This is a major change on how we represent nested name qualifications in
the AST.
* The nested name specifier itself and how it's stored is changed. The
prefixes for types are handled within the type hierarchy, which makes
canonicalization for them super cheap, no memory allocation required.
Also translating a type into nested name specifier form becomes a no-op.
An identifier is stored as a DependentNameType. The nested name
specifier gains a lightweight handle class, to be used instead of
passing around pointers, which is similar to what is implemented for
TemplateName. There is still one free bit available, and this handle can
be used within a PointerUnion and PointerIntPair, which should keep
bit-packing aficionados happy.
* The ElaboratedType node is removed, all type nodes in which it could
previously apply to can now store the elaborated keyword and name
qualifier, tail allocating when present.
* TagTypes can now point to the exact declaration found when producing
these, as opposed to the previous situation of there only existing one
TagType per entity. This increases the amount of type sugar retained,
and can have several applications, for example in tracking module
ownership, and other tools which care about source file origins, such as
IWYU. These TagTypes are lazily allocated, in order to limit the
increase in AST size.
This patch offers a great performance benefit.
It greatly improves compilation time for
[stdexec](https://github.com/NVIDIA/stdexec). For one datapoint, for
`test_on2.cpp` in that project, which is the slowest compiling test,
this patch improves `-c` compilation time by about 7.2%, with the
`-fsyntax-only` improvement being at ~12%.
This has great results on compile-time-tracker as well:
This patch also further enables other optimziations in the future, and
will reduce the performance impact of template specialization resugaring
when that lands.
It has some other miscelaneous drive-by fixes.
About the review: Yes the patch is huge, sorry about that. Part of the
reason is that I started by the nested name specifier part, before the
ElaboratedType part, but that had a huge performance downside, as
ElaboratedType is a big performance hog. I didn't have the steam to go
back and change the patch after the fact.
There is also a lot of internal API changes, and it made sense to remove
ElaboratedType in one go, versus removing it from one type at a time, as
that would present much more churn to the users. Also, the nested name
specifier having a different API avoids missing changes related to how
prefixes work now, which could make existing code compile but not work.
How to review: The important changes are all in
`clang/include/clang/AST` and `clang/lib/AST`, with also important
changes in `clang/lib/Sema/TreeTransform.h`.
The rest and bulk of the changes are mostly consequences of the changes
in API.
PS: TagType::getDecl is renamed to `getOriginalDecl` in this patch, just
for easier to rebasing. I plan to rename it back after this lands.
Fixes#136624
Fixes https://github.com/llvm/llvm-project/issues/43179
Fixes https://github.com/llvm/llvm-project/issues/68670
Fixes https://github.com/llvm/llvm-project/issues/92757
Currently, the library-internal feature test macros are only defined if
the feature is not available, and always have the prefix
`_LIBCPP_HAS_NO_`. This patch changes that, so that they are always
defined and have the prefix `_LIBCPP_HAS_` instead. This changes the
canonical use of these macros to `#if _LIBCPP_HAS_FEATURE`, which means
that using an undefined macro (e.g. due to a missing include) is
diagnosed now. While this is rather unlikely currently, a similar change
in `<__configuration/availability.h>` caught a few bugs. This also
improves readability, since it removes the double-negation of `#ifndef
_LIBCPP_HAS_NO_FEATURE`.
The current patch only touches the macros defined in `<__config>`. If
people are happy with this approach, I'll make a follow-up PR to also
change the macros defined in `<__config_site>`.
This patch adds a large number of missing includes in the libc++ headers
and the test suite. Those were found as part of the effort to move
towards a mostly monolithic top-level std module.
This takes 1m40s to run when testing picolib on qemu. This isn't the end
of the world but that's on an AArch64 server. So if someone felt the
need to mark this unsupported in the first place, it's likely much
slower on average hardware.
This patch removes many annotations that are not relevant anymore since
we don't support or test back-deploying to macOS < 10.13. It also cleans
up raw usage of target triples to identify versions of dylibs shipped on
prior versions of macOS, and uses the target-agnostic Lit features
instead. Finally, it reorders both the Lit backdeployment features and
the corresponding availability macros in the library in a way that makes
more sense, and reformulates the Lit backdeployment features in terms of
when a version of LLVM was introduced instead of encoding the system
versions on which it hasn't been introduced yet. Although one can be
derived from the other, encoding the negative form is extremely
error-prone.
