2019 19:26:37 +0000 (19:26 +0000)]
Add a __has_extension check for '#pragma clang attribute' as an external-declaration
This was added in r356075.
llvm-svn: 356600
Erik Pilkington [Wed, 20 Mar 2019 19:26:33 +0000 (19:26 +0000)]
[Sema] Deduplicate some availability checking logic
Before this commit, we emit unavailable errors for calls to functions during
overload resolution, and for references to all other declarations in
DiagnoseUseOfDecl. The early checks during overload resolution aren't as good as
the DiagnoseAvailabilityOfDecl based checks, as they error on the code from
PR40991. This commit fixes this by removing the early checking.
llvm.org/PR40991
rdar://
48564179
Differential revision: https://reviews.llvm.org/D59394
llvm-svn: 356599
Rafael Auler [Wed, 20 Mar 2019 19:22:24 +0000 (19:22 +0000)]
Recommit "Support attribute used in member funcs of class templates"
This diff previously exposed a bug in LLVM's IRLinker, breaking
buildbots that tried to self-host LLVM with monolithic LTO.
The bug is now in LLVM by D59552
Original commit message:
As PR17480 describes, clang does not support the used attribute
for member functions of class templates. This means that if the member
function is not used, its definition is never instantiated. This patch
changes clang to emit the definition if it has the used attribute.
Test Plan: Added a testcase
Reviewed By: aaron.ballman
Differential Revision: https://reviews.llvm.org/D56928
llvm-svn: 356598
Rafael Auler [Wed, 20 Mar 2019 19:20:07 +0000 (19:20 +0000)]
[Linker] Fix crash handling appending linkage
Summary:
When linking two llvm.used arrays, if the resulting merged
array ends up with duplicated elements (with the same name) but with
different types, the IRLinker was crashing. This was supposed to be
legal, as the IRLinker bitcasts elements to match types in these
situations.
This bug was exposed by D56928 in clang to support attribute used
in member functions of class templates. Crash happened when self-hosting
with LTO. Since LLVM depends on attribute used to generate code
for the dump() method, ubiquitous in the code base, many input bc
had a definition of this method referenced in their llvm.used array.
Some of these classes got optimized, changing the type of the first
parameter (this) in the dump method, leading to a scenario with a
pool of valid definitions but some with a different type, triggering
this bug.
This is a memory bug: ValueMapper depends on (calls) the materializer
provided by IRLinker, and this materializer was freely calling RAUW
methods whenever a global definition was updated in the temporary merged
output file. However, replaceAllUsesWith may or may not destroy
constants that use this global. If the linked definition has a type
mismatch regarding the new def and the old def, the materializer would
bitcast the old type to the new type and the elements of the llvm.used
array, which already uses bitcast to i8*, would end up with elements
cascading two bitcasts. RAUW would then indirectly call the
constantfolder to update the constant to the new ref, which would,
instead of updating the constant, destroy it to be able to create
a new constant that folds the two bitcasts into one. The problem is that
ValueMapper works with pointers to the same constants that may be
getting destroyed by RAUW. Obviously, RAUW can update references in the
Module to do not use the old destroyed constant, but it can't update
ValueMapper's internal pointers to these constants, which are now
invalid.
The approach here is to move the task of RAUWing old definitions
outside of the materializer.
Test Plan:
Added LIT test case, tested clang self-hosting with D56928 and
verified it works
Reviewed By: efriedma
Differential Revision: https://reviews.llvm.org/D59552
llvm-svn: 356597
Alina Sbirlea [Wed, 20 Mar 2019 19:18:55 +0000 (19:18 +0000)]
[NFC] Fix brace indentation.
llvm-svn: 356596
Louis Dionne [Wed, 20 Mar 2019 19:12:27 +0000 (19:12 +0000)]
[libc++] Use the compiler that CMake found when running lit for back-deployment tests
llvm-svn: 356595
Robert Lougher [Wed, 20 Mar 2019 19:08:18 +0000 (19:08 +0000)]
Resubmit r356511 "[TailCallElim] Add tailcall elimination pass to LTO pipelines"
Failing LLD tests have been fixed in r356593.
llvm-svn: 356594
Robert Lougher [Wed, 20 Mar 2019 19:05:05 +0000 (19:05 +0000)]
[TailCallElim] Update tests for LTO pipeline change
D58391 changed the LTO pipelines to add the tailcall elimination pass.
This caused three LLD tests to fail.
Differential Revision: https://reviews.llvm.org/D59604
llvm-svn: 356593
Raphael Isemann [Wed, 20 Mar 2019 19:00:25 +0000 (19:00 +0000)]
Remove the unused return value in ASTImporter::Imported [NFC]
Summary:
`ASTImporter::Imported` currently returns a Decl, but that return value is not used by the ASTImporter (or anywhere else)
nor is it documented.
