Djordje Todorovic [Wed, 20 Nov 2019 11:20:53 +0000 (12:20 +0100)]
[DebugInfo] Remove the DIFlagArgumentNotModified debug info flag
Due to changes in D68206, we remove the DIFlagArgumentNotModified
and its usage.
Differential Revision: https://reviews.llvm.org/D68207
LLVM GN Syncbot [Wed, 20 Nov 2019 12:10:51 +0000 (12:10 +0000)]
gn build: Merge
ea8678d1c78
Raphael Isemann [Wed, 20 Nov 2019 11:56:42 +0000 (12:56 +0100)]
[lldb][NFC] Remove unused ClangASTContext::GetUnknownAnyType
Sam McCall [Tue, 19 Nov 2019 15:57:31 +0000 (16:57 +0100)]
[clangd] Treat UserDefinedLiteral as a leaf in SelectionTree, sidestepping tokenization issues
Summary: Fixes https://github.com/clangd/clangd/issues/203
Reviewers: kadircet
Subscribers: ilya-biryukov, MaskRay, jkorous, arphaman, usaxena95, cfe-commits
Tags: #clang
Differential Revision: https://reviews.llvm.org/D70446
Serge Pavlov [Thu, 29 Aug 2019 12:29:11 +0000 (19:29 +0700)]
Move floating point related entities to namespace level
This is recommit of commit
e6584b2b7b2d, which was reverted in
30e7ee3c4bac together with
af57dbf12e54.
Original message is below.
Enumerations that describe rounding mode and exception behavior were
defined inside ConstrainedFPIntrinsic. It makes sense to use the same
definitions to represent the same properties in other cases, not only
in constrained intrinsics. It was however inconvenient as required to
include constrained intrinsics definitions even if they were not needed.
Also using long scope prefix reduced readability.
This change moves these definitioins to the namespace llvm::fp.
No functional changes.
Differential Revision: https://reviews.llvm.org/D69552
Raphael Isemann [Wed, 20 Nov 2019 11:40:08 +0000 (12:40 +0100)]
[lldb][NFC] Simplify ClangASTContext::GetBasicTypes
static convenience methods that do the clang::ASTContext -> ClangASTContext
conversion and handle errors by simply ignoring them are not a good idea.
Raphael Isemann [Wed, 20 Nov 2019 11:17:48 +0000 (12:17 +0100)]
[lldb][NFC] Remove ClangASTContext::GetAsDeclContext
Everything we pass to this function is already a DeclContext.
Raphael Isemann [Wed, 20 Nov 2019 11:09:19 +0000 (12:09 +0100)]
[lldb][NFC] Remove ClangASTContext::FieldIsBitfield overload
Raphael Isemann [Wed, 20 Nov 2019 10:36:15 +0000 (11:36 +0100)]
[lldb][NFC] Remove ClangASTContext::GetUniqueNamespaceDeclaration overload
This overload is only used in one place and having static overloads for
all methods that only do an additional clang::ASTContext -> ClangASTContext
conversion is just not sustainable.
Sameer Sahasrabuddhe [Tue, 19 Nov 2019 08:35:39 +0000 (14:05 +0530)]
[AMDGPU] add support for hostcall buffer pointer as hidden kernel argument
Hostcall is a service that allows a kernel to submit requests to the
host using shared buffers, and block until a response is
received. This will eventually replace the shared buffer currently
used for printf, and repurposes the same hidden kernel argument. This
change introduces a new ValueKind in the HSA metadata to represent the
hostcall buffer.
Differential Revision: https://reviews.llvm.org/D70038
Martin Storsjö [Wed, 20 Nov 2019 10:19:50 +0000 (12:19 +0200)]
[ExecutionEngine] Add a missing break to avoid warnings
This fixes buildbot errors since
dc3ee330891c2.
Georgii Rymar [Tue, 12 Nov 2019 09:46:44 +0000 (12:46 +0300)]
[llvm-readobj] - Improve dumping of the SHT_LLVM_LINKER_OPTIONS sections.
I've added a few tests that shows how the current code could overrun the section data
buffer while dumping. I had to rewrite the code to fix this.
Differential revision: https://reviews.llvm.org/D70112
Djordje Todorovic [Tue, 19 Nov 2019 13:47:08 +0000 (14:47 +0100)]
Reland "[clang] Remove the DIFlagArgumentNotModified debug info flag"
It turns out that the ExprMutationAnalyzer can be very slow when AST
gets huge in some cases. The idea is to move this analysis to the LLVM
back-end level (more precisely, in the LiveDebugValues pass). The new
approach will remove the performance regression, simplify the
implementation and give us front-end independent implementation.
Differential Revision: https://reviews.llvm.org/D68206
Adam Kallai [Wed, 20 Nov 2019 08:51:03 +0000 (10:51 +0200)]
ExecutionEngine: add preliminary support for COFF ARM64
Differential Revision: https://reviews.llvm.org/D69434
Georgii Rymar [Mon, 18 Nov 2019 14:43:50 +0000 (17:43 +0300)]
[llvm-readobj/llvm-readelf] - Improve dumping of versioning sections.
Our elf-versioninfo.test is not perfect. It does not properly test how
flags are dumped and also we have a bug: they are dumped as enums in
LLVM style now, i.e not dumped properly.
GNU style uses a `versionFlagToString` method to build a string from flags
which seems is consistent with GNU readelf.
In this patch I fixed the issues mentioned.
