Re: [gentoo-user] KDE weirdness after upgrade past Sunday.
J. Roeleveld wrote: > On Wednesday, 4 March 2020 14:41:31 CET Dale wrote: >> Howdy, >> >> I did my usual Sunday upgrades the other day. There was a lot of >> upgrades to plasma and elogind it seems. Let's add LOo in there as well >> just for giggles. Anyway, I have a few oddities going on here. I'm not >> quite sure what to make of it but wondering if anyone else has ran into >> this. >> >> First weirdness. When I lock my screen, CTRL + shift + L. It locks the >> screen just fine. The weird part happens when I poke the mouse or hit a >> key to wake the screen back up. Instead of a screen asking for my >> password with the goofy looking user avatar, I get a black screen with >> the mouse pointer visible. The background and the little box for my >> password, nowhere to be found. After a bit, I type in the password, >> blindly, and it sits there for a while and then my desktop comes back. >> It takes a while and could be related to other problems coming up. > I investigated this a bit more and found it has to do with " kde-plasma/ > kscreenlocker ". > This process was using up 100% CPU (gladly only 1 core) and killing (-9) this > cleared the black screen. > > Can this also be related to still using consolekit? Do I really need to > migrate to "elogind" to be able to unlock my laptop after hibernate? > > Many thanks, > > Joost > I migrated to elogind a few months ago, there's a thread about it too. It went well. No problems. Still, when something goes wrong with elogind, I'm trying to figure out how to recognize it. The biggest thing I don't like, it being in the boot runlevel. It requires me to remember to restart it after a upgrade. If I forget, I run into the issues in this thread. From my understanding tho, consolekit is leaving the building. At some point, switching will have to be done. When I was restarting elogind yesterday, I had a process that was taking up one core as well. I went back to the boot runlevel, stopped elogind, killed any straggling processes and then restarted elogind and back to default runlevel. So far, everything is working fine. It could be a fluke, it could repeat and be a bug of some sort. I'll know next time, if I remember to check. lol Dale :-) :-)
Re: [gentoo-user] KDE weirdness after upgrade past Sunday.
n952162 wrote: > > On 2020-03-04 17:16, Dale wrote: >> >> So elogind is a pretty good suspect. One reason I'm asking about this, >> I'm trying to figure out how elogind fails. After all, if I'm stuck on >> a console, I can't use Seamonkey or anything to find help. > > You could use w3m > > > > I have fluxbox installed but it wouldn't start. I'm glad you mentioned this because I need to either figure out what is broken or find another backup GUI, that I can start from a console of course. Thanks. Dale :-) :-)
Re: [gentoo-user] KDE weirdness after upgrade past Sunday.
On 2020-03-04 17:16, Dale wrote: So elogind is a pretty good suspect. One reason I'm asking about this, I'm trying to figure out how elogind fails. After all, if I'm stuck on a console, I can't use Seamonkey or anything to find help. You could use w3m
Re: [gentoo-user] KDE weirdness after upgrade past Sunday.
On Wednesday, 4 March 2020 14:41:31 CET Dale wrote: > Howdy, > > I did my usual Sunday upgrades the other day. There was a lot of > upgrades to plasma and elogind it seems. Let's add LOo in there as well > just for giggles. Anyway, I have a few oddities going on here. I'm not > quite sure what to make of it but wondering if anyone else has ran into > this. > > First weirdness. When I lock my screen, CTRL + shift + L. It locks the > screen just fine. The weird part happens when I poke the mouse or hit a > key to wake the screen back up. Instead of a screen asking for my > password with the goofy looking user avatar, I get a black screen with > the mouse pointer visible. The background and the little box for my > password, nowhere to be found. After a bit, I type in the password, > blindly, and it sits there for a while and then my desktop comes back. > It takes a while and could be related to other problems coming up. I investigated this a bit more and found it has to do with " kde-plasma/ kscreenlocker ". This process was using up 100% CPU (gladly only 1 core) and killing (-9) this cleared the black screen. Can this also be related to still using consolekit? Do I really need to migrate to "elogind" to be able to unlock my laptop after hibernate? Many thanks, Joost
Re: [gentoo-user] KDE weirdness after upgrade past Sunday.
