Gentoo Forums
Gentoo Forums
Gentoo Forums
Quick Search: in
[einit] 0.21.0 -- now with gentoo-compat-0.99.0 (superseded)
View unanswered posts
View posts from last 24 hours

Goto page Previous  1, 2, 3, 4, 5, 6  Next  
Reply to topic    Gentoo Forums Forum Index Unsupported Software
View previous topic :: View next topic  
Author Message
jvale
n00b
n00b


Joined: 23 Jan 2006
Posts: 74
Location: Porto, Portugal

PostPosted: Tue Apr 03, 2007 11:49 pm    Post subject: Reply with quote

You missed the 'entrance' and 'entranced' modules when you split the configuration files. ;P

Cheers!

EDIT:
Oh, and I just emerged from SVN and the "mount"-module tries to acess my (empty) floppy drive like crazy on startup.
Back to top
View user's profile Send private message
modified_bessel
Tux's lil' helper
Tux's lil' helper


Joined: 16 Jul 2004
Posts: 101
Location: Cote d'Azur, France

PostPosted: Wed Apr 04, 2007 8:46 am    Post subject: Reply with quote

rmh3093 wrote:
modified_bessel wrote:
Now it is clear that with the X/KDE thing fixed (9 seconds lost!), this will be a spectacular boot! (~27s to KDE ??...) (and all this is on a flash card!)

this also happens with gdm.... its been pissing me off, I cant figure it out


Code:

  <prefer>
   <xdm s="daemon-xdm" />
   <syslog s="daemon-metalog" />
  </prefer>


along with disabling the dbus service in the service list has essentially done the trick! Although the X/KDE load itself is still 4 seconds slower than that provided by a conventional boot, it is a remarkable total improvement! (~31 seconds, grub to fully loaded KDE!) PLUS: no more arbitrary reboots while loading X/KDE....
Back to top
View user's profile Send private message
mdeininger
Veteran
Veteran


Joined: 15 Jun 2005
Posts: 1740
Location: Emerald Isles, observing Dublin's docklands

PostPosted: Wed Apr 04, 2007 9:34 am    Post subject: Reply with quote

modified_bessel:
so it IS probably a problem with dbus and either of the *DMs... wonder what we're doing wrong for dbus =)

jvale:
i think ryan put those back in later that night ;)

there must be something wrong with the bootup scans in the mounter, takes ages on my box (i get a delay of pretty much exactly 30 seconds somewhere, and i can't figure out for the life of me where the hell that delay comes from... might be the mounter doing stuff it's not supposed to :D)
_________________
"Confident, lazy, cocky, dead." -- Felix Jongleur, Otherland

( Twitter | Blog | GitHub )
Back to top
View user's profile Send private message
rmh3093
Advocate
Advocate


Joined: 06 Aug 2003
Posts: 2138
Location: Albany, NY

PostPosted: Wed Apr 04, 2007 1:55 pm    Post subject: Reply with quote

mdeininger wrote:
modified_bessel:
so it IS probably a problem with dbus and either of the *DMs... wonder what we're doing wrong for dbus =)

jvale:
i think ryan put those back in later that night ;)

there must be something wrong with the bootup scans in the mounter, takes ages on my box (i get a delay of pretty much exactly 30 seconds somewhere, and i can't figure out for the life of me where the hell that delay comes from... might be the mounter doing stuff it's not supposed to :D)

mine hangs on `creating device nodes` for ages
_________________
Do not meddle in the affairs of wizards, for they are subtle and quick to anger.
Back to top
View user's profile Send private message
mdeininger
Veteran
Veteran


Joined: 15 Jun 2005
Posts: 1740
Location: Emerald Isles, observing Dublin's docklands

PostPosted: Wed Apr 04, 2007 3:17 pm    Post subject: Reply with quote

rmh3093 wrote:
mine hangs on `creating device nodes` for ages

you mean, it's stuck there every now and then for a batch of releases already, or it's taking a lot of time at that step? ;)
_________________
"Confident, lazy, cocky, dead." -- Felix Jongleur, Otherland

( Twitter | Blog | GitHub )
Back to top
View user's profile Send private message
rmh3093
Advocate
Advocate


Joined: 06 Aug 2003
Posts: 2138
Location: Albany, NY

PostPosted: Wed Apr 04, 2007 3:18 pm    Post subject: Reply with quote

mdeininger wrote:
rmh3093 wrote:
mine hangs on `creating device nodes` for ages

you mean, it's stuck there every now and then for a batch of releases already, or it's taking a lot of time at that step? ;)


I mean it takes a lot of time at that step, often....

