Re: 2.0.31 shutdown after heavy load + core dumps on heavy load
I'm seeing the same thing with the prefork mpm under linux with lots of load.
Possibly it
is something outside of the mpms themselves? The processes definitly don't go away
until
you kill them with SIGKILL a few times. One thing I noticed is that, under load,
processes
seem to spend a lot of time in the 'close connection' state. By the time you have to
issue
the kills, mod_status shows idle children. Perhaps there is something that is not
getting
cleaned up properly?
On a possibly related note, I am seeing segfaults in prefork children under load when
the
number of children is high (over 1000). The stacktrace looks like so and doesn't make
a lot of sense:
#0 pthread_sighandler (signo=11, ctx=
{gs = 0, __gsh = 0, fs = 0, __fsh = 0, es = 43, __esh = 0, ds = 43, __dsh = 0,
edi = 1076506184, esi = 0, ebp = 3221223720, esp = 3221223680, ebx = 1076516568, edx =
1, ecx = 0, eax = 0, trapno = 13, err = 0, eip = 1076467940, cs = 35, __csh = 0,
eflags = 66070, esp_at_signal = 3221223680, ss = 43, __ssh = 0, fpstate = 0xb680,
oldmask = 2147483648, cr2 = 0}) at signals.c:87
87 signals.c: No such file or directory.
in signals.c
(gdb) where
#0 pthread_sighandler (signo=11, ctx=
{gs = 0, __gsh = 0, fs = 0, __fsh = 0, es = 43, __esh = 0, ds = 43, __dsh = 0,
edi = 1076506184, esi = 0, ebp = 3221223720, esp = 3221223680, ebx = 1076516568, edx =
1, ecx = 0, eax = 0, trapno = 13, err = 0, eip = 1076467940, cs = 35, __csh = 0,
eflags = 66070, esp_at_signal = 3221223680, ss = 43, __ssh = 0, fpstate = 0xb680,
oldmask = 2147483648, cr2 = 0}) at signals.c:87
#1
#2 __pthread_reset_main_thread () at internals.h:372
#3 0x402990e5 in __fork () at ptfork.c:92
#4 0x0809b2cd in ap_graceful_stop_signalled () at eval.c:41
#5 0x0809b643 in ap_graceful_stop_signalled () at eval.c:41
#6 0x0809ba62 in ap_mpm_run () at eval.c:41
#7 0x080a24ee in main () at eval.c:41
#8 0x402c2177 in __libc_start_main (main=0x80a1b9c , argc=1, ubp_av=0xbb44,
init=0x80637a8 <_init>,
fini=0x80c2f50 <_fini>, rtld_fini=0x4000e184 <_dl_fini>, stack_end=0xbb3c) at
../sysdeps/generic/libc-start.c:129
Anyone have an idea how to track this down? The ap_graceful_stop_signalled function
does *nothing* in the prefork mpm.
-adam
On Mon, Feb 04, 2002 at 02:13:52PM -0500, MATHIHALLI,MADHUSUDAN (HP-Cupertino,ex1)
wrote:
> This was with the worker MPM.. I initially suspected the latency - 'waited
> for about 5 minutes, and then nothing happened till I issued a "kill -9"
> command. I'll try again today, and will probably post the stack trace of the
> parent process when it happens.
>
> Thanks
> -Madhu
>
> -Original Message-
> From: Greg Ames [mailto:[EMAIL PROTECTED]]
> Sent: Monday, February 04, 2002 9:58 AM
> To: [EMAIL PROTECTED]
> Subject: Re: 2.0.31 shutdown after heavy load
>
>
> "MATHIHALLI,MADHUSUDAN (HP-Cupertino,ex1)" wrote:
> >
> > Hi,
> > I'm getting the following message when I try to stop apache after
> a
> > stress-test on HPUX (using webstone).. Inspite of the SIGKILL message, the
> > process does not exit.. A second attempt is successful.. Any clues
> regarding
> > what may be happening is appreciated..
> >
> > [error] child process 5891 still did not exit, sending a SIGKILL.
>
> which MPM? and have you tried this test with different results before?
>
> I didn't think we could do anything to block SIGKILL. Could this be just
> dispatching latency?
>
> Greg
--
"I believe in Kadath in the cold waste, and Ultima Thule. But you
cannot prove to me that Harvard Law School actually exists."
- Theodora Goss
"I'm not like that, I have a cat, I don't need you.. My cat, and
about 18 lines of bourne shell code replace you in life."
- anonymous
Adam Sussman
Vidya Media Ventures
[EMAIL PROTECTED]
mod_rewrite.h:133:18: ndbm.h: No such file or directory
Hi All, It's that problem again, this time with apache_1.3.23 on linux (rh7.2). I am building it with php4 and most of other modules. Here's my config: #! /bin/sh EAPI_MM=SYSTEM export EAPI_MM INCLUDES="-I/usr/include/db1 -I/usr/local/BerkeleyDB.4.0/include -I/usr/local/include" \ LDFLAGS="-L/usr/local/BerkeleyDB.4.0/lib -L/usr/local/lib -L/usr/local/src/openldap/lib" \ EAPI_MM=SYSTEM \ CC=gcc LIBS="-lldap -llber -ldb" OPTIM="-O2" \ ./configure \ --prefix=/usr/local/apache_1.3.23+mod_ssl+php \ --activate-module=src/modules/php4/libphp4.a \ --enable-module=ssl \ --enable-module=php4 \ --enable-module=most Error: gcc -c -I../../os/unix -I../../include -O2 -DLINUX=22 -DMOD_SSL=208106 -I/usr/local/src/php -I/usr/local/src/php/main -I/usr/local/src/php/main -I/usr/local/src/php/Zend -I/usr/local/src/php/Zend -I/usr/local/src/php/TSRM -I/usr/local/src/php/TSRM -I/usr/local/src/php -DEAPI -DEAPI_MM -DNO_DL_NEEDED `../../apaci` mod_rewrite.c In file included from mod_rewrite.c:93: mod_rewrite.h:133:18: ndbm.h: No such file or directory make[4]: *** [mod_rewrite.o] Error 1 make[3]: *** [all] Error 1 make[2]: *** [subdirs] Error 1 make[2]: Leaving directory `/usr/local/src/apache_1.3.23+mod_ssl+php/src' make[1]: *** [build-std] Error 2 make[1]: Leaving directory `/usr/local/src/apache_1.3.23+mod_ssl+php' make: *** [build] Error 2 Without LIBS=, apache will add -I/usr/include/db1 and there would be no problem with mod_rewrite; otherwise, apache won't be able to find the ldap library files. I added INCLUDES=, but it gets ignored. Thanks, Jie
Re: more minor 2.0.31 breakage
From: "Stas Bekman" <[EMAIL PROTECTED]> Sent: Monday, February 04, 2002 11:18 PM > >>I think it's a bug. The sub-directory can be password-protected. Does it > >>mean that mod_autoindex won't display it, since it'll fail to run the > >>sub-request? > >> > > > > By design, for better or worse [feel free to debate here.] > > > > The new logic [perhaps 1.3 as well, you would have to look] does a subrequest > > on the /~stas/ url ... and decides it's broken [or inaccessable, or whatever] > > and autoindex won't display it. What's broken about not displaying things that > > the user cannot access? > > I gave an example of a password protected sub-directory. You say that > it's correct not to display it then? How do you know whether the user > cannot access this directory? May be he does have the needed access > credentials. Fair enough - we can rescope the rules. Right now 200s and 300s pass. The test is right there in autoindex if you would like to change it. Personally, I never want to see 500'ed resources return OK ;-/
Re: more minor 2.0.31 breakage
William A. Rowe, Jr. wrote: > From: "Stas Bekman" <[EMAIL PROTECTED]> > Sent: Monday, February 04, 2002 8:39 PM > > > >>Rodent of Unusual Size wrote: >> >> >>>Greg Ames wrote: >>> >>> I just noticed that autoindex listings from apache.org no longer display the folder icon for directories. You see a '?' icon instead. The Rodent tells me there's some kind of magic type that mod_autoindex uses to figure this out. Not debugged yet. >>>As an unsubstantiated WAG, I would guess that the subrequest >>>that's sent off to determine the content-type is running >>>afoul of mod_dir somehow, and S_ISDIR isn't set correctly >>>when it comes back. >>> >> >>BTW, since you are talkinga about mod_autoindex, I don't recall that it >>was working this way before: >> >>I've uploaded a dir with files to apache.org/~stas/. as of that moment >>apache.org was running 2.0.28. I was trying to view apache.org/~stas/, >>all the files/dirs showed up, but not this new sub-dir. When I've tried >>to enter the new directory manually >>http://apache.org/~stas/modperl-site/ I've got 500 error. Looking at >>error_log revealed that I had a broken .htaccess in that directory. I >>don't understand why mod_autoindex won't display the directory in the >>index's listing. >> >>I think it's a bug. The sub-directory can be password-protected. Does it >>mean that mod_autoindex won't display it, since it'll fail to run the >>sub-request? >> > > By design, for better or worse [feel free to debate here.] > > The new logic [perhaps 1.3 as well, you would have to look] does a subrequest > on the /~stas/ url ... and decides it's broken [or inaccessable, or whatever] > and autoindex won't display it. What's broken about not displaying things that > the user cannot access? I gave an example of a password protected sub-directory. You say that it's correct not to display it then? How do you know whether the user cannot access this directory? May be he does have the needed access credentials. -- _ Stas Bekman JAm_pH -- Just Another mod_perl Hacker http://stason.org/ mod_perl Guide http://perl.apache.org/guide mailto:[EMAIL PROTECTED] http://ticketmaster.com http://apacheweek.com http://singlesheaven.com http://perl.apache.org http://perlmonth.com/
Re: Wondering what I'm missing...
Recognize that rmm_malloc returns apr_rmm_off_t's into an apr_rmm_t... you need to port all of the calls to deref through apr_rmm_baseaddr_get(), but be careful, don't store dereferenced real addresses in the rmm itself! - Original Message - From: "Aaron Bannert" <[EMAIL PROTECTED]> To: <[EMAIL PROTECTED]> Sent: Monday, February 04, 2002 10:02 PM Subject: Re: Wondering what I'm missing... > On Mon, Feb 04, 2002 at 09:27:53PM -0500, Paul J. Reder wrote: > > Perhaps I'm just really dense, but I don't see where (in the > > mod_auth_digest.c code in httpd-2.0 HEAD) the following variable: > ... > > What obvious point am I missing? > > Unless we're both blind, it's missing a call to apr_rmm_init(). > > -aaron >
Re: Wondering what I'm missing...
On Mon, Feb 04, 2002 at 09:27:53PM -0500, Paul J. Reder wrote: > Perhaps I'm just really dense, but I don't see where (in the > mod_auth_digest.c code in httpd-2.0 HEAD) the following variable: ... > What obvious point am I missing? Unless we're both blind, it's missing a call to apr_rmm_init(). -aaron
Re: versioning process
According to Roy T. Fielding: > And, personally, I have always hated the stupid alpha/beta/GA distinction. > Our release process became constipated on the day that was added. I tend to agree. And our users get (or area already) confused by this release process. ciao... -- Lars Eilebrecht- vuja de: The feeling that you've *never*, [EMAIL PROTECTED] - *ever* been in this situation before.
Re: Apache2 and php
In further trial and error, even that is wrong. You must have both the
AddType directive as well as the Filter section when trying to use php
with Apache2.
Now that I have the php engine going, I need to find and fix a segfault.
I installed phpSysInfo 1.9 in default state. Two things happen.
a) segfault when index.php is loaded and a new child is formed which
then serves the page
b) if i try to change the template theme, httpd gobbles memory as fast
as it can before the OOM killer slaps it sideways.
(gdb) bt
#0 0x080989a0 in ap_save_brigade (f=0x81cbd20, saveto=0x81c8304,
b=0xb364, p=0x81c79e0) at util_filter.c:473
#1 0x402785e0 in php_output_filter (f=0x81cbd20, bb=0x81cbe80)
at sapi_apache2.c:330
#2 0x0809891a in ap_pass_brigade (next=0x81cbd20, bb=0x81cbe80)
at util_filter.c:445
#3 0x0809eff6 in default_handler (r=0x81c7a18) at core.c:2992
#4 0x0808df56 in ap_run_handler (r=0x81c7a18) at config.c:185
#5 0x0808e3f1 in ap_invoke_handler (r=0x81c7a18) at config.c:359
#6 0x08075569 in ap_process_request (r=0x81c7a18) at http_request.c:290
#7 0x08071a49 in ap_process_http_connection (c=0x818df10) at http_core.c:287
---Type to continue, or q to quit---
#8 0x08096dd6 in ap_run_process_connection (c=0x818df10) at connection.c:86
#9 0x0808ca57 in child_main (child_num_arg=1) at prefork.c:717
#10 0x0808cc00 in make_child (s=0x818cce0, slot=1) at prefork.c:807
#11 0x0808cde5 in perform_idle_server_maintenance (p=0x81003b8)
at prefork.c:948
#12 0x0808d31f in ap_mpm_run (_pconf=0x81003b8, plog=0x8138498, s=0x818cce0)
at prefork.c:1105
#13 0x08091f19 in main (argc=1, argv=0xb6a4) at main.c:501
(gdb) p *e
$2 = {link = {next = 0x819de98, prev = 0x81f4248}, type = 0x814a5e8,
length = 136101408, start = 136101504, data = 0x80a9e00,
free = 0x80a9e00 }
(gdb) p rv
$3 = 1094927173
(gdb) p /x rv
$4 = 0x41434345
(gdb) p *f
$6 = {frec = 0x8186e08, ctx = 0x0, next = 0x81cbd58, r = 0x81c7a18,
c = 0x818df10}
(gdb) p *saveto
$8 = (apr_bucket_brigade *) 0x41434345
(gdb) p **b
$9 = {p = 0x81c79e0, list = {next = 0x819de98, prev = 0x81f4248}}
(gdb) p *p
$10 = {parent = 0x818de08, child = 0x0, sibling = 0x0, ref = 0x818de0c,
cleanups = 0x81cbe90, subprocesses = 0x0, abort_fn = 0, user_data = 0x0,
tag = 0x0, allocator = 0x8153108, active = 0x81cba70, self = 0x81c79d0,
self_first_avail = 0x81c7a18 "?y\034\b\020?\030\b??\030\b"}
Again, your indulgence in helping is appreciated. This is HEAD of httpd2
and php 4.1.1. php42 HEAD simply does not work,
David
David Ford wrote:
> My apologies, I realized that after I sent it. Actually that is
> incorrect. This seems to be a fairly unknown answer as nobody was able
> to answer it for me in the past on either the php or apache groups.
