[BRLTTY] 4.0

[Samuel Thibault]

高生旺, le Mon 20 Apr 2009 22:58:01 +0800, a écrit :
> On Mon, 20 Apr 2009, Samuel Thibault wrote:
> 
> > 高生旺, le Mon 20 Apr 2009 21:51:51 +0800, a ecrit :
> > > On Sun, 19 Apr 2009, Samuel Thibault wrote:
> > > > 高生旺, le Sun 19 Apr 2009 09:50:00 +0800, a ecrit :
> > > > > On Sat, 18 Apr 2009, Samuel Thibault wrote:
> > > > > >> 1. Zh_tw.ctb doesn't work on kernel 2.6.28 and upper.
> > > > > >> 2. Console's unicode problem isn't resolved yet.
> > > > > > That's still a kernel-side issue that brltty can't solve itself.
> > > > > > About the 2.6.28 break, would you be able to recompile a kernel? I think
> > > > > > I know which change broke it.
> > > > > 
> > > > > Yes. Which options should be on or ooff?
> > > > 
> > > > It's not an option, it's a patch, attached to this mail.
> > > 
> > > First problem resolved. Thanks!
> > 
> > Ok.  I hope it will get applied to 2.6.30.  As for the 128+27 character,
> > the attached patch can be used to fix it.  Alan Cox however said that he
> > was reluctant to apply it.
> Compile failed.

Oops, sorry, here is a fixed patch.

Samuel
-------------- next part --------------
--- drivers/char/vt.c.orig	2009-04-20 16:25:24.000000000 +0200
+++ drivers/char/vt.c	2009-04-20 16:25:36.000000000 +0200
@@ -1708,9 +1708,6 @@
 	case 127:
 		del(vc);
 		return;
-	case 128+27:
-		vc->vc_state = ESsquare;
-		return;
 	}
 	switch(vc->vc_state) {
 	case ESesc:
@@ -2256,8 +2252,7 @@
                 ok = tc && (c >= 32 ||
 			    !(vc->vc_disp_ctrl ? (CTRL_ALWAYS >> c) & 1 :
 				  vc->vc_utf || ((CTRL_ACTION >> c) & 1)))
-			&& (c != 127 || vc->vc_disp_ctrl)
-			&& (c != 128+27);
+			&& (c != 127 || vc->vc_disp_ctrl);
 
 		if (vc->vc_state == ESnormal && ok) {
 			if (vc->vc_utf && !vc->vc_disp_ctrl) {