This behavior masksĪpple's bug - when NPN_ForceRedraw() gets called on a destroyed Re-used to instantiate another plugin instance. Instead it gets "cached", and potentially NsNPAPIPluginInstance object) doesn't get deleted when the plugin When plugin caching is on, a plugin instance (its Which (prior to bug 670079's removal of support for it) turns on In any case, Apple's current Java updates and the previous versionsĪll call NPN_SetValue() to turn on NPPVpluginKeepLibraryInMemory, (I can't find anything in the plugin-futures mailing list archives),Īnd this variable remains documented (and undeprecated) on MDN But we apparently didn't announce this anywhere NPPVpluginKeepLibraryInMemory variable (set by the plugin with a call
FIREFOX MAC OS X 10.6.8 PATCH
The patch for bug 670079 dropped support for the Previous versions of Apple's Java plugin never call NPN_ForceRedraw() The deleted nsNPAPIPluginInstance object. (mozilla::plugins::parent::_forceredraw()) dereferences a pointer to The crashes happen when NPN_ForceRedraw() Updates can call NPN_ForceRedraw() on a plugin instance after theīrowser has called NPP_Destroy() on the plugin instance, and theīrowser has deleted the nsNPAPIPluginInstance object corresponding to On both OS X 10.7 and 10.6, the Java plugin from Apple's latest Java Splinter Review Turns out this is Apple's bug: But I'm confident I'llįind a reasonable fix (or workaround) today or tomorrow. Smallest change that will "fix" this bug's crashes.Įven then I might not have found the cure. I'll be playing with various changes to that patch, to find the
FIREFOX MAC OS X 10.6.8 TRIAL
When you do that you find this bug's crashes start with the followingĪnd I've already found (by trial and error) that the patch for bugĦ70069 is the one (in that regression range) that triggered them. But youĬan programmatically turn off jemalloc by setting the NO_MAC_JEMALLOCĮnvironment variable (I set it to "1"). Problems with jemalloc, those builds are extremely crashy. Testing the mozilla-central nightlies from (where the patchįor bug 414946 was landed) to (just before the patch forīug 414946 was backed out) is difficult on OS X 10.7 - thanks to mozilla-central nightly seems to "fix" the crashes (atĪnd I also found that disabling jemalloc (backing out the patch forīug 414946) in the mozilla-central nightly seems to haveīut, like I also said, we don't see these crashes in FF 7.0.1 on OS Xġ0.7, and FF 7.0.1 doesn't have jemalloc. Yes, enabling jemalloc (landing the patch for bug 694896) in the The whole jemalloc business seems to have been a red herring.
Triggered these crashes (at least on OS X 10.7):īug 670079: Stop caching plugin instances. I'm making progress - I've now found the patch that actually
0 kommentar(er)
