This reverts commit 02c6845c762dfd0a19d4a2f997990e160f392dae,
reapplying bfabc958c7c0d7ddc15f23383d9da836e8c6093f.
The patch was reverted due to the test failing on MacOS and Windows
where type units aren't supported. This is addressed by limiting type
unit flag/test coverage to Linux.
Complete C++ type information can be quite expensive - and there's
limited value in representing every member function, even those that
can't be called (we don't do similarly for every non-member function
anyway). So add a flag to opt out of this behavior for experimenting
with this more terse behavior.
I think Sony already does this by default, so perhaps with a change to
the defaults, Sony can migrate to this rather than a downstream patch.
This breaks current debuggers in some expected ways - but those
breakages are visible without this feature too. Consider member function
template instantiations - they can't be consistently enumerated in every
translation unit:
a.h:
```
struct t1 {
template <int i>
static int f1() {
return i;
}
};
namespace ns {
template <int i>
int f1() {
return i;
}
} // namespace ns
```
a.cpp:
```
void f1() {
t1::f1<0>();
ns::f1<0>();
}
```
b.cpp:
```
void f1();
int main() {
f1();
t1::f1<1>();
ns::f1<1>();
}
```
```
(gdb) p ns::f1<0>()
$1 = 0
(gdb) p ns::f1<1>()
$2 = 1
(gdb) p t1::f1<0>()
Couldn't find method t1::f1<0>
(gdb) p t1::f1<1>()
$3 = 1
(gdb) s
f1 () at a.cpp:3
3 t1::f1<0>();
(gdb) p t1::f1<0>()
$4 = 0
(gdb) p t1::f1<1>()
Couldn't find method t1::f1<1>
(gdb)
```
(other similar non-canonical features are implicit special members
(copy/move ctor/assignment operator, default ctor) and nested types (eg:
pimpl idiom, where the nested type is declared-but-not-defined in one
TU, and defined in another TU))
lldb can't parse the template expressions above, so I'm not sure how to
test it there, but I'd guess it has similar problems. (
https://stackoverflow.com/questions/64602475/how-to-print-value-returned-by-template-member-function-in-gdb-lldb-debugging
so... I guess that's just totally not supported in lldb, how
unfortunate. And implicit special members are instantiated implicitly by
lldb, so missing those doesn't tickle the same issue)
Some very rudimentary numbers for a clang debug build:
.debug_info section size:
-g: 476MiB
-g -fdebug-types-section: 357MiB
-g -gomit-unreferenced-members: 340MiB
Though it also means a major reduction in .debug_str size,
-fdebug-types-section doesn't reduce string usage (so the first two
examples have the same .debug_str size, 247MiB), down to 175MiB.
So for total clang binary size (I don't have a quick "debug section size
reduction" on-hand): 1.45 (no type units) GiB -> 1.34 -> 1.22, so it
saves about 120MiB of binary size.
Original Differential Revision: https://reviews.llvm.org/D152017
These changes slighly modify the output of the unittests so that they better
match GTest, so that utilities that parse the expected output from GTest (such
as Android's unit test harness) can read the output from our unit tests.
This allows our unit tests to be run on Android devices.
Add very primitive command line parsing to:
- support --gtest_color=no to disable printing terminal colors.
- recognize --gtest_print_time and print the test time in milliseconds.
- most of our unit tests run on the order of microseconds, so its useful to
preserve the existing behavior. But upsteram GTest ONLY prints time tests
in milliseconds, and Android's atest expects to be able to parse exactly
that. Atest always passes --gtest_print_time. The word `took` is removed as
that also differs from upstream GTest, tripping up parsers.
- ignore other --gtest_* flags
Do so so that atest can parse the output correctly.
Print the test number count before
each run, so that atest can parse this value correctly.
Link: https://android-review.googlesource.com/c/platform/external/llvm-libc/+/3107252
Link: https://google.github.io/googletest/advanced.html#colored-terminal-output
Link: https://google.github.io/googletest/advanced.html#suppressing-the-elapsed-time
This patch fixes:
flang/lib/Semantics/check-do-forall.cpp:731:9: error: default label
in switch which covers all enumeration values
[-Werror,-Wcovered-switch-default]
Problem: the printer used to ignore all but the first declarator for
unbraced language linkage declarators. Furthemore, that one would be
printed without the final semicolon.
