platform/upstream/llvm.git
2 years ago[mips][lld] Add test case to check symbol index reading on mips64el. NFC
Simon Atanasyan [Tue, 11 Jan 2022 16:06:40 +0000 (19:06 +0300)]
[mips][lld] Add test case to check symbol index reading on mips64el. NFC

2 years ago[mips] Use `push_back` to insert element at the end of a container. NFC
Simon Atanasyan [Tue, 11 Jan 2022 05:56:26 +0000 (08:56 +0300)]
[mips] Use `push_back` to insert element at the end of a container. NFC

2 years ago[X86] Tag existing shuffle test case as PR53124
Simon Pilgrim [Tue, 11 Jan 2022 15:11:06 +0000 (15:11 +0000)]
[X86] Tag existing shuffle test case as PR53124

The poor AVX1 codegen matches the core issue raised in PR53124

2 years ago[SCEV] `getSequentialMinMaxExpr()`: look into `umin` when deduplicating operands
Roman Lebedev [Tue, 11 Jan 2022 15:23:02 +0000 (18:23 +0300)]
[SCEV] `getSequentialMinMaxExpr()`: look into `umin` when deduplicating operands

We could just merge all umin into umin_seq, but that is likely
a pessimization, so don't do that, but pretend that we did
for the purpose of deduplication.

2 years ago[NFC][SCEV] More tests with operand-wise redundant operands of umin of umin_seq
Roman Lebedev [Tue, 11 Jan 2022 15:25:53 +0000 (18:25 +0300)]
[NFC][SCEV] More tests with operand-wise redundant operands of umin of umin_seq

2 years ago[compiler-rt] Silence warnings when building with MSVC
Alexandre Ganea [Tue, 11 Jan 2022 15:36:46 +0000 (10:36 -0500)]
[compiler-rt] Silence warnings when building with MSVC

Differential Revision: https://reviews.llvm.org/D116872

2 years ago[libc++] Use TEST_HAS_NO_UNICODE instead of _LIBCPP_HAS_NO_UNICODE in the test suite
Louis Dionne [Mon, 10 Jan 2022 22:14:01 +0000 (17:14 -0500)]
[libc++] Use TEST_HAS_NO_UNICODE instead of _LIBCPP_HAS_NO_UNICODE in the test suite

Differential Revision: https://reviews.llvm.org/D116973

2 years ago[ASan] Driver changes to always link-in asan_static library.
Kirill Stoimenov [Wed, 5 Jan 2022 17:14:40 +0000 (17:14 +0000)]
[ASan] Driver changes to always link-in asan_static library.

This enables the changes from D116182.

Reviewed By: vitalybuka

Differential Revision: https://reviews.llvm.org/D116670

2 years ago[GlobalStatus] Look through non-constexpr casts
Nikita Popov [Tue, 11 Jan 2022 15:00:13 +0000 (16:00 +0100)]
[GlobalStatus] Look through non-constexpr casts

analyzeGlobal() looks through non-constexpr cast instructions when
looking for users. However, this particular place only strips the
casts again if they are constexprs. We should be looking through all
casts here.

2 years ago[NFC][SCEV] Add more tests for umin_seq with redundant operands
Roman Lebedev [Tue, 11 Jan 2022 14:33:15 +0000 (17:33 +0300)]
[NFC][SCEV] Add more tests for umin_seq with redundant operands

2 years ago[gn build] (manually) port 8503c688d555
Nico Weber [Tue, 11 Jan 2022 14:47:28 +0000 (09:47 -0500)]
[gn build] (manually) port 8503c688d555

2 years ago[GlobalOpt] Regenerate test checks (NFC)
Nikita Popov [Tue, 11 Jan 2022 14:34:16 +0000 (15:34 +0100)]
[GlobalOpt] Regenerate test checks (NFC)

2 years ago[clang][lex] Keep references to `DirectoryLookup` objects up-to-date
Jan Svoboda [Tue, 11 Jan 2022 13:47:03 +0000 (14:47 +0100)]
[clang][lex] Keep references to `DirectoryLookup` objects up-to-date

The elements of `SearchPath::SearchDirs` are being referenced to by their indices. This proved to be error-prone: `HeaderSearch::SearchDirToHSEntry` was accidentally not being updated in `HeaderSearch::AddSearchPath()`. This patch fixes that by referencing `SearchPath::SearchDirs` elements by their address instead, which is stable thanks to the bump-ptr-allocation strategy.

Reviewed By: ahoppen

Differential Revision: https://reviews.llvm.org/D116750

2 years ago[mlir][linalg] Use cast instead of dyn_cast that's always dereferenced
Benjamin Kramer [Tue, 11 Jan 2022 14:05:45 +0000 (15:05 +0100)]
[mlir][linalg] Use cast instead of dyn_cast that's always dereferenced

This turns a random nullptr deref into an assertion failure in case
`tensor::registerInferTypeOpInterfaceExternalModels` isn't called.

2 years ago[SCEV] `getSequentialMinMaxExpr()`: keep only the first instance of an operand
Roman Lebedev [Tue, 11 Jan 2022 13:45:22 +0000 (16:45 +0300)]
[SCEV] `getSequentialMinMaxExpr()`: keep only the first instance of an operand

Having the same operand more than once doesn't change the outcome here,
neither reduction-wise nor poison-wise.
We must keep the first instance specifically though.

2 years ago[SCEV] Add test for umin_seq with duplicate operands
Roman Lebedev [Tue, 11 Jan 2022 13:24:29 +0000 (16:24 +0300)]
[SCEV] Add test for umin_seq with duplicate operands

2 years ago[SPIR-V] Remove unused variable
Anastasia Stulova [Tue, 11 Jan 2022 13:45:33 +0000 (13:45 +0000)]
[SPIR-V] Remove unused variable

2 years ago[InstSimplify] Add additional GEP tests with undef bases.
Florian Hahn [Tue, 11 Jan 2022 13:43:53 +0000 (13:43 +0000)]
[InstSimplify] Add additional GEP tests with undef bases.