>
> http://marc.theaimsgroup.com/?l=php-dev&m=101180222102820&w=2
>
>
>/you must replace :
>
>LoadModule php4_module modules/libphp4.so
>AddType application/x-httpd-php .php
>
>By
>
>LoadModule php4_module modules/libphp4.so
>
>SetOutputFilter PHP
>SetInputFilter PHP
>/
>
> Thank you,
> David
Re: more minor 2.0.31 breakage
From: "Stas Bekman" <[EMAIL PROTECTED]> Sent: Monday, February 04, 2002 8:39 PM > Rodent of Unusual Size wrote: > > > Greg Ames wrote: > > > >>I just noticed that autoindex listings from apache.org no > >>longer display the folder icon for directories. You see > >>a '?' icon instead. The Rodent tells me there's some kind > >>of magic type that mod_autoindex uses to figure this out. > >>Not debugged yet. > >> > > > > As an unsubstantiated WAG, I would guess that the subrequest > > that's sent off to determine the content-type is running > > afoul of mod_dir somehow, and S_ISDIR isn't set correctly > > when it comes back. > > > BTW, since you are talkinga about mod_autoindex, I don't recall that it > was working this way before: > > I've uploaded a dir with files to apache.org/~stas/. as of that moment > apache.org was running 2.0.28. I was trying to view apache.org/~stas/, > all the files/dirs showed up, but not this new sub-dir. When I've tried > to enter the new directory manually > http://apache.org/~stas/modperl-site/ I've got 500 error. Looking at > error_log revealed that I had a broken .htaccess in that directory. I > don't understand why mod_autoindex won't display the directory in the > index's listing. > > I think it's a bug. The sub-directory can be password-protected. Does it > mean that mod_autoindex won't display it, since it'll fail to run the > sub-request? By design, for better or worse [feel free to debate here.] The new logic [perhaps 1.3 as well, you would have to look] does a subrequest on the /~stas/ url ... and decides it's broken [or inaccessable, or whatever] and autoindex won't display it. What's broken about not displaying things that the user cannot access? Bill
Re: Wondering what I'm missing...
* On 2002-02-04 at 22:07, Paul J. Reder <[EMAIL PROTECTED]> excited the electrons to say: > > Perhaps I'm just really dense, but I don't see where (in the > mod_auth_digest.c code in httpd-2.0 HEAD) the following variable: > > static apr_rmm_t *client_rmm = NULL; > > ever gets initialized before it is used in the 5 calls to: Uh, that declaration right there is initialising it to NULL.. -- #kenP-)} Ken Coar, Sanagendamgagwedweinini http://Golux.Com/coar/ Author, developer, opinionist http://Apache-Server.Com/ "Millennium hand and shrimp!"
Re: versioning process
* On 2002-02-04 at 22:07, Roy T. Fielding <[EMAIL PROTECTED]> excited the electrons to say: > > After the group spends enough time pounding on a release to call it GA, > there is no reason why we cannot take the code and reversion it a 2.1.0 > and start treating that as the stable branch. It may be a little > confusing to the Linux folks, but we can always skip to 2.2.0 if that > makes people happier. Personally, I tend to think that even-numbered > releases are less reliable than odd ones, because that has been my > experience with the quality of commercial releases. Yar. Lotsa people say, "I never install an x.0 release; I always wait at least for x.1". Some manufacturers would make .even release 'feature' releases, and .odd ones 'fix' releases. > Of course, the real problem is that httpd developers these days just have > it too easy. When I was an RM, we had to get up at 4am and walk through > snow in our underwear And it was uphill both ways. :-) -- #kenP-)} Ken Coar, Sanagendamgagwedweinini http://Golux.Com/coar/ Author, developer, opinionist http://Apache-Server.Com/ "Millennium hand and shrimp!"
Re: [PATCH] mod_disk_cache.c mod_cache.c
Committed. Thanks Joe.
Bill
>
> I made a mimimal set of changes to mod_disk_cache.c so that I could
> simply try out the current mod_disk_cache functionality. This patch
> fixes a few core dumps related to usage of the apr_pstrcat method and
> some minor compiler warnings.
>
>
> Index: modules/experimental/mod_cache.c
> ===
> RCS file: /home/cvspublic/httpd-2.0/modules/experimental/mod_cache.c,v
> retrieving revision 1.25
> diff -u -r1.25 mod_cache.c
> --- modules/experimental/mod_cache.c5 Feb 2002 00:15:42 - 1.25
> +++ modules/experimental/mod_cache.c5 Feb 2002 01:56:28 -
> @@ -931,8 +931,8 @@
> conf->complete_set = 1;
> return NULL;
> }
> -static cache_post_config(apr_pool_t *p, apr_pool_t *plog,
> -apr_pool_t *ptemp, server_rec *s)
> +static int
> +cache_post_config(apr_pool_t *p, apr_pool_t *plog, apr_pool_t *ptemp, server_rec *s)
> {
> /* This is the means by which unusual (non-unix) os's may find alternate
> * means to run a given command (e.g. shebang/registry parsing on Win32)
> Index: modules/experimental/mod_disk_cache.c
> ===
> RCS file: /home/cvspublic/httpd-2.0/modules/experimental/mod_disk_cache.c,v
> retrieving revision 1.17
> diff -u -r1.17 mod_disk_cache.c
> --- modules/experimental/mod_disk_cache.c 4 Feb 2002 14:43:11 - 1.17
> +++ modules/experimental/mod_disk_cache.c 5 Feb 2002 01:56:28 -
> @@ -59,11 +59,13 @@
> #include "mod_cache.h"
> #include "apr_file_io.h"
> #include "apr_strings.h"
> -#include "http_config.h"
> -#include "http_log.h"
> #include "util_filter.h"
> #include "util_script.h"
>
> +#if APR_HAVE_UNISTD_H
> +#include /* needed for unlink/link */
> +#endif
> +
> /*
> * disk_cache_object_t
> * Pointed to by cache_object_t::vobj
> @@ -116,7 +118,7 @@
> {
> char *hashfile;
> hashfile = generate_name(p, dirlevels, dirlength, name);
> -return apr_pstrcat(p, root, "/", hashfile, CACHE_HEADER_SUFFIX);
> +return apr_pstrcat(p, root, "/", hashfile, CACHE_HEADER_SUFFIX, NULL);
> }
>
> static char *data_file(apr_pool_t *p, int dirlevels, int dirlength,
> @@ -124,7 +126,7 @@
> {
> char *hashfile;
> hashfile = generate_name(p, dirlevels, dirlength, name);
> -return apr_pstrcat(p, root, "/", hashfile, CACHE_DATA_SUFFIX);
> +return apr_pstrcat(p, root, "/", hashfile, CACHE_DATA_SUFFIX, NULL);
> }
>
> static int mkdir_structure(char *file, const char *root)
> @@ -304,7 +306,7 @@
> obj->complete = 0; /* Cache object is not complete */
>
> info = apr_pcalloc(r->pool, sizeof(cache_info));
> -info->name = key;
> +info->name = (char *) key;
> obj->info = *(info);
>
> #ifdef AS400
> @@ -351,7 +353,6 @@
> {
> disk_cache_conf *conf = ap_get_module_config(r->server->module_config,
> &disk_cache_module);
> -apr_status_t ret = DECLINED;
> apr_status_t rc;
> char *data = data_file(r->pool, conf->dirlevels, conf->dirlength,
> conf->cache_root, key);
> @@ -366,12 +367,12 @@
> }
>
> obj = apr_pcalloc(r->pool, sizeof(cache_object_t));
> -obj->key = key;
> +obj->key = (char *) key;
>
> rc = apr_file_open(&fd, data, APR_WRITE | APR_READ | APR_BINARY, 0, r->pool);
> if (rc == APR_SUCCESS) {
> info = apr_pcalloc(r->pool, sizeof(cache_info));
> -info->name = key;
> +info->name = (char *) key;
> /* XXX log message */
> info->fd = fd;
> info->datafile = data;
> @@ -403,6 +404,7 @@
>
> static int remove_url(const char *type, char *key)
> {
> + return OK;
> }
>
> static int remove_entity(cache_handle_t *h)
> @@ -458,7 +460,7 @@
> }
>
> /* XXX log */
> -if(rv = file_cache_read_mydata(fd, info, r) != APR_SUCCESS) {
> +if((rv = file_cache_read_mydata(fd, info, r)) != APR_SUCCESS) {
> /* XXX log message */
> apr_file_close(fd);
> return rv;
> @@ -530,9 +532,9 @@
> else {
> /* XXX log message */
> }
> -if(rc = apr_file_open(&hfd, info->hdrsfile,
> +if((rc = apr_file_open(&hfd, info->hdrsfile,
>APR_WRITE | APR_CREATE | APR_BINARY | APR_EXCL, /*
>XXX:?
| APR_INHERIT | APR_NONQSYS, */
> - 0, r->pool) != APR_SUCCESS) {
> + 0, r->pool)) != APR_SUCCESS) {
> /* XXX log message */
> return rc;
> }
> @@ -555,7 +557,7 @@
> buf = apr_pstrcat(r->pool, statusbuf, CRLF, NULL);
> amt = strlen(buf);
> apr_file_write(hfd, buf, &amt);
> -buf = apr_pstrcat(r->pool, r->status_line, '\n', NULL);
> +buf = apr_pstrcat(r->pool, r->status_line, "\n", NULL);
> amt = strlen(buf);
> a
Re: more minor 2.0.31 breakage
Rodent of Unusual Size wrote: > Greg Ames wrote: > >>I just noticed that autoindex listings from apache.org no >>longer display the folder icon for directories. You see >>a '?' icon instead. The Rodent tells me there's some kind >>of magic type that mod_autoindex uses to figure this out. >>Not debugged yet. >> > > As an unsubstantiated WAG, I would guess that the subrequest > that's sent off to determine the content-type is running > afoul of mod_dir somehow, and S_ISDIR isn't set correctly > when it comes back. BTW, since you are talkinga about mod_autoindex, I don't recall that it was working this way before: I've uploaded a dir with files to apache.org/~stas/. as of that moment apache.org was running 2.0.28. I was trying to view apache.org/~stas/, all the files/dirs showed up, but not this new sub-dir. When I've tried to enter the new directory manually http://apache.org/~stas/modperl-site/ I've got 500 error. Looking at error_log revealed that I had a broken .htaccess in that directory. I don't understand why mod_autoindex won't display the directory in the index's listing. I think it's a bug. The sub-directory can be password-protected. Does it mean that mod_autoindex won't display it, since it'll fail to run the sub-request? _ Stas Bekman JAm_pH -- Just Another mod_perl Hacker http://stason.org/ mod_perl Guide http://perl.apache.org/guide mailto:[EMAIL PROTECTED] http://ticketmaster.com http://apacheweek.com http://singlesheaven.com http://perl.apache.org http://perlmonth.com/
Wondering what I'm missing...
Perhaps I'm just really dense, but I don't see where (in the mod_auth_digest.c code in httpd-2.0 HEAD) the following variable: static apr_rmm_t *client_rmm = NULL; ever gets initialized before it is used in the 5 calls to: apr_rmm_malloc(client_rmm, ...); and the single call to: apr_rmm_free(client_rmm, (apr_rmm_off_t)entry); call). Neither of these calls initialize it. It is not referenced anywhere else. It is static. The functions (apr_rmm_malloc/free) assume that client_rmm points to a vald struct and they dereference into it. Some of the calls to malloc are within an #if APR_HAS_SHARED_MEMORY block, but most of the calls are not. What obvious point am I missing? Thanks, -- Paul J. Reder --- "The strength of the Constitution lies entirely in the determination of each citizen to defend it. Only if every single citizen feels duty bound to do his share in this defense are the constitutional rights secure." -- Albert Einstein
Insecure installations of cgi wrappers (RTFM people!)