Differential revision: https://reviews.llvm.org/D70399
Alexander Shaposhnikov [Wed, 20 Nov 2019 07:30:52 +0000 (23:30 -0800)]
Introduce llvm-install-name-tool
This diff adds a new "driver" for llvm-objcopy
which is supposed to emulate the behavior of install-name-tool.
This is a recommit of
b5913e6 with ubsan, test dependencies issues fixed.
Differential revision: https://reviews.llvm.org/D69146
Test plan: make check-all
Fangrui Song [Wed, 20 Nov 2019 07:09:07 +0000 (23:09 -0800)]
[FEnv] Fix AddingConstrainedIntrinsics.rst after
llvmorg-10-init-10282-g0c50c0b0552
Serge Pavlov [Tue, 5 Nov 2019 13:42:16 +0000 (20:42 +0700)]
[FEnv] File with properties of constrained intrinsics
Summary
In several places we need to enumerate all constrained intrinsics or IR
nodes that should be represented by them. It is easy to miss some of
the cases. To make working with these intrinsics more convenient and
robust, this change introduces file containing definitions of all
constrained intrinsics and some of their properties. This file can be
included to generate constrained intrinsics processing code.
Reviewers: kpn, andrew.w.kaylor, cameron.mcinally, uweigand
Subscribers: hiraditya, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D69887
Austin Kerbow [Mon, 18 Nov 2019 00:43:59 +0000 (16:43 -0800)]
AMDGPU/GlobalISel: Legalize FDIV64
Reviewers: arsenm
Reviewed By: arsenm
Subscribers: kzhuravl, jvesely, wdng, nhaehnle, yaxunl, rovka, dstuttard, tpr, t-tye, hiraditya, Petar.Avramovic, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D70403
Rui Ueyama [Tue, 19 Nov 2019 06:04:21 +0000 (15:04 +0900)]
Use lld::make<T> to make TpiSource objects
In lld we rarely use std::unique_ptr but instead allocate new instances
using lld::make<T>() so that they are deallocated at the end of linking.
This patch changes existing code so that that follows the convention.
Differential Revision: https://reviews.llvm.org/D70420
Eric Schweitz [Wed, 20 Nov 2019 03:28:12 +0000 (19:28 -0800)]
Extends the tblgen macro to allow mlir-tblgen to be installed
The mlir-tblgen tool was not getting installed. This change allows
the MLIR project to be installed along with llvm-tblgen.
Differential Revision: https://reviews.llvm.org/D69824
Puyan Lotfi [Wed, 20 Nov 2019 02:59:50 +0000 (21:59 -0500)]
Revert "[clang][IFS] Driver Pipeline: generate stubs after standard pipeline (2)"
This reverts commit
ea8e02822341e2421b94167d828d3f224e767424.
Puyan Lotfi [Wed, 20 Nov 2019 02:59:39 +0000 (21:59 -0500)]
Revert "[clang][IFS] Fixing failing bots that do not have PPC target or "orbis-ld""
This reverts commit
377d70cdea733e36107e99d9148864d24797d51c.
Puyan Lotfi [Wed, 20 Nov 2019 02:59:31 +0000 (21:59 -0500)]
Revert "[clang][IFS] Attempting to fix missing 'orbis-ld' on scei-ps4-ubuntu bot."
This reverts commit
29fd1e1f4a372f3870e054da24b57a4f45861808.
Puyan Lotfi [Wed, 20 Nov 2019 02:59:22 +0000 (21:59 -0500)]
Revert "[clang][IFS] Fixing unsupported emulation mode on clang-ppc64be-linux bot."
This reverts commit
1b387484b9b38a4a1e98a9d22a9a26065b0d184e.
Puyan Lotfi [Wed, 20 Nov 2019 02:59:10 +0000 (21:59 -0500)]
Revert "[clang][IFS][test] Removing driver-test.c. Test is still too brittle."
This reverts commit
f37356d6f60ae5db978611621d3a375ed87ec0f0.
Puyan Lotfi [Wed, 20 Nov 2019 02:42:17 +0000 (21:42 -0500)]
[clang][IFS][test] Removing driver-test.c. Test is still too brittle.
Removing this test because if I add a triple then there are link falures
on targets like ppc and s390x. If I don't add a triple then on PS4
targets the clang driver tries to invoke orbis-ld which ends up being
not found.
Duncan P. N. Exon Smith [Wed, 20 Nov 2019 02:10:04 +0000 (18:10 -0800)]
clang/Modules: Early return in CompilerInstance::createModuleManager, NFC
Reduce nesting with an early `return`.
Reid Kleckner [Mon, 18 Nov 2019 23:31:37 +0000 (15:31 -0800)]
[musttail] Don't forward AL on Win64
AL is only used for varargs on SysV platforms. Don't forward it on
Windows. This allows control flow guard to set up an extra hidden
parameter in RAX, as described in PR44049.
This also has the effect of freeing up RAX for use in virtual member
pointer thunks, which may also be a nice little code size improvement on
Win64.
Fixes PR44049
Reviewers: ajpaverd, efriedma, hans
Differential Revision: https://reviews.llvm.org/D70413
Francis Visoiu Mistrih [Tue, 19 Nov 2019 22:04:59 +0000 (14:04 -0800)]
[LTO][Legacy] Add API for passing LLVM options separately
In order to correctly pass options to LLVM, including options containing
spaces which are used as delimiters for multiple options in
lto_codegen_debug_options, add a new API:
lto_codegen_debug_options_array.
Unfortunately, tools/lto has no testing infrastructure yet, so there are
no tests associated with this patch.
