On Fri, Nov 07, 2003 at 03:07:01PM -0700, Kevin Rosenberg wrote:
> Daniel Jacobowitz wrote:
> > Yes, substantially.  But LD_ASSUME_KERNEL=2.4.1 should undo it.  You'll
> > have dig further.
> 
> Well, it's not fixing this issue:
> 
> [EMAIL PROTECTED]:~> LD_ASSUME_KERNEL=2.4.1 df .
> Filesystem           1K-blocks      Used Available Use% Mounted on
> df: `/home': Bad address
> 
> > It's not really a testsuite, just a lot of little test programs.  I
> > doubt it will tell you anything new.
> 
> Most things seem to work well. So far, statfs is the only reproducible
> problem.

I thought you said gcc was broken?

If it's just statfs, look to the kernel's 32-bit wrapper for that
syscall for the bug, probably.

-- 
Daniel Jacobowitz
MontaVista Software                         Debian GNU/Linux Developer


Reply via email to