Linux kernel 2.6 smbfs bugs

Michael B Allen mba2000 at ioplex.com
Mon Jan 12 00:53:15 GMT 2004


What doesn't Urban Widmark say? Is he still the smbfs maintainer?

> (I sent this message to the samba list last week, but didn't receive
>  any responses. Probably samba-technical is more appropriate anyway)
>
> Has anyone taken a look at these bugs?
>
> http://bugme.osdl.org/show_bug.cgi?id=1671
> http://bugme.osdl.org/show_bug.cgi?id=1732

No, but just from looking at this now my *guess* is that they are indeed related
and because it has a timing element something could be getting corrupted because
the locking isn't right. But this is *total speculation* on my part. For example
consider the in-situ initializer at the top of smb_readdir:

struct smb_sb_info *server = server_from_dentry(dentry);

This hasn't changed since 2.4 but 2.6 now has a lock_kernel() *after* this. If the
locking semantics have changed in such a way that initialization needs to be moved
into the critical section this is the kind of thing that could cause a problem.

Based on what I know about the kernel this probably isn't really the problem but I
can still make wild guesses :-)

Mike

-- 
A program should be written to  model the concepts of the task it
performs rather than the physical world or a process because this
maximizes the  potential for it  to be applied  to tasks that are
conceptually similar and, more  important, to tasks that have not
yet been conceived.


More information about the samba-technical mailing list