Re: Startup disk check
Re: Startup disk check
- Subject: Re: Startup disk check
- From: Peter Lovell <email@hidden>
- Date: Wed, 30 Apr 2003 22:45:58 -0400
Hi Eric (again :),
one thought is to use your (faster) desktop machine with a small
external disk for this testing. Don't have your large (internal) disk
mounted. Just have the minimal stuff needed to test.
Loading and unloading isn't the problem. The problem is that the test
case
produces a panic. I have to reboot after that. I tried a lot of
things to
try and isolate the trigger. What I found is that if I set the
kextcb->e_fcb value to anything other than null in my test kext's
socreate()
hook, the third party extension will call panic(). I'm thinking, what
business is it of theirs what that value is... besides the issue of
why a
kernel extension would ever call panic(). Seems to me if their kext is
confused it should just bow out gracefully in any case.
That's rather tacky. Why on earth would a kext be walking the kextcb
chain to see what other kexts have put in that field of their own
kextcb? The other kext can't just be accidentally stumbling across
yours, as it is given a kextcb of its very own. So it has to be
starting back at the head of the list and walking down.
What this is really saying is that the other kext will panic() if
there's any kext other than itself (perhaps, before itself).
Regards.....Peter
_______________________________________________
darwin-kernel mailing list | email@hidden
Help/Unsubscribe/Archives:
http://www.lists.apple.com/mailman/listinfo/darwin-kernel
Do not post admin requests to the list. They will be ignored.