latest Apache2 version work together?
If so, and I believe they do, then that should be the only recommended
version(s) at this time. What use is there documenting what older
combos work together (especially given the few security fixes lately).
-aaron
p.s. why is this cc: so huge!?
--
PHP Developme
ms with large input bodies [IIRC] at
the moment.
2) Any 3rd party library that is linked with PHP and/or Apache2 must
be threadsafe (if Apache2 is configured to run multithreaded).
-aaron
--
PHP Development Mailing List <http://www.php.net/>
To unsubscribe, visit: http://www.php.net/unsub.php
I'd like to work on the documentation, cleaning up some of the comments and updating a
few of the pages. So I guess that means phpdoc access :-) I might also get involved
with the translations.
Who sent me? Well... I just got back from PHPCon where I met Rasmus, yay! Jim Winstead
suggested th
Apologies if you receive this twice. My posts dont appear to be going
through...
Cheers.
Aa.
-- Forwarded message --
Date: Mon, 21 Oct 2002 11:57:09 -0700 (PDT)
From: Aaron Gowatch <[EMAIL PROTECTED]>
To: "[EMAIL PROTECTED]" <[EMAIL PROTECTED]>
Sub
further it'd be much appreciated.
Thanks,
Aa.
On Mon, 14 Oct 2002, Andi Gutmans wrote:
> What we'd really need is a short 10-20 line reproducing script.
>
> Andi
>
> At 11:20 AM 10/14/2002 -0700, Aaron Gowatch wrote:
> >Heres a backtrace from one of my SIGSEGVs. Th
I'll leave my gdb attached in case more information is needed.
Thanks in advance,
Aa.
(Apologies if this is received twice)
On Sun, 13 Oct 2002, Aaron Gowatch wrote:
> My apache children have been segfaulting for a few weeks now. My attempts
> to debug them leads me to believe tha
My apache children have been segfaulting for a few weeks now. My attempts
to debug them leads me to believe that its happening in php somewhere.
The SIGSEGV actually occurs in chunk_alloc, after a malloc and several
calls to execute. (I'll fwd a backtrace per the bugs docs as soon as my
gdb
is released. Until such a change is made there is really no need to
> restrict the functionality of Apache 2 IMHO.
+1
-aaron
--
PHP Development Mailing List <http://www.php.net/>
To unsubscribe, visit: http://www.php.net/unsub.php
es in the same
> manner PHP s Apache 1.X code does.
>
> If there are no objections to the patch, I would like to commit to CVS, so
> that the PHP 4.3.0 (whenever it comes out) support of Apache 2.X is slightly
> more complete.
Awesome! Thanks Ilia!
-aaron
--
PH
other 3rd party modules.
-aaron
p.s. I'm using the threaded "worker" MPM under Linux 2.2.
--
PHP Development Mailing List <http://www.php.net/>
To unsubscribe, visit: http://www.php.net/unsub.php
handler to be called, which means the process dies. Quite
frustrating.
-aaron
On Fri, Aug 02, 2002 at 02:54:17PM -0700, Rasmus Lerdorf wrote:
> I'm surprised it worked at all. Start with CVS versions of both and go
> from there.
>
> On Fri, 2 Aug 2002, Cody Sherr wrote:
>
>
You guys aren't the only ones who have to deal with the reprocussions
of changing the register_globals default...
(I've already sent the reporter in the right direction and closed
this bug, but something tells me this won't be the last time we hear
about this.)
-aaron
- Fo
willing to code it up. What is the
current feeling/policy on extending the SAPI?
-aaron
--
PHP Development Mailing List <http://www.php.net/>
To unsubscribe, visit: http://www.php.net/unsub.php
On Mon, Jun 10, 2002 at 09:29:58AM -0700, Aaron Bannert wrote:
>- creation (called once per server lifetime)
>- malloc (called many times per request)
>- free (called many times per request)
>- end-of-request (called many times per request)
(Whoops, that shou
r all our bases? For example, when using pools, the
free() call would do nothing, and the end-of-request call would simply
call apr_pool_clear(). Note that this only applies to dynamically
allocated memory required for the lifetime of a request. For memory
with longer lifetimes we could make the crea
he memory allocation for an application reaches a
"steady state". In PHP this could be accomplished by simply using
the per-request pool that is already available when the internal PHP
functions are called from httpd. At the end of a request this pool is
"cleared", and then a
ient
> in lifetime so apr_brigade_write() will be forced to copy it, but that's
> better than sending out 11 byte packets by a long shot.
