Messages in this thread | | | Date | Wed, 23 Dec 2009 14:43:14 +0100 | From | Stefan Richter <> | Subject | Re: workqueue thing |
| |
Stefan Richter wrote: > Ingo Molnar wrote: >> (Plus reduction in driver complexity can be measured as well, in the diffstat >> space.) >> >> So there's no leap of faith needed really, IMHO. > > To measure this properly by means of diffstat, first convert typical > create_singlethread_workqueue users to <linux/slow-work.h>, then convert > them back to <linux/workqueue.h>. Then look at the diffstat of the > _second_ step.
PS: Or easier, just do the first step and take its diffstat --- the reverse of that diffstat is more or less the benefit that cmwq would provide. -- Stefan Richter -=====-==--= ==-- =-=== http://arcgraph.de/sr/
| |