Fixes#80901
We were not making any distinction between e.g. the "Apple-flavored"
libc++ built from trunk and the system-provided standard library on
Apple platforms. For example, any test that would be XFAILed on a
back-deployment target would unexpectedly pass when run on that
deployment target against the tip of trunk Apple-flavored libc++. In
reality, that test would be expected to pass because we're running
against the latest libc++, even if it is Apple-flavored.
To solve this issue, we introduce a new feature that describes whether
the Standard Library in use is the one provided by the system by
default, and that notion is different from the underlying standard
library flavor. We also refactor the existing Lit features to make a
distinction between availability markup and the library we're running
against at runtime, which otherwise limit the flexibility of what we can
express in the test suite. Finally, we refactor some of the
back-deployment versions that were incorrect (such as thinking that LLVM
10 was introduced in macOS 11, when in reality macOS 11 was synced with
LLVM 11).
Fixes#82107
Adjust some of the [rand.dist] critical values that are too strict
- Most critical values are determined empirically by running each test
51
times with a different PRNG seed and finding the smallest symmetric
interval
around the median that contains 90% of the sample means, variances, etc.
- For the Kolmogorov-Smirnov tests, the alpha=0.1 critical value for
large N
is 1.224/sqrt(N).
- For normally distributed variates, the sample kurtosis is distributed
as
Normal(0, 24/N). For N=1e5, this gives a 90% confidence interval of
0+/-0.0255. For Binomial(40, 0.25), which is approximately normal, the
kurtosis is -0.0167, so the relative 90% CI is large, on the order of
0.0255/0.0167 = 153%. In most cases the distribution of the sample
kurtosis
isn't known analytically, but similarly large relative tolerances can be
expected if the kurtosis is near zero.
This PR is a followup to #81080.
This PR makes two major changes to how the LCG operation is computed:
The first is that I added an additional case where `ax + c` might
overflow the intermediate variable, but `ax` by itself won't. In this
case, it's much better to use `(ax mod m) + c mod m` than the previous
behavior of falling back to Schrage's algorithm. The addition modulo is
done in the same way as when using Schrage's algorithm (i.e. `x += c -
(x >= m - c)*m`), but the multiplication modulo is calculated directly,
which is faster.
The second is that I added handling for the case where the `ax`
intermediate might overflow, but Schrage's algorithm doesn't apply (i.e.
r > q). In this case, the only real option is to increase the precision
of the intermediate values. The good news is that - for `x`, `a`, and
`c` being n-bit values - `ax + c` will never overflow a 2n-bit
intermediary, meaning this promotion can only happen once, and will
always be able to use the simplest implementation. This is already the
case for 16-bit LCGs, as libcxx chooses to compute them with 32-bit
intermediate values. For 32-bit LCGs, I simply added code similar to the
16-bit case to use the existing 64-bit implementations. Lastly, for
64-bit LCGs, I wrote a case that calculates it using `unsigned __int128`
if it is available to use.
While this implementation covers a *lot* of the missing cases from
#81080, this still won't compile **every** possible
`linear_congruential_engine`. Specifically, if `a`, `c`, and `m` are
chosen such that it needs 128-bit integers, but the platform doesn't
support `__int128` (eg. 32-bit x86), then it will fail to compile.
However, this is a fairly rare case to see actually used, and libcxx
would be in good company with this, as [libstdc++ also fails to compile
under these
circumstances](https://gcc.gnu.org/bugzilla/show_bug.cgi?id=87744).
Fixing **this** gap would require even **more** work of further
complexity, so that would probably be best handled by a different PR
(I'll put more details on what that PR would entail in a comment).
This fixes two major mistakes in the implementation of
`linear_congruential_engine` that allowed it to produce incorrect
output. Specifically, these mistakes are in `__lce_alg_picker`, which is
used to determine whether Schrage's algorithm is valid and needed.
The first mistake is in the definition of `_OverflowOK`. The code
comment and the description of [D65041](https://reviews.llvm.org/D65041)
both indicate that it's supposed to be true iff `m` is a power of two.
