Philip Reames [Fri, 17 Dec 2021 17:20:10 +0000 (09:20 -0800)]
[capturetracking] Explicitly check for callee operand [NFC]
Pull out an explicit check rather than relying on the fact that the callee operand is not a data operand. The only real value is it gives us a clear place to move the comment, and makes the code slightly more understandable.
Craig Topper [Fri, 17 Dec 2021 16:58:28 +0000 (08:58 -0800)]
[RISCV} Add FSGNJ_H to isAsCheapAsAMove and isCopyInstrImpl.
This matches FSGNJ_S and FSGNJ_D.
Simon Pilgrim [Fri, 17 Dec 2021 17:01:17 +0000 (17:01 +0000)]
[DAG] Constant fold + canonicalize fp binops before SimplifyVBinOp call
Replace custom constant scalar/splat folding with FoldConstantArithmetic call and canonicalize commutative constant ops to the RHS before the SimplifyVBinOp call
Louis Dionne [Thu, 16 Dec 2021 21:59:56 +0000 (16:59 -0500)]
[libc++] Add a bunch of missing inline and _LIBCPP_HIDE_FROM_ABI in __threading_support
The inline keyword is required on those functions because they are defined
in the headers, so we need them to be inline to avoid ODR violations.
While we're at it, slap _LIBCPP_HIDE_FROM_ABI on them because they are
implementation details and we don't want them to be part of our ABI under
any circumstances.
Differential Revision: https://reviews.llvm.org/D115906
Philip Reames [Fri, 17 Dec 2021 16:32:43 +0000 (08:32 -0800)]
[funcattrs] Consistently treat calling a function pointer as a non-capturing read
We were being wildly inconsistent about what memory access was implied by an indirect function call. Depending on the call site attributes, you could get anything from a read, to unknown, to none at all. (The last was a miscompile.)
We were also always traversing the uses of a readonly indirect call. This is entirely unneeded as the indirect call does not capture. The callee might capture itself internally, but that has no implications for this caller. (See the nice explanation in the CaptureTracking comments if that case is confusing.)
Note that elsewhere in the same file, we were correctly computing the nocapture attribute for indirect calls. The changed case only resulted in conservatism when computing memory attributes if say the return value was written to.
Differential Revision: https://reviews.llvm.org/D115916
David Truby [Fri, 17 Dec 2021 15:46:19 +0000 (15:46 +0000)]
[AArch64][SVE] Improve code generation for VLS i1 masks
This patch partially resolves an issue for VLS code generation
where a mask is generated from a smaller width integer comparison
than the instruction using the mask requires.
Instead of sign extending a p register by converting it to a z
register, extending that, and converting back, we instead just
do an unpack of the p register.
A separate issue causes the code generation to still be poor when
the mask generation would fit in a neon register, as we then use
a neon comparison operation and have to convert that to a p register.
This will be resolved in a separate patch.
Reviewed By: peterwaller-arm
Differential Revision: https://reviews.llvm.org/D111221
Simon Pilgrim [Fri, 17 Dec 2021 16:00:30 +0000 (16:00 +0000)]
[DAG] Move foldConstantFPMath() inside FoldConstantArithmetic
Further merging of integer and fp constant folding paths.
This allows us to handle undef vector arguments the same as scalar cases.
Simon Pilgrim [Fri, 17 Dec 2021 15:54:53 +0000 (15:54 +0000)]
[X86] Rename funnel-shift X32 check prefixes to X86
We try to use X32 for gnux32 triple checks only
Simon Pilgrim [Fri, 17 Dec 2021 15:53:26 +0000 (15:53 +0000)]
[X86] Rename lzcnt/tzcnt X32 check prefixes to X86
We try to use X32 for gnux32 triple checks only
Nikita Popov [Fri, 17 Dec 2021 09:31:50 +0000 (10:31 +0100)]
[ConstantFolding] Unify handling of load from uniform value
There are a number of places that specially handle loads from a
uniform value where all the bits are the same (zero, one, undef,
poison), because we a) don't care about the load offset in that
case and b) it bypasses casts that might not be legal generally
but do work with uniform values.
We had multiple implementations of this, with a different set of
supported values each time, as well as incomplete type checks in
some cases. In particular, this fixes the assertion reported in
https://reviews.llvm.org/D114889#3198921, as well as a similar
assertion that could be triggered via constant folding.
Differential Revision: https://reviews.llvm.org/D115924
Carlo Bertolli [Fri, 17 Dec 2021 15:56:57 +0000 (15:56 +0000)]
[OpenMP][libomptarget] Fix __tgt_rtl_run_target_team_region_async API with missing parameter
I missed the async info parameter in the first version of this API.
Reviewed By: JonChesterfield
Differential Revision: https://reviews.llvm.org/D115887
Sanjay Patel [Fri, 17 Dec 2021 15:38:43 +0000 (10:38 -0500)]
[x86] add RUN line to test file for 32-bit target; NFC
More coverage for D115885
Momchil Velikov [Fri, 17 Dec 2021 15:07:19 +0000 (15:07 +0000)]
[AA] Correctly maintain the sign of PartiaAlias offset
Preserve the invariant that offset reported in the case of a
`PartialAlias` between `Loc1` and `Loc2`, is such that
`Loc1 + Offset = Loc2`, where `Loc1` and `Loc2` are the first and
the second argument, respectively, in alias queries.