Big +1
-aaron
--
PHP Development Mailing List <http://www.php.net/>
To unsubscribe, visit: http://www.php.net/unsub.php
t note is you really should use the newest version of AB (and report
the version you're using for these tests). I made some significant
changes to AB recently that dramatically improved its ability to
multiplex multiple connections.
Thanks for the reports,
-aaron
--
PHP Development
d that this has been added.
> (Accepting a buffer instead of a file handle.)
Oh? That would be awesome! Zeev, care to elaborate or point us at
some code?
-aaron
--
PHP Development Mailing List <http://www.php.net/>
To unsubscribe, visit: http://www.php.net/unsub.php
till
performs. Another thing that is gained in the new Apache 2.0 multithreaded
architecture is memory conservation -- I've run a 500-thread system on
under 20MB resident memory. Under Apache 1.3 that would require
hundreds of MB of memory.
-aaron
--
PHP Development Mailing List <http://www.php.net/>
To unsubscribe, visit: http://www.php.net/unsub.php
On Tue, Jun 04, 2002 at 10:44:17PM -0500, Andrei Zmievski wrote:
> In case you've missed it, I have been putting up unofficial patches for
> PHP on my website. The latest one changes some operators. See the
> website for more details.
Is there a reason why you don't just post
few weeks. The apache2filter still has
yet to prove itself in production, but more of a concern at this point
are 3rd party libraries and other extensions that may not be entirely
threadsafe or reentrant. Let us know how it goes.
-aaron
--
PHP Development Mailing List <http://www.php.net/>
would I be able to do, apart from coding C, to make it easier for
> you guys to fix and track down the problems.
Thanks for the offer. Just use it however you see fit an report any problems
(or successes). :)
-aaron
--
PHP Development Mailing List <http://www.php.net/>
To unsubscribe, visit: http://www.php.net/unsub.php
oon as that becomes
available the apache2filter will take advantage of it.
-aaron
--
PHP Development Mailing List <http://www.php.net/>
To unsubscribe, visit: http://www.php.net/unsub.php
ave put some great effort into
making it usable, and I have been using it on my site for the last few
weeks with no problems. It will become stable when it has been used in
production, not when it's been sitting stale in CVS for 6 months.
-aaron
--
PHP Development Mailing List <http://ww
I haven't read this entire thread nor heard all sides of
the argument, at this point I'm -0.9 to the whole idea. Unless
there is a *very* good reason for including 3rd party libraries
in PHP, I beleve that should be strongly avoided.
-aaron
* Email me privately if you don't know wh
g
for a committment on our part to some guarantee of code quality, I
suggest you review the license. I don't mean this to be discouraging,
since I do believe that in general projects like PHP produce much higher
quality code than a proprietary counterpart, and I think you may find
success w
ologise if this was discussed before, but why do we want to default
to having AcceptPathInfo enabled for a PHP resource? This seems like we
are taking the conservative approach in having this disabled unless the
admin knows what they're doing.
-aaron
--
PHP Development Mailing List <htt
re of Apache 2 and how
that will interact with 3rd party libraries that may not be reentrant
or threadsafe (as Rasmus points out).
If you encounter any bugs, please report them. More importantly, if
you have success please let us know. :)
-aaron (one of the apache2filter maintainers)
--
PHP Developm
Actually this isn't as bad as it sounds. I've been doing some of the
work necessary to make perchild portable lately. And as for making it
threadsafe/reentrant, I think those kinds of problems are going to show
up with the worker MPM as well (which seems to be working fairly well
with PHP
On Thu, May 09, 2002 at 06:53:39PM -0700, Aaron Bannert wrote:
> On Fri, May 10, 2002 at 10:11:14AM +0900, Masaki Fujimoto wrote:
> > This should be because of my patch, which committed by Rui 2 days ago.
> > Sorry for sparing your time...
> >
> > Anyway, attached
;zend_istdiostream.h"' in zend_multibyte.c. If this patch
> works, would someone commit this to HEAD please ? And if not, would you
> please show me your configure line ?
Cliff pointed out this fix to me, I tested and committed it.
-aaron
--
PHP Development Mailing List <http
ll to apr_pool_userdata_get() with a different pointer than
the original setn() call finds no entry and return NULL.
4) we don't initialize because we never find the entry set to (void*)1.
Thanks for pointing this out and thanks for the patch! :)
-aaron
--
PHP Development Mailing List <http://www.php.net/>
To unsubscribe, visit: http://www.php.net/unsub.php
o both apache and
php?
