On Saturday, October 25, 2003, at 03:27  PM, der Mouse wrote:

>>> I'm not sure I completely understand the logic, but I cannot see why
>>> it would be necessary to run the final child cleanup (work done by
>>> reaper) from process context.
>> The reaper was specifically added (by me) because these code paths
>> CAN block.  Any time you make a call into UVM, you are likely to take
>> a lock, which could block.
>
> I _must_ be missing something.  What's wrong with their blocking?  I
> see nothing wrong with blocking during process exit.

There were actually several bug reports filed against it at the time (I 
seem to recall that the m68k platforms were nailed particularly hard by 
it, but I encountered it on i386 at the time, as well).  It's been a 
while, so I don't quite remember all of the details.  But the bug 
caused the system to crash.

         -- Jason R. Thorpe <thorpej@wasabisystems.com>