Differential Revision: https://reviews.llvm.org/D115927
Simon Pilgrim [Fri, 17 Dec 2021 15:19:13 +0000 (15:19 +0000)]
[DAG] foldConstantFPMath - fold vector splats as well as scalar constants
Simon Pilgrim [Fri, 17 Dec 2021 15:03:31 +0000 (15:03 +0000)]
Revert rG22dbc7a48bf7a3942a7e5ff57977ef828d240bd3 "[DAG] foldConstantFPMath - fold vector splats as well as scalar constants"
A followup patch uncovered an issue with allowing undef elements in the splat - I will reapply this with a fixed implementation.
Nikita Popov [Fri, 17 Dec 2021 15:12:07 +0000 (16:12 +0100)]
[CodeGen] Fix element type for sret argument
Fix a mistake in
9bf917394eba3ba4df77cc17690c6d04f4e9d57f: sret
arguments use ConvertType, not ConvertTypeForMem, see the handling
in CodeGenTypes::GetFunctionType().
This fixes fp-matrix-pragma.c on s390x.
Benjamin Kramer [Fri, 17 Dec 2021 15:05:40 +0000 (16:05 +0100)]
Silence unused variable warning in release builds
lldb/source/Core/DataFileCache.cpp:278:10: warning: unused variable 'pos' [-Wunused-variable]
auto pos = m_string_to_offset.find(s);
^
lldb/source/Core/DataFileCache.cpp:277:18: warning: unused variable 'stroff' [-Wunused-variable]
const size_t stroff = encoder.GetByteSize() - strtab_offset;
^
Matthew Devereau [Thu, 19 Aug 2021 10:42:20 +0000 (11:42 +0100)]
[AArch64][SVE] Teach cost model that masked loads/stores are cheap
Reduce the cost of VLS masked loads/stores to make the vectorizor emit them more frequently.
David Truby [Fri, 17 Dec 2021 14:46:49 +0000 (14:46 +0000)]
[DAG][sve] Lowering for VLS masked truncating stores
This extends the custom lowering for truncating stores on
fixed length vectors in SVE to support masked truncating stores.
It also adds a DAG combine for truncates followed by masked
stores.
Reviewed By: peterwaller-arm, paulwalker-arm
Differential Revision: https://reviews.llvm.org/D108115
Nico Weber [Fri, 17 Dec 2021 15:02:26 +0000 (10:02 -0500)]
[gn build] (manually) port
fb9a075c813c5
Nico Weber [Fri, 17 Dec 2021 14:50:59 +0000 (09:50 -0500)]
[lld/test] List one test dep per line
Matches llvm's and clang's /test/CMakeLists.txt, makes it easier to
see in diffs which deps get added, and makes it easier to see if
a given dependency is present or not.
No behavior change.
Nico Weber [Fri, 17 Dec 2021 14:36:16 +0000 (09:36 -0500)]
[asan/mac] Fix remaining -Wformat warnings
AARCH64_GET_REG() is used to initialize uptrs, and after D79132
the ptrauth branch of its implementation explicitly casts to uptr.
The non-ptrauth branch returns ucontext->uc_mcontext->__ss.__fp (etc),
which has either type void* or __uint64_t (ref usr/include/mach/arm/_structs.h)
where __uint64_t is a unsigned long long (ref usr/include/arm/_types.h).
uptr is an unsigned long (ref
compiler-rt/lib/sanitizer_common/sanitizer_internal_defs.h). So explicitly
cast to uptr in this branch as well, so that AARCH64_GET_REG() has a
well-defined type.
Then change DUMPREGA64() tu use %lx instead of %llx since that's the right type
for uptr. (Most other places in compiler-rt print uptrs as %p and cast the arg
to (void*), but there are explicit 0x%016 format strings in the surroundings,
so be locally consistent with that.)
No behavior change, in the end it's just 64-bit unsigneds by slightly different
names.
Simon Pilgrim [Fri, 17 Dec 2021 14:24:26 +0000 (14:24 +0000)]
[DAG] foldConstantFPMath - fold vector splats as well as scalar constants
Nico Weber [Fri, 17 Dec 2021 14:16:07 +0000 (09:16 -0500)]
[asan/mac] Fix a few -Wformat warnings
All other places cast uptrs to (void*) when printing it via
%p after D113099, so do that here too.
Nico Weber [Fri, 17 Dec 2021 14:15:06 +0000 (09:15 -0500)]
[asan/mac] Fix a -Wformat warning
internal_getpid() is manually cast to int in all other places that
pass its result to a printf string, so do that here too for now.
Erich Keane [Mon, 13 Dec 2021 16:40:03 +0000 (08:40 -0800)]
Implement some constexpr vector unary operators, fix boolean-ops
As requested in the review, this implements unary +,-,~, and ! for
vector types.