... and I still dont see any fsck feedback :wink:
_________________
Do not meddle in the affairs of wizards, for they are subtle and quick to anger.
Back to top
View user's profile Send private message
mdeininger
Veteran
Veteran


Joined: 15 Jun 2005
Posts: 1740
Location: Emerald Isles, observing Dublin's docklands

PostPosted: Wed Apr 04, 2007 4:04 pm    Post subject: Reply with quote

rmh3093 wrote:
mdeininger wrote:
rmh3093 wrote:
mine hangs on `creating device nodes` for ages

you mean, it's stuck there every now and then for a batch of releases already, or it's taking a lot of time at that step? ;)


I mean it takes a lot of time at that step, often....

... and I still dont see any fsck feedback :wink:
ah i get it =)

err, it *should* say that it's now doing an fsck........... :)
_________________
"Confident, lazy, cocky, dead." -- Felix Jongleur, Otherland

( Twitter | Blog | GitHub )
Back to top
View user's profile Send private message
rmh3093
Advocate
Advocate


Joined: 06 Aug 2003
Posts: 2138
Location: Albany, NY

PostPosted: Thu Apr 05, 2007 12:24 am    Post subject: Reply with quote

Alright everyone, there has been a big change in the network templates to listen up this effect you all!

NOTE: this is not 100% perfect (but still functional), things will probably change regarding how this all works but how interfaces are configured will try to remain constant

Starting with einit-9999-r18
1) depends on sys-apps/iproute2
2) all of the old net templates are obsolete, remove any reference to them from local.xml, they might break things
3) there is no more need to hack together modules to provide a particular net-, they are generated on the fly

Configuring network interfaces is as simple as the examples below.
There are 3 required componets, 1. the interface name, 2. the control= attribute and 3. the ip= attribute.
[*]control= specifies the manager for the network interfaces, possible options now are (none, wpa_supplicant). I will make an iwconfig template soon.
[*]ip= specifies how the network interface should gets its ip address, possible options now are (dhcpcd, pump, dhclient, iproute2), with iproute2 for static address
[*]kernel-module= specifies the required kernel module for the network interface, this is optional

All other options an interface might need are configured within the interfaces brackets.

EXAMPLE
local.xml:
   <interfaces>
    <lo control="none" ip="iproute2" address="127.0.0.1/8" network="127.0.0.0/8" />
    <eth0 control="none" ip="pump" kernel-module="e1000" />
    <wlan0 control="wpa_supplicant" driver="wext" config="/etc/wpa_supplicant/wpa_supplicant.conf" ip="pump" kernel-module="ipw3945" />
   </interfaces>

_________________
Do not meddle in the affairs of wizards, for they are subtle and quick to anger.
Back to top
View user's profile Send private message
modified_bessel
Tux's lil' helper
Tux's lil' helper


Joined: 16 Jul 2004
Posts: 101
Location: Cote d'Azur, France

PostPosted: Thu Apr 05, 2007 3:33 pm    Post subject: Reply with quote

Don't know if this is simple or not, but it's about the last thing holding up my system from being 100% functional under einit.

I have to call some serial ports and a touchscreen driver; under sysvinit the following works just like I need it to:

/etc/conf.d/local.start
Code:

# activate the raw serio port for the touchscreen:
echo -n "serio_raw" > /sys/bus/serio/devices/serio1/drvctl
mknod --mode 660 /dev/touchscreen1 c 10 62
# change serial port permissions for awhile:
chmod 777 /dev/ttyS*


None of these are executed during the start of the einit "local" service, so I guess I need to write a module?
Shouldn't be difficult?

Thanks for your consideration!
Back to top
View user's profile Send private message
modified_bessel
Tux's lil' helper
Tux's lil' helper


Joined: 16 Jul 2004
Posts: 101
Location: Cote d'Azur, France

PostPosted: Thu Apr 05, 2007 5:44 pm    Post subject: Reply with quote

Typo, last line? Vairables --> Variables ?? Line 338

/lib/einit/einit.xml v. 0.21
Code:
<!-- modified version of the gpm module by teppic from the wiki -->
<!-- as has been suggested by him, i have modified this to use variables, hope it works... -->
    <shell id="shell-gpm"
     name="Console mouse services (GPM)"
     provides="gpm"
     requires="tty"
     enable="/usr/sbin/gpm -m ${configuration_peripherals_pointing_device_gpm_device} -t ${configuration_peripherals_pointing_device_gpm_type}"
     disable="killall gpm"
     restart="yes"
     vairables="configuration-peripherals-pointing-device-gpm/.*" />
Back to top
View user's profile Send private message
rmh3093
Advocate
Advocate


Joined: 06 Aug 2003
Posts: 2138
Location: Albany, NY

PostPosted: Thu Apr 05, 2007 6:22 pm    Post subject: Reply with quote

modified_bessel wrote:
Don't know if this is simple or not, but it's about the last thing holding up my system from being 100% functional under einit.

