headers
P.V.Anthony | 1 Apr 2006 06:28
Picon
Picon

php-fastcgi error

Hi,

I have a problem with my php-fastcgi.

Please note that there is nothing wrong with lighttpd.

I have compiled php with fastcgi on server 01 and it is worknig fine.
I tried the same thing on server 02 and I am getting an error. The strange part 
is that the error only happens when I login as a normal user. When I login as 
root, there is no error.

The error is as follows.
# php-fastcgi -v
PHP Warning:  Unknown(): Unable to load dynamic library 
'/usr/lib/20020429/pgsql.so' - /usr/lib/20020429/pgsql.so: undefined symbol: 
php_regcomp in Unknown on line 0
PHP 4.3.11 (cgi-fcgi) (built: Apr  1 2006 03:00:55)
Copyright (c) 1997-2004 The PHP Group
Zend Engine v1.3.0, Copyright (c) 1998-2004 Zend Technologies

I do not get this error on server 01 when I run as a normal user.

Where have I gone wrong? I really do not have a clue.

My permissions on pgsql.so has others set to read.

Please do advise.

P.V.Anthony
(Continue reading)

Permalink | Reply | 1 comment |
headers
Charles M. Gerungan | 1 Apr 2006 16:46

vqadmin works on .8/.9, not on .11

Has anything changed wrt cgi handling from .8 to .11? I just tried  
upgrading my lighty but it breaks vqadmin (a cgi app written in C-- 
see http://inter7.com/?page=vqadmin). It works with .9 too and  
unfortunately I wasn't able to get .10 finish `make`.

The error I'm getting from the app is: Error No Query String.

root <at> coruscant /var/svc.d/www-example # ./run
2006-04-01 16:39:29: (log.c.75) server started
2006-04-01 16:39:44: (request.c.293) fd: 7 request-len: 876
GET /cgi-bin/vqadmin/vqadmin.cgi HTTP/1.1
Host: secure.example.com
User-Agent: Mozilla/5.0 (Macintosh; U; PPC Mac OS X Mach-O; en-US; rv: 
1.8) Gecko/20051111 Firefox/1.5
Accept: text/xml,application/xml,application/xhtml+xml,text/ 
html;q=0.9,text/plain;q=0.8,image/png,*/*;q=0.5
Accept-Language:  
en,com;q=0.9,ja;q=0.9,fr;q=0.8,de;q=0.7,es;q=0.7,it;q=0.6,sv;q=0.5,nb;q= 
0.5,da;q=0.4,fi;q=0.3,pt;q=0.3,zh-Hans;q=0.2,zh-Hant;q=0.1,ko;q=0.1
Accept-Encoding: gzip,deflate
Accept-Charset: ISO-8859-1,utf-8;q=0.7,*;q=0.7
Keep-Alive: 300
Connection: keep-alive
Cookie: TRACKID=5c851d10a4aa2034da349cbe1cc2b35e
Authorization: Digest username="username", realm="realm",  
nonce="wKsGIl0QBAA=2be3bc130eb75eae3fd894d4e63a980f350258a6", uri="/ 
cgi-bin/vqadmin/vqadmin.cgi", algorithm=MD5,  
response="b328247379883600dce5a15d4a83d933", qop=auth, nc=0000001a,  
cnonce="55d98796b776e203"
(Continue reading)

Permalink | Reply | 2 comments |
headers
P.V.Anthony | 3 Apr 2006 06:00
Picon
Picon

Re: php-fastcgi error

P.V.Anthony wrote:
> Hi,
> 
> I have a problem with my php-fastcgi.
> 
> Please note that there is nothing wrong with lighttpd.
> 
> I have compiled php with fastcgi on server 01 and it is worknig fine.
> I tried the same thing on server 02 and I am getting an error. The 
> strange part is that the error only happens when I login as a normal 
> user. When I login as root, there is no error.
> 
> The error is as follows.
> # php-fastcgi -v
> PHP Warning:  Unknown(): Unable to load dynamic library 
> '/usr/lib/20020429/pgsql.so' - /usr/lib/20020429/pgsql.so: undefined 
> symbol: php_regcomp in Unknown on line 0
> PHP 4.3.11 (cgi-fcgi) (built: Apr  1 2006 03:00:55)
> Copyright (c) 1997-2004 The PHP Group
> Zend Engine v1.3.0, Copyright (c) 1998-2004 Zend Technologies
> 
> I do not get this error on server 01 when I run as a normal user.
> 
> Where have I gone wrong? I really do not have a clue.
> 
> My permissions on pgsql.so has others set to read.

I think I got it solved.

I changed the pgsql.so with newly compiled one. Now it seems to work.
(Continue reading)

Permalink | Reply | 0 comments |
headers
Xuefer | 4 Apr 2006 07:04
Picon

one more php opcode cacher for your choice

http://blog.lighttpd.net/articles/2006/04/04/one-more-opcache-for-php-preview
we might make it available online in a separated svn/trac, by now u
can download and try the preview package.
any comment is welcomed.
Permalink | Reply | 0 comments |
headers
Mark van Eijk | 4 Apr 2006 13:07

Re: mod_redirect and url encoding

On Tue, Mar 21, 2006 at 11:15:14AM +0100, Spill Mailing Lists wrote:
> Mark van Eijk wrote:
> > Or do you mean that lighttpd first decodes the URL-encoding and then
> > redirects?
> 
> Location:
> http://www2.mydomain.com/myscript.php?path=domain.com/&ref=http://myreferer/
> 
> As you can see, the 'Location' URL is not properly encoded. I think this
> should be the resulting URL:
> 
> http://www2.mydomain.com/myscript.php?path=domain.com/&ref=http%3A%2F%2Fmyreferer%2F
> 
> So I think the decoding pass is made by Lighttpd, but mod_redirect and
> mod_rewrite should make sure to make the encoding pass when placing
> parameters in a URL.

