This is because a subsequent patch will propose obtaining the VRAI from
the advisor, which will enable feature caching for the ML advisor, for
better compile time. Making this change first as it's both innocuous and
keeps the future patch to be reviewed small.
Summary:
Previously this test used a hard-coded triple value in the check lines
wihch failed on other architectures. This patch changes that to accept
any host triple.
We plan to pass the MachineFunction& to APIs that expect it non-const
(for legitimate reasons). The advisor still holds the ref as a const
ref, though, so we keep most of the maintainability value of that.
For "USE, INTRINSIC", search only for intrinsic modules;
for "USE, NON_INTRINSIC", do not recognize intrinsic modules.
Allow modules of both kinds with the same name to be used in
the same source file (but not in the same scoping unit, a
constraint of the standard that is now enforced).
The symbol table's scope tree now has a single instance of
a scope with a new kind, IntrinsicModules, whose children are
the USE'd intrinsic modules (explicit or not). This separate
"top-level" scope is a child of the single global scope and
it allows both intrinsic and non-intrinsic modules of the same
name to exist in the symbol table. Intrinsic modules' scopes'
symbols now have the INTRINSIC attribute set.
The search path directories need to make a distinction between
regular directories and the one(s) that point(s) to intrinsic
modules. I allow for multiple intrinsic module directories in
the second search path, although only one is needed today.
Differential Revision: https://reviews.llvm.org/D118631
When upgrading a loop of load/store to a memcpy, the existing pass does not keep existing aliasing information. This patch allows existing aliasing information to be kept.
Reviewed By: jeroen.dobbelaere
Differential Revision: https://reviews.llvm.org/D108221
The old `__regex_word` aliased the mask for `xdigit`, causing stray
test failures.
The diff may look surprising, as if the previous faulty value had
been set specifically for Windows - but this is due to a restructuring
in 411c630bae. Prior to that, there
were OS specific settings for some OSes, and one fallback used for
the rest (which turns out to not work for Windows).
Differential Revision: https://reviews.llvm.org/D118188
Re-add filtering options with fixes for failed tests. We were not passing the
is_filtered argument in all check generator calls in update_cc_test_checks.py
Enhance the various update_*_test_checks.py tools to allow filtering the tool
output with regular expressions. The --filter option will emit only tool output
lines matching the given regular expression while the --filter-out option will
emit only tools output lines not matching the given regular expression. Filters
are applied in order of appearance on the command line (or in UTC_ARGS) and the
first matching filter terminates the search.
This allows test authors to create more focused tests by removing irrelevant
tool output and checking only the pieces of output necessary to test the desired
functionality.
Differential Revision: https://reviews.llvm.org/D117694
This removes `HasPAUTH` from `AArch64SubTarget`, as it seems to be a
redundant, unused copy of `HasPAuth`.
Differential Revision: https://reviews.llvm.org/D117782
New target SDNodes are added: AArch64ISD::MOPS_MEMSET, etc.
Each intrinsic is translated to one of these in SelectionDAGBuilder
via EmitTargetCodeForMOPS.
A custom lowering routine for INTRINSIC_W_CHAIN is added to handle
llvm.aarch64.mops.memset.tag. This takes a separate path from the common
intrinsics but ultimately ends up in the same EmitMOPS().
This is part 4/4 of a series of patches split from
https://reviews.llvm.org/D117405 to facilitate reviewing.
Patch by Tomas Matheson, Lucas Prates and Son Tuan Vu.
Differential Revision: https://reviews.llvm.org/D117764
This patch introduces a linker wrapper tool that allows us to preprocess
files before they are sent to the linker. This adds a dummy action and
job to the driver stage that builds the linker command as usual and then
replaces the command line with the wrapper tool.
Depends on D116543
Reviewed By: JonChesterfield
Differential Revision: https://reviews.llvm.org/D116544
This patch adds support for embedding the device object files into the
host IR to create a fat binary. Each offloading file will be inserted
into a section with the following naming format
`.llvm.offloading.<triple>.<arch>.<filename>`.
Depends on D116542
Reviewed By: JonChesterfield
Differential Revision: https://reviews.llvm.org/D116543
This patch adds support for a flag `-fembed-offload-binary` to embed a
file as an ELF section in the output by placing it in a global variable.
This can be used to bundle offloading files with the host binary so it
can be accessed by the linker. The section is named using the
`-fembed-offload-section` option.
Depends on D116541
Reviewed By: JonChesterfield
Differential Revision: https://reviews.llvm.org/D116542
This patch introduces the `-fopenmp-new-driver` option which instructs
the compiler to use a new driver scheme for producing offloading code.
In this scheme we create a complete offloading object file and then pass
it as input to the host compilation phase. This will allow us to embed
the object code in the backend phase.
This is the start of a series of commits to rework the OpenMP offloading driver
pipeline. The goal of this is to simplify the steps required for creating an
offloading program. This patch changes the driver's configuration to simply pass
the device file back to the host as an input so it can be embedded as an LLVM IR
global during the backend, then simply passes that object file to the linker.