Solution: when there is more than one declarator, we print them in a
braced `extern <lang>` block. If the original declaration was unbraced
and there is one or less declarator, we omit the braces, but add the
semicolon.
**N.B.** We are printing braces which were, in some cases, absent from
the original CST. If that's an issue, I'll work on it. See the tests for
the examples.
The `FirstClauses` is a vector of pointer-bool pairs, and the pointer
part of the pair is never used. Replace the vector with std::bitset, and
rename it to `SeenClauses` to make the purpose of it a bit clearer.
So far, these tests have been disabled in mingw build configurations
(built as asan-dynamic), but these were enabled in
246234ac70faa1e3281a2bb83dfc4dd206a7d59c, exposing the issue.
(That commit is currently reverted, but will probably be relanded in
some form soon.)
- In `Header::readFromBuffer`, read the buffer in the forward direction by using `readNext`.
- When compute the header size, spell out the constant.
With the changes above, we can remove `offsetOf` in InstrProf.cpp
---------
Co-authored-by: Kazu Hirata <kazu@google.com>
- added unittests for the raw_fd_stream output case.
- the `BitstreamWriter` ctor was confusing, the relationship between the buffer and the file stream wasn't clear and in fact there was a potential bug in `BitcodeWriter` in the mach-o case, because that code assumed in-buffer only serialization. The incremental flushing behavior of flushing at end of block boundaries was an implementation detail that meant serializers not using blocks (for example) would need to know to check the buffer and flush. The bug was latent - in the sense that, today, because the stream being passed was not a `raw_fd_stream`, incremental buffering never kicked in.
The new design moves the responsibility of flushing to the `BitstreamWriter`, and makes it work with any `raw_ostream` (but incrementally flush only in the `raw_fd_stream` case). If the `raw_ostream` is over a buffer - i.e. a `raw_svector_stream` - then it's equivalent to today's buffer case. For all other `raw_ostream` cases, buffering is an implementation detail. In all cases, the buffer is flushed (well, in the buffer case, that's a moot statement).
This simplifies the state and state transitions the user has to track: you have a raw_ostream -> BitstreamWrite in it -> destroy the writer => the bitstream is completely written in your raw_ostream. The "buffer" case and the "raw_fd_stream" case become optimizations rather than imposing state transition concerns to the user.
The function was meant to find the Developer/ dir, but it found a
Developer directory nested deep inside the top-level Developer dir.
The new implementation rejects everything in Xcode.app/Developer in
broad strokes.
rdar://128571037
Derived from #92480. This PR supports parsing of the DO CONCURRENT
REDUCE clause in Fortran 2023. Following the style of the OpenMP parser
in MLIR, the front end accepts both arbitrary operations and procedures
for the REDUCE clause. But later Semantics can notify type errors and
resolve procedure names.
New pull request for https://github.com/llvm/llvm-project/issues/89443
The previous PR was reverted after breaking fullbuild due to a missing
struct declaration, which I forgot to commit.
Reverts revert and adds the missing pthread_rwlockattr_getkind_np /
pthread_rwlockattr_setkind_np functions and tests respecitvely.
This patch adds a new special type of VPBasicBlock that wraps an
existing IR basic block. Recipes of the block get added before the
terminator of the wrapped IR basic block. Making it a subclass of
VPBasicBlock avoids duplicating various APIs to manage recipes in a
block, as well as makes sure the traversals filtering VPBasicBlocks
automatically apply as well.
Initially VPIRBasicBlock are only used for the pre-preheader (wrapping
the original preheader of the scalar loop).
As follow-up, this will be used to move more parts of the skeleton
inside VPlan, starting with the branch and condition in the middle
block.
Separated out of https://github.com/llvm/llvm-project/pull/92651
PR: https://github.com/llvm/llvm-project/pull/93398
Re-Apply: 246234ac70faa1e3281a2bb83dfc4dd206a7d59c
Originally #81677
The static asan runtime on windows had various buggy hacks to ensure loaded dlls got the executable's copy of asan, these never worked all that well, so we have eliminated the static runtime altogether and made the dynamic runtime work for applications linking any flavor of the CRT.
Among other things this allows non-asan-instrumented applications to load asan-instrumented dlls that link against the static CRT.
Co-authored-by: Amy Wishnousky <amyw@microsoft.com>
Follow-up to a previous simplification
2473b1af085ad54e89666cedf684fdf10a84f058.