All of our boolean operations on vector types should be using something
like vcmpeqd, which results in a mask of '-1' for the 'truth' type. We are
currently instead using '1', which results in some incorrect
calculations when used later (note that it does NOT result in a boolean
vector, as that is not really a thing).
This patch corrects that 1 to be a -1, and updates the affected tests.
Differential Revision: https://reviews.llvm.org/D115670
Simon Pilgrim [Fri, 17 Dec 2021 13:59:19 +0000 (13:59 +0000)]
[AArch64] Regenerate concat_vector-truncate-combine.ll tests
Paul Walker [Thu, 9 Dec 2021 18:30:22 +0000 (18:30 +0000)]
[NFC][SVE] Add missing tests for i32 INC/DEC patterns.
D111441 included trunc isel patterns for sve_int_pred_pattern_a
but no accompanying tests. This patch adds the missing tests and
also simplifies the isel patterns that use sve_cnt_shl_imm.
Differential Revision: https://reviews.llvm.org/D115512
Jan Svoboda [Fri, 17 Dec 2021 12:00:52 +0000 (13:00 +0100)]
[clang][deps] NFC: Unify ErrorOr patterns
This patch canonicalized some code into repetitive ErrorOr pattern. This will make refactoring easier if we ever come up with a way to simplify this.
Jan Svoboda [Fri, 17 Dec 2021 12:02:34 +0000 (13:02 +0100)]
[clang][deps] NFC: Take and store entry as reference
Jan Svoboda [Fri, 17 Dec 2021 12:43:12 +0000 (13:43 +0100)]
[clang][deps] NFC: Remove explicit call to implicit constructor
Jan Svoboda [Fri, 17 Dec 2021 12:30:38 +0000 (13:30 +0100)]
[clang][deps] NFC: Rename member variable
Jan Svoboda [Fri, 17 Dec 2021 12:34:56 +0000 (13:34 +0100)]
[clang][deps] NFC: Fix whitespace formatting
Pavel Labath [Fri, 17 Dec 2021 12:51:58 +0000 (13:51 +0100)]
[lldb/qemu] Add emulator-env-vars setting
This setting is for variables we want to pass to the emulator only --
then will be automatically removed from the target environment by our
environment diffing code. This variable can be used to pass various
QEMU_*** variables (although most of these can be passed through
emulator-args as well), as well as any other variables that can affect
the operation of the emulator (e.g. LD_LIBRARY_PATH).
Simon Pilgrim [Fri, 17 Dec 2021 12:33:54 +0000 (12:33 +0000)]
[DAG] foldConstantFPMath - use APFloat& for read-only constant fold arg. NFC.
We just need to copy the 1st arg (which we use for the constant fold result) - use a cheaper const reference for the 2nd arg.
Nico Weber [Fri, 17 Dec 2021 12:26:34 +0000 (07:26 -0500)]
[gn build] (semiautomatically) port
3f5f687e2e8b
Sjoerd Meijer [Fri, 17 Dec 2021 12:02:44 +0000 (12:02 +0000)]
[FuncSpec] Create helper to update state. NFC.
This creates a helper function updateSpecializedFuncs and is a NFC just to make
the function that drives the transformation easier to read.
Simon Pilgrim [Fri, 17 Dec 2021 12:01:56 +0000 (12:01 +0000)]
[DAG] Constant fold + canonicalize integer binops before SimplifyVBinOp call
SimplifyVBinOp still has a FoldConstantArithmetic call, which now it isn't vector specific we should be able to remove (once fp binops are tidied up); but we can at least clean up the integer opcodes to perform the basic constant/undef handling in common code first.
OCHyams [Fri, 17 Dec 2021 11:39:35 +0000 (11:39 +0000)]
[NFC] Fix type-units-maybe-unused-types.ll from D115325 (
78d15a112cbd)
78d15a112cbd adds llvm/test/DebugInfo/Generic/type-units-maybe-unused-types.ll
Move the test into llvm/test/DebugInfo/X86 and add -mtriple=x86_64-linux-gnu
because not all platforms support type units.
Example of failing bot: type-units-maybe-unused-types.ll
Original review: https://reviews.llvm.org/D115325
Nikita Popov [Fri, 17 Dec 2021 10:00:25 +0000 (11:00 +0100)]
[CodeGen] Avoid more pointer element type accesses
Denys Petrov [Thu, 16 Dec 2021 16:48:30 +0000 (18:48 +0200)]
[analyzer][NFC] Change return value of StoreManager::attemptDownCast function from SVal to Optional<SVal>
Summary: Refactor return value of `StoreManager::attemptDownCast` function by removing the last parameter `bool &Failed` and replace the return value `SVal` with `Optional<SVal>`. Make the function consistent with the family of `evalDerivedToBase` by renaming it to `evalBaseToDerived`. Aligned the code on the call side with these changes.