This driver implementation will currently create the following phases,
```
$ clang input.c -fopenmp -fopenmp-targets=nvptx64 -fopenmp-new-driver -ccc-print-phases
+- 0: input, "input.c", c, (host-openmp)
+- 1: preprocessor, {0}, cpp-output, (host-openmp)
+- 2: compiler, {1}, ir, (host-openmp)
| | +- 3: input, "input.c", c, (device-openmp)
| | +- 4: preprocessor, {3}, cpp-output, (device-openmp)
| |- 5: compiler, {4}, ir, (device-openmp)
| +- 6: offload, "host-openmp (x86_64-unknown-linux-gnu)" {2}, "device-openmp (nvptx64)" {5}, ir
| +- 7: backend, {6}, assembler, (device-openmp)
|- 8: assembler, {7}, object, (device-openmp)
+- 9: offload, "host-openmp (x86_64-unknown-linux-gnu)" {2}, "device-openmp (nvptx64)" {8}, ir
+- 10: backend, {9}, assembler, (host-openmp)
+- 11: assembler, {10}, object, (host-openmp)
12: clang-linker-wrapper, {11}, image, (host-openmp)
```
Which will map to the following bindings
```
# "x86_64-unknown-linux-gnu" - "clang", inputs: ["input.c"], output: "/tmp/input-bae62e.bc"
# "nvptx64" - "clang", inputs: ["input.c", "/tmp/input-bae62e.bc"], output: "/tmp/input-76784e.s"
# "nvptx64" - "NVPTX::Assembler", inputs: ["/tmp/input-76784e.s"], output: "/tmp/input-8f29db.o"
# "x86_64-unknown-linux-gnu" - "clang", inputs: ["/tmp/input-bae62e.bc", "/tmp/input-8f29db.o"], output: "/tmp/input-545450.o"
# "x86_64-unknown-linux-gnu" - "Offload::Linker", inputs: ["/tmp/input-545450.o"], output: "a.out"
```
Reviewed By: JonChesterfield
Differential Revision: https://reviews.llvm.org/D116541
This implements codegen for Armv8.8/9.3 Memory Operations extension (MOPS).
Any memcpy/memset/memmov intrinsics will always be emitted as a series
of three consecutive instructions P, M and E which perform the
operation. The SelectionDAG implementation is split into a separate
patch.
AArch64LegalizerInfo will now consider the following generic opcodes
if +mops is available, instead of legalising by expanding them to
libcalls: G_BZERO, G_MEMCPY_INLINE, G_MEMCPY, G_MEMMOVE, G_MEMSET
The s8 value of memset is legalised to s64 to match the pseudos.
AArch64O0PreLegalizerCombinerInfo will still be able to combine
G_MEMCPY_INLINE even if +mops is present, as it is unclear whether it is
better to generate fixed length copies or MOPS instructions for the
inline code of small or zero-sized memory operations, so we choose to be
conservative for now.
AArch64InstructionSelector will select the above as new pseudo
instructions: AArch64::MOPSMemory{Copy/Move/Set/SetTagging} These are
each expanded to a series of three instructions (e.g. SETP/SETM/SETE)
which must be emitted together during code emission to avoid scheduler
reordering.
This is part 3/4 of a series of patches split from
https://reviews.llvm.org/D117405 to facilitate reviewing.
Patch by Tomas Matheson and Son Tuan Vu
Differential Revision: https://reviews.llvm.org/D117763
According to the specification, MOPS instructions define/use NZCV flags as
part of their semantics (see discussion in
https://reviews.llvm.org/D116157).
More specifically, the specification of the MOPS extension states that
each memcpy/memset/memmov operation will be performed by a series
of three MOPS instructions P, M and E. The P instruction writes to the
NZCV flags, while the others (M and E) reads from the NZCV flags.
This is part 2/4 of a series of patches split from
https://reviews.llvm.org/D117405 to facilitate reviewing.
Differential Revision: https://reviews.llvm.org/D117757
When RECL= is set on OPEN(), ensure that it:
1) enforces a max output record payload size
(not including header+footer or newline), and
2) causes padding of short output records only
for ACCESS='DIRECT'
The previous code was causing some false overrun errors
and applying padding to sequential/stream output files.
Differential Revision: https://reviews.llvm.org/D118630
In ancient Microsoft C runtimes, there might only have been
a nonstandard `_vsnprintf` instead of the standard `vsnprintf`, but
in modern versions (the only ones relevant for libc++), both
are available.
In MinGW configurations built with `__USE_MINGW_ANSI_STDIO=1` (as it
is built in CI), `vsnprintf` provides a more standards compliant
behaviour than what Microsoft's CRT provides, while `_vsnprintf` retains
the Microsoft C runtime specific quirks.
Differential Revision: https://reviews.llvm.org/D118187
Compact Unwind is a subsection, but that was lost in rGff9feeb520a32d076c3095468208ae116c428285
Reviewed By: void
Differential Revision: https://reviews.llvm.org/D118499
This reverts commit 00bf4755e9.
