Ok, thanks for the details.
I'm not too happy about scanning NCPUS 'mm' pointers every context
wrap, that can be up to 4096 so the cost is not exactly trivial.
Hi Dave,
Thank you for accepting the changes.
Going through NCPU proportional loop for xcall purpose or 'mm' pointers
is a fundamental limitation when there is only one global context
counter per system. To solve this limitation would be a different
project: we will need multiple context counters per system: up-to one
per core. Bob Picco is working on adding this support. In that case the
loop sizes are going to be bound by:
NCPU / (# context domains)
Pasha
--
To unsubscribe from this list: send the line "unsubscribe sparclinux" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
