On Tue, 11 May 2004 15:16:01 -0700
Greg KH <[EMAIL PROTECTED]> wrote:

> On Tue, May 11, 2004 at 03:09:55PM -0700, Pete Zaitcev wrote:
> > I think I get the idea.
> > 
> > What do you think about moving 2.4 on to the 2.6.6 scheme? Not worth it?
> 
> I think the disconnect issues are fixed in 2.4 for now, right?  If so, I
> doubt it is worth it.

I forgot the typical check in the serial_write, thus it is possible
to fail there (one of Rudiger's patches added it). In practical terms
it's somewhat hard to hit, because a line discipline first calls
->write_room, which does the check. If you're on UP, and copy_from_user
does not block in a wrong time, there's no chance for oops. But
even if the check were added to serial_write, we would have been
vulnerable on SMP still.

I did not want to monkey with this coming up to 2.4.26 because
I was afraid to derail the post_helper patch by introducing an
unrelated regression. Now that it's safely in, I am more enthusiastic
about taking good stuff from 2.6.

-- Pete


-------------------------------------------------------
This SF.Net email is sponsored by Sleepycat Software
Learn developer strategies Cisco, Motorola, Ericsson & Lucent use to 
deliver higher performing products faster, at low TCO.
http://www.sleepycat.com/telcomwpreg.php?From=osdnemail3
_______________________________________________
[EMAIL PROTECTED]
To unsubscribe, use the last form field at:
https://lists.sourceforge.net/lists/listinfo/linux-usb-devel

Reply via email to