2 years ago[lldb] Remove non address bits from memory read arguments
David Spickett [Wed, 3 Nov 2021 12:10:40 +0000 (12:10 +0000)]
[lldb] Remove non address bits from memory read arguments

Addresses on AArch64 can have top byte tags, memory tags and pointer
authentication signatures in the upper bits.

While testing memory tagging I found that memory read couldn't
read a range if the two addresses had different tags. The same
could apply to signed pointers given the right circumstance.

(lldb) memory read mte_buf_alt_tag mte_buf+16
error: end address (0x900fffff7ff8010) must be greater than the start
address (0xa00fffff7ff8000).

Or it would try to read a lot more memory than expected.

(lldb) memory read mte_buf mte_buf_alt_tag+16
error: Normally, 'memory read' will not read over 1024 bytes of data.
error: Please use --force to override this restriction just once.
error: or set target.max-memory-read-size if you will often need a
larger limit.

Fix this by removing non address bits before we calculate the read
range. A test is added for AArch64 Linux that confirms this by using
the top byte ignore feature.

This means that if you do read with a tagged pointer the output
does not include those tags. This is potentially confusing but I think
overall it's better that we don't pretend that we're reading memory
from a range that the process is unable to map.

(lldb) p ptr1
(char *) $4 = 0x3400fffffffff140 "\x80\xf1\xff\xff\xff\xff"
(lldb) p ptr2
(char *) $5 = 0x5600fffffffff140 "\x80\xf1\xff\xff\xff\xff"
(lldb) memory read ptr1 ptr2+16
0xfffffffff140: 80 f1 ff ff ff ff 00 00 38 70 bc f7 ff ff 00 00  ........8p......

Reviewed By: omjavaid, danielkiss

Differential Revision: https://reviews.llvm.org/D103626

2 years ago[SPIR-V] Add linking using spirv-link.
Anastasia Stulova [Tue, 11 Jan 2022 11:21:30 +0000 (11:21 +0000)]
[SPIR-V] Add linking using spirv-link.

Add support of linking files compiled into SPIR-V objects
using spirv-link.

Command line inteface examples:

clang --target=spirv64 test1.cl test2.cl

clang  --target=spirv64 test1.cl -o test1.o
clang  --target=spirv64 test1.o test2.cl -o test_app.out

This works independently from the SPIR-V generation method
(via an external tool or an internal backend) and applies
to either approach that is being used.

Differential Revision: https://reviews.llvm.org/D116266

2 years ago[lldb/qemu] Implement GetMmapArgumentList
Pavel Labath [Mon, 10 Jan 2022 15:34:39 +0000 (16:34 +0100)]
[lldb/qemu] Implement GetMmapArgumentList

By forwarding it to the host platform.

2 years ago[SCEV] Reenable umin_seq support and fix the `computeSCEVAtScope()`
Roman Lebedev [Tue, 11 Jan 2022 12:51:43 +0000 (15:51 +0300)]
[SCEV] Reenable umin_seq support and fix the `computeSCEVAtScope()`

This reverts commit f62f47f5e1f641b41d3b7d593c058ebec2883534.

2 years ago[NFC][SCEV] Add reproducers for umin_seq crashes
Roman Lebedev [Tue, 11 Jan 2022 12:33:53 +0000 (15:33 +0300)]
[NFC][SCEV] Add reproducers for umin_seq crashes

As reported in https://reviews.llvm.org/D116766#3233042

2 years ago[LSR] Use pointer args instead of undef for uglygep*.ll tests.
Florian Hahn [Tue, 11 Jan 2022 12:35:55 +0000 (12:35 +0000)]
[LSR] Use pointer args instead of undef for uglygep*.ll tests.

Make the test more robust by replacing undef by pointer arguments. This
ensures that the GEPs cannot be folded to undef.

2 years agoRevert "[Clang][AArch64][ARM] PMUv3.4 Option Added"
David Green [Tue, 11 Jan 2022 12:33:53 +0000 (12:33 +0000)]
Revert "[Clang][AArch64][ARM] PMUv3.4 Option Added"

It turns out this is conflating a few different PMU extensions. And on
Arm ended up breaking M-Profile code generation. Reverting for the
moment whilst we sort out the details.

This reverts commit d17fb46e894501568a1bf3b11a5d920817444630.

2 years ago[Clang][Sema] Fix attribute mismatch warning for ObjC class properties
Egor Zhdan [Wed, 29 Dec 2021 22:02:21 +0000 (22:02 +0000)]
[Clang][Sema] Fix attribute mismatch warning for ObjC class properties

If a class declares an instance property, and an inheritor class declares a class property with the same name, Clang Sema currently treats the latter as an overridden property, and compares the attributes of the two properties to check for a mismatch. The resulting diagnostics might be misleading, since neither of the properties actually overrides the another one.

rdar://86018435

Differential Revision: https://reviews.llvm.org/D116412

2 years ago[CodeGen] Avoid deprecated Address constructor
Nikita Popov [Tue, 11 Jan 2022 12:01:28 +0000 (13:01 +0100)]
[CodeGen] Avoid deprecated Address constructor

2 years ago[SVE][CodeGen] Use splice instruction when lowering VECTOR_SPLICE
David Sherwood [Thu, 16 Dec 2021 08:57:18 +0000 (08:57 +0000)]
[SVE][CodeGen] Use splice instruction when lowering VECTOR_SPLICE

For certain negative indices passed to the VECTOR_SPLICE operation
we can actually directly use the SVE splice instruction by creating
the appropriate predicate. The predicate needs to be constructed in
such a way that all but the last -idx elements are false. We can do
this efficiently using a combination of 'ptrue' (with the appropriate
fixed pattern, e.g. vl1, vl2, etc.) and 'rev'. The advantage of using
these instructions to generate the predicate is they do not set any
flags, unlike the whilelo instruction. This is critical when the splice
operation is in a loop, since we want MachineLICM to hoist the
predicate generation out of the loop.

