Possibility for FreeBSD 4.11 Extended Support
tom at samplonius.org
Sat Dec 23 23:35:37 PST 2006
> > But kernel panic issues are being fixed right up to the last
> > minute in the 6.2 release train (these and em and socket change
> > issues are probably what has delayed the final 6.2). There is a
> > lot of work getting done, but clearly a lot of work to do. I wonder
> > if this is an area where the Foundation can do something. FreeBSD
> > needs someone to troubleshoot all of the panics and LOR issues.
> > Bug hunting is no fun (for most), and no one is going to do it.
> Actually, I raised hell when the decision was made to release 6.1
> when it was KNOWN that there were bugs. ISTR that the response
> was "we gotta ship and can't be bothered to hold up the schedule
> to fix bugs." I admit that at that point I pretty much gave up.
First of all, knowing their are bugs, and finding bugs are not the same. Just because you know you can cause a panic under some circumstance, does not translate into a fix. It might take 2 to 3 weeks of work to find the cause of that panic.
But this is the problem, everyone just bails out when they see a bug. As I stated, almost no one is really looking for bugs. There are lists of bugs all over the place. But where are the back traces? Where is the analysis?
But instead, more postings to the mailing list. Unless more people start trying to re-create these panics, and post usable data to the lists. It is not insane that releases are being made (like 6.1) with some known panic conditions. It is insane that this situation just generates more content-free e-mails to the mailing lists.
More information about the freebsd-stable