Differential Revision: https://reviews.llvm.org/D70463
Puyan Lotfi [Wed, 20 Nov 2019 00:24:12 +0000 (19:24 -0500)]
[clang][IFS] Fixing unsupported emulation mode on clang-ppc64be-linux bot.
I am in another pickle here where if I specify a triple, I get the wrong elf
target arch on the PPC bot (error from the PPC elf Linker). To avoid this I am
going to turn this test off on the PPC bots for now.
Alexander Shaposhnikov [Wed, 20 Nov 2019 00:19:08 +0000 (16:19 -0800)]
Revert "Introduce llvm-install-name-tool"
This temporarily reverts the commit
88f00aef684ff84a6494e1f17d5466c5678f703d.
The change broke the buildbot http://lab.llvm.org:8011/builders/sanitizer-x86_64-linux-fast/builds/36836
Jonas Devlieghere [Wed, 20 Nov 2019 00:18:19 +0000 (16:18 -0800)]
[Reproducer] Namespace the reproducer dump options.
Make it clear that the current reproducer options are for dumping.
Craig Topper [Wed, 20 Nov 2019 00:07:50 +0000 (16:07 -0800)]
[LegalizeDAG][X86] Enable STRICT_FP_TO_SINT/UINT to be promoted
Differential Revision: https://reviews.llvm.org/D70220
Peter Collingbourne [Tue, 19 Nov 2019 21:54:12 +0000 (13:54 -0800)]
scudo: Switch from std::random_shuffle to std::shuffle in a test.
This lets the test build with C++17.
Differential Revision: https://reviews.llvm.org/D70471
Duncan P. N. Exon Smith [Tue, 19 Nov 2019 22:44:22 +0000 (14:44 -0800)]
clang/Modules: Sink ASTReadResult in ReadControlBlock, NFC
Simplify the code by avoiding some state that wasn't being used. The
function-level `Result` was only assigned a value other than `Success`
in the handler for `OPTIONS_BLOCK_ID`, but in that case it also hits an
early return. Remove it at the function-level to make it obvious that
the normal case always returns `Success`.
Craig Topper [Tue, 19 Nov 2019 23:51:19 +0000 (15:51 -0800)]
[X86] Add custom type legalization and lowering for scalar STRICT_FP_TO_SINT/UINT
This is a first pass at Custom lowering for these operations. I also updated some of the vector code where it was obviously easy and straightforward. More work needed in follow up.
This enables these operations to be handled with X87 where special rounding control adjustments are needed to perform a truncate.
Still need to fix Promotion in the target independent code in LegalizeDAG.
llrint/llround split into separate test file because we can't make a strict libcall properly yet either and we need to do that when i64 isn't a legal type.
This does not include any isel support. So we still rely on the mutation in SelectionDAGIsel to remove the strict from this stuff later. Except for the X87 stuff which goes through custom nodes that already had chains.
Differential Revision: https://reviews.llvm.org/D70214
Philip Reames [Wed, 20 Nov 2019 00:01:43 +0000 (16:01 -0800)]
[tests] Autogen a test to eliminate spurious diff from following patch
Pete Couperus [Tue, 19 Nov 2019 23:48:24 +0000 (15:48 -0800)]
[ARC] Add InitializePasses header to fix ARC build.
Puyan Lotfi [Tue, 19 Nov 2019 23:32:31 +0000 (18:32 -0500)]
[clang][IFS] Attempting to fix missing 'orbis-ld' on scei-ps4-ubuntu bot.
I want this test to run end to end, but I am still having trouble with
missing linkers on the scei-ps4 bot. Will remove this test if it
continues to be a source of brittle failures. Sorry for the noise.
Jordan Rupprecht [Tue, 19 Nov 2019 23:06:35 +0000 (15:06 -0800)]
[lldb][test] Prevent \n in calls to lldb's expect() test helper.
Summary:
expect() forwards its command to sendline(). This can be problematic if the command already contains a newline: sendline() unconditionally adds a newline to the command, which causes the command to run twice (hitting enter in lldb runs the previous command). The expect() helper looks for the prompt and finds the first one, but because the command has run a second time, the buffer will contain the contents of the second time the command ran, causing potential erroneous matching.
Simplify the editline test, which was using different commands to workaround this misunderstanding.
Reviewers: labath
Reviewed By: labath
Subscribers: merge_guards_bot, lldb-commits
Tags: #lldb
Differential Revision: https://reviews.llvm.org/D70324
Alexander Shaposhnikov [Mon, 18 Nov 2019 23:44:58 +0000 (15:44 -0800)]
Introduce llvm-install-name-tool
This diff adds a new "driver" for llvm-objcopy
which is supposed to emulate the behavior of install-name-tool.
This is a recommit of
b5913e6d2 with ubsan issues fixed.
Differential revision: https://reviews.llvm.org/D69146
Test plan: make check-all
Philip Reames [Tue, 19 Nov 2019 23:14:41 +0000 (15:14 -0800)]
[GuardWidening] Remove WidenFrequentBranches transform
This code has never been enabled. While it is tested, it's complicating some refactoring. If we decide to re-implement this, doing it in SimplifyCFG would probably make more sense anyways.
Puyan Lotfi [Tue, 19 Nov 2019 23:12:07 +0000 (18:12 -0500)]
[clang][IFS] Fixing failing bots that do not have PPC target or "orbis-ld"
Puyan Lotfi [Tue, 19 Nov 2019 08:07:47 +0000 (03:07 -0500)]
[clang][IFS] Driver Pipeline: generate stubs after standard pipeline (2)
Second Landing Attempt:
Up until now, clang interface stubs has replaced the standard
PP -> C -> BE -> ASM -> LNK pipeline. With this change, it will happen in
conjunction with it. So what when you build your code you will get an
a.out or lib.so as well as an interface stub file.
