"The tragedy of PIC" summary
Locked
"The tragedy of PIC" summary
 
|
Hi eliot So I was reading the PhD of Holzle and I wanted to know - I remember that at southampton (yes that far) you did a presentation on the tragedy of the PIC. And I would not find the slides (because there were none). What was the pitch? That megamorphic site are killing PIC? - does VW use type feedback? if yes, what were the strategies to decide what and when to recompile. - I was wondering how type feedback and PIC interact (I'm at chapter 5 of the PhD so may be this is explain in the self implementation part) the first case of the PIC could be based on type feedback? Stef |
Re: "The tragedy of PIC" summary
|
|
What is PIC? :) Doru On Sun, Jun 30, 2013 at 10:06 AM, stephane ducasse <[hidden email]> wrote:
"Every thing has its own flow"
|
Re: "The tragedy of PIC" summary
|
|
http://en.wikipedia.org/wiki/Inline_caching#Polymorphic_inline_caching frank On 2 July 2013 10:12, Tudor Girba <[hidden email]> wrote: > > What is PIC? :) > > Doru > > > On Sun, Jun 30, 2013 at 10:06 AM, stephane ducasse <[hidden email]> wrote: >> >> >> Hi eliot >> >> So I was reading the PhD of Holzle and I wanted to know >> >> - I remember that at southampton (yes that far) you did a presentation on the tragedy of the PIC. >> And I would not find the slides (because there were none). >> What was the pitch? That megamorphic site are killing PIC? >> >> - does VW use type feedback? >> if yes, what were the strategies to decide what and when to recompile. >> >> - I was wondering how type feedback and PIC interact (I'm at chapter 5 of the PhD so may be this is explain in the self implementation part) >> the first case of the PIC could be based on type feedback? >> >> >> >> Stef > > > > > -- > www.tudorgirba.com > > "Every thing has its own flow" > |
Re: "The tragedy of PIC" summary
|
|
I should add, this is Urs Holzle's thesis: http://selflanguage.org/documentation/published/urs-thesis.html, "Adaptive Optimization for Self: reconciling high performance with exploratory programming" frank On 2 July 2013 10:18, Frank Shearar <[hidden email]> wrote: > http://en.wikipedia.org/wiki/Inline_caching#Polymorphic_inline_caching > > frank > > On 2 July 2013 10:12, Tudor Girba <[hidden email]> wrote: >> >> What is PIC? :) >> >> Doru >> >> >> On Sun, Jun 30, 2013 at 10:06 AM, stephane ducasse <[hidden email]> wrote: >>> >>> >>> Hi eliot >>> >>> So I was reading the PhD of Holzle and I wanted to know >>> >>> - I remember that at southampton (yes that far) you did a presentation on the tragedy of the PIC. >>> And I would not find the slides (because there were none). >>> What was the pitch? That megamorphic site are killing PIC? >>> >>> - does VW use type feedback? >>> if yes, what were the strategies to decide what and when to recompile. >>> >>> - I was wondering how type feedback and PIC interact (I'm at chapter 5 of the PhD so may be this is explain in the self implementation part) >>> the first case of the PIC could be based on type feedback? >>> >>> >>> >>> Stef >> >> >> >> >> -- >> www.tudorgirba.com >> >> "Every thing has its own flow" >> |
Re: "The tragedy of PIC" summary
|
|
Ahh :) Thanks, Doru On Tue, Jul 2, 2013 at 11:23 AM, Frank Shearar <[hidden email]> wrote:
"Every thing has its own flow"
|
Re: "The tragedy of PIC" summary
|
|
In reply to this post by Tudor Girba-2
PIC = polymorphic inline caches. When you optimize the lookup in the VM for speed you implement the so called inline caches: at each send site, you put information about the result of the previous lookup done here. In most case, a send site is monomorphic, which mean it is always the same method that is called there (the receiver has always the same class), so the information put is the method always called and this method is triggered instead of doing a lookup to find the method (to increase the execution speed by avoiding lookup).
Some other send sites use polymorphism, which means there are a few different class possible for the receiver of the send site. Inline caches optimizations on this sites are called PIC, and the optimization is done differently.
For more informations, there's Eliot's blog about cog here: or the thesis of hölzle here: The thesis of hölzle explains the basics of inline caches and their implementation in self 91, the cog blog explains their implementation in Cog. Hope I could help you,
2013/7/2 Tudor Girba <[hidden email]>
Clément Béra Mate Virtual Machine Engineer Bâtiment B 40, avenue Halley 59650 Villeneuve d'Ascq
|
«
Return to Squeak VM
|
1 view|%1 views
| Free forum by Nabble | Edit this page |