This isn't really a reporting of a vulnerability, it's more a reporting of mind-bogglingly foolish administrators that refuse to follow installation instructions and read the documentation. (I've cc'd this to both the cgiwrap and apache development mailing lists, but I'm sure certain it's not news to readers of either.) Note the following from cgiwrap documentation: --- *VERY IMPORTANT* - Do NOT allow any non-trusted user to run scripts directly out of the main cgi-bin directory, as this will allow them to use cgiwrap to run any of the other users scripts. The reason for this is that if they can run scripts as the same userid as the web server, they can subvert some of cgiwrap's security checks to allow them to run other users scripts. I recommend not running ANY scripts on the web server directly, once you have cgiwrap installed. --- I FREQUENTLY receive messages like this: --- Hi : My web host provides us with CgiWrap access. However they only treat scripts installed inside cgi-bin to run as user me and not nobody. I wanted to know if there is a way to get CgiWrap to get scripts installed outside cgi-bin to run as user me, and not nobody ? --- What that tell's me is that web host is a security disaster waiting to happen because they are allowing both cgiwrap and scripts run directly from cgi-bin. It won't necessarily give root or anything like that, but it allows cgi scripts to have their environment COMPLETELY subverted. If there are any scripts that rely upon the authentication or access control provided by the web server (such as scripts to administer the contents of databases), they can be subverted simply because all of that information is passed via environment variables. I hate to see cgiwrap or apache/suexec or any of the other wrappers get the blame for administrators not reading the documentation. About the only way I can think of getting around this problem would be to have some sort of web-server -> cgi-wrapper token passing taking place with a shared secret compiled into the wrapper executable, combined with non-readable wrapper executables and web server config. (And I haven't thought about it enough to be sure that wouldn't be exploitable. With some of the ptrace stuff, I'd bet it probably could be exploited pretty quick.) To my knowledge, none of the wrappers are currently doing anything like this. CGIwrap most certainly isn't. -- Nathan (Author of CGIwrap) Nathan Neulinger EMail: [EMAIL PROTECTED] University of Missouri - Rolla Phone: (573) 341-4841 Computing Services Fax: (573) 341-4216
versioning process
> I know that there have been lots of changes in .32-dev since .31 was > tagged that have challenged the stability of the current tree. If > we took .32 and made it as close to .31 as we could but with the > *minimum* required changes to make .32, we could be close to a real > beta. (This seems to mimic the APACHE_BEST tag I think was proposed > before my time.) Actually, that one was during your time. The fact of the matter is that the RM doesn't even need to tag the tree. If they *really* want to test a configuration (set of revisions) on all platforms, then simply checkout HEAD (or whatever) and work within the checked-out tree, avoiding the use of cvs update except for specific file changes. A -dev tarball can be created (with whatever non-version name you want to give it) from the checked-out tree, which can then be tested by people with access to the other platforms. Only after they come back with a +1 is the checked-out tree tagged with the next version number, which impacts only those revisions in that tree and not whatever has happened to HEAD. I mention this every time the release process is argued: it is the responsiblility of the RM to tag a version of the tree that they *think* is stable. How they come to think that will entirely depend on how much time, energy, and resources they can throw at the code *before* it is tagged with a version number. They can do that by managing a checked-out tree or by using a temporary non-version tag to manage a branch of the real tree -- either one is better than tagging the tree with a public version number and them moving that tag around to suit the testers. Once it is tagged with a version number, we need to simply live with the fact that the version number is HISTORY and stop arguing about whether one change or another should be included. BTW, the version number means major.minor.patchlevel After the group spends enough time pounding on a release to call it GA, there is no reason why we cannot take the code and reversion it a 2.1.0 and start treating that as the stable branch. It may be a little confusing to the Linux folks, but we can always skip to 2.2.0 if that makes people happier. Personally, I tend to think that even-numbered releases are less reliable than odd ones, because that has been my experience with the quality of commercial releases. And, personally, I have always hated the stupid alpha/beta/GA distinction. Our release process became constipated on the day that was added. I guess the point is that we shouldn't let the process of tagging the tree interfere with getting a complete-revision-set of the server that is stable enough for a release. There is really no reason (aside from a desire to get three +1 votes) for the RM to need to coordinate with everyone else in the group when a release is ready to be tarballed, nor is there any reason (aside from the release votes) for the RM to include everything on HEAD when they do tag a release version. The only thing that is really necessary is that the final tarball has at least three +1 (and a majority positive) before it is placed in www.apache.org/dist/httpd/. If I were the RM, I'd just work in a tree of relatively stable code and tag the version when I bloody well felt like it was good enough to stake my reputation on as being better than the last version released. No, I don't have time to be an RM right now. OTOH, there is also very little reason why we can't have multiple RMs working in parallel, provided they work with private tag names and let the group decide which tags to promote to an official version. Of course, the real problem is that httpd developers these days just have it too easy. When I was an RM, we had to get up at 4am and walk through snow in our underwear in order get enough carrier pigeons together to distribute a tarball, and then wait a week for enough pigeons to come back alive to count it a release. When it's that hard to do a release, the RM spends a lot more of their own time proofing a tarball before it gets out, and makes a habit of doing a release before there are more than 30 or so changes since the last one, because the probability of that many changes being error-free is approaching zero. I think people just respect the process too much. We are here to maintain and release code -- the process only exists to allow us to get there without killing each other. If it isn't working, determine the part that doesn't work and replace it with something new. The only things that you cannot replace are that code must be committed before it is released and the tarball must be voted on before it is released to the public. Roy
Re: cvs commit: httpd-2.0 STATUS
From: <[EMAIL PROTECTED]> Sent: Monday, February 04, 2002 7:01 PM > + > +* Win32 allows multiple processes to attach to the same port. This > + means that I can open three DOS prompts, and run the same Apache > + binary with the same config file three different times, and the > + server won't complain. This becomes more serious ... My XP Home bundled with this new PC seems to have installed a port 8080 - our 'default' in development. Definately can't continue this behavior - it goes back to 1.3.x and should be fixed on both platforms.
[PATCH] mod_disk_cache.c mod_cache.c
I made a mimimal set of changes to mod_disk_cache.c so that I could
simply try out the current mod_disk_cache functionality. This patch
fixes a few core dumps related to usage of the apr_pstrcat method and
some minor compiler warnings.
Index: modules/experimental/mod_cache.c
===
RCS file: /home/cvspublic/httpd-2.0/modules/experimental/mod_cache.c,v
retrieving revision 1.25
diff -u -r1.25 mod_cache.c
--- modules/experimental/mod_cache.c5 Feb 2002 00:15:42 - 1.25
+++ modules/experimental/mod_cache.c5 Feb 2002 01:56:28 -
@@ -931,8 +931,8 @@
conf->complete_set = 1;
return NULL;
}
-static cache_post_config(apr_pool_t *p, apr_pool_t *plog,
-apr_pool_t *ptemp, server_rec *s)
+static int
+cache_post_config(apr_pool_t *p, apr_pool_t *plog, apr_pool_t *ptemp, server_rec *s)
{
/* This is the means by which unusual (non-unix) os's may find alternate
* means to run a given command (e.g. shebang/registry parsing on Win32)
Index: modules/experimental/mod_disk_cache.c
===
RCS file: /home/cvspublic/httpd-2.0/modules/experimental/mod_disk_cache.c,v
retrieving revision 1.17
diff -u -r1.17 mod_disk_cache.c
--- modules/experimental/mod_disk_cache.c 4 Feb 2002 14:43:11 - 1.17
+++ modules/experimental/mod_disk_cache.c 5 Feb 2002 01:56:28 -
@@ -59,11 +59,13 @@
#include "mod_cache.h"
#include "apr_file_io.h"
#include "apr_strings.h"
-#include "http_config.h"
-#include "http_log.h"
#include "util_filter.h"
#include "util_script.h"
+#if APR_HAVE_UNISTD_H
+#include /* needed for unlink/link */
+#endif
+
/*
* disk_cache_object_t
* Pointed to by cache_object_t::vobj
@@ -116,7 +118,7 @@
{
char *hashfile;
hashfile = generate_name(p, dirlevels, dirlength, name);
-return apr_pstrcat(p, root, "/", hashfile, CACHE_HEADER_SUFFIX);
+return apr_pstrcat(p, root, "/", hashfile, CACHE_HEADER_SUFFIX, NULL);
}
static char *data_file(apr_pool_t *p, int dirlevels, int dirlength,
@@ -124,7 +126,7 @@
{
char *hashfile;
hashfile = generate_name(p, dirlevels, dirlength, name);
-return apr_pstrcat(p, root, "/", hashfile, CACHE_DATA_SUFFIX);
+return apr_pstrcat(p, root, "/", hashfile, CACHE_DATA_SUFFIX, NULL);
}
static int mkdir_structure(char *file, const char *root)
@@ -304,7 +306,7 @@
obj->complete = 0; /* Cache object is not complete */
info = apr_pcalloc(r->pool, sizeof(cache_info));
-info->name = key;
+info->name = (char *) key;
obj->info = *(info);
#ifdef AS400
@@ -351,7 +353,6 @@
{
disk_cache_conf *conf = ap_get_module_config(r->server->module_config,
&disk_cache_module);
-apr_status_t ret = DECLINED;
apr_status_t rc;
char *data = data_file(r->pool, conf->dirlevels, conf->dirlength,
conf->cache_root, key);
@@ -366,12 +367,12 @@
}
obj = apr_pcalloc(r->pool, sizeof(cache_object_t));
-obj->key = key;
+obj->key = (char *) key;
rc = apr_file_open(&fd, data, APR_WRITE | APR_READ | APR_BINARY, 0, r->pool);
if (rc == APR_SUCCESS) {
info = apr_pcalloc(r->pool, sizeof(cache_info));
-info->name = key;
+info->name = (char *) key;
/* XXX log message */
info->fd = fd;
info->datafile = data;
@@ -403,6 +404,7 @@
static int remove_url(const char *type, char *key)
{
+ return OK;
}
static int remove_entity(cache_handle_t *h)
@@ -458,7 +460,7 @@
}
/* XXX log */
-if(rv = file_cache_read_mydata(fd, info, r) != APR_SUCCESS) {
+if((rv = file_cache_read_mydata(fd, info, r)) != APR_SUCCESS) {
/* XXX log message */
apr_file_close(fd);
return rv;
@@ -530,9 +532,9 @@
else {
/* XXX log message */
}
-if(rc = apr_file_open(&hfd, info->hdrsfile,
+if((rc = apr_file_open(&hfd, info->hdrsfile,
APR_WRITE | APR_CREATE | APR_BINARY | APR_EXCL, /*
XXX:? | APR_INHERIT | APR_NONQSYS, */
- 0, r->pool) != APR_SUCCESS) {
+ 0, r->pool)) != APR_SUCCESS) {
/* XXX log message */
return rc;
}
@@ -555,7 +557,7 @@
buf = apr_pstrcat(r->pool, statusbuf, CRLF, NULL);
amt = strlen(buf);
apr_file_write(hfd, buf, &amt);
-buf = apr_pstrcat(r->pool, r->status_line, '\n', NULL);
+buf = apr_pstrcat(r->pool, r->status_line, "\n", NULL);
amt = strlen(buf);
apr_file_write(hfd, buf, &amt);
buf = apr_pstrcat(r->pool, CRLF, NULL);
@@ -608,7 +610,7 @@
* mod_disk_cache configuration directives handlers.
*/
static const char
-*set_cache_root(cmd_parms *parms, void *struct_ptr, char *arg)
+*set_cache_root(cmd_parms *pa
Re: 2.0.31 on Windows
I guess I'm alittle confused... I don't see any errors related to NULL file handles. And the thing about the ServerRoot starting with a /, it seems that the process found the config file okay, so wouldn't it also know what the ServerRoot should be? And why doesn't Apache complain about the invalid ServerRoot variable when started from the command line? Even with the -t option, the error is unrecognized. I tried adding the -d option through the Service Manager to use when starting the service manually, with no change in behavior. I'll keep looking... Tks, Dwayne William A. Rowe, Jr. wrote: >From: "Dwayne Miller" <[EMAIL PROTECTED]> >Sent: Monday, February 04, 2002 5:13 PM > > >>>Which Win32 OS are you using, exactly? >>> >>I'm on Windows 2000, build 2195, SP2 >> > >Win2000/XP have NULL stdin/stdout/stderr handles... > >the create_process() invokes ap_open_stderr_log() which in turn invokes >apr_file_open_stderr() attempting to open a NULL STDERR handle. > >The patch I just committed reinvokes ap_open_stderr_log() after we jump into >the service control manager track with a stderr pipe that we capture to the >application event log. > >The right solution is to finish the move of the service manager code in the >apr_app_initialize code (apr_initialize + apr_app_main) which I hadn't finished >yet, and didn't plan to for a week, maybe two =-/ > >>I was looking at mpm_winnt.c, where the apr_filepath_merge() call is >>made. It looks like it is trying to build the -d argument to use when >>starting the service, but it's obviously not having the desired effect. >> > >That's where things are still goofy... I'm still debugging. > >More updates as they become available. > >Bill >
Re: --with-module annoyingness
On Mon, 4 Feb 2002, Ryan Bloom wrote: > Would anybody mind if --with-module created a symbolic link in the > modules/foo/ directory to the specified module? Right now, we copy the > module to that directory, which means that if you edit the .c file, it > gets out of sync with the version in the original directory. If you > have the .c file under revision control, that gets annoying quickly. I would keep in mind that the module could reside on a read-only file system and make sure it plays well with --srcdir/--shadow. Cheers, -- Sander van Zoest [EMAIL PROTECTED] San Diego, CA, US http://Sander.vanZoest.com/
Re: 2.0.31 on Windows
From: "Dwayne Miller" <[EMAIL PROTECTED]> Sent: Monday, February 04, 2002 5:13 PM > >Which Win32 OS are you using, exactly? > > > I'm on Windows 2000, build 2195, SP2 Win2000/XP have NULL stdin/stdout/stderr handles... the create_process() invokes ap_open_stderr_log() which in turn invokes apr_file_open_stderr() attempting to open a NULL STDERR handle. The patch I just committed reinvokes ap_open_stderr_log() after we jump into the service control manager track with a stderr pipe that we capture to the application event log. The right solution is to finish the move of the service manager code in the apr_app_initialize code (apr_initialize + apr_app_main) which I hadn't finished yet, and didn't plan to for a week, maybe two =-/ > I was looking at mpm_winnt.c, where the apr_filepath_merge() call is > made. It looks like it is trying to build the -d argument to use when > starting the service, but it's obviously not having the desired effect. That's where things are still goofy... I'm still debugging. More updates as they become available. Bill
Re: Thoughts on 2.0.32
From: "Ryan Bloom" <[EMAIL PROTECTED]> Sent: Monday, February 04, 2002 5:52 PM > > Since it sounds like Win32 is busted (I'd say it'd be fine to > > say that .31 is beta but doesn't work on Win32, but wrowe doesn't > > like that), > > IMNSHO, .31 should be released as a beta, with the understanding that it > doesn't work on Win32. That is a PITA, but it wouldn't be the first > time that a release of Apache didn't work the same on all platforms. .31 is an alpha already. What is the point of releasing something with such issues? I would consider a similar bug exclusively on Linux or Solaris also a showstopper. > > how about just taking the .31 tree, tagging *that* as > > .32, and carefully bumping files that resolve issues into .32 and > > the RM can carefully decide which other recent commits/changes are > > worthy of inclusion in .32? > > I strongly disagree with doing this. Watching each commit to determine > if it belongs in .32 is completely against our release model. Either > the code keeps moving forward or it doesn't, let's not bastardize the > release model. Ok, so we shouldn't work through our yet-unproven release model? If we can't get the release model right in early development, we will end up on the same treadmill as 1.3.x. When I joined the list, there hadn't been a release for a few months. There wasn't a new release for almost a year after I started hacking on the server. The first bug of Win32 is fixed (error logging to the event viewer), the second is not a showstopper (path names cannot start with /) although I need to finish testing to assure that this bug isn't harming other aspects of the server, and the third I will wrap up in a few hours (bad arguments created by apache -k install). There is little reason not to wait for these two/three patches - checkout 31, cvs up -A server/mpm/winnt, and then release that code. This cannot harm Unix (reasonably well tested for a first pass) and the code can be tested by everyone. Is anyone besides Ryan concerned about releasing a .32 with alpha-bugs only fixed, and then .33 with all the new bells and whistles (or destablizing code, depending on how you look at it :) ... then .34 with its fixes? Or are we all happy with the release model as it stands? Bill
RE: cvs commit: httpd-2.0/server/mpm/worker worker.c
I finally had time to review this patch, and I have some comments.
