Messages in this thread | | | Date | Tue, 29 Jul 2014 16:05:33 +0200 (CEST) | From | Jiri Kosina <> | Subject | Re: kGraft to -next [was: 00/21 kGraft] |
| |
On Sat, 5 Jul 2014, Jiri Kosina wrote:
> > > The same holds for the kernel threads -- until all (or most of) the > > > kthreads are converted to workqueues, the obivous choice, that should > > > cover most of the use-cases, has been made. > > > > But, this is different. > > Quite frankly, I have to say that I don't personally see *that* big > difference. In both cases we are making assumptions about semantics, and > are trying to get "as close as possible" to the point in time where the > set of assumptions can be made as minimal as possible. > > With userspace thread, this is kernel/userspace boundary. With kthread, > this is starting of new iteration of the main loop / try_to_freeze(). > > We were not able to come up with anything better. Alan, you were > stating "I don't see why it can't use the kernel freeze functionality?". > What exactly was your suggestion, if not this, please?
Seems like there was no further activity on this wile I was on vacation, so let me resurrect this.
Alan, either we are already doing what you suggested, or I misunderstood your proposal above. Could you please elaborate?
Thanks,
-- Jiri Kosina SUSE Labs
| |