However, the definition used does not work out to that, and instead is
true whenever `m` is even. This could result in
`linear_congruential_engine` using an invalid implementation, as it
would incorrectly assume that any integer overflow can't change the
result. I changed the implementation to one that accurately checks if
`m` is a power of two. Technically, this implementation has an edge case
where it considers `0` to be a power of two, but in this case this is
actually accurate behavior, as `m = 0` indicates a modulus of 2^w where
w is the size of `result_type` in bits, which *is* a power of two.
The second mistake is in the static assert. The original static assert
erroneously included an unnecessary `a != 0 || m != 0`. Combined with
the `|| !_MightOverflow`, this actually resulted in the static assert
being impossible to fail. Applying De Morgan's law and expanding
`_MightOverflow` gives that the only way this static assert can be
triggered is if `a == 0 && m == 0 && a != 0 && m != 0 && ...`, which
clearly cannot be true. I simply removed the explicit checks against `a`
and `m`, as the intended checks are already included in `_MightOverflow`
and `_SchrageOK`, and their inclusion doesn't provide any obvious
semantic benefit.
This should fix all the current instances where
`linear_congruential_engine` uses an invalid implementation. This
technically isn't a complete implementation, though, since the static
assert will cause some instantiations of `linear_congruential_engine`
not disallowed by the standard from compiling. However, this should
still be an improvement, as all compiling instantiations of
`linear_congruential_engine` should use a valid implementation. Fixing
the cases where the static assert triggers will require adding
additional implementations, some of which will be fairly non-trivial, so
I'd rather leave those for another PR so they don't hold up these more
important fixes.
Fixes#33554
Found while running libc++'s tests with MSVC's STL. This fixes 3 kinds of warnings:
- Add void-casts to fix `-Wunused-variable` warnings.
- Avoid sign/truncation warnings in `ConvertibleToIntegral.h`.
- Add `TEST_STD_AT_LEAST_23_OR_RUNTIME_EVALUATED` to avoid mixing preprocessor
and runtime tests.
- Cleanup: Add `TEST_STD_AT_LEAST_20_OR_RUNTIME_EVALUATED` for
consistency.
While trying to work around MSAN/TSAN build timeouts, we enabled
optimizations on some tests. This caused GCC to start complaining that
some values may be uninitialized. I believe GCC is wrong, but more
investigation is needed.
The values are initialized when the variable `__g` is either < 0 or >=
0. Which only leaves out NaN I believe, which is likely well into UB
land anyway.
However, more investigation needed.
This attempts to fix flakes on the bots where the modified test times
out while running under sanitizers.
Turning on the optimizer for just this test appears to mostly fix the
issue.
This patch actually runs the tests for picolibc behind an emulator,
removing a few workarounds and increasing coverage.
Differential Revision: https://reviews.llvm.org/D155521
I made sure they all had some expected-error output in them. Many of
these tests would be better implemented as a positive test using SFINAE,
but this is beyond the scope of this patch.
Differential Revision: https://reviews.llvm.org/D153980
Some tests in our test suite are unbelievably slow on GCC due to the
use of the always_inline attribute. See [1] for more details.
This patch introduces the GCC-ALWAYS_INLINE-FIXME lit feature to
disable tests that are plagued by that issue. At the same time, it
moves several existing tests from ad-hoc `UNSUPPORTED: gcc-12` markup
to the new GCC-ALWAYS_INLINE-FIXME feature, and marks the slowest tests
reported by the CI as `UNSUPPORTED: GCC-ALWAYS_INLINE-FIXME`.
[1]: https://discourse.llvm.org/t/rfc-stop-supporting-extern-instantiations-with-gcc/71277/1
Differential Revision: https://reviews.llvm.org/D152736
We changed the `abort` calls when trying to throw exceptions in `-fno-exceptions` mode to `__verbose_abort` calls, which removes the dependency in most files.
Reviewed By: ldionne, #libc
Spies: dim, emaste, mikhail.ramalho, smeenai, libcxx-commits
Differential Revision: https://reviews.llvm.org/D146076
The use_system_cxx_lib Lit feature was only used for back-deployment
testing. However, one immense hole in that setup was that we didn't
have a proper way to test Apple's own libc++ outside of back-deployment,
which was embodied by the fact that we needed to define _LIBCPP_DISABLE_AVAILABILITY
when testing (see change in libcxx/utils/libcxx/test/params.py).
This led to the apple-system testing configuration not checking for
availability markup, which is obviously quite bad since the library
we ship actually has availability markup.
