| View previous topic :: View next topic |
| Author |
Message |
lflores
n00b
Joined: 03 Aug 2005
Posts: 3
|
 Posted: Fri Sep 02, 2005 4:18 pm Post subject: Cannot open root device "ram0" |
 |
|
I can generate succesfully my stage2..
It creates the iso image....
But when i use it (in vmware) it starts loading and it stucks...
I keep getting the following error:
VFS - Cannot open root device "ram0" or unkown-block(0,0)
Please append a coorect "root=" boot option
Kernel panic - not syncing: VFS unable to mount root fs on unkown-block(0,0)
In my stage2 spec file the cdtar is set to /usr/lib/catalyst/livecd/cdtar/isolinux-3.09-cdtar.tar.bz2 but i have tried too with isolinux-3.09-memtest86+-cdtar.tar.bz2.
What am i missing??? |
|
| Back to top |
|
 |
MrWonderful
n00b
Joined: 30 Jul 2005
Posts: 10
Location: Bartonville, Illinois
|
| Posted: Sat Sep 10, 2005 8:09 am Post subject: |
|
|
I started a thread here. I can't understand it. I'm pretty sure this is because I'm using devfs instead of udev. I have 2.4 kernel, so udev is out of the question though. I'll post whatever anyone needs. I've got a week before I get this finished or my all-gentoo LUG demo will be useless, and I'll have to go with a pre-made cluster cd.
>> Activating devfs
mount: Mounting devfs on /dev failed: No such device
Error opening file: ".devfsd" No such file or directory
>> Making tmpfs for /newroot
>> No bootable medium found. Waiting for new devices...
!! Could not find CD to boot, something else needed!
>> Determining root device...
!! The root block device is unspecified or not detected.
Please specify a device to boot, or "shell" for a shell... |
|
| Back to top |
|
|
Nodlehs
n00b
Joined: 21 May 2005
Posts: 3
|
| Posted: Thu Sep 15, 2005 8:35 pm Post subject: Gensplash |
|
|
What is the correct way to get gensplash working with a livecd using catalyst-1.1.10.10?
My livecd stage 1 builds fine, but when I attempt to build my stage 2 I get an error...
| Quote: | | Error, cannot setup splash theme livecd-2005.1 |
relevant parts from my livecd stage 2 spec file
| Quote: |
livecd/splash_type: gensplash
livecd/splash_theme: livecd-2005.1
boot/kernel/gentoo/packages: splashutils splash-themes-livecd
|
Am I missing something? When I check the stage2 files it creates there is no /etc/splash directory where splash-themes-livecd would put the livecd-2005.1 so I can understand it not working and getting the error I specified. But I have it set up to emerge the splash-themes-livecd. |
|
| Back to top |
|
|
drescherjm
Advocate
Joined: 05 Jun 2004
Posts: 2790
Location: Pittsburgh, PA, USA
|
| Posted: Thu Feb 02, 2006 4:41 am Post subject: |
|
|
Anyone on earth know why catalyst does not like my distfiles folder mounted at /mnt/vg/distfiles? I also tried /usr/portage/distfiles and made that a normal folder instead of the symlink but got the same error with both catalyst1 and catalyst2.
| Code: | catalyst2 -f livecd-stage1_template.spec
Gentoo Catalyst, version 2.0_rc26
Copyright 2003-2005 The Gentoo Foundation
Distributed under the GNU General Public License version 2
Using default Catalyst configuration file, /etc/catalyst2/catalyst2.conf
Setting sharedir to config file value "/usr/lib/catalyst2"
Setting snapshot_cache to default value "/var/tmp/catalyst2/snapshot_cache"
Setting hash_function to config file value "md5"
Setting storedir to config file value "/var/tmp/catalyst2"
Setting portdir to config file value "/usr/portage"
Setting distdir to config file value "/mnt/vg/distfiles/"
Setting options to config file value "autoresume ccache kerncache pkgcache seedcache snapcache"
Compiler cache support enabled.
Package cache support enabled.
Snapshot cache support enabled.
Seed cache support enabled.
Kernel cache support enabled.
Autoresuming support enabled.
Using target: livecd-stage1
Building natively for x86
Source path set to /var/tmp/catalyst2/builds/default/stage3-x86-2005.1.tar.bz2
Caching snapshot to /var/tmp/catalyst2/snapshot_cache/2006.0/
The autoresume path is /var/tmp/catalyst2/tmp/default/.autoresume-livecd-stage1-athlon-xp-2006.0/
Location of the package cache is /tmp/packages
Location of the kerncache is /var/tmp/catalyst2/kerncache/default/livecd-stage1-athlon-xp-2006.0/
Checking for processes running in chroot and killing them.
Running action sequence: unpack
Resume point detected, skipping unpack operation...
Running action sequence: unpack_snapshot
Valid snapshot cache, skipping unpack of portage tree ...
Running action sequence: config_profile_link
Resume point detected, skipping config_profile_link operation...
Running action sequence: setup_confdir
Resume point detected, skipping setup_confdir operation...
Running action sequence: portage_overlay
Running action sequence: bind
Running action sequence: chroot_setup
Resume point detected, skipping chroot_setup operation...
Running action sequence: setup_environment
Running action sequence: build_packages
copying livecd-stage1-chroot.sh to /var/tmp/catalyst2/tmp/default/livecd-stage1-athlon-xp-2006.0/tmp/
copying chroot-functions.sh to /var/tmp/catalyst2/tmp/default/livecd-stage1-athlon-xp-2006.0/tmp/
Running livecd-stage1-chroot.sh in chroot /var/tmp/catalyst2/tmp/default/livecd-stage1-athlon-xp-2006.0/
>>> Regenerating /etc/ld.so.cache...
Calculating dependencies ...done!
>>> emerge (1 of 1) dev-util/ccache-2.3 to /
!!! File system problem. (Bad Symlink?)
!!! Fetching may fail: [Errno 17] File exists: '/usr/portage/distfiles'
!!! No write access to /usr/portage/distfiles/
!!! File ccache-2.3.tar.gz isn't fetched but unable to get it.
