Svante Signell, le Mon 22 Aug 2011 08:59:13 +0200, a écrit :
> On Mon, 2011-08-22 at 05:06 +0200, Samuel Thibault wrote:
> > I've found the issues with ruby1.9.1. Some stack position detection
> > issue as well as readdir_r correctness. I'm letting it build during the
> > night and will upload this monday so it'll be available on debian-ports
> > later this monday or tuesday.
> Do you mean that the build problems are due to ruby itself, not
Both. Too strong assertion from ruby on __libc_stack_end which is not
updated by libpthread, and bogus usage of struct dirent from ruby.
> How come it built w/o problems on other archs?
Linux doesn't change the stack position, and has a PATH_MAX-like struct