Re: SEGV in mount?

[Jeff Dike]

> I've put some debugging code into my kernels to spit out a syslog
> message anytime "do_page_fault" enters "bad_area".  Something that
> I've noticed recently is that "mount" appears to seg-fault every time
> it is called.

This is not a problem.  It's the result of some cleanup in -pre9.  
copy_mount_options was changed to use copy_from_user and let it do the pointer 
validity checking instead of doing it by hand.

The mount options can be up to a page of data, so a page is allocated and 
copied in from userspace.  In general, there won't be that much, and so there 
is likely to be a fault involved.

				Jeff