[PATCH] vt: Expose system-wide UTF-8 default setting via sysfs

[PATCH] vt: Expose system-wide UTF-8 default setting via sysfs

[Antonino A. Daplas]

Create a variable, default_utf8, that defines the system-wide default UTF-8
setting.  This variable can be altered via sysfs. If the variable is properly
set, this should mimimize breakage of UTF-8 encoded consoles when doing a
reset or echo -e '\033c' and of newly opened/allocated consoles.

This is based from patches by Jan Engelhardt and Paul LeoNerd Evans.

Signed-off-by: Antonino Daplas <adaplas@gmail.com>
---
> I think you're missing the whole point of console reset.  Its purpose is 
> to force the console into a known-good state.  The fewer pieces of state 
> it leaves unset, the better.  To some degree it's less important what 
> that state actually is.

Okay, you convinced me. Hopefully this is acceptable to all parties.

Andrew,

If everybody agrees, can you drop the previous patch I sent to you, and use
this instead?

Tony

 drivers/char/vt.c |    4 +++-
 1 files changed, 3 insertions(+), 1 deletions(-)

diff --git a/drivers/char/vt.c b/drivers/char/vt.c
index 1bbb45b..8aca96f 100644
--- a/drivers/char/vt.c
+++ b/drivers/char/vt.c
@@ -157,6 +157,8 @@ static void blank_screen_t(unsigned long
 static void set_palette(struct vc_data *vc);
 
 static int printable;		/* Is console ready for printing? */
+static int default_utf8;
+module_param(default_utf8, int, S_IRUGO | S_IWUSR);
 
 /*
  * ignore_poke: don't unblank the screen when things are typed.  This is
@@ -1497,7 +1499,7 @@ static void reset_terminal(struct vc_dat
 	vc->vc_charset		= 0;
 	vc->vc_need_wrap	= 0;
 	vc->vc_report_mouse	= 0;
-	vc->vc_utf		= 0;
+	vc->vc_utf              = default_utf8;
 	vc->vc_utf_count	= 0;
 
 	vc->vc_disp_ctrl	= 0;

Re: [PATCH] vt: Expose system-wide UTF-8 default setting via sysfs

[Alexander E. Patrakov]

Antonino A. Daplas wrote:
> Create a variable, default_utf8, that defines the system-wide default UTF-8
> setting.  This variable can be altered via sysfs. If the variable is properly
> set, this should mimimize breakage of UTF-8 encoded consoles when doing a
> reset or echo -e '\033c' and of newly opened/allocated consoles.
> 
> This is based from patches by Jan Engelhardt and Paul LeoNerd Evans.
> 
> Signed-off-by: Antonino Daplas <adaplas@gmail.com>
> ---
>> I think you're missing the whole point of console reset.  Its purpose is 
>> to force the console into a known-good state.  The fewer pieces of state 
>> it leaves unset, the better.  To some degree it's less important what 
>> that state actually is.
> 
> Okay, you convinced me. Hopefully this is acceptable to all parties.
> 
> Andrew,
> 
> If everybody agrees, can you drop the previous patch I sent to you, and use
> this instead?
> 
> Tony
> +static int default_utf8;
> +module_param(default_utf8, int, S_IRUGO | S_IWUSR);

Module parameter without description and documentation? Yes, I understand 
that it is impossible to make vt a module. How about adding a line to 
Documentation/kernel-parameters.txt?

Other than that, the patch looks like a useful change.

-- 
Alexander E. Patrakov

Re: [PATCH] vt: Expose system-wide UTF-8 default setting via sysfs

[Antonino A. Daplas]

On Tue, 2007-04-03 at 10:06 +0600, Alexander E. Patrakov wrote:
> Antonino A. Daplas wrote:
> > Create a variable, default_utf8, that defines the system-wide default UTF-8
> > setting.  This variable can be altered via sysfs. If the variable is properly
> > set, this should mimimize breakage of UTF-8 encoded consoles when doing a
> > reset or echo -e '\033c' and of newly opened/allocated consoles.
> > 
> > This is based from patches by Jan Engelhardt and Paul LeoNerd Evans.
> > 
> > Signed-off-by: Antonino Daplas <adaplas@gmail.com>
> > ---
> >> I think you're missing the whole point of console reset.  Its purpose is 
> >> to force the console into a known-good state.  The fewer pieces of state 
> >> it leaves unset, the better.  To some degree it's less important what 
> >> that state actually is.
> > 
> > Okay, you convinced me. Hopefully this is acceptable to all parties.
> > 
> > Andrew,
> > 
> > If everybody agrees, can you drop the previous patch I sent to you, and use
> > this instead?
> > 
> > Tony
> > +static int default_utf8;
> > +module_param(default_utf8, int, S_IRUGO | S_IWUSR);
> 
> Module parameter without description and documentation? Yes, I understand 
> that it is impossible to make vt a module. How about adding a line to 
> Documentation/kernel-parameters.txt?

I'll do that (and I'll also include Jan's palette patch) once I'm sure
there's no violent objection against the change.

Tony 

Re: [PATCH] vt: Expose system-wide UTF-8 default setting via sysfs

[Pavel Machek]

Hi!

> Create a variable, default_utf8, that defines the system-wide default UTF-8
> setting.  This variable can be altered via sysfs. If the variable is properly
> set, this should mimimize breakage of UTF-8 encoded consoles when doing a
> reset or echo -e '\033c' and of newly opened/allocated consoles.

HPA is right... this should be fixed in userland. Reset should reset a
console, and if you want utf-8, do \ec\ewhatever to get it.