Differential Revision: https://reviews.llvm.org/D115863

2 years agoARM: make FastISel & GISel pass -1 to ADJCALLSTACKUP to signal no callee pop.
Tim Northover [Fri, 7 Jan 2022 13:49:18 +0000 (13:49 +0000)]
ARM: make FastISel & GISel pass -1 to ADJCALLSTACKUP to signal no callee pop.

The interface for these instructions changed with support for mandatory tail
calls, and now -1 indicates the CalleePopAmount argument is not valid.
Unfortunately I didn't realise FastISel or GISel did calls at the time so
didn't update them.

2 years ago[SemaTemplateInstantiate] Use cast<> instead of dyn_cast<> to avoid dereference of...
Simon Pilgrim [Tue, 11 Jan 2022 11:12:12 +0000 (11:12 +0000)]
[SemaTemplateInstantiate] Use cast<> instead of dyn_cast<> to avoid dereference of nullptr

The pointer is always dereferenced immediately below, so assert the cast is correct instead of returning nullptr

2 years ago[MemoryBuiltins] Remove unused isOpNewLikeFn() (NFC)
Nikita Popov [Tue, 11 Jan 2022 11:27:23 +0000 (12:27 +0100)]
[MemoryBuiltins] Remove unused isOpNewLikeFn() (NFC)

This function is no longer used since
2cafbcb560d9e6e2300941d088e754b01d56595b.

2 years ago[MemoryBuiltins] Remove unused isStrdupLikeFn() function (NFC)
Nikita Popov [Tue, 11 Jan 2022 11:25:39 +0000 (12:25 +0100)]
[MemoryBuiltins] Remove unused isStrdupLikeFn() function (NFC)

This function is no longer used after
dcbc91f40c2e6ff578667020f7c6a05c25149865.

2 years ago[DSE] Check for noalias calls rather than alloc functions
Nikita Popov [Tue, 11 Jan 2022 11:08:44 +0000 (12:08 +0100)]
[DSE] Check for noalias calls rather than alloc functions

For these "visible on unwind/ret" checks we only care about the
fact that no other code has access to the pointer (unless it
escapes). A noalias call is sufficient for this, it does not
have to be a known allocation function.

This is basically the same change as D116728, but for DSE rather
than LICM.

2 years ago[LSR] Remove duplicated test address-space-loop.ll.
Florian Hahn [Tue, 11 Jan 2022 11:18:28 +0000 (11:18 +0000)]
[LSR] Remove duplicated test address-space-loop.ll.

llvm/test/Transforms/LoopStrengthReduce/uglygep-address-space.ll has
exactly the same checks and input. Remove the duplicated test.

2 years ago[mlir][linalg][bufferize] Fix CallOp bufferization
Matthias Springer [Tue, 11 Jan 2022 11:04:55 +0000 (20:04 +0900)]
[mlir][linalg][bufferize] Fix CallOp bufferization

Previously, CallOps did not have any aliasing OpResult/OpOperand pairs. Therefore, CallOps were mostly ignored by the analysis and buffer copies were not inserted when necessary.

This commit introduces the following changes:
* Function bbArgs writable by default. A function can now be bufferized without inspecting its callers.
* Callers must introduce buffer copies of function arguments when necessary. If a function is external, the caller must conservatively assume that a function argument is modified by the callee after bufferization. If the function is not external, the caller inspects the callee to determine if a function argument is modified.

Differential Revision: https://reviews.llvm.org/D116457

2 years ago[DSE] Add additional tests for noalias calls (NFC)
Nikita Popov [Tue, 11 Jan 2022 10:55:59 +0000 (11:55 +0100)]
[DSE] Add additional tests for noalias calls (NFC)

Currently this is special-cased to TLI alloc functions only.

2 years agoReland "[AST] Add RParen loc for decltype AutoTypeloc."
Haojian Wu [Mon, 10 Jan 2022 14:19:05 +0000 (15:19 +0100)]
Reland "[AST] Add RParen loc for decltype AutoTypeloc."

Reland 55d96ac and 37ec65e with a clang-tidy fix.

2 years ago[ADT] Add an in-place version of toHex()
Hans Wennborg [Mon, 10 Jan 2022 18:45:13 +0000 (19:45 +0100)]
[ADT] Add an in-place version of toHex()

and use that to simplify MD5's hex string code which was previously
using a string stream, as well as Clang's
CGDebugInfo::computeChecksum().

Differential revision: https://reviews.llvm.org/D116960

2 years ago[DSE] Make test more robust (NFC)
Nikita Popov [Tue, 11 Jan 2022 10:49:08 +0000 (11:49 +0100)]
[DSE] Make test more robust (NFC)

If the allocation is not captured, then all the stores before the
ret are dead anyway.

2 years ago[ADT] Use a lookup table in hexdigit() and call that from toHex()
Hans Wennborg [Tue, 11 Jan 2022 09:08:25 +0000 (10:08 +0100)]
[ADT] Use a lookup table in hexdigit() and call that from toHex()

A lookup table, which toHex() was using, seems like the better approach.
Having two implementations is redundant, so put the lookup table in
hexdigit() and make toHex() call that.

Differential revision: https://reviews.llvm.org/D116960

2 years ago[SemaOverload] compareConversionFunctions - use castAs<> instead of getAs<> to avoid...
Simon Pilgrim [Mon, 10 Jan 2022 17:44:18 +0000 (17:44 +0000)]
[SemaOverload] compareConversionFunctions - use castAs<> instead of getAs<> to avoid dereference of nullptr

The pointer is dereferenced immediately below, so assert the cast is correct instead of returning nullptr

2 years ago[SemaOverload] Use castAs<> instead of getAs<> to avoid dereference of nullptr
Simon Pilgrim [Mon, 10 Jan 2022 17:31:35 +0000 (17:31 +0000)]
[SemaOverload] Use castAs<> instead of getAs<> to avoid dereference of nullptr

The pointer is always dereferenced inside BuildSimilarlyQualifiedPointerType, so assert the cast is correct instead of returning nullptr

2 years ago[clang] [test] Fix clang-cl unused argument tests on paths that start with /U
Martin Storsjö [Tue, 11 Jan 2022 10:22:42 +0000 (12:22 +0200)]
[clang] [test] Fix clang-cl unused argument tests on paths that start with /U

This reinstates a test that was temporarily removed in
e26bbae30218a35d76a79fe90b0e41dd0f71b779, in a form that works on
Darwin.

