On Fri, Aug 11, 2017 at 3:08 PM, Greg Gorsuch <[email protected]> wrote: > Your description sounds about right. Since I do see a message ("Entering > energy saver mode") come up on the monitor when I move the mouse as I > mentioned before, this does suggest it has been awakened and becomes > unrigistered at this point and sees no input anymore (hence the message). > All the windows are already moved to the other monitor at this point and the > screen saver seems like it has more or less died (though it does prompt for > the password and allows me to log in). > > I did what you asked, but it didn't really give any useful info since it > normally takes a while for the monitor to fail to come back up (maybe 5 or > 10 minutes after I lock the screen even with the sleep time as 1 minutes or > so). I kind of already generated the info you may be seeking in my original > post. > > I am wondering if there is some way to just "lock" the setup so that the > monitor is always there if it is a registration problem, or change things so > the monitor is not unrigestering itself.
Does the monitor have an option to disable input polling? Some monitors rotate through their inputs when they are not getting a signal to see if one attaches so they can automatically switch to that input. This often confuses drivers because the input switching on the monitor often triggers a hotplug event which looks like the display has been disconnected to the driver since the monitor has switched to another input. Alex > > > On Thu, Aug 10, 2017 at 9:55 AM, Thomas Lübking <[email protected]> > wrote: >> >> On Wed, Aug 09, 2017 at 11:39:06AM -0400, Greg Gorsuch wrote: >>> >>> Where are the sleep logistics handled? Is it with the X server, the >>> NVIDIA >>> drivers, the Linux Kernel, or somewhere else? Is it possible to capture >>> the >>> communications on Windows 10 to identify how it is being handled. I don't >>> get the feeling it should be that complex to fix or at least implement a >>> work around since xrander is able to bring the monitor back online. >> >> >> It sounds like when you go dpms off, the monitor unregisters and when >> you go dpms on, the monitor re-registers. >> Then some semi-smart randr daemon kicks in and adjusts the layout, but >> does not so when the monitor comes back. >> >> Try to log the setup in the various modes: >> >> xset dpms force standby; sleep 10; xrandr --current > ~/randr.standby; >> xset dpms force suspend; sleep 10; xrandr --current > ~/randr.suspend; >> xset dpms force off; sleep 10; xrandr --current > ~/randr.off; xset >> dpms force on >> >> Also try the behavior on a "naked" X11 server (no desktop session, only >> an xterm) to see whether some client or the server adjusts the randr >> setup. >> >> Cheers, >> Thomas > > > > > -- > Greg Gorsuch > > _______________________________________________ > [email protected]: X.Org support > Archives: http://lists.freedesktop.org/archives/xorg > Info: https://lists.x.org/mailman/listinfo/xorg > Your subscription address: %(user_address)s _______________________________________________ [email protected]: X.Org support Archives: http://lists.freedesktop.org/archives/xorg Info: https://lists.x.org/mailman/listinfo/xorg Your subscription address: %(user_address)s
