Re: NSCursor mismatched push and pop

[David Chisnall]

On 4 Oct 2013, at 19:51, Germán Arias <address@hidden> wrote:

> El jue, 03-10-2013 a las 18:07 +0100, David Chisnall escribió:
>> Hi,
>> 
>> I managed to reproduce the NSCursor mismatched push / pop issue, with 
>> tracing enabled.  The log is here:
>> 
>> http://theravensnest.org/NSCursor.1.log
>> 
>> I haven't yet done more than glance at it, but perhaps someone who knows 
>> this code better than me could gain more insight from it.  The trace is from 
>> launching the application until exit.  I exited after I was sure that the 
>> bug had been triggered.
>> 
>> For each NSCursor push or pop[1], there is a stack trace.  Ignore any 
>> functions in the trace that start .tesla or __tesla in the stack traces, 
>> these are artefacts of the instrumentation.
>> 
>> David
>> 
>> [1] Unfortunately, in this log it isn't totally clear whether the class or 
>> instance method is being called, so it looks like there are some duplicate 
>> pops that are actually -pop calling +pop.  I'll improve the trace output and 
>> try to recreate it.
>> 
>> -- Send from my Jacquard Loom
>> 
>> 
> 
> Would be helpful know the events processed alongside the nscursor
> events. The type and subtype of these.

http://theravensnest.org/NSCursor.1.log

This has the event delivered with every sendEvent: message logged, and is 
showing the mouse-over-text-field-when-window-created-gets-ibeam-cursor-stuck 
bug.

David


-- Sent from my brain