[VDPAU] [PATCH] Detect libflashplayer through /proc/self/maps

[Jeroen Roovers]

On Tue, 19 Nov 2013 15:41:11 -0800
Aaron Plattner <aplattner at nvidia.com> wrote:

> On 11/12/13 15:44, Jeroen Roovers wrote:
> > Parsing /proc/self/cmdline does not always detect libflashplayer
> > since not every browser includes that string. /proc/self/maps will
> > always map the library we are looking for, so we use that instead.
> 
> Thanks for the patch, Jeroen.
> 
> Which browser, specifically, does this patch fix?  All I can glean
> from the discussion above is that this doesn't break Firefox.

I previously tried with any Linux-supported web browser that I could
find. However, the patch was based on a Gentoo bug report[1] about the
Opera web browser[2].

Seamonkey names libflashplayer in argv:
/usr/lib64/seamonkey/plugin-container /usr/lib64/nsbrowser/plugins/libflashplayer.so
-greomni /usr/lib64/seamonkey/omni.ja -appdir /usr/lib64/seamonkey
10321 plugin

As does Firefox:
/opt/firefox/plugin-container /usr/lib64/nsbrowser/plugins/libflashplayer.so
-greomni /opt/firefox/omni.ja -appomni /opt/firefox/browser/omni.ja
-appdir /opt/firefox/browser 10479 true plugin

Whilst Opera (probably) does some IPC here after forking:
/usr/lib64/opera/pluginwrapper/operapluginwrapper-native -multiproc
2,0.1.2201799,51,85,91 -logfolder /home/jeroen/.opera/logs/ -type native

I'll test some more with WebKitGTK+/qtwebkit based browsers. Are there
any others I should be thinking of?


Regards,
     jer


[1] https://bugs.gentoo.org/show_bug.cgi?id=473084
[2] http://www.opera.com/