There doesn't seem to be much sense in defaulting "on" unwind tables on
amd64 and not on other arches. It causes surprising differences between
platforms, such as the PR below[1].
Prior to this change, FreeBSD inherited the default implementation of the
method from the Gnu.h Generic_Elf => Generic_GCC parent class, which
returned true only for amd64 targets. Override that and opt on always,
similar to, e.g., NetBSD's driver.
[1] https://bugs.freebsd.org/241562
Patch by cem (Conrad Meyer).
Reviewed By: dim
Differential Revision: https://reviews.llvm.org/D70110
Summary:
we have a few places using `ASTCtx.getLangOpts().IsHeaderFile` to
determine a header file, but it relies on "-x c-header" compiler flag,
if the compilation command doesn't have this flag, we will get a false
positive. We are encountering this issue in bazel build system.
To solve this problem, we infer the file from file name, actual changes will
come in follow-ups.
Reviewers: sammccall
Subscribers: ilya-biryukov, MaskRay, jkorous, arphaman, kadircet, usaxena95, cfe-commits
Tags: #clang
Differential Revision: https://reviews.llvm.org/D70235
Summary:
Clang/LLVM is a cross-compiler, and so we don't have to make a choice
about `-march`/`-mabi` at build-time, but we may have to compute a
default `-march`/`-mabi` when compiling a program. Until now, each
place that has needed a default `-march` has calculated one itself.
This patch adds a single place where a default `-march` is calculated,
in order to avoid calculating different defaults in different places.
This patch adds a new function `riscv::getRISCVArch` which encapsulates
this logic based on GCC's for computing a default `-march` value
when none is provided. This patch also updates the logic in
`riscv::getRISCVABI` to match the logic in GCC's build system for
computing a default `-mabi`.
This patch also updates anywhere that `-march` is used to now use the
new function which can compute a default. In particular, we now
explicitly pass a `-march` value down to the gnu assembler.
GCC has convoluted logic in its build system to choose a default
`-march`/`-mabi` based on build options, which would be good to match.
This patch is based on the logic in GCC 9.2.0. This commit's logic is
different to GCC's only for baremetal targets, where we default
to rv32imac/ilp32 or rv64imac/lp64 depending on the target triple.
Tests have been updated to match the new logic.
Reviewers: asb, luismarques, rogfer01, kito-cheng, khchen
Reviewed By: asb, luismarques
Subscribers: sameer.abuasal, rbar, johnrusso, simoncook, apazos, sabuasal, niosHD, shiva0217, jrtc27, MaskRay, zzheng, edward-jones, MartinMosbeck, brucehoult, the_o, rkruppe, PkmX, jocewei, psnobl, benna, Jim, s.egerton, pzheng, cfe-commits
Tags: #clang
Differential Revision: https://reviews.llvm.org/D69383
Summary:
This is shorter, shouldn't be confusing (is consistent with how they're declared),
and avoids messy cases that are printed as myclass<type-param-0-0>(int) in the
case of partial specialization.
Fixes part of https://github.com/clangd/clangd/issues/76
Reviewers: hokein, lh123
Subscribers: ilya-biryukov, kadircet, usaxena95, cfe-commits
Tags: #clang
Differential Revision: https://reviews.llvm.org/D70307
Summary:
As well as vector/vector compare instructions, MVE also has a family
of comparisons taking a vector and a scalar, which compare every lane
of the vector against the same value. We generate those at isel time
using isel patterns that match `(ARMvcmp vector, (ARMvdup scalar))`.
This commit adds corresponding patterns for the operand-reversed form
`(ARMvcmp (ARMvdup scalar), vector)`, with condition codes swapped as
necessary. That way, we can still generate the vector/scalar compare
instruction if the IR happens to have been rearranged to put the
operands the other way round, which can happen in some optimization
phases. Previously, a vcmp the other way round was handled by emitting
a `vdup` instruction to //explicitly// replicate the scalar input into
a vector, and then doing a vector/vector comparison.