Use -LD instead of -link as a linker argument that is unused when
compiling, that produces warnings normally. -LD can be placed anywhere
in the command line, so that the command line ends with "-- %s", making
paths starting with /U correctly interpreted as paths, not options.

2 years ago[LICM] Regenerate test checks (NFC)
Nikita Popov [Tue, 11 Jan 2022 09:55:50 +0000 (10:55 +0100)]
[LICM] Regenerate test checks (NFC)

2 years ago[MLIR] Update allocs to memref.allocs in documentation.
Julian Gross [Tue, 11 Jan 2022 10:04:15 +0000 (11:04 +0100)]
[MLIR] Update allocs to memref.allocs in documentation.

Changed the remaining appearances of alloc to memref.alloc in several
documentation sections, since they lead to misunderstandings, if they
are used.

Differential Revision: https://reviews.llvm.org/D116999

2 years ago[RISCV] Generate 32 bits jumptable entries when code model is small
wangpc [Tue, 11 Jan 2022 10:19:05 +0000 (18:19 +0800)]
[RISCV] Generate 32 bits jumptable entries when code model is small

The code can only address the whole RV32 address space or the lower 2 GiB
of the RV64 address space in small code model, so 32 bits entry is enough.
Cache hit ratio and code size have some improvements.

Reviewed By: asb

Differential Revision: https://reviews.llvm.org/D116435

2 years ago[flang] Add tests for converting arrays and refs to arrays. NFC
Diana Picus [Wed, 22 Dec 2021 09:06:37 +0000 (09:06 +0000)]
[flang] Add tests for converting arrays and refs to arrays. NFC

Cover more of the code paths from LLVMTypeConverter::convertPointerLike
and LLVMTypeConverter::convertSequenceType.

Differential Revision: https://reviews.llvm.org/D116927

2 years ago[clangd] Save more getFileID in Selection
Sam McCall [Tue, 11 Jan 2022 10:01:02 +0000 (11:01 +0100)]
[clangd] Save more getFileID in Selection

This saves about 10% of SelectionVisitor::pop().

2 years ago[SCEVExpander] Use IntToPtr for temporary instruction.
Florian Hahn [Tue, 11 Jan 2022 09:40:21 +0000 (09:40 +0000)]
[SCEVExpander] Use IntToPtr for temporary instruction.

Use PtrToInt instead Add when creating temporary instructions. The add
might get folded away with more sophisticated folding.

2 years ago[IR] Change vector.splice intrinsic to reject out-of-bounds indices
David Sherwood [Fri, 17 Dec 2021 09:39:21 +0000 (09:39 +0000)]
[IR] Change vector.splice intrinsic to reject out-of-bounds indices

I've changed the definition of the experimental.vector.splice
instrinsic to reject indices that are known to be or possibly
out-of-bounds. In practice, this means changing the definition so that
the index is now only valid in the range [-VL, VL-1] where VL is the
known minimum vector length. We use the vscale_range attribute to
take the minimum vscale value into account so that we can permit
more indices when the attribute is present.

The splice intrinsic is currently only ever generated by the vectoriser,
which will never attempt to splice vectors with out-of-bounds values.
Changing the definition also makes things simpler for codegen since we
can always assume that the index is valid.

This patch was created in response to review comments on D115863

Differential Revision: https://reviews.llvm.org/D115933

2 years ago[clangd] Small optimization in SelectionTree
Sam McCall [Tue, 11 Jan 2022 09:15:12 +0000 (10:15 +0100)]
[clangd] Small optimization in SelectionTree

This seems to be strictly faster in all cases. Before fixing D116978 it
was one of the hot paths, and may become one again.

2 years ago[clang] [test] Remove newly added tests that fail on Darwin
Martin Storsjö [Tue, 11 Jan 2022 09:06:29 +0000 (11:06 +0200)]
[clang] [test] Remove newly added tests that fail on Darwin

These tests were recently added in
50ec1306d060e46e0d53c9f5d8a052e1b0d10d3b. The clang-cl invocations
interpret the source path, %s, which begins with /Users, as a cl
option /U. Normally this is worked around by passing -- before
the arguments that must be interpreted as files, not options, but
in the current test, the -link option must be passed last, after any
file names.

2 years ago[LoopVersioning] Check the full generated runtime checks in lcssa.ll.
Florian Hahn [Tue, 11 Jan 2022 08:37:29 +0000 (08:37 +0000)]
[LoopVersioning] Check the full generated runtime checks in lcssa.ll.

This makes improvements in follow-up patches more obvious.

2 years ago[mlir][GPU] Fix attribute name of DL specification
Diego Caballero [Tue, 11 Jan 2022 08:30:33 +0000 (08:30 +0000)]
[mlir][GPU] Fix attribute name of DL specification

D115722 added a DL spec to GPU modules. It happens that the DL
default interface implementation is sensitive to the name of the
DL spec attribute. This patch is fixing the name of the attribute
to be the expected one.

Reviewed By: ftynse

Differential Revision: https://reviews.llvm.org/D116956

2 years ago[CodeGen] Make element type in emitArrayDestroy() predictable
Nikita Popov [Tue, 4 Jan 2022 14:24:14 +0000 (15:24 +0100)]
[CodeGen] Make element type in emitArrayDestroy() predictable

When calling emitArrayDestroy(), the pointer will usually have
ConvertTypeForMem(EltType) as the element type, as one would expect.
However, globals with initializers sometimes don't use the same
types as values normally would, e.g. here the global uses
{ double, i32 } rather than %struct.T as element type.

