Lighttpd

2006-07-20

15:28 Bug #759: mmap fails on NSLU2 with offset >= 512k

I just add a strace file. Hope it's helpful.
-- xummxxu Anonymous

11:34 Bug #759: mmap fails on NSLU2 with offset >= 512k

Yes the:
server.network-backend = "write"
does workaround the issue, thank you!
I found the section in the wiki ... Anonymous

07:20 Bug #759 (Assigned): mmap fails on NSLU2 with offset >= 512k

Can you attach a strace (see the wiki for how to report a bug) to this bug-report ?
Does using
server.network-... jan

12:39 Revision 1213 (svn): the .h is already included, the backend needs some more time

jan

12:13 Revision 1212 (svn): moved the http-backend stuff to mod_proxy_backend_http

* src/mod_proxy_core_rewrites.c
- free all memory before shutdown
* the rest
- moved the HTTP-specific code to mo... jan

2006-07-19

22:09 Revision 1211 (svn): added #include "config.h" to fixed nasty crash in mod_cgi

jan

22:06 Bug #759 (Invalid): mmap fails on NSLU2 with offset >= 512k

I am using lighttpd 1.4.11 w/ php 5.1.4 and fcgi on a Linksys NSLU2 running Unslung 6.8. Whenever a file is download... Anonymous

21:37 Revision 1210 (svn): fixed non-critical mem-leaks

jan

14:32 Revision 1209 (svn): enabled the config-settings for the load-balancers

jan

14:31 Revision 1208 (svn): moved the status-counters out from base.h and mod_fastcgi.c

jan

11:21 Revision 1207 (svn): if we get a close instead of chunk-header, request more content

jan

11:04 Revision 1206 (svn): add -std=gnu99 to the compile options if we are with GCC

jan

11:03 Revision 1205 (svn): support host:port notation (IPv4 and IPv6)

jan

11:02 Revision 1204 (svn): enable IPv6 if we have inet_ntop

jan

10:54 Revision 1203 (svn): use ->sock instead of ->fd to make sure we didn't missed a conversion to iosocket

jan

06:34 Bug #499: FastCGI stderr output doesn't go to error log

Has this bug been fixed since 1.4.9 ?
I'm using 1.4.11_1 (ssl) on FreeBSD 6.1, the bug seems to be persist here...... Anonymous

2006-07-18

22:40 Bug #753: Repeated headers produce 400 Bad Request

More bugs in If-None-Match processing. Looks like lighttpd reads If-None-Match as a string header and not a list head... Anonymous

14:07 Bug #393 (Fixed): Invilid RTSP OPTIONS request cause lighttpd crash.

we have a testcase for this request since a while and it looks like this bug got fixed.
jan

14:05 Feature #433 (Assigned): please add support for X-LIGHTTPD-send-file to mod_proxy as well

This will be added to mod_proxy_core which currently grows in SVN.
jan

14:03 Bug #396 (Fixed): Can't send Status Header with PHP

<?php
header("Status: 301");
header("Location: http://www.example.org/");
?>
works as expected.
jan

14:00 Bug #754 (Fixed): after update from svn fastcgi can't handle responses >8k???

This should be fixed since r1182.
jan

13:25 Revision 1202 (svn): only try to handle the request if we really have backends available

jan

12:59 Revision 1201 (svn): added rewrite support

jan

12:55 Revision 1200 (svn): removed debug and added some assertions

jan

09:43 Revision 1199 (svn): added function check for strptime and added support for --with_xattr

jan

09:42 Revision 1198 (svn): added mod_proxy_core

jan

09:41 Revision 1197 (svn): ported to iosocket

jan

09:41 Revision 1196 (svn): include log.c for SEGFAULT()