Differential Revision: https://reviews.llvm.org/
rkorsa [Fri, 17 Dec 2021 07:26:38 +0000 (12:56 +0530)]
[AMDGPU] Fixes in ISelDAG path and GlobalISel path for 'bias' operand with A16 bit on
The LOD bias operand is of type 'half' when the A16-bit is ON' for MIMG instructions.
'bias' is only 16-bit but occupies 32-bits with upper 16-bits containing junk.
The patch fixes both the paths(ISelDAG and GlobalISel) for proper encoding of LOD bias operand.
Differential Revision: https://reviews.llvm.org/D111754
Andrzej Warzynski [Tue, 23 Nov 2021 18:13:51 +0000 (18:13 +0000)]
[flang][codegen] Add a conversion for `fir.coordinate_of` - part 2
This patch extends the `FIRToLLVMLowering` pass in Flang by extending
the hook to transform `fir.coordinate_of` into a sequence of LLVM MLIR
instructions (i.e. `CoordinateOpConversion::doRewrite`). The following
case is added:
3.1 the input object is inside `!fir.ref` (e.g. `!fir.ref<!fir.array>` or
`!fir.ref<!fir.type>`).
3.2 the input object is inside `!fir.ptr` (e.g. `!fir.ptr<!fir.array>` or
`!fir.ptr<!fir.type>`).
From the point of view of the conversion, 3.1 and 3.2 are currently identical.
This is part of the upstreaming effort from the `fir-dev` branch in [1].
[1] https://github.com/flang-compiler/f18-llvm-project
Originally written by:
Co-authored-by: Jean Perier <jperier@nvidia.com>
Co-authored-by: Eric Schweitz <eschweitz@nvidia.com>
Co-authored-by: V Donaldson <vdonaldson@nvidia.com>
Depends on: D114159
Differential Revision: https://reviews.llvm.org/D115333
Nikolas Klauser [Thu, 16 Dec 2021 13:55:03 +0000 (14:55 +0100)]
[libc++] Disable _LIBCPP_DEBUG_ASSERT during constant evaluation
Disable `_LIBCPP_DEBUG_ASSERT` and debug iterators in <string> during constant evaluation
Reviewed By: ldionne, #libc
Spies: goncharov, libcxx-commits
Differential Revision: https://reviews.llvm.org/D115788
Florian Hahn [Mon, 13 Dec 2021 19:12:56 +0000 (19:12 +0000)]
[LV] Pass VectorHeader block to emitTransformedIndex (NFC).
Pass in the vector header instead of relying on ILV::LoopVectorBody.
This reduces the dependence on state from ILV. Where VPTransformState is
available, State.CFG.PrevBB can be used.
OCHyams [Fri, 17 Dec 2021 09:25:52 +0000 (09:25 +0000)]
[DWARF] Fix PR51087 Extraneous enum record in DWARF with type units
Fixes https://llvm.org/PR51087: Extraneous enum record in DWARF with type units.
As explained in PR51087 we sometimes get skeleton DIEs for enums in a Dwarf
Compile Unit (CU) that are not referenced from any CU and are already described
by a type unit.
Types for enums are emitted whether used or not, all together before most types
in the CU. Mechanically, the extraneous CU records are generated because the
enum types are generated with a call to CU->getOrCreateTypeDIE. This function
will recursively get-or-create the parent DIE (in the CU) and the type unit for
each. We don't need the CU-side DIEs if the type units are sucesfully
emitted. Fix by only emitting the type units for enums if possible, falling back
to a call to getOrCreateTypeDIE if not. Do the same for retained types.
Reviewed By: dblaikie
Differential Revision: https://reviews.llvm.org/D115325
Andrzej Warzynski [Thu, 14 Oct 2021 08:07:30 +0000 (08:07 +0000)]
[flang] Make the frontend driver error out when requesting multiple actions
With this change, the following invocations will be treated as errors
(multiple actions are specified):
```
$ flang-new -fc1 -E -fsyntax-only file.95
$ flang-new -fc1 -fsyntax-only -fdebug-dump-symbols file.95
```
In the examples above it is not clear whether it is `-fsyntax-only` or
the other action that is run (i.e. `-E` or `-fdebug-dump-symbols`). It
makes sense to disallow such usage. This should also lead to cleaner and
clearer tests (the `RUN` lines using `%flang_fc1` will only allow one
action).
This change means that `flang-new -fc1` and `clang -cc1` will behave
differently when multiple action options are specified. As frontend
drivers are mostly used by compiler developers, this shouldn't affect or
confuse the compiler end-users. Also, `flang-new` and `clang` remain
consistent.
Tests are updated accordingly. More specifically, I've made sure that
every test specifies only one action. I've also taken the opportunity to
simplify "multiple-input-files.f90" a bit.
Differential Revision: https://reviews.llvm.org/D111781
Sjoerd Meijer [Fri, 17 Dec 2021 09:25:45 +0000 (09:25 +0000)]
[FuncSpec] Respect MaxConstantsThreshold
This is a follow up of D115458 and truncates the worklist of actual arguments
that can be specialised to 'MaxConstantsThreshold' candidates if
MaxConstantsThreshold was exceeded. Thus, this changes the behaviour of option
-func-specialization-max-constants. Before it didn't specialise at all when
this threshold was exceeded, but now it specialises up to MaxConstantsThreshold
candidates from the sorted worklist.