I have to call some serial ports and a touchscreen driver; under sysvinit the following works just like I need it to:

/etc/conf.d/local.start
Code:

# activate the raw serio port for the touchscreen:
echo -n "serio_raw" > /sys/bus/serio/devices/serio1/drvctl
mknod --mode 660 /dev/touchscreen1 c 10 62
# change serial port permissions for awhile:
chmod 777 /dev/ttyS*


None of these are executed during the start of the einit "local" service, so I guess I need to write a module?
Shouldn't be difficult?

Thanks for your consideration!


Ahh, thanks for reminding me that einit does not have an "local" module for executing cli commands like that, I will get on that, look for it in the svn later today

....however, you probably want to create udev rules for the touchscreen module so that it creates the device node and set group permissions when you modprobe touchscreen
_________________
Do not meddle in the affairs of wizards, for they are subtle and quick to anger.
Back to top
View user's profile Send private message
modified_bessel
Tux's lil' helper
Tux's lil' helper


Joined: 16 Jul 2004
Posts: 101
Location: Cote d'Azur, France

PostPosted: Thu Apr 05, 2007 7:56 pm    Post subject: Reply with quote

rmh3093 wrote:
modified_bessel wrote:
Don't know if this is simple or not, but it's about the last thing holding up my system from being 100% functional under einit.

I have to call some serial ports and a touchscreen driver; under sysvinit the following works just like I need it to:

/etc/conf.d/local.start
Code:

# activate the raw serio port for the touchscreen:
echo -n "serio_raw" > /sys/bus/serio/devices/serio1/drvctl
mknod --mode 660 /dev/touchscreen1 c 10 62
# change serial port permissions for awhile:
chmod 777 /dev/ttyS*


None of these are executed during the start of the einit "local" service, so I guess I need to write a module?
Shouldn't be difficult?

Thanks for your consideration!


Ahh, thanks for reminding me that einit does not have an "local" module for executing cli commands like that, I will get on that, look for it in the svn later today

....however, you probably want to create udev rules for the touchscreen module so that it creates the device node and set group permissions when you modprobe touchscreen


Infinite thanks for looking at the "local module"; will check the SVM ..........

Yes you are right, but the problem here is that there are no kernel modules for this particular touchscreen; hence I am using the raw serio1 device to intercept the signals, then use an X-module to translate the intercepted signals via the device node created above. The /dev/ttyS* permissions are for yet another serial device which I access through python serial-2.2 modules.

<bump> I think the typo mentioned earlier is relevant for correct operation of the gpm module!
Back to top
View user's profile Send private message
mdeininger
Veteran
Veteran


Joined: 15 Jun 2005
Posts: 1740
Location: Emerald Isles, observing Dublin's docklands

PostPosted: Thu Apr 05, 2007 8:50 pm    Post subject: Reply with quote

heh, i think you're right -- sorry, been at the gym the last three hours or so ;)
_________________
"Confident, lazy, cocky, dead." -- Felix Jongleur, Otherland

( Twitter | Blog | GitHub )
Back to top
View user's profile Send private message
mdeininger
Veteran
Veteran


Joined: 15 Jun 2005
Posts: 1740
Location: Emerald Isles, observing Dublin's docklands

PostPosted: Thu Apr 05, 2007 10:40 pm    Post subject: Reply with quote

something else, i changed the services with mount/* to mount-*

pros: you can now create modules to take their part, and prefer them over the compiled modules; cons, you need to change your configuration and also /CUSTOM MODULES/ to reflect this
;)

so: change the following:
mount/system -> mount-system
mount/local -> mount-local
mount/critical -> mount-critical
mount/remote -> mount-remote
_________________
"Confident, lazy, cocky, dead." -- Felix Jongleur, Otherland

( Twitter | Blog | GitHub )
Back to top
View user's profile Send private message
UberPinguin
Guru
Guru


