Seems that 10.5.1 fixed this runaway memory growth problem. I can't
reproduce it once this update is applied.
On Nov 14, 2007 11:07 PM, Eric Gouriou <email@hidden> wrote:
>
> On Nov 14, 2007, at 7:21 PM, David Walthour wrote:
> > Based upon the output of heap (see below) that it looks like mostly
> > NSCFString and NSCFData objects. These are the objects that increment
> > in count between successive runs of rdesktop.
>
> Would it be possible for you to file a Radar at <http://bugreporter.apple.com/
> > with that info ?
>
> Greg's request is right on the money (not surprisingly)
> ------
> * `heap quartz-wm`.
> * `vmmap quartz-wm`.
> * Quartz Debug (in /Developer/Applications/Graphics Tools). First
> "Show Window List", then look for quartz-wm windows that are created
> and never destroyed every cycle.
> ------
>
> If you want to go the extra mile, getting the associated stack
> traces would
> probably be ideal to save time for Ben:
>
> 1) Find a way to launch quartz-wm with the environment variable
> MallocStackLogging=1
>
> Given the launchd approach of launching X11, this will likely
> involve
> adding the appropriate key/value to the corresponding launch plist,
> /System/Library/LaunchAgents/org.x.X11.plist
> ---
> <key>EnvironmentVariables</key>
> <dict>
> <key>MallocStackLogging</key> <string>1</string>
> </dict>
> ---
> 2) Collect a heap and leak output
> 3) Run 10 cycles of rdesktop start/exit
> 4) Collect heap and leak outputs anew
>
> Given this data, the unfreed allocations should jump out with their
> associated stack trace.
>
> Eric
>
>
_______________________________________________
Do not post admin requests to the list. They will be ignored.
X11-users mailing list (email@hidden)
Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/x11-users/email@hidden
This email sent to email@hidden