> Log:
> This patch restores most of Ryan's patch (11/12/2001) to remove the
> client_socket from the conn_rec. Diffs from Ryan's patch include:
>
> - rename the create_connection hook to install_transport_filters
> - move the point of invocation of the hook till after the call to
> after ap_update_vhost_given_ip to enable the hook to use vhost
> config info in its decision making.
>
> -AP_CORE_DECLARE(void) ap_process_connection(conn_rec *c)
> +AP_CORE_DECLARE(void) ap_process_connection(conn_rec *c,
apr_socket_t *csd)
I was very careful to keep the csd a void * instead of an apr_socket_t
*, because you have no way of knowing that the module that did the
accept will be giving us a socket, or at least not an apr_socket_t.
Right now, Aaron and I are hacking Apache to allow it to accept
connections over a Unix Domain Socket. This patch will break that
abstraction completely.
>{
>ap_update_vhost_given_ip(c);
>
> +ap_run_install_transport_filters(c, csd);
> +
>ap_run_pre_connection(c);
Why did we create a new hook to put here? This is the last place a new
hook is needed, in five lines of code in this function, we call three
hooks. Why not just add an argument to the pre_connection hook that
gives people access to the socket?
Unless there are any complaints, I'll be making both of those changes
today or tomorrow.
Ryan
RE: Thoughts on 2.0.32
> Since it sounds like Win32 is busted (I'd say it'd be fine to > say that .31 is beta but doesn't work on Win32, but wrowe doesn't > like that), IMNSHO, .31 should be released as a beta, with the understanding that it doesn't work on Win32. That is a PITA, but it wouldn't be the first time that a release of Apache didn't work the same on all platforms. > how about just taking the .31 tree, tagging *that* as > .32, and carefully bumping files that resolve issues into .32 and > the RM can carefully decide which other recent commits/changes are > worthy of inclusion in .32? I strongly disagree with doing this. Watching each commit to determine if it belongs in .32 is completely against our release model. Either the code keeps moving forward or it doesn't, let's not bastardize the release model. Ryan
Thoughts on 2.0.32
Since it sounds like Win32 is busted (I'd say it'd be fine to say that .31 is beta but doesn't work on Win32, but wrowe doesn't like that), how about just taking the .31 tree, tagging *that* as .32, and carefully bumping files that resolve issues into .32 and the RM can carefully decide which other recent commits/changes are worthy of inclusion in .32? I know that there have been lots of changes in .32-dev since .31 was tagged that have challenged the stability of the current tree. If we took .32 and made it as close to .31 as we could but with the *minimum* required changes to make .32, we could be close to a real beta. (This seems to mimic the APACHE_BEST tag I think was proposed before my time.) And, if the only substantive changes in .32 are related to Win32, we don't lose the testing that has already occurred on Unix platforms. Food for thought. FWIW, I'd volunteer to be RM, but I won't have the time until Wednesday at the earliest. -- justin
Re: 2.0.31 on Windows
William A. Rowe, Jr. wrote: >From: "Dwayne Miller" <[EMAIL PROTECTED]> >To: <[EMAIL PROTECTED]> >Sent: Monday, February 04, 2002 3:57 PM >Subject: Re: 2.0.31 on Windows > > >>I tracked down my problem with not being able to run Apache as a >>service. The ServerRoot param was not set correctly. It was /Apache2 >>instead of D:/Apache2. >>This configuration file error was never logged anywhere. Not only that, >>even with this setting, Apache would run fine from the command line. >> > >The 'service' mode has all sorts of peculiar behaviors. You've identified >exactly what I'm seeing today. > >Which Win32 OS are you using, exactly? > I'm on Windows 2000, build 2195, SP2 > >>But trying to start as a server, the ap_walk_config() routine, called in >>server/config.c returned the error and called exit(1). >> >>Somewhere, this error should be presented to the user. I'm just not >>sure where, or how to do it. >> > >Definately the biggest 'bug' of the hour is the service logging. It looks >like some changes a little ways back in main.c borked us. > >And the second bug is the lack of -d -f args in the service's Parameters >registry key ... ConfigArgs is coming up empty. > I was looking at mpm_winnt.c, where the apr_filepath_merge() call is made. It looks like it is trying to build the -d argument to use when starting the service, but it's obviously not having the desired effect. > >The final bug is the simple one. Why paths can't merge right in the service. >I suspect that's the whole LocalService/non-Network situation going on... >but it's worthless to debug without resolving the bigger error logging fish :) > >The only real answer is moving the service code into apr, pre-main() or in >tandem with apr_initialize() [starting with apr_app_main and getting the >service stuff tied in.] But I wanted to wait for .33 on those changes. > >Bill >
Re: Apache2 and php
My apologies, I realized that after I sent it. Â Actually that is incorrect. Â This seems to be a fairly unknown answer as nobody was able to answer it for me in the past on either the php or apache groups. http://marc.theaimsgroup.com/?l=php-dev&m=101180222102820&w=2 you must replace :LoadModule php4_module modules/libphp4.soAddType application/x-httpd-php .phpByLoadModule php4_module modules/libphp4.soSetOutputFilter PHPSetInputFilter PHP Thank you, David Daniel Lopez wrote: [EMAIL PROTECTED]"> This question should be directed to the [EMAIL PROTECTED] listYou need to add the following:AddType application/x-httpd-php .phpDanielOn Mon, Feb 04, 2002 at 05:20:25PM -0500, David Ford wrote: What does apache2 need in httpd.conf to properly shunt php files through the php engine before serving them?Apparently it needs more than just Load libphp4.soNote, I'm using php 4.1.1 with httpd_2_0 HEAD.Thank you,David
Re: cvs commit: httpd-2.0/modules/proxy proxy_ftp.c
On Sun, Feb 03, 2002 at 04:11:38PM -0800, Greg Stein wrote: > On Sun, Feb 03, 2002 at 05:18:19PM -, [EMAIL PROTECTED] wrote: > > jerenkrantz02/02/03 09:18:19 > > > > Modified:modules/proxy proxy_ftp.c > > Log: > > Make sure we include time.h if it is there. (Other mojo may be needed for > > other platforms.) > > Should that code even be using time.h? If it isn't there, then what happens? > Isn't that what APR is for? I wish, but Martin added a bunch of code that calls the non-portable time code directly. I don't have the time to figure out what he was trying to do and I just wanted a buildable tree. -- justin
--with-module annoyingness
Would anybody mind if --with-module created a symbolic link in the modules/foo/ directory to the specified module? Right now, we copy the module to that directory, which means that if you edit the .c file, it gets out of sync with the version in the original directory. If you have the .c file under revision control, that gets annoying quickly. If people would rather I didn't create a symlink, I could also just edit the Makefile to refer to the .c file in the other directory directly. Either approach is fine with me, but the copy is bothering me to no end. Ryan -- Ryan Bloom [EMAIL PROTECTED] 645 Howard St. [EMAIL PROTECTED] San Francisco, CA
Re: Apache2 and php
This question should be directed to the [EMAIL PROTECTED] list You need to add the following: AddType application/x-httpd-php .php Daniel On Mon, Feb 04, 2002 at 05:20:25PM -0500, David Ford wrote: > What does apache2 need in httpd.conf to properly shunt php files through > the php engine before serving them? > > Apparently it needs more than just Load libphp4.so > > Note, I'm using php 4.1.1 with httpd_2_0 HEAD. > > Thank you, > David > >
Re: 2.0.31 on Windows
From: "Dwayne Miller" <[EMAIL PROTECTED]> To: <[EMAIL PROTECTED]> Sent: Monday, February 04, 2002 3:57 PM Subject: Re: 2.0.31 on Windows > I tracked down my problem with not being able to run Apache as a > service. The ServerRoot param was not set correctly. It was /Apache2 > instead of D:/Apache2. > This configuration file error was never logged anywhere. Not only that, > even with this setting, Apache would run fine from the command line. The 'service' mode has all sorts of peculiar behaviors. You've identified exactly what I'm seeing today. Which Win32 OS are you using, exactly? > But trying to start as a server, the ap_walk_config() routine, called in > server/config.c returned the error and called exit(1). > > Somewhere, this error should be presented to the user. I'm just not > sure where, or how to do it. Definately the biggest 'bug' of the hour is the service logging. It looks like some changes a little ways back in main.c borked us. And the second bug is the lack of -d -f args in the service's Parameters registry key ... ConfigArgs is coming up empty. The final bug is the simple one. Why paths can't merge right in the service. I suspect that's the whole LocalService/non-Network situation going on... but it's worthless to debug without resolving the bigger error logging fish :) The only real answer is moving the service code into apr, pre-main() or in tandem with apr_initialize() [starting with apr_app_main and getting the service stuff tied in.] But I wanted to wait for .33 on those changes. Bill
Apache2 and php
What does apache2 need in httpd.conf to properly shunt php files through the php engine before serving them? Apparently it needs more than just Load libphp4.so Note, I'm using php 4.1.1 with httpd_2_0 HEAD. Thank you, David
Re: 2.0.31 on Windows
I tracked down my problem with not being able to run Apache as a service. The ServerRoot param was not set correctly. It was /Apache2 instead of D:/Apache2. This configuration file error was never logged anywhere. Not only that, even with this setting, Apache would run fine from the command line. But trying to start as a server, the ap_walk_config() routine, called in server/config.c returned the error and called exit(1). Somewhere, this error should be presented to the user. I'm just not sure where, or how to do it. Dwayne Miller wrote: > I'm not familiar with the -n apache2 option, but I was just trying to > figure out why I can't get apache to run as a service. I'm using the > -k start option, a message that it's starting the service appears, but > nothing happens. In the service.c file, I added a line to print the > 'Failed to start the service' message as I was not seeing this message > in any of my log files. It in fact fails to start the service. > > Bill Stoddard wrote: > >> Couple of problems noticed in a few minutes of testing >> >> 1. apache -k restart -n apache2 is broken. hangs forever and a new >> child process is never >> started. >> >> 2. If I kill off the child process (by attaching and detaching a >> debugger for instance), >> the parent should detect the child process exit and start a new child >> process. Instead, >> the parent goes belly up. >> >> Beta killers (for Windows at least). >> >> Bill >> > >
Re: more minor 2.0.31 breakage
Rodent of Unusual Size wrote: > > Greg Ames wrote: > > > > I just noticed that autoindex listings from apache.org no > > longer display the folder icon for directories. You see > > a '?' icon instead. The Rodent tells me there's some kind > > of magic type that mod_autoindex uses to figure this out. > > Not debugged yet. > > As an unsubstantiated WAG, I would guess that the subrequest > that's sent off to determine the content-type is running > afoul of mod_dir somehow, mod_dir is innocent this time. > and S_ISDIR isn't set correctly when it comes back. close...dirent->filetype isn't set on Unix when you do a plain ol' apr_dir_read. I believe you can ask for it to be set, but that causes extra stat()s. veto'ed and backed out the patch which broke this. Greg
Re: Apache 2_0_31 is now rolled (take 2)
Victor J. Orlikowski wrote: > > Ugh. Behind on mail. > > My fault; needed it for cadaver for testing... ;) > Why do I feel like I just entered an episode of 6 Feet Under? :) -- === Jim Jagielski [|] [EMAIL PROTECTED] [|] http://www.jaguNET.com/ "A society that will trade a little liberty for a little order will lose both and deserve neither" - T.Jefferson
Re: cvs commit: httpd-2.0 STATUS
From: "Aaron Bannert" <[EMAIL PROTECTED]> Sent: Monday, February 04, 2002 12:59 PM > > wrowe 02/02/04 10:51:59 > > > > Modified:.STATUS > > Log: > > Brrr... it's frigid in Chicago today! Hope FirstBill doesn't mind if > > I speak for us both here. > > > > Revision ChangesPath > > 1.453 +7 -5 httpd-2.0/STATUS > > > > + > > +* FirstBill reports problem [re]starting as-a-service, shared > > + score is suspect. OtherBill is investigating, definately > > + a showstopper for a .31/.32 beta. Progress report later today. > > This is a BETA, it is not perfect. The biggest benefit we get from > releasing a beta is _Beta Testers_. It seems to me that we are *very* > close to GA, but I will not be in support of a gold release without > significantly more beta visibility and testing than we've had. Aaron, it doesn't even friggin work [for the usual Win32 application] :) > This particular problem aparently has a workaround: Don't run it as a > service on win32. That's not an acceptable workaround - we need the BETA to test the robustness of the service. > Although it looks like we have a comfortable majority in favor of > releasing this as a beta, I'd feel better if there were no -1 votes. Can > I pursuade you to change your mind for the above reasons? Ahmmm... no, and your vote is invalid at the moment; I'm certain if you give folks 8 hours to comment; some will agree this is a showstopper. However, I expect to have this fixed and tested tonight - we should be able to move on from here fairly quickly. Critical faults include; 1. failure to log to the WinNT system event log. 2. failure to recognize the usual DocRoot/ServerRoot. It doesn't seem like we have any score problem, but I haven't gotten that far yet... Bill
new mod_proxy + multiple set-cookie
Just a heads up: There are multiple bug reports that claim that the new mod_proxy is messing up when multiple set-cookie headers are present, which prevents some sites like hotmail from working correctly. Anyone who is interested should check the bug database. Joshua.