!!! catalyst: Error in attempt to build packages
Traceback (most recent call last):
File "modules/generic_stage_target.py", line 1152, in build_packages
" build_packages "+mypack,"Error in attempt to build packages",env=self.env)
File "/usr/lib/catalyst2/modules/catalyst_support.py", line 485, in cmd
raise CatalystError,myexc
CatalystError: <unprintable instance object>
None
!!! catalyst: livecdbuild aborting due to error.
Catalyst aborting....
|
_________________
John
My gentoo overlay
Instructons for overlay |
|
| Back to top |
|
|
drescherjm
Advocate
Joined: 05 Jun 2004
Posts: 2790
Location: Pittsburgh, PA, USA
|
| Posted: Thu Feb 02, 2006 8:47 pm Post subject: |
|
|
With the help of a coworker I solved that one. When I made a portage snapshot it copied the link to /mnt/vg/distfiles so in the chroot jail this folder did not exist so their was no change of emerging anything with an invalid distfiles folder.
_________________
John
My gentoo overlay
Instructons for overlay |
|
| Back to top |
|
|
linoleum
Tux's lil' helper
Joined: 15 Nov 2004
Posts: 129
|
| Posted: Sun Feb 19, 2006 11:23 am Post subject: genpacthes-2.6.12-10.extras.tar.bz2 |
|
|
Hi,
I'm trying to create my livecd (with X) and I have the following problem with genpatches-2.6.12-10.extras.tar.bz2:
I tryied to add a
GENTOO_MIRRORS="..."
in /var/tmp/catalyst/tmp/default/livecd-stage1-x86-2005.1/etc/make.conf without success.
Any help?
Thanks
Here my stage1.spec
| Quote: |
subarch: x86
version_stamp: 2005.1
target: livecd-stage1
rel_type: default
profile: default-linux/x86/2005.1
snapshot: 20060216
source_subpath: default/stage3-x86-2005.1
livecd/use:
-kde
-doc
-arts
-gnome
-esd
-qt
ipv6
socks5
livecd
fbcon
ncurses
readline
ssl
atm
png
gpm
atlas
blas
lapack
livecd/packages:
livecd-tools
gentoo-sources
dhcpcd
acpid
apmd
coldplug
fxload
irssi
gpm
syslog-ng
parted
links
raidtools
dosfstools
nfs-utils
jfsutils
xfsprogs
alsa-utils
e2fsprogs
reiserfsprogs
ntfsprogs
pwgen
popt
dialog
rp-pppoe
screen
mirrorselect
penggy
iputils
hwsetup
lvm2
evms
vim
pptpclient
mdadm
ethtool
expat
cairo
xfce4
wireless-tools
# prism54-firmware
zd1201-firmware
|
and my stage2
| Quote: |
subarch: x86
version_stamp: 2005.1
target: livecd-stage2
rel_type: default
profile: default-linux/x86/2005.1
snapshot: official
source_subpath: default/livecd-stage1-x86-2005.1
livecd/cdfstype: squashfs
livecd/archscript: /usr/lib/catalyst/livecd/runscript/x86-archscript.sh
livecd/runscript: /usr/lib/catalyst/livecd/runscript/default-runscript.sh
livecd/cdtar: /usr/lib/catalyst/livecd/cdtar/isolinux-3.09-memtest86+-cdtar.tar.bz2
livecd/iso: /usr/local/bose-x86-2005.1-universal.iso
livecd/splash_type: gensplash
livecd/splash_theme: livecd-2005.1
livecd/type: gentoo-release-universal
livecd/modblacklist: 8139cp
#livecd/overlay: /usr/portage
livecd/bootargs: dokeymap
livecd/gk_mainargs: --lvm2 --dmraid --evms2
boot/kernel: gentoo
boot/kernel/gentoo/sources: gentoo-sources
#boot/kernel/gentoo/sources: =sys-kernel/gentoo-sources-2.6.15-r1
boot/kernel/gentoo/config: /usr/share/livecd-kconfigs-2005.1/x86/2.6.12-smp.config
#boot/kernel/gentoo/config: /usr/src/linux-2.6.15-gentoo-r1/.config
boot/kernel/gentoo/use: pcmcia usb png truetype
boot/kernel/gentoo/packages:
splashutils
splash-themes-livecd
pcmcia-cs
# Removed in favor of in-kernel drivers
# speedtouch
slmodem
globespan-adsl
hostap-driver
hostap-utils
# These versions have been edited to work together for the release.
# =ipw2100-1.0.5
# =ipw2200-1.0.1
fritzcapi
fcdsl
cryptsetup
# at76c503a
# rt2500
# rtl8180
# adm8211
# acx100
orinoco
livecd/unmerge:
acl
addpatches
attr
autoconf
automake
bc
bin86
binutils
bison
bison
ccache
cpio
cronbase
diffutils
distcc
ed
# expat
# flex
gcc
gcc-config
gcc-sparc64
genkernel
gentoo-sources
gettext
gnuconfig
groff
grub
help2man
lcms
ld.so
ld.so
lib-compat
libmng
libperl
libtool
linux-headers
m4
make
man
man-pages
miscfiles
patch
perl
rsync
sash
sysklogd
texinfo
ucl
vanilla-sources
livecd/empty:
/etc/bootsplash/gentoo
/etc/bootsplash/gentoo-highquality
/etc/cron.daily
/etc/cron.hourly
/etc/cron.monthly
/etc/cron.weekly
/etc/logrotate.d
/etc/rsync
/etc/skel
/etc/splash/emergence
/etc/splash/gentoo
/root/.ccache
/tmp
/usr/diet/include
/usr/diet/man
/usr/i386-gentoo-linux-uclibc
/usr/i386-pc-linux-gnu
/usr/i386-pc-linux-uclibc
/usr/include
/usr/lib/X11/config
/usr/lib/X11/doc
/usr/lib/X11/etc
/usr/lib/awk
/usr/lib/ccache
/usr/lib/gcc-config
/usr/lib/gconv
/usr/lib/nfs
/usr/lib/perl5
/usr/lib/portage
/usr/lib/python2.2
/usr/local
/usr/portage
/usr/share/aclocal
/usr/share/baselayout
/usr/share/consolefonts/partialfonts