Add an early cast to the global destruction path to avoid this
special case. The cast would happen lateron anyway, it only gets
moved to an earlier point.

Differential Revision: https://reviews.llvm.org/D116219

2 years ago[CSKY] Add CSKYConstantIslands Pass to lift or duplicate constant pool entry
Zi Xuan Wu [Tue, 11 Jan 2022 08:13:21 +0000 (16:13 +0800)]
[CSKY] Add CSKYConstantIslands Pass to lift or duplicate constant pool entry

Loading constants inline is expensive on CSKY and it's in general better
to place the constant nearby in code space and then it can be loaded with a
simple 16/32 bit load instruction like lrw.

It needs lift or duplicates constant pool entry to make constant nearby so that lrw instruction can reach.

2 years ago[RISCV] Override TargetLowering::BuildSDIVPow2 to generate SELECT
wangpc [Tue, 11 Jan 2022 07:52:00 +0000 (15:52 +0800)]
[RISCV] Override TargetLowering::BuildSDIVPow2 to generate SELECT

When `Zbt` is enabled, we can generate SELECT for division by power
of 2, so that there is no data dependency.

Reviewed By: craig.topper

Differential Revision: https://reviews.llvm.org/D114856

2 years ago[clang] Add --start-no-unused-arguments/--end-no-unused-arguments to silence some...
Martin Storsjö [Sat, 1 Jan 2022 13:54:29 +0000 (15:54 +0200)]
[clang] Add --start-no-unused-arguments/--end-no-unused-arguments to silence some unused argument warnings

When passing a set of flags to configure defaults for a specific
target (similar to the cmake settings `CLANG_DEFAULT_RTLIB`,
`CLANG_DEFAULT_UNWINDLIB`, `CLANG_DEFAULT_CXX_STDLIB` and
`CLANG_DEFAULT_LINKER`, but without hardcoding them in the binary),
some of the flags may cause warnings (e.g. `-stdlib=` when compiling C
code). Allow requesting selectively ignoring unused arguments among
some of the arguments on the command line, without needing to resort
to `-Qunused-arguments` or `-Wno-unused-command-line-argument`.

Fix up the existing diagnostics.c testcase. It was added in
response to PR12181 to fix handling of
`-Werror=unused-command-line-argument`, but the command line option
in the test (`-fzyzzybalubah`) now triggers "error: unknown argument"
instead of the intended warning. Change it into a linker input
(`-lfoo`) which triggers the intended diagnostic. Extend the
existing test case to check more cases and make sure that it keeps
testing the intended case.

Add testing of the new option to this existing test.

Differential Revision: https://reviews.llvm.org/D116503

2 years ago[gn build] Port 8d23b7420c92
LLVM GN Syncbot [Tue, 11 Jan 2022 06:50:34 +0000 (06:50 +0000)]
[gn build] Port 8d23b7420c92

2 years ago[libc++][ranges] Implement `uninitialized_copy{,_n}` and `uninitialized_move{,_n}`.
Konstantin Varlamov [Tue, 11 Jan 2022 06:49:37 +0000 (22:49 -0800)]
[libc++][ranges] Implement `uninitialized_copy{,_n}` and `uninitialized_move{,_n}`.

Also implement `in_out_result` which is a prerequisite.

Differential Revision: https://reviews.llvm.org/D116023

2 years ago[NFC][XCOFF][llvm-readobj] replace binaries with YAMLs (only tests for Symbols).
Esme-Yi [Tue, 11 Jan 2022 06:17:16 +0000 (06:17 +0000)]
[NFC][XCOFF][llvm-readobj] replace binaries with YAMLs (only tests for Symbols).

Summary: Since yaml2obj now supports converting AuxSymbols, we can use YAMLs as input for symbol-related tests of llvm-readobj instead of binaries.

Reviewed By: jhenderson, shchenz

Differential Revision: https://reviews.llvm.org/D114434

2 years agoallow llvm-reduce, if asked, to run its set of passes more than once, taking longer...
John Regehr [Tue, 11 Jan 2022 05:24:23 +0000 (22:24 -0700)]
allow llvm-reduce, if asked, to run its set of passes more than once, taking longer to finish but also potentially resulting in a smaller reduced file.

2 years ago[ELF] Move OffsetGetter before some static functions. NFC
Fangrui Song [Tue, 11 Jan 2022 04:16:02 +0000 (20:16 -0800)]
[ELF] Move OffsetGetter before some static functions. NFC

to prepare for D116881.

2 years ago[lld-macho] Change some global pointers to unique_ptr
Fangrui Song [Tue, 11 Jan 2022 03:39:14 +0000 (19:39 -0800)]
[lld-macho] Change some global pointers to unique_ptr

Similar to D116143. My x86-64 `lld` is ~8KiB smaller.

Reviewed By: keith

Differential Revision: https://reviews.llvm.org/D116902

2 years ago[lld-macho][test] Add missing coverage for archive/dylib resolution after D115092
Fangrui Song [Tue, 11 Jan 2022 03:36:24 +0000 (19:36 -0800)]
[lld-macho][test] Add missing coverage for archive/dylib resolution after D115092

When `file->fetch(sym)` is replaced with a no-op, no test fails.

The new test catches the case.

Reviewed By: #lld-macho, oontvoo

Differential Revision: https://reviews.llvm.org/D116916

2 years ago[RISCV] Block vmsgeu.vi with 0 immediate in Isel
Chenbing.Zheng [Tue, 11 Jan 2022 03:00:09 +0000 (03:00 +0000)]
[RISCV] Block vmsgeu.vi with 0 immediate in Isel

For vmsgeu.vi with 0, we know this is always true. So we can replace
it with vmset.m (unmasked) or vmset.m+vmand.mm (masked).

