> I take it that :
> 
>     panic: pool_get: extent: page empty
>     syncing disks...
> 
>     <disk timeouts ...>
> 
> is related to the other panics that people are seeing?  Is there a fix in
> -current?   (1.4 GENERIC_SCSI3 on a Sparc 4 - two panics in two days).


I have fixed five instances of missing spl*()/splx() bracketing in various
kernel subsystems which were all liable to cause a failure such as this one.

All of the fixes are scheduled to appear in 1.4.1

-pk