Using stdlib=<VENDOR>-libc++ instead to encode back-deployment restrictions
on tests is simpler and it makes it possible to naturally support tests
such as availability markup checking even in the tip-of-trunk Apple-libc++
configuration.
Differential Revision: https://reviews.llvm.org/D146366
Instead of writing something like `XFAIL: use_system_cxx_lib && target=...`
to XFAIL back-deployment tests, introduce named Lit features like
`availability-shared_mutex-missing` to represent those. This makes the
XFAIL annotations leaner, and solves the problem of XFAIL comments
potentially getting out of sync. This would also make it easier for
another vendor to add their own annotations to the test suite by simply
changing how the feature is defined for their OS releases, instead
of having to modify hundreds of tests to add repetitive annotations.
This doesn't touch *all* annotations -- only annotations that were widely
duplicated are given named features (e.g. when filesystem or shared_mutex
were introduced). I still think it probably doesn't make sense to have a
named feature for every single fix we make to the dylib.
This is in essence a revert of 2659663, but since then the test suite
has changed significantly. Back when I did 2659663, the configuration
files we have for the test suite right now were being bootstrapped and
it wasn't clear how to provide these features for back-deployment in
that context. Since then, we have a streamlined way of defining these
features in `features.py` and that doesn't impact the ability for a
configuration file to stay minimal.
The original motivation for this change was that I am about to propose
a change that would touch essentially all XFAIL annotations for back-deployment
in the test suite, and this greatly reduces the number of lines changed
by that upcoming change, in addition to making the test suite generally
better.
Differential Revision: https://reviews.llvm.org/D146359
This has been done using the following command
find libcxx/test -type f -exec perl -pi -e 's|^([^/]+?)((?<!::)size_t)|\1std::\2|' \{} \;
And manually removed some false positives in std/depr/depr.c.headers.
The `std` module doesn't export `::size_t`, this is a preparation for that module.
Reviewed By: ldionne, #libc, EricWF, philnik
Differential Revision: https://reviews.llvm.org/D146088
We pretty consistently don't define those cause they are not needed,
and it removes the potential pitfall to think that these tests are
being run. This doesn't touch .compile.fail.cpp tests since those
should be replaced by .verify.cpp tests anyway, and there would be
a lot to fix up.
As a fly-by, I also fixed a bit of formatting, removed a few unused
includes and made some very minor, clearly NFC refactorings such as
in allocator.traits/allocator.traits.members/allocate.verify.cpp where
the old test basically made no sense the way it was written.
Differential Revision: https://reviews.llvm.org/D146236
This has been done using the following command
find libcxx/test -type f -exec perl -pi -e 's|^([^/]+?)((?<!::)(?<!::u)u?int(_[a-z]+)?[0-9]{1,2}_t)|\1std::\2|' \{} \;
And manually removed some false positives in std/depr/depr.c.headers.
Reviewed By: ldionne, #libc
Differential Revision: https://reviews.llvm.org/D145880
It seems these includes are still provided by the sub headers, so it only
removes the duplicates.
There is no change in the list of includes, but the change affects the
modular build. By not having the includes in the top-level header the
module map has changed. This uncovers missing includes in the tests
and missing exports in the module map. This causes the huge amount of
changes in the patch.
Reviewed By: #libc, ldionne
Differential Revision: https://reviews.llvm.org/D133252
frederick-vs-ja noticed that https://github.com/microsoft/STL/pull/2976#issuecomment-1201926893
while we are working on updating LLVM submodule for MS STL:
[...]\std\numerics\rand\rand.dist\rand.dist.samp\rand.dist.samp.discrete\eval.pass.cpp(33): error C2220: the following warning is treated as an error
[...]\std\numerics\rand\rand.dist\rand.dist.samp\rand.dist.samp.discrete\eval.pass.cpp(287): note: see reference to function template instantiation 'void tests<__int64>(void)' being compiled
[...]\std\numerics\rand\rand.dist\rand.dist.samp\rand.dist.samp.discrete\eval.pass.cpp(33): warning C4244: 'argument': conversion from '__int64' to 'const unsigned int', possible loss of data
Differential Revision: https://reviews.llvm.org/D130963
This patch rewords the static assert diagnostic output. Failing a
_Static_assert in C should not report that static_assert failed. This
changes the wording to be more like GCC and uses "static assertion"
when possible instead of hard coding the name. This also changes some
instances of 'static_assert' to instead be based on the token in the
source code.