Also, for anyone who has had problems getting this to work, please
do the same.
The two prevailing theories right now are:
- php is tickling a broken pthread implementation.
- there are some php/apache configure parameters that when combined can
cause this.
-aaron
-
roduceable by me) that we have already
spent much time on and hope to have solved soon.
Brian,
I suggest you join the volunteer effort by providing detailed bug
reports or even patches to fix problems that you come across. Only
then can you ensure that your issues are properly addressed.
-a
e function, but there is
no way that I know of for sending a meaningful message back to the
console/logs. Maybe we should just write a better PHPINIDir error message?
-aaron
--
PHP Development Mailing List <http://www.php.net/>
To unsubscribe, visit: http://www.php.net/unsub.php
dule startup time (for us this is
post_config) we don't know which vhost we'd like to serve.
I think the problem will become more apparent as the perchild MPM
becomes more mature, so for now I just kept it as simple as possible
and disallowed attempts at per-vhost configs.
-aaron
--
PH
On Fri, Apr 19, 2002 at 06:17:16PM +0200, Marcus Börger wrote:
> What about leaving -O2 but adding -finline-functions
You can add all of these things to CFLAGS while running configure,
I believe.
I normally run with:
CFLAGS="-Wall -g" ./configure ...
-aaron
--
PHP Development
> default?
Is making these kinds of optimizations by default such a good idea? I've
seen many a broken gcc install produce incorrect binaries with the larger
-O optimizations turned on (unfortunately).
-aaron
--
PHP Development Mailing List <http://www.php.net/>
To unsubscribe, visit: http://www.php.net/unsub.php
I'm not sure I understand this post. I didn't add these files.
What do you mean by "harmut"?
-aaron
On Fri, Apr 19, 2002 at 03:26:28AM +0100, James Cox wrote:
> this was also harmut with his .cvsignore commits. (i believe)
>
> -Original Message-
> Fr
d object and the last few members of that
struct aren't expliclty initialized.
-aaron
--
PHP Development Mailing List <http://www.php.net/>
To unsubscribe, visit: http://www.php.net/unsub.php
tp://bugs.php.net/bug.php?id=16622
Thanks for reporting this, I'll try to reproduce it this evening.
-aaron
--
PHP Development Mailing List <http://www.php.net/>
To unsubscribe, visit: http://www.php.net/unsub.php
ed. (Thanks for Cliff for pointing this message out to me.)
-aaron
Index: sapi_apache2.c
===
RCS file: /repository/php4/sapi/apache2filter/sapi_apache2.c,v
retrieving revision 1.61.2.9
diff -u -u -r1.61.2.9 sapi_apache2.c
--- sapi_a
->server,
>"\"Set%sFilter PHP\" already configured for %s",
>output ? "Output" : "Input", r->uri);
> return;
> }
> f = f->next;
> }
Naw, we don't want to add the PHP filter twice.
-aaron
--
PHP Development Mailing List <http://www.php.net/>
To unsubscribe, visit: http://www.php.net/unsub.php
nce a milestone in the php-apache front, especially regarding
> *security*.
What kind of security issues are you refering to? We're not quite ready
for full Apache 2 support, but I hope to have all known issues fixed
today (in the next couple hours).
-aaron
--
PHP Development Mailing Lis
he JPEG is
> written in full rather than just the first 4k.
Yes, after looking at other modules it appears you are correct and
it was just broken before. I'll change it to return the number of
consumed bytes.
-aaron
--
PHP Development Mailing List <http://www.php.net/>
To unsubscribe, visit: http://www.php.net/unsub.php
should get us working on all known
MPMs, including windows and other platform-specific MPMs. Static module
support is something that probably needs to be solved on a larger scale
from inside apache, but we can address that too.
-aaron
--
PHP Development Mailing List <http://www.php.net/&
mise! */
>
> Why did you change this to return 0? It breaks ImageJpeg
> (and likely other things) as GD checks that the return value
> matches the passed length & aborts if it doesn't.
>
> Otherwise, this patch seems to work much better than the HEAD
> code.
It seems to me that it returns any remaining data, not how much was
written. In the old code there was no way that it would ever return
non-zero, so I decided to make it explicit. Did I interpret the
wb_write function incorrectly?
-aaron
--
PHP Development Mailing List <http://www.php.net/>
To unsubscribe, visit: http://www.php.net/unsub.php
s change
occured, but the part of me that spent *way* too long on this
has to ask: Why did this change? Doesn't this violate the
Principle of Least Astonishment?
