2016-09-28 23:41 GMT+02:00 Holger Freyther <holger@freyther.de>:

Hi,

Magritte and my TagLengthValue (TLV) both use >>#instVarNamed:/>>#instVarNamed:put: to read and write from an object. I was just running >>#bench on my SMPP library and noticed that besides Spur Pharo5 is slower than Pharo3.

I added this to PointerLayout:

>># instVarIndexFor: aString ifAbsent: aBlockClosure
�� | idx |
�� idx := 1.
�� slotScope do: [:each |
�� each isVisible ifTrue: [
�� each name = aString ifTrue: [^idx].
�� idx := idx +1]].
�� ^aBlockClosure value

and modified ClassDescription/TClassDescription to use it:

"protocol: instance variables"
instVarIndexFor: instVarName ifAbsent: aBlock
�� "Answer the index of the named instance variable."

�� | index |
�� index := self classLayout instVarIndexFor: instVarName ifAbsent: [0].
�� index = 0 ifTrue:
�� [^self superclass == nil
�� ifTrue: [aBlock value]
�� ifFalse: [self superclass instVarIndexFor: instVarName ifAbsent: aBlock]].
�� ^self superclass == nil
�� ifTrue: [index]
�� ifFalse: [index + self superclass instSize]

The speed-up comes from filtering allSlots to allVisibleSlots (and creating an Array), then collecting the slot names and finally searching the name. Does it make sense to integrate such speed-ups?

cheers
�� holger