Trace trees
mikong
Here's the abstract for the paper on trace trees entitled "One Method
at a Time is Quite a Waste of Time":
Most just-in-time compilers for object-oriented languages operate at
the granularity of methods. Unfortunately, even “hot” methods often
contain “cold”
code paths. As a consequence, just-in-time compilers waste time
compiling code
that will be executed only rarely, if at all. We discuss an
alternative approach in
which only truly “hot” code is ever compiled.
This is the link to the article:
http://www.ics.uci.edu/%7Efranz/Site/pubs-pdf/C44Prepub.pdf
Just a disclaimer, I am not knowledgeable about trace trees, VMs,
compilers, etc. I only read Steve Yegge's post on dynamic languages,
and read a couple of papers linked from that article about type
inference and trace trees.
It would have been nice if I have forked Rubinius and implemented a
code to demo trace trees but I'm not really at that level yet. Right
now, I'm just at the stage of looking for a good book on compilers so
that I can contribute to Rubinius in the future.
William Morgan
> Here's the abstract for the paper on trace trees entitled "One Method
> at a Time is Quite a Waste of Time":
I posted an Andreas Gal paper a few weeks ago that you might want to
check out too. They're both about the same idea (basically optimize at
the loop level instead of at the method level) but I think that one had
more technical content. (I also started looking into this stuff thanks
to the Steve Yegge post.)
> Right now, I'm just at the stage of looking for a good book on
> compilers so that I can contribute to Rubinius in the future.
The "dragon book" is the go-to compilers book, but you'll have to keep
in mind the dynamically-typed, late-bound, anything-can-be-redefined
nature of Ruby, especially when it talks about optimization.
--
William <wmo...@masanjin.net>
mikong
On May 30, 9:20 pm, William Morgan <wmor...@masanjin.net> wrote:
> I posted an Andreas Gal paper a few weeks ago that you might want to
> check out too. They're both about the same idea (basically optimize at
> the loop level instead of at the method level) but I think that one had
> more technical content. (I also started looking into this stuff thanks
> to the Steve Yegge post.)
> The "dragon book" is the go-to compilers book, but you'll have to keep
> in mind the dynamically-typed, late-bound, anything-can-be-redefined
> nature of Ruby, especially when it talks about optimization.
A question though, is it important that I get the purple dragon book
(which I find a bit expensive)? I wonder, when one mentions 'the
dragon book' do they mean the 2nd edition or any of them?
William Morgan
> I got curious about the 'dragon book' when I read about it a week ago.
> A question though, is it important that I get the purple dragon book
> (which I find a bit expensive)? I wonder, when one mentions 'the
> dragon book' do they mean the 2nd edition or any of them?
I have no idea what the differences are between editions. My personal
experience is that I owned some version of it in college, sold it
immediately afterwards, and now wish I had kept it.
--
William <wmo...@masanjin.net>
Samuel Tesla
purple dragon). They're both very good, but I would strongly recommend
picking up the 3rd edition (which is probably what you're going to
find in stores anyway). It's been updated for a lot of more modern
stuff. It actually has content that is relevant to dynamic interpreted
languages such as Ruby.
-- Samuel
mikong
> They're both very good, but I would strongly recommend
> picking up the 3rd edition (which is probably what you're going to
> find in stores anyway). It's been updated for a lot of more modern
> stuff. It actually has content that is relevant to dynamic interpreted
> languages such as Ruby.
Werner Schuster (murphee)
> I got curious about the 'dragon book' when I read about it a week ago.
dragon book is what you want - a lot of it is about parsing/lexing
technology. (Mind you, I don't know the latest edition).
For dynamic optimization technology, I'd say you're better off reading
research papers which are available for free online.
Eg. a good starting point is the 'Publications' page of the JikesRVM
project (a JVM + JIT written in Java):
http://jikesrvm.org/Publications
For Garbage Collection, there's really only one book:
http://www.cs.kent.ac.uk/people/staff/rej/gc.html
Again - like the Dragon book - the price is a bit steep, and you might
get a lot of the info from the linked papers in the bibliography:
http://www.cs.kent.ac.uk/people/staff/rej/gcbib/gcbib.html