-aaron
--
PHP Development Mailing List <http://www.php.net/>
To unsubscribe, visit: http://www.php.net/unsub.php
I'd like to work on the apache2filter (I'm an active
developer on Apache 2.0). In addition to the fixes
I posted today, I have some good ideas how the
performance of the apache2filter can be greatly
improved.
(p.s. my requested password is intentionally lame, I'd
prefer to change it throug
On Thu, Apr 11, 2002 at 02:20:00AM +0100, Wez Furlong wrote:
> the php configuration directive register_globals defaults to off
> in PHP 4.2; are you sure it's not just due to misconfiguration?
I'll be ecstatic if that's the problem. :)
-aaron
--
PHP Developme
fail
to retrieve the same values it set earlier. I'm really hoping that
someone experienced in Zend and the SAPI will spot the problem right
away and point me in the right direction.
-aaron
--
PHP Development Mailing List <http://www.php.net/>
To unsubscribe, visit: http://www.php.net/unsub.php
this list and others are fixed with
> > this patch, as well as some other potential problems.
>
> Great work. Could you please send the patch as an email attachment? You
> should probably apply for the CVS account if you didn't have one already.
Attached, thanks.
-aaron
Index: sa
On Wed, Apr 10, 2002 at 04:45:53PM -0400, Cliff Woolley wrote:
> FWIW, Aaron and I are trying really hard to get the Apache 2 support
> ironed out in the very near term (like the next day or so). It's not far
> off. Serious improvements can still be made, and those could wait u
E, it steps out of
the way and passes all other brigades down the chain.
The SEGVs reported earlier on this list and others are fixed with
this patch, as well as some other potential problems.
-aaron
p.s. this patch applies to the php4_2_0RC2 branch, and includes some
changes from HEAD that weren&
patch corrects intermittent SEGVs in the apache2filter. The
patch is against the 4.0.2RC2 branch/tag, and includes changes
from HEAD that were made to work with new bucket allocator code in
apache 2.
-aaron
Index: sapi/apache2filter/sapi_apache2.c
=
wrong
here and it's threadsafe already :).
OTOH, I see that the nsapi.c implementation isn't using any
obvious mutexes, but it is using the SG(request_info) in the same
way. What am I missing?
-aaron
--
PHP Development Mailing List <http://www.php.net/>
To unsubscribe, visit: http://www.php.net/unsub.php
ex to call a callback
registered by the SAPI module. This will allow the lexer to pull
input data in whatever method is most efficient to the SAPI module.
(Cliff and I want this specifically for the apache2filter so that
we can pull data straight from the bucket brigade provided in
Apache 2.)
-
On Sat, Apr 06, 2002 at 07:56:56AM -0500, Jim Jagielski wrote:
> Aaron Bannert wrote:
> >
> > I think a good way to take advantage of the filter system would be
> > to break the input stream into buckets of two types: constant strings
> > and PHP script code. As flex
ed to me, but I can't find all the bugs on
my own. The best way to find those edge cases is by getting user exposure.
Just my 2c,
-aaron
--
PHP Development Mailing List <http://www.php.net/>
To unsubscribe, visit: http://www.php.net/unsub.php
is
not consistently reproduceable.
Anyone had any luck with another version?
-aaron
--
PHP Development Mailing List <http://www.php.net/>
To unsubscribe, visit: http://www.php.net/unsub.php
feasible in the SAPI domain? I know very little about the
SAPI, but will it allow me to override the equivalent of the yyinput
function?
-aaron
--
PHP Development Mailing List <http://www.php.net/>
To unsubscribe, visit: http://www.php.net/unsub.php
$i\n";
}
?>
Anyone else working with the apache2filter?
-aaron
--
PHP Development Mailing List <http://www.php.net/>
To unsubscribe, visit: http://www.php.net/unsub.php
er options)
to make sure it is of type integer rather than type double?
Aaron
--
PHP Development Mailing List <http://www.php.net/>
To unsubscribe, visit: http://www.php.net/unsub.php
u:/www/bug-11006.php
http://www.uvm.edu:/www/bug-11006.phps
aaron
--
Edit bug report at http://bugs.php.net/?id=15374&edit=1
--
Fixed in CVS:http://bugs.php.net/fix.php?id=15374&r=fixedcvs
Fixed in release:http://bugs.php.net/fix.php?id=15374&r=alread
ID: 15011
Updated by: [EMAIL PROTECTED]
-Reported By: [EMAIL PROTECTED]
+Reported By: [EMAIL PROTECTED]
Status: Open
Bug Type: *Configuration Issues
Operating System: Linux
PHP Version: 4.1.1
New Comment:
can now:
http://cvs.php.net/co.
reported as "not being
found"..u...