/usr/share/consoletrans
/usr/share/dict
/usr/share/doc
/usr/share/emacs
/usr/share/et
/usr/share/gcc-data
/usr/share/genkernel
/usr/share/gettext
/usr/share/glib-2.0
/usr/share/gnuconfig
/usr/share/gtk-doc
/usr/share/i18n
/usr/share/info
/usr/share/lcms
/usr/share/locale
/usr/share/man
/usr/share/perl
/usr/share/rfc
/usr/share/ss
/usr/share/state
/usr/share/texinfo
/usr/share/unimaps
/usr/share/zoneinfo
/usr/sparc-unknown-linux-gnu
/usr/src
/var/cache
/var/db
/var/empty
/var/lib/portage
/var/lock
/var/log
/var/run
/var/spool
/var/state
/var/tmp
livecd/rm:
/boot/System*
/boot/initr*
/boot/kernel*
/etc/*-
/etc/*.old
/etc/default/audioctl
/etc/dispatch-conf.conf
/etc/env.d/05binutils
/etc/env.d/05gcc
/etc/etc-update.conf
/etc/hosts.bck
/etc/issue*
/etc/genkernel.conf
/etc/make.conf
/etc/make.conf.example
/etc/make.globals
/etc/make.profile
/etc/man.conf
/etc/resolv.conf
/etc/splash/livecd-2005.1/12*
/etc/splash/livecd-2005.1/14*
/etc/splash/livecd-2005.1/16*
/etc/splash/livecd-2005.1/19*
/etc/splash/livecd-2005.1/6*
/etc/splash/livecd-2005.1/8*
/etc/splash/livecd-2005.1/images/background-12*
/etc/splash/livecd-2005.1/images/background-14*
/etc/splash/livecd-2005.1/images/background-16*
/etc/splash/livecd-2005.1/images/background-19*
/etc/splash/livecd-2005.1/images/background-6*
/etc/splash/livecd-2005.1/images/background-8*
/etc/splash/livecd-2005.1/images/verbose-12*
/etc/splash/livecd-2005.1/images/verbose-14*
/etc/splash/livecd-2005.1/images/verbose-16*
/etc/splash/livecd-2005.1/images/verbose-19*
/etc/splash/livecd-2005.1/images/verbose-6*
/etc/splash/livecd-2005.1/images/verbose-8*
/lib/*.a
/lib/security/pam_access.so
/lib/security/pam_chroot.so
/lib/security/pam_debug.so
/lib/security/pam_ftp.so
/lib/security/pam_issue.so
/lib/security/pam_mail.so
/lib/security/pam_mkhomedir.so
/lib/security/pam_motd.so
/lib/security/pam_postgresok.so
/lib/security/pam_rhosts_auth.so
/lib/security/pam_userdb.so
/root/.viminfo
/sbin/fsck.cramfs
/sbin/fsck.minix
/sbin/mkfs.bfs
/sbin/mkfs.cramfs
/sbin/mkfs.minix
/usr/bin/addr2line
/usr/bin/ar
/usr/bin/as
/usr/bin/audioctl
/usr/bin/c++*
/usr/bin/elftoaout
/usr/bin/gprof
/usr/bin/i386-gentoo-linux-uclibc-*
/usr/bin/i386-pc-linux-*
/usr/bin/ld
/usr/bin/nm
/usr/bin/objcopy
/usr/bin/objdump
/usr/bin/piggyback*
/usr/bin/ranlib
/usr/bin/readelf
/usr/bin/size
/usr/bin/sparc-unknown-linux-*
/usr/bin/sparc64-unknown-linux-*
/usr/bin/strings
/usr/bin/strip
/usr/lib/*.a
/usr/lib/gcc-lib/*/*/libgcj*
/usr/sbin/bootsplash*
/usr/sbin/fb*
/usr/share/consolefonts/1*
/usr/share/consolefonts/7*
/usr/share/consolefonts/8*
/usr/share/consolefonts/9*
/usr/share/consolefonts/A*
/usr/share/consolefonts/C*
/usr/share/consolefonts/E*
/usr/share/consolefonts/G*
/usr/share/consolefonts/L*
/usr/share/consolefonts/M*
/usr/share/consolefonts/R*
/usr/share/consolefonts/a*
/usr/share/consolefonts/c*
/usr/share/consolefonts/dr*
/usr/share/consolefonts/g*
/usr/share/consolefonts/i*
/usr/share/consolefonts/k*
/usr/share/consolefonts/l*
/usr/share/consolefonts/r*
/usr/share/consolefonts/s*
/usr/share/consolefonts/t*
/usr/share/consolefonts/v*
/usr/share/misc/*.oldwpa_supplicant
# vlock
|
[/quote] |
|
| Back to top |
|
|
drescherjm
Advocate
Joined: 05 Jun 2004
Posts: 2790
Location: Pittsburgh, PA, USA
|
| Posted: Mon Feb 20, 2006 12:40 am Post subject: |
|
|
| Quote: | I tryied to add a
GENTOO_MIRRORS="..."
in /var/tmp/catalyst/tmp/default/livecd-stage1-x86-2005.1/etc/make.conf without success.
Any help? |
Did you try using the mirrors from your main system's make.conf?
Also you can chroot to the /var/tmp/catalyst/tmp/default/livecd-stage1-x86-2005.1/ folder and make sure you have network connectivity. Possibly you do not have a valid resolv.conf.
[EDIT]Wait a minute you do have network connectivity. I believe the problem is you are tyring to use a file that is not in portage anymore, I'm going to check that after dinner (my girlfriend is calling me to eat)...
[/EDIT]
[EDIT]
I just looked and that file (ftp://cudlug.cudenver.edu/pub/mirrors/distributions/gentoo/distfiles/genpatches-2.6.12-10.extras.tar.bz2 ) is no longer in portage. There are several ways to solve this. One is to get a newer portage snapshot (possibly by making it yourself) a second is using a higher version of the kernel. This can be done by using /etc/portage/package.keywords and /etc/portage/package.mask
[/EDIT]
_________________
John
My gentoo overlay
Instructons for overlay |
|
| Back to top |
|
|
linoleum
Tux's lil' helper
Joined: 15 Nov 2004
Posts: 129
|
| Posted: Mon Feb 20, 2006 11:37 am Post subject: |
|
|
SOLVED
Thanks drescherjm, I'm trying.