Example:
clang -shared -o libfoo.so -emit-interface-stubs ...
will generate both a libfoo.so and a libfoo.ifso. The .so file will
contain the code from the standard compilation pipeline and the .ifso
file will contain the ELF stub library.
Differential Revision: https://reviews.llvm.org/D70274
Philip Reames [Tue, 19 Nov 2019 22:43:13 +0000 (14:43 -0800)]
[NFC] Factor out utilities for manipulating widenable branches
With the widenable condition construct, we have the ability to reason about branches which can be 'widened' (i.e. made to fail more often). We've got a couple o transforms which leverage this. This patch just cleans up the API a bit.
This is prep work for generalizing our definition of a widenable branch slightly. At the moment "br i1 (and A, wc()), ..." is considered widenable, but oddly, neither "br i1 (and wc(), B), ..." or "br i1 wc(), ..." is. That clearly needs addressed, so first, let's centralize the code in one place.
Philip Reames [Tue, 19 Nov 2019 22:06:36 +0000 (14:06 -0800)]
[LoopPred] Generalize profitability check to handle unswitch output
Unswitch (and other loop transforms) like to generate loop exit blocks with unconditional successors, and phi nodes (LCSSA, or simple multiple exiting blocks sharing an exit). Generalize the "likely very rare exit" check slightly to handle this form.
Amara Emerson [Tue, 19 Nov 2019 21:38:55 +0000 (13:38 -0800)]
[AArch64] Fix MIR test instruction to not have invalid operand.
In anticipation of an improved verifier in D63973.
Benjamin Kramer [Tue, 19 Nov 2019 18:12:45 +0000 (19:12 +0100)]
[ValueTracking] Add a basic version of isKnownNonInfinity and use it to detect more NoNaNs
Duncan P. N. Exon Smith [Thu, 14 Nov 2019 21:57:57 +0000 (13:57 -0800)]
Wrap C APIs with pragmas enforcing -Werror=strict-prototypes
Force `-Werror=strict-prototypes` so that C API tests fail to compile if
we add a non-prototype declaration. This should help avoid regressions
like
bddecba4b333f7772029b4937d2c34f9f2fda6ca was fixing.
https://reviews.llvm.org/D70285
rdar://problem/
57203137
Philip Reames [Tue, 19 Nov 2019 19:48:51 +0000 (11:48 -0800)]
Precommit test showing oppurtunity when computing exit tests of unsimplified IR
If we partially unswitch a loop, we leave around the (and i1 X, true) or (or i1 X, false) forms. At the moment, this inhibits SCEVs ability to compute trip counts, patch forthcoming.
Alexey Bataev [Tue, 19 Nov 2019 17:07:54 +0000 (12:07 -0500)]
[OPENMP50]Add if clause in simd directive.
According to OpenMP 5.0, if clause can be used in simd directive. If
condition in the if clause if false, the non-vectorized version of the
loop must be executed.
Vedant Kumar [Tue, 19 Nov 2019 20:51:10 +0000 (12:51 -0800)]
[profile] Address unused function warnings on Windows after D69586
This '#ifdef's out two functions which are unused on Windows, to prevent
-Wunused-function warnings.
Vedant Kumar [Thu, 14 Nov 2019 02:19:32 +0000 (18:19 -0800)]
[CGDebugInfo] Emit subprograms for decls when AT_tail_call is understood (reland with fixes)
Currently, clang emits subprograms for declared functions when the
target debugger or DWARF standard is known to support entry values
(DW_OP_entry_value & the GNU equivalent).
Treat DW_AT_tail_call the same way to allow debuggers to follow cross-TU
tail calls.
Pre-patch debug session with a cross-TU tail call:
```
* frame #0: 0x0000000100000fa4 main`target at b.c:4:3 [opt]
frame #1: 0x0000000100000f99 main`main at a.c:8:10 [opt]
```
Post-patch (note that the tail-calling frame, "helper", is visible):
```
* frame #0: 0x0000000100000fa4 main`target at b.c:4:3 [opt]
frame #1: 0x0000000100000f80 main`helper [opt] [artificial]
frame #2: 0x0000000100000f99 main`main at a.c:8:10 [opt]
```
This was reverted in
5b9a072c because it attached declaration
subprograms to inlinable builtin calls, which interacted badly with the
MergeICmps pass. The fix is to not attach declarations to builtins.
rdar://
46577651
Differential Revision: https://reviews.llvm.org/D69743
Tim Northover [Tue, 19 Nov 2019 20:46:20 +0000 (20:46 +0000)]
[docs] Remove dangling parenthesis from documentation
Patch by leiteg.
Reid Kleckner [Fri, 15 Nov 2019 19:22:02 +0000 (11:22 -0800)]
Add a key method to Sema to optimize debug info size
It turns out that the debug info describing the Sema class is an
appreciable percentage of the total object file size of objects in Sema.
By adding a key function, clang is able to optimize the debug info size
by emitting a forward declaration in TUs that do not define the key
function.
On Windows, with clang-cl, these are the total sizes of object files in
Sema before and after this change, compiling with optimizations and
debug info:
before: 335,012 KB
after: 278,116 KB
delta: -56,896 KB
percent: -17.0%
The effect on link time was negligible, despite having ~56MB less input.
On Linux, with clang, these are the same sizes using DWARF -g and
optimizations:
before: 603,756 KB
after: 515,340 KB
delta: -88,416 KB
percent: -14.6%
I didn't use type units, DWARF-5, fission, or any other special flags.
