The subject tells the story. Running a fresh NetBSD 10.1 system on a Lenovo T480s laptop. I am currently using the system without a login manager. After I installed xfce4-battery-plugin, on reboot, using the startx command was refused with an error message about a missing authorization file in /home/user. I assume that this is a bug, but might have something to do with me not using a login manager???

I think it's a coincidence. I'm pretty sure this is not related to xfce4-battery-plugin. `rm ~/.serverauth*` , make sure you have write access in your home directory, and make sure you `startx` as you. If that doesn't work, edit your `startx` and make sure it looks normal. (Maybe that package does mess with that??)

@"anarchosax"#p18580 is your user in operator group? Also what's your rc.conf ?

@"Jay"#p18582 Yes, my user is in operator group. Here is my rc.conf: dhcpcd=YES dhcpcd_flags="-qM iwm0" wpa_supplicant=YES wpa_supplicant_flags="-B -s -i iwm0 -D bsd -c /etc/wpa_supplicant.conf" hostname=ironhorse dhcpcd=YES dhcpcd_flags="-qM iwm0" wpa_supplicant=YES wpa_supplicant_flags="-B -s -i iwm0 -D bsd -c /etc/wpa_supplicant.conf" cgd=NO raidframe=NO wscons=YES dbus=YES avahidaemon=YES cups_browsed=YES cupsd=YES

@"kc9udx"#p18581 Not finding any startx file. Here is a more precise quote of the error message: `xauth: /home/user/.serverauth.1367 (or 1517) does not exist`

@"anarchosax"#p18584 what are the contents of `/etc/hosts`? **EDIT:** Here's mine removing my IP-address. It need to be formatted using `domain hostname` ``` # $NetBSD: hosts,v 1.9 2013/11/24 07:20:01 dholland Exp $ # # Host name database. # # This file contains addresses and aliases for local hosts whose names # need to be resolvable during system boot; typically this includes only # the address and FQDN for this machine's hostname. # # By default this file is consulted before DNS, so adding additional # material here that then becomes out of date can lead to confusion. # See nsswitch.conf(5). # ::1 mybox.my.domain mybox xxx.x.x.x mybox.my.domain mybox # # RFC 1918 specifies that these networks are "internal": # 10.0.0.0 - 10.255.255.255 (10/8 prefix) # 172.16.0.0 - 172.31.255.255 (172.16/12 prefix) # 192.168.0.0 - 192.168.255.255 (192.168/16 prefix) ```

@"anarchosax"#p18584 `which startx`

@"kc9udx"#p18586 I think this is a `xauth` issues because `/etc/hosts` doesn't match domain and host names.

@"pin"#p18585 The only difference that I see is my file has "localhost" instead of "mybox"...

@"kc9udx"#p18586 Thanks for that. I had a look at the file, but to be honest, I have no idea what it is supposed to look like, but nothing obvious stood out...

Hmm. Definitely irritating. I tried installing the plugin again to confirm behavior, which was the same. I removed again with `pkg_delete -r`. I tried running `rm ~/.serverauth*` which reported that there was nothing. I then rebooted and now every time that I try to startx, I get the `xauth: /home/user/.serverauth.1367 (or 1517) does not exist` and am unable to startx at all. Last time, I was able to save the situation by plugging in the laptop to a wall socket, but that doesn't work this time.

xfce4-battery-plugin renounces my user’s ability to startx.

kc9udx

pin why not? Dynamic IP.

anarchosax

kc9udx I always use DHCP.

kc9udx ?!?

rvp

kc9udx why not? Dynamic IP.

Do what I do if you run DHCP: use the vastly empty 127/8 address space:

printf '127.0.1.1\tCoreBook.local\tCoreBook\n' >> /etc/hosts

anarchosax I am again experiencing the xauth: /home/user/.serverauth.xxx does not exist error.

This is a harmless warning--just ignore it.

What's your actual issue, here?

anarchosax

I have reinstalled NetBSD on the T480s, with CWM as a minimal window manager. the package script desktop-installer was not used, and after the first boot, I am again experiencing the xauth: /home/user/.serverauth.xxx does not existerror. I tried deleting rm ~/.serverauth* again, but without any luck.

Enabling xdm in /etc/rc.conf gets rid of the xauth error, but only ctwm will appear at "startx", despite my .xinitrc and .xsessions files being exactly the same (both contain "exec cwm" of course). I will have to try this install again with xdm enabled at install, but it looks like xdm is the variable that prevents the error. desktop-installer also enables xdm automatically.

kc9udx

xdm runs as root so it's looking for a different magic cookie and finding it. It isn't really solving your 'problem'. As @pin writes, using a static IP and putting your hostname in /etc/hosts should "solve" that.

You don't need to reinstall anything to use or not use xdm. Just change /etc/rc.conf xdm=YES or xdm=NO
Really all xdm does for you is give you graphical login and an express ticket to X. Unless you always want to boot and go straight to X, it's easier to not use it. If you want to run it but not at startup, set it to NO in rc.conf and doas service xdm onestart or just doas xdm

You can startx cwm, no?
There are a dozen or so scripts that run at X startup; any one of them could be causing ctwm to start.