Meanwhile on a different branch fo my attempt to create a livecd I get the following error on the stage1 (which appears in previous posts of this thread without solutions)
| Quote: |
duby2291 wrote:
I get this error after I almost get to the finish of the stage 1
Code:
Running command "/bin/bash /usr/lib/catalyst/targets/livecd-stage1/livecd-stage1.sh preclean"
umount: /var/tmp/catalyst/tmp/default/livecd-stage1-x86-20041023/dev: device is busy
umount: /var/tmp/catalyst/tmp/default/livecd-stage1-x86-20041023/dev: device is busy
!!! catalyst: Couldn't umount bind mount: /var/tmp/catalyst/tmp/default/livecd-stage1-x86-20041023/dev
!!! catalyst: Couldn't umount one or more bind-mounts; aborting for safety.
!!! catalyst: could not complete build
Any Idea's????
I'm having the exact same problem. I'm also using stage3-x86-2004.3.tar.bz2 as my seed for a livecd-stage1 target.
|
Any help on this?
Last edited by linoleum on Mon Feb 20, 2006 3:48 pm; edited 2 times in total |
|
| Back to top |
|
|
linoleum
Tux's lil' helper
Joined: 15 Nov 2004
Posts: 129
|
| Posted: Mon Feb 20, 2006 2:12 pm Post subject: |
|
|
I found a thread here with a solution for
| Quote: |
Code:
Running command "/bin/bash /usr/lib/catalyst/targets/livecd-stage1/livecd-stage1.sh preclean"
umount: /var/tmp/catalyst/tmp/default/livecd-stage1-x86-20041023/dev: device is busy
umount: /var/tmp/catalyst/tmp/default/livecd-stage1-x86-20041023/dev: device is busy
!!! catalyst: Couldn't umount bind mount: /var/tmp/catalyst/tmp/default/livecd-stage1-x86-20041023/dev
!!! catalyst: Couldn't umount one or more bind-mounts; aborting for safety.
!!! catalyst: could not complete build
|
For me it worked. |
|
| Back to top |
|
|
linoleum
Tux's lil' helper
Joined: 15 Nov 2004
Posts: 129
|
| Posted: Wed Feb 22, 2006 11:35 am Post subject: |
|
|
SOLVED: eliminating atlas from the USE variable so it does not pretend to emerge blas-atlas anymore
Still working on my livecd.
I need to include atlas-blas libraries. Generally I emerge this package with
| Quote: |
interactive=1 emerge blas-atlas
|
Is there a way to do that during the stage1? |
|
| Back to top |
|
|
linoleum
Tux's lil' helper
Joined: 15 Nov 2004
Posts: 129
|
| Posted: Mon Feb 27, 2006 11:59 am Post subject: |
|
|
New problem with my livecd. On a working stage1.spec file I added just "doc" in the use variable and I got this error while emerging pango
| Quote: |
i386-pc-linux-gnu-gcc -DPANGO_ENABLE_BACKEND -DPANGO_ENABLE_ENGINE -I.. -I/usr/include/cairo -pthread -I/usr/include/glib-2.0 -I/usr/lib/glib-2.0/include -I/usr/include/freetype2 -O2 -mcpu=i686 -Wall -c pango-scan.c -fPIC -DPIC -o .libs/pango-scan.o
In file included from ../pango/pangoxft.h:30,
from pango-scan.c:10:
../pango/pangoxft-render.h:31:22: X11/Xlib.h: No such file or directory
../pango/pangoxft-render.h:32:25: X11/Xft/Xft.h: No such file or directory
../pango/pangoxft-render.h:35:2: #error "must have Xft version 2 or newer"
In file included from ../pango/pangoxft.h:30,
from pango-scan.c:10:
../pango/pangoxft-render.h:64: error: syntax error before "Display"
../pango/pangoxft-render.h:64: warning: no semicolon at end of struct or union
../pango/pangoxft-render.h:66: error: syntax error before '*' token
../pango/pangoxft-render.h:66: warning: type defaults to `int' in declaration of `draw'
../pango/pangoxft-render.h:66: warning: data definition has no type or storage class
../pango/pangoxft-render.h:69: error: syntax error before '}' token
../pango/pangoxft-render.h:91: error: syntax error before "XTrapezoid"
../pango/pangoxft-render.h:94: error: syntax error before "XftFont"
../pango/pangoxft-render.h:101: error: syntax error before '*' token
../pango/pangoxft-render.h:104: error: syntax error before "XftDraw"
../pango/pangoxft-render.h:108: error: syntax error before '*' token
../pango/pangoxft-render.h:114: error: syntax error before '*' token
../pango/pangoxft-render.h:121: error: syntax error before '*' token
../pango/pangoxft-render.h:128: error: syntax error before '*' token
../pango/pangoxft-render.h:133: error: syntax error before '*' token
In file included from pango-scan.c:10:
../pango/pangoxft.h:51: error: syntax error before '*' token
../pango/pangoxft.h:53: error: syntax error before '*' token
../pango/pangoxft.h:55: error: syntax error before '*' token
../pango/pangoxft.h:58: error: syntax error before '*' token
../pango/pangoxft.h:63: error: syntax error before '*' token
../pango/pangoxft.h:78: error: syntax error before '*' token
../pango/pangoxft.h:78: warning: type defaults to `int' in declaration of `pango_xft_font_get_font'
../pango/pangoxft.h:78: warning: data definition has no type or storage class
../pango/pangoxft.h:79: error: syntax error before '*' token
../pango/pangoxft.h:79: warning: type defaults to `int' in declaration of `pango_xft_font_get_display'
../pango/pangoxft.h:79: warning: data definition has no type or storage class
Compilation of scanner failed
make[2]: *** [scan-build.stamp] Error 1
make[2]: Leaving directory `/var/tmp/portage/pango-1.10.2/work/pango-1.10.2/docs'
make[1]: *** [all-recursive] Error 1
make[1]: Leaving directory `/var/tmp/portage/pango-1.10.2/work/pango-1.10.2'
make: *** [all] Error 2
|
I do not understand why it complains about Xft since libXft should be included in xorg-x11 as far as I know.
Any suggestion?
Thanks |
|
| Back to top |
|
|
Bob P
Advocate
Joined: 20 Oct 2004
Posts: 3355
Location: Jackass! Development Labs
|
| Posted: Mon Feb 27, 2006 3:42 pm Post subject: |
|
|
Can anyone tell me -- Do Catalyst 1 and Catalyst 2 finally work? I've seen bugs that have gone unanswered for weeks on the catalyst mailing list, and both versions of Catalyst have either been unsupported or b0rken (and in either case totally unuseable) for a long time. 