Reviewed By: thakis
Differential Revision: https://reviews.llvm.org/D70340
Vedant Kumar [Tue, 19 Nov 2019 20:29:56 +0000 (12:29 -0800)]
[profile] Unbreak Fuchsia/Windows after D68351
Continuous mode is not yet supported on Fuchsia/Windows, however an
error should not be reported unless the user attempted to actually
enable continuous mode.
diggerlin [Tue, 19 Nov 2019 20:17:32 +0000 (15:17 -0500)]
The patch is the compiler error specific on the compile error on CMVC
SUMMARY:
CMVC has a compiler error on the
const uint64_t OffsetToRaw = is64Bit()
? toSection64(Sec)->FileOffsetToRawData
: toSection32(Sec)->FileOffsetToRawData;
while gcc compiler do not have the problem.
I have to change the code to
uint64_t OffsetToRaw;
if (is64Bit())
OffsetToRaw = toSection64(Sec)->FileOffsetToRawData;
else
OffsetToRaw = toSection32(Sec)->FileOffsetToRawData;
Reviewers: Sean Fertile
Subscribers: rupprecht, seiyai,hiraditya
Differential Revision: https://reviews.llvm.org/D70255
Tyker [Tue, 19 Nov 2019 20:05:30 +0000 (21:05 +0100)]
fixe leak found by asan build bot
Vedant Kumar [Tue, 19 Nov 2019 19:58:14 +0000 (11:58 -0800)]
[DebugInfo] Describe size of spilled values in call site params
A call site parameter description of a memory operand needs to
unambiguously convey the size of the operand to prevent incorrect entry
value evaluation.
Thanks for David Stenberg for pointing this issue out!
Duncan P. N. Exon Smith [Mon, 18 Nov 2019 01:58:01 +0000 (17:58 -0800)]
llvm/ObjCARC: Eliminate inlined AutoreleaseRV calls
Pair up inlined AutoreleaseRV calls with their matching RetainRV or
ClaimRV.
- RetainRV cancels out AutoreleaseRV. Delete both instructions.
- ClaimRV is a peephole for RetainRV+Release. Delete AutoreleaseRV and
replace ClaimRV with Release.
This avoids problems where more aggressive inlining triggers memory
regressions.
This patch is happy to skip over non-callable instructions and non-ARC
intrinsics looking for the pair. It is likely sound to also skip over
opaque function calls, but that's harder to reason about, and it's not
relevant to the goal here: if there's an opaque function call splitting
up a pair, it's very unlikely that a handshake would have happened
dynamically without inlining.
Note that this patch also subsumes the previous logic that looked
backwards from ReleaseRV.
https://reviews.llvm.org/D70370
rdar://problem/
46509586
Sanjay Patel [Tue, 19 Nov 2019 17:24:55 +0000 (12:24 -0500)]
[SLP] fix miscompile on min/max reductions with extra uses (PR43948) (2nd try)
The 1st attempt was reverted because it revealed an existing
bug where we could produce invalid IR (use of value before
definition). That should be fixed with:
rG39de82ecc9c2
The bug manifests as replacing a reduction operand with an undef
value.
The problem appears to be limited to cases where a min/max reduction
has extra uses of the compare operand to the select.
In the general case, we are tracking "ExternallyUsedValues" and
an "IgnoreList" of the reduction operations, but those may not apply
to the final compare+select in a min/max reduction.
For that, we use replaceAllUsesWith (RAUW) to ensure that the new
vectorized reduction values are transferred to all subsequent users.
Differential Revision: https://reviews.llvm.org/D70148
JonChesterfield [Tue, 19 Nov 2019 19:41:59 +0000 (19:41 +0000)]
[nfc][libomptarget] Remove casts of string literals to char*
Evgenii Stepanov [Fri, 15 Nov 2019 21:29:54 +0000 (13:29 -0800)]
MTE: add more unchecked instructions.
Summary:
In particular, 1- and 2-byte loads and stores ignore the pointer tag
when using SP as the base register.
Reviewers: pcc, ostannard
Subscribers: hiraditya, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D70341
Tom Stellard [Tue, 19 Nov 2019 18:42:56 +0000 (10:42 -0800)]
test-release.sh: Update to fetch source from GitHub
Summary:
This also changes the test-release.sh script to build using the monorepo
layout instead of copying sub-projects into llvm/tools or llvm/projects.
Reviewers: jdoerfert, hans
Reviewed By: hans
Subscribers: hans, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D70353
Alex Lorenz [Tue, 19 Nov 2019 19:07:55 +0000 (11:07 -0800)]
[ADT][Expensive checks] Create a std::random_device seed only once when shuffling before sorting
This speeds up the build of compiler-rt with an expensive checks enabled clang by an order of
1 or 2 magnitudes on my machine. I was hoping this would also fix the 'large.test' libFuzzer
timeout on the expensive checks bot on green dragon http://lab.llvm.org:8080/green/job/clang-stage1-cmake-RA-expensive/,
but the fuzzer test still takes too long to compile because of other IR/MIR verification inefficiencies.
Differential Revision: https://reviews.llvm.org/D70288
Adrian Prantl [Tue, 19 Nov 2019 16:53:15 +0000 (08:53 -0800)]
Add a "Using LLDB" section to the welcome page of the website
This is an attempt to feature the user-facing resources more
prominently on the LLDB website by calling out the tutorial and the
GDB command map wight on the start page.
I also moved the "Why a new debugger" section to the "Goals"
subpage. Given that LLDB's first release is almost a decade in the
past now, the title is a bit of an anachronism.
