1. Minor reworks in two core functions.
2. Set proper time slice for idle tasks.
This should be the last BMQ release in this kernel cycle. Sync-up work from next week. Enjoy BMQ for your linux kernel.
Full kernel tree repository can be found at https://gitlab.com/alfredchen/linux-bmq
And all-in-one patch can be found at gitlab.
Bug report at https://gitlab.com/alfredchen/bmq/issues
Full kernel tree repository can be found at https://gitlab.com/alfredchen/linux-bmq
And all-in-one patch can be found at gitlab.
Bug report at https://gitlab.com/alfredchen/bmq/issues
This comment has been removed by the author.
ReplyDeleteThank you very much.
ReplyDelete@Alfred & @Oleksandr:
ReplyDeleteIs the item 2 ("Set proper time slice for idle tasks.") a resolution for
https://gitlab.com/alfredchen/bmq/-/issues/16 ?
I just don't understand why Oleksandr dropped BMQ inclusion to his -pf patchset.
Anyway, also from my side: Thank you very very much for your great work!
Manuel
None. It's just an improvement for recent idle task investigation.
DeleteIf you drop BMQ it's your own fault.
DeleteIt's the most responsive, albeit with some hiccups now and then.
Let's wait some time until this one gets more and more refined.
It has very good potential.
Actually, there is just one thing, which mostly is not used on desktops, but if some installed package plays with nice and SCHED stuff, it may shoot itself in the foot, e.g. not getting CPU time at all in case already running stuff uses CPU a lot.
DeleteYeah. Like I have explained previously, users of BMQ need to pay attention to the difference between BMQ and other scheduler. And tools based on CFS assumption may failed at some edge cases. I believe that's also true for other non-CFS schedulers.
DeleteThis the one of the main character of BMQ, it's unlikely to be changed.
I believe other maintainer making their decision for their projects/users, must have their own rasons, just like I have done. They should be understood.