Aaron M Ucko <[EMAIL PROTECTED]> writes: > Package: openafs-modules-source > Version: 1.4.0-4 > Followup-For: Bug #358203
> AFAICT, KBUILD_BASENAME is actually supposed to have been defined on > the command line; I worked around that problem by tweaking all the > configure tests accordingly, and it looks like 1.4.1 will do the same. > However, there are two further problems, one at compile time > and one at runtime (at least on amd64). > The other compile-time problem is relatively trivial: osi_file.c > expects struct inode to have a semaphore named i_sem rather than a > mutex named i_mutex. 1.4.1rc10 fixes this problem too. Sounds like rather than try to patch this for 1.4.0, I should just accelerate the packaging of 1.4.1. I was planning to start working on that anyway, so that's not a problem. Hopefully in the next couple of days I can get started. > However, even with the compilation errors out of the way, I still have > a serious problem: I get an oops in osi_syscall_init when loading the > module, both with a minimally-tweaked version of 1.4.0-4 and with > 1.4.1rc10 merged in. This lossage may be architecture-dependent, as I > couldn't find any similar reports in the OpenAFS mailing list archives. I haven't had a chance to look at this yet, but I forwarded it along to the openafs-devel mailing list in case someone gets a chance to look before I do. Thank you! -- Russ Allbery ([EMAIL PROTECTED]) <http://www.eyrie.org/~eagle/> -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
