[BUG] User interrupts no longer interrupt locked UI process

John M McIntosh johnmci at smalltalkconsulting.com
Fri Mar 19 04:55:56 UTC 2004


Ok, well l added that interruptCheckCounterFeedbackReset many years ago  
to moderate the calling of
quickCheckForInterrupts which was being done each time that  
interruptCheckCounter default of 1000 decremented down to
zero. What I found on 33Mhz macs was that it was being called 2000  
times a second, hence we needed something to better indicate how often  
to call.

Because of the tight loop, I think the 116,998,772 means  we decrement  
interruptCheckCounter that  much in interruptChecksEveryNms (2  
milliseconds).  Mmm I'm wondering about the side effects when it gets  
so large because of a tight loop, which we exit and then need to change
down to oh say 20,000

PS also see where interruptCheckCounter gets set to zero, I wonder if  
it's masking this issue.

On Mar 18, 2004, at 4:46 PM, Ned Konz wrote:

> On Thursday 18 March 2004 3:53 pm, Ned Konz wrote:
>> On Thursday 18 March 2004 3:46 pm, I wrote:
>
>> but I can't seem to break out of
>>
>>
>> recursionTest
>> 	^self recursionTest
>>
>>
>> for some reason.
>>
>> Wonder why...
>
>
> Well, it appears to be because quickCheckForInterrupts is missing the  
> boat.
> 	interruptCheckCounter is 115411591
> 	interruptCheckCounterFeedbackReset is 116998772
>
> Though something as simple as causing a checkForInterrupts() after  
> every
> fullGC() might catch the infinite recursion.
>
> I tried zeroing interruptCheckCounter after a fullGC() and I could  
> break out
> of tight recursion.
>
> Without that change, if you let the recursion go for a while, there's  
> no
> breaking out of it.
>
> Oh yeah...
> 	reclaimableContextCount is 1587135
>
> Wonder if that could also be used as a trigger for checking the  
> interrupts.
>
> --  
> Ned Konz
> http://bike-nomad.com/squeak/
>
>
>
--
======================================================================== 
===
John M. McIntosh <johnmci at smalltalkconsulting.com> 1-800-477-2659
Corporate Smalltalk Consulting Ltd.  http://www.smalltalkconsulting.com
======================================================================== 
===




More information about the Squeak-dev mailing list