Lastly, I moved the Architecture sub-page from "use" to "resources",
since end-users do not care about the source code layout.
Differential Revision: https://reviews.llvm.org/D70449
David Green [Tue, 19 Nov 2019 18:37:21 +0000 (18:37 +0000)]
[ARM] MVE interleaving load and stores.
Now that we have the intrinsics, we can add VLD2/4 and VST2/4 lowering
for MVE. This works the same way as Neon, recognising the load/shuffles
combination and converting them into intrinsics in a pre-isel pass,
which just calls getMaxSupportedInterleaveFactor, lowerInterleavedLoad
and lowerInterleavedStore.
The main difference to Neon is that we do not have a VLD3 instruction.
Otherwise most of the code works very similarly, with just some minor
differences in the form of the intrinsics to work around. VLD3 is
disabled by making isLegalInterleavedAccessType return false for those
cases.
We may need some other future adjustments, such as VLD4 take up half the
available registers so should maybe cost more. This patch should get the
basics in though.
Differential Revision: https://reviews.llvm.org/D69392
David Green [Tue, 19 Nov 2019 18:17:46 +0000 (18:17 +0000)]
[ARM] Add and update a lot of VLDn tests. NFC
diggerlin [Tue, 19 Nov 2019 18:29:09 +0000 (13:29 -0500)]
implement printing out raw section data of xcoff objectfile for llvm-objdump
SUMMARY:
implement printing out raw section data of xcoff objectfile for llvm-objdump
and option -D --disassemble-all option for llvm-objdump
Reviewers: Sean Fertile
Subscribers: rupprecht, seiyai,hiraditya
Differential Revision: https://reviews.llvm.org/D70255
Alexandre Ganea [Tue, 19 Nov 2019 18:30:14 +0000 (13:30 -0500)]
[C-index] Fix test when using Debug target & MSVC STL
Avoids a deadlock in "clang/test/Index/crash-recovery-modules.m" when building with the MSVC STL & _ITERATOR_DEBUG_LEVEL == 2 (meaning a DEBUG build)
Differential Revision: https://reviews.llvm.org/D69959
Jonas Devlieghere [Tue, 19 Nov 2019 18:23:35 +0000 (10:23 -0800)]
[LLDB] Fix formatting in the driver (NFC)
Dan Liew [Tue, 29 Oct 2019 06:04:27 +0000 (23:04 -0700)]
[Symbolizers] On Darwin compute function offset when possible.
Summary:
The sanitizer symbolizers support printing the function offset
(difference between pc and function start) of a stackframe using the
`%q` format specifier.
Unfortunately this didn't actually work because neither the atos
or dladdr symbolizer set the `AddressInfo::function_offset` field.
This patch teaches both symbolizers to try to compute the function
offset. In the case of the atos symbolizer, atos might not report the
function offset (e.g. it reports a source location instead) so in this
case it fallsback to using `dladdr()` to compute the function offset.
Two test cases are included.
rdar://problem/
56695185
Reviewers: kubamracek, yln
Subscribers: #sanitizers, llvm-commits
Tags: #sanitizers, #llvm
Differential Revision: https://reviews.llvm.org/D69549
Matt Arsenault [Sun, 27 Oct 2019 00:15:10 +0000 (17:15 -0700)]
clang: Add -fconvergent-functions flag
The CUDA builtin library is apparently compiled in C++ mode, so the
assumption of convergent needs to be made in a typically non-SPMD
language. The functions in the library should still be assumed
convergent. Currently they are not, which is potentially incorrect and
this happens to work after the library is linked.
Fangrui Song [Tue, 19 Nov 2019 17:47:17 +0000 (09:47 -0800)]
[ELF] Disallow out-of-range section group indices after D70146
Exposed by invalid/sht-group-wrong-section.test
http://45.33.8.238/win/2613/step_9.txt
Alexey Bataev [Tue, 19 Nov 2019 17:12:23 +0000 (12:12 -0500)]
[OPENMP]Rename function, NFC.
Change the name of the CGOpenMPRuntime::emitOMPIfClause to CGOpenMPRuntime::emitIfClause.
JonChesterfield [Tue, 19 Nov 2019 17:23:46 +0000 (17:23 +0000)]
[nfc][libomptarget] Write amdgcn macros in terms of compiler intrinsics
Tyker [Sun, 17 Nov 2019 10:41:55 +0000 (11:41 +0100)]
[NFC] Refactor representation of materialized temporaries
Summary:
this patch refactor representation of materialized temporaries to prevent an issue raised by rsmith in https://reviews.llvm.org/D63640#inline-612718
Reviewers: rsmith, martong, shafik
Reviewed By: rsmith
Subscribers: thakis, sammccall, ilya-biryukov, rnkovacs, arphaman, cfe-commits
Tags: #clang
Differential Revision: https://reviews.llvm.org/D69360
Yitzhak Mandelbaum [Wed, 13 Nov 2019 14:04:13 +0000 (09:04 -0500)]
[libTooling] Extend `buildASTFromCodeWithArgs` to take files argument.
Summary:
Adds an optional parameter to `buildASTFromCodeWithArgs` that allows the user to
pass additional files that the main code needs to compile. This change makes
`buildASTFromCodeWithArgs` consistent with `runToolOnCodeWithArgs`.
Patch by Alexey Eremin.
