Re: dtrace return hook missing for some iokit funcs?
Re: dtrace return hook missing for some iokit funcs?
- Subject: Re: dtrace return hook missing for some iokit funcs?
- From: Andrew Gallatin <email@hidden>
- Date: Sat, 24 Apr 2010 08:50:43 -0400
Terry Lambert wrote:
On Apr 23, 2010, at 8:41 AM, Andrew Gallatin wrote:
Terry,
Thanks. Commenting the holding-a-simple-lock-while-entering-mutex
assertion got me booting. I managed to get my code working, thanks
to an assertion that triggered in the debug kernel when I passed
the wrong task into IOMemoryDescriptor::withOptions using physical
memory.
Thanks again,
No problem, glad to be of assistance!
Can I ask for something in return?
It would probably be worthwhile for you (an outside person) to request
that the default kernel one uses to do KEXT work be a DEBUG kernel, and
that we support this by adding one to the kernel debug kit, along with
the other necessary pieces. If this caught the problem for you, it
would make sense that it could catch problems for other people.
Will do. Or even better would be to have debug flags like solaris.
I'd love to be able to on-demand enable certain debugging features
and not others, especially since switching kernels is such a PITA
on MacOSX and can easily leave a machine unbootable.
It would also be useful to know the criminal KEXT; if this one is too
hard to track down with your setup, I'd understand if you said not to
it, but it can never hurt to ask...
I can track down the bad KEXT if you can send a patch to turn that
panic() into a stack trace that will fire just once. Eg, so I
can boot up multi-user and pull it out of dmesg, rather than
copy down 15 hex addresses at the limit of my vision.
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