diff options
author | Joseph Myers <joseph@codesourcery.com> | 2015-10-27 10:48:05 +0000 |
---|---|---|
committer | Joseph Myers <joseph@codesourcery.com> | 2015-10-27 10:48:05 +0000 |
commit | 7a4cb4ad1130e43e1256f85cea38a3227320f944 (patch) | |
tree | 94076fad4cc4e997011f65d0b6e2a0b9044bbc8d /posix/tst-cpucount.c | |
parent | e4f5eae8f004f236db11325df2adf338766c0543 (diff) | |
download | glibc-7a4cb4ad1130e43e1256f85cea38a3227320f944.tar.gz glibc-7a4cb4ad1130e43e1256f85cea38a3227320f944.tar.xz glibc-7a4cb4ad1130e43e1256f85cea38a3227320f944.zip |
Remove TLS configure tests.
There seemed to be support in response to <https://sourceware.org/ml/libc-alpha/2015-10/msg00510.html> for removing configure tests that exist only to produce errors, where we expect that tool versions failing the tests would also fail the tests of minimum GCC / binutils versions. This patch removes the tests for TLS support as one instance of such tests. Since the addition of TLS emulation support in GCC 4.3, I don't think these tests would have failed even if proper TLS support (as required by glibc) was missing in that architecture's GCC back end, so any new glibc ports wanting a substantive test (if there are actual GCC or binutils versions for those architectures, with supported version numbers, missing TLS support) would have needed an architecture-specific test anyway. Tested for x86_64 and x86 (testsuite, and that installed shared libraries are unchanged by the patch). * configure.ac (libc_cv_gcc___thread): Remove configure test. (libc_cv_gcc_tls_model_attr): Likewise. * configure: Regenerated.
Diffstat (limited to 'posix/tst-cpucount.c')
0 files changed, 0 insertions, 0 deletions