Power Management Bugs Hold Up Some Linux Laptops Due To Regulatory Requirements

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts
  • partcyborg
    Senior Member
    • Apr 2019
    • 288

    #21
    Originally posted by OmniNegro View Post
    Anyone remember back when vehicular emissions were facing "new emissions standards? They would supposedly reduce pollution from cars by 50% overnight. Yeah sure. What happened? Nothing at all. Cars became more expensive, big oil used this as an excuse to raise the prices, and the pollution remained unchanged, because they wrote in loopholes for old big trucks that carried "essential" goods.
    Absolute nonsense. Try doing at least 5 minutes of research next time. At least in the US, emission regulations have had a transformative effect on pollution and air quality. Read a high level summary at https://www.epa.gov/transportation-a...transportation

    Comment

    • QwertyChouskie
      Senior Member
      • Nov 2017
      • 637

      #22
      Originally posted by archkde View Post

      I can certainly see s2idle being a great option, but I don't want it as an unconditional S3 substitute.
      My laptop (2021 Asus Zephyrus G14 with the Ryzen 9 5900HS) uses s2idle sleep rather than S3 sleep. The first few months I owned it during 2021 it wouldn't sleep due to bugs/missing support in the kernel, but a few months later, when the kernel drastically improved its support for s2idle on AMD platforms, it started working just fine, and it's been pretty much rock solid since. Obviously, it's not using any of the fancy features that s2idle can provide, but if I had bought the laptop a few months later, I probably wouldn't even know it was using s2idle sleep rather than S3 sleep.

      Comment

      • Old Grouch
        Senior Member
        • Apr 2020
        • 695

        #23
        Originally posted by archkde View Post
        Wow, shutdown not working properly is certainly a new one.
        Please don't use the BLINK attribute on your invisible sarcasm tags: it hurts my eyes.

        Comment

        • primary
          Junior Member
          • Nov 2023
          • 26

          #24
          My older thinkpad just works, If i close the lid, it sleeps and goes multiple days without charging, wakes up just fine. But with newer laptops I've had (thinkpad and dell) so much troubles, Suspend basically wakes up automatically and even on hibernation battery is sometimes consumed in hours.

          Comment

          • Smurphy
            Phoronix Member
            • Apr 2008
            • 110

            #25
            Amazing. I tend to disable all S power states in the BIOS except S3, as I actually tend to always shutdown my computer at the end of the day.
            When going to lunch, I make a suspend to RAM, and that's it. All other power modes are disabled. Makes less wierd issues happening in general.
            Linuxer since the early beginnings...

            Comment

            • archkde
              Senior Member
              • May 2019
              • 684

              #26
              Originally posted by Old Grouch View Post

              Please don't use the BLINK attribute on your invisible sarcasm tags: it hurts my eyes.
              No sarcasm intended at all. Complaints about S3 or s2idle not working properly are frequent. But before this article, I wasn't aware that the OS is even able to shut down the system improperly in the way described.

              Comment

              • Adarion
                Senior Member
                • Mar 2009
                • 2063

                #27
                Holy shit, I would not have expected this.
                S3 issues, yes, all the way. On some boxes it works, other mess up rarely on entering, but some, sadly, mess up on resume.
                But Soft-off, ATX-off or S5, whatever we call it?

                I thought INIT makes all the programs shut down (correctly). Then it flushes all buffers/caches to the drives. Unmounts the drives (on can follow that during shutdown/reboot procedure) so everything is synced, and finally it remounts / read-only and sends the hardware the shutdown or reboot command.
                So does the kernel have to prepare hardware?
                I thought the last part is matter of CPU or SIO/EC and the SIO/EC (incl. clock) is the only thing to remain powered by the PSU (as one can see when you measure power consumption that there is something like 0.5 Watts being pulled). And the SIO listenes to possible magic key from the keyboard (so yeah, kbd controller will also be on then) or to magic packet arriving for WOL. And in some BIOS setups you can configure that, if the computer can be switched on in soft-off by keyboard or if you really have to shortcut the pins with the power-on key to signal the PSU to send power to the mainboard.