RE: Apache 2_0_31 is now rolled (take 2)
My apologies. The context was to provide a semi-humorous reply to the breakage I caused Jeff by installing Expat on the AIX box we use. Been a while since the mail, been a while since I posted last. Grrr. ;) Victor -- Victor J. Orlikowski | The Wall is Down, But the Threat Remains! == [EMAIL PROTECTED] | [EMAIL PROTECTED] | [EMAIL PROTECTED]
RE: Apache 2_0_31 is now rolled (take 2)
> From: Victor J. Orlikowski [mailto:[EMAIL PROTECTED]] > Ugh. Behind on mail. > > My fault; needed it for cadaver for testing... ;) Victor: You need a little context on your mail. I don't know about others, but I have no idea what you are replying to. Joshua.
Re: Apache 2_0_31 is now rolled (take 2)
Ugh. Behind on mail. My fault; needed it for cadaver for testing... ;) Victor -- Victor J. Orlikowski | The Wall is Down, But the Threat Remains! == [EMAIL PROTECTED] | [EMAIL PROTECTED] | [EMAIL PROTECTED]
Re: more minor 2.0.31 breakage
Greg Ames wrote: > > I just noticed that autoindex listings from apache.org no > longer display the folder icon for directories. You see > a '?' icon instead. The Rodent tells me there's some kind > of magic type that mod_autoindex uses to figure this out. > Not debugged yet. As an unsubstantiated WAG, I would guess that the subrequest that's sent off to determine the content-type is running afoul of mod_dir somehow, and S_ISDIR isn't set correctly when it comes back. -- #kenP-)} Ken Coar, Sanagendamgagwedweinini http://Golux.Com/coar/ Author, developer, opinionist http://Apache-Server.Com/ "Millenium hand and shrimp!"
RE: cvs commit: httpd-2.0 STATUS
FWIW 2.0.31 compiles fine on HPUX also.. -Madhu -Original Message- From: Aaron Bannert [mailto:[EMAIL PROTECTED]] Sent: Monday, February 04, 2002 10:59 AM To: [EMAIL PROTECTED] Subject: Re: cvs commit: httpd-2.0 STATUS On Mon, Feb 04, 2002 at 06:51:59PM -, [EMAIL PROTECTED] wrote: > wrowe 02/02/04 10:51:59 > > Modified:.STATUS > Log: > Brrr... it's frigid in Chicago today! Hope FirstBill doesn't mind if > I speak for us both here. > > Revision ChangesPath > 1.453 +7 -5 httpd-2.0/STATUS > > Index: STATUS > === > RCS file: /home/cvs/httpd-2.0/STATUS,v > retrieving revision 1.452 > retrieving revision 1.453 > diff -u -r1.452 -r1.453 > --- STATUS 4 Feb 2002 18:41:45 - 1.452 > +++ STATUS 4 Feb 2002 18:51:59 - 1.453 > @@ -1,5 +1,5 @@ >APACHE 2.0 STATUS: -*-text-*- > -Last modified at [$Date: 2002/02/04 18:41:45 $] > +Last modified at [$Date: 2002/02/04 18:51:59 $] > >Release: > > @@ -45,9 +45,9 @@ >running on Daedalus since 02-Feb-2002 7:58 PST (need 3 days) >Compiles on : AIX 4.3, Solaris, FreeBSD 3.4 & 4.5, Win32, Linux 2.2 & 2.4 >for beta > -+1 : Justin, Ian, Jeff Trawick, BillS, BrianP, Aaron, Jim > ++1 : Justin, Ian, Jeff Trawick, BrianP, Aaron, Jim >+0 : Lars > --1 : > +-1 : BillS, BillR >bumps since original tag: >* mod-dir patch >* scoreboard x2 : 1 to fix gracefull restarts > @@ -77,8 +77,10 @@ >+1: Justin, Cliff, Aaron, BillS, Jim, trawick >+0: >-1: > -* hasn't run for 3 days on Daedalus > -* the release has underscores instead of periods. > + > +* FirstBill reports problem [re]starting as-a-service, shared > + score is suspect. OtherBill is investigating, definately > + a showstopper for a .31/.32 beta. Progress report later today. This is a BETA, it is not perfect. The biggest benefit we get from releasing a beta is _Beta Testers_. It seems to me that we are *very* close to GA, but I will not be in support of a gold release without significantly more beta visibility and testing than we've had. This particular problem aparently has a workaround: Don't run it as a service on win32. Although it looks like we have a comfortable majority in favor of releasing this as a beta, I'd feel better if there were no -1 votes. Can I pursuade you to change your mind for the above reasons? -aaron
[PATCH] mod_mime.c
In theme with the "broken module" being allowed in ap_directory_walk()
and ap_file_walk(), find_ct() of mod_mime.c needs the following change
to allow a missing filename to pass through to the handlers. I am pretty
sure that DECLINED is the correct return, but I don't think it matters.
Index: mod_mime.c
===
RCS file: /home/cvs/httpd-2.0/modules/http/mod_mime.c
retrieving revision 1.76
--- mod_mime.cSat Dec 8 02:02:51 2001
+++ mod_mime.cMon Feb 4 09:30:46 2002
@@ -736,6 +736,14 @@
const char *fn, *type, *charset = NULL;
int found_metadata = 0;
+/* To allow broken modules to proceed, we allow missing filenames to
pass.
+ * We will catch it later if it's heading for the core handler.
+ * directory_walk already posted an INFO note for module debugging.
+ */
+if (!r->filename) {
+return DECLINED;
+}
+
if (r->finfo.filetype == APR_DIR) {
r->content_type = DIR_MAGIC_TYPE;
return OK;
Kent Bruinsma
[EMAIL PROTECTED]
RE: 2.0.31 shutdown after heavy load
This was with the worker MPM.. I initially suspected the latency - 'waited for about 5 minutes, and then nothing happened till I issued a "kill -9" command. I'll try again today, and will probably post the stack trace of the parent process when it happens. Thanks -Madhu -Original Message- From: Greg Ames [mailto:[EMAIL PROTECTED]] Sent: Monday, February 04, 2002 9:58 AM To: [EMAIL PROTECTED] Subject: Re: 2.0.31 shutdown after heavy load "MATHIHALLI,MADHUSUDAN (HP-Cupertino,ex1)" wrote: > > Hi, > I'm getting the following message when I try to stop apache after a > stress-test on HPUX (using webstone).. Inspite of the SIGKILL message, the > process does not exit.. A second attempt is successful.. Any clues regarding > what may be happening is appreciated.. > > [error] child process 5891 still did not exit, sending a SIGKILL. which MPM? and have you tried this test with different results before? I didn't think we could do anything to block SIGKILL. Could this be just dispatching latency? Greg
Re: [PATCH] Caseless sorting for mod_autoindex
++1 for 2.0 [just saw such a complaint this week.]
-0 for 1.3
Forewarned that
> o abc
> o Abc
> o Bcd
> o bcd
> o xyz
> o Xyz
is equally possible with this patch, it doesn't sort by case at all.
When you look at 2.0, you will notice most of these cosmetics can be
controlled by the query string. A C=0/C=1[default] would be nice here :)
Bill
- Original Message -
From: "Rodent of Unusual Size" <[EMAIL PROTECTED]>
To: "Apache Developers" <[EMAIL PROTECTED]>
Sent: Monday, February 04, 2002 12:56 PM
Subject: [PATCH] Caseless sorting for mod_autoindex
> This got sent to me privately a long time ago, but got
> lost in the maze of twisty little passages that is my mail.
> It's against 1.3, but if approved I'll bring it forward to
> 2.0. I've reworked the original patch from Tullio Andreatta
> to apply against HEAD.
>
> The patch adds the IgnoreCase keyword to the IndexOptions directive.
> If active, directory listings are displayed case-insensitively.
> On systems with case-aware/case-sensitive filesystems, this would
> change
>
> o Abc
> o Bcd
> o Xyz
> o abc
> o bcd
> o xyz
>
> to
>
> o Abc
> o abc
> o Bcd
> o bcd
> o Xyz
> o xyz
>
> which can be a lot easier for humans to scan when looking for
> something specific.
>
> Index: src/modules/standard/mod_autoindex.c
> ===
> RCS file: /home/cvs/apache-1.3/src/modules/standard/mod_autoindex.c,v
> retrieving revision 1.121
> diff -u -r1.121 mod_autoindex.c
> --- src/modules/standard/mod_autoindex.c17 Nov 2001 03:27:09 - 1.121
> +++ src/modules/standard/mod_autoindex.c4 Feb 2002 18:08:56 -
> @@ -98,6 +98,7 @@
> #define NO_OPTIONS 256
> #define FOLDERS_FIRST 512
> #define TRACK_MODIFIED 1024
> +#define SORT_NOCASE 2048
>
> #define K_PAD 1
> #define K_NOPAD 0
> @@ -411,6 +412,9 @@
> else if (!strcasecmp(w, "TrackModified")) {
> option = TRACK_MODIFIED;
> }
> + else if (!strcasecmp(w, "IgnoreCase")) {
> +option = SORT_NOCASE;
> + }
> else if (!strcasecmp(w, "None")) {
> if (action != '\0') {
> return "Cannot combine '+' or '-' with 'None' keyword";
> @@ -732,6 +736,7 @@
> int ascending;
> int isdir;
> int checkdir;
> +int ignorecase;
> char key;
> };
>
> @@ -1222,6 +1227,7 @@
> * rather than CPU.
> */
> p->checkdir = ((d->opts & FOLDERS_FIRST) != 0);
> +p->ignorecase = ((d->opts & SORT_NOCASE) != 0);
> p->key = ap_toupper(keyid);
> p->ascending = (ap_toupper(direction) == D_ASCENDING);
>
> @@ -1592,7 +1598,14 @@
> }
> break;
> }
> -return strcmp(c1->name, c2->name);
> +
> +if (c1->ignorecase) {
> +result = strcasecmp(c1->name, c2->name);
> +}
> +else {
> +result = strcmp(c1->name, c2->name);
> +}
> +return result;
> }
>
>
> --
> #ken P-)}
>
> Ken Coar, Sanagendamgagwedweinini http://Golux.Com/coar/
> Author, developer, opinionist http://Apache-Server.Com/
>
> "Millenium hand and shrimp!"
>
Re: cvs commit: httpd-2.0 STATUS
On Mon, Feb 04, 2002 at 06:51:59PM -, [EMAIL PROTECTED] wrote: > wrowe 02/02/04 10:51:59 > > Modified:.STATUS > Log: > Brrr... it's frigid in Chicago today! Hope FirstBill doesn't mind if > I speak for us both here. > > Revision ChangesPath > 1.453 +7 -5 httpd-2.0/STATUS > > Index: STATUS > === > RCS file: /home/cvs/httpd-2.0/STATUS,v > retrieving revision 1.452 > retrieving revision 1.453 > diff -u -r1.452 -r1.453 > --- STATUS 4 Feb 2002 18:41:45 - 1.452 > +++ STATUS 4 Feb 2002 18:51:59 - 1.453 > @@ -1,5 +1,5 @@ >APACHE 2.0 STATUS: -*-text-*- > -Last modified at [$Date: 2002/02/04 18:41:45 $] > +Last modified at [$Date: 2002/02/04 18:51:59 $] > >Release: > > @@ -45,9 +45,9 @@ >running on Daedalus since 02-Feb-2002 7:58 PST (need 3 days) >Compiles on : AIX 4.3, Solaris, FreeBSD 3.4 & 4.5, Win32, Linux 2.2 & 2.4 >for beta > -+1 : Justin, Ian, Jeff Trawick, BillS, BrianP, Aaron, Jim > ++1 : Justin, Ian, Jeff Trawick, BrianP, Aaron, Jim >+0 : Lars > --1 : > +-1 : BillS, BillR >bumps since original tag: >* mod-dir patch >* scoreboard x2 : 1 to fix gracefull restarts > @@ -77,8 +77,10 @@ >+1: Justin, Cliff, Aaron, BillS, Jim, trawick >+0: >-1: > -* hasn't run for 3 days on Daedalus > -* the release has underscores instead of periods. > + > +* FirstBill reports problem [re]starting as-a-service, shared > + score is suspect. OtherBill is investigating, definately > + a showstopper for a .31/.32 beta. Progress report later today. This is a BETA, it is not perfect. The biggest benefit we get from releasing a beta is _Beta Testers_. It seems to me that we are *very* close to GA, but I will not be in support of a gold release without significantly more beta visibility and testing than we've had. This particular problem aparently has a workaround: Don't run it as a service on win32. Although it looks like we have a comfortable majority in favor of releasing this as a beta, I'd feel better if there were no -1 votes. Can I pursuade you to change your mind for the above reasons? -aaron
[PATCH] Caseless sorting for mod_autoindex
This got sent to me privately a long time ago, but got
lost in the maze of twisty little passages that is my mail.
It's against 1.3, but if approved I'll bring it forward to
2.0. I've reworked the original patch from Tullio Andreatta
to apply against HEAD.
The patch adds the IgnoreCase keyword to the IndexOptions directive.
If active, directory listings are displayed case-insensitively.
On systems with case-aware/case-sensitive filesystems, this would
change
o Abc
o Bcd
o Xyz
o abc
o bcd
o xyz
to
o Abc
o abc
o Bcd
o bcd
o Xyz
o xyz
which can be a lot easier for humans to scan when looking for
something specific.
Index: src/modules/standard/mod_autoindex.c
===
RCS file: /home/cvs/apache-1.3/src/modules/standard/mod_autoindex.c,v
retrieving revision 1.121
diff -u -r1.121 mod_autoindex.c
--- src/modules/standard/mod_autoindex.c17 Nov 2001 03:27:09 - 1.121
+++ src/modules/standard/mod_autoindex.c4 Feb 2002 18:08:56 -
@@ -98,6 +98,7 @@
#define NO_OPTIONS 256
#define FOLDERS_FIRST 512
#define TRACK_MODIFIED 1024
+#define SORT_NOCASE 2048
#define K_PAD 1
#define K_NOPAD 0
@@ -411,6 +412,9 @@
else if (!strcasecmp(w, "TrackModified")) {
option = TRACK_MODIFIED;
}
+ else if (!strcasecmp(w, "IgnoreCase")) {
+option = SORT_NOCASE;
+ }
else if (!strcasecmp(w, "None")) {
if (action != '\0') {
return "Cannot combine '+' or '-' with 'None' keyword";
@@ -732,6 +736,7 @@
int ascending;
int isdir;
int checkdir;
+int ignorecase;
char key;
};
@@ -1222,6 +1227,7 @@
* rather than CPU.