Differential Revision: https://reviews.llvm.org/D129048
In D125283, we ensured that integer distributions would not compile when
used with arbitrary unsupported types. This effectively enforced what
the Standard mentions here: http://eel.is/c++draft/rand#req.genl-1.5.
However, this also had the effect of breaking some users that were
using integer distributions with unsupported types like int8_t. Since we
already support using __int128_t in those distributions, it is reasonable
to also support smaller types like int8_t and its unsigned variant. This
commit implements that, adds tests and documents the extension. Note that
we voluntarily don't add support for instantiating these distributions
with bool and char, since those are not integer types. However, it is
trivial to replace uses of these random distributions on char using int8_t.
It is also interesting to note that in the process of adding tests
for smaller types, I discovered that our distributions sometimes don't
provide as faithful a distribution when instantiated with smaller types,
so I had to relax a couple of tests. In particular, we do a really bad
job at implementing the negative binomial, geometric and poisson distributions
for small types. I think this all boils down to the algorithm we use in
std::poisson_distribution, however I am running out of time to investigate
that and changing the algorithm would be an ABI break (which might be
reasonable).
As part of this patch, I also added a mitigation for a very likely
integer overflow bug we were hitting in our tests in negative_binomial_distribution.
I also filed http://llvm.org/PR56656 to track fixing the problematic
distributions with int8_t and uint8_t.
Supersedes D125283.
Differential Revision: https://reviews.llvm.org/D126823
Looks like we again are going to have problems with libcxx tests that
are overly specific in their dependency on clang's diagnostics.
This reverts commit 6542cb55a3eb115b1c3592514590a19987ffc498.
This patch is basically the rewording of the static assert statement's
output(error) on screen after failing. Failing a _Static_assert in C
should not report that static_assert failed. It’d probably be better to
reword the diagnostic to be more like GCC and say “static assertion”
failed in both C and C++.
consider a c file having code
_Static_assert(0, "oh no!");
In clang the output is like:
<source>:1:1: error: static_assert failed: oh no!
_Static_assert(0, "oh no!");
^ ~
1 error generated.
Compiler returned: 1
Thus here the "static_assert" is not much good, it will be better to
reword it to the "static assertion failed" to more generic. as the gcc
prints as:
<source>:1:1: error: static assertion failed: "oh no!"
1 | _Static_assert(0, "oh no!");
| ^~~~~~~~~~~~~~
Compiler returned: 1
The above can also be seen here. This patch is about rewording
the static_assert to static assertion.
Differential Revision: https://reviews.llvm.org/D129048
Display 'static_assert failed: message' instead of
'static_assert failed "message"' to be consistent
with other implementations and be slightly more
readable.
Reviewed By: #libc, aaron.ballman, philnik, Mordante
Differential Revision: https://reviews.llvm.org/D128844
Since those features are general properties of the environment, it makes
sense to use them from libc++abi too, and so the name libcpp-has-no-xxx
doesn't make sense.
Differential Revision: https://reviews.llvm.org/D126482
All supported compilers that support C++20 now support concepts. So, remove
`_LIB_LIBCPP_HAS_NO_CONCEPTS` in favor of `_LIBCPP_STD_VER > 17`. Similarly in
the tests, remove `// UNSUPPORTED: libcpp-no-concepts`.
Differential Revision: https://reviews.llvm.org/D121528
This should fix the regressions detected in D117992.
This lands before D117992 to avoid breaking main.
Reviewed By: #libc, ldionne
Differential Revision: https://reviews.llvm.org/D118056
https://cplusplus.github.io/LWG/issue3422
Also add a static_assert to check the "Mandates:" on the
iterator-pair constructor. Oddly, the `InputIterator` parameter
itself is merely preconditioned, not constrained, to satisfy the
input iterator requirements.
Also drive-by rename `init` to `__init`.
Differential Revision: https://reviews.llvm.org/D117962
... it's easier to suppress warnings internally, where we can detect the compiler.
* Rename `TEST_COMPILER_C1XX` to `TEST_COMPILER_MSVC`
* Rename all `TEST_WORKAROUND_C1XX_<meow>` to `TEST_WORKAROUND_MSVC_<meow>`
Differential Revision: https://reviews.llvm.org/D117422