                And then there is Microsoft with their privacy-invading shitty "connected standby" and the likes. You computer never really is off, it seems off, but there is full network stack and everything and it will talk to the net and send and fetch data.
                But now people complain that Linux-driven devices would not fulfill all S5 criteria...
                Stop TCPA, stupid software patents and corrupt politicians!

                Comment

                • Old Grouch
                  Senior Member
                  • Apr 2020
                  • 695

                  #28
                  Originally posted by archkde View Post

                  No sarcasm intended at all. Complaints about S3 or s2idle not working properly are frequent. But before this article, I wasn't aware that the OS is even able to shut down the system improperly in the way described.
                  OK. I'll say that some of the more frequent problems I've had are to do with shutdowns. I've had systems go into tight cpu loops with fans screaming until either hard powered-off or the batteries run out, systems that sit waiting indefinitely for network hardware, or graphics hardware, or unable to work out how to dismount /var that was on a separate, encrypted partition.

                  I went though a phase of being routinely unable to recover from suspend states (that was wireless network card related), but able to hibernate; and vice-versa.

                  I think that shutting down (and suspending) exercises rarely used code-paths, and rarely-used hardware facilities that are by no means bug-free. I don't think such problems are prioritised for resolution because it is 'always'* possible to 'just cut the power', which mostly has few after-effects,


                  *Not in phones that have unremovable batteries. Sometimes you have to wait for them to run flat. And not in remote devices, unless you have remotely-accessible power relays.

                  Comment

                  • cj.wijtmans
                    Senior Member
                    • Mar 2016
                    • 1404

                    #29
                    Originally posted by caligula View Post

                    In this case they actually improve the situation for customers by letting them buy higher quality systems with non-shit firmware.
                    its almost as if UEFI and firmware was a terrible idea compared to a BIOS with open standards. Not that BIOS was perfect, but it could have been replaced with open source BIOS in the long term. Now we have proprietary firmware full of bugs spyware and backdoors loading code into the operating system. The fact that firmware is not considered kernel taint is surprising since nvidia out of kernel module was considered kernel taint.

                    Comment

                    • cj.wijtmans
                      Senior Member
                      • Mar 2016
                      • 1404

                      #30
                      Originally posted by Adarion View Post
                      Holy shit, I would not have expected this.
                      S3 issues, yes, all the way. On some boxes it works, other mess up rarely on entering, but some, sadly, mess up on resume.
                      But Soft-off, ATX-off or S5, whatever we call it?

                      I thought INIT makes all the programs shut down (correctly). Then it flushes all buffers/caches to the drives. Unmounts the drives (on can follow that during shutdown/reboot procedure) so everything is synced, and finally it remounts / read-only and sends the hardware the shutdown or reboot command.
                      So does the kernel have to prepare hardware?
                      I thought the last part is matter of CPU or SIO/EC and the SIO/EC (incl. clock) is the only thing to remain powered by the PSU (as one can see when you measure power consumption that there is something like 0.5 Watts being pulled). And the SIO listenes to possible magic key from the keyboard (so yeah, kbd controller will also be on then) or to magic packet arriving for WOL. And in some BIOS setups you can configure that, if the computer can be switched on in soft-off by keyboard or if you really have to shortcut the pins with the power-on key to signal the PSU to send power to the mainboard.

                      And then there is Microsoft with their privacy-invading shitty "connected standby" and the likes. You computer never really is off, it seems off, but there is full network stack and everything and it will talk to the net and send and fetch data.
                      But now people complain that Linux-driven devices would not fulfill all S5 criteria...
                      WOL is important to me for local networking, but apart from that you are completely correct. I dont like my devices to connect to the outside world at all, let alone while its suspended. It is obvious they want to push updates and "telemetry" control even while your device is "off". I have all my automatic updates OFF for EVERYTHING. I have one day in the week where i run all updates manually. When anything goes wrong during an update i can have the rest of the day to fix it. If nothing goes wrong i can do other chores or resume work as i wish. It is a planned day so i will never have surprises and be frustrated. I can also track what the updates actually update, a useful new feature, or nothing useful at all.

                      Comment

                      Working...
                      X