help-smalltalk
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

[Help-smalltalk] ByteCode to ByteCode optimizations?


From: Holger Hans Peter Freyther
Subject: [Help-smalltalk] ByteCode to ByteCode optimizations?
Date: Mon, 15 Sep 2014 19:58:05 +0200
User-agent: Mutt/1.5.23 (2014-03-12)

Hi,

partially inspired by the (Cog) Sista slides. On our slow ARM hardware
I was looking into ways to speed things up. So I was having code like
this:


next
        aStream atEnd ifFalse: [aStream uncheckedPeek = 123]

and

SomeStream>>#atEnd
        ^a > b


I was wondering if we could create additional ByteCode that would
allow us to 'replace' the receiver/inline without creating a new
context? Something that allows us to quickly assert that the new
receiver is a of specific class/oop and the execute the inlined code.

What do you think? Is it feasible with the current code? In the
next step we could look into adding code to check the PIC and count
and learn about the types. :}


What do you think?

        holger



reply via email to

[Prev in Thread] Current Thread [Next in Thread]