james wrote: > > Hello Dale, > > Truthfully, I learned, after much pain, decades ago, to keep at least > (2) working gentoo systems, just to avoid catastrophic situations > > Used PC's, if running a minimalistic desktop, are pretty fast. Cross > compile and copy over the updates, or there are many ways to keep old > gentoo systems, active. > > I also picked up a recent laptop, with a 1G traditional Hard Drive, > for under $500.00. It's way faster than my '8350' AMD systems, until > the Kernel-10.0 + starts using the video cards for compiling > *everything* automagically. Kernel-10.x is suppose to be a > 'game-changer' for older hardware, speeding up compiling, tremendously. > > I put 'thermaltake-watercoolers 3.0' on all my chassis based gentoo > systems, at about $80/unit. Keeping the cpu cooled, allows for faster > speeds. My 8350's run at 4GHz and can easily gain 25% clocking to > 5GHz. I use sensors > to monitor temperatures: > > 'watch -n5 sensors -f' > > If I were you, I'd figure out a second, graphical Gentoo system, on > the cheap; as it can be used to fix most 'borked gentoo' systems, > particularly if the best tools are setup and ready, just for > gentoo_repair. > > > hth, > James > > > I've got a spare mobo, memory for it and even my old 4 core CPU. I need a case and a video card to build a system. It's on my todo list, sort of. I got some hard drives laying around that should work as well. They are older small drives, 2 or 3TB drives. I guess I could use a stock CPU cooler but I'd likely buy a new one of those too. May even take the cooler on my current system and put on the spare and put the newer one on this one. I could get a more efficient one and run this 8 core CPU a bit cooler. Anyway. Maybe one day. Oh crap, power supply. Must have that too. I was going to watch a show while my download finished. I feel asleep. Anyway, I finally logged out and restarted elogind. Everything is back to normal so far. Another thing I noticed, when I went to login on the console, note console and Konsole, it to was slow to log me in. Console = ctrl + Alt + F1, 2 etc. Konsole = KDEs commandline program. They appear the same but sometimes work differently. Anyway, both are back to normal. I haven't tested the screen lock part yet but I have to go to a neighbors house and will lock the screen when I leave. I'll know when I get back later. So, if anyone runs up on this thread and are experiencing this type of issues, restart elogind. Rebooting would work if a windoze fan or going to a lower runlevel with the GUI not running and then restart it, using whatever tool manages services. Thanks to all. Dale :-) :-)
Re: [gentoo-user] KDE weirdness after upgrade past Sunday.
On 3/4/20 11:16 AM, Dale wrote: Jack wrote: On 3/4/20 8:41 AM, Dale wrote: All the timing of the above problems are very similar.� I believe they have the same cause.� When I finished my updates, I logged out, went to boot runlevel, used checkrestart to make sure everything that needed to be restarted was clean, restarted any that weren't and then when back to default runlevel.� In the past this has always worked fine.� Thing is, elogind is in the boot runlevel.� I'm going to have to get used to restarting it manually I guess.� Could elogind be the cause of all this?� Would it be safe to put elogind in the default runlevel?� That would solve the problem of me forgetting to restart it after upgrades. Or would some other service in the boot runlevel start it as a dependency anyway?? When I get to a point where I can logout and back in, I'll test restarting elogind to see if it helps.� Thing is, I'm not really sure what all elogind does but from what little I know, it sounds like a good place to start.� Thing that confuses me, checkrestart not showing it needed to be restarted.� It's never failed me before. I was recently in a similar position (but not such serious effects) but discovered that elogind refused to restart.� The message was that it wouldn't start because it was already started, but that implies that it simply failed to stop, without producing any error message.� I didn't want to fight it any further, so I just rebooted. Jack So elogind is a pretty good suspect.� One reason I'm asking about this, I'm trying to figure out how elogind fails.� After all, if I'm stuck on a console, I can't use Seamonkey or anything to find help.� I need to be able to recognize what is failing.� So far, I've yet to find where elogind problems are logged. When you run into the problem with a stuck script, don't forget the zap option.� I'd recommend using ps aux | grep to make sure it is killed and if not kill it with kill first.� After you use the zap option, it should start it normally, the start option.� In case you have never heard of this, it looks like this: /etc/init.d/chronyd zap Hope that helps.