Linux 5.13 To Fix Its Handling Of Unused ACPI Power Resources

[andre30correia]

98% of the barely functional microsoft morons at MSI can't even say spec, much less code firmware to it.

Trust me. Doing it yourself, in the kernel, is the safest outcome for everyone. Until a US based fab comes online, this is what we're stuck with.

But their complete incompetence idoes have a silver lining every so often.

https://www.kotaku.com.au/2020/07/msi-ceo-dies-after-falling-out-of-a-building/

https://www.pcmag.com/news/msi-subsidiary-caught-scalping-rtx-3080-3090-cards-on-ebay

us fab? the country who makes a monopolity of everything and closes everything?

[cyring]

My gigabyte x570 can't wake up from sleep or suspend...maybe this will help

Even with SMT enabled ?

[muncrief]

There have been many longstanding problems with AMD cards, old and new, not properly resuming from suspend, so it will be interesting to see if these patches fix any of them.

I used to have a GT 710 for my VM and R9 390 for linux (and sometimes visa-versa) and I would have resume problems that would come and go with various kernel versions, always having to do with the R9 390 failing to power on. And now I have two RX 580s (with one dedicated to a VM), and until kernel 5.12 rc2 it failed to resume about 30% of the time. However since 5.12 rc2 it's only failed to resume once, which is a great improvement but proves the basic problem still exists. By the way I filed a bug for my specific problem back in October. You can find it at https://bugzilla.kernel.org/show_bug.cgi?id=209457

[bdincman]

German heise.de had an article on a PC "Bauvorschlag" which led to some problems with the SSD involved and turned out to be a problem with the lowest power saving state (autonomous) if added to a list already present in the code the problem would go my question should this be related to the early non-powering-off mentioned

https://www.heise.de/hintergrund/c-t-wirkt-Problematische-SSD-im-Bauvorschlag-fuehrt-zu-Bugfix-im-Linux-Kernel-5074713.html

[Guest]

There have been many longstanding problems with AMD cards, old and new, not properly resuming from suspend, so it will be interesting to see if these patches fix any of them.

I used to have a GT 710 for my VM and R9 390 for linux (and sometimes visa-versa) and I would have resume problems that would come and go with various kernel versions, always having to do with the R9 390 failing to power on. And now I have two RX 580s (with one dedicated to a VM), and until kernel 5.12 rc2 it failed to resume about 30% of the time. However since 5.12 rc2 it's only failed to resume once, which is a great improvement but proves the basic problem still exists. By the way I filed a bug for my specific problem back in October. You can find it at https://bugzilla.kernel.org/show_bug.cgi?id=209457

Now for me the AMD Linux driver works fine, while Windows driver is facing problems like display blanking in the middle of the game, unable to resume from suspend, or not resuming after screen idle.

Of course AMD Linux driver also had it's own problems like memory clock stubbornly staying at low speed, Zero RPM not working and fan at constant 1600 RPM etc.