Differential Revision: https://reviews.llvm.org/D115509
David Green [Fri, 17 Dec 2021 09:16:28 +0000 (09:16 +0000)]
[ARM] Handle splats of constants for MVE qr instruction
Some MVE instructions have qr variants that take a Q and R register,
splatting the R register for each lane. This is usually handled fine for
standard splats as we sink the splat into the loop and combine the
resulting dup into the qr instruction. It does not work for constant
splats though, as we generate a vmovimm or constant pool load instead.
This intercepts that, generating a vdup of the constant instead where we
can turn the result into a qr instruction variant.
Differential Revision: https://reviews.llvm.org/D115242
Pavel Labath [Thu, 28 Oct 2021 11:33:58 +0000 (13:33 +0200)]
[lldb] Fix matchers for char array formatters
They were being applied too narrowly (they didn't cover signed char *,
for instance), and too broadly (they covered SomeTemplate<char[6]>) at
the same time.
Differential Revision: https://reviews.llvm.org/D112709
Sjoerd Meijer [Thu, 16 Dec 2021 16:06:23 +0000 (16:06 +0000)]
Recommit "[FuncSpec] Decouple cost/benefit analysis, allowing sorting of candidates."
Replaced llvm:sort with llvm::stable_sort, this was failing on the bot with
expensive checks enabled.
seven-mile [Fri, 17 Dec 2021 08:55:42 +0000 (19:55 +1100)]
[examples][BuildingAJIT] Fix use-after-move bug in Chapter 4.
Reviewed By: lhames
Differential Revision: https://reviews.llvm.org/D115260
Nikita Popov [Thu, 16 Dec 2021 15:56:51 +0000 (16:56 +0100)]
[CodeGen] Store element type in RValue
For aggregates, we need to store the element type to be able to
reconstruct the aggregate Address. This increases the size of this
packed structure (as the second value is already used for alignment
in this case), but I did not observe any compile-time or memory
usage regression from this change.
Nikita Popov [Thu, 16 Dec 2021 08:36:42 +0000 (09:36 +0100)]
[llvm-c] Accept GEP operators in some APIs
As requested in D115787, I've added a test for LLVMConstGEP2 and
LLVMConstInBoundsGEP2. However, to make this work in the echo test,
I also had to change a couple of APIs to work on GEP operators,
rather than only GEP instructions.
Differential Revision: https://reviews.llvm.org/D115858
Fangrui Song [Fri, 17 Dec 2021 07:42:50 +0000 (23:42 -0800)]
[ELF] InputSection::writeTo: reorder type checks and add LLVM_UNLIKELY
Aaron DeBattista [Fri, 17 Dec 2021 07:24:47 +0000 (23:24 -0800)]
[mlir][tosa] Move tosa canonicalizers to optional optimization pass
TOSA's canonicalizers that change dense operations should be moved to a
seperate optimization pass to avoid canonicalizing to operations not supported
for relevant backends.
Reviewed By: rsuderman
Differential Revision: https://reviews.llvm.org/D115890
Vitaly Buka [Thu, 16 Dec 2021 05:54:53 +0000 (21:54 -0800)]
Passthrough LLVM_USE_LLD LLVM_USE_LINKER into runtimes
Fixes https://github.com/llvm/llvm-project/issues/42339
Reviewed By: phosek, sylvestre.ledru
Differential Revision: https://reviews.llvm.org/D115852
Clemens Wasser [Thu, 16 Dec 2021 19:56:29 +0000 (11:56 -0800)]
Port __sanitizer::StopTheWorld to Windows
This also makes the sanitizer_stoptheworld_test cross-platform by using the STL, rather than pthread.
Reviewed By: vitalybuka
Differential Revision: https://reviews.llvm.org/D115204
Vitaly Buka [Thu, 16 Dec 2021 22:14:09 +0000 (14:14 -0800)]
[sanitizer] Simplify sanitizer_stoptheworld_test
David Blaikie [Fri, 17 Dec 2021 06:22:54 +0000 (22:22 -0800)]
DWARFVerifier: Skip resolution failures for locations in dwo files
When reading location lists in dwo files the addresses cannot be
resolved, but that's not a problem.
Long term this probably should be fixed with a different API that
exposes location expressions without the need to resolve the address
ranges, since that's all the verifier (in its current state) requires.
(though the verifier should probably also eventually verify the address
ranges in location lists are a subset of the enclosing scope's address
range)
David Blaikie [Fri, 17 Dec 2021 05:32:40 +0000 (21:32 -0800)]
DWARFVerifier: Test failures to parse DW_AT_locations
Fangrui Song [Fri, 17 Dec 2021 05:22:59 +0000 (21:22 -0800)]
[ELF] Optimize MergeInputSection::splitNonStrings. NFC
Fangrui Song [Fri, 17 Dec 2021 05:17:02 +0000 (21:17 -0800)]
[ELF] Speed up MergeInputSection::split*. NFC
Fangrui Song [Fri, 17 Dec 2021 05:07:39 +0000 (21:07 -0800)]
[ELF] Use SmallVector for MergeInputSection::pieces. NFC
sizeof(pieces) decreases from 24 to 16 on ELF64.