Reviewed By: craig.topper

Differential Revision: https://reviews.llvm.org/D116584

2 years ago[RISCV] Select vl op to X0 when it is equal to ~0.
jacquesguan [Fri, 7 Jan 2022 08:35:23 +0000 (16:35 +0800)]
[RISCV] Select vl op to X0 when it is equal to ~0.

Now the backend will select ~0 vl to a register and load instruction, we could use X0 to replace it.

Differential Revision: https://reviews.llvm.org/D116798

2 years ago[RISCV] Add precommit test for select vl op that equal to ~0.
jacquesguan [Fri, 7 Jan 2022 08:53:02 +0000 (16:53 +0800)]
[RISCV] Add precommit test for select vl op that equal to ~0.

Precommit test for D116798

Differential Revision: https://reviews.llvm.org/D116799

2 years ago[M68k][test][NFC] Reorder the tests for MxBTST_MI and MxBTST_MR
Jim Lin [Tue, 11 Jan 2022 02:16:01 +0000 (10:16 +0800)]
[M68k][test][NFC] Reorder the tests for MxBTST_MI and MxBTST_MR

Follow the order of definitions in M68kInstrBits.td

2 years ago[RISCV] Use shift for zero extension when Zbb and Zbp are not enabled
Haocong.Lu [Tue, 11 Jan 2022 02:32:25 +0000 (02:32 +0000)]
[RISCV] Use shift for zero extension when Zbb and Zbp are not enabled

Now AND is used for zero extension when both Zbb and Zbp are not enabled.
It may be better to use shift operation if the trailing ones mask exceeds simm12.

This patch optimzes LUI+ADDI+AND to SLLI+SRLI.

Reviewed By: craig.topper

Differential Revision: https://reviews.llvm.org/D116720

2 years agogit-clang-format HEAD~
Nick Desaulniers [Tue, 11 Jan 2022 02:34:30 +0000 (18:34 -0800)]
git-clang-format HEAD~

2 years ago[TargetLowering] precommit refactor from D115688 NFC
Nick Desaulniers [Tue, 11 Jan 2022 02:31:29 +0000 (18:31 -0800)]
[TargetLowering] precommit refactor from D115688 NFC

Signed-off-by: Nick Desaulniers <ndesaulniers@google.com>
2 years ago[RISCV] Use vmv.s.x to build one element splat vector.
jacquesguan [Sat, 25 Dec 2021 10:09:27 +0000 (18:09 +0800)]
[RISCV] Use vmv.s.x to build one element splat vector.

When we want to create an splat vector that only the first element is initialized, we could use vmv.s.x or vfmv.s.f to build it.

Differential Revision: https://reviews.llvm.org/D116277

2 years agoPartial revert of 82fb4f4
Philip Reames [Tue, 11 Jan 2022 02:17:34 +0000 (18:17 -0800)]
Partial revert of 82fb4f4

Two crashes have been reported.  This change disables the new logic while leaving the new node in tree.  Hopefully, that's enough to allow investigation without breakage while avoiding massive churn.

2 years agoDelete a stale comment
Philip Reames [Tue, 11 Jan 2022 00:15:39 +0000 (16:15 -0800)]
Delete a stale comment

2 years ago[AST] lookup in parent DeclContext for transparent DeclContext
Chuanqi Xu [Tue, 11 Jan 2022 02:15:46 +0000 (10:15 +0800)]
[AST] lookup in parent DeclContext for transparent DeclContext

The compiler would crash if we lookup for name in transparent decl
context. See the tests attached for example.

I think this should make sense since the member declared in transparent
DeclContext are semantically defined in the enclosing (non-transparent)
DeclContext, this is the definition for transparent DeclContext.

Reviewed By: erichkeane

Differential Revision: https://reviews.llvm.org/D116792

2 years ago[SelectionDAG] Add additional documentation to FP_TO_S/UINT_SAT to justify D116870...
Craig Topper [Tue, 11 Jan 2022 02:07:12 +0000 (18:07 -0800)]
[SelectionDAG] Add additional documentation to FP_TO_S/UINT_SAT to justify D116870. NFC

As noted in post-commit review, the value of the bits between
the saturating type and result type were not defined defined. Define
them to be sign extended for FP_TO_SINT_SAT and zero extended for
FP_TO_UINT_SAT.

2 years ago[AST] Don't consider 'ExportDecl' when calculating DeclContext 'Encloses'
Chuanqi Xu [Tue, 11 Jan 2022 01:54:57 +0000 (09:54 +0800)]
[AST] Don't consider 'ExportDecl' when calculating DeclContext 'Encloses'

This mimics the style of 90010c2e1 (Don't consider 'LinkageSpec' when
calculating DeclContext 'Encloses'). Since ExportDecl and LinkageSpec
are transparent DeclContext, they share some similarity.

Reviewed By: erichkeane

Differential Revision: https://reviews.llvm.org/D116911

2 years agoAdd a `qualified` directive to the Op, Attribute, and Type declarative assembly format
Mehdi Amini [Tue, 11 Jan 2022 01:26:44 +0000 (01:26 +0000)]
Add a `qualified` directive to the Op, Attribute, and Type declarative assembly format

This patch introduces a new directive that allow to parse/print attributes and types fully
qualified.
This is a follow-up to ee0908703d29 which introduces the eliding of the `!dialect.mnemonic` by default and allows to force to fully qualify each type/attribute
individually.

Differential Revision: https://reviews.llvm.org/D116905

2 years ago[clang][OpenMP5.1] Initial parsing/sema for 'indirect' clause
Jennifer Yu [Tue, 11 Jan 2022 00:20:02 +0000 (16:20 -0800)]
[clang][OpenMP5.1] Initial parsing/sema for 'indirect' clause

Differential Revision: https://reviews.llvm.org/D116764

2 years ago[Tooling] When transferring compile commands between files, always use '--'
Sam McCall [Thu, 6 Jan 2022 04:00:11 +0000 (05:00 +0100)]
[Tooling] When transferring compile commands between files, always use '--'

"driver <flags> -- <input>" is a particularly convenient form of the
compile command to manipulate, with fewer special cases to handle.