Now that the new 2006.0 media is out, I was hoping to find that Catalyst had miraculously been fixed in time for the release.  But still no joy -- Catalyst engine and spec file updates aren't in the Portage tree.
When asking this question during the Gentoo 2006.0 development phase I was told that Gentoo won't release any of the components used to build Gentoo until the final release is finished. Ok. Now we have the 2006.0 Gentoo release, so I'm hoping that its safe to stick my toe in the water...
If Catalyst is now working and it was used to produce the 2006.0 media, will we be seeing a release of the working version of Catalyst 2 and the 2006.0 media file specs?
Or is it safe to assume that Catalyst 2 is still so horribly broken that RelEng had to rely on some other software to build the 2006.0 media?
So far I've been rolling my own scripts for everything that I do, partly because Catalyst just hasn't been functional. I'm hoping that Catalyst actually works well enough that I can take it for a test drive. Any insights would be appreciated.
_________________
.
Stage 1/3 | Jackass! | Rockhopper! | Thanks | Google Sucks |
|
| Back to top |
|
|
96140
Retired Dev
Joined: 23 Jan 2005
Posts: 1324
|
| Posted: Mon Feb 27, 2006 6:42 pm Post subject: |
|
|
Well, releng has been using catalyst2 for the last few releases (at least!), including 2006.0. Pretty generic specs (take a look at the 2005.1 files for examples), though the installer CD was less generic. From what one of the devs was saying, everything is in the snapshot. Check here at sources.gentoo.org for the installer code. If you want Catalyst specfiles or similar, mail wolf31o2@gentoo.org , though he won't be back until mid-March.
In the mean time, the source code for catalyst2 is available.
I looked through quite a bit of code for that installer while I was working on the handbooks. It seems to be mostly (entirely?) agaffney and codeman's hand-written Python code.
HTH |
|
| Back to top |
|
|
todw1fd
Tux's lil' helper
Joined: 21 Mar 2003
Posts: 93
|
| Posted: Tue Feb 28, 2006 5:17 pm Post subject: |
|
|
| nightmorph wrote: | Well, releng has been using catalyst2 for the last few releases (at least!), including 2006.0. Pretty generic specs (take a look at the 2005.1 files for examples), though the installer CD was less generic. From what one of the devs was saying, everything is in the snapshot. Check here at sources.gentoo.org for the installer code. If you want Catalyst specfiles or similar, mail wolf31o2@gentoo.org , though he won't be back until mid-March.
In the mean time, the source code for catalyst2 is available.
I looked through quite a bit of code for that installer while I was working on the handbooks. It seems to be mostly (entirely?) agaffney and codeman's hand-written Python code.
HTH |
Have you used catalyst recently? Just for fun, emerge it and try and build something with those 2005.1 spec files. They're surely a nice starting point to get some ideas. But that's about it. Plenty of not quite documented changes in play now. If you happen to be stuck doing documentation for this package, i'd suggest sitting down with one of the devs and letting them correct you as you try to follow the current examples and docs to build something. You'll be surprised.
I sure wish the error I started getting with the latest incarnation, rc39, would turn out to be my imagination. Probably too the others that got the same error. I haven't been able to generate a portage snapshot to use with the other stages since the change. Here's the error:
| Code: | is-1 catalyst2 # catalyst2 -s 2006.latest
Gentoo Catalyst, version 2.0_rc39
Copyright 2003-2005 The Gentoo Foundation
Distributed under the GNU General Public License version 2
Using default Catalyst configuration file, /etc/catalyst2/catalyst2.conf
Setting sharedir to config file value "/usr/lib/catalyst2"
Setting snapshot_cache to default value "/var/tmp/catalyst2/snapshot_cache"
Setting hash_function to config file value "md5"
Setting storedir to config file value "/var/tmp/catalyst2"
Setting portdir to default value "/usr/portage"
Setting distdir to config file value "/usr/portage/distfiles"
Setting options to config file value "autoresume ccache kerncache pkgcache seedcache snapcache"
Compiler cache support enabled.
Package cache support enabled.
Snapshot cache support enabled.
Seed cache support enabled.
Kernel cache support enabled.
Autoresuming support enabled.
Envscript support enabled.
Creating Portage tree snapshot 2006.latest from /usr/portage...
Compressing Portage snapshot tarball...
Catalyst aborting....
Traceback (most recent call last):
File "/usr/bin/catalyst2", line 380, in ?
build_target(addlargs, targetmap)
File "/usr/bin/catalyst2", line 198, in build_target
mytarget.run()
File "/usr/lib/catalyst2/modules/snapshot_target.py", line 63, in run
self.gen_digest_file(self.settings["snapshot_path"])
AttributeError: snapshot_target instance has no attribute 'gen_digest_file'
|
Same on all 4 of my boxen. rc38 didn't seem to have the same problem. Wonder what the releng team used to do their work for the release. Haven't found any fixes on the forums or bugs site. And yet somehow the release team got their product out the door. Even if I had the spec files from the latest release, I still think I'd have the same error when trying to create the portage snapshot.
In any case, i'm sure the fix is something simple. And I certainly look forward to fresh documentation when a final working build is released. |
|
| Back to top |
|
|
Bad Penguin
Guru
Joined: 18 Aug 2004
Posts: 507
|
| Posted: Wed Mar 01, 2006 2:47 am Post subject: |
|
|
| Bob P wrote: | | If Catalyst is now working and it was used to produce the 2006.0 media, will we be seeing a release of the working version of Catalyst 2 and the 2006.0 media file specs? |
Bob, I look here. They don't appear to be there yet. I wonder why the big mystery about release specs.