One BumpPtrAllocator can store more MergeInputSections.
The lld executable becomes smaller.
Fangrui Song [Fri, 17 Dec 2021 04:50:06 +0000 (20:50 -0800)]
[ELF] Internalize createMergeSynthetic. NFC
Only called once. Moving to OutputSections.cpp can make it inlined.
finalizeInputSections can be very hot, especially in -O1 links with much debug info.
David Blaikie [Fri, 17 Dec 2021 04:30:33 +0000 (20:30 -0800)]
DWARFVerifier: Cleanup a test, remove redundant abbreviation
David Blaikie [Fri, 17 Dec 2021 04:06:07 +0000 (20:06 -0800)]
DWARFVerifier: Don't error on missing ranges in Split DWARF
When verifying dwo files address ranges won't be able to be resolved due
to missing debug_addr (or missing debug_ranges in the case of DWARFv4
Split DWARF).
wlei [Thu, 16 Dec 2021 03:14:50 +0000 (19:14 -0800)]
[llvm-profgen] Fix warning of enumerated and non-enumerated type in conditional expression
Differential Revision: https://reviews.llvm.org/D115842
Mogball [Fri, 17 Dec 2021 02:44:56 +0000 (02:44 +0000)]
[mlir][ods] Added EnumAttr, an AttrDef implementation of enum attributes
`EnumAttr` is a pure TableGen implementation of enum attributes using `AttrDef`. This is meant as a drop-in replacement for `StrEnumAttr`, which is soon to be deprecated. `StrEnumAttr` is often used over `IntEnumAttr` because its more readable in MLIR assembly formats. However, storing and manipulating strings is not efficient. Defining `StrEnumAttr` can also be awkward and relies on a lot of special logic in `EnumsGen`, and has some hidden sharp edges.
Also, `EnumAttr` stores the enum directly, removing the need to convert to/from integers when calling attribute getters on ops.
Reviewed By: mehdi_amini
Differential Revision: https://reviews.llvm.org/D115181
Chuanqi Xu [Fri, 17 Dec 2021 02:35:17 +0000 (10:35 +0800)]
[Modules] Incorrect ODR detection for unresolved using type
Implement `getUnresolvedUsingType()` and don't create a new
`UnresolvedUsingType` when there is already canonical declaration.
This solved an incorrect ODR detection in modules for uresolved using
type.
Reviewed By: rjmccall
Differential Revision: https://reviews.llvm.org/D115792
David Blaikie [Thu, 16 Dec 2021 23:20:21 +0000 (15:20 -0800)]
DWARFVerifier: Don't try to resolve decl_file in split CUs
Since they refer to the debug_line in the skeleton unit, they can't be
resolved from the dwo CU.
But they can be resolved for split TUs, since those refer to
.debug_line.dwo, which is available in the dwo file.
LLVM GN Syncbot [Fri, 17 Dec 2021 01:57:11 +0000 (01:57 +0000)]
[gn build] Port
09103807e7dc
Mircea Trofin [Tue, 14 Dec 2021 06:49:57 +0000 (22:49 -0800)]
[NFC][regalloc] Introduce the RegAllocEvictionAdvisorAnalysis
This patch introduces the eviction analysis and the eviction advisor,
the default implementation, and the scaffolding for introducing the
other implementations of the advisor.
Differential Revision: https://reviews.llvm.org/D115707
Nico Weber [Fri, 17 Dec 2021 01:46:51 +0000 (20:46 -0500)]
Revert "[analyzer][ctu] Fix wrong 'multiple definitions' errors caused by space characters in lookup names when parsing the ctu index file"
This reverts commit
333d66b09494b7ebc1a89f2befa79128a56f77e3.
Breaks tests on macOS, see comments on https://reviews.llvm.org/D102669
Philip Reames [Fri, 17 Dec 2021 01:27:54 +0000 (17:27 -0800)]
Extract a helper function for computing estimate trip count of an exiting branch
Plan to use this in following change to support estimated trip counts derived from multiple loop exits.
Heejin Ahn [Thu, 16 Dec 2021 02:10:57 +0000 (18:10 -0800)]
[WebAssembly] Support clang -fwasm-exceptions for bitcode
This supports bitcode compilation using `clang -fwasm-exceptions`.
---
The current situation:
Currently the backend requires two options for Wasm EH:
`-wasm-enable-eh` and `-exception-model=wasm`. Wasm SjLj requires two
options as well: `-wasm-enable-sjlj` and `-exception-model=wasm`. When
using Wasm EH via Emscripten, you only need to pass `-fwasm-exceptions`,
and these options will be added within the clang driver. This
description will focus on the case of Wasm EH going forward, but Wasm
SjLj's case is similar.