Reviewers: balazske, martong, a.sidorin, shafik
Reviewed By: balazske, martong
Subscribers: rnkovacs, cfe-commits
Tags: #clang
Differential Revision: https://reviews.llvm.org/D59595
llvm-svn: 356592
Tim Renouf [Wed, 20 Mar 2019 18:47:21 +0000 (18:47 +0000)]
[AMDGPU] Added MsgPack format PAL metadata
Summary:
PAL metadata now supports both the old linear reg=val pairs format and
the new MsgPack format.
The MsgPack format uses YAML as its textual representation. On output to
YAML, a mnemonic name is provided for some hardware registers.
Differential Revision: https://reviews.llvm.org/D57028
Change-Id: I2bbaabaaca4b3574f7e03b80fbef7c7a69d06a94
llvm-svn: 356591
Philip Reames [Wed, 20 Mar 2019 18:44:58 +0000 (18:44 +0000)]
Simplify operands of masked stores and scatters based on demanded elements
If we know we're not storing a lane, we don't need to compute the lane. This could be improved by using the undef element result to further prune the mask, but I want to separate that into its own change since it's relatively likely to expose other problems.
Differential Revision: https://reviews.llvm.org/D57247
llvm-svn: 356590
Zinovy Nis [Wed, 20 Mar 2019 18:37:04 +0000 (18:37 +0000)]
[clang-tidy] Fix redundant check breaking the test on many platforms.
Differential Revision: https://reviews.llvm.org/D57662
llvm-svn: 356589
Alina Sbirlea [Wed, 20 Mar 2019 18:33:37 +0000 (18:33 +0000)]
[LICM & MemorySSA] Don't sink/hoist stores in the presence of ordered loads.
Summary:
Before this patch, if any Use existed in the loop, with a defining
access in the loop, we conservatively decide to not move the store.
What this approach was missing, is that ordered loads are not Uses, they're Defs
in MemorySSA. So, even when the clobbering walker does not find that
volatile load to interfere, we still cannot hoist a store past a
volatile load.
Resolves PR41140.
Reviewers: george.burgess.iv
Subscribers: sanjoy, jlebar, Prazek, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D59564
llvm-svn: 356588
Louis Dionne [Wed, 20 Mar 2019 18:16:24 +0000 (18:16 +0000)]
[libc++][CMake] Clean up some of the libc++ re-exporting logic
Summary:
This change allows specifying the version of libc++abi's ABI to re-export
when configuring CMake. It also clearly identifies which ABI version of
libc++abi each export file contains.
Finally, it removes hardcoded knowledge about the 10.9 SDK for MacOS,
since that knowledge is not relevant anymore. Indeed, libc++ can't be
built with the toolchain that came with the 10.9 SDK anyway because
the version of Clang it includes is too old (for example if you want
to build a working libc++.dylib, you need bugfixes to visibility
attributes that are only in recent Clangs).
Reviewers: dexonsmith, EricWF
Subscribers: mgorny, christof, jkorous, arphaman, libcxx-commits
Differential Revision: https://reviews.llvm.org/D59489
llvm-svn: 356587
Nikita Popov [Wed, 20 Mar 2019 18:16:02 +0000 (18:16 +0000)]
[ValueTracking] Compute range for abs without nsw
This is a small followup to D59511. The code that was moved into
computeConstantRange() there is a bit overly conversative: If the
abs is not nsw, it does not compute any range. However, abs without
nsw still has a well-defined contiguous unsigned range from 0 to
SIGNED_MIN. This is a lot less useful than the usual 0 to SIGNED_MAX
range, but if we're already here we might as well specify it...
Differential Revision: https://reviews.llvm.org/D59563
llvm-svn: 356586
Marshall Clow [Wed, 20 Mar 2019 18:13:23 +0000 (18:13 +0000)]
Make to_chars/from_chars work back to C++11. This means that we can use them to implement to_string as well. Reviewed as https://reviews.llvm.org/D59598.
llvm-svn: 356585
Nikita Popov [Wed, 20 Mar 2019 18:00:27 +0000 (18:00 +0000)]
[InstCombine] Fold add nuw + uadd.with.overflow
Fold add nuw and uadd.with.overflow with constants if the
addition does not overflow.
Part of https://bugs.llvm.org/show_bug.cgi?id=38146.
Patch by Dan Robertson.
Differential Revision: https://reviews.llvm.org/D59471
llvm-svn: 356584
Jordan Rupprecht [Wed, 20 Mar 2019 17:44:24 +0000 (17:44 +0000)]
[Remarks] Fix mismatched delete due to missing virtual destructor
This fixes an asan failure introduced in r356519.
llvm-svn: 356583
Tim Renouf [Wed, 20 Mar 2019 17:42:00 +0000 (17:42 +0000)]
[AMDGPU] Factored PAL metadata handling out into its own class
Summary:
This commit introduces a new AMDGPUPALMetadata class that:
* is inside the AMDGPU target;
* keeps an in-memory representation of PAL metadata;
* provides a method to read the frontend-supplied metadata from LLVM IR;
* provides methods for the asm printer to set metadata items;
* provides methods to write the metadata as a binary blob to put in a
.note record or as an asm directive;
* provides a method to read the metadata as a binary blob from a .note
record.