Reviewers: gribozavr
Subscribers: cfe-commits
Tags: #clang
Differential Revision: https://reviews.llvm.org/D70175
Joel E. Denny [Tue, 19 Nov 2019 15:54:21 +0000 (10:54 -0500)]
[FileCheck] Use lit's internal shell for the test suite
An advantage is that there are less portability concerns when writing
tests. For example, `-u` is not supported by all implementations of
`env`, but lit's internal shell provides its own `env` that supports
`-u`.
A disadvantage is that some shell constructs, such as parentheses, are
not supported, but FileCheck's test suite currently doesn't require
such constructs.
For comparison, lit configures its test suite in the same manner. See
`llvm/utils/lit/tests/lit.cfg`.
Reviewed By: rnk
Differential Revision: https://reviews.llvm.org/D70278
Fangrui Song [Tue, 19 Nov 2019 05:56:58 +0000 (21:56 -0800)]
[ELF] Improve --gc-sections compatibility with GNU ld regarding section groups
Based on D70020 by serge-sans-paille.
The ELF spec says:
> Furthermore, there may be internal references among these sections that would not make sense if one of the sections were removed or replaced by a duplicate from another object. Therefore, such groups must be included or omitted from the linked object as a unit. A section cannot be a member of more than one group.
GNU ld has 2 behaviors that we don't have:
- Group members (nextInSectionGroup != nullptr) are subject to garbage collection.
This includes non-SHF_ALLOC SHT_NOTE sections.
In particular, discarding non-SHF_ALLOC SHT_NOTE sections is an expected behavior by the Annobin
project. See
https://developers.redhat.com/blog/2018/02/20/annobin-storing-information-binaries/
for more information.
- Groups members are retained or discarded as a unit.
Members may have internal references that are not expressed as
SHF_LINK_ORDER, relocations, etc. It seems that we should be more conservative here:
if a section is marked live, mark all the other member within the
group.
Both behaviors are reasonable. This patch implements them.
A new field InputSectionBase::nextInSectionGroup tracks the next member
within a group. on ELF64, this increases sizeof(InputSectionBase) froms
144 to 152.
InputSectionBase::dependentSections tracks section dependencies, which
is used by both --gc-sections and /DISCARD/. We can't overload it for
the "next member" semantic, because we should allow /DISCARD/ to discard
sections independent of --gc-sections (GNU ld behavior). This behavior
may be reasonably used by `/DISCARD/ : { *(.ARM.exidx*) }` or `/DISCARD/
: { *(.note*) }` (new test `linkerscript/discard-group.s`).
Reviewed By: ruiu
Differential Revision: https://reviews.llvm.org/D70146
AndreyChurbanov [Tue, 19 Nov 2019 16:45:06 +0000 (19:45 +0300)]
Fix openmp on PowerPC64-BE-ELFv2 ABI on FreeBSD.
Patch by adalava (Alfredo Dal'Ava J.nior)
Differential Revision: https://reviews.llvm.org/D67190
Mitchell Balan [Tue, 19 Nov 2019 16:39:13 +0000 (11:39 -0500)]
[NFC] Attempting to fix sphinx build failure with badly encoded characters from incoming patch.
LLVM GN Syncbot [Tue, 19 Nov 2019 16:34:22 +0000 (16:34 +0000)]
gn build: Merge
7fe9435dc88
Matt Arsenault [Tue, 29 Oct 2019 23:16:05 +0000 (16:16 -0700)]
Work on cleaning up denormal mode handling
Cleanup handling of the denormal-fp-math attribute. Consolidate places
checking the allowed names in one place.
This is in preparation for introducing FP type specific variants of
the denormal-fp-mode attribute. AMDGPU will switch to using this in
place of the current hacky use of subtarget features for the denormal
mode.
Introduce a new header for dealing with FP modes. The constrained
intrinsic classes define related enums that should also be moved into
this header for uses in other contexts.
The verifier could use a check to make sure the denorm-fp-mode
attribute is sane, but there currently isn't one.
Currently, DAGCombiner incorrectly asssumes non-IEEE behavior by
default in the one current user. Clang must be taught to start
emitting this attribute by default to avoid regressions when this is
switched to assume ieee behavior if the attribute isn't present.
Pavel Labath [Tue, 12 Nov 2019 14:19:21 +0000 (15:19 +0100)]
[cmake] Disable GCC 9's -Winit-list-lifetime warning in ArrayRef
Summary:
This is a new warning which fires when one stores a reference to the
initializer_list contents in a way which may outlive the
initializer_list which it came from. In llvm this warning is triggered
whenever someone uses the initializer_list ArrayRef constructor.
This is indeed a dangerous thing to do (I myself was bitten by that at
least once), but it is not more dangerous than calling other ArrayRef
constructors with temporary objects -- something which we are used to
and have accepted as a tradeoff for ArrayRef's efficiency.
Currently, this warnings generates so much output that it completely
obscures any actionable warnings, so this patch disables it.
Reviewers: rnk, aaron.ballman
Subscribers: mgorny, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D70122
Thomas Preud'homme [Tue, 19 Nov 2019 11:18:51 +0000 (11:18 +0000)]
[clang][NFC] Make various uses of Regex const
The const-correctness of match() was fixed in rL372764, which allows
uses of Regex objects to be const in cases they couldn't be before. This
patch tightens up the const-ness of Regex in various such cases.
Reviewers: thopre
Reviewed By: thopre
Subscribers: cfe-commits
Tags: #clang
Differential Revision: https://reviews.llvm.org/D68155
Mitchell Balan [Tue, 19 Nov 2019 16:13:27 +0000 (11:13 -0500)]
[NFC] Attempting to fix sphinx build failure with badly encoded characters from incoming patch.
jasonliu [Tue, 19 Nov 2019 15:09:07 +0000 (15:09 +0000)]
[AIX][XCOFF] Write Function descriptors and TOC base to data section
This patch implements writing function descriptors and TOC base into
data section, and also add function descriptors(both csect and label)
and TOC base symbols to the symbol table.