Guaranteeing that the output command is of that form is cheap and makes
it easier to consume the result in some cases.

Differential Revision: https://reviews.llvm.org/D116721

2 years ago[Diagnostics] Don't drop a statically set NoWarningAsError flag during option processing
Wolfgang Pieb [Fri, 17 Sep 2021 00:22:31 +0000 (17:22 -0700)]
[Diagnostics] Don't drop a statically set NoWarningAsError flag during option processing

When a -W<diag> option is given on the command line, and the corresponding diagnostic has
the NoWarnOnError flag set, prevent the flag from being dropped when the severity is reevaluated.
This fixes PR51837.

Reviewed By: dexonsmith

Differential Revision: https://reviews.llvm.org/D109981

2 years ago[libcxx][test] Make LIBCPP_STATIC_ASSERT usable at namespace scope
Casey Carter [Tue, 11 Jan 2022 00:35:31 +0000 (16:35 -0800)]
[libcxx][test] Make LIBCPP_STATIC_ASSERT usable at namespace scope

... even when `!defined(_LIBCPP_VERSION)`. (Note that the previous definition for this case - `((void)0);` - is ill-formed at namespace scope.) Ditto for `LIBCPP_ASSERT`, `LIBCPP_ASSERT_NOEXCEPT`, `LIBCPP_ASSERT_NOT_NOEXCEPT`, and `LIBCPP_ONLY`.

Differential Revision: https://reviews.llvm.org/D116880

2 years agoEmit the C++ dialect in -gmodules .pcm files.
Adrian Prantl [Mon, 10 Jan 2022 22:55:03 +0000 (14:55 -0800)]
Emit the C++ dialect in -gmodules .pcm files.

Because of commit: https://reviews.llvm.org/D104291 the -gmodules .pcm
files do not have the same DW_AT_language dialect as the .o file. This
was a simple matter of passing the DebugStrictDwarf flag to the
PCHContainerGenerator object's CodeGenOpts from the CompilerInstance
passed in to it.

Before this change if you ran dwarfdump on the gmodule cache folder
you would get DW_AT_language (DW_LANG_C_plus_plus) even when using
-std=c++14 with clang

Patch by Shubham Rastogi!

Differential Revision: https://reviews.llvm.org/D116790

2 years agodocs: update some bug tracker references (NFC)
Keith Smiley [Sun, 9 Jan 2022 02:36:49 +0000 (18:36 -0800)]
docs: update some bug tracker references (NFC)

Fixes https://github.com/llvm/llvm-project/issues/53091

Differential Revision: https://reviews.llvm.org/D116898

2 years ago[CodeCompletion] Complete designators for fields in anonymous structs/unions
Sam McCall [Thu, 6 Jan 2022 02:34:41 +0000 (03:34 +0100)]
[CodeCompletion] Complete designators for fields in anonymous structs/unions

Fixes https://github.com/clangd/clangd/issues/836

Differential Revision: https://reviews.llvm.org/D116717

2 years agoAdd coverage of GlobalsModRef's indirect global case
Philip Reames [Mon, 10 Jan 2022 23:54:18 +0000 (15:54 -0800)]
Add coverage of GlobalsModRef's indirect global case

2 years ago[libc] Add linux aarch64 syscall implementation.
Siva Chandra [Mon, 10 Jan 2022 07:02:41 +0000 (23:02 -0800)]
[libc] Add linux aarch64 syscall implementation.

Add mmap and munmap to the linux aarch64 entrypoint list as the first
user of these syscalls.

Reviewed By: michaelrj

Differential Revision: https://reviews.llvm.org/D116949

2 years ago[NFC][regalloc] Pull out some AllocationOrder/CostPerUseLimit eviction logic
Mircea Trofin [Tue, 21 Dec 2021 04:03:40 +0000 (20:03 -0800)]
[NFC][regalloc] Pull out some AllocationOrder/CostPerUseLimit eviction logic

We are reusing that logic in the ML implementation.

Differential Revision: https://reviews.llvm.org/D116075

2 years ago[MemoryBuiltin] Add an API for checking if an unused allocation can be removed [NFC]
Philip Reames [Mon, 10 Jan 2022 17:50:01 +0000 (09:50 -0800)]
[MemoryBuiltin] Add an API for checking if an unused allocation can be removed [NFC]

Not all allocation functions are removable if unused.  An example of a non-removable allocation would be a direct call to the replaceable global allocation function in C++.  An example of a removable one - at least according to historical practice - would be malloc.

2 years ago[LowerMatrixIntrinsics] Call getRegisterClassForType before getNumberOfRegisters.
Craig Topper [Mon, 10 Jan 2022 22:42:23 +0000 (14:42 -0800)]
[LowerMatrixIntrinsics] Call getRegisterClassForType before getNumberOfRegisters.

getNumberOfRegisters takes a ClassID as it's argument. It shouldn't be passed a bool. Assuming the bool meant vector or not, we should call getRegisterClassForType first.

Reviewed By: fhahn

Differential Revision: https://reviews.llvm.org/D116903

2 years ago[clang-tidy] Refactor: Extract Class CheckRunner on check_clang_tidy.py
Richard [Sat, 1 Jan 2022 22:17:08 +0000 (15:17 -0700)]
[clang-tidy] Refactor: Extract Class CheckRunner on check_clang_tidy.py

Break up the huge function by extracting a class, storing intermediate
state as class members and breaking up the big function into a group
of class methods all at the same level of abstraction.

Differential Revision: https://reviews.llvm.org/D56343

2 years ago[HIP] Fix device only linking for -fgpu-rdc
Yaxun (Sam) Liu [Fri, 7 Jan 2022 18:33:00 +0000 (13:33 -0500)]
[HIP] Fix device only linking for -fgpu-rdc

Currently when -fgpu-rdc is specified, HIP toolchain always does host linking even
if --cuda-device-only is specified.

