On Wednesday 10 September 2003 06:42 pm, Sean Dague wrote:

> The rh9 patch is just a NPTL fixup, right?  If so, shouldn't that logic
> just be added to the build process for SI?  This would be outside the scope
> of the spec file, but would go into SI proper.

Right.  The changes to jfsutils are just burning out 'extern int errno' and 
adding back #include <errno.h>,  but the change to dosfstools is burning out 
the in-line version of llseek.  It would be possible to just apply the 
patches across the board, but others would have to re-test other distros.  In 
particular, I get nervous that I'm going to break another distro when I'm 
torching out code that has a comment like this one:

/* Use the _llseek system call directly, because there (once?) was a bug in
 * the glibc implementation of it. */

I have no idea whether any of the other distros the team wants to support have 
the buggy glibc, or even if it matters. Since the code runs against the 
uClibc version of the libraries, I'm probably just worrying unnecessarily.

The changes required are modeled after changes that redhat made in their 
distro source rpms.

If the others agree with this outlook, I'll just go ahead and submit the 
patches and expect others to test against debian, suse, gentoo, etc.


-------------------------------------------------------
This sf.net email is sponsored by:ThinkGeek
Welcome to geek heaven.
http://thinkgeek.com/sf
_______________________________________________
Sisuite-devel mailing list
[EMAIL PROTECTED]
https://lists.sourceforge.net/lists/listinfo/sisuite-devel

Reply via email to