On Tue, 29 Jun 2004, Bjoern A. Zeeb wrote: > On Tue, 29 Jun 2004, Julian Elischer wrote: > > > > +> do you have a 4.x version? > > > > > > I don't have 4.x boxes, so I can't prepare one for 4.x. Could you try to > > > port it? It should be easy. > > > > I was looking at the original patch for 4.x (the one that is a kld > > module). > > the one from pjd I think this is as I do not have an 4.x verison > online. > > > here is my (I hope) fixed version... > > why do you work on this ? wouldn't backporting the actual patch be > better instead of s.th. different ? I need a solution for my "users" here now (today) so I thought that teh fact that pjd had a 4.x lkm module already available would allow me to give them that solution immediatly while we work on getting a permanent solution in the tree(s). As it happenned our app complained because teh number of filesystems reported when buf == NULL was not the same as the number of filesystems for which data was actually provided. While fixing that I noticed the fact that 'copyout' was not being used to manipulate the strings so I fixed that two.. I will now look at the newer code, and my users are happy :-) pjd can now at least replace teh version he has on his page with teh fixed version, except that I just realised that '[f]statfs()' also probably needs to be fixed to not write directly to userland. > > there are of course some differences in both struct prison and statfs > but it should be doable; making a KLD then is s.th. very simple from > what I had learned. I'll be looking at the new code in a few days.. I have to MFC USB stuff today (we need it :-/ ) > > -- > Bjoern A. Zeeb bzeeb at Zabbadoz dot NeT >Received on Tue Jun 29 2004 - 18:31:37 UTC
This archive was generated by hypermail 2.4.0 : Wed May 19 2021 - 11:37:59 UTC