Bug#619711: console-setup: breaks copying keymap to initramfs
 

reassign 619711 initramfs-tools
thanks

On Sat, Mar 26, 2011 at 12:49:24PM +0100, Mario 'BitKoenig' Holbe wrote:
>
> if a system's keymap needs to be loaded during the initramfs stage,
> initramfs-tools' /usr/share/initramfs-tools/hooks/keymap looks for
> /etc/console-setup/cached.kmap.gz
> and copies it to the initramfs.

I had no idea that initramfs-tools does this. I think this was wrong
even with old version of console-setup because in some situations
cached.kmap.gz will not correspond to the actual configuration and in
some versions of the package it was even possible for this file not to
exist.

> console-setup 1.71 changed the name of this file to
> /etc/console-setup/cached_${CHARMAP}_$backspace$VARIANT.kmap.gz
> i.e. something like
> /etc/console-setup/cached_ISO-8859-15_del.kmap.gz

The new version of console-setup tries to guess good values for several
configuration options if values are not provided in its configuration
files. Because of this, it is possible (when the system environment
changes - locale and/or terminal settings) that a new keymap will be
required even when the user has not changed anything in the
configuration files of console-setup. Hence, it would be impossible for
console-setup to tell whether cached.kmap.gz can be used or the keymap
has to be regenerated.

> There are several alternatives to fix this bug like
> * symlinking the new name to the old
> * moving the keymap-copying from initramfs-tools to console-setup
> * updating initramfs-tools to honor the new keymap name

It will be difficult for initramfs-tools to tell what is the correct
name of the cached keymap. Symlinking the new name to the old is
unreliable so I'd prefer not to implement this. Because of this in
version 1.72 of console-setup a new option of setupcon is implemented
--save-keyboard. Suppose you want to save the keymap in
/tmp/initrd/etc/console-setup/cached.kmap.gz. Then simply use the
following command:

setupcon --save-keyboard /tmp/initrd/etc/console-setup/cached.kmap.gz

Alternatively, instead of --save-keyboard you can use --setup-dir. I
will explain this second option in a separate bug report. Both new
options of setupcon are for now undocumented because I want to know
first that they will be useful.

Anton Zinoviev


--
To UNSUBSCRIBE, email to debian-kernel-REQUEST@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
Archive: 20110329120457.GA25963@debian.lan">http://lists.debian.org/20110329120457.GA25963@debian.lan

Bug#619711: console-setup: breaks copying keymap to initramfs
 

severity 619711 important
stop

default is KEYMAP=n.

On Tue, 29 Mar 2011, Anton Zinoviev wrote:

> I had no idea that initramfs-tools does this. I think this was wrong
> even with old version of console-setup because in some situations
> cached.kmap.gz will not correspond to the actual configuration and in
> some versions of the package it was even possible for this file not to
> exist.

people using cryptsetup asked for it in order to be able to pass
their luks passphrase.

>
> It will be difficult for initramfs-tools to tell what is the correct
> name of the cached keymap. Symlinking the new name to the old is
> unreliable so I'd prefer not to implement this. Because of this in
> version 1.72 of console-setup a new option of setupcon is implemented
> --save-keyboard. Suppose you want to save the keymap in
> /tmp/initrd/etc/console-setup/cached.kmap.gz. Then simply use the
> following command:
>
> setupcon --save-keyboard /tmp/initrd/etc/console-setup/cached.kmap.gz
>
> Alternatively, instead of --save-keyboard you can use --setup-dir. I
> will explain this second option in a separate bug report. Both new
> options of setupcon are for now undocumented because I want to know
> first that they will be useful.

I'd prerfer that console-setup would ship an proper initramfs hook
telling what to add to initramfs, will followup on the other report
on that soonest.

thanks for your input.

--
maks


--
To UNSUBSCRIBE, email to debian-kernel-REQUEST@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
Archive: 20110405210049.GA5925@stro.at">http://lists.debian.org/20110405210049.GA5925@stro.at

Bug#619711: console-setup: breaks copying keymap to initramfs
 

* maximilian attems [Die Apr 05, 2011 at 11:00:49 +0200]:
> On Tue, 29 Mar 2011, Anton Zinoviev wrote:

[...]

> > It will be difficult for initramfs-tools to tell what is the correct
> > name of the cached keymap. Symlinking the new name to the old is
> > unreliable so I'd prefer not to implement this. Because of this in
> > version 1.72 of console-setup a new option of setupcon is implemented
> > --save-keyboard. Suppose you want to save the keymap in
> > /tmp/initrd/etc/console-setup/cached.kmap.gz. Then simply use the
> > following command:

> > setupcon --save-keyboard /tmp/initrd/etc/console-setup/cached.kmap.gz

> > Alternatively, instead of --save-keyboard you can use --setup-dir. I
> > will explain this second option in a separate bug report. Both new
> > options of setupcon are for now undocumented because I want to know
> > first that they will be useful.