I haven't added a new test, because it turned out that several
existing tests were already exhibiting that failure mode. So just
updating the expected output in the existing MVE codegen tests
demonstrates what's been improved.
Reviewers: ostannard, MarkMurrayARM, dmgreen
Reviewed By: dmgreen
Subscribers: kristof.beyls, hiraditya, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D70296
Summary:
ASTImporter makes now difference between class templates with same
name in different translation units if these are not visible outside.
Reviewers: martong, a.sidorin, shafik
Reviewed By: martong
Subscribers: rnkovacs, dkrupp, Szelethus, gamesh411, teemperor, cfe-commits
Tags: #clang
Differential Revision: https://reviews.llvm.org/D67543
ValueInfo has user-defined 'operator bool' which allows incorrect implicit conversion
to GlobalValue::GUID (which is unsigned long). This causes bugs which are hard to
track and should be removed in future.
TestFormatters.py has a sequence of three 'next' commands to get past
all the initializations in the test function. On AArch64 (and
potentially other platforms), this was one 'next' too many and we ended
up outside our frame.
This patch replaces the sequence with a 'thread until ' the line of the
return from the function, so we should stop after all the
initializations but before actually returning.
Differential Revision: https://reviews.llvm.org/D70303
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
Summary:
Using c-index-test is fragile since it does not parse all the clang
arguments that are used in the RUN: line. This can result in incorrect
mangled names that do not match any of the generated IR.
For example macOS triples include a leading underscore (which was handled
with a hack in the current script). For the CHERI target we have added
new qualifiers which affect C++ name mangling, but will be included added
by update_cc_test_checks since it parses the source file with the host
triple because it ignores the -triple= argument passed to clang -cc1.
Using the new feature of including the mangled name in the JSON AST dump
(see D69564), we can parse the output of the RUN: command with
"-fsyntax-only -ast-dump=json" appended.
This should make the script less fragile and also forks one process less.
Reviewers: MaskRay, xbolva00
Reviewed By: MaskRay
Subscribers: llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D69565
I am planning to use this feature to make update_cc_test_checks.py less fragile
by obtaining the mangled names directly from -ast-dump=json. Currently,
it uses c-index-test which ignores the -triple=, etc. arguments that are
in the RUN: line and therefore does not generate checks for some targets.
The AST dump tests were updated using the following command:
`python $LLVM_BINDIR/gen_ast_dump_json_test.py --update --source $LLVM_SRC/clang/test/AST/*-json.*`
Reviewers: aaron.ballman
Reviewed By: aaron.ballman
Subscribers: rsmith, MaskRay, cfe-commits
Tags: #clang
Differential Revision: https://reviews.llvm.org/D69564
This was done by running `python $LLVM_BINDIR/gen_ast_dump_json_test.py --update --source $LLVM_SRC/clang/test/AST/*-json.*`
Only changes are whitespace and line endings.
Differential Revision: https://reviews.llvm.org/D70119
With this change it is possible to update all JSON dump tests using the
following command:
python $LLVM_BINDIR/gen_ast_dump_json_test.py --update --source $LLVM_SRC/clang/test/AST/*-json.*
See https://reviews.llvm.org/D70119
The script will now check if a clang binary exists in the same directory
and default to that instead of requiring a --clang argument. The script
is copied to the clang build directory using CMake configure_file() with
COPYONLY. This ensures that the version in the build directory is updated
any time the source version changes.
See https://reviews.llvm.org/D70119
This will allow updating the JSON tests for new format changes. Instead of
simply appending the JSON to the input file, the script will now make a
copy of the input file up to the "CHECK lines have been autogenerated"
disclaimer and then append the new JSON.
See https://reviews.llvm.org/D70119
Summary:
This adds a visitLocationList function to the DWARF v4 location lists,
similar to what already exists for DWARF v5. It follows the approach
outlined in previous patches (D69672), where the parsed form is always
stored in the DWARF v5 format, which makes it easier for generic code to
be built on top of that. v4 location lists are "upgraded" during
parsing, and then this upgrade is undone while dumping.