Because llvm-readobj cannot call directly into a target, I had to remove
llvm-readobj's ability to dump PAL metadata, pending a resolution to
https://reviews.llvm.org/D52821
Differential Revision: https://reviews.llvm.org/D57027
Change-Id: I756dc830894fcb6850324cdcfa87c0120eb2cf64
llvm-svn: 356582
Sterling Augustine [Wed, 20 Mar 2019 17:37:23 +0000 (17:37 +0000)]
Make __cpu_model a hidden symbol, to match libgcc.
Also hide __cpu_inicator_init and __cpu_features2
for similar reasons.
Summary: Make __cpu_model a hidden symbol, to match libgcc.
Subscribers: llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D59561
llvm-svn: 356581
Craig Topper [Wed, 20 Mar 2019 17:26:51 +0000 (17:26 +0000)]
[X86] Remove getCPUKindCanonicalName which is unused.
Differential Revision: https://reviews.llvm.org/D59578
llvm-svn: 356580
Nico Weber [Wed, 20 Mar 2019 17:26:11 +0000 (17:26 +0000)]
Remove HAVE_REALPATH from config.h
Its last use was removed in r352916.
No behavior change.
Differential Revision: https://reviews.llvm.org/D59601
llvm-svn: 356579
Roman Lebedev [Wed, 20 Mar 2019 17:15:47 +0000 (17:15 +0000)]
[NFC][ASTMatchers] Alphabetically sort REGISTER_MATCHER() macros in RegistryMaps::RegistryMaps()
As noted in https://reviews.llvm.org/D59453#inline-526253
llvm-svn: 356578
Roman Lebedev [Wed, 20 Mar 2019 17:14:49 +0000 (17:14 +0000)]
[AST] Disable ast-dump-openmp-parallel-master-XFAIL.c test
Fails on MSVC buildbot (but not locally).
Not important as it is 'testing' something that isn't supported yet anyway:
https://bugs.llvm.org/show_bug.cgi?id=41022
llvm-svn: 356577
Dmitry Preobrazhensky [Wed, 20 Mar 2019 17:13:58 +0000 (17:13 +0000)]
[AMDGPU][MC] Corrected checks for DS offset0 range
See bug 40889: https://bugs.llvm.org/show_bug.cgi?id=40889
Reviewers: artem.tamazov, arsenm
Differential Revision: https://reviews.llvm.org/D59313
llvm-svn: 356576
Paul Hoad [Wed, 20 Mar 2019 17:10:23 +0000 (17:10 +0000)]
[clang-format] structured binding in range for detected as Objective C
Summary:
Sometime after 6.0.0 and the current trunk 9.0.0 the following code would be considered as objective C and not C++
Reported by: https://twitter.com/mattgodbolt/status/
1096188576503644160
$ clang-format.exe test.h
Configuration file(s) do(es) not support Objective-C: C:\clang\build\.clang-format
--- test.h --
```
std::vector<std::pair<std::string,std::string>> C;
void foo()
{
for (auto && [A,B] : C)
{
std::string D = A + B;
}
}
```
The following code fixes this issue of incorrect detection
Reviewers: djasper, klimek, JonasToth, reuk
Reviewed By: klimek
Subscribers: cfe-commits
Tags: #clang-tools-extra
Differential Revision: https://reviews.llvm.org/D59546
llvm-svn: 356575
Louis Dionne [Wed, 20 Mar 2019 17:05:52 +0000 (17:05 +0000)]
[NFC] Fix a couple of typos in libc++'s __config
llvm-svn: 356574
Greg Clayton [Wed, 20 Mar 2019 16:50:17 +0000 (16:50 +0000)]
Fix UUID decoding from minidump files
This patch fixes:
UUIDs now don't include the age field from a PDB70 when the age is zero. Prior to this they would incorrectly contain the zero age which stopped us from being able to match up the UUID with real files.
UUIDs for Apple targets get the first 32 bit value and next two 16 bit values swapped. Breakpad incorrectly swaps these values when it creates darwin minidump files, so this must be undone so we can match up symbol files with the minidump modules.
UUIDs that are all zeroes are treated as invalid UUIDs. Breakpad will always save out a UUID, even if one wasn't available. This caused all files that have UUID values of zero to be uniqued to the first module that had a zero UUID. We now don't fill in the UUID if it is all zeroes.
Added tests for PDB70 and ELF build ID based CvRecords.
Differential Revision: https://reviews.llvm.org/D59433
llvm-svn: 356573
Sanjay Patel [Wed, 20 Mar 2019 16:47:53 +0000 (16:47 +0000)]
[CGP] fix formatting; NFC
llvm-svn: 356572