jan

09:40 Revision 1195 (svn): ported to iosocket

jan

09:39 Revision 1194 (svn): removed WP() as it is gone now

jan

09:39 Revision 1193 (svn): don't care about the errorlog

jan

09:38 Revision 1192 (svn): use closesocket() to close a socket() on unix too

jan

09:37 Revision 1191 (svn): ported to iosocket and new fdevent functions

- added 502 and 504 as known response codes
- removed some debug code jan

09:36 Revision 1190 (svn): introduce BUF_STR() to access the NUL-terminated string-part of the buffer

moved SEGFAULT() to log.h jan

09:35 Revision 1189 (svn): added chunkqueue_print() and chunkqueue_remove_empty_last_chunk()

jan

09:34 Revision 1188 (svn): moved the mainloop() out of the main() function

use iosocket and the new fdevent interface jan

09:34 Revision 1187 (svn): replace strcmp() by our own buffer_compare_... functions

jan

09:32 Revision 1186 (svn): fixed mem-leaks

use TRACE() from log.c instead of fprintf() jan

09:31 Revision 1185 (svn): moved errorlog pointers to log.h

moved SSL from connection to iosocket.h jan

09:30 Revision 1184 (svn): use the http_response parser

use network_backend_write and _read jan

09:29 Revision 1183 (svn): ported to iosocket

jan

09:29 Revision 1182 (svn): ported iosocket

* network_linux_sendfile.c
- send the whole file via sendfile() and not just the first chunk
- proper return code... jan

09:27 Revision 1181 (svn): ported fdevent to iosocket

- instead if get_ndx, get_next_fd ... build a full revent-array and return it
this is cleaner and removes a race-co... jan

09:25 Revision 1180 (svn): added a wrapper around 'int fd; int fde_ndx; ...'

jan

09:25 Revision 1179 (svn): get rid of 'server *' in log.[ch]

- added TRACE() and ERROR() jan

09:22 Revision 1178 (svn): disabled the mod_proxy tests for a while

jan

09:20 Revision 1177 (svn): 200 ? This is a broken response-code, expect 502 (Bad Gateway) instead

jan

09:19 Revision 1176 (svn): enforce ETag and Last-Modified headers

jan

09:19 Revision 1175 (svn): use php-5.1.4 for testing

jan

09:18 Revision 1174 (svn): added a external TRACEME env to start a test(suite) with strace or valgrind

jan

2006-07-17

21:36 Bug #758 (Fixed): memory fragmentation leads to high memory usage after peaks

*Summary: lighttpd of course had some memory leaks (and perhaps even has today), but this bug is not about these prob... Anonymous

13:42 Bug #757: internal server error gives 200 OK sometimes

after some time running (10 minutes), the problem came back, it randomly gives 200/500
-- jimmy Anonymous

13:17 Bug #757: internal server error gives 200 OK sometimes

after enabling it back again, the problem was not back, which is odd...
-- jimmy Anonymous

13:15 Bug #757: internal server error gives 200 OK sometimes

I just notices mod_rewrite was enabled, after disabling it, the problem was away
-- jimmy Anonymous

13:12 Bug #757 (Duplicate): internal server error gives 200 OK sometimes

Config: OSX/freebsd-kqueue/cgi/setenv
Perl: v5.8.6
If the CGI script fails, it somitimes gives a 200 with empty p... Anonymous

13:30 Bug #756: transfer-encoding chunked broken (OSX)

after a few minutes, i also get this in the log:... Anonymous

13:26 Bug #756: transfer-encoding chunked broken (OSX)

The problem persists after several restarts and config changes.
My config:... Anonymous

12:56 Bug #756 (Fixed): transfer-encoding chunked broken (OSX)

Config: OSX/freebsd-kqueue/CGI/setenv
Test script:... Anonymous

2006-07-16

22:58 Bug #755 (Fixed): 206 partial response fails over fastcgi connection

If a fastcgi app tries to output some lengthy response with 206 partial response status then lighty will eat just the... Anonymous

08:51 Bug #754 (Fixed): after update from svn fastcgi can't handle responses >8k???

Just updated lighty sources from svn, and each response from fastcgi
backend over 8kbytes fails. I made simple php ... chernousov