Can you offer any assistance?
Regards,
Aaron
--
Edit bug report at: http://bugs.php.net/?id=14667&edit=1
--
PHP Development Mailing List <http://www.php.net/>
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands,
From: [EMAIL PROTECTED]
Operating system: Linux
PHP version: 4.1.0
PHP Bug Type: Apache2 related
Bug description: configure script looks in the wrong place for httpd.h
I'm trying to configure php 4.1.0 on my server running turbolinux 6.0 with
Apache 2.0.28
By default Ap
,
aaron
*** Socket.php.old Fri Sep 7 18:19:12 2001
--- Socket.php Fri Sep 7 19:13:33 2001
***
*** 291,299
$timeout = time() + $this->timeout;
while (!feof($this->fp) && (!$this->timeout || time() < $timeout)) {
fine.
Aaron.
Previous Comments:
---
[2001-03-19 14:39:43] [EMAIL PROTECTED]
both production and development environments
are using identical versions and configurations
of apache web server 1.3.14 and php 4.0.4 pl1
D]
Woops!
No it still doesnt work.
Aaron.
---
[2001-03-19 14:16:04] [EMAIL PROTECTED]
both production and development environments
are using identical versions and configurations
of apache web server 1.3.14 an
ID: 9390
User Update by: [EMAIL PROTECTED]
Old-Status: Closed
Status: Open
Bug Type: *Session related
Description: PHPSESSID adds itself to $HTTP_POST_VARS
Woops!
No it still doesnt work.
Aaron.
Previous Comments
ID: 9390
User Update by: [EMAIL PROTECTED]
Old-Status: Open
Status: Closed
Bug Type: *Session related
Description: PHPSESSID adds itself to $HTTP_POST_VARS
both production and development environments
are using identical versions and configurations
of apache web server 1.3.14 and php 4.0.4 p
ID: 9390
User Update by: [EMAIL PROTECTED]
Old-Status: Feedback
Status: Open
Bug Type: *Session related
Description: PHPSESSID adds itself to $HTTP_POST_VARS
Thanx Jani,
As suggested:
In php.ini:
url_rewriter.tags = "a=href,area=href,frame=src,input=src,form=fakeentry"
Remove the last
ID: 9390
User Update by: [EMAIL PROTECTED]
Old-Status: Closed
Status: Open
Bug Type: *Session related
Description: PHPSESSID adds itself to $HTTP_POST_VARS
Thanx Jani,
As suggested:
In php.ini:
url_rewriter.tags = "a=href,area=href,frame=src,input=src,form=fakeentry"
Remove the last s
ID: 9727
User Update by: [EMAIL PROTECTED]
Old-Status: Closed
Status: Open
Bug Type: Filesystem function related
Description: Unable to use @ in URL passwords with fopen()
Here's my exact code with only the username and password changed (oops - i sent a
slightly altered version before, to change
From: [EMAIL PROTECTED]
Operating system: rh 6.1
PHP version: 4.0.4pl1
PHP Bug Type: Filesystem function related
Bug description: Unable to use @ in URL passwords with fopen()
I'm running 4.0.4pl1 on redhat 6.1. I've tried (raw and regular) urlencoding my ftp
password to u
From: [EMAIL PROTECTED]
Operating system: HP-UX 11.00
PHP version: 4.0.4pl1
PHP Bug Type: *Session related
Bug description: PHPSESSID adds itself to $HTTP_POST_VARS
Problem:
For some reason PHP adds PHPSESSID to $HTTP_POST_VARS
At login, I use:
session_start():
$sid = sess
ID: 9106
User Update by: [EMAIL PROTECTED]
Status: Open
Bug Type: Reproduceable crash
Description: Floating Point and Arithmetic Exception crashes
the change helps, but today i managed another FPE.
so there's something else to be fixed here.
Previous Comments:
--
From: [EMAIL PROTECTED]
Operating system: FreeBSD 4.2-STABLE
PHP version: 4.0.4pl1
PHP Bug Type: Reproduceable crash
Bug description: Floating Point and Arithmetic Exception crashes
I have suffered from
/kernel: pid 69000 (httpd), uid 99: exited on signal 8
repeatedly in
80 matches
Mail list logo