When you pass `-fwasm-exceptions` to emcc and clang driver, the clang
driver adds these options to the command line that calls the clang
frontend (`clang -cc1`): `-mllvm -wasm-enable-eh` and
`-exception-model=wasm`. `-wasm-enable-eh` is prefixed with `-mllvm`, so
it is passed as is to the backend. But `-exception-model` is parsed and
processed within the clang frontend and stored in `LangOptions` class.
This info is later transferred to `TargetOptions` class, and then
eventually passed to `MCAsmInfo` class. All LLVM code queries this
`MCAsmInfo` to get the exception model.
---
Problem:
The problem is the whole `LangOptions` processing is bypassed when
compiling bitcode, so the information transfer of `LangOptions` ->
`TargetOptions` -> `MCAsmInfo` does not happen. They are all set to
`ExceptionHandling::None`, which is the default value.
---
What other targets do, and why we can't do the same:
Other targets support bitcode compilation by the clang driver, but they
can do that by using different triples. For example, X86 target supports
multiple triples, each of which has its own subclass of `MCAsmInfo`, so
it can hardcode the appropriate exception model within those subclasses'
constructors. But we don't have separate triples for each exception
mode: none, emscripten, and wasm.
---
What this CL does:
If we can figure out whether `-wasm-enable-eh` is passed to the backend,
we can programatically set the exception model from the backend, rather
than requiring it to be passed.
So we check `WasmEnableEH` and `WasmEnableSjLj` variables, which are
`cl::opt` for `-wasm-enable-eh` and `-wasm-enable-sjlj`, in
`WebAssemblyMCAsmInfo` constructor, and if either of them is set, we set
`MCAsmInfo.ExceptionType` to Wasm. `TargetOptions` cannot be updated
there, so we make sure they are the same later.
Fixes https://github.com/emscripten-core/emscripten/issues/15712.
Reviewed By: dschuff
Differential Revision: https://reviews.llvm.org/D115893
Philip Reames [Fri, 17 Dec 2021 00:38:40 +0000 (16:38 -0800)]
[funcattrs] Add some additional indirect call tests
LLVM GN Syncbot [Fri, 17 Dec 2021 00:28:54 +0000 (00:28 +0000)]
[gn build] Port
bdc68ee70fa9
Ellis Hoag [Fri, 17 Dec 2021 00:27:08 +0000 (16:27 -0800)]
Revert "[InstrProf] Add Correlator class to read debug info"
Also reverts an attempt to fix the build errors https://reviews.llvm.org/D115911
The original diff https://reviews.llvm.org/D114566 causes some build
errors that I need to investigate.
https://lab.llvm.org/buildbot/#/builders/118/builds/7037
This reverts commit
95946d2f8589b5450d8f9a9e9b4f7adf44386f8b.
Reviewed By: kyulee
Differential Revision: https://reviews.llvm.org/D115913
Greg Clayton [Fri, 17 Dec 2021 00:13:58 +0000 (16:13 -0800)]
Fix windows buildbots after https://reviews.llvm.org/D115324
Windows has trouble deleting the executable due to permissions.
Ellis Hoag [Thu, 16 Dec 2021 23:58:51 +0000 (15:58 -0800)]
[InstrProf] Fix build error in Correlator class
For some reason this did not fail to compile for me locally, but did
fail on one of the builders. Here is the error:
https://lab.llvm.org/buildbot/#/builders/58/builds/14969
```
/home/buildbot/as-builder-4/lld-x86_64-ubuntu-fast/llvm-project/llvm/lib/ProfileData/InstrProfCorrelator.cpp: In static member function ‘static llvm::Expected<std::unique_ptr<llvm::InstrProfCorrelator::Context> > llvm::InstrProfCorrelator::Context::get(std::unique_ptr<llvm::MemoryBuffer>, const llvm::object::ObjectFile&)’:
/home/buildbot/as-builder-4/lld-x86_64-ubuntu-fast/llvm-project/llvm/lib/ProfileData/InstrProfCorrelator.cpp:44:10: error: could not convert ‘C’ from ‘std::unique_ptr<llvm::InstrProfCorrelator::Context>’ to ‘llvm::Expected<std::unique_ptr<llvm::InstrProfCorrelator::Context> >’
return C;
```
Reviewed By: kyulee
Differential Revision: https://reviews.llvm.org/D115911
Nico Weber [Thu, 16 Dec 2021 23:31:56 +0000 (18:31 -0500)]
[gn build] Port
95946d2f8589
Nico Weber [Thu, 16 Dec 2021 23:37:30 +0000 (18:37 -0500)]
[gn build] port
78523516bc (cut MC->ProfileData dep)
Rob Suderman [Thu, 16 Dec 2021 23:30:45 +0000 (15:30 -0800)]
[mlir][tosa] Handle rescale case where shift > 63
It is possible for the shift value to exceed the number of bits. In these
cases we can just multiply by zero. This is relatively rare occurence but
should be handled.
