On Tue, Oct 13, 2009 at 02:21, Roman Kyrylych wrote:

Upstream update. . . . Please signoff.

I also prepared an update to klibc-kbd. Due to options=('force') not working with versioned depends mkinitcpio has to be updated as well. I don't have i686 at the moment, so someone please update klibc-kbd (i686 and x86_64) and mkinitcpio (any) from SVN trunk at once so that -Syu from testing won't complain about deps. Time to get some sleep, 03:05 here. :-P P.S.: can this "versioned deps vs. options=(force)" be solved somehow? -- Roman Kyrylych (Роман Кирилич)

On Wed, Oct 14, 2009 at 06:29, Allan McRae wrote:

Roman Kyrylych wrote:

...

Upstream update.

Changelog:

...

Please signoff.

...

sudo /bin/loadkeys -q -u us Password: KDSKBENT: Invalid argument failed to bind key 6 to value 61604 KDSKBENT: Invalid argument failed to bind key 46 to value 61602 KDSKBDIACRUC: Invalid argument

The 1.15 version of kbd works fine.

I'll investigate this. Did anyone experience a similar issue? -- Roman Kyrylych (Роман Кирилич)

Roman Kyrylych wrote:

On Wed, Oct 14, 2009 at 12:05, Roman Kyrylych wrote:

...

On Wed, Oct 14, 2009 at 06:29, Allan McRae wrote:

...

Roman Kyrylych wrote:

...

Upstream update.

Changelog:

Please signoff.

sudo /bin/loadkeys -q -u us

Password: KDSKBENT: Invalid argument failed to bind key 6 to value 61604 KDSKBENT: Invalid argument failed to bind key 46 to value 61602 KDSKBDIACRUC: Invalid argument

The 1.15 version of kbd works fine.

I'll investigate this. Did anyone experience a similar issue?

The keymap was not changed, but there were numerous changes in the parser. Do you use UTF-8 locale?

/etc/rc.conf: LOCALE="en_AU.UTF-8" KEYMAP="us" Allan

On Wed, Oct 14, 2009 at 13:10, Roman Kyrylych wrote:

On Wed, Oct 14, 2009 at 12:20, Allan McRae wrote:

...

Roman Kyrylych wrote:

...

On Wed, Oct 14, 2009 at 12:05, Roman Kyrylych wrote:

...

On Wed, Oct 14, 2009 at 06:29, Allan McRae wrote:

...

 sudo /bin/loadkeys -q -u us Password: KDSKBENT: Invalid argument failed to bind key 6 to value 61604 KDSKBENT: Invalid argument failed to bind key 46 to value 61602 KDSKBDIACRUC: Invalid argument

The 1.15 version of kbd works fine.

I'll investigate this. Did anyone experience a similar issue?

The keymap was not changed, but there were numerous changes in the parser. Do you use UTF-8 locale?

/etc/rc.conf: LOCALE="en_AU.UTF-8" KEYMAP="us"

I cannot reproduce the problem. :-/

Anyone else can test this please? This is a must-have package. ;-) -- Roman Kyrylych (Роман Кирилич)

On Wed, Oct 14, 2009 at 12:10, Roman Kyrylych wrote:

On Wed, Oct 14, 2009 at 12:20, Allan McRae wrote:

...

Roman Kyrylych wrote:

...

On Wed, Oct 14, 2009 at 12:05, Roman Kyrylych wrote:

...

On Wed, Oct 14, 2009 at 06:29, Allan McRae wrote:

...

 sudo /bin/loadkeys -q -u us Password: KDSKBENT: Invalid argument failed to bind key 6 to value 61604 KDSKBENT: Invalid argument failed to bind key 46 to value 61602 KDSKBDIACRUC: Invalid argument

The 1.15 version of kbd works fine.

I'll investigate this. Did anyone experience a similar issue?

The keymap was not changed, but there were numerous changes in the parser. Do you use UTF-8 locale?

/etc/rc.conf: LOCALE="en_AU.UTF-8" KEYMAP="us"

I cannot reproduce the problem. :-/

Okay, found the reason.

From http://bugs.gentoo.org/289265 : "kbd 1.15 used to temporarily set the kbd_mode to unicode with loadkeys -u. kbd 1.15.1 doesn't. Changing the init script to run kbd_mode before loadkeys gets rid of the errors."

