The Testing Move has begun

Gullible Jones · 2006-06-23 20:40:14

slackhack wrote:

smooth as silk here. another great job from the arch team!

The main mirror's fully synced now?

Skyscraper · 2006-06-23 20:47:22

tpowa wrote:

install nvidia modprobe -i nvidia ; /etc/udev/scripts/create-nvidia-nodes.sh ; true
this entry is done by nvidia-legacy and is only needed for legacy drivers.
so modprobe.conf is fine when legacy package is installed

I thought so, but when I reinstalled nvidia-legacy package on a current system with empty modprobe.conf - those lines weren't add So i had to manually edit modprobe.conf

Skyscraper · 2006-06-23 20:51:37

Hmm... i think that "nvidia-legacy-utils" should be reinstalled instead of "nvidia-legacy" to recreate entries in modprobe.conf.... What You think?

slackhack · 2006-06-23 20:53:59

Gullible Jones wrote:

slackhack wrote:
smooth as silk here. another great job from the arch team!
The main mirror's fully synced now?

oops, i guess you're right, current might still not be synced. better hold off on the congrats for a while longer.

tpowa · 2006-06-23 20:54:49

yes you are right it's the utils package, that does the entries, but normall modprobe.conf is not updated on upgrading, because it's set to noupgrade in pacman.conf by default.

scottlc · 2006-06-23 20:54:53

Actually, upon closer inspection, I do seem to be having the same libssl trouble explained earlier.

Skyscraper · 2006-06-23 20:58:51

tpowa wrote:

yes you are right it's the utils package, that does the entries, but normall modprobe.conf is not updated on upgrading, because it's set to noupgrade in pacman.conf by default.

So it means that now nvidia-legacy users should keep an eye on the modprobe.conf during upgrade and be sure to have nvidia line in it?

EDIT
-----
Or maybe when i upgrade (manually or not) nvidia-legacy package, pacman shouls automatically upgrade mvidia-legace-utils to check modprobe.conf?

tpowa · 2006-06-23 21:03:41

argh no, modprobe.conf is not touched by pacman, if you do not tell it explicitly to replace it by default.

Skyscraper · 2006-06-23 21:06:48

tpowa wrote:

argh no, modprobe.conf is not touched by pacman, if you do not tell it explicitly to replace it by default.

Yes i know, but nvidia-legacy-utiles touches modprobe.conf and if I accidentally/or not (after upgrade) replace modprobe.conf with an empty one and try to reinstall nvidia-legacy - this should "repaire" my modprobe.conf. I thinnk

tpowa · 2006-06-23 21:08:21

this is more a technically problem. imho the utils package should do the entry because it's the base depend for all legacy modules.

twiistedkaos · 2006-06-23 21:16:35

Hmmm,... I guess I upgraded too soon. The update breaks the nvidia modules and now I can't get on X. Yay!? Thank god fo Links so I can console browse the web. When will the nvidia modules be fixed?

tpowa · 2006-06-23 21:18:55

you can always use 'nv' or 'vesa' driver if nvidia is broken to start X.

Skyscraper · 2006-06-23 21:19:00

tpowa wrote:

this is more a technically problem. imho the utils package should do the entry because it's the base depend for all legacy modules.

And it does but when I upgrade nvidia-legacy, this one doesn't require "nvidia-legacy-utils" to be reinstalled/upgraded to eventually repair modprobe.conf and udev script.
I did something like this today. I replaced modprobe.conf with empty one and when Xorg told me that it has problems with nvidia i automatically tried to reinstall only nvidia-legacy. Everything went fine but xorg complained again... so I restored old modprobe.conf. And I didn't remember/know that nvidia-legacy-utils are responsible for this. So in my opinion, in such occasion, nvidia-legacy during upgrade should call nvidia-legacy-utils to be upgraded too. What You think?

Skyscraper · 2006-06-23 21:20:15

tpowa wrote:

you can always use 'nv' or 'vesa' driver if nvidia is broken to start X.

that's not the elegant solution, You know this

tpowa · 2006-06-23 21:25:44

my opinion is if a user decides to move system files, he knows what he is doing
we cannot handle all possibilities users might do to their system, we try to keep it going for normal users.
and the nv/vesa hint was for the user with Links browser, with the standard Xorg drivers you can always launch X to use your normal browser.

matersci · 2006-06-23 21:34:01

The upgrade went quite fine. Only problem, Cups stopped working.

An error occurred while loading http://localhost:631/:
Could not connect to host localhost (port 631).

twiistedkaos · 2006-06-23 21:43:35

Hmmm the vesa module doesn't work on my systme, just a black screen. But that's alright. I suppose I can wait it out, so I should wait 24-48hr's and the nvidia package should be fixed and I can just do a 'pacman -Suy' to fix my problem? Or is it more serious than that?

Gullible Jones · 2006-06-23 21:51:30

matersci wrote:

The upgrade went quite fine. Only problem, Cups stopped working.
An error occurred while loading http://localhost:631/:
Could not connect to host localhost (port 631).