This change broke the emscripten builder (among other things):
https://ci.chromium.org/ui/p/emscripten-releases/builders/try/linux/b8823500584349280721/overview
Sample failure:
```
test_unistd_unlink (test_core.core0) ...
wasm-ld: error: symbol type mismatch: __stdio_write
>>> defined as WASM_SYMBOL_TYPE_FUNCTION in /usr/local/google/home/sbc/dev/wasm/emscripten/cache/sysroot/lib/wasm32-emscripten/libc-debug.a(__stdio_write.o)
>>> defined as WASM_SYMBOL_TYPE_DATA in /usr/local/google/home/sbc/dev/wasm/emscripten/cache/sysroot/lib/wasm32-emscripten/libc-debug.a(stderr.o)
```
This patch changes the error message to instead mention the
documentation page for the debugging options provided by libomptarget
and the bitcode runtimes. Add some extra information to the documentation to
help users more quickly identify debugging resources.
Reviewed By: jdoerfert
Differential Revision: https://reviews.llvm.org/D118626
Reduces the shared memory size used for globalization to 512 bytes from
2048 to reduce the pressure on shared memory. This patch ado adds a
debug mesage to indicate when the shared memory was insufficient.
Reviewed By: jdoerfert
Differential Revision: https://reviews.llvm.org/D118625
See:
https://green.lab.llvm.org/green/view/LLDB/job/lldb-cmake/40636/consoleFull#-39956214149ba4694-19c4-4d7e-bec5-911270d8a58c
```
llvm/lib/Support/Valgrind.cpp:37:63: error: missing '#include <stddef.h>'; 'size_t' must be declared before it is used
void llvm::sys::ValgrindDiscardTranslations(const void *Addr, size_t Len) {
^
/Applications/Xcode.app/Contents/Developer/Toolchains/XcodeDefault.xctoolchain/usr/lib/clang/13.0.0/include/stddef.h:46:23: note: declaration here is not visible
typedef __SIZE_TYPE__ size_t;
^
1 error generated.
```
rdar://88049280
We do support building with a default target unspecified. This fixes
two small build issues that prevented LLVM's unit tests from building
and libSupport from building on Windows.
std::result_of creates problems when building with C++20 because it's
deprecated there.
The solution is to remove it and get return value type for a function
with decltype.
Substitute std::invoke_result for std::result_of is unnecessary because
we don't have std::invoke semantics within the function - we don't work
with pointer-to-member's.
Reviewed by: ldionne, MikeDvorskiy, #libc
Differential Revision: https://reviews.llvm.org/D118457
As discovered in D117817, `std::ranges::input_range<Holder<Incomplete>*[10]>`
hard-errored before this patch. That's because `input_range` requires
`iter_rvalue_reference_t`, which requires `iter_move`, which was
not ADL-proofed.
Add ADL-proofing tests to all the range refinements.
`output_range` and `common_range` shouldn't be affected,
and all the others subsume `input_range` anyway, but we might as
well be thorough.
Differential Revision: https://reviews.llvm.org/D118213
Added support for alternate ops vectorization of the cmp instructions.
It allows to vectorize either cmp instructions with same/swapped
predicate but different (swapped) operands kinds or cmp instructions
with different predicates and compatible operands kinds.
Differential Revision: https://reviews.llvm.org/D115955
We can have a scenario where multiple CUs share an abbrev table.
We modify or don't modify one CU, which leads to other CUs having invalid abbrev section.
Example that caused it.
All of CUs shared the same abbrev table. First CU just had compile_unit and sub_program.
It was not modified. Next CU had DW_TAG_lexical_block with
DW_AT_low_pc/DW_AT_high_pc converted to DW_AT_low_pc/DW_AT_ranges.
We used unmodified abbrev section for first and subsequent CUs.
So when parsing subsequent CUs debug info was corrupted.
In this patch we will now duplicate all sections that are modified and are different.
This also means that if .debug_types is present and it shares Abbrev table, and
they usually are, we now can have two Abbrev tables. One for CU that was modified,
and unmodified one for TU.
Reviewed By: maksfb
Differential Revision: https://reviews.llvm.org/D118517
- Remove the `{Op,Attr,Type}Trait` TableGen classes and replace with `Trait`
- Rename `OpTraitList` to `TraitList` and use it in a few places
The bulk of this change is a mechanical s/OpTrait/Trait/ throughout the codebase.
Reviewed By: rriddle, jpienaar, herhut
Differential Revision: https://reviews.llvm.org/D118543
Rosetta crashlogs can have their own thread register state. Unlike the
other registers which ware directly listed under "threadState", the
Rosetta registers are nested under their own key in the JSON, as
illustrated below:
{
"threadState":
{
"rosetta":
{
"tmp2":
{
"value": 4935057216
},
"tmp1":
{
"value": 4365863188
},
"tmp0":
{
"value": 18446744073709551615
}
}
}
}
The optimization added in D118139 causes a crash on the added test case
while trying to zero extend an vector of floats.
Fix the crash by bailing out for floating point operands.
Reviewed By: DavidTruby
Differential Revision: https://reviews.llvm.org/D118615
Mircea Trofin