Re: Trying to update gd

["Peter Müller" <peter.mueller@ipfire.org>]

[-- Attachment #1: Type: text/plain, Size: 1934 bytes --]

Hello Adolf,

reading this thread, I just stumbled across the part below, so I thought I slip into the conversation. :-)

> With memory.cgi I get an Internal Server Error. The log gives the following:-
> 
> Bareword "collectd" not allowed while "strict subs" in use at /srv/web/ipfire/cgi-bin/memory.cgi line 59.
> Bareword "localhost" not allowed while "strict subs" in use at /srv/web/ipfire/cgi-bin/memory.cgi line 59.
> Bareword "swap" not allowed while "strict subs" in use at /srv/web/ipfire/cgi-bin/memory.cgi line 59.
> Execution of /srv/web/ipfire/cgi-bin/memory.cgi aborted due to compilation errors.
> 
> Looking at line 59 and checking in git this looks to be related to a change done by Stefan Schantl 8 days ago
> https://git.ipfire.org/?p=ipfire-2.x.git;a=commitdiff;h=4353c579248b3016740252009b4ed15a4c6f8fe3
> 
> I don't want to commit the patch for gd until I have memory.cgi confirmed working and enough firewall data to confirm the pie charts are working but it looks close to completion.

This is because of bug #12616, which required massive changes to nearly all of the CGIs. Stefan and Michael
worked on this one, and their changes was merged into the "next" branch a few days ago.

Therefore, copying CGIs from "next" onto an already installed IPFire running Core Update 156 is not suitable
for testing, as C156 lacks additional Perl methods introduced by Stefan.

Unless your changes to a CGI hit a part of it that has been modified, you can still generate a patch of your
changes, and apply that on an existing IPFire installation. This would look like:

(on your workstation)

> git diff memory.cgi > memory-cgi-fixes.patch
> scp memory-cgi-fixes.patch [FQDN or IP address of your IPFire]:/tmp/.

(on the IPFire machine)

> cd /srv/web/ipfire/cgi-bin/
> patch < /tmp/memory-cgi-fixes.patch

Good luck, and happy patching. :-)

Thanks, and best regards,
Peter Müller