The xor difference between a SHT_NOTE and a read-only SHT_PROGBITS
(previously >=NOT_SPECIAL) should be smaller than RF_EXEC. Otherwise,
for the following section layout, `findOrphanPos` would place .text
before note.
```
// simplified from linkerscript/custom-section-type.s
non orphans:
progbits 0x8060c00 NOT_SPECIAL
note 0x8040003
orphan:
.text 0x8061000 NOT_SPECIAL
```
rw-text.lds in orphan.s (added by
73e07e924470ebab76a634e41fadf425a859e0ea) demonstrates a similar case.
The new behavior is more similar to GNU ld.
#93763 fixed BOLT's brittle reliance on the previous .interp behavior.
Currently, lld assigns RF_NOT_SPECIAL so that orphan .interp and
SHT_NOTE are always before other sections. GNU ld doesn't do so. The
next change will remove RF_NOT_SPECIAL.
This adds new SB API calls and classes to allow a user of the SB API to obtain an address range from SBFunction and SBBlock. This is a second attempt to land the reverted PR #92014.
Try to avoid referring to specific operand names, except in the special
case. The special case for hasNamedOperand(Op32, sdst) seems to have
been dead code.
Then two tests rely on .interp being the first section.
llvm-bolt would crash if lld places .interp after .got
(f639b57f7993cadb82ee9c36f04703ae4430ed85).
For best portability, when a linker scripts specifies a SECTIONS
command, the first section for each PT_LOAD segment should be specified
with a MAXPAGESIZE alignment. Otherwise, linkers have freedom to decide
how to place orphan sections, which might break intention.
To keep the test filenames consistent, this patch:
* removes "test-" from file names (there used to be a mix of
"test-feature-1.mlir" and "feature-2.mlir"),
* replaces "_" with "-" (there used to be a mix of "feature-3.mlir"
and "feature_4.mlir").
Only files under test/Integration/Dialect/Vector/CPU are updated.
This fixes
https://github.com/llvm/llvm-project/pull/92328#issuecomment-2139339444
by not differentiating `DW_TAG_class_type` and `DW_TAG_structure_type`
in `UniqueDWARFASTTypeList`, because it's possible that DIE for a type
is `DW_TAG_class_type` in one CU but is `DW_TAG_structure_type` in a
different CU.
---------
Co-authored-by: Michael Buch <michaelbuch12@gmail.com>
Although the patch that enables debug records by default has been
temporarily reverted, it will (eventually) be reverted and everyone's
code will be subjected to the new debug info format. Although this is
broadly a good thing, it is important that the documentation has enough
information to guide users through the update; this patch adds what
should hopefully be enough detail for most users to either find the
answers, or find out how to find those answers.
This patch introduces `SemaAMDGPU`, `SemaARM`, `SemaBPF`, `SemaHexagon`,
`SemaLoongArch`, `SemaMIPS`, `SemaNVPTX`, `SemaPPC`, `SemaSystemZ`,
`SemaWasm`. This continues previous efforts to split Sema up. Additional
context can be found in #84184 and #92682.
I decided to bundle target-specific components together because of their
low impact on `Sema`. That said, their impact on `SemaChecking.cpp` is
far from low, and I consider it a success.
Somewhat accidentally, I also moved Wasm- and AMDGPU-specific function
from `SemaDeclAttr.cpp`, because they were exposed in `Sema`. That went
well, and I consider it a success, too. I'd like to move the rest of
static target-specific functions out of `SemaDeclAttr.cpp` like we're
doing with built-ins in `SemaChecking.cpp` .
Recently we have disabled this test for Windows host and Linux target.
Now we faced the same issue #92419 in case of Linux x86_64 host and
Linux Aarch64 target.
The TestGdbRemoteLibrariesSvr4Support test failed in case of Linux
x86_64 host and Linux Aarch64 target. Installing libraries to the remote
target is not enough. This test actively uses self.getBuildDir() and
os.path.realpath() which does not work in case of the remote target. So,
disable this test for remote target now.
Handle lowering of non optional inquired argument in custom lowering.
Also fix an issue in the lowering of associated optional argument where
a box was emboxed again which led to weird result.
Better design to put semantics on the ops, and in this case the ntt/intt
op can lower in multiple ways depending on the polynomial ring modulus
(it can need an nth root of unity for cyclic polymul -> ntt, or a 2nth
root for negacyclic polymul -> ntt)
---------
Co-authored-by: Jeremy Kun <j2kun@users.noreply.github.com>