Reviewed By: not-jenni
Differential Revision: https://reviews.llvm.org/D115779
not-jenni [Thu, 16 Dec 2021 23:20:34 +0000 (15:20 -0800)]
[mlir][tosa] Add tosa.max_pool2d as no-op canonicalization
When the input and output of a pool2d op are both 1x1, it can be canonicalized to a no-op
Reviewed By: rsuderman
Differential Revision: https://reviews.llvm.org/D115908
Vitaly Buka [Thu, 16 Dec 2021 23:24:14 +0000 (15:24 -0800)]
[NFC][sanitizer] Remove unused variable
Ellis Hoag [Thu, 16 Dec 2021 23:21:54 +0000 (15:21 -0800)]
[dSYM][NFC] Add error test case for llvm-dwarfdump
Add a test to `llvm-dwarfdump` to simply test that the error messages
make sense when passing bad `.dSYM`s.
Reviewed By: clayborg
Differential Revision: https://reviews.llvm.org/D115889
Ellis Hoag [Thu, 16 Dec 2021 22:45:54 +0000 (14:45 -0800)]
[InstrProf] Add Correlator class to read debug info
Extend `llvm-profdata` to read in a `.proflite` file and also a debug info file to generate a normal `.profdata` profile. This reduces the binary size by 8.4% when building an instrumented Clang binary without value profiling (164 MB vs 179 MB).
This work is part of the "lightweight instrumentation" RFC: https://groups.google.com/g/llvm-dev/c/r03Z6JoN7d4
Reviewed By: kyulee
Differential Revision: https://reviews.llvm.org/D114566
Rob Suderman [Thu, 16 Dec 2021 21:53:33 +0000 (13:53 -0800)]
[mlir][tosa] Minor cleanup of tosa.conv2d canonicalizer
Slight rename and better variable type usage in tosa.conv2d to
tosa.fully_connected lowering. Included disabling pass for padded
convolutions.
Reviewed By: not-jenni
Differential Revision: https://reviews.llvm.org/D115776
Konstantin Varlamov [Thu, 16 Dec 2021 22:47:08 +0000 (14:47 -0800)]
[libc++][ranges][NFC] Remove extraneous cleanup checks.
There is no need to check the counters on `Counted` after destroying
elements in the range because these tests are not testing `destroy`.
Differential Revision: https://reviews.llvm.org/D115839
David Blaikie [Thu, 16 Dec 2021 22:39:51 +0000 (14:39 -0800)]
DebugInfoVerifier: Verify dwo units
Initial pass - follow-up patches to address various limitations (debug
ranges, location lists... )
Ellis Hoag [Thu, 16 Dec 2021 22:19:03 +0000 (14:19 -0800)]
[Try2][InstrProf] Attach debug info to counters
Add the llvm flag `-debug-info-correlate` to attach debug info to instrumentation counters so we can correlate raw profile data to their functions. Raw profiles are dumped as `.proflite` files. The next diff enables `llvm-profdata` to consume `.proflite` and debug info files to produce a normal `.profdata` profile.
Part of the "lightweight instrumentation" work: https://groups.google.com/g/llvm-dev/c/r03Z6JoN7d4
The original diff https://reviews.llvm.org/D114565 was reverted because of the `Instrumentation/InstrProfiling/debug-info-correlate.ll` test, which is fixed in this commit.
Reviewed By: kyulee
Differential Revision: https://reviews.llvm.org/D115693
Michael Jones [Fri, 10 Dec 2021 18:55:38 +0000 (10:55 -0800)]
[libc] fix memcpy builtin looping
previously, memcpy could get stuck in a loop, calling __builtin_memcpy
which would redirect to itself. This patch removes that path.
Reviewed By: gchatelet, sivachandra, lntue
Differential Revision: https://reviews.llvm.org/D115542
LLVM GN Syncbot [Thu, 16 Dec 2021 21:40:46 +0000 (21:40 +0000)]
[gn build] Port
8a85be807bd4
Ron Lieberman [Thu, 16 Dec 2021 21:39:04 +0000 (21:39 +0000)]
Revert "AMDGPU: Update pass pipeline test"
needed to match revert of
rG2b4876157562: AMDGPU: Remove AMDGPUFixFunctionBitcasts pass
This reverts commit
7ca355225d530acb24b50a5fcccf5c40938b30bf.
Anton Korobeynikov [Thu, 16 Dec 2021 21:31:19 +0000 (00:31 +0300)]
No need to checkout a repo
Anton Korobeynikov [Thu, 16 Dec 2021 21:30:01 +0000 (00:30 +0300)]
Send to llvm-bugs
Vitaly Buka [Thu, 16 Dec 2021 21:19:13 +0000 (13:19 -0800)]
[sanitizer] Define SANITIZER_ARM SANITIZER_X64 SANITIZER_ARM64
Vitaly Buka [Thu, 16 Dec 2021 21:21:06 +0000 (13:21 -0800)]
[sanitizer] Clang-format sanitizer_platform.h
Anton Korobeynikov [Thu, 16 Dec 2021 21:25:55 +0000 (00:25 +0300)]
Install test mailer for github bugs