-- 
(english) http://www.livejournal.com/~pavelmachek
(cesky, pictures) http://atrey.karlin.mff.cuni.cz/~pavel/picture/horses/blog.html

Re: [PATCH] vt: Expose system-wide UTF-8 default setting via sysfs

[Paul LeoNerd Evans]

[-- Attachment #1: Type: text/plain, Size: 929 bytes --]

On Tue, Apr 03, 2007 at 04:20:52PM +0000, Pavel Machek wrote:
> HPA is right... this should be fixed in userland. Reset should reset a
> console, and if you want utf-8, do \ec\ewhatever to get it.

As I've already said elsewhere, does anything say that "reset" means
that UTF-8 is turned off, or merely reset back to some known state? I
personally would consider that UTF-8 being on is the default, and if
something wants it off it should ask specifically..

Consider any other terminal emulator, such as those running on X11; when
they're reset, they know to keep UTF-8 mode turned on, because it said
so in their config, or in the locale. Kernel doesn't really have nice
access to the LANG environment variable, but we can give it something
similar; namely, via a module parameter.

-- 
Paul "LeoNerd" Evans

leonerd@leonerd.org.uk
ICQ# 4135350       |  Registered Linux# 179460
http://www.leonerd.org.uk/

[-- Attachment #2: Digital signature --]
[-- Type: application/pgp-signature, Size: 189 bytes --]

Re: [PATCH] vt: Expose system-wide UTF-8 default setting via sysfs

[Antonino A. Daplas]

On Tue, 2007-04-03 at 17:35 +0100, Paul LeoNerd Evans wrote:
> On Tue, Apr 03, 2007 at 04:20:52PM +0000, Pavel Machek wrote:
> > HPA is right... this should be fixed in userland. Reset should reset a
> > console, and if you want utf-8, do \ec\ewhatever to get it.
> 
> As I've already said elsewhere, does anything say that "reset" means
> that UTF-8 is turned off, or merely reset back to some known state? I
> personally would consider that UTF-8 being on is the default, and if
> something wants it off it should ask specifically..
> 
> Consider any other terminal emulator, such as those running on X11; when
> they're reset, they know to keep UTF-8 mode turned on, because it said
> so in their config, or in the locale. Kernel doesn't really have nice
> access to the LANG environment variable, but we can give it something
> similar; namely, via a module parameter.

I agree with Paul.  Reset or RIS is (R)eset to (I)nitial (S)tate,
represented as ESC 06/03. "Initial state" is defined in ECMA-48 as the
state of the device after it was made operational. It doesn't
necessarily mean that the "initial state" is the safest configuration.
So, if this "initial state" includes enabling UTF-8, then the device has
to go into UTF-8 upon invoking RIS.

Tony

Re: [PATCH] vt: Expose system-wide UTF-8 default setting via sysfs

[Pavel Machek]

Hi!

> > > HPA is right... this should be fixed in userland. Reset should reset a
> > > console, and if you want utf-8, do \ec\ewhatever to get it.
> > 
> > As I've already said elsewhere, does anything say that "reset" means
> > that UTF-8 is turned off, or merely reset back to some known state? I
> > personally would consider that UTF-8 being on is the default, and if
> > something wants it off it should ask specifically..
> > 
> > Consider any other terminal emulator, such as those running on X11; when
> > they're reset, they know to keep UTF-8 mode turned on, because it said
> > so in their config, or in the locale. Kernel doesn't really have nice
> > access to the LANG environment variable, but we can give it something
> > similar; namely, via a module parameter.
> 
> I agree with Paul.  Reset or RIS is (R)eset to (I)nitial (S)tate,
> represented as ESC 06/03. "Initial state" is defined in ECMA-48 as the
> state of the device after it was made operational. It doesn't
> necessarily mean that the "initial state" is the safest configuration.
> So, if this "initial state" includes enabling UTF-8, then the device has
> to go into UTF-8 upon invoking RIS.

You could do that... during development series. But this is
userspace-visible API change, and those are not allowed during stable.

...and having /sysctl configuring inital state for reset is just plain
ugly.


									Pavel
-- 
(english) http://www.livejournal.com/~pavelmachek
(cesky, pictures) http://atrey.karlin.mff.cuni.cz/~pavel/picture/horses/blog.html

Re: [PATCH] vt: Expose system-wide UTF-8 default setting via sysfs

[Paul LeoNerd Evans]

[-- Attachment #1: Type: text/plain, Size: 757 bytes --]

On Wed, Apr 04, 2007 at 10:01:40AM +0200, Pavel Machek wrote:
> You could do that... during development series. But this is
> userspace-visible API change, and those are not allowed during stable.
> 
> ...and having /sysctl configuring inital state for reset is just plain
> ugly.

Well, how else would you suggest to do it? Bearing in mind all other
terminal emulators (xterm, gnome-terminal, konsole, and so on) all have
other out-of-band ways to determine this - config files, menu options,
locale variables,... None of those are particularlly viable for the
kernel. I vote this is the closest thing we've got.

-- 
Paul "LeoNerd" Evans

leonerd@leonerd.org.uk
ICQ# 4135350       |  Registered Linux# 179460
http://www.leonerd.org.uk/

[-- Attachment #2: Digital signature --]
[-- Type: application/pgp-signature, Size: 189 bytes --]

Re: [PATCH] vt: Expose system-wide UTF-8 default setting via sysfs

[H. Peter Anvin]

Pavel Machek wrote:
> 
> You could do that... during development series. But this is
> userspace-visible API change, and those are not allowed during stable.
> 

Only if you change the default setting.

	-hpa