> I'd prerfer that console-setup would ship an proper initramfs hook
> telling what to add to initramfs, will followup on the other report
> on that soonest.

So, how could we resolve this issue?

Anton, is there any chance you could provide an initramfs hook which
takes care of that? (The advantage of this approach would be that
the hook can use features of console-setup according to its version
and you'd have full control over it.)

regards,
-mika-

Bug#619711: console-setup: breaks copying keymap to initramfs
 

On Wed, Nov 23, 2011 at 12:40:05PM +0100, Michael Prokop wrote:
>
> > I'd prerfer that console-setup would ship an proper initramfs hook
> > telling what to add to initramfs, will followup on the other report
> > on that soonest.
>
> So, how could we resolve this issue?
>
> Anton, is there any chance you could provide an initramfs hook which
> takes care of that? (The advantage of this approach would be that
> the hook can use features of console-setup according to its version
> and you'd have full control over it.)

Well, the situation is that I am ready to devote as much time as it is
required in order to develop whatever is required. The problem is I am
not prepared and I don't want to maintain this afterwards (for personal
rather than technical reasons).

There are some other people helping with the development of
console-setup. They are members of debian-boot list, hopefuly an email
to this list will be helpful?

So it seems there are three options:

1. A current contributor of console-setup volunteers to help.

2. Someone from the kernel team maintains the hooks in console-setup (in
this case I can do any requested initial development, documentation,
etc.)

3. Something like --save-keyboard or --setup-dir is used.

Anton Zinoviev




--
To UNSUBSCRIBE, email to debian-kernel-REQUEST@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
Archive: 20111123143700.GB10734@logic.fmi.uni-sofia.bg">http://lists.debian.org/20111123143700.GB10734@logic.fmi.uni-sofia.bg

Bug#619711: console-setup: breaks copying keymap to initramfs
 

Package: initramfs-tools
Version: 0.99
Followup-For: Bug #619711

Hello,

please find attached a patch for the keymap hook, which fixes the bug as
described by Anton Zinoviev <anton@lml.bas.bg> in the email 'Re: Bug#619711:
console-setup: breaks copying keymap to initramfs' from Tue, 29 Mar 2011
15:04:57 +0300.

Regards,
Andreas



-- Package-specific info:
-- initramfs sizes
-- /proc/cmdline
BOOT_IMAGE=/vmlinuz-3.2.0-1-amd64 root=/dev/mapper/andreas--notebook-root ro quiet splash

-- resume
RESUME=/dev/mapper/andreas--notebook-swap_1
-- /proc/filesystems
ext4
ext2
fuseblk

