#89 CDE hangs at loading screen on openbsd 6.7
As of a few days ago CDE no longer works on openbsd..
Following the instructions, I built and installed it without issue on my VM and it was working.
I tried again a few days later, and it is no longer working.
Trying to start CDE via xenodm results in hanging at the loading screen for a long time before crashing, and dtlogin does not work at all. Following the instructions for enabling/making dtlogin and disabling xemodm simply results in being spammed with "cannot open on display: 0" and booting into a tty.
I have tried everything I can think of, reinstalling, different machines ,earlier versions of openbsd, earlier and from what I can see there has not been any changes to CDE since the time I tried it when it worked..
This is driving me crazy, I cannot understand why something that worked no longer does after no discernible changes
Running(?) the latest CDE release on openbsd 6.7 x64
Tried posting on the mailing list about this, had some trouble getting verified on there but was ultimately ignored anyways.
Related
Check your spam box - I responded to this on the mailing list as follows:
Keep in mind that we all have day jobs and are not just sitting around waiting for support calls.
It looks like my mail provider is not receiving emails from the mailing list, that doesn't surprise me, I had great difficulty getting it to work initially. I apologize for my rudeness, in that case.
The guide for building this on openbsd does not mention rpcbind, I do however have inetd and portmap running.
I was only able to find one of the logs (errorlog), the others did not exist. i will link to it below
https://dpaste.com/453736LQQ.txt
a mirror ...
https://pastebin.com/raw/fcif0K5q
As well as a rather useless picture.
Thank you for replying
I feel I should make another ticket now that I am getting a little further although it is still the same issue.
I am very confused. It was working on this VM okay, now all my machines have this same issue.. what could i have possibly done? these errors seem so ambigious
Last edit: hoobertuder 2020-08-13
To answer your question, openbsd does not use rpcbind, but portmap. Thhe guide advises you on running the portmap service for this.
I have verified it us running.
On 8/14/20 9:47 AM, hoobertuder wrote:
Hi,
does rpcinfo -p
show anything? I wonder if it's a network problem (hostname resolution
and the like) - ttsession is pretty finicky about that, but I thought
those issues had been resolved. And you said you had it working once,
what changed?
-jon
--
Jon Trulson
"Entropy. It isn't what it used to be."
-- Sheldon
Related
Tickets:
#89I am wondering this as well. I have been baffling myself trying to figure this out all week now, I have no clue what changed. The system was untouched after CDE was working, I don't know what could have been changed.
rpcinfo -p output while cde is (attempting) to run ...
rpcinfo -p
program vers proto port
100000 2 tcp 111 portmapper
100000 2 udp 111 portmapper
1342177279 4 tcp 46371
1342177279 1 tcp 46371
1342177280 4 tcp 35092
1342177280 1 tcp 35092
1342177281 4 tcp 3296
1342177281 1 tcp 3296
1342177282 4 tcp 9350
1342177282 1 tcp 9350
1342177283 4 tcp 33320
1342177283 1 tcp 33320
1342177284 4 tcp 40085
1342177284 1 tcp 40085
1342177285 4 tcp 8837
1342177285 1 tcp 8837
1342177286 4 tcp 39926
1342177286 1 tcp 39926
1342177287 4 tcp 35314
1342177287 1 tcp 35314
1342177288 4 tcp 23398
1342177288 1 tcp 23398
and the contents of /etc/hosts
tester:6$ cat /etc/hosts
127.0.0.1 localhost
::1 localhost
I do have pretty finicky internet, but I don't imagine that would be the issue here.
On 8/15/20 10:13 AM, hoobertuder wrote:
Yeah, this look fine...
Also looks fine...
Does your host have an external IP?
Also, if you login to some other X session and open up a terminal, can
you do something like:
/usr/dt/bin/ttsession -S -c /bin/bash
? Does it give you a shell? Does it core? or?
Sorry I can't be of more help. I last tried CDE on OBSD 6.5 and it
seemed to work fine, other than you had to use their login manager
(xenodm?) to start CDE - dtlogin does not work on later BSD's from what
I hear. That worked fine for me, though it was in a VM.
