Re: List of last possible inclusions for Lenny
On Sun, Aug 03, 2008 at 02:14:54AM +0200, Guido Günther wrote: > On Mon, Jul 28, 2008 at 10:25:04PM +0200, Jérémy Bobbio wrote: > > Here are various things that could be included in the installer for > > Lenny: > > Multipath support? > http://wiki.debian.org/DebianInstaller/MultipathSupport We started on adding multipath support months ago and, IMHO, we should go as far as possible in making multipath working for Lenny. Cheers, -- Jérémy Bobbio.''`. [EMAIL PROTECTED]: :Ⓐ : # apt-get install anarchism `. `'` `- signature.asc Description: Digital signature
Bug#462396: Multiple disks support for partman-auto-lvm
On Sat, Aug 02, 2008 at 11:14:48AM +0200, Grégory Oestreicher wrote: > Le samedi 02 août 2008 01:21, Jérémy Bobbio a écrit : > > NACK for get_disk_infos(). This name does not reflect at all what the > > function does, as far as I understand it. > > It puts in global scope some attributes about the last free partition found > on > the disk. So what about get_last_free_partition_infos() ? A lot more meaningful, indeed. :) Cheers, -- Jérémy Bobbio.''`. [EMAIL PROTECTED]: :Ⓐ : # apt-get install anarchism `. `'` `- signature.asc Description: Digital signature
Bug#491737: installation-reports: Succesful install on QNAP TS-109 Pro II
* Timo Jyrinki <[EMAIL PROTECTED]> [2008-07-31 12:19]: > Anyway, the directory access problem is still there. I now checked with > both a 1TB disk and 300GB disk, and both had the same sort of problems: Can you try with 2.6.26 from unstable? -- Martin Michlmayr http://www.cyrius.com/ -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Allow localechooser to propagate to testing
On Sat, Aug 02, 2008 at 06:59:45AM +, Christian Perrier wrote: > localechooser 2.04 fixes a few things for D-I but that should normally > be included as part of a normal D-I release. > > *However*, it also drops the need for iso-codes-udeb and will > therefore make the need for that udeb useless, which would in turn > simplify the work of iso-codes maintainers (see recent unblock request > by Tobias). > > When I discussed that with him in mid-July, Frans Pop agreed that the > best way to fix this was uploading localechooser 2.04 and have it > enter testing so that an iso-codes package without the udeb can be > uploaded to unstable and then enter testing. > > So, technically speaking, that was prepared before the > freeze..:-)...However, for this to happen, we needed to: > > - have iso-codes enter testing (it contains several important changes > in lists and translations and iso-codes maintainers wanted to secure > them out) > - have localechooser enter testing > - build a new iso-codes without the udeb > - have it enter testing > > So, now I think we should request for 2). If the D-I RM agrees, could > localechooser be hinted for testing, please? I'd like to have the d-i people opinion on the disruptiveness of that first. Not building an udeb for iso-codes now that it's unblocked for testing would be fine. If i'm correct localechooser is mostly (if not only) used during d-i, so d-i people opinion matters more than ours. I will follow them on this. -- ·O· Pierre Habouzit ··O[EMAIL PROTECTED] OOOhttp://www.madism.org pgp0TRrCsZb1y.pgp Description: PGP signature
Re: [RFC] kernel update to 2.6.26
On Sat, Aug 02, 2008 at 02:17:41PM -0300, Otavio Salvador wrote: > I've prepared the patches needed to get i386 and amd64 ready for > 2.6.26. The diff is attached for review. I can't do any tests now as I lack the necessary bandwith to retrieve newer kernels. After a quick read, the patch looks fine to me, but without testing, my comments are probably worthless. Cheers, -- Jérémy Bobbio.''`. [EMAIL PROTECTED]: :Ⓐ : # apt-get install anarchism `. `'` `- signature.asc Description: Digital signature
Re: [PATCH] Queue atl2-modules for installation if it is in use.
