On 17-02-03 11:08:38, Dietmar Eggemann wrote:
On 02/03/2017 10:34 AM, Juri Lelli wrote:
On 02/02/17 15:59, Olav Haugan wrote:
On 17-02-01 10:12:54, Juri Lelli wrote:
On 31/01/17 13:23, Dietmar Eggemann wrote:
On 26/01/17 00:03, Olav Haugan wrote:
On 17-01-25 13:26:12, Dietmar Eggemann wrote: > On 12/01/17 19:28, Stephen Boyd wrote:
[...]
Yes, lets not get too complicated here. Its not that critical. I can check against ONLN at parsetime.
Maybe I managed to confuse myself, but don't we actually want to use CONF for the static parsing and checking once before start time?
Yes, we should use sysconf(_SC_NPROCESSORS_CONF) for boundary check only (no hotplug).
On my Juno with [0 1 2 3 5] online I get:
SYSC_sched_getaffinity=8 _SC_NPROCESSORS_CONF=6 _SC_NPROCESSORS_ONLN=5
So it seems that only CONF works here.
[...]
I would prefer not to have this restriction. Sometimes I really want the rt-app thread itself run on a separate CPU to not conflict with the actual usecase that I am trying to test using the json.
Right, +1.
Agreed.
Great, I think we are in agreement. I will submit another patch for review.
Thanks.