Re: chown(2) vs lchown(2) and application breakage
>I may be missing something basic, but this seems to imply that
>dpkg breaks on newer kernels; but (for i386 at least) the breakage
>was for a few 2.1.8x kernels; the newer 2.1.9x and 2.1.10X kernels
>and dpkg work well together.
It still breaks on the alpha (which is how all this started). What
was done in the recent 2.1.9x kernels for i386 and m68k (at least)
was to change the syscall numbers so that calls to the old chown()
to to the new lchown(). This will work fine until libc figures out
that there is an lchown() and chown() has moved.
>If I recall correctly (and I could be wrong here); 2.1.8x
>kernels broke POSIX semantics; and the new kernels unbroke 'em?
Not really, 2.1.8x changed chown() to follow symlinks and introduced
lchown(). The newer kernels switched the calls so old programs that
expected chown() not to follow symlinks would not immediately break.
Regards,
/Anders
--
-- Of course I'm crazy, but that doesn't mean I'm wrong.
Anders Hammarquist | Mud at Kingdoms | iko@netg.se
NetGuide Scandinavia | telnet kingdoms.se 1812 | Fax: +46 31 50 79 39
http://www.netg.se | | Tel: +46 31 50 79 40
--
To UNSUBSCRIBE, email to debian-devel-request@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
Reply to: