We already use a mechanism to atomically queue kernel requests
via a second BH queue and those routines could kmalloc. I'm generally against
dynamic resource allocation on the RT side: it is RT, it can't wait for
resources, if not, it can run under Linux. Most RT kernels are filled with
bad implementations of the mechanisms of a non-rt kernel.
I'd much prefer to have the Linux side do all the complex stuff.
The other issue is how mm will work. My initial theory is that one could
allocate a chunk for the RT system (in linux kernel mode)
and put all RT tasks in it rather than alloc a map and space for each RT task.
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.tux.org/lkml/