*/
p->checkdir = ((d->opts & FOLDERS_FIRST) != 0);
+p->ignorecase = ((d->opts & SORT_NOCASE) != 0);
p->key = ap_toupper(keyid);
p->ascending = (ap_toupper(direction) == D_ASCENDING);
@@ -1592,7 +1598,14 @@
}
break;
}
-return strcmp(c1->name, c2->name);
+
+if (c1->ignorecase) {
+result = strcasecmp(c1->name, c2->name);
+}
+else {
+result = strcmp(c1->name, c2->name);
+}
+return result;
}
--
#kenP-)}
Ken Coar, Sanagendamgagwedweinini http://Golux.Com/coar/
Author, developer, opinionist http://Apache-Server.Com/
"Millenium hand and shrimp!"
Re: Ongoing 2.0.31 and php hiccups
Aren't the APIs in apache HEAD significantly different preventing PHP
4.1 from being used? I gathered that you -must- use PHP42 HEAD.
Thank you,
David
MATHIHALLI,MADHUSUDAN (HP-Cupertino,ex1) wrote:
>I've been using PHP 4.1.1 and things seem to be okay for me (on HPUX).. I'm
>using PHP 4.1.1 with the following configure command :
>
>configure --prefix=/opt/apache2/modules \
>--enable-so --with-apache=/opt/apache2
>--with-apxs2=/opt/apache2/bin/apxs
>
>-Madhu
>
>-Original Message-
>From: David Ford [mailto:[EMAIL PROTECTED]]
>Sent: Sunday, February 03, 2002 2:48 PM
>To: [EMAIL PROTECTED]
>Subject: Ongoing 2.0.31 and php hiccups
>
>
>Ok, has -anyone- gotten php and apache running together the last week?
> If so, would you mind terribly if you posted the configure/make lines
>please? I'm getting kind of fidgity and needing to get it going :)
> Apache by itself seems fine.
>
>I've gotten down to a barebones ./configure
>--with-apxs2=/usr/local/apache2/bin/apxs for php and even that causes a
>segfault just after startup. Unfortunately I can't convince gdb to
>follow the child no matter what the settings so I can't post a bt just
>yet. Best I have is the following:
>
>strace -f httpd
>...
>open("/usr/local/apache2/conf/mime.types", O_RDONLY) = 6
>fstat64(6, {st_mode=S_IFREG|0644, st_size=12322, ...}) = 0
>read(6, "# This is a comment. I love comments.\n\n#...
>...
>read(6, "ovie\nx-conference/x-cooltalk\t\tice\n", 4096) = 34
>read(6, "", 4096) = 0
>close(6)= 0
>getpid()= 5906
>open("/tmp/aprMaKOvt", O_RDWR|O_CREAT|O_EXCL, 0600) = 6
>unlink("/tmp/aprMaKOvt")= 0
>getcwd("/", 4096) = 2
>--- SIGSEGV (Segmentation fault) ---
>
>ltrace ...
>...
>read(6, "", 4096) = 0
>pthread_mutex_unlock(0x08189a5c, 0x4000b870, 0, 0, 1) = 0
>close(6) = 0
>pthread_mutex_unlock(0x08189a5c, 0, 0xbfffd448, 0x400b3167, 0) = 0
>pthread_mutex_destroy(0x08189a5c, 0x4000b870, 6, 0, 0x081889e8) = 0
>strcmp("sapi_apache2.c", "mod_proxy.c") = 1
>strcmp("mod_so.c", "mod_proxy.c") = 1
>strcmp("mod_rewrite.c", "mod_proxy.c")= 1
>strcmp("mod_alias.c", "mod_proxy.c") = -1
>strcmp("mod_userdir.c", "mod_proxy.c")= 1
>strcmp("mod_speling.c", "mod_proxy.c")= 1
>strcmp("mod_actions.c", "mod_proxy.c")= -1
>strcmp("mod_imap.c", "mod_proxy.c") = -1
>strcmp("mod_dir.c", "mod_proxy.c")= -1
>strcmp("mod_negotiation.c", "mod_proxy.c")= -1
>strcmp("mod_vhost_alias.c", "mod_proxy.c")= 1
>strcmp("mod_cgi.c", "mod_proxy.c")= -1
>strcmp("mod_info.c", "mod_proxy.c") = -1
>strcmp("mod_asis.c", "mod_proxy.c") = -1
>strcmp("mod_autoindex.c", "mod_proxy.c") = -1
>strcmp("mod_status.c", "mod_proxy.c") = 1
>strcmp("mod_mime.c", "mod_proxy.c") = -1
>strcmp("http_core.c", "mod_proxy.c") = -1
>strcmp("prefork.c", "mod_proxy.c")= 1
>strcmp("mod_setenvif.c", "mod_proxy.c") = 1
>strcmp("mod_unique_id.c", "mod_proxy.c") = 1
>strcmp("mod_usertrack.c", "mod_proxy.c") = 1
>strcmp("mod_headers.c", "mod_proxy.c")= -1
>strcmp("mod_expires.c", "mod_proxy.c")= -1
>strcmp("mod_mime_magic.c", "mod_proxy.c") = -1
>strcmp("mod_env.c", "mod_proxy.c")= -1
>strcmp("mod_log_config.c", "mod_proxy.c") = -1
>strcmp("mod_include.c", "mod_proxy.c")= -1
>strcmp("mod_deflate.c", "mod_proxy.c")= -1
>strcmp("mod_case_filter_in.c", "mod_proxy.c") = -1
>strcmp("mod_case_filter.c", "mod_proxy.c")= -1
>strcmp("mod_file_cache.c", "mod_proxy.c") = -1
>strcmp("mod_auth_digest.c", "mod_proxy.c")= -1
>strcmp("mod_auth.c", "mod_proxy.c") = -1
>strcmp("mod_access.c", "mod_proxy.c") = -1
>strcmp("core.c", "mod_proxy.c") = -1
>memset(0x08151538, '\000', 80)= 0x08151538
>strlen(0x080ed7c4, 0x080f735c, 1793, 0x08054f5c, 0x400c65ec) = 14
>memcpy(0x08151588, "/tmp/aprXX", 15) = 0x08151588
>mkstemp(0x08151588, 0x40016a70, 1160, 0x080b3fd5, 0x08151588) = 6
>memset(0x08151598, '\000', 72)= 0x08151598
>strlen(0x08151588, 0x08151588, 0xb408, 0x080b8976, 0x081003f8) = 14
>memcpy(0x081515e0, "/tmp/aprVYqD12", 15) = 0x081515e0
>unlink("/tmp/aprVYqD12") = 0
>pthread_mutexattr_init(0xb458, 24, 0x080be520, 0x080c2810,
>0x08151588) = 0
>pthread_mutex_init(0x08151610, 0xb458, 0x080be520, 0x080c2810,
>0x08151588) = 0
>pthread_mutexattr_destroy(0xb458, 0xb458, 0x080be520,
>0x080c2810, 0x08151588) = 0
>--- SIGSEGV (Segmentation fault) ---
>+++ killed by SIGSEGV +++
>
># find -type f -exec grep "/
RE: cvs commit: httpd-2.0/server/mpm/worker worker.c
> Modified:.CHANGES STATUS >server/mpm/perchild perchild.c >server/mpm/prefork prefork.c >server/mpm/worker worker.c > Log: > Not being able to bind to a socket is a fatal error. This makes all > MPMs treat it as such. We now print a message to the console, and > return > a non-zero status code. That should have read "all Unix MPMs". If I can figure out the rest of the MPMs, I will port this change to them as well. Ryan
Re: cvs commit: apache-1.3/src Configure CHANGES
[EMAIL PROTECTED] writes: > martin 02/02/04 05:57:40 > > Modified:.configure >src Configure CHANGES > Log: > Fix the longstanding bug that errors (returned by src/Configure) > would not be noticed by the top level configure script. > That was bad for automated production environments, as errors would > go thru unnoticed, and caused havoc much later in the production. Thanks, folks! I just love it when the elves make a visit! [trawick@amadeus apache-1.3]$ CFLAGS=jkalsjflkajdsflkjasdf ./configure Configuring for Apache, Version 1.3.24-dev ... ** Apache requires an ANSI C Compiler, such as gcc. Error Output for sanity check cd ..; gcc -DSOLARIS2=280 -DUSE_EXPAT -I./lib/expat-lite -DNO_DL_NEEDED jkalsjflkajdsflkjasdf `./apaci` -o helpers/dummy helpers/dummy.c -lsocket -lnsl -lpthread gcc: jkalsjflkajdsflkjasdf: No such file or directory make: *** [dummy] Error 1 = End of Error Report = Aborting! [trawick@amadeus apache-1.3]$ echo $? 1<-cool! [trawick@amadeus apache-1.3]$
Re: cvs commit: httpd-2.0/server/mpm/worker worker.c
From: "Bill Stoddard" <[EMAIL PROTECTED]> Sent: Monday, February 04, 2002 12:01 PM > Actually, this patch is not the only (or even main) culprit. The switch to use the >shared > scoreboard is fatally broken. I am in the process of keeping the spirit of Ryan's >patch to > only use pre_mpm in the parent and revert back to non shared scoreboard for Windows. Agreed. But I think you are seeing a problem reported some time back, before the shared score was even implemented. Seems [perhaps] that some users can't run as a service with the bin/ pathing we implemented several months ago. Users report that moving everything from bin/ into the apache root works. I've never seen this, so I can't comment. But please do _not_ go reverting patches without reviewing them... would you kindly explain exactly _what_ is fatally broken? Bill
Re: 2.0.31 on Windows
From: "Bill Stoddard" <[EMAIL PROTECTED]> Sent: Monday, February 04, 2002 10:50 AM > Couple of problems noticed in a few minutes of testing > > 1. apache -k restart -n apache2 is broken. hangs forever and a new child process is >never > started. > > 2. If I kill off the child process (by attaching and detaching a debugger for >instance), > the parent should detect the child process exit and start a new child process. >Instead, > the parent goes belly up. > > Beta killers (for Windows at least). Yup. Can we have a bit of patience on .32, a day or two, so I can get these fixed? Bill - which version of Win32? Bill
RE: cvs commit: httpd-2.0/server/mpm/worker worker.c
Ahhh, I hadn't tried that yet. Ryan -- Ryan Bloom [EMAIL PROTECTED] 645 Howard St. [EMAIL PROTECTED] San Francisco, CA > -Original Message- > From: Bill Stoddard [mailto:[EMAIL PROTECTED]] > Sent: Monday, February 04, 2002 10:04 AM > To: [EMAIL PROTECTED]; [EMAIL PROTECTED] > Subject: Re: cvs commit: httpd-2.0/server/mpm/worker worker.c > > Start apache as a service. That seems to make a difference. > > Bill > > > I'm not having this problem. I just started Apache, and killed off the > > child process five times on my XP box. Each time, a new child was > > created to take its place. > > > > I stopped the child by opening the Task Manager, and killing the > > process. It took some time to make sure I wasn't killing the parent > > process, because for some reason, the pids don't just go in increasing > > order like you would expect. > > > > I would help debug this problem, but since I can't duplicate, I am kind > > of in the dark. The best I can suggest, is take a look at the > > information that is being set about the generation in the scoreboard. > > It is possible that some of the fields in the scoreboard aren't being > > initialized in the parent anymore. > > > > Ryan > > > > -- > > Ryan Bloom [EMAIL PROTECTED] > > 645 Howard St. [EMAIL PROTECTED] > > San Francisco, CA > > > > > -Original Message- > > > From: Bill Stoddard [mailto:[EMAIL PROTECTED]] > > > Sent: Monday, February 04, 2002 8:58 AM > > > To: [EMAIL PROTECTED] > > > Subject: Fw: cvs commit: httpd-2.0/server/mpm/worker worker.c > > > > > > This patch seems to have broken restarts and child recovery on > > Windows. > > > Investigating... > > > > > > Bill > > > >
Re: cvs commit: httpd-2.0/server/mpm/worker worker.c
On Mon, Feb 04, 2002 at 01:01:49PM -0500, Bill Stoddard wrote: > Actually, this patch is not the only (or even main) culprit. The switch to use the >shared > scoreboard is fatally broken. I am in the process of keeping the spirit of Ryan's >patch to > only use pre_mpm in the parent and revert back to non shared scoreboard for Windows. How is it fatally broken? -aaron
Re: cvs commit: httpd-2.0/server/mpm/worker worker.c
Start apache as a service. That seems to make a difference. Bill > I'm not having this problem. I just started Apache, and killed off the > child process five times on my XP box. Each time, a new child was > created to take its place. > > I stopped the child by opening the Task Manager, and killing the > process. It took some time to make sure I wasn't killing the parent > process, because for some reason, the pids don't just go in increasing > order like you would expect. > > I would help debug this problem, but since I can't duplicate, I am kind > of in the dark. The best I can suggest, is take a look at the > information that is being set about the generation in the scoreboard. > It is possible that some of the fields in the scoreboard aren't being > initialized in the parent anymore. > > Ryan > > -- > Ryan Bloom [EMAIL PROTECTED] > 645 Howard St. [EMAIL PROTECTED] > San Francisco, CA > > > -Original Message- > > From: Bill Stoddard [mailto:[EMAIL PROTECTED]] > > Sent: Monday, February 04, 2002 8:58 AM > > To: [EMAIL PROTECTED] > > Subject: Fw: cvs commit: httpd-2.0/server/mpm/worker worker.c > > > > This patch seems to have broken restarts and child recovery on > Windows. > > Investigating... > > > > Bill > >
Re: cvs commit: httpd-2.0/server/mpm/worker worker.c
Actually, this patch is not the only (or even main) culprit. The switch to use the
shared
scoreboard is fatally broken. I am in the process of keeping the spirit of Ryan's
patch to
only use pre_mpm in the parent and revert back to non shared scoreboard for Windows.
Bill
- Original Message -
From: "Bill Stoddard" <[EMAIL PROTECTED]>
To: <[EMAIL PROTECTED]>
Sent: Monday, February 04, 2002 11:57 AM
Subject: Fw: cvs commit: httpd-2.0/server/mpm/worker worker.c
> This patch seems to have broken restarts and child recovery on Windows.