� Thanks for the reply. Dale :-)� :-) Hello Dale, Truthfully, I learned, after much pain, decades ago, to keep at least (2) working gentoo systems, just to avoid catastrophic situations Used PC's, if running a minimalistic desktop, are pretty fast. Cross compile and copy over the updates, or there are many ways to keep old gentoo systems, active. I also picked up a recent laptop, with a 1G traditional Hard Drive, for under $500.00. It's way faster than my '8350' AMD systems, until the Kernel-10.0 + starts using the video cards for compiling *everything* automagically. Kernel-10.x is suppose to be a 'game-changer' for older hardware, speeding up compiling, tremendously. I put 'thermaltake-watercoolers 3.0' on all my chassis based gentoo systems, at about $80/unit. Keeping the cpu cooled, allows for faster speeds. My 8350's run at 4GHz and can easily gain 25% clocking to 5GHz. I use sensors to monitor temperatures: 'watch -n5 sensors -f' If I were you, I'd figure out a second, graphical Gentoo system, on the cheap; as it can be used to fix most 'borked gentoo' systems, particularly if the best tools are setup and ready, just for gentoo_repair. hth, James
Re: [gentoo-user] KDE weirdness after upgrade past Sunday.
Jack wrote: > On 3/4/20 8:41 AM, Dale wrote: >> All the timing of the above problems are very similar. I believe they >> have the same cause. When I finished my updates, I logged out, went to >> boot runlevel, used checkrestart to make sure everything that needed to >> be restarted was clean, restarted any that weren't and then when back to >> default runlevel. In the past this has always worked fine. Thing is, >> elogind is in the boot runlevel. I'm going to have to get used to >> restarting it manually I guess. Could elogind be the cause of all >> this? Would it be safe to put elogind in the default runlevel? That >> would solve the problem of me forgetting to restart it after upgrades. >> Or would some other service in the boot runlevel start it as a >> dependency anyway?? >> >> When I get to a point where I can logout and back in, I'll test >> restarting elogind to see if it helps. Thing is, I'm not really sure >> what all elogind does but from what little I know, it sounds like a good >> place to start. Thing that confuses me, checkrestart not showing it >> needed to be restarted. It's never failed me before. > > I was recently in a similar position (but not such serious effects) > but discovered that elogind refused to restart. The message was that > it wouldn't start because it was already started, but that implies > that it simply failed to stop, without producing any error message. I > didn't want to fight it any further, so I just rebooted. > > Jack > > > So elogind is a pretty good suspect. One reason I'm asking about this, I'm trying to figure out how elogind fails. After all, if I'm stuck on a console, I can't use Seamonkey or anything to find help. I need to be able to recognize what is failing. So far, I've yet to find where elogind problems are logged. When you run into the problem with a stuck script, don't forget the zap option. I'd recommend using ps aux | grep to make sure it is killed and if not kill it with kill first. After you use the zap option, it should start it normally, the start option. In case you have never heard of this, it looks like this: /etc/init.d/chronyd zap Hope that helps. Thanks for the reply. Dale :-) :-)
Re: [gentoo-user] KDE weirdness after upgrade past Sunday.
I'm running Kubuntu but have seen the black screen thing once or twice about a month ago. It hasn't returned for awhile. The problem I see about 1 out of every 3 days is ksplashqml crashes when first logging in. It doesn't seem to cause any long=term problems but it's frustrating. That's been going on for about a month, maybe a bit longer but it corresponded to a full upgrade awhile back. On Wed, Mar 4, 2020 at 8:49 AM Jack wrote: > On 3/4/20 8:41 AM, Dale wrote: > > All the timing of the above problems are very similar. I believe they > > have the same cause. When I finished my updates, I logged out, went to > > boot runlevel, used checkrestart to make sure everything that needed to > > be restarted was clean, restarted any that weren't and then when back to > > default runlevel. In the past this has always worked fine. Thing is, > > elogind is in the boot runlevel. I'm going to have to get used to > > restarting it manually I guess. Could elogind be the cause of all > > this? Would it be safe to put elogind in the default runlevel? That > > would solve the problem of me forgetting to restart it after upgrades. > > Or would some other service in the boot runlevel start it as a > > dependency anyway?? > > > > When I get to a point where I can logout and back in, I'll test > > restarting elogind to see if it helps. Thing is, I'm not really sure > > what all elogind does but from what little I know, it sounds like a good > > place to start. Thing that confuses me, checkrestart not showing it > > needed to be restarted. It's never failed me before. > > I was recently in a similar position (but not such serious effects) but > discovered that elogind refused to restart. The message was that it > wouldn't start because it was already started, but that implies that it > simply failed to stop, without producing any error message. I didn't > want to fight it any further, so I just rebooted. > > Jack > > >
Re: [gentoo-user] KDE weirdness after upgrade past Sunday.
