I have released klibc-1.0.9. This version fixes signal handing on x86-64, and on i386 with Red Hat's execshield patch. There might be other architectures affected (ppc is known to *not* be affected), but right at the moment I don't have a good way to test those. The architectures affected are ones for which signal handing *without* SA_RESTORER set is broken. This is also the first release after switching to git for development, so there might be some screwage there. I plan to re-release this as 1.1 assuming there aren't any unforseen problems. -hpa
Jeff Bailey wrote:> > Should we consider adding a testsuite to klibc? If there's something I > can run at build-time, it'll get checked automatically when I do the > upload to Ubuntu on i386, x86_64, powerpc, sparc, parisc and ia64. > > Once Sarge releases, it can get testing on the rest of Debian's archs > too (alpha, arm, m68k, mips, and s390) >Yes, a real test suite for klibc would be a major win. Right now there is a bunch of "tests" but they're not self-checking, and fairly random. -hpa
Le jeudi 19 mai 2005 ? 23:48 -0700, H. Peter Anvin a ?crit :> I have released klibc-1.0.9. This version fixes signal handing on > x86-64, and on i386 with Red Hat's execshield patch. > > There might be other architectures affected (ppc is known to *not* be > affected), but right at the moment I don't have a good way to test > those. The architectures affected are ones for which signal handing > *without* SA_RESTORER set is broken.Should we consider adding a testsuite to klibc? If there's something I can run at build-time, it'll get checked automatically when I do the upload to Ubuntu on i386, x86_64, powerpc, sparc, parisc and ia64. Once Sarge releases, it can get testing on the rest of Debian's archs too (alpha, arm, m68k, mips, and s390) Tks, Jeff Bailey -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.vas.nu/pipermail/klibc/attachments/20050520/b3a015e0/attachment.html
Reasonably Related Threads
- [PATCH klibc 1/3] Revert " Kbuild: Tell gas we don't want executable stacks"
- ExecShield in C6 or C7 kernels?
- NX (execshield)
- process '/usr/bin/rsync' started with executable stack
- [PATCH/RFC] klibc/kbuild: use separate kbuild files for each klibc subdirectory