-- lsmod
Module Size Used by
ip6table_filter 12540 0
ip6_tables 22175 1 ip6table_filter
ebtable_nat 12580 0
ebtables 26235 1 ebtable_nat
ipt_MASQUERADE 12594 3
iptable_nat 12928 1
nf_nat 18242 2 ipt_MASQUERADE,iptable_nat
nf_conntrack_ipv4 14078 4 iptable_nat,nf_nat
nf_defrag_ipv4 12483 1 nf_conntrack_ipv4
xt_state 12503 1
nf_conntrack 52720 5 ipt_MASQUERADE,iptable_nat,nf_nat,nf_conntrack_ipv 4,xt_state
ipt_REJECT 12502 2
xt_CHECKSUM 12471 1
iptable_mangle 12536 1
acpi_cpufreq 12935 1
mperf 12453 1 acpi_cpufreq
xt_tcpudp 12570 5
cpufreq_stats 12866 0
cpufreq_powersave 12454 0
cpufreq_userspace 12576 0
cpufreq_conservative 13147 0
iptable_filter 12536 1
ip_tables 22042 3 iptable_nat,iptable_mangle,iptable_filter
x_tables 19073 12 ip6table_filter,ip6_tables,ebtables,ipt_MASQUERADE ,iptable_nat,xt_state,ipt_REJECT,xt_CHECKSUM,iptab le_mangle,xt_tcpudp,iptable_filter,ip_tables
bridge 70728 0
stp 12392 1 bridge
hdaps 13616 1
parport_pc 22364 0
ppdev 12763 0
lp 17149 0
parport 31858 3 parport_pc,ppdev,lp
bnep 17567 2
rfcomm 33656 0
binfmt_misc 12957 1
uinput 17440 1
fuse 61981 1
nfsd 211858 2
nfs 312062 0
lockd 67328 2 nfsd,nfs
fscache 36739 1 nfs
auth_rpcgss 37143 2 nfsd,nfs
nfs_acl 12511 2 nfsd,nfs
sunrpc 173671 6 nfsd,nfs,lockd,auth_rpcgss,nfs_acl
kvm_intel 121968 0
kvm 287572 1 kvm_intel
ext2 59231 1
tp_smapi 23531 0
thinkpad_ec 12857 2 hdaps,tp_smapi
loop 22641 0
snd_hda_codec_hdmi 30783 1
snd_hda_codec_conexant 45245 1
btusb 17502 0
bluetooth 119406 11 bnep,rfcomm,btusb
uvcvideo 57744 0
videodev 70889 1 uvcvideo
snd_hda_intel 26345 2
snd_hda_codec 77994 3 snd_hda_codec_hdmi,snd_hda_codec_conexant,snd_hda_ intel
media 18148 2 uvcvideo,videodev
arc4 12458 2
v4l2_compat_ioctl32 16655 1 videodev
joydev 17266 0
hid_logitech_dj 17247 0
snd_hwdep 13186 1 snd_hda_codec
snd_pcm 63900 3 snd_hda_codec_hdmi,snd_hda_intel,snd_hda_codec
snd_seq 45093 0
snd_timer 22917 2 snd_pcm,snd_seq
iwlwifi 166703 0
snd_seq_device 13176 1 snd_seq
mac80211 192768 1 iwlwifi
ac 12624 0
thinkpad_acpi 61270 0
battery 13109 0
power_supply 13475 2 ac,battery
nvram 13049 1 thinkpad_acpi
iTCO_wdt 17081 0
iTCO_vendor_support 12704 1 iTCO_wdt
snd 52850 14 snd_hda_codec_hdmi,snd_hda_codec_conexant,snd_hda_ intel,snd_hda_codec,snd_hwdep,snd_pcm,snd_seq,snd_ timer,snd_seq_device,thinkpad_acpi
cfg80211 137140 2 iwlwifi,mac80211
i2c_i801 16870 0
rfkill 19012 5 bluetooth,thinkpad_acpi,cfg80211
psmouse 64455 0
serio_raw 12931 0
evdev 17562 23
soundcore 13065 1 snd
snd_page_alloc 13003 2 snd_hda_intel,snd_pcm
wmi 13243 0
processor 28059 1 acpi_cpufreq
ext4 350557 1
mbcache 13065 2 ext2,ext4
jbd2 62015 1 ext4
crc16 12343 2 bluetooth,ext4
sha256_generic 16797 2
aesni_intel 50643 24
cryptd 14517 9 aesni_intel
aes_x86_64 16796 1 aesni_intel
aes_generic 33026 2 aesni_intel,aes_x86_64
cbc 12754 0
dm_crypt 22586 1
dm_mod 63545 9 dm_crypt
sr_mod 21899 0
sd_mod 36136 3
cdrom 35401 1 sr_mod
crc_t10dif 12348 1 sd_mod
usbhid 36379 1 hid_logitech_dj
hid 81288 2 hid_logitech_dj,usbhid
i915 351781 4
drm_kms_helper 27227 1 i915
drm 167670 5 i915,drm_kms_helper
i2c_algo_bit 12841 1 i915
ahci 24997 2
libahci 22860 1 ahci
xhci_hcd 73208 0
libata 140589 2 ahci,libahci
ehci_hcd 40215 0
scsi_mod 162417 3 sr_mod,sd_mod,libata
button 12937 1 i915
e1000e 124918 0
sdhci_pci 17976 0
sdhci 27053 1 sdhci_pci
mmc_core 72460 2 sdhci_pci,sdhci
usbcore 128498 6 btusb,uvcvideo,usbhid,xhci_hcd,ehci_hcd
usb_common 12354 1 usbcore
thermal 17383 0
i2c_core 23876 6 videodev,i2c_i801,i915,drm_kms_helper,drm,i2c_algo _bit
video 17628 1 i915
thermal_sys 18040 3 processor,thermal,video

-- /etc/initramfs-tools/modules
intel_agp
drm
i915 modeset=1

-- /etc/kernel-img.conf
# Kernel image management overrides
# See kernel-img.conf(5) for details
do_symlinks = yes
do_bootloader = no
do_initrd = yes
link_in_boot = no

-- /etc/initramfs-tools/initramfs.conf
MODULES=most
BUSYBOX=y
KEYMAP=y
COMPRESS=gzip
BOOT=local
DEVICE=
NFSROOT=auto

-- /etc/initramfs-tools/update-initramfs.conf
update_initramfs=yes
backup_initramfs=no

-- /etc/crypttab
sda5_crypt UUID=7326340f-32f9-4ece-84dc-e309b3b82454 none luks

-- mkinitramfs hooks
/etc/initramfs-tools/hooks/:

/usr/share/initramfs-tools/hooks:
busybox
cryptgnupg
cryptkeyctl
cryptopenct
cryptopensc
cryptpassdev
cryptroot
dmsetup
fuse
keymap
klibc
lvm2
ntfs_3g
plymouth
thermal
udev


