Re: Eject problem with optical disk
Re: Eject problem with optical disk
- Subject: Re: Eject problem with optical disk
- From: Jeremy Pereira <email@hidden>
- Date: Tue, 9 May 2006 17:21:23 +0100
On May 8, 2006, at 23:01, Terry Lambert wrote:
The best thing for this is two machine debugging and the command
"showallstacks". This will tell you exactly who has things open,
and why.
I've tried that, but it's more or less impossible to capture the
machine state in the window during which the unmount syscall is
occurring.
But it's most likely that you just need to support the forced
unmount option in your VFS's unmount vfsop.
That would be a nice idea, but I know for a fact that the unmount
operation does not get as far as the vfsop. I inserted a call to
panic() as the first line of the vfsop and there was no panic.
At least not until I rebooted the machine where the panic did occur
at the expected line either because the force flag is set when the
machine is being shut down or because none of the other readdir
activities are happening.
Basically, this will cause the vnodes to be ripped out from under
whoever is trying to use them, and cause further operations to be
invalid.
-- Terry
_______________________________________________
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