Investigating...
>
> Bill
>
>
> > rbb 02/01/30 14:35:57
> >
> > Modified:include scoreboard.h
> >server scoreboard.c
> >server/mpm/prefork prefork.c
> >server/mpm/winnt mpm_winnt.c
> >server/mpm/worker worker.c
> > Log:
> > Change the Windows MPM to only use the pre_mpm phase in the parent process.
> > The child processes use the child_init phase to reattach to the shared
> > memory. This makes Windows work like Unix, which should make it easier
> > for module authors to write portable modules.
> >
> > Revision ChangesPath
> > 1.39 +3 -3 httpd-2.0/include/scoreboard.h
> >
> > Index: scoreboard.h
> > ===
> > RCS file: /home/cvs/httpd-2.0/include/scoreboard.h,v
> > retrieving revision 1.38
> > retrieving revision 1.39
> > diff -u -r1.38 -r1.39
> > --- scoreboard.h 28 Jan 2002 00:41:31 - 1.38
> > +++ scoreboard.h 30 Jan 2002 22:35:56 - 1.39
> > @@ -73,6 +73,7 @@
> >#include "apr_hooks.h"
> >#include "apr_thread_proc.h"
> >#include "apr_portable.h"
> > +#include "apr_shm.h"
> >
> >/* Scoreboard info on a process is, for now, kept very brief ---
> > * just status value and pid (the latter so that the caretaker process
> > @@ -113,8 +114,7 @@
> > */
> >typedef enum {
> >SB_NOT_SHARED = 1,
> > -SB_SHARED = 2, /* PARENT */
> > -SB_SHARED_CHILD = 3
> > +SB_SHARED = 2
> >} ap_scoreboard_e;
> >
> >#define SB_WORKING 0 /* The server is busy and the child is useful. */
> > @@ -185,7 +185,7 @@
> >AP_DECLARE(void) ap_increment_counts(ap_sb_handle_t *sbh, request_rec *r);
> >
> >int ap_create_scoreboard(apr_pool_t *p, ap_scoreboard_e t);
> > -apr_status_t reopen_scoreboard(apr_pool_t *p, int detached);
> > +apr_status_t ap_reopen_scoreboard(apr_pool_t *p, apr_shm_t **shm, int detached);
> >void ap_init_scoreboard(void *shared_score);
> >int ap_calc_scoreboard_size(void);
> >apr_status_t ap_cleanup_scoreboard(void *d);
> >
> >
> >
> > 1.53 +9 -17 httpd-2.0/server/scoreboard.c
> >
> > Index: scoreboard.c
> > ===
> > RCS file: /home/cvs/httpd-2.0/server/scoreboard.c,v
> > retrieving revision 1.52
> > retrieving revision 1.53
> > diff -u -r1.52 -r1.53
> > --- scoreboard.c 28 Jan 2002 00:41:31 - 1.52
> > +++ scoreboard.c 30 Jan 2002 22:35:56 - 1.53
> > @@ -143,7 +143,8 @@
> >{
> >char *more_storage;
> >int i;
> > -
> > +
> > + ap_calc_scoreboard_size();
> >ap_scoreboard_image =
> >calloc(1, sizeof(scoreboard) + server_limit * sizeof(worker_score *));
> >more_storage = shared_score;
> > @@ -200,7 +201,7 @@
> >/* If detach is non-zero, this is a seperate child process,
> > * if zero, it is a forked child.
> > */
> > -apr_status_t reopen_scoreboard(apr_pool_t *p, int detached)
> > +apr_status_t ap_reopen_scoreboard(apr_pool_t *p, apr_shm_t **shm, int detached)
> >{
> >#if APR_HAS_SHARED_MEMORY
> >if (!detached) {
> > @@ -215,6 +216,9 @@
> >}
> >/* everything will be cleared shortly */
> >#endif
> > +if (*shm) {
> > +*shm = ap_scoreboard_shm;
> > +}
> >return APR_SUCCESS;
> >}
> >
> > @@ -260,14 +264,6 @@
> >memset(sb_shared, 0, scoreboard_size);
> >ap_init_scoreboard(sb_shared);
> >}
> > -else if (sb_type == SB_SHARED_CHILD) {
> > -void *sb_shared;
> > -rv = reopen_scoreboard(p, 1);
> > -if (rv || !(sb_shared = apr_shm_baseaddr_get(ap_scoreboard_shm))) {
> > -return HTTP_INTERNAL_SERVER_ERROR;
> > -}
> > -ap_init_scoreboard(sb_shared);
> > -}
> >else
> >#endif
> >{
> > @@ -282,14 +278,10 @@
> >ap_init_scoreboard(sb_mem);
> >}
> >}
> > -/* can't just memset() */
> > -if (sb_type != SB_SHARED_CHILD) {
> > -ap_scoreboard_image->global->sb_type = sb_type;
> > -ap_scoreboard_image->global->running_generation = running_gen;
> > -apr_pool_cleanup_register(p, NULL, ap_cleanup_scoreboard,
> > -
Re: 2.0.31 shutdown after heavy load
"MATHIHALLI,MADHUSUDAN (HP-Cupertino,ex1)" wrote: > > Hi, > I'm getting the following message when I try to stop apache after a > stress-test on HPUX (using webstone).. Inspite of the SIGKILL message, the > process does not exit.. A second attempt is successful.. Any clues regarding > what may be happening is appreciated.. > > [error] child process 5891 still did not exit, sending a SIGKILL. which MPM? and have you tried this test with different results before? I didn't think we could do anything to block SIGKILL. Could this be just dispatching latency? Greg
more minor 2.0.31 breakage
I just noticed that autoindex listings from apache.org no longer display the folder icon for directories. You see a '?' icon instead. The Rodent tells me there's some kind of magic type that mod_autoindex uses to figure this out. Not debugged yet. Greg
RE: cvs commit: httpd-2.0/server/mpm/worker worker.c
I'm not having this problem. I just started Apache, and killed off the child process five times on my XP box. Each time, a new child was created to take its place. I stopped the child by opening the Task Manager, and killing the process. It took some time to make sure I wasn't killing the parent process, because for some reason, the pids don't just go in increasing order like you would expect. I would help debug this problem, but since I can't duplicate, I am kind of in the dark. The best I can suggest, is take a look at the information that is being set about the generation in the scoreboard. It is possible that some of the fields in the scoreboard aren't being initialized in the parent anymore. Ryan -- Ryan Bloom [EMAIL PROTECTED] 645 Howard St. [EMAIL PROTECTED] San Francisco, CA > -Original Message- > From: Bill Stoddard [mailto:[EMAIL PROTECTED]] > Sent: Monday, February 04, 2002 8:58 AM > To: [EMAIL PROTECTED] > Subject: Fw: cvs commit: httpd-2.0/server/mpm/worker worker.c > > This patch seems to have broken restarts and child recovery on Windows. > Investigating... > > Bill
Re: 2.0.31 on Windows
I'm not familiar with the -n apache2 option, but I was just trying to figure out why I can't get apache to run as a service. I'm using the -k start option, a message that it's starting the service appears, but nothing happens. In the service.c file, I added a line to print the 'Failed to start the service' message as I was not seeing this message in any of my log files. It in fact fails to start the service. Bill Stoddard wrote: >Couple of problems noticed in a few minutes of testing > >1. apache -k restart -n apache2 is broken. hangs forever and a new child process is >never >started. > >2. If I kill off the child process (by attaching and detaching a debugger for >instance), >the parent should detect the child process exit and start a new child process. >Instead, >the parent goes belly up. > >Beta killers (for Windows at least). > >Bill >
Fw: cvs commit: httpd-2.0/server/mpm/worker worker.c
This patch seems to have broken restarts and child recovery on Windows.
Investigating...
Bill
> rbb 02/01/30 14:35:57
>
> Modified:include scoreboard.h
>server scoreboard.c
>server/mpm/prefork prefork.c
>server/mpm/winnt mpm_winnt.c
>server/mpm/worker worker.c
> Log:
> Change the Windows MPM to only use the pre_mpm phase in the parent process.
> The child processes use the child_init phase to reattach to the shared
> memory. This makes Windows work like Unix, which should make it easier
> for module authors to write portable modules.
>
> Revision ChangesPath
> 1.39 +3 -3 httpd-2.0/include/scoreboard.h
>
> Index: scoreboard.h
> ===
> RCS file: /home/cvs/httpd-2.0/include/scoreboard.h,v
> retrieving revision 1.38
> retrieving revision 1.39
> diff -u -r1.38 -r1.39
> --- scoreboard.h 28 Jan 2002 00:41:31 - 1.38
> +++ scoreboard.h 30 Jan 2002 22:35:56 - 1.39
> @@ -73,6 +73,7 @@
>#include "apr_hooks.h"
>#include "apr_thread_proc.h"
>#include "apr_portable.h"
> +#include "apr_shm.h"
>
>/* Scoreboard info on a process is, for now, kept very brief ---
> * just status value and pid (the latter so that the caretaker process
> @@ -113,8 +114,7 @@
> */
>typedef enum {
>SB_NOT_SHARED = 1,
> -SB_SHARED = 2, /* PARENT */
> -SB_SHARED_CHILD = 3
> +SB_SHARED = 2
>} ap_scoreboard_e;
>
>#define SB_WORKING 0 /* The server is busy and the child is useful. */
> @@ -185,7 +185,7 @@
>AP_DECLARE(void) ap_increment_counts(ap_sb_handle_t *sbh, request_rec *r);
>
>int ap_create_scoreboard(apr_pool_t *p, ap_scoreboard_e t);
> -apr_status_t reopen_scoreboard(apr_pool_t *p, int detached);
> +apr_status_t ap_reopen_scoreboard(apr_pool_t *p, apr_shm_t **shm, int detached);
>void ap_init_scoreboard(void *shared_score);
>int ap_calc_scoreboard_size(void);
>apr_status_t ap_cleanup_scoreboard(void *d);
>
>
>
> 1.53 +9 -17 httpd-2.0/server/scoreboard.c
>
> Index: scoreboard.c
> ===
> RCS file: /home/cvs/httpd-2.0/server/scoreboard.c,v
> retrieving revision 1.52
> retrieving revision 1.53
> diff -u -r1.52 -r1.53
> --- scoreboard.c 28 Jan 2002 00:41:31 - 1.52
> +++ scoreboard.c 30 Jan 2002 22:35:56 - 1.53
> @@ -143,7 +143,8 @@
>{
>char *more_storage;
>int i;
> -
> +
> + ap_calc_scoreboard_size();
>ap_scoreboard_image =
>calloc(1, sizeof(scoreboard) + server_limit * sizeof(worker_score *));
>more_storage = shared_score;
> @@ -200,7 +201,7 @@
>/* If detach is non-zero, this is a seperate child process,
> * if zero, it is a forked child.
> */
> -apr_status_t reopen_scoreboard(apr_pool_t *p, int detached)
> +apr_status_t ap_reopen_scoreboard(apr_pool_t *p, apr_shm_t **shm, int detached)
>{
>#if APR_HAS_SHARED_MEMORY
>if (!detached) {
> @@ -215,6 +216,9 @@
>}
>/* everything will be cleared shortly */
>#endif
> +if (*shm) {
> +*shm = ap_scoreboard_shm;
> +}
>return APR_SUCCESS;
>}
>
> @@ -260,14 +264,6 @@
>memset(sb_shared, 0, scoreboard_size);
>ap_init_scoreboard(sb_shared);
>}
> -else if (sb_type == SB_SHARED_CHILD) {
> -void *sb_shared;
> -rv = reopen_scoreboard(p, 1);
> -if (rv || !(sb_shared = apr_shm_baseaddr_get(ap_scoreboard_shm))) {
> -return HTTP_INTERNAL_SERVER_ERROR;
> -}
> -ap_init_scoreboard(sb_shared);
> -}
>else
>#endif
>{
> @@ -282,14 +278,10 @@
>ap_init_scoreboard(sb_mem);
>}
>}
> -/* can't just memset() */
> -if (sb_type != SB_SHARED_CHILD) {
> -ap_scoreboard_image->global->sb_type = sb_type;
> -ap_scoreboard_image->global->running_generation = running_gen;
> -apr_pool_cleanup_register(p, NULL, ap_cleanup_scoreboard,
> - apr_pool_cleanup_null);
> -}
> +ap_scoreboard_image->global->sb_type = sb_type;
> +ap_scoreboard_image->global->running_generation = running_gen;
>ap_restart_time = apr_time_now();
> +apr_pool_cleanup_register(p, NULL, ap_cleanup_scoreboard,
>apr_pool_cleanup_null);
>return OK;
>}
>
>
>
>
> 1.236 +1 -1 httpd-2.0/server/mpm/prefork/prefork.c
>
> Index: prefork.c
> ===
> RCS file: /home/cvs/httpd-2.0/server/mpm/prefork/prefork.c,v
> retrieving revision 1.235
> retrieving revision 1.236
> diff -u -r1.235 -r1.236
> --- prefork.c 29 Jan 2002 22:31:25 - 1.
