ID: 24567 User updated by: php at grooky dot com Reported By: php at grooky dot com -Status: Feedback +Status: Open Bug Type: Apache2 related Operating System: RedHat 9 PHP Version: 4.3.3RC1 New Comment:
No differences with php4-STABLE-200307100730 and disable-all --disable-cli In fact, this bug migh be a dup of #17414. Similar backtrace and exactly same problem. I have the bug on RedHat, but #17414 was found on debian also. According to the bt, it really seems to be PHP's fault. Previous Comments: ------------------------------------------------------------------------ [2003-07-09 16:50:01] [EMAIL PROTECTED] And try it with this configure script: ./configure --with-apxs2=/tmp/apache2/bin/apxs --disable-all --disable-cli --without-pear ------------------------------------------------------------------------ [2003-07-09 16:48:41] [EMAIL PROTECTED] Please try using this CVS snapshot: http://snaps.php.net/php4-STABLE-latest.tar.gz For Windows: http://snaps.php.net/win32/php4-win32-STABLE-latest.zip This doesn't sound like something we could really fix in PHP but do try the latest CVS just in case. ------------------------------------------------------------------------ [2003-07-09 16:30:38] php at grooky dot com The segfault comes out with any kernel I build (whom by the way works fine with any other apps, including apache without PHP) I wasn't able to reproduce it on my standard RH9 kernel (2.4.20-8) ------------------------------------------------------------------------ [2003-07-09 16:10:48] [EMAIL PROTECTED] I can't reproduce this (I have kernel 2.4.19 SMP). What is the version of the stock RH kernel this works with? ------------------------------------------------------------------------ [2003-07-09 15:34:47] php at grooky dot com Description: ------------ Apache 2 root process core when it received a signal (for example HUP or USR1 at logrotation) when php module loaded. Bug is reproductible at least with Apache 2.0.44 .45 .46 or .47 and with php 4.3.1, 4.3.2, 4.3.3RC1, CVS-STABLE. Bug is reproductible with linux 2.4.20 or 2.4.21 but NOT with the standard RedHat 9 kernel. To reproduce the segfault : kill -HUP "pid of apache root process" apachectl graceful Error log show : "[notice] seg fault or similar nasty error detected in the parent process" Reproduce code: --------------- php configure : ./configure --enable-debug --with-mysql --with-ldap --with-gettext --enable-force-cgi-redirect --with-apxs2=/tmp/apache2/bin/apxs --enable-track-vars --with-config-file-path=/etc standard php.ini As apache core only with any custom kernel I tried (but not with redHat one), here is my kernel configuration : My kernel configuration : http://www.grooky.com/linux.config I found few people on newgroup's archive with exaclty the same segfault. None got solution. Expected result: ---------------- Expected result : apache should reload it's configration Actual result: -------------- (gdb) bt #0 0x26a5cc20 in sapi_initialize_empty_request (tsrm_ls=0x8b776452) at /usr/src/php-4.3.3RC1/main/SAPI.c:424 #1 0x26a57210 in php_module_startup (sf=0x26ad5e20, additional_modules=0x26ad6100, num_additional_modules=1) at /usr/src/php-4.3.3RC1/main/main.c:1094 #2 0x26a93fdb in php_apache2_startup (sapi_module=0x26ad5e20) at /usr/src/php-4.3.3RC1/sapi/apache2handler/sapi_apache2.c:297 #3 0x26a9410b in php_apache_server_startup (pconf=0x80da508, plog=0x81165f8, ptemp=0x811d810, s=0x26ad5e20) at /usr/src/php-4.3.3RC1/sapi/apache2handler/sapi_apache2.c:396 #4 0x08094240 in ap_run_post_config (pconf=0x80da508, plog=0x81165f8, ptemp=0x811d810, s=0x80de450) at config.c:131 #5 0x080986e2 in main (argc=3, argv=0x5db89cd4) at main.c:649 #6 0x267dba07 in __libc_start_main () from /lib/i686/libc.so.6 ------------------------------------------------------------------------ -- Edit this bug report at http://bugs.php.net/?id=24567&edit=1