On 3/4/20 8:41 AM, Dale wrote: All the timing of the above problems are very similar. I believe they have the same cause. When I finished my updates, I logged out, went to boot runlevel, used checkrestart to make sure everything that needed to be restarted was clean, restarted any that weren't and then when back to default runlevel. In the past this has always worked fine. Thing is, elogind is in the boot runlevel. I'm going to have to get used to restarting it manually I guess. Could elogind be the cause of all this? Would it be safe to put elogind in the default runlevel? That would solve the problem of me forgetting to restart it after upgrades. Or would some other service in the boot runlevel start it as a dependency anyway?? When I get to a point where I can logout and back in, I'll test restarting elogind to see if it helps. Thing is, I'm not really sure what all elogind does but from what little I know, it sounds like a good place to start. Thing that confuses me, checkrestart not showing it needed to be restarted. It's never failed me before. I was recently in a similar position (but not such serious effects) but discovered that elogind refused to restart. The message was that it wouldn't start because it was already started, but that implies that it simply failed to stop, without producing any error message. I didn't want to fight it any further, so I just rebooted. Jack
[gentoo-user] KDE weirdness after upgrade past Sunday.
Howdy, I did my usual Sunday upgrades the other day. There was a lot of upgrades to plasma and elogind it seems. Let's add LOo in there as well just for giggles. Anyway, I have a few oddities going on here. I'm not quite sure what to make of it but wondering if anyone else has ran into this. First weirdness. When I lock my screen, CTRL + shift + L. It locks the screen just fine. The weird part happens when I poke the mouse or hit a key to wake the screen back up. Instead of a screen asking for my password with the goofy looking user avatar, I get a black screen with the mouse pointer visible. The background and the little box for my password, nowhere to be found. After a bit, I type in the password, blindly, and it sits there for a while and then my desktop comes back. It takes a while and could be related to other problems coming up. Second weirdness. When I turn on my external hard drive to do my backups, the device notifier won't let me mount it. It says I don't have permission to mount. I ended up creating the directory and mounting it manually to do my backups. I did a google search but only found the problem back when KDE5 was first coming out, 2018 or so. Given the passage of time, most likely not the same cause. One link even said the bug had been fixed. Third weirdness. When I start Krusader as root, I use it to edit files during updates etc, the window that pops up and asks for my root password takes a long time to show up. Sort of like the lock screen except that it does show up, just really slow to get there. Also, when I type in my password in Konsole to login there, it also takes a long time. By long time, a minute or so. Usually, it is almost instant. All the timing of the above problems are very similar. I believe they have the same cause. When I finished my updates, I logged out, went to boot runlevel, used checkrestart to make sure everything that needed to be restarted was clean, restarted any that weren't and then when back to default runlevel. In the past this has always worked fine. Thing is, elogind is in the boot runlevel. I'm going to have to get used to restarting it manually I guess. Could elogind be the cause of all this? Would it be safe to put elogind in the default runlevel? That would solve the problem of me forgetting to restart it after upgrades. Or would some other service in the boot runlevel start it as a dependency anyway?? When I get to a point where I can logout and back in, I'll test restarting elogind to see if it helps. Thing is, I'm not really sure what all elogind does but from what little I know, it sounds like a good place to start. Thing that confuses me, checkrestart not showing it needed to be restarted. It's never failed me before. Thoughts?? Dale :-) :-)