-- System Information:
Debian Release: wheezy/sid
APT prefers testing
APT policy: (500, 'testing')
Architecture: amd64 (x86_64)

Kernel: Linux 3.2.0-1-amd64 (SMP w/4 CPU cores)
Locale: LANG=de_DE.utf8, LC_CTYPE=de_DE.utf8 (charmap=UTF-8)
Shell: /bin/sh linked to /bin/dash

Versions of packages initramfs-tools depends on:
ii cpio 2.11-7
ii findutils 4.4.2-4
ii klibc-utils 1.5.25-1.1
ii module-init-tools 3.16-1
ii udev 175-3

Versions of packages initramfs-tools recommends:
ii busybox 1:1.19.3-5

Versions of packages initramfs-tools suggests:
ii bash-completion 1:1.99-3

-- Configuration Files:
/etc/initramfs-tools/initramfs.conf changed [not included]

-- no debconf information

Bug#619711: console-setup: breaks copying keymap to initramfs
 

On Wed, Feb 15, 2012 at 02:28:13PM +0100, Andreas Altergott wrote:
>
> please find attached a patch for the keymap hook, which fixes the bug as
> described by Anton Zinoviev <anton@lml.bas.bg> in the email 'Re: Bug#619711:
> console-setup: breaks copying keymap to initramfs' from Tue, 29 Mar 2011
> 15:04:57 +0300.
>
> Regards,
> Andreas

thank you very much for looking into this.

could you please send the patch in unified form, (diff -u)
the one below is unreadble or directly use git:
http://anonscm.debian.org/gitweb/?p=kernel/initramfs-tools.git


best regards,
max



--
To UNSUBSCRIBE, email to debian-kernel-REQUEST@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
Archive: 20120215144025.GA30020@vostochny.stro.at">http://lists.debian.org/20120215144025.GA30020@vostochny.stro.at

Bug#619711: console-setup: breaks copying keymap to initramfs
 

Hi,

Am Mittwoch, den 15.02.2012, 14:40 +0000 schrieb maximilian attems:
> could you please send the patch in unified form, (diff -u)
> the one below is unreadble or directly use git:
> http://anonscm.debian.org/gitweb/?p=kernel/initramfs-tools.git

the patch generated with git --diff in unified form is attached to this
mail.


regards,
Andreas

Bug#619711: console-setup: breaks copying keymap to initramfs
 

On Wed, Feb 15, 2012 at 04:57:19PM +0100, Andreas Altergott wrote:
>
> +/bin/setupcon --save-keyboard ${DESTDIR}/etc/boottime.kmap.gz

I am not familiar with the method these scripts work (so I don't know if
this is important) but in principle it can be expected that in some
occasions setupcon will return non-zero exit code. This means something
has prevented it from saving the keyboard. In such cases an error
message will be printed to stderr.

Anton Zinoviev




--
To UNSUBSCRIBE, email to debian-kernel-REQUEST@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
Archive: 20120219154322.GA8735@debian.lan">http://lists.debian.org/20120219154322.GA8735@debian.lan

Bug#619711: console-setup: breaks copying keymap to initramfs
 

Hi,

is there any progress on this one? This recently bit me quite hard when
I installed a system with full-disk encryption, using a password which
required the correct keyboard layout to be entered.

Kind regards,
Ralf


--
To UNSUBSCRIBE, email to debian-kernel-REQUEST@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
Archive: 5051FCDE.2080006@ralfj.de">http://lists.debian.org/5051FCDE.2080006@ralfj.de

Bug#619711: console-setup: breaks copying keymap to initramfs
 

[I am sending CC also to 532852@bugs.debian.org]

On Thu, Sep 13, 2012 at 05:33:50PM +0200, Ralf Jung wrote:
>
> is there any progress on this one? This recently bit me quite hard when
> I installed a system with full-disk encryption, using a password which
> required the correct keyboard layout to be entered.

As far as I can understand, this bug can be fixed in two ways - by code
provided by console-setup, or by code provided by initramfs-tools.

Almost all of the code required in order to fix this bug has been
implemented in console-setup for quite some time. Unfortunately there
are two problems:

1. I (as one of the maintainers of console-setup) can not guarantee
that I will be able to maintain this code in future. Not because this
will be too complex or too time consuming (no, it won't be...), but
rather because I don't feel myself wanting to keep myself updated with
any future changes in the initrd build process.

2. I suppose the maintainers of initramfs-tools can say something
analogous...

I am willing to provide any required development, education, whatever,
the moment there appears a developer willing to maintain this.

Anton Zinoviev


--
To UNSUBSCRIBE, email to debian-kernel-REQUEST@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
Archive: 20120915114404.GA20544@logic.fmi.uni-sofia.bg">http://lists.debian.org/20120915114404.GA20544@logic.fmi.uni-sofia.bg