walter harms | 1 Sep 09:11 2008
Picon

Re: [patch] fix for autodocifier.pl uninitialized value in join


Cristian Ionescu-Idbohrn schrieb:
> On Sun, 31 Aug 2008, Denys Vlasenko wrote:
> 
>> On Saturday 30 August 2008 10:06, Cristian Ionescu-Idbohrn wrote:
>>
>> Applied, thanks!
> 
> Thanks
> 
>>> There's also some irritating colorization problem (I use emacs and
>>> font-lock) caused by unescaped double-quote characters in substitution
>>> expressions inside the beautify subroutine.  I have a patch that fixes
>>> that.  Would you be interested?
>> Sounds like emacs' bug in handling complex cases.
> 
> Yes.  emacs22 handles double-quote characters inside substitution
> expressions better than emacs21.  Still, double-quote characters _are_
> special.  Should be escaped, IMO.
> 
>> You decide whether it worth working around, considering that not many
>> people would edit autodocifier.pl in the course of hacking on busybox.
> 
> Not a big deal.  As said, I can provide such a patch if there's intrest.
> 

IMHO,
please provide the test. time will tell if its important but adding that beast
will mean nothing as long as you have it already there.
I would like to see a small comment why you did escape the double quotes.
(Continue reading)

Tobias Poschwatta | 1 Sep 09:42 2008
Picon

Re: insmod broken in 1.12.0

On Sun, Aug 31, 2008 at 11:58:41PM +0200, Denys Vlasenko wrote:
> Looks like "standard" insmod works the same?
> 
> # /app/module-init-tools-0.9.15-pre2/sbin/insmod camellia
> insmod: can't read 'camellia': No such file or directory

When was 0.9.15-pre2 released?

I have module-init-tools-3.4 and insmod works as expected, even after
moving /lib/modules/`uname -r` out of the way, copying the module file
to e.g. /tmp, and renaming it to xxx.ko.

insmod in busybox up to 1.11.2 works as expected. 1.12.0 is the first
version that doesn't.

Tobias
Vladimir Dronnikov | 1 Sep 10:20 2008
Picon

Re: insmod broken in 1.12.0

There is a patch by Timo for BB modutils, which is to be sent in a week, I hope.

--
Vladimir

_______________________________________________
busybox mailing list
busybox <at> busybox.net
http://busybox.net/cgi-bin/mailman/listinfo/busybox
Vladimir Dronnikov | 1 Sep 13:06 2008
Picon

Re: sv

Well, this immediately provides you with an alternative:
just use full directory name or cd to your servoce directory
and use "sv <cmd> ./<service_name>".

As might the others do, I use to treat service mostly as name, not path.
Executing "sv u dialup" from anywhere is nicer than "( cd /etc/init; sv u ./dialup )", IMO.
 
Having it configurable means that you cannot know beforehand
whaty is the default directory in some custom built busybox binary.

We face _exactly_ the same problem with /lib/modules of modutils, Denys. And modutils are of much harder use, right?
There is no problem to symlink /etc/init as /var/service ( if recursive_action won't go wild :), though...

--
Vladimir

_______________________________________________
busybox mailing list
busybox <at> busybox.net
http://busybox.net/cgi-bin/mailman/listinfo/busybox
dronnikov | 1 Sep 17:10 2008
Picon

halt: cosmetic

* FEATURE_WTMP is used to control -w option.
* indentation fixed

Please, consider applying.

TIA,
--
Vladimir

Attachment (halt.patch): application/octet-stream, 2679 bytes
_______________________________________________
busybox mailing list
busybox <at> busybox.net
http://busybox.net/cgi-bin/mailman/listinfo/busybox
Bernhard Reutner-Fischer | 1 Sep 17:26 2008
Picon

Re: halt: cosmetic

On Mon, Sep 01, 2008 at 08:10:50AM -0700, dronnikov <at> gmail.com wrote:
>* FEATURE_WTMP is used to control -w option.
>* indentation fixed
>
>Please, consider applying.

Applied, thanks!

PS: Can you please fix/extend your "From:" address to your full name?
Makes copy'n pasting the attribution simpler :)

cheers,
Bernhard Fischer | 1 Sep 17:32 2008
Picon

Re: [PATCH] httpd Accept Header Patch

On Tue, Aug 26, 2008 at 10:03:46PM +0200, Alina Friedrichsen wrote:
> With this patch the BusyBox httpd pass the "Accept:" and
> "Accept-Language:" header by the environment variables to the
> CGI-Script, so this can make Content Negotiation to deliver the page in
> the language, which was selected by the user in the browser settings,
> and/or serve the XHTML page with the right MIME-Type
> application/xhtml+xml to user agents which support it und text/html
> which don't.
>
> (Needed e.g. for OpenWrt LuCI.)
>
> Signed-off-by: Alina Friedrichsen <x-alina <at> gmx.net>

Sorry for the delay. I have added your patch as r23295.
Thanks!