Weird, that never happened to me when I was using the version from Testing.

zaterio · 2006-06-23 21:56:09

Well my laptop is my production machine and runs only arch..i Have the last beyond kernel from the reposiroties, the update procces says:

:: Replace alsa-driver with kernel26 from "current"? [Y/n] n
:: Replace pwc with kernel26 from "current"? [Y/n] n

Is this a problem for the kernel-beyond users?

tpowa · 2006-06-23 22:18:05

just remove those packages by hand then you were not asked again

matersci · 2006-06-23 22:46:29

Gullible Jones wrote:

matersci wrote:
The upgrade went quite fine. Only problem, Cups stopped working.
An error occurred while loading http://localhost:631/:
Could not connect to host localhost (port 631).
Weird, that never happened to me when I was using the version from Testing.

/etc/rc.d/cups start
:: Starting CUPS Daemon                                                  [FAIL]

and

I [24/Jun/2006:01:48:17 +0300] Saving remote.cache...
E [24/Jun/2006:01:48:17 +0300] Hostname lookup for "localhost" failed!
E [24/Jun/2006:01:48:17 +0300] Bad Listen address localhost:631 at line 18.
I [24/Jun/2006:01:48:17 +0300] Listening to /var/run/cups/cups.sock (Domain)
I [24/Jun/2006:01:48:17 +0300] Loaded configuration file "/etc/cups/cupsd.conf"
I [24/Jun/2006:01:48:17 +0300] Cleaning out old temporary files in "/var/spool/cups/tmp"...
I [24/Jun/2006:01:48:17 +0300] Configured for up to 100 clients.
I [24/Jun/2006:01:48:17 +0300] Allowing up to 100 client connections per host.
I [24/Jun/2006:01:48:17 +0300] Using policy "default" as the default!
I [24/Jun/2006:01:48:17 +0300] Full reload is required.
I [24/Jun/2006:01:48:17 +0300] Saving job cache file "/var/cache/cups/job.cache"...
I [24/Jun/2006:01:48:17 +0300] Loaded MIME database from '/etc/cups': 34 types, 39 filters...
I [24/Jun/2006:01:48:17 +0300] Loading job cache file "/var/cache/cups/job.cache"...
I [24/Jun/2006:01:48:17 +0300] Full reload complete.
I [24/Jun/2006:01:48:17 +0300] Listening to /var/run/cups/cups.sock on fd 1...

Any ideas? Or I will have to downgrade to previous version?

Gullible Jones · 2006-06-24 00:05:47

Hmmm. You sure you have lo up? Also, check your /etc/hosts file.

BTW, I've noticed a bug in mkinitcpio - the filesystem autodetection doesn't work if it's run during the installation, it only works when you run it from within a working Arch Linux system. See here.

(FWIW though, booting with the initcpio fallback image is almost as fast as booting with the customized image. Awesome work, guys.)

jason · 2006-06-24 00:10:18

fglrx is borked.

(II) Loading /usr/lib/xorg/modules/drivers/fglrx_drv.so
dlopen: libstdc++.so.5: cannot open shared object file: No such file or directory
(EE) Failed to load /usr/lib/xorg/modules/drivers/fglrx_drv.so
(II) UnloadModule: "fglrx"
(EE) Failed to load module "fglrx" (loader failed, 7)

This blows, I can't use X without using old ATI drivers boo.

matersci · 2006-06-24 00:12:53

Yes, it was host problem (But why now?)
Thanks
Solved

http://bbs.archlinux.org/viewtopic.php?p=172615

Evetything OK with the upgrade, very good work guys.
My system is faster than ever.

detto · 2006-06-24 00:13:04

ive got a weird" modules.dep" problem. at boot he complains that "......./modules.dep" doesnt exist. i dont have the exact path right now, im on another pc, but i cant load any modules bcoz of this error, and network, mouse, and nvidia driver dont work now! :shock: :?
i only did a "pacman -Syu".
any ideas?

best regards,
detto

Arch Linux

#26 2006-06-23 20:40:14

Re: The Testing Move has begun

#27 2006-06-23 20:47:22

Re: The Testing Move has begun

#28 2006-06-23 20:51:37

Re: The Testing Move has begun

#29 2006-06-23 20:53:59

Re: The Testing Move has begun

#30 2006-06-23 20:54:49

Re: The Testing Move has begun

#31 2006-06-23 20:54:53

Re: The Testing Move has begun

#32 2006-06-23 20:58:51

Re: The Testing Move has begun

#33 2006-06-23 21:03:41

Re: The Testing Move has begun

#34 2006-06-23 21:06:48

Re: The Testing Move has begun

#35 2006-06-23 21:08:21

Re: The Testing Move has begun

#36 2006-06-23 21:16:35

Re: The Testing Move has begun

#37 2006-06-23 21:18:55

Re: The Testing Move has begun

#38 2006-06-23 21:19:00

Re: The Testing Move has begun

#39 2006-06-23 21:20:15

Re: The Testing Move has begun

#40 2006-06-23 21:25:44

Re: The Testing Move has begun

#41 2006-06-23 21:34:01

Re: The Testing Move has begun

#42 2006-06-23 21:43:35

Re: The Testing Move has begun

#43 2006-06-23 21:51:30

Re: The Testing Move has begun

#44 2006-06-23 21:56:09

Re: The Testing Move has begun

#45 2006-06-23 22:18:05

Re: The Testing Move has begun

#46 2006-06-23 22:46:29

Re: The Testing Move has begun

#47 2006-06-24 00:05:47

Re: The Testing Move has begun

#48 2006-06-24 00:10:18

Re: The Testing Move has begun

#49 2006-06-24 00:12:53

Re: The Testing Move has begun

#50 2006-06-24 00:13:04

Re: The Testing Move has begun

Board footer