Both "inline" and section-based dumping is rewritten to reuse the
existing "generic" location list dumper. This means that the output
format is consistent for all location lists (the only thing one needs to
implement is the function which prints the "raw" form of a location
list), and that debug_loc dumping correctly processes base address
selection entries, etc.
The previous existing debug_loc functionality (e.g.,
parseOneLocationList) is rewritten on top of the new API, but it is not
removed as there is still code which uses them. This will be done in
follow-up patches, after I build the API to access the "interpreted"
location lists in a generic way (as that is what those users really
want).
Reviewers: dblaikie, probinson, JDevlieghere, aprantl, SouraVX
Subscribers: hiraditya, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D69847
The SmallVector reserve() call in
MachineInstrExpressionTrait::getHashValue accounted for over 3% of all
calls to malloc() when I compiled a bunch of graphics shaders for the
AMDGPU target. Its initial size was only enough for machine instructions
with up to 7 operands, but for AMDGPU 8 and 10 operands are very common.
Here's a histogram of number of operands for each call to getHashValue,
gathered from the same collection of shaders:
1 13503
2 254273
3 135781
4 422508
5 614997
6 194953
7 287248
8 1517255
9 31218
10 1191269
11 70731
12 24
13 77
15 84
17 4692
27 16
33 705
49 6
Typical instructions with 8 and 10 operands are floating point
arithmetic and multiply-accumulate instructions like:
%83:vgpr_32 = V_MUL_F32_e64 0, killed %82:vgpr_32, 0, killed %81:vgpr_32, 0, 0, implicit $exec
%330:vgpr_32 = V_MAC_F32_e64 0, killed %327:vgpr_32, 0, killed %329:sgpr_32, 0, %328:vgpr_32(tied-def 0), 0, 0, implicit $exec
Differential Revision: https://reviews.llvm.org/D70301
This is a follow up of d90804d, to also flag fmcp instructions as instructions
that we do not support in tail-predicated vector loops.
Differential Revision: https://reviews.llvm.org/D70295
Introduce IntImmLeaf version of PatLeaf immZExt16 for 32-bit immediates.
Change immZExt16 with imm32ZExt16 for andi, ori and xori.
This keeps same behavior for SDAG and allows for GlobalISel selectImpl
to select 'G_CONSTANT imm' + G_AND, G_OR, G_XOR into ANDi, ORi, XORi,
respectively, when 32-bit imm satisfies imm32ZExt16 predicate: zero
extending 16 low bits of imm is equal to imm.
Large number of test changes comes from zero extending of small types
which is transformed into 'and' with bitmask in legalizer.
Differential Revision:https://reviews.llvm.org/D70185
Introduce IntImmLeaf version of PatLeaf immSExt16 for 32-bit immediates.
Change immSExt16 with imm32SExt16 for addiu.
This keeps same behavior for SDAG and allows for GlobalISel selectImpl
to select 'G_CONSTANT imm' + G_ADD into ADDIu when 32-bit imm satisfies
imm32SExt16 predicate: sign extending 16 low bits of imm is equal to imm.
Differential Revision: https://reviews.llvm.org/D70184
Summary:
When scalarizing PHI nodes we might try to examine/rewrite
InsertElement nodes in predecessors. If those predecessors
are unreachable from entry, then the IR in those blocks could
have unexpected properties resulting in infinite loops in
Scatterer::operator[].
By simply treating values originating from instructions in
unreachable blocks as undef we do not need to analyse them
further.
This fixes PR41723.
Reviewers: bjope
Reviewed By: bjope
Subscribers: bjope, hiraditya, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D70171
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
This adds the `vgetq_lane` and `vsetq_lane` families, to copy between
a scalar and a specified lane of a vector.
One of the new `vgetq_lane` intrinsics returns a `float16_t`, which
causes a compile error if `%clang_cc1` doesn't get the option
`-fallow-half-arguments-and-returns`. The driver passes that option to
cc1 already, but I've had to edit all the explicit cc1 command lines
in the existing MVE intrinsics tests.
