BMQ 0.95 is released with the following changes
1. Fix rq missing on arm arch.
2. Rework/Remove code to reduce overhead.
Enjoy BMQ 0.95 for v5.1 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.
Please report bugs at https://gitlab.com/alfredchen/bmq/issues.
PS, there was a respined before this official release. Pls re-pull/fetch the git or all-in-one file if you have used the 0.95 before this blog.
ReplyDeleteYeah, thank you very much for the rework and respin.
DeleteThe new BMQ 0.95 looks stable ~28h. (The old one led to unclear hardlocks or cpu core overload.)
Best regards,
Manuel
Runs fine on 7 different machines, thanks.
ReplyDeleteRuns well on 5.1.6 with 3ms time slice, is quite strict with usage allocation when it comes to heavy usage. I am compiling for 4ms because current desktop usage is not ideal. What is the reason for 4ms being highly recommended?
ReplyDeleteIt is still the earlier stage of BMQ, so 4ms should be kept to avoid mismatch issues. That's the only reason it is recommended.
Delete@Alfred Chen and @ all other BMQ users:
ReplyDeleteHi all,
can someone of you test (in-kernel) hibernation thoroughly with BMQ (real use, more than two or three runs)?
After resuming from TOI hibernation I again noticed a segfault and very shortly afterwards a complete overload of my 2nd of 2 cores (cpu1), leading to make the desktop (KDE) unusable: Displaying continued, but no input possible, so no logs to reach. This behaviour is the same like I reported for the preliminary BMQ 0.95).
Unfortunately, my kernel seemed to have gotten dangerous again, as I use:
* BMQ 0.95
* "bmq-5.1: re-introduce READ/WRITE_ONCE" from pf (https://gitlab.com/post-factum/pf-kernel/commit/1d9c80521be0f51873ea9de59f65ca380beeb47e)
* O3 commit from pf
* current 5.1 tuxonice/TOI (made from local diff)
Thank you in advance for your ideas,
best regards,
Manuel