chicken-users
[Top][All Lists]
Advanced

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

Re: [Chicken-users] jbogenturfa'i compile time down to 12 minutes


From: Felix
Subject: Re: [Chicken-users] jbogenturfa'i compile time down to 12 minutes
Date: Sat, 04 Dec 2010 15:01:35 +0100 (CET)

From: Alex Shinn <address@hidden>
Subject: Re: [Chicken-users] jbogenturfa'i compile time down to 12 minutes
Date: Fri, 3 Dec 2010 14:55:37 -0800

> On Wed, Dec 1, 2010 at 12:45 PM, Felix
> <address@hidden> wrote:
>> From: Alan Post <address@hidden>
>> Subject: [Chicken-users] jbogenturfa'i compile time down to 12 minutes
>> Date: Wed, 1 Dec 2010 09:33:07 -0700
>>
>>> I will be more than happy to thoroughly test any changes mode to
>>> Chicken's compiler that attempt to optimize large letrec forms
>>> with mutually recursive definitions, should any work be done in
>>> that area.
>>
>> Working on it - I'll get back to about this.

Well, it turns out this is not that easy. The right thing would
be a good lambda-lifter, but this is tricky to implement (The
existing lambda-lifter is crap and I removed it, since it's
awfully complicated and mostly ineffective). Another interesting
optimization would be to "globalize" let[rec]-bound variables
by lifting them to toplevel, but that again may retain more
garbage than necessary.

> I have a macro called let-machine which is basically
> a letrec that automatically does the transformations
> needed to compile efficiently under Chicken.  I thought
> I made an egg of it, but apparently not.  It must be
> on my home machine, which I won't have access to
> for another two weeks.

This is nice but only really effective for tight code (as
you have pointed out in your article).


cheers,
felix



reply via email to

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