Sorry for the late response, but I think you are right.
So it should encode in case the URL is placed in one of the parameters,
but it shouldn't if it redirects to another URL without parameters?

Mark
Permalink | Reply | 0 comments |
headers
Jan Kneschke | 5 Apr 2006 16:23
Picon
Gravatar

SSL uploads fixed


Dear lighties,

perhaps you have experiences this bug too. Sometimes the SSL uploads
(POST requests) where stalling without any obvious reasons.

This bug was reported as http://trac.lighttpd.net/trac/ticket/607 and
had a wonderful, reproducable test-case which made it easy to track it
down.

The fix itself is a one-line

http://trac.lighttpd.net/trac/changeset/1095

In case you can't wait for 1.4.12 and are running into this problem,
apply this fix. It should apply to basicly all releases as this part is
the same since ... a long time.

Thanks go to jtate <at> rpath.com for the testcase.

    Jan

--

-- 
Jan Kneschke                                     http://jan.kneschke.de/
Perhaps you want to say 'thank you, jan':    http://jk.123.org/wishlist/
Permalink | Reply | 0 comments |
headers
James Nachlin | 5 Apr 2006 17:26

environment for include_shell

lighttpd is failing on startup when I use include_shell to include a 
shell script that checks for the presence of a maintenance file and adds 
the line
url.rewrite-once = ( "^.*$" => "/system/maintenance.html" )
if it exists.

The error is:

Starting lighttpd: env SHELL is required2006-05-07 02:00:15: 
(configfile.c.882) opening 
/u/apps/motionbox/current/config/os/web/lighttpd/maintenance.sh failed: 
Success

This happens starting or restarting lighty with the init script. 
Putting SHELL="/bin/bash" in that script does not help.  Is there a way 
to get this to work?

Thanks,
Jim
Permalink | Reply | 5 comments |
headers
Krzysztof Krzyzaniak | 5 Apr 2006 17:33
Gravatar

Re: environment for include_shell

James Nachlin wrote:
> lighttpd is failing on startup when I use include_shell to include a 
> shell script that checks for the presence of a maintenance file and adds 
> the line
> url.rewrite-once = ( "^.*$" => "/system/maintenance.html" )
> if it exists.
> 
> The error is:
> 
> Starting lighttpd: env SHELL is required2006-05-07 02:00:15: 
> (configfile.c.882) opening 
> /u/apps/motionbox/current/config/os/web/lighttpd/maintenance.sh failed: 
> Success
> 
> This happens starting or restarting lighty with the init script. 
> Putting SHELL="/bin/bash" in that script does not help.  Is there a way 
> to get this to work?

Take look at patch in debian:

diff -urNad lighttpd-1.4.8~/src/proc_open.c lighttpd-1.4.8/src/proc_open.c
--- lighttpd-1.4.8~/src/proc_open.c	2005-08-11 00:26:39.000000000 +0200
+++ lighttpd-1.4.8/src/proc_open.c	2005-11-26 12:12:02.000000000 +0100
 <at>  <at>  -223,12 +223,13  <at>  <at> 
 /* {{{ proc_open */
 int proc_open(proc_handler_t *proc, const char *command) {
 	pid_t child;
-	const char *shell;
+	const char *shell = "/bin/sh";
(Continue reading)

Permalink | Reply | 0 comments |
headers
Pete Lavetsky | 5 Apr 2006 23:46
Favicon

Ajax Requests and mod_auth Issues

Hi everyone,

I believe I've come across an issue using mod_auth and Ajax requests ... 
here's my relative mod_auth config:

auth.backend                            = "htdigest"
auth.backend.htdigest.userfile          = "lighttpd.user.htdigest"

auth.require = ( "/views/" =>
(
   "method"  => "digest",
   "realm"   => test",
   "require" => "user=test"
 )
)

server.modules = (
    "mod_auth",
    "mod_rewrite",
    "mod_redirect",
    "mod_status",
    "mod_proxy",
    "mod_accesslog"
)

Browsing to a site like http://example.com/views/test brings up the 
authentication box and I can authenticate correctly ... everything is 
fine until I browse to a page that makes an Ajax request ... the request 
looks something like this:
(Continue reading)

Permalink | Reply | 0 comments |
headers
Mark van Eijk | 6 Apr 2006 01:59

Re: vqadmin works on .8/.9, not on .11

On Sat, Apr 01, 2006 at 04:46:19PM +0200, Charles M. Gerungan wrote:
> Has anything changed wrt cgi handling from .8 to .11? I just tried  
> upgrading my lighty but it breaks vqadmin (a cgi app written in C-- 
> see http://inter7.com/?page=vqadmin). It works with .9 too and  
> unfortunately I wasn't able to get .10 finish `make`.
> 
> The error I'm getting from the app is: Error No Query String.

What you are seeing is the result of a behavioural change in lighttpd
since changeset 1006 [1]. Not quite sure which releases it concearns,
but .11 is certainly affected.

What happened is that lighttpd only adds the QUERY_STRING environment
variable if the query information is non-empty.

According to the CGI-specification [2] this is appropriate behaviour:
"QUERY_STRING

The information which follows the ? in the URL which referenced this
script. This is the query information. It should not be decoded in any
fashion. This variable should always be set when there is query
information, regardless of command line decoding."

So it should always be set when there is query information. When there
is no query information, servers can choose not to include the
QUERY_STRING environment variable.

One could see this as an unwanted behavioural change of lighttpd, but
essentially your vqadmin CGI-app is broken. It should make sure it still
functions if no query information is available (which is the case
(Continue reading)

Permalink | Reply | 1 comment |

Gmane