[Vm-dev] [squeak-dev] Bug in PolygonMorph>>#filledForm

[Nicolas Cellier]

Several candidates, maybe I missed some:
https://source.squeak.org/VMMaker/VMMaker.oscog-eem.2461.diff
https://source.squeak.org/VMMaker/VMMaker.oscog-eem.2460.diff (Oct 2018)
https://source.squeak.org/VMMaker/VMMaker.oscog-eem.2284.diff
https://source.squeak.org/VMMaker/VMMaker.oscog- nice.2281.diff

Le ven. 21 févr. 2020 à 23:13, Nicolas Cellier <
nicolas.cellier.aka.nice at gmail.com> a écrit :

> I do not observe the artifacts at same smear step (nor any other) with
> VM [CoInterpreterPrimitives VMMaker.oscog-eem.2266] 5.0.201708312323
>
> That's just 222 commits to bissect...
>
> Le ven. 21 févr. 2020 à 23:00, Nicolas Cellier <
> nicolas.cellier.aka.nice at gmail.com> a écrit :
>
>> There's already a problem with end 2018 VM
>> VM [CoInterpreterPrimitives VMMaker.oscog-eem.2488] 5.0.201812040127
>>
>> Le ven. 21 févr. 2020 à 22:48, Nicolas Cellier <
>> nicolas.cellier.aka.nice at gmail.com> a écrit :
>>
>>> Interesting case to debug...
>>> In 64bits VM, a problem first appears in first smear:distance: operation
>>> (direction 1 at 0 to the right) at skew=16...
>>>
>>> [image: Capture d’écran 2020-02-21 à 22.24.03.png]
>>>
>>>
>>> It seems that we copy bits too far to the right (hence we overwrite the
>>> bottom rows)
>>> Fortunately, the artefacts outside polygon bounds disappear at final
>>> bitAnd operation, but that's scary!
>>>
>>> It could be:
>>> - case of overlap detection failure (#checkSourceOverlap), but the code
>>> did not change
>>> - bad initialisation in #sourceSkewAndPointerInit (the one I corrected
>>> in 2019)
>>>  https://github.com/OpenSmalltalk/opensmalltalk-vm/issues/426
>>> - bug in #copyLoop (which I also corrected in 2019)
>>> https://github.com/OpenSmalltalk/opensmalltalk-vm/issues/441
>>>
>>> Weird kind of integer overflow? (I've made most of the int unsigned, so
>>> this might not be detected by UB sanitizer).
>>>
>>> Or could it be that the extra buffer overrun that I removed did have a
>>> side effect of filling more gaps?
>>> It would be interesting to observe what happens at same #smear:distance:
>>> step with older VM.
>>>
>>> Le ven. 21 févr. 2020 à 20:39, Stéphane Rollandin <
>>> lecteur at zogotounga.net> a écrit :
>>>
>>>> > Hmmm... the form was still filled in Squeak 5.1 (Update 16549),
>>>> > VM 201701311639 (32-bit). Not so in Squeak 5.2.
>>>>
>>>> Yes, but that same 5.1 image with a current VM has the problem...
>>>>
>>>> Stef
>>>>
>>>>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.squeakfoundation.org/pipermail/vm-dev/attachments/20200221/c73807a0/attachment-0001.html>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: Capture d’écran 2020-02-21 à 22.24.03.png
Type: image/png
Size: 18235 bytes
Desc: not available
URL: <http://lists.squeakfoundation.org/pipermail/vm-dev/attachments/20200221/c73807a0/attachment-0001.png>