[libc++] Define all CMake configuration features in the same location #115361
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
llvmbot
added
the
libc++
|
@llvm/pr-subscribers-libcxx Author: Louis Dionne (ldionne) ChangesThis moves the configuration of the CMake features to turn off RTTI, exceptions and friends to the beginning of the CMake file, where we configure other optional parts of the library. This fixes an important bug where we would disable the benchmarks because these options were not defined yet, leading to the build thinking they were defined to OFF. Full diff: https://github.com/llvm/llvm-project/pull/115361.diff 1 Files Affected:
diff --git a/libcxx/CMakeLists.txt b/libcxx/CMakeLists.txt
index d699135774ee0b..abe12c2805a7cf 100644
--- a/libcxx/CMakeLists.txt
+++ b/libcxx/CMakeLists.txt
@@ -47,6 +47,10 @@ include(HandleCompilerRT)
# Basic options ---------------------------------------------------------------
option(LIBCXX_ENABLE_SHARED "Build libc++ as a shared library." ON)
option(LIBCXX_ENABLE_STATIC "Build libc++ as a static library." ON)
+option(LIBCXX_ENABLE_EXCEPTIONS "Enable exceptions in the built library." ON)
+option(LIBCXX_ENABLE_RTTI
+ "Use runtime type information.
+ This option may only be set to OFF when LIBCXX_ENABLE_EXCEPTIONS=OFF." ON)
option(LIBCXX_ENABLE_FILESYSTEM
"Whether to include support for parts of the library that rely on a filesystem being
available on the platform. This includes things like most parts of <filesystem> and
@@ -97,6 +101,10 @@ option(LIBCXX_ENABLE_WIDE_CHARACTERS
support the C functionality for wide characters. When wide characters are
not supported, several parts of the library will be disabled, notably the
wide character specializations of std::basic_string." ON)
+option(LIBCXX_ENABLE_THREADS "Build libc++ with support for threads." ON)
+option(LIBCXX_ENABLE_MONOTONIC_CLOCK
+ "Build libc++ with support for a monotonic clock.
+ This option may only be set to OFF when LIBCXX_ENABLE_THREADS=OFF." ON)
# To use time zone support in libc++ the platform needs to have the IANA
# database installed. Libc++ will fail to build if this is enabled on a
@@ -284,14 +292,6 @@ if (LIBCXX_BUILD_32_BITS)
endif()
# Feature options -------------------------------------------------------------
-option(LIBCXX_ENABLE_EXCEPTIONS "Use exceptions." ON)
-option(LIBCXX_ENABLE_RTTI
- "Use runtime type information.
- This option may only be set to OFF when LIBCXX_ENABLE_EXCEPTIONS=OFF." ON)
-option(LIBCXX_ENABLE_THREADS "Build libc++ with support for threads." ON)
-option(LIBCXX_ENABLE_MONOTONIC_CLOCK
- "Build libc++ with support for a monotonic clock.
- This option may only be set to OFF when LIBCXX_ENABLE_THREADS=OFF." ON)
option(LIBCXX_HAS_MUSL_LIBC "Build libc++ with support for the Musl C library" OFF)
option(LIBCXX_HAS_PTHREAD_API "Ignore auto-detection and force use of pthread API" OFF)
option(LIBCXX_HAS_WIN32_THREAD_API "Ignore auto-detection and force use of win32 thread API" OFF)
|
This moves the configuration of the CMake features to turn off RTTI, exceptions and friends to the beginning of the CMake file, where we configure other optional parts of the library. This fixes an important bug where we would disable the benchmarks because these options were not defined yet, leading to the build thinking they were defined to OFF.