Mitchell Balan [Tue, 19 Nov 2019 15:57:16 +0000 (10:57 -0500)]
[clang-tidy] Give readability-redundant-member-init an option IgnoreBaseInCopyConstructors to avoid breaking code with gcc -Werror=extra
Summary:
readability-redundant-member-init removes redundant / unnecessary member and base class initialization. Unfortunately for the specific case of a copy constructor's initialization of a base class, gcc at strict warning levels warns if "base class is not initialized in the copy constructor of a derived class".
This patch adds an option `IgnoreBaseInCopyConstructors` defaulting to 0 (thus maintaining current behavior by default) to skip the specific case of removal of redundant base class initialization in the copy constructor. Enabling this option enables the resulting code to continue to compile successfully under `gcc -Werror=extra`. New test cases `WithCopyConstructor1` and `WithCopyConstructor2` in clang-tools-extra/test/clang-tidy/readability-redundant-member-init.cpp show that it removes redundant members even from copy constructors.
Reviewers: malcolm.parsons, alexfh, hokein, aaron.ballman, lebedev.ri
Patch by: poelmanc
Subscribers: mgehre, lebedev.ri, cfe-commits
Tags: #clang, #clang-tools-extra
Differential revision: https://reviews.llvm.org/D69145
Sanjay Patel [Tue, 19 Nov 2019 15:47:07 +0000 (10:47 -0500)]
[SLP] fix insertion point for min/max reduction
As discussed in D70148 (and caused a revert of the original commit):
if we insert at the select, then we can produce invalid IR because
the replacement for the compare may have uses before the select.
LLVM GN Syncbot [Tue, 19 Nov 2019 15:33:25 +0000 (15:33 +0000)]
gn build: Merge
765b1250f68
David Bozier [Tue, 19 Nov 2019 13:56:29 +0000 (13:56 +0000)]
Fixup AVR tests to reflect changes in addend format in llvm-objdump
Summary: Changes to llvm-objdump made in D69997
Reviewers: thakis, jhenderson, grimar
Reviewed By: thakis
Subscribers: dylanmckay, Jim, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D70438
Sanjay Patel [Tue, 19 Nov 2019 15:24:32 +0000 (10:24 -0500)]
[SLP] add test for reduction miscompile; NFC
See D70148 for discussion.
Raphael Isemann [Tue, 19 Nov 2019 14:21:15 +0000 (15:21 +0100)]
[lldb] Add logging to IRExecutionUnit::GetStaticInitializers
Simon Tatham [Tue, 19 Nov 2019 14:47:07 +0000 (14:47 +0000)]
[ARM,MVE] Add intrinsics for scalar shifts.
This fills in the small family of MVE intrinsics that have nothing to
do with vectors: they implement bit-shift operations on 32- or 64-bit
values held in one or two general-purpose registers. Most of these
shift operations saturate if shifting left, and round to nearest if
shifting right, although LSLL and ASRL behave like ordinary shifts.
When these instructions take a variable shift count in a register,
they pay attention to its sign, so that (for example) LSLL or UQRSHLL
will shift left if given a positive number but right if given a
negative one. That makes even LSLL and ASRL different enough from
standard LLVM IR shift semantics that I couldn't see any better
alternative than to simply model the whole family as a set of
MVE-specific IR intrinsics.
(The //immediate// forms of LSLL and ASRL, on the other hand, do
behave exactly like a standard IR shift of a 64-bit value. In fact,
those forms don't have ACLE intrinsics defined at all, because you can
just write an ordinary C shift operation if you want one of those.)
The 64-bit shifts have to be instruction-selected in C++, because they
deliver two output values. But the 32-bit ones are simple enough that
I could write a DAG isel pattern directly into each Instruction
record.
Reviewers: ostannard, MarkMurrayARM, dmgreen
Reviewed By: dmgreen
Subscribers: kristof.beyls, hiraditya, cfe-commits, llvm-commits
Tags: #clang, #llvm
Differential Revision: https://reviews.llvm.org/D70319
Sam McCall [Sat, 16 Nov 2019 21:15:05 +0000 (22:15 +0100)]
[clangd] Show values of more expressions on hover
Reviewers: kadircet
Subscribers: ilya-biryukov, MaskRay, jkorous, arphaman, usaxena95, cfe-commits
Tags: #clang
Differential Revision: https://reviews.llvm.org/D70359
Matt Arsenault [Fri, 1 Nov 2019 01:50:30 +0000 (18:50 -0700)]
AMDGPU: Refactor treatment of denormal mode
Start moving towards treating this as a property of the calling
convention, and not the subtarget. The default denormal mode should
not be part of the subtarget, and be moved into a separate function
attribute.
This patch is still NFC. The denormal mode remains as a subtarget
feature for now, but make the necessary changes to switch to using an
attribute.
Matt Arsenault [Fri, 1 Nov 2019 16:44:56 +0000 (09:44 -0700)]
AMDGPU: Be explicit about denormal mode in MIR tests
Start checking the machine function in GlobalISel instead of the
target directly.
This temporarily breaks fcanonicalize selection in GlobalISel.
Balázs Kéri [Tue, 19 Nov 2019 14:22:38 +0000 (15:22 +0100)]
[clang-tidy] DefaultOperatorNewCheck test fixes.