>Index: networking/httpd.c
>===================================================================
>--- networking/httpd.c	(Revision 23211)
>+++ networking/httpd.c	(Arbeitskopie)
> <at>  <at>  -254,6 +254,8  <at>  <at> 
> 	USE_FEATURE_HTTPD_BASIC_AUTH(char *remoteuser;)
> 	USE_FEATURE_HTTPD_CGI(char *referer;)
> 	USE_FEATURE_HTTPD_CGI(char *user_agent;)
>+	USE_FEATURE_HTTPD_CGI(char *http_accept;)
>+	USE_FEATURE_HTTPD_CGI(char *http_accept_language;)
> 
> 	off_t file_size;        /* -1 - unknown */
> #if ENABLE_FEATURE_HTTPD_RANGES
> <at>  <at>  -299,6 +301,8  <at>  <at> 
> #define remoteuser        (G.remoteuser       )
> #define referer           (G.referer          )
> #define user_agent        (G.user_agent       )
>+#define http_accept       (G.http_accept      )
>+#define http_accept_language (G.http_accept_language)
> #define file_size         (G.file_size        )
> #if ENABLE_FEATURE_HTTPD_RANGES
> #define range_start       (G.range_start      )
> <at>  <at>  -1384,6 +1388,10  <at>  <at> 
> 		}
> 	}
> 	setenv1("HTTP_USER_AGENT", user_agent);
>+	if (http_accept)
>+		setenv1("HTTP_ACCEPT", http_accept);
>+	if (http_accept_language)
>+		setenv1("HTTP_ACCEPT_LANGUAGE", http_accept_language);
> 	if (post_len)
> 		putenv(xasprintf("CONTENT_LENGTH=%d", post_len));
> 	if (cookie)
> <at>  <at>  -2005,6 +2013,10  <at>  <at> 
> 				referer = xstrdup(skip_whitespace(iobuf + sizeof("Referer:")-1));
> 			} else if (STRNCASECMP(iobuf, "User-Agent:") == 0) {
> 				user_agent = xstrdup(skip_whitespace(iobuf + sizeof("User-Agent:")-1));
>+			} else if (STRNCASECMP(iobuf, "Accept:") == 0) {
>+				http_accept = xstrdup(skip_whitespace(iobuf + sizeof("Accept:")-1));
>+			} else if (STRNCASECMP(iobuf, "Accept-Language:") == 0) {
>+				http_accept_language = xstrdup(skip_whitespace(iobuf + sizeof("Accept-Language:")-1));
> 			}
> #endif
> #if ENABLE_FEATURE_HTTPD_BASIC_AUTH
>

>_______________________________________________
>busybox mailing list
>busybox <at> busybox.net
>http://busybox.net/cgi-bin/mailman/listinfo/busybox
Cristian Ionescu-Idbohrn | 1 Sep 18:56 2008
Picon

Re: [patch] fix for autodocifier.pl uninitialized value in join

On Mon, 1 Sep 2008, walter harms wrote:

> please provide the test.

Please note: what I'm talking about is _not_ a test, it's not a functional
change, but just a colorization problem caused by a possible bug in
emacs21 perl-mode (see attached picture).

> I would like to see a small comment why you did escape the double
> quotes.

I changed my mind about that, as emacs21 cperl-mode does not suffer of
this and neither does emacs22.  Still, this is a short example of what I
think is a more obvious/better special character protection:

-               $text =~ s/SKIP_\w+\(.*?"\s*\)//sxg;
+               $text =~ s/SKIP_\w+\(.*?\"\s*\)//sxg;
                                        ^^

Cheers,

--

-- 
Cristian
_______________________________________________
busybox mailing list
busybox <at> busybox.net
http://busybox.net/cgi-bin/mailman/listinfo/busybox
Bernhard Fischer | 1 Sep 19:22 2008
Picon

Re: [patch] fix for autodocifier.pl uninitialized value in join

On Mon, Sep 01, 2008 at 06:56:21PM +0200, Cristian Ionescu-Idbohrn wrote:

>I changed my mind about that, as emacs21 cperl-mode does not suffer of
>this and neither does emacs22.  Still, this is a short example of what I
>think is a more obvious/better special character protection:
>
>-               $text =~ s/SKIP_\w+\(.*?"\s*\)//sxg;
>+               $text =~ s/SKIP_\w+\(.*?\"\s*\)//sxg;

my copy of vi(m) gets it right, so if current emacs gets it right too,
i suggest to leave it alone..

If you're really bored then let me point you to ./TODO :P
I suggest ar write-support ff.

have fun,
Bernhard
Vladimir Dronnikov | 1 Sep 19:55 2008
Picon

Ответ: [patch] fix for autodocifier.pl uninitialized value in join

And -D option for install, please :^) I can't compile buildroot using
BB coreutils...

--
Vladimir

2008/9/1, Bernhard Fischer <rep.dot.nop <at> gmail.com>:
> On Mon, Sep 01, 2008 at 06:56:21PM +0200, Cristian Ionescu-Idbohrn wrote:
>
>>I changed my mind about that, as emacs21 cperl-mode does not suffer of
>>this and neither does emacs22.  Still, this is a short example of what I
>>think is a more obvious/better special character protection:
>>
>>-               $text =~ s/SKIP_\w+\(.*?"\s*\)//sxg;
>>+               $text =~ s/SKIP_\w+\(.*?\"\s*\)//sxg;
>
> my copy of vi(m) gets it right, so if current emacs gets it right too,
> i suggest to leave it alone..
>
> If you're really bored then let me point you to ./TODO :P
> I suggest ar write-support ff.
>
> have fun,
> Bernhard
> _______________________________________________
> busybox mailing list
> busybox <at> busybox.net
> http://busybox.net/cgi-bin/mailman/listinfo/busybox
>

Gmane