| Bob P wrote: | | So far I've been rolling my own scripts for everything that I do, partly because Catalyst just hasn't been functional. I'm hoping that Catalyst actually works well enough that I can take it for a test drive. Any insights would be appreciated. |
I tried to use it a couple of months ago. The only way to customize the stages it generates is to alter profiles, which defeats the purpose of using it as far as I am concerned. |
|
| Back to top |
|
|
Bob P
Advocate
Joined: 20 Oct 2004
Posts: 3355
Location: Jackass! Development Labs
|
| Posted: Wed Mar 01, 2006 5:53 pm Post subject: |
|
|
| Bad Penguin wrote: | | Bob P wrote: | | If Catalyst is now working and it was used to produce the 2006.0 media, will we be seeing a release of the working version of Catalyst 2 and the 2006.0 media file specs? |
Bob, I look here. They don't appear to be there yet. I wonder why the big mystery about release specs. |
well, i'm also at a loss to explain why the ebuild has not been updated and why the spec files have not been published ... so i thought i'd ask.
everyone i've spoken to has told me that attempting to use Catalyst2 has been very frustrating of late. for a while catalyst was being aggressively worked on -- there were 2 or 3 updates to catalyst coming out every day. the penultimate version of catalyst (rc38) worked fine to build a live CD, though there were some problems with udevstart. the ultimate version of catalyst (rc39) came out -- which supposedly fixed those problems -- but as more than one catalyst user has already mentioned on the mailing list, the latest version of catalyst has targeting problems and won't even build a portage snapshot.
that's pretty sad. catalyst, the primary tool for building gentoo, won't even build a portage snapshot. the default catalyst configuration is committed to building a portage snapshot. out of the box, the tool won't perform its primary function.
interestingly, gentoo has managed to release 2006.0 -- in spite of the fact that Catalyst has been broken, has not been updated, and there have been several unaddressed problems on the mailing list.
a couple of weeks ago it looked like Catalyst was pretty close to being finished. work on catalyst was aggressive, and it was being updated 2 to 3 times daily. then, when catalyst was working well enough to build an ISO image and appeared to be almost ready for prime time, the next release broke it, and all catalyst development appears to have abruptly stopped. two weeks later we have a gentoo release, even though the tool used to build gentoo remains broken.
just so that people don't get the impression that i'm some sort of whacked-out conspiracy theorist -- i'm not suggesting that the working version of catalyst and the x86 spec files have been intentionally withheld. i'm just asking for some clarification. is the latest release of catalyst the working release, which would suggest that everyone who is having problems on the mailing list is just too stupid to use the tool and that it works fine in the hands of competent people? or is catalyst really just broken, making it a waste of everyone's time to attempt to use this tool? if the later is the case, then we would all be better served by looking somewhere else.
in my overly simplistic view of the situation, one of the following has to be true:
1. catalyst works
2. catalyst is broken
if catalyst works, then there are an awful lot of users who are reporting problems on the mailing list are just too too stupid to use the tool to get the job done. they need support, and its not forthcoming, either in the form of responses on the mailing list or in the form of releases of the software and the spec files.
if catalyst doesn't work, it would be nice to know what tools gentoo used for the 2006.0 release.
personally, i haven't been using catalyst though i have been watching the mailing list. before i spend countless hours working with catalyst, i'd like to hear that its working. i'd like to know whether the current release in portage is the one that was used to build 2006.0. and i'd like to know if the x86 spec files are going to be made availble.
its interesting to note that catalyst 1 and the 2005.x spec files are no longer supported because the portage tree that was in existence at that time no longer exists. in that case, obsolescence of the portage tree has rendered the older spec files unsupportable. it would be nice if catalyst and the x86 spec files used to build 2006.0 were to be released before the portage tree changed sufficiently to render the 2006.0 spec files equally worthless. 
_________________
.
Stage 1/3 | Jackass! | Rockhopper! | Thanks | Google Sucks |
|
| Back to top |
|
|
Bad Penguin
Guru
Joined: 18 Aug 2004
Posts: 507
|
| Posted: Wed Mar 01, 2006 6:10 pm Post subject: |
|
|
| Bob P wrote: | | Bad Penguin wrote: | | Bob P wrote: | | If Catalyst is now working and it was used to produce the 2006.0 media, will we be seeing a release of the working version of Catalyst 2 and the 2006.0 media file specs? |
Bob, I look here. They don't appear to be there yet. I wonder why the big mystery about release specs. |
everyone i've spoken to has told me that attempting to use Catalyst2 has been very frustrating of late. for a while catalyst was being aggressively worked on -- there were 2 or 3 updates to catalyst coming out every day. the penultimate version of catalyst (rc38) worked fine to build a live CD, though there were some problems with udevstart. the ultimate version of catalyst (rc39) came out -- which supposedly fixed those problems -- but as more than one catalyst user has already mentioned on the mailing list, the latest version of catalyst has targeting problems and won't even build a portage snapshot. |
In all fairness, catalyst2 is unstable and still being developed. Whether or not "official" gentoo releases and livecds can be created with catalyst at all is an entirely different question  It appears that some of the teams (ppc) did release 2006.0 specs, perhaps the those spec files might indicate which version of catalyst might have been used.
Like you I have had to resort to writing my own build scripts (gentoo-buildhoster). I hope some day to be able to use catalyst instead of my own hacky stuff. I have noticed recently that my own stuff is having problems unmounting bind mounted stuff because of udev, for some reason I have to umount -l to get it to let go (umount -f fails also). I don't know if the problem is udev or the latest kernel versions, but I am almost certain catalyst would run into the same problem...
Anyone know if pentium3/pentium4 stage3s are a thing of the past? |
|
| Back to top |
|
|
Bob P
Advocate
Joined: 20 Oct 2004
Posts: 3355
Location: Jackass! Development Labs
|
| Posted: Wed Mar 01, 2006 7:02 pm Post subject: |
|
|
| Bad Penguin wrote: | | In all fairness, catalyst2 is unstable and still being developed. |
well, this is an interesting situation. Catalyst 1 is the official release, and it is broken and unsupported. Catalyst 2 is in development and it is broken and unsupported. So I guess that means that no working tool is available. the question remains whether rc39 is the final build that was used to build the release media for 2006.0. inquiring minds want to know.
| Quote: | | Anyone know if pentium3/pentium4 stage3s are a thing of the past? |
i don't have the answer, but here's what little i have to offer on the idea. with Jackass! and Rockhopper! i decided not to build every tarball for every arch. as more people came to understand the logic behind the Stage 1/3 install and its conceptual approach to rebuilding the toolkit gained critical mass, rebuilding the toolkit in the Stage 1/3 fasion came to become established as a standard building method. now that everyone understands how to do it, its simpler from a support perspective to just build for CHOST and to let the user do the final architecture-specific tweaking.