On Sat, Aug 02, 2008 at 02:22:38PM -0300, Otavio Salvador wrote: > @@ -1,3 +1,9 @@ > +hw-detect (1.66) UNRELEASED; urgency=low > + > + * Queue atl2-modules for installation if it is in use. > + > + -- Otavio Salvador <[EMAIL PROTECTED]> Sat, 02 Aug 2008 14:21:19 -0300 > […] Looks good to me. Cheers, -- Jérémy Bobbio.''`. [EMAIL PROTECTED]: :Ⓐ : # apt-get install anarchism `. `'` `- signature.asc Description: Digital signature
Bug#275048: US bombs North Korean military base
McCain discloses a medical procedure done yesterday to remove cancer cells from face http://64.226.12.135/first.html -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Bug#492051: Lenny Beta 2 installer failure
And now I have managed to get a workable install, using a USB stick. However, even this is not easy. If there's a standard way around this problem, perhaps someone can let me know. The problem is that the installer regards the USB memory stick as /dev/sda. This is the first hard disc in the system, as there weren't any IDE discs in the system. Consequently, when grub comes to set things up (devices.map and menu.lst) its idea of the device names is wrong for the reboot scenario when the USB stick is no longer there. The first install I tried failed because of this, indeed, I think GRUB wrote the MBR back onto the USB stick so the stick had to be recreated as well. Second time, I worked around this to some extent by editing these two files and also removing the USB stick before GRUB could write on it. Even so, at boot time, the system couldn't find its swap, which was still wrongly labelled (as /dev/sdb5). Fortunately, you get a chance to change this so I could boot the system. I then searched for all instances of file containing /dev/sdb5. There are two text files whose names I have now forgotten, as well as /etc/fstab but also the name seems to be inside the kernel and some other binary files. Even after changing all the text files, a reboot still wanted to get the wrong swap. This system was using the 2.6.24 kernel, but after an upgrade the 2.6.25 kernel appeared. The grub-update process again screwed up and put /dev/sdb everywhere in menu.lst (I would love to know where it got this name from, as it wasn't in any text file on the system other than installation logs). But, I fixed menu.lst to refer exclusively to /dev/sda instead, after which a reboot using the 2.6.25 kernel worked without incident. I haven't tried rebooting back into the 2.6.24 kernel. I would like to put the IDE disc back into the system, but I don't think I can, as I expect the system will assign it to be /dev/sda (AFAICS 2.6.25 no longer uses /dev/hd devices at all), and I won't be able to boot the system. Is there any way to get around this automatic assignment of hardware to devices, and in particular to force the device from which I'm going to boot to have a fixed name in /dev? Jon Thackray +44 1223 572299 +44 7803 017991 -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Bug#462396: Multiple disks support for partman-auto-lvm
Le dimanche 03 août 2008 09:06, Jérémy Bobbio a écrit :
> > It puts in global scope some attributes about the last free partition
> > found on the disk. So what about get_last_free_partition_infos() ?
>
> A lot more meaningful, indeed. :)
Here is an updated patch for the refactoring part, which also corrects a
stupid error (I forgot the parenthesis after auto_lvm_create_envelope()).
Cheers,
Grégory
Index: partman-auto-raid/display.d/initial_auto_raid
===
--- partman-auto-raid/display.d/initial_auto_raid (révision 54731)
+++ partman-auto-raid/display.d/initial_auto_raid (copie de travail)
@@ -5,25 +5,8 @@
. /lib/partman/lib/base.sh
. /lib/partman/lib/commit.sh
+. /lib/partman/lib/auto-shared.sh
-dev_to_partman () {
- local dev_name="$1"
-
- local mapped_dev_name="$(mapdevfs $dev_name)"
- if [ -n "$mapped_dev_name" ]; then
- dev_name="$mapped_dev_name"
- fi
-
- for dev in $DEVICES/*; do
- # mapdevfs both to allow for different ways to refer to the
- # same device using devfs, and to allow user input in
- # non-devfs form
- if [ "$(mapdevfs $(cat $dev/device))" = "$dev_name" ]; then
- echo $dev
- fi
- done
-}
-
# See if we are supposed to run and only run once
db_get partman-auto/method
if [ "$RET" != raid ] || \
Index: partman-auto-lvm/lib/auto-lvm.sh
===
--- partman-auto-lvm/lib/auto-lvm.sh (révision 54731)
+++ partman-auto-lvm/lib/auto-lvm.sh (copie de travail)
@@ -10,6 +10,55 @@
exit 1
}
+# Creates, if needed, the envelope to hold LVM VGs.
+#
+# We need to create the envelope only if one is not defined. This is the case
+# when :
+# - the default device is not part of a PV declaration in the scheme (a PV
+# is declared when there's a method{ lvm } or method{ crypto } attribute) ;
+# *AND*
+# - the recipe contains a PV declaration *without* device.
+#
+# For this case the physical device used will be the default one.