Joined: 20 Nov 2005
Posts: 510
Location: 2416.94 Miles From Home

PostPosted: Fri Apr 06, 2007 6:08 pm    Post subject: Reply with quote

I updated einit last night, but now at boot I get this:
Code:
eINIT 0.21.1-svn-1081: Initializing: Linux
 >> initializing in-core modules:[module-so]Floppy drive(s):fd0 is 1.44m
 FDC0 is a post-1991 82077
end_request: I/O error, dev fd0 sector 0
Buffer I/O error on device fd0p124, logical block 0
It then increments through several sectors and blocks before incrementing the fd0p device and starting over at sector 0/block 0; as far as I can tell, this continues indefinitely, though I only let it run for about 5 minutes.
_________________
aidanjt wrote:
You see, instead of arguing from ignorance, and fear, there is only one way to verify a theory. And that's not by clutching a black book and begging the sky fairy for deliverance from the mad scientists and their big machines.
Back to top
View user's profile Send private message
modified_bessel
Tux's lil' helper
Tux's lil' helper


Joined: 16 Jul 2004
Posts: 101
Location: Cote d'Azur, France

PostPosted: Fri Apr 06, 2007 6:36 pm    Post subject: Reply with quote

and I got, for (einit-9999-r19.ebuild)

Code:
eINIT 0.21.1-svn-1087: Initializing: Linux
 >> initializing in-core modules:[module-so]eINIT terminated by a signal...
Kernel panic-not syncing: Attempted to kill init


I managed to hack in the new udev, net modules, and other new modules into an earlier ebuild, and my goodness; the kernel boot is now on the order of 3 seconds (without X, wherein kdm doesnt start (neither do the login tty's...) ! but on the positive side, really unbelievable speed for the rest .........

Really (desperately?) need a new ebuild with the new modules you have improved in the last day (like the new local.xml, THANKS rmh3093 and mdeininger).

Keep it up!
Back to top
View user's profile Send private message
rmh3093
Advocate
Advocate


Joined: 06 Aug 2003
Posts: 2138
Location: Albany, NY

PostPosted: Fri Apr 06, 2007 8:10 pm    Post subject: Reply with quote

magnus: ever since you moved module-so from internal to external things have been weird I think
_________________
Do not meddle in the affairs of wizards, for they are subtle and quick to anger.
Back to top
View user's profile Send private message
modified_bessel
Tux's lil' helper
Tux's lil' helper


Joined: 16 Jul 2004
Posts: 101
Location: Cote d'Azur, France

PostPosted: Sat Apr 07, 2007 7:04 am    Post subject: Reply with quote

0.21.1-svn-1094 now (somewhat) duplicating the 1081 results of UberPenguin. tty12 in an endless loop reporting udevd[16259] error. tty11 shows many broken services (dbus, hal, xdm, alsa, logging .......). After several minutes, can finally get to a login terminal, but the endless udevd loop is still active, and the filesystem is all Read-Only. /dev shows many tty devices, but no hard disks, etc....
Back to top
View user's profile Send private message
mdeininger
Veteran
Veteran


Joined: 15 Jun 2005
Posts: 1740
Location: Emerald Isles, observing Dublin's docklands

PostPosted: Sat Apr 07, 2007 5:23 pm    Post subject: Reply with quote

might be something about the new udev module... i added a dont-close-stdin option, that could fix it (no stdin has a tendency to kill udev utils... ohwell :-/)

if udev keeps dying, then maybe that's due to that stdin problem, just commited a fix for that. alternatively, it might be it needs >= udev-093...

as for the fd0... that's kind of weird, did you try the fbsplash plugin? (fb0 can be both a floppy and the framebuffer i think... meh..)

rmh3093: didn't move module-so from internal to external, actually i specifically made sure that module-so is built into the core in the build script...
_________________
"Confident, lazy, cocky, dead." -- Felix Jongleur, Otherland

( Twitter | Blog | GitHub )
Back to top
View user's profile Send private message
rmh3093
Advocate
Advocate


Joined: 06 Aug 2003
Posts: 2138
Location: Albany, NY

PostPosted: Sun Apr 08, 2007 1:20 pm    Post subject: Reply with quote

is there anyone still having udev issues?
_________________
Do not meddle in the affairs of wizards, for they are subtle and quick to anger.
Back to top
View user's profile Send private message
modified_bessel
Tux's lil' helper
Tux's lil' helper


Joined: 16 Jul 2004
Posts: 101
Location: Cote d'Azur, France

PostPosted: Sun Apr 08, 2007 10:18 pm    Post subject: Reply with quote

I upgraded from udev-087 to 104-r11 and that seemed to work; it is somewhat slower than 087 was under einit-0.21.0, but it seens to load fine now; no more endless floppy and hard disk loops....