Gentoo appears to have adopted this concept and the 2006.0 x86 media appears to include CHOST specific builds instead of arch-specific builds. this is a significant paradigm shift, even though its been implemented very quietly and without fanfare. all that's required is to select the correct CHOST, change your march setting, and perform two emerge -e systems. for anyone with a fast processor like a P3, P4 or XP the processors are fast enough that this should not be a concern.
it would appear that after i've spent 16 months of pounding on the table, my favorite method of rebuilding every architecture via a Stage 3 tarball been assimilated into the standard Gentoo idealogy. 
_________________
.
Stage 1/3 | Jackass! | Rockhopper! | Thanks | Google Sucks |
|
| Back to top |
|
|
todw1fd
Tux's lil' helper
Joined: 21 Mar 2003
Posts: 93
|
| Posted: Thu Mar 02, 2006 4:09 pm Post subject: |
|
|
Me? Too Stupid to use the tool? Hey, I resemble that remark! Found out last night that catalyst 1 still is able to build a portage snapshot successfully. Can't seem to understand why version 2 should be having any problem in that dept. while ver1 still works. Anyway, looks like i'll be using 1 to build my snapshot, copying it over to the relevant catalyst2 directory and see what happens from there. Not sure what else may not be working since gave up after snapshots failed. Yes, sure, could have downloaded a snapshot from the mirrors. Should have thought of that earlier.
Was surprised to see the Gentoo 2006.0 release out without a bump to a final Catalyst2. Mailing list awful quiet since I posted the problem with the snapshot creation. Figured there'd be a trivial fix for that one...not so trivial I could find it and fix it though. It would be great is this tool would be finalized ( if that is possible with a ever changing portage tree ). Spec files for the new release would have been nice to look at. Curious why always getting an Alsa-Drivers error on the new livecd. It looks like the alsa-driver package wasn't installed...yet the kernel doesn't have the sound modules built there either. Found a /lib/modules/ directory for the 2.6.15-r4 kernel on the livecd too, but looks like 2.6.15-r5 was used for the build. Not much wasted space, but hey 14 megs is 14 megs times lots of downloads. |
|
| Back to top |
|
|
Bob P
Advocate
Joined: 20 Oct 2004
Posts: 3355
Location: Jackass! Development Labs
|
| Posted: Thu Mar 02, 2006 8:04 pm Post subject: |
|
|
yes, you are right! there's definitely 14 MB of garbage on the Live Cd, presumably because a kernel upgrade was made and somebody just forgot to purge the drivers contained in /lib/modules from the older kernel build. if there's ever a need to make an -r1 version of the CD, it would be worth adding a cleanup of the modules directory to the list of thigns to do. after all 14 MB is 14 MB, and there's no sense in having everyone download 14 MB of junk as it just wastes bandwidth for everyone involved.
as you mentioned, the other funny thing about the live CD is that alsa isn't even installed on this one. it looks like the udev device starting problems never got solved in time for the release, and that the fix to make the alsa loading errors go away was to not install alsa on the system.
correct me if i'm wrong, but when you were using the most recent version of catalyst 2, didn't it work this way: the init script would test for alsa, find it installed, attempt to start it, fail, and issue an error message. on the 2006.0 gentoo CD it looks like the init script tests for alsa installation, doesn't find it, doesn't attempt to start it, and doesn't issue the error becasue alsa isn't even installed. so it looks like the solution that gentoo used to avoid the catalyst2 problem of alsa/udev failing to initialize the device was to simply remove all alsa support from the build, taking sound support out of the CD. that's too bad, as i could see reasons to have sound support enabled if you're going to all of the trouble to have a GUI installer.
_________________
.
Stage 1/3 | Jackass! | Rockhopper! | Thanks | Google Sucks |
|
| Back to top |
|
|
todw1fd
Tux's lil' helper
Joined: 21 Mar 2003
Posts: 93
|
| Posted: Sun Mar 05, 2006 3:20 pm Post subject: |
|
|
Finally managed to get a snapshot out of catalyst...and then eventually a livecd. To get the portage snapshot to complete I commented out the line "self.gen_digest_file(self.settings["snapshot_path"])" from /usr/lib/catalyst2/modules/snapshot_target.py. | Code: | print "Compressing Portage snapshot tarball..."
cmd("tar cjf "+self.settings["snapshot_path"]+" -C "+mytmp+" portage",\
"Snapshot creation failure",env=self.env)
#self.gen_digest_file(self.settings["snapshot_path"])
self.cleanup()
print "snapshot: complete!" |
Snapshot completed, but without the DIGEST file.
Another fix from bugs site: Change generic_target to generic_stage_target from the same snapshot_target.py file : | Code: | class snapshot_target(generic_stage_target):
def __init__(self,myspec,addlargs): |
You get a DIGEST file with this fix.
Big problem appears to be with trying to use seed file that's using testing branch portage 2.1. pre whatever. Dropping back to a seed with stable 2.0.54 portage seemed to make catalyst2 happy again...although still couldn't get a portage snapshot to build without one of the fixes above. |
|
| Back to top |
|
|
Bob P
Advocate
Joined: 20 Oct 2004
Posts: 3355
Location: Jackass! Development Labs
|
| Posted: Sun Mar 12, 2006 9:17 pm Post subject: |
|
|
just in case anyone is still having problems with the catalyst2 snapshot generation problem, if you replace your /usr/lib/catalyst2/modules/snapshot_target.py with the following code, the problem with the failed snapshots and broken digests goes away. 
| Code: |
"""
Builder class for snapshots.
