From Bugzilla Helper: User-Agent: Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.7.5) Gecko/20041111 Firefox/1.0 Description of problem: When trying to start httpd after updating the system w/ yum on 1 Feb. 2005, httpd startup process fails complaining that it cannot start due to an undefined symbol in a curl lib. [root@localhost log]# /sbin/service httpd start Starting httpd: Syntax error on line 6 of /etc/httpd/conf.d/php.conf: Cannot load /etc/httpd/modules/libphp4.so into server: /usr/lib/libcurl.so.3: undefined symbol: tld_strerror The version of curl installed: curl-7.12.3-2 curl-devel-7.12.3-2 Version-Release number of selected component (if applicable): httpd-2.0.52-3.1 How reproducible: Always Steps to Reproduce: 1. Try to start httpd w/ any method: service, /etc/init.d, /usr/sbin/httpd 2. Get syntax error on undefined symbol 3. Actual Results: Received error stating that libphp4.so could not be loaded due to an undefined symbol, tld_strerror, in /usr/lib/libcurl.so.3. Expected Results: httpd should've started Additional info: httpd worked until updating curl and curl-devel today.
Works fine for me. Did you --nodeps an upgrade? Was this system upgraded from Fedora Core 2? What's the output of: # ldd /usr/lib/libcurl.so.3 # rpm -q libidn # ls -l /usr/lib/libidn.so.* # ldd /usr/lib/libidn.so.*
Created attachment 110547 [details] libcurl, libidn debug output
This was a fresh install of Fedora Core 3. I did the update of curl and curl-devel w/ 'yum -y update curl curl-devel'. I noticed on a couple of other FC3 systems that I cannot recreate the bug. They all have the distinction of not having been rebooted. I'm not sure if that makes a difference, but it's the one main difference I can think of between those systems and the problem system.
I saw that libidn and libidn-devel were not the same version as the other systems. Not sure how that happened since I only use yum to update them and they're all pulling from the same repository. Anywho, I updated libidn and libidn-devel using rpm instead of yum and all seems to work now. I'll mark this as 'Not a bug' and note that the problem is between the keyboard and chair.