Any advice on getting the network back up? I have tried endless permutations with the new network module (9999-r20, latest svn as of last night). can not get eth0 up under any circumstance; net-lo reports:
Code:

[ENAB] Network Interface (lo) RTNETLINK answers: Invalid argument

ifconfig shows it's up, but not eth0 (even tried to configure an eth1, and add it to the group as described; no luck)

I have iproute2 and dhcpcd installed, and am using a blank /etc/conf.d/net configuration; works fine under sysvinit.
Back to top
View user's profile Send private message
jvale
n00b
n00b


Joined: 23 Jan 2006
Posts: 74
Location: Porto, Portugal

PostPosted: Mon Apr 09, 2007 12:19 am    Post subject: Reply with quote

About udev:
eINIT: 0.21.1-svn-1107
udev: 104-r12

When booting, it hangs for 10 seconds when enabling udev (got the same 10 seconds in several runs).


When enabling eth0, strangely enough dhcpcd is not setting the default gateway. What puzzles me is that dhcpcd is working fine under Init, but under eINIT it doesn't set the default gateway, neither by enabling net-eth0 nor calling it directly ('dhcpcd eth0'). Gonna look into it to see if it's some freak-problem in my setup as it seems strange that eINIT would cause this. This might be the same issue as modified_bessel tough.

Update:
Did some more testing and the dhcpcd issue stands.
Back to top
View user's profile Send private message
UberPinguin
Guru
Guru


Joined: 20 Nov 2005
Posts: 510
Location: 2416.94 Miles From Home

PostPosted: Mon Apr 09, 2007 2:55 am    Post subject: Reply with quote

mdeininger wrote:
might be something about the new udev module... i added a dont-close-stdin option, that could fix it (no stdin has a tendency to kill udev utils... ohwell :-/)

if udev keeps dying, then maybe that's due to that stdin problem, just commited a fix for that. alternatively, it might be it needs >= udev-093...
I have udev-108, and was previously on 107-r1. Have not had any issues with udev dying, so this may be correct.
mdeininger wrote:
as for the fd0... that's kind of weird, did you try the fbsplash plugin? (fb0 can be both a floppy and the framebuffer i think... meh..)
No, I haven't had the opportunity to play with the framebuffer splash yet. I'm nearly done with an emerge -uDN world, so once that's finished I'll be able to do some more testing. An interesting note, after updating from udev 107-r1 to 108, the fd0 message seems to have vanished, but einit still hangs. Hitting sysrq-p shows that it's stuck on modprobing something, but I can't see what. More info as it's available.
_________________
aidanjt wrote:
You see, instead of arguing from ignorance, and fear, there is only one way to verify a theory. And that's not by clutching a black book and begging the sky fairy for deliverance from the mad scientists and their big machines.
Back to top
View user's profile Send private message
modified_bessel
Tux's lil' helper
Tux's lil' helper


Joined: 16 Jul 2004
Posts: 101
Location: Cote d'Azur, France

PostPosted: Mon Apr 09, 2007 7:03 am    Post subject: Reply with quote

I am getting precisely the same results as jvale; same hang (~10s) on udev, and same problem with setting the gateway, eth0, etc.

I have reinstalled back to einit-0.21.0, same system as before except with udev 104 instead of the previous 087, and the results are just as with 087: no 10s hang on udev, mounts fast, gateway and eth0 activated correctly; no problems!
Back to top
View user's profile Send private message
mdeininger
Veteran
Veteran


Joined: 15 Jun 2005
Posts: 1740
Location: Emerald Isles, observing Dublin's docklands

PostPosted: Mon Apr 09, 2007 8:20 am    Post subject: Reply with quote

heh, i'll be able to do more thorough tests again today, so i'll probably be able to figure out what's wrong later today... lots of stuff did change, so sorry if it broke stuff :-/
_________________
"Confident, lazy, cocky, dead." -- Felix Jongleur, Otherland

( Twitter | Blog | GitHub )
Back to top
View user's profile Send private message
Display posts from previous:   
Reply to topic    Gentoo Forums Forum Index Unsupported Software All times are GMT
Goto page Previous  1, 2, 3, 4, 5, 6  Next
Page 3 of 6

 
Jump to:  
You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot vote in polls in this forum