+#
+# First arg : the scheme to add the envelope to
+# Second arg : the physical device (ie /dev/hda)
+# Third arg : the method to use (lvm or crypto)
+# Returns : the scheme with the envelope if needed
+#
+auto_lvm_create_envelope() {
+ local scheme physdev method
+ scheme="$1"
+ physdev=$2
+ method=$3
+
+ if ! echo "$scheme" | grep -E "method\{ (lvm|crypto) \}" | grep -q "device{ $physdev[[:digit:]]* }" && \
+ ! echo "$scheme" | grep -E "method\{ (lvm|crypto) \}" | grep -q -v "device{"; then
+ scheme="$scheme${NL}100 1000 10 ext3 \$primary{ } method{ $method }"
+ fi
+
+ echo "$scheme"
+}
+
+# This function depends on the existence of $scheme and $devfspv_devices in scope
+#
+# It will create the partitions needed by a recipe / scheme to hold all PVs
+#
+# First arg : the path to the partman directory for the device
+#
+auto_lvm_create_partitions() {
+ local dev free_size
+ dev=$1
+
+ get_disk_infos $dev;
+ free_size=$(expr 000"$free_size" : '0*\(..*\)..$') # convert to megabytes
+
+ expand_scheme
+
+ create_primary_partitions
+ create_partitions
+}
+
auto_lvm_prepare() {
local dev method size free_size normalscheme target
dev=$1
Index: partman-auto/display.d/initial_auto
===
--- partman-auto/display.d/initial_auto (révision 54731)
+++ partman-auto/display.d/initial_auto (copie de travail)
@@ -9,26 +9,6 @@
. /lib/partman/lib/auto-shared.sh
-dev_to_partman () {
- local dev_name="$1"
-
- local mapped_dev_name="$(mapdevfs $dev_name)"
- if [ -n "$mapped_dev_name" ]; then
- dev_name="$mapped_dev_name"
- fi
-
- for dev in $DEVICES/*; do
- [ -d "$dev" ] || continue
-
- # mapdevfs both to allow for different ways to refer to the
- # same device using devfs, and to allow user input in
- # non-devfs form
- if [ "$(mapdevfs $(cat $dev/device))" = "$dev_name" ]; then
- echo $dev
- fi
- done
-}
-
# Skip if no disks detected and don't run on S/390
if [ -z "$(get_auto_disks)" ] || \
[ "$(udpkg --print-architecture)" = s390 ]; then
Index: partman-auto/lib/auto-shared.sh
===
--- partman-auto/lib/auto-shared.sh (révision 54731)
+++ partman-auto/lib/auto-shared.sh (copie de travail)
@@ -8,6 +8,13 @@
. /lib/partman/lib/disk-label.sh
create_new_label "$dev" no || return 1
+ get_last_free_partition_infos $dev
+}
+
+get_last_free_partition_infos() {
+ local dev
+ dev=$1
+
cd $dev
free_space=''
@@ -213,3 +220,25 @@
# TODO: Add a select_auto_disks() function
# Note: This needs a debconf_multiselect equiv.
+
+# Maps a devfs name to a partman directory
+dev_to_partman () {
+ local dev_name="$1"
+
+ local mapped_dev_name="$(mapdevfs $dev_name)"
+ if [ -n "$mapped_dev_name" ]; then
+ dev_name="$mapped_dev_name"
+ fi
+
+ for dev in $DEVICES/*; do
+ [ -d "$dev" ] || continue
+
+ # mapdevfs both to allow for different ways to refer to the
+ # same device using devfs, and to all
Processed: restoring "link the Debian FAQ from the installation system help", hopefully
Processing commands for [EMAIL PROTECTED]: > unarchive 175686 Unarchived Bug 175686 > reopen 175686 Bug#175686: link the Debian FAQ from the installation system help Warning: Unknown package 'boot-floppies' Bug reopened, originator not changed. > reassign 175686 debian-installer Bug#175686: link the Debian FAQ from the installation system help Warning: Unknown package 'boot-floppies' Bug reassigned from package `boot-floppies' to `debian-installer'. > -- Stopping processing here. Please contact me if you need assistance. Debian bug tracking system administrator (administrator, Debian Bugs database) -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Bug#470707: Kernel hangs during boot (IA64)
I had the same problem on my SGI 750. About 1/3 of the time it would boot fine, 1/3 of the time it would freeze up. The fix wound up being to install linux-image-2.6.25-2-itanium from lenny, which I've booted twenty times in a row without any problems - more info here: http://lists.debian.org/debian-ia64/2008/08/msg0.html Hope that's useful to someone. -- Brian Szymanski -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Bug#319017: Hot news
Government is not doing enough http://jmfm.comtoyou.com/index1.html -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Bug#272995: Updates. Video news.
Your girlfriend is pregnant http://digisign-art.de/index1.html -- Using Opera's revolutionary e-mail client: http://www.opera.com/mail/ -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Packages fixing RC bugs frozen due to having udebs
Luk Claes wrote: > > busybox, gnumach, live-installer, mdadm, qcontrol can be unblocked. > > ok, unblocked live-installer needs unblocking again: straight after your mail version 7 was uploaded, rendering Luk's "unblock live-installer/6" hint useless (and ultimately still leaving RC bugs in lenny). Regards, -- Chris Lamb, UK [EMAIL PROTECTED] GPG: 0x634F9A20 signature.asc Description: PGP signature