A couple of fixes are included for the code I wrote up front in
MveEmitter to support lane-index immediates (and which nothing has
tested until now): the type was wrong (`uint32_t` instead of `int`)
and the range was off by one.
I've also added a method of bypassing the default promotion to `i32`
that is done by the MveEmitter code generation: it's sensible to
promote short scalars like `i16` to `i32` if they're going to be
passed to custom IR intrinsics representing a machine instruction
operating on GPRs, but not if they're going to be passed to standard
IR operations like `insertelement` which expect the exact type.
Reviewers: ostannard, MarkMurrayARM, dmgreen
Reviewed By: dmgreen
Subscribers: kristof.beyls, cfe-commits
Tags: #clang
Differential Revision: https://reviews.llvm.org/D70188
This batch of intrinsics includes lots of things that move vector data
around or change its type without really affecting its value very
much. It includes the `vreinterpretq` family (cast one vector type to
another); `vuninitializedq` (create a vector of a given type with
don't-care contents); and `vcreateq` (make a 128-bit vector out of two
`uint64_t` halves).
These are all implemented using completely standard IR that's already
tested in existing LLVM unit tests, so I've just written a clang test
to check the IR is correct, and left it at that.
I've also added some richer infrastructure to the MveEmitter Tablegen
backend, to make it specify the exact integer type of integer
arguments passed to IR construction functions, and wrap those
arguments in a `static_cast` in the autogenerated C++. That was
necessary to prevent an overloading ambiguity when passing the integer
literal `0` to `IRBuilder::CreateInsertElement`, because otherwise, it
could mean either a null pointer `llvm::Value *` or a zero `uint64_t`.
Reviewers: ostannard, MarkMurrayARM, dmgreen
Subscribers: kristof.beyls, cfe-commits
Tags: #clang
Differential Revision: https://reviews.llvm.org/D70133
Summary:
This doesn't cover decls in diagnostics, which use NamedDecl::getNameForDiagnostic().
(That should also be fixed later I think).
This covers some cases of https://github.com/clangd/clangd/issues/76
(hover, but not outline or sighelp)
Reviewers: hokein
Subscribers: ilya-biryukov, jkorous, arphaman, kadircet, usaxena95, cfe-commits
Tags: #clang
Differential Revision: https://reviews.llvm.org/D70236
It was failing with
llvm/lib/ExecutionEngine/Orc/DebugUtils.cpp:56:10:
error: could not convert ‘Obj’ from ‘std::unique_ptr<llvm::MemoryBuffer>’
to ‘llvm::Expected<std::unique_ptr<llvm::MemoryBuffer> >’
return Obj;
^
The usage of target boolean checks is overly inflexible, since sext
and zext of a compare are equally cheap. The choice is arbitrary, but
using 0/1 to some degree is the choice of lower resistance since
that's what most targets use. This enables a few combines that don't
bother to support ZeroOrNegativeOneBooleanContent.
Adds a DumpObjects utility that can be used to dump JIT'd objects to disk.
Instances of DebugObjects may be used by ObjectTransformLayer as no-op
transforms.
This patch also adds an ObjectTransformLayer to LLJIT and an example of how
to use this utility to dump JIT'd objects in LLJIT.
getFirstNonPHI iterates over all the instructions in a block until it
finds a non-PHI.
Then, the loop starts from the beginning of the block and goes through
all the instructions until it reaches the instruction found by
getFirstNonPHI.
Instead of doing that, just stop when a non-PHI is found.
This reduces the compile-time of a test case discussed in
https://reviews.llvm.org/D47023 by 13x.
Not entirely sure how to come up with a test case for this since it's a
compile time issue that would significantly slow down running the tests.
Differential Revision: https://reviews.llvm.org/D70016
Summary: This is a bug fix for further issues in PR43585.
Reviewers: rnk, RKSimon, craig.topper, andrew.w.kaylor
Subscribers: hiraditya, llvm-commits, annita.zhang
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D70224
Sergey Dmitriev