Re: mbuf corruption tracking advice
Re: mbuf corruption tracking advice
- Subject: Re: mbuf corruption tracking advice
- From: Andrew Gallatin <email@hidden>
- Date: Tue, 24 Oct 2006 14:51:35 -0400 (EDT)
Andrew Gallatin writes:
>
> Does anybody have any advice about how to track down an
> mbuf data corruption panic?
FWIW, I finally stopped looking at my receive routine, and found a bug
in my transmit routine that explained the problem. I was (rarely)
calling mbuf_pullup() and erroneously retaining the original, not the
new, mbuf in my transmit data structures. This presumably caused the
original mbuf to be freed twice (once a bit later by transmit done
interrupt routine, and previously by mbuf_pullup()). Rarely, I'd get
a "freeing free mbuf" panic, and more often the frees would happen far
enough apart that somebody else had allocated it, and things went
downhill fast.
Drew
_______________________________________________
Do not post admin requests to the list. They will be ignored.
Darwin-kernel mailing list (email@hidden)
Help/Unsubscribe/Update your Subscription:
This email sent to email@hidden