To Apache Flood development team:
As part of our work on the Flood code (in our usage of it as par of an
enterprise product), I have been instructed to contribute back our changes
on a daily basis so we can both contribute them back to the product and
get back inportant feedback on these changes.
2.0 already has this fixed
Index: http_core.c
===
RCS file: /home/cvs/apache-1.3/src/main/http_core.c,v
retrieving revision 1.325
diff -u -r1.325 http_core.c
--- http_core.c 7 Jul 2003 13:02:28 - 1.325
+++ http_core.c 14
Stas Bekman wrote:
Also we need to have the vhost info in the score, otherwise request URI
info is not quite useful on the system with several vhosts, since you
can't tell which vhost it was invoked from.
hmmm? AFAIK vhost is working fine - see http://apache.org/server-status
Greg
[Tue Oct 14 08:24:57 2003] [error] [client 213.243.186.233] Client sent
malformed Host header
[Tue Oct 14 08:24:57 2003] [notice] child pid 29013 exit signal
Segmentation fault (11)
Every time I get a malformed header, I get a segfault. Any ideas? I
have UseCannonicalName On, and I alwasy use
Brian Akins wrote:
[Tue Oct 14 08:24:57 2003] [error] [client 213.243.186.233] Client sent
malformed Host header
[Tue Oct 14 08:24:57 2003] [notice] child pid 29013 exit signal
Segmentation fault (11)
Every time I get a malformed header, I get a segfault. Any ideas?
next step is to get backtrace
On Tue, 2003-10-14 at 09:56, Jeff Trawick wrote:
next step is to get backtrace from coredump... do you know how to
enable core dumps and get backtraces? (if not, what OS is this and what
version of the web server?)
Yes. But I can never get them when I let httpd change id (ie, from root
Brian Akins wrote:
On Tue, 2003-10-14 at 09:56, Jeff Trawick wrote:
next step is to get backtrace from coredump... do you know how to
enable core dumps and get backtraces? (if not, what OS is this and what
version of the web server?)
Yes. But I can never get them when I let httpd change
Thanks.
How can I simulated a malformed host header? I set all kinds of
garbage, but it never complains.
--
Brian Akins [EMAIL PROTECTED]
CNN Internet Technologies
On Tue, 2003-10-14 at 10:28, Brian Akins wrote:
Thanks.
How can I simulate a malformed host header? I set all kinds of
garbage, but it never complains.
I figure out if I send a / in the Host header this will be triggered.
It seems to be in mod_include can anyone else reproduce this?
--
On Tue, 2003-10-14 at 12:16, Brian Akins wrote:
On Tue, 2003-10-14 at 10:28, Brian Akins wrote:
Thanks.
How can I simulate a malformed host header? I set all kinds of
garbage, but it never complains.
I figure out if I send a / in the Host header this will be triggered.
It seems to
On Tue, 2003-10-14 at 13:39, Brian Akins wrote:
I figure out if I send a / in the Host header this will be triggered.
It seems to be in mod_include can anyone else reproduce this?
Someone else confirmed on a couple of installs. This seems to only
happen when AddOutputFilterByType
WinXP, Apache 1.3.28 crashed when browse I configure the
proxy inside the SSL VirtualHost:443 config directive.
Similar crash happen when I include proxy+mod_gzip
or mod_php inside SSL config too.
The crash seems to happen regularly with framed pages.
Is this know issue? Am I required to
1.3 ap_die() is simpler, so use that as an example when referring to code.
case 1:
ErrorDocument 413 /index.html
module returns 413 from handler and doesn't set r-status to 413
GOOD: /index.html is used for the error document
case 2:
ErrorDocument 413 /index.html
module returns 413 from handler
[EMAIL PROTECTED] wrote:
Stas Bekman wrote:
Also we need to have the vhost info in the score, otherwise request
URI info is not quite useful on the system with several vhosts, since
you can't tell which vhost it was invoked from.
hmmm? AFAIK vhost is working fine - see
Geoffrey Young wrote:
ap_die() is definitely confused about whether or not it is a recursive
error. Any time r-status is set, it thinks it is a recursive error.
On a side note, is it bad for modules to set r-status?
it was because of the former that I thought it was bad to do the latter.
15 matches
Mail list logo