[ATfL] Restore the use of -DLLVM_ENABLE_LIBCXX=ON in build.sh #370
Conversation
This file contains hidden or 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
Recently we removed -DLLVM_ENABLE_LIBCXX=ON from build.sh to prevent compiler warnings and sudden relocation errors at the link time. This move was not justified properly. With removal of this setting, there is no CMake-level governance and reassurance that the LLVM's libc++ is astually being used as the C++ standard library. Currently we are ensuring that by using the -stdlib++-isystem flag fed directly to the compiler flags, along with -stdlib=libc++ fed directly to the linker flags. Although this is the right thing to do, it does not effect in a proper awareness at CMake level that libc++ is being used. An observation has been made that restoring -DLLVM_ENABLE_LIBCXX=ON with current set of compiler flags result in CMake dropping the use of -fPIC from the compiler invocations, this happens despide our explicit use of -DLLVM_ENABLE_PIC=ON. As a result, the following relocation error is occuring at the link time: ld.lld: error: relocation R_AARCH64_TLSLE_ADD_TPREL_HI12 against CurrentThreadTaskGroups cannot be used with -shared To address it, this patch adds explicit use of the -fPIC flag to the compiler flags.
pawosm-arm
requested review from
DavidTruby,
tblah,
pratlucas,
dcandler,
voltur01 and
schander-arm
|
We are still a long way away from the ATfL release so I don't think we need hacks like this. Either dropping -fPIC is a bug in upstream llvm (and should be fixed there), or there is a good reason why llvm is doing this and so we shouldn't be doing it either. |
|
I'll keep it open anyway, as I don't like the situation where something suspicious happens without a good explanation. |
|
Slightly (un)related: #374 |
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.
Recently we removed -DLLVM_ENABLE_LIBCXX=ON from build.sh to prevent compiler warnings and sudden relocation errors at the link time.
This move was not justified properly. With removal of this setting, there is no CMake-level governance and reassurance that the LLVM's libc++ is astually being used as the C++ standard library. Currently we are ensuring that by using the -stdlib++-isystem flag fed directly to the compiler flags, along with -stdlib=libc++ fed directly to the linker flags. Although this is the right thing to do, it does not effect in a proper awareness at CMake level that libc++ is being used.
An observation has been made that restoring -DLLVM_ENABLE_LIBCXX=ON with current set of compiler flags results in CMake dropping the use of -fPIC from the compiler invocations, this happens despide our explicit use of -DLLVM_ENABLE_PIC=ON. As a result, the following relocation error is occuring at the link time:
To address it, this patch adds explicit use of the -fPIC flag to the compiler flags.