2.0.31 on Windows
Couple of problems noticed in a few minutes of testing 1. apache -k restart -n apache2 is broken. hangs forever and a new child process is never started. 2. If I kill off the child process (by attaching and detaching a debugger for instance), the parent should detect the child process exit and start a new child process. Instead, the parent goes belly up. Beta killers (for Windows at least). Bill
Re: Connection timed out: read_request_line() failed
The details... It's a ColdFusion script, so the processing module is mod_coldfusion.so. The script runs, it just doesn't see any of the form data. And I see the error logged with every post. I'm using SSL, so I'm also configured with mod_ssl. Interesting part is that it seems to be somehow related to which browser I use. Noticed the problem with Mozilla 0.9.6. Seems to work fine when client is IE. Same module, same POST worked with previous builds of Apache and SSL and all client browsers. Bill Stoddard wrote: >POSTs to CGIs look okay in my testing. Can you provide a bit more info about your >application? Is it a CGI? > >Bill > >>I'm getting this error in my log file when I post a form, but the code >>looks like it should be ignoring timeout errors. Something is >>definitely wrong with the handling of the request because it appears >>that all of the form data is missing. >> >>The full error message is... >>[Mon Feb 04 09:00:03 2002] [error] [client 127.0.0.1] (32560)Connection >>timed out: read_request_line() failed >> >>This is a build of the latest tarball (2.0.31) running on Win2000 system. >> >>Any thoughts on how to track down what is going on? >> >> >> >
Re: Connection timed out: read_request_line() failed
POSTs to CGIs look okay in my testing. Can you provide a bit more info about your application? Is it a CGI? Bill > I'm getting this error in my log file when I post a form, but the code > looks like it should be ignoring timeout errors. Something is > definitely wrong with the handling of the request because it appears > that all of the form data is missing. > > The full error message is... > [Mon Feb 04 09:00:03 2002] [error] [client 127.0.0.1] (32560)Connection > timed out: read_request_line() failed > > This is a build of the latest tarball (2.0.31) running on Win2000 system. > > Any thoughts on how to track down what is going on? > > >
Re: Apache 2_0_31 is now rolled
Ryan Bloom wrote: > > > I think we all agree that once a tarball > > is "public" (meaning that non-developers have the *potential* of > > grabbing it) if there's something wrong with it, we have to bump > > before we reroll. If it makes it to /www.apache.org/dist/httpd/, then I agree. If the only place it resides is intended strictly for developers, then I think that's too conservative, especially if it's just a roll problem. > I would vote in favor of having a developer only place to grab it then. +1 > The mail to ask people to test should go to the list of httpd > committers. That excludes people like Madhu who isn't a committer but has given us valuable feedback on our roll procedures. My vote would be something like Justin suggested, with a few tweaks: put the preliminary tarballs in a developer only directory, put appropriate disclaimers in a HEADER file (so you see them in an autoindex listing), and ask people to test on the dev@httpd list. Once we know the fate of the tarball, remove it from this dir so people don't get used to finding stuff there. We probably wouldn't want any links on our site to this directory either. If in spite of these precautions, any lurkers grab the preliminary tarballs and have problems: * they have been warned, * if it's a code problem, we are going to patch, re-tag, and re-roll soon anyway, * if it's a roll problem, doing ./buildconf will often fix (we could mention that in the HEADER), * I would think most responsible lurkers would de-cloak and speak up if they are the first to see a problem. If not, how much should I care about their problem? Greg
bugdb userdatabase
Problem here: [Mon Feb 04 07:21:15 2002] [error] [client 65.94.4.72] (2)No such file or directory: could not open dbm auth file: /home/apmail/bugdbaccounts I can't access the gnats bug database in "private" mode to edit bug reports. The database seems to be there. I'm guessing that this is a problem with the recently introduced AuthDBMType not being set correctly on the removal of mod_auth_db. Joshua
Connection timed out: read_request_line() failed
I'm getting this error in my log file when I post a form, but the code looks like it should be ignoring timeout errors. Something is definitely wrong with the handling of the request because it appears that all of the form data is missing. The full error message is... [Mon Feb 04 09:00:03 2002] [error] [client 127.0.0.1] (32560)Connection timed out: read_request_line() failed This is a build of the latest tarball (2.0.31) running on Win2000 system. Any thoughts on how to track down what is going on?
Re: WXP
From: "Stipe Tolj" <[EMAIL PROTECTED]> Sent: Monday, February 04, 2002 8:24 AM > Ben Hyde wrote: > > > > Rodent of Unusual Size wrote: > > > There's a FAQ I'm getting more and more to which I don't > > > know the answer; to wit, does Apache 1.3.* run on Windows XP? > > > What *is* the answer? > > > > "Not known not to work." > > at least our Cygwin 1.3 based port of 1.3.x works on Windows XP. So this > is something like a "yeno". Stipe, how many boxes had you tested yet, before you make that assertion? About one dozen XP users (+ those who haven't checked in with us through either bugs or c.i.s.w.ms-windows) have a very serious sockets bug that results in corrupted -script- output (e.g. SHTML, PGP, etc.) Cygwin now essentially shares sockets code, so I'd consider your statement suspect. FirstBill and I have banged our heads together and come up with nothing useful yet. Nobody in Bill's or my own work groups can reproduced the bug. It's probably with a specific transport provider, but which one nobody can isolate yet. So, does it work on Cygwin? About as well as XP - in that there is a buggy driver out there to bite 'em in the arse. And no, turning of QoS or other driver extensions does not clear the bug, and no, you won't see it on any localhost loopback requests. The mystery... nobody [on the very same boxes] with Apache 2.0.28 can reproduce =-/ Bill
RE: WXP
I had 1.3.22 working on one of my XP boxes a few months ago. It is a real pity that I now have multiple XP machines, and just one lonely Linux VM. :-( Ryan -- Ryan Bloom [EMAIL PROTECTED] 645 Howard St. [EMAIL PROTECTED] San Francisco, CA > -Original Message- > From: Stipe Tolj [mailto:[EMAIL PROTECTED]] > Sent: Monday, February 04, 2002 6:24 AM > To: [EMAIL PROTECTED] > Subject: Re: WXP > > Ben Hyde wrote: > > > > Rodent of Unusual Size wrote: > > > There's a FAQ I'm getting more and more to which I don't > > > know the answer; to wit, does Apache 1.3.* run on Windows XP? > > > What *is* the answer? > > > > "Not known not to work." > > at least our Cygwin 1.3 based port of 1.3.x works on Windows XP. So this > is something like a "yeno". > > Stipe
Re: WXP
Ben Hyde wrote: > > Rodent of Unusual Size wrote: > > There's a FAQ I'm getting more and more to which I don't > > know the answer; to wit, does Apache 1.3.* run on Windows XP? > > What *is* the answer? > > "Not known not to work." at least our Cygwin 1.3 based port of 1.3.x works on Windows XP. So this is something like a "yeno". Stipe
Re: WXP
Rodent of Unusual Size wrote: > There's a FAQ I'm getting more and more to which I don't > know the answer; to wit, does Apache 1.3.* run on Windows XP? > What *is* the answer? "Not known not to work."
Re: pre-compiled binaries of apache for NextStep or OpenStep
Rodent of Unusual Size wrote: > Not acked. Is this a PMC decision? ... > From: Chris Swenson <[EMAIL PROTECTED]> ... > > I noticed you no longer have pre-compiled binaries of Apache for NextStep > or OpenStep available on the Apache website. If a group of people (myself > included) volunteered to compile the binaries of each Apache release for > these OSes, would you be interested in distributing them on your website? > > Thanks. > Chris Swenson ... The policy: We, the HTTP project, distribute precompiled binaries only from project committers or ASF members. Trojan horse risks are one reason for this paranoia, but even small tinkered with the as shipped build can create a problems. - ben
Re: cvs commit: httpd-2.0/modules/experimental mod_disk_cache.c
On Mon, Feb 04, 2002 at 01:06:45PM -, [EMAIL PROTECTED] wrote:
>static const command_rec disk_cache_cmds[] =
>{
> -AP_INIT_FLAG("CacheRoot", set_cache_root, NULL, RSRC_CONF,
> +AP_INIT_FLAG1("CacheRoot", set_cache_root, NULL, RSRC_CONF,
> "The directory to store cache files"),
Uhm? Isn't that AP_INIT_TAKE1 ?
Martin
--
<[EMAIL PROTECTED]> | Fujitsu Siemens
Fon: +49-89-636-46021, FAX: +49-89-636-47655 | 81730 Munich, Germany
Re: I'd like to release 1.3.24...
>Here's a patch to bail out on ./Configure errors. Sorry about the >&3 >redirection, but without it, we would have to create an intermediate >file (and need to prevent race conditions, need to clean up & all that). Let me see look into this... I think there's a clearer way... >Also, I added a 2>&1 redirection to Configure's sanity check report >(without it, the "echo ===" output will not always be where the >error output is). > >Should I commit? +1 on this part -- === Jim Jagielski [|] [EMAIL PROTECTED] [|] http://www.jaguNET.com/ "A society that will trade a little liberty for a little order will lose both and deserve neither" - T.Jefferson
Re: Apache 2_0_31 is now rolled
> > From: Jim Jagielski [mailto:[EMAIL PROTECTED]] > > > > Not sure what this "Roy himself" comment means... like it's some sort > > of Voice From On High. When did the opinion of any one person, no > > matter who, become Gospel? > Just to be clear, the above wasn't, in any way, a slam on Roy. If Ryan had said "Bill" or "Greg" or "Jim" or "Zipperhead" I would have responded the same way :) -- === Jim Jagielski [|] [EMAIL PROTECTED] [|] http://www.jaguNET.com/ "A society that will trade a little liberty for a little order will lose both and deserve neither" - T.Jefferson
Re: [PATCH] SSL_* in suexec safe env list
On Sun, Feb 03, 2002 at 12:37:58PM -0500, Joshua Slive wrote:
>
> > From: Zvi Har'El [mailto:[EMAIL PROTECTED]]
>
> > triple rather then
> > double strncmp:
> >
> > -if (!strncmp(*ep, "HTTP_", 5)) {
> > +if (!strncmp(*ep, "HTTP_", 5) || !strncmp(*ep, "HTTPS", 5) ||
(no, not like that. not strncmp, but strcmp(*ep, "HTTPS")
> > + !strncmp(*ep, "SSL_", 4)) {
>
> As I said, I'm not going to commit without two "+1"s.
>
>
> Index: suexec.c
> ===
> RCS file: /home/cvs/httpd-2.0/support/suexec.c,v
> retrieving revision 1.17
> diff -u -d -b -r1.17 suexec.c
> --- suexec.c22 Nov 2001 07:42:13 - 1.17
> +++ suexec.c3 Feb 2002 17:30:13 -
> @@ -136,6 +136,7 @@
> "DOCUMENT_URI",
> "FILEPATH_INFO",
> "GATEWAY_INTERFACE",
> +"HTTPS",
> "LAST_MODIFIED",
> "PATH_INFO",
> "PATH_TRANSLATED",
Yep, that's cleaner. +1.
Martin
--
<[EMAIL PROTECTED]> | Fujitsu Siemens
Fon: +49-89-636-46021, FAX: +49-89-636-47655 | 81730 Munich, Germany
[PATCH] Re: I'd like to release 1.3.24...
On Mon, Feb 04, 2002 at 01:58:25PM +0100, Kraemer, Martin wrote: > Here's a patch to bail out on ./Configure errors. Sorry about the >&3 > redirection, but without it, we would have to create an intermediate > file (and need to prevent race conditions, need to clean up & all that). Only I forgot to add [PATCH] to the subject line. For the patch, see the previous mail. Martin -- <[EMAIL PROTECTED]> | Fujitsu Siemens Fon: +49-89-636-46021, FAX: +49-89-636-47655 | 81730 Munich, Germany
Re: I'd like to release 1.3.24...
Here's a patch to bail out on ./Configure errors. Sorry about the >&3
redirection, but without it, we would have to create an intermediate
file (and need to prevent race conditions, need to clean up & all that).
Also, I added a 2>&1 redirection to Configure's sanity check report
(without it, the "echo ===" output will not always be where the
error output is).
Should I commit?
Martin
--
<[EMAIL PROTECTED]> | Fujitsu Siemens
Fon: +49-89-636-46021, FAX: +49-89-636-47655 | 81730 Munich, Germany
Index: configure
===
RCS file: /home/cvs/apache-1.3/configure,v
retrieving revision 1.133
diff -u -r1.133 configure
--- configure 27 Sep 2001 18:12:03 - 1.133
+++ configure 4 Feb 2002 12:54:24 -
@@ -1582,14 +1582,15 @@
if [ "x$verbose" = "xyes" ]; then
vflag="-v";
fi
-if [ "x$quiet" = "xyes" ]; then
-(cd $src; ./Configure ${vflag} -file Configuration.apaci >/dev/null);
+rc=`if [ "x$quiet" = "xyes" ]; then
+(cd $src; ./Configure ${vflag} -file Configuration.apaci >/dev/null; echo $? >&3;
+);
else
-(cd $src; ./Configure ${vflag} -file Configuration.apaci |\
+(cd $src; ( ./Configure ${vflag} -file Configuration.apaci; echo $? >&3; ) |\
sed -e '/^Using config file:.*/d' \
-e "s:Makefile in :Makefile in $src\\/:" \
-e "s:Makefile\$:Makefile in $src:")
-fi
+fi 3>&1 1>&2`
+[ $rc = 0 ] || exit 1
##
## final hints
Index: src/Configure
===
RCS file: /home/cvs/apache-1.3/src/Configure,v
retrieving revision 1.449
diff -u -r1.449 Configure
--- src/Configure 1 Feb 2002 02:27:04 - 1.449
+++ src/Configure 4 Feb 2002 12:54:25 -
@@ -2398,7 +2398,7 @@
echo "** Apache requires an ANSI C Compiler, such as gcc. "
echo ""
echo " Error Output for sanity check "
- (./helpers/TestCompile -v sanity)
+ (./helpers/TestCompile -v sanity) 2>&1
echo "= End of Error Report ="
echo ""
echo " Aborting!"
Re: I'd like to release 1.3.24...
On Sat, Feb 02, 2002 at 04:34:14PM -0500, Rodent of Unusual Size wrote:
>
> if [ "x$quiet" = "xyes" ]; then
> (cd $src; \
> ./Configure ${vflag} -file Configuration.apaci >/dev/null || \
> exit $?);
> else
Nope -- this will do the same as it did: terminate the subshell
with an error (which is ignored).
You could do...
if [ "x$quiet" = "xyes" ]; then
(cd $src; ./Configure ${vflag} -file Configuration.apaci >/dev/null) || exit;
-either-
else
(cd $src; ./Configure ${vflag} -file Configuration.apaci |\
sed -e '/^Using config file:.*/d' \
-e "s:Makefile in :Makefile in $src\\/:" \
-e "s:Makefile\$:Makefile in $src:")
fi || exit
or-^^^
But the "else" branch is more tricky, as you said.
Martin
--
<[EMAIL PROTECTED]> | Fujitsu Siemens
Fon: +49-89-636-46021, FAX: +49-89-636-47655 | 81730 Munich, Germany