Until you get familiar with ironing out these issues, I wouldn't use desktop installer (mostly because I don't know what it all does)

anarchosax

kc9udx You can startx cwm, no?

Not really. I have checked both .xinitrc and .cwmrc. Both are fine, however I get a very weird desktop:

The white terminal in the upper left is xterm, which for some reason everything that I do through the CWM menu, plus any programs I run, goes through that xterm instance.

The other window is lxterminal, which is set as my default terminal and appears when I run scrot, for taking a screenshot.

After a reboot, my setup came back, but according to xrandr -q at 1920x1080, rather than what I set it to with $ xrandr -s 1680x1050...

kc9udx

I just installed cwm and
startx cwm
works fine. But I don't know the first thing about cwm and it didn't like the .cwmrc I quickly typed up. So all I get is the same output window you get in the upper left. The only way I can make anything else happen is by starting X capable programs with DISPLAY=:0 from a different virtual terminal.
What little I've seen of cwm, it looks less capable than twm.

anarchosax

kc9udx Here is my .cwmrc and my .xinitrc. Enjoy.

kc9udx

rvp so any IP address assigned to the machine in question would work. Thank you.

anarchosax

rvp When it happens, I never get an X environment. I have to shutdown and try again.

anarchosax

rvp Do what I do if you run DHCP: use the vastly empty 127/8 address space:
printf '127.0.1.1\tCoreBook.local\tCoreBook\n' >> /etc/hosts

That looks pretty straight forward, but I am wondering what happens when the laptop goes traveling… Doesn’t/etc/hosts file up quickly?

rvp

kc9udx so any IP address assigned to the machine in question would work.

Yeah, any address in the 127/8 range should work (it does on Linux); even 127.0.0.1 is OK. After all, no packets sent by the system to any of its own addresses (even external) will ever go out on the wire.

anarchosax That looks pretty straight forward, but I am wondering what happens when the laptop goes traveling… Doesn’t/etc/hosts file up quickly?

Don't get this--how does it fill up? We're just saying that the hostname is also 127.0.0.1.

anarchosax I added a -t4value in the xbattbar line of my ~/.xinitrc and again, I am getting xauth errors.

Correlation--not causation, mate. xauth is a red-herring (really! I can give you chapter and verse if you want...).

I think the root cause here is, either:

an issue with your HW (less likely).
Run memtest+ for a day or so to rule out the major cause of this, ie. memory.
KabyLake CPU issue (even less likely).
Update the BIOS and install the latest intel-microcode-netbsd package for this one.
this is an DRMKMS driver problem.
There were some fixes just a few days ago for ~~KabyLake~~ Gen6 HW.

For the latter, either try the latest -HEAD image (with the recommended modesetting(4) Xorg display-driver), or if that too is unstable use the safe wsfb(4) driver. You could also run a 10.1 userland on the latest -HEAD kernel/modules/graphics-blobs to keep using the quarterly binary packages.

rvp

anarchosax Show us your /var/log/Xorg.0.log and/or ~/.xession-errors

rvp

anarchosax You need a comment char. (#) after &
(Actually, you don't need a & after commands which exit right away):

xsetroot -solid \#5F9EA0 &   (or #11acdb) (or #1fb2b7)

Does ~/.xinitrc have execute perms.? Add a #!/bin/sh as the first line, too.

anarchosax

rvp

OK, I got hit by the xauth: /home/user/.serverauth.xxx does not existerror again, when I added an xrandr change xrandr -s 1680x1050 & that works manually, to my ~/.xinitrc file. This seems like an appropriate time to post the requested files:
xorg_log.txt

The ~/.xsession_errors file is just one line, and apparently reflects my having to poweroff the computer and restart a couple of times before I could get to cwm:
X connection to :0 broken (explicit kill or server shutdown).

anarchosax

rvp Does ~/.xinitrc have execute perms.?

Yes.

kc9udx

It sure seems to me like cwm halted, but because you're using exec in .xinitrc you never see it. I wouldn't use exec there.

anarchosax

kc9udx

I have tried using startx cwm both from the command line and in the ~/.xinitrc file. In both cases, I consistently get xauth errors and then a black screen that doesn't recover. I end up having to shutdown the computer and restart every time.

kc9udx

That sure sounds (I could be wrong) like X is running but cwm fails. You should be able to switch to a different virtual terminal and pkill startx. You shouldn't have to reboot.

Did you make the change @rvp requested?
Did you change exec cwm to just cwm?
Don't put startx in .xinitrc

You can safely ignore all the xauth: /home/user/.serverauth.xxx does not exist errors. They really don't have anything to do with whatever is happening.

anarchosax

kc9udx Did you make the change @rvp requested?
Did you change exec cwm to just cwm?
Don't put startx in .xinitrc

Ahh. OK, now I get it. X seems to be working fine.

Does dropping the "exec" also count for other desktop environments and window managers as well?

« Previous Page