When I activate internal screensaver with blanking on (e.g. xset s activate s blank), monitor is turned off. It is able to resume correctly, but I think this behaviour is not wanted, because I can use dpms for this and at least basic KDE screensaver sets blanking on, so it is hard to turn it off. I got this behaviour after upgrading from xf86-video-ati-6.6.3 and xorg-1.2 to xf86-video-ati-6.7.195 and xorg-1.4. When I use xf86-video-ati-6.6.3 with xorg-1.4, internal screensaver is working as with xorg-1.2. I am using CRT monitor and ATI Radeon 9250. I was able to track it down to RADEONBlank function in radeon_display.c in xf86-video-ati. When I comment out lines with ...->funcs->dpms(...,DPMSModeOff), screensaver don't anymore turn monitor off. Obviously, it is not fix. I think this is also triggered when X server is started, reset and when switching between X server and text console - in all these cases is monitor turned off and on again - it disappears with my "not-fix".
I was going to report the same thing, as old behaviour was pretty nice: "xset s 300 300": make the screen go black in 5 minutes "xset dpms 600 600 600": and then make it turn off in another 5 (10 from start) Now "s" will let you choose between off with "xset s blank" or the moire inducing cross pattern with X logo with "xset s noblank", but no way to make it black but still give time to avoid a start/stop cycle. Old behaviour, even if what the man page says contradicts it, was the most complete (if I want animations eating CPU, even if slow, I will install xlock, xscreensaver or something like that, which I just had to do to get a black screensaver). New behaviour forces CRT tubes and LCD backlights a bit too much without option to have the two phase power off. And in the case of CRTs, it is even worse, as I doubt a high frequency pattern is nice for the driving circuits (at least I saw it moves slightly, so no burn in if using fast cycle time as "xset s 300 30"). A less extreme pattern would be better (or even just show the X logo in faded colours, without any bg pattern), if there is no way to go back to "s makes screen black". Yes, I tried "xset dpms 100 900 900" to test if first level would be OK, but it seems this monitor does the same for all modes.
Created attachment 14313 [details] [review] revert monitor blanking to pre 6.7 behaviour With this patch applied, "xset s activate s blank" will blank the video output but will not turn monitor off. So it should work like in version 6.6.3 and before. Only tested with one CRT monitor, so I don't know if it will work with different configurations.
The patch still works for xf86-video-ati-6.8.0.
turning off saves more power.
That is what DPMS was doing. Now the only way to have a progressive turn off is running yet another process.
Hmm, this is sad, you just ignore CRT monitors. Using DPMS is not option for CRT, because after power off/on it takes some time before colors stabilize.
Use of freedesktop.org services, including Bugzilla, is subject to our Code of Conduct. How we collect and use information is described in our Privacy Policy.