-jon
--
Jon Trulson
"Entropy. It isn't what it used to be."
-- Sheldon
Related
Tickets:
#89On 8/15/20 10:13 AM, hoobertuder wrote:
[...]
One other thing I forgot to mention...
What happens if you add 'localhost.localdomain to the end for the
127.0.0.1 line in your /etc/hosts? Does it improve anything?
-jon
--
Jon Trulson
"Entropy. It isn't what it used to be."
-- Sheldon
Related
Tickets:
#89Hi, sorry for the delay in reply.
Yes, I do have an external IP. I am on a fairly normal network, one modem, one router with one private network. No additio vlans or anything.
Performing the suggested edits to /etc/hosts does not seem to solve the issue. It does howerver start faster, with the same errors though.
if it helps, my hostname is (and please forgive me) trumpdadboomer420.home
My host file is exactly this following the suggested changes ...
127.0.0.1 localhost.home
::1 localhost
One user on reddit told me to change my /etc/hosts to
127.0.0.1 hostname
::1 hostname
It also did nothing.
Now, for your last suggestion, if I understood you correctly I booted into a working wm and tried executing ...
/usr/dt/bin/ttsession -S -c /usr/local/bin/bash (because /bin/bash does not exist)
And, it seems nothing happened. Nothing seems to start, or even appear in top. Nothing seems to crash, either. My dummy wm just stays the way it is
This is so strange. Off the top of your head, what are some reasons this couldbe happening? if it was working before, then suddenly does not want to work.. could this be due to changes in the network? anything?
I hope that tooltalk is less finicky about conditions that don't matter in the future
On 8/17/20 9:59 AM, hoobertuder wrote:
Actually what I was suggesting was something like:
127.0.0.1 localhost localhost.localdomain
exactly as written above.
Yeah, that would be wrong anyway...
So it does not appear to be crashing either... What that command did was
simply start a new TT session and fire off a bash shell as it's parent.
Only a couple things - something with the network, and/or something with
the way it is being started.
I just do not know, off hand.
-jon
--
Jon Trulson
"Entropy. It isn't what it used to be."
-- Sheldon
Related
Tickets:
#89If I could have this working on just one of the bsds I would be happy.
On the advice of a redditor I have changed my hostname to ...
127.0.0.1 localhost
127.0.0.1 spinnerhost.home
::1 localhost
::1 spinnerhost.home
Specifying the actual hostname instead of the variable that resolves to it.
I am pissed off.
Today, it works. It is working today, I have changed nothing. It's working on all my VMs, even the old ones and on my laptop too, I have not touched any of them. Why??
I'm mad I don't understand, anger stems from lack of understanding. It's going to stop working again and i'll never know why, CDE is such a mystery...
On 8/23/20 7:21 PM, hoobertuder wrote:
Well, that would have to be network related. CDE certainly isn't
magically changing itself, especially on multiple machines. Tooltalk is
sensitive to the network environment, so that's where I would look if
things just stop working again. If you have bad routes, bogus DNS or
other issues, tooltalk will be displeased.
So, clearly something changed in your network environment. As to what,
I have no idea. I wish you luck in figuring it out though, should
thinks stop working again.
-jon
--
Jon Trulson
"Entropy. It isn't what it used to be."
-- Sheldon
Related
Tickets:
#89Could you tell me what you mean by bad routes and bogus dns?
The network environment, it would be my local network correct?
Which network is this, my internet, my local network? or simply whatever goes on inside my openbsd install?
Thanks for replying, I love CDE and would like to resolve this, if I can learn more of the issue I can look into the network problem or use another router/modem, although much of the time I was trying to get this to work my VMs were not connected to the internet.. not that connecting made a difference
Last edit: hoobertuder 2020-08-24
I am not entirely sure where to look, for openbsd as far as my network environment goes, I have PF enables, not that disabling it made a difference. My network environment is somewhat stock, I have dhcp and ipv6 autoconf enables on both wireless and ethernet
My hosts file didn't even contain the changes when it started to work, it was just stock as well.
To this day, it still works. All I can really think of is that I uninstalled sakura, p9port and their dependencies as well as a bunch of gtk dog shit