ldionne
force-pushed
the
review/fix-skipping-benchmarks
branch
from
1e15e50
to
0f10a25
Compare
|
LLVM Buildbot has detected a new failure on builder Full details are available at: https://lab.llvm.org/buildbot/#/builders/193/builds/3136 Here is the relevant piece of the build log for the reference |
|
LLVM Buildbot has detected a new failure on builder Full details are available at: https://lab.llvm.org/buildbot/#/builders/52/builds/3715 Here is the relevant piece of the build log for the reference |
|
LLVM Buildbot has detected a new failure on builder Full details are available at: https://lab.llvm.org/buildbot/#/builders/38/builds/833 Here is the relevant piece of the build log for the reference |
|
@vitalybuka Do you know who owns the |
|
@vvereschaka Sorry for breaking your windows bot, that failure seems legitimate. It looks like we're now trying to build the benchmarks (on Windows) but we weren't previously. I suspect this is because building GoogleBenchmark won't work in a windows-cross-linux configuration, I imagine you have flags that make it unhappy. IMO the way to go here might be (sadly) to use |
|
LLVM Buildbot has detected a new failure on builder Full details are available at: https://lab.llvm.org/buildbot/#/builders/55/builds/3560 Here is the relevant piece of the build log for the reference |
|
Hey @ldionne - possibly related to #115361 (comment), we're also seeing this error propagate on our clang-linux-arm64/clang-linux-x64 down stream builders as well[0]. It sounds like this change possibly triggered something that was broken with how |
@ldionne thank you for letting me know. I'll test the build configuration with UPDATE: unfortunately passing |
|
LLVM Buildbot has detected a new failure on builder Full details are available at: https://lab.llvm.org/buildbot/#/builders/85/builds/2783 Here is the relevant piece of the build log for the reference |
I am seeing this in your error message, which seems to be the problem: I guess just using
Can you try doing a clean build? We definitely don't try to configure GoogleBenchmarks if llvm-project/libcxx/test/CMakeLists.txt Line 65 in 0fd6f68 Can you share more context around the failure you're still seeing? |
petrhosek
added a commit
to petrhosek/llvm-project
that referenced
this pull request
Nov 18, 2024
This addresses the issue uncovered by llvm#115361. Previously, we weren't building benchmarks in many cases due to the following block: https://github.com/llvm/llvm-project/blob/e58949632e91477af58d983f3b66369e6a2c8233/libcxx/CMakeLists.txt#L162-L172 We need to passthrough the necessary variables into the benchmarks subbuild and use correct syntax.
|
akshayrdeodhar
pushed a commit
to akshayrdeodhar/llvm-project
that referenced
this pull request
Nov 18, 2024
…llvm#115361) This moves the configuration of the CMake features to turn off RTTI, exceptions and friends to the beginning of the CMake file, where we configure other optional parts of the library. This fixes an important bug where we would disable the benchmarks because these options were not defined yet, leading to the build thinking they were defined to OFF.
Good point, they include more heavy steps which likely fail for reasons other than libc++. Solution is to extract libc++ into dedicated builder. llvm/llvm-zorg#310 for tracking. |
|
However this one looks very relevant @ldionne |
I was quick with judgement. It will not help. They are just slow, but far from slowest bots on LLVM. Slow bots mean large blame list, more unrelated patches blamed with broken patch. I don't think that they are any how special. Bots a very useful catching sanitizer related miscompilations early. |
I think this is because we're now building and running benchmarks that we weren't building before due to: llvm-project/libcxx/CMakeLists.txt Lines 162 to 172 in e589496 Those benchmarks run too slowly and timeout when built with instrumentation. |
Thanks, there is obviously no point to do so with sanitizers. |
vitalybuka
added a commit
to llvm/llvm-zorg
that referenced
this pull request
Nov 19, 2024
petrhosek
added a commit
that referenced
this pull request
Nov 19, 2024
…6644) This addresses the issue uncovered by #115361. Previously, we weren't building benchmarks in many cases due to the following block: https://github.com/llvm/llvm-project/blob/e58949632e91477af58d983f3b66369e6a2c8233/libcxx/CMakeLists.txt#L162-L172 We need to passthrough the necessary variables into the benchmarks subbuild and use correct syntax.
swatheesh-mcw
pushed a commit
to swatheesh-mcw/llvm-project
that referenced
this pull request
Nov 20, 2024
…m#116644) This addresses the issue uncovered by llvm#115361. Previously, we weren't building benchmarks in many cases due to the following block: https://github.com/llvm/llvm-project/blob/e58949632e91477af58d983f3b66369e6a2c8233/libcxx/CMakeLists.txt#L162-L172 We need to passthrough the necessary variables into the benchmarks subbuild and use correct syntax.
|
@ldionne ,
sorry, it was my bad with some mistype during the quick check. |
|
@petrhosek @vitalybuka FWIW I suspect that you folks are running the tests using |
Can't say for @petrhosek issues But bootstrap bots run |
|
We're using |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
This moves the configuration of the CMake features to turn off RTTI, exceptions and friends to the beginning of the CMake file, where we configure other optional parts of the library.
This fixes an important bug where we would disable the benchmarks because these options were not defined yet, leading to the build thinking they were defined to OFF.