This patch fixes that. Only device linking is performed when --cuda-device-only
is specified.

Reviewed by: Artem Belevich

Differential Revision: https://reviews.llvm.org/D116840

2 years ago[lldb] fix memory leak in "GetGDBServerRegisterInfoXMLAndProcess"
Lirong Yuan [Mon, 10 Jan 2022 22:32:15 +0000 (14:32 -0800)]
[lldb] fix memory leak in "GetGDBServerRegisterInfoXMLAndProcess"

While running heap checker on a test that uses LLDB API, the following memory leak is found:

RAW: HeapChecker started...
RAW: Leak check _main_ detected leaks of 34 bytes in 4 objects
RAW: The 2 largest leaks:
RAW: Leak of 17 bytes in 2 objects allocated from:
@ 0x7fb93bd20166 NewHook()
@ 0x7fb929372a73 absl::base_internal::MallocHook::InvokeNewHookSlow()
@ 0x5600d1046093 libc_malloc
@ 0x7fb974529c03 xmlStrdup
@ 0x7fb9744c2a0b xmlGetProp
@ 0x7fb9749d9ed6 lldb_private::XMLNode::GetAttributeValue()
@ 0x7fb979043001 std::u::function::policy_invoker<>::__call_impl<>()
@ 0x7fb9749da06d lldb_private::XMLNode::ForEachChildElement()
@ 0x7fb97903c54d lldb_private::process_gdb_remote::ProcessGDBRemote::GetGDBServerRegisterInfoXMLAndProcess()
@ 0x7fb97902cfe4 lldb_private::process_gdb_remote::ProcessGDBRemote::GetGDBServerRegisterInfo()
@ 0x7fb97902c1d0 lldb_private::process_gdb_remote::ProcessGDBRemote::BuildDynamicRegisterInfo()
@ 0x7fb97902e92a lldb_private::process_gdb_remote::ProcessGDBRemote::SetThreadStopInfo()
@ 0x7fb97902db18 lldb_private::process_gdb_remote::ProcessGDBRemote::DoConnectRemote()
@ 0x7fb97584965e lldb_private::Process::ConnectRemote()
@ 0x7fb975839fa6 lldb_private::Platform::DoConnectProcess()
@ 0x7fb97583a39e lldb_private::Platform::ConnectProcessSynchronous()
@ 0x7fb97545b28b CommandObjectProcessConnect::DoExecute()
@ 0x7fb9755a70c9 lldb_private::CommandObjectParsed::Execute()
@ 0x7fb97559c0e9 lldb_private::CommandInterpreter::HandleCommand()
@ 0x7fb975460145 lldb_private::CommandObjectRegexCommand::DoExecute()
@ 0x7fb9755a72d2 lldb_private::CommandObjectRaw::Execute()
@ 0x7fb97559c0e9 lldb_private::CommandInterpreter::HandleCommand()
@ 0x7fb997a5f22e lldb::SBCommandInterpreter::HandleCommand()
@ 0x7fb997a5ef9b lldb::SBCommandInterpreter::HandleCommand()

This change fixes the memory leaks by freeing memory after it is no
longer in use. Tested with "ninja check-lldb".

Differential revision: https://reviews.llvm.org/D116707

2 years agoDon't fail if unable to promote loops during unrolling
Tyler Augustine [Mon, 10 Jan 2022 21:42:03 +0000 (21:42 +0000)]
Don't fail if unable to promote loops during unrolling

When the unroll factor is 1, we should only fail "unrolling" when the trip count also is determined to be 1 and it is unable to be promoted.

Reviewed By: bondhugula

Differential Revision: https://reviews.llvm.org/D115365

2 years ago[libc++][NFC] Use _LIBCPP_DEBUG_ASSERT in <list>
Nikolas Klauser [Mon, 10 Jan 2022 15:31:04 +0000 (16:31 +0100)]
[libc++][NFC] Use _LIBCPP_DEBUG_ASSERT in <list>

Use `_LIBCPP_DEBUG_ASSERT` in `<list>`

Reviewed By: Quuxplusone, ldionne, #libc

Spies: libcxx-commits

Differential Revision: https://reviews.llvm.org/D116938

2 years ago[openmp][amdgpu] Replace unsigned long with uint64_t
Jon Chesterfield [Mon, 10 Jan 2022 22:19:29 +0000 (22:19 +0000)]
[openmp][amdgpu] Replace unsigned long with uint64_t

Some types need to be 64 bit. Unsigned long is a hazard there.

Reviewed By: arsenm

Differential Revision: https://reviews.llvm.org/D116963

2 years ago[libcxx][test] Auto-detect _LIBCPP_HAS_NO_UNICODE for MSVC STL
Casey Carter [Mon, 10 Jan 2022 21:56:22 +0000 (13:56 -0800)]
[libcxx][test] Auto-detect _LIBCPP_HAS_NO_UNICODE for MSVC STL

... by examining `_MSVC_EXECUTION_CHARACTER_SET` in the force-include header.

Differential Revision: https://reviews.llvm.org/D116883

2 years agoFix a missed opportunity to merge stores.
Nadav Rotem [Mon, 10 Jan 2022 20:51:37 +0000 (12:51 -0800)]
Fix a missed opportunity to merge stores.

This commit fixes a missed opportunity in merging consecutive stores.
The code that searches for stores skipped the case of stores that
directly connect to the root. The comment above the implementation lists
this case but the code did not handle it. I found this pattern when
looking into the shared_ptr destructor. GCC generates the right
sequence. Here is a small repo:

    int foo(int* buff) {
        buff[0] = 0;
        int x = buff[1];
        buff[1] = 0;
        return x;
    }

Differential Revision: https://reviews.llvm.org/D116895

2 years ago[llvm] Fix the module build
Jonas Devlieghere [Mon, 10 Jan 2022 21:39:49 +0000 (13:39 -0800)]
[llvm] Fix the module build

Fix the module build after 2c0fb96254fef2509b66d75290fedafd4adede95