"""
import os
from catalyst_support import *
from generic_stage_target import *
class snapshot_target(generic_stage_target):
def __init__(self,myspec,addlargs):
self.required_values=["version_stamp","target"]
self.valid_values=["version_stamp","target","portdir_overlay"]
generic_target.__init__(self,myspec,addlargs)
self.settings=myspec
self.settings["target_subpath"]="portage"
st=self.settings["storedir"]
self.settings["snapshot_path"]=normpath(st+"/snapshots/portage-"+self.settings["version_stamp"]\
+".tar.bz2")
self.settings["tmp_path"]=normpath(st+"/tmp/"+self.settings["target_subpath"])
if self.settings.has_key("portdir_overlay"):
print "\nWarning!!!!"
print "\tThis feature is deprecated no overlay will be added to the snapshot."
print "\tIf you need an overlay feature please use portage_overlay in the other spec files"
print "\tOtherwise add your files to the normal snapshot and redigest .. you should know what"
print "\t\tto do."
print "\tThis was removed due to digesting issues that are incompatible with portage."
def setup(self):
x=normpath(self.settings["storedir"]+"/snapshots")
if not os.path.exists(x):
os.makedirs(x)
def mount_safety_check(self):
pass
def run(self):
self.setup()
print "Creating Portage tree snapshot "+self.settings["version_stamp"]+\
" from "+self.settings["portdir"]+"..."
mytmp=self.settings["tmp_path"]
if not os.path.exists(mytmp):
os.makedirs(mytmp)
cmd("rsync -a --delete --exclude /packages/ --exclude /distfiles/ --exclude /local/ --exclude CVS/ "+\
self.settings["portdir"]+"/ "+mytmp+"/portage/","Snapshot failure",env=self.env)
#if self.settings.has_key("portdir_overlay"):
#print "Adding Portage overlay to the snapshot..."
#cmd("rsync -a --exclude /packages/ --exclude /distfiles/ --exclude /local/ --exclude CVS/ "+\
# self.settings["portdir_overlay"]+"/ "+mytmp+"/portage/","Snapshot/ overlay addition failure",\
# env=self.env)
print "Compressing Portage snapshot tarball..."
cmd("tar cjf "+self.settings["snapshot_path"]+" -C "+mytmp+" portage",\
"Snapshot creation failure",env=self.env)
# commenting out the following line aborts the digest build and an error
self.gen_digest_file(self.settings["snapshot_path"])
self.cleanup()
print "snapshot: complete!"
def kill_chroot_pids(self):
pass
def cleanup(self):
print "Cleaning up..."
def register(foo):
foo.update({"snapshot":snapshot_target})
return foo |
this code makes both a snapshot AND a digest. you should also be able to use testing branch portage to create snapshots/digests without any problems.
_________________
.
Stage 1/3 | Jackass! | Rockhopper! | Thanks | Google Sucks |
|
| Back to top |
|
|
slycordinator
Advocate
Joined: 31 Jan 2004
Posts: 3065
Location: Korea
|
| Posted: Sat Mar 18, 2006 6:22 am Post subject: |
|
|
Everytime I try to use the stage1 spec file it attempts to install pam-login which is blocked by shadow.
Here's the spec file I used:
| Code: | ## livecd-stage1 template for Gentoo Linux release bootable LiveCDs
## John Davis
## Modified by: Stephan Hermann <sh@sourcecode.de>
## Modified by: Zac Slade <krakrjak@gmail.com>
## Modified to make compatible with newer portage:
## Now using 2004.3 profile
## Added -opengl and changed gtk to -gtk
## Removed xorg firefox and xfce4 to prevent X from building
## Barrier for entry should now be lower to create a CD that works
subarch: x86
version_stamp: 2004.3
target: livecd-stage1
rel_type: default
profile: default-linux/x86/2006.0
snapshot: 2006.0
source_subpath: stage3-x86-2006.0
livecd/use:
-* ipv6 socks5 livecd fbcon ncurses readline ssl
livecd/packages:
livecd-tools dhcpcd acpid apmd gentoo-sources coldplug fxload irssi gpm syslog-ng parted links raidtools dosfstools nfs-utils jfsutils xfsprogs e2fsprogs reiserfsprogs ntfsprogs pwgen rp-pppoe screen mirrorselect penggy iputils hwdata-gentoo hwsetup lvm2 evms vim pptpclient mdadm ethtool wireless-tools prism54-firmware wpa_supplicant ndiswrapper |
named it livecd_gentoo_wiki_howto-stage1.spec
Then I ran "catalyst2 -s 2006.0"
Then I downloaded the stage3-x86-2006.0.tar.bz2 and put it in /var/tmp/catalyst2/builds
Then did "catalyst2 -f livecd_gentoo_wiki_howto-stage1.spec"
And it eventually crapped out because it tried installing pam-login (and it's installation is blocked by shadow).
If this looks familiar, I'm using http://gentoo-wiki.com/HOWTO_build_a_LiveCD as my guide. And for all I can tell this would work if catalyst weren't attempting to build something blocked.
| Code: | | dev-util/catalyst-2.0_rc40 USE="ccache cdr" |
|
|
| Back to top |
|
|
drescherjm
Advocate
Joined: 05 Jun 2004
Posts: 2790
Location: Pittsburgh, PA, USA
|
| Posted: Sat Mar 18, 2006 9:28 pm Post subject: |
|
|
I had the same problem, with catalyst and also with a gentoo install. With catalyst you have to chroot to the livecd and see exactly what versions of pam-login and shadow that are trying to be installed. pam-login requires shadow but only certain versions of pam-login and shadow are compatible. I believe I had to mask a version of shadow to get my catalyst to build for me. You can examine the ebuilds at /usr/portage/sys-apps/pam-login to see what versions of shadow are needed for pam-login.
_________________
John
My gentoo overlay
Instructons for overlay |
|
| Back to top |
|
|
slycordinator
Advocate
Joined: 31 Jan 2004
Posts: 3065
Location: Korea
|
| Posted: Sat Mar 18, 2006 9:39 pm Post subject: |
|
|
| drescherjm wrote: | | I had the same problem, with catalyst and also with a gentoo install. With catalyst you have to chroot to the livecd and see exactly what versions of pam-login and shadow that are trying to be installed. pam-login requires shadow but only certain versions of pam-login and shadow are compatible. I believe I had to mask a version of shadow to get my catalyst to build for me. You can examine the ebuilds at /usr/portage/sys-apps/pam-login to see what versions of shadow are needed for pam-login. |
I see what was wrong. The stage tarball had pam-login already installed and it tried to update it too.
edit:
So what seems to be the "fix" for the problem is just unpacking the stage tarball anywhere, bind-mounting portage into that unpacked area, chroot in, remove pam-login, and then unmount stuff and create a new tarball.
This way I can automate it in the future. Though I'll have to see if this works on another day (as I work soon).
But it no longer gave the pam-login message so that's good. |
|
| Back to top |
|
|
|
|
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
|
|
Gentoo Chat