But in our rc.sysinit there is this code: if echo "$LOCALE" | /bin/grep -qi utf ; then stat_busy "Setting Consoles to UTF-8 mode" # UTF-8 consoles are default since 2.6.24 kernel # this code is needed not only for older kernels, # but also when user has set vt.default_utf8=0 but LOCALE is *.UTF-8. for i in /dev/tty[0-9]*; do /usr/bin/kbd_mode -u < ${i} printf "\033%%G" > ${i} done # the $CONSOLE check helps us avoid this when running scripts from cron echo 'if [ "$CONSOLE" = "" -a "$TERM" = "linux" -a -t 1 ]; then printf "\033%%G"; fi' >>/etc/profile.d/locale.sh stat_done [ -n "$KEYMAP" ] && status "Loading Keyboard Map: $KEYMAP" /bin/loadkeys -q -u $KEYMAP else stat_busy "Setting Consoles to legacy mode" # make non-UTF-8 consoles work on 2.6.24 and newer kernels for i in /dev/tty[0-9]*; do /usr/bin/kbd_mode -a < ${i} printf "\033%%@" > ${i} done # the $CONSOLE check helps us avoid this when running scripts from cron echo 'if [ "$CONSOLE" = "" -a "$TERM" = "linux" -a -t 1 ]; then printf "\033%%@"; fi' >>/etc/profile.d/locale.sh stat_done [ -n "$KEYMAP" ] && status "Loading Keyboard Map: $KEYMAP" /bin/loadkeys -q $KEYMAP fi So it looks like the problem should not happen on Arch when you have 'utf' (case insensitive) in your LOCALE. To Allan and everyone who have the "KDSKBENT: Invalid argument" issue: * does your system says 'Setting Consoles to UTF-8 mode' during boot? * what does the output of `kbd_mode` runned from a tty says? The patch that makes sure loadkeys -u first sets console in UTF-8 mode has been submitted upstream, but I am not sure if there is a need to apply it to our package immediately, because there is no reason why correct Arch config should have this problem. P.S.: Note that /usr/bin/kbd_mode -u < ${i} may look weird, but from man kbd_mode : "kbd_mode operates on the console specified by the "-C" option; if there is none, the console associated with stdin is used." And according to my checks < ${i} works, however it may be better to use -C. -- Roman Kyrylych (Роман Кирилич)

On Mon, Nov 16, 2009 at 02:24, Allan McRae wrote:

Roman Kyrylych wrote:

...

So it looks like the problem should not happen on Arch when you have 'utf' (case insensitive) in your LOCALE.

To Allan and everyone who have the "KDSKBENT: Invalid argument" issue: * does your system says 'Setting Consoles to UTF-8 mode' during boot? * what does the output of `kbd_mode` runned from a tty says?

I get the UTF-8 message during boot and kbd_mode says the keyboards is in UTF-8 mode.

But something has chnaged since I last looked at this.  I no longer get the KDSKBENT message, just

...

sudo /bin/loadkeys -q -u us KDSKBDIACRUC: Invalid argument

Google finds only this: http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=540534 Seems to be i686 only. Can anyone else reproduce this? It is weird that there is only one referrence to the issue on the whole internets (two if counting this thread). :-/

I have no idea what changed... (new initscripts?)

Only one relevant change was made: http://projects.archlinux.org/initscripts.git/commit/?id=65f4102d48726afa432... -- Roman Kyrylych (Роман Кирилич)

Roman Kyrylych wrote:

On Mon, Nov 16, 2009 at 12:33, Roman Kyrylych wrote:

...

On Mon, Nov 16, 2009 at 02:24, Allan McRae wrote:

...

Roman Kyrylych wrote:

...

So it looks like the problem should not happen on Arch when you have 'utf' (case insensitive) in your LOCALE.

To Allan and everyone who have the "KDSKBENT: Invalid argument" issue: * does your system says 'Setting Consoles to UTF-8 mode' during boot? * what does the output of `kbd_mode` runned from a tty says? I get the UTF-8 message during boot and kbd_mode says the keyboards is in UTF-8 mode.

But something has chnaged since I last looked at this. I no longer get the KDSKBENT message, just

...

sudo /bin/loadkeys -q -u us KDSKBDIACRUC: Invalid argument Google finds only this: http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=540534 Seems to be i686 only.

Do you have i686 with x86_64 kernel?

Umm... yes! I should probably install the x86_64 version for that it seems. Allan

Roman Kyrylych wrote:

On Mon, Nov 16, 2009 at 12:42, Allan McRae wrote:

...

Roman Kyrylych wrote:

...

On Mon, Nov 16, 2009 at 12:33, Roman Kyrylych wrote:

...

On Mon, Nov 16, 2009 at 02:24, Allan McRae wrote:

...

KDSKBDIACRUC: Invalid argument Google finds only this: http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=540534 Seems to be i686 only. Do you have i686 with x86_64 kernel? Umm... yes! I should probably install the x86_64 version for that it seems.

Okay, so this seems like a minor issue (by number of users who may be affected). Is it okay to move the package to Core in this case? I will report the problem to upstream.

BTW, does previous version work on i686 system with x86_64 kernel or not? (judging from Debian's report it shouldn't but who knows)

The version in [core] works for me. So it seems to be something in Debian's patchset that made it into 1.15.1. Anyway, I will signoff both. I only know two other people stupid enough to do this... :P Allan

Roman Kyrylych wrote:

On Mon, Nov 16, 2009 at 13:50, Roman Kyrylych wrote:

...

On Mon, Nov 16, 2009 at 12:42, Allan McRae wrote:

...

Roman Kyrylych wrote:

...

On Mon, Nov 16, 2009 at 12:33, Roman Kyrylych wrote:

...

On Mon, Nov 16, 2009 at 02:24, Allan McRae wrote:

...

KDSKBDIACRUC: Invalid argument Google finds only this: http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=540534 Seems to be i686 only. Do you have i686 with x86_64 kernel? Umm... yes! I should probably install the x86_64 version for that it seems. Okay, so this seems like a minor issue (by number of users who may be affected). Is it okay to move the package to Core in this case? I will report the problem to upstream.

Got this reply on kbd mailing list: <quote>

...

This is a kernel issue (KDSKBDIACRUC is not in ioctl32()’s compatibility list). I’m seeing about it.

Try upgrading to 2.6.32; the problem is already fixed there. </quote>

I was sure I had replied saying that the kernel upgrade had fixed this... it appears not! Signoff both, Allan