[mercury-users] Destructive list operations

[Julien Fischer]

On Mon, 12 Jan 2009, Julien Fischer wrote:

>
> On Mon, 12 Jan 2009, Michael Day wrote:
>
>> By the way, the --optimize-constructor-last-call flag allows us to drop a 
>> lot of the reverse-accumulator tricks we were using and get a speed boost 
>> into the bargain. It's great! But it doesn't seem to be enabled by default 
>> for -O5, is that correct? Is there any reason why it wouldn't be enabled 
>> for this optimisation level? It seems pretty useful; we only didn't use it 
>> in the past because it made the 2006 compiler crash.
>
> The last time it was benchmarked on our usual benchmark program (the
> compiler compiling its six largest modules), it sometimes resulted in
> a slowdown - that's why it's not currently enabled by one of the -O
> options.
>
> We should re-run those benchmarks (in fact I will), since there have
> been a number of bugfixes made to lco since then.

The results were posted on mercury-developers:
<http://www.mercury.csse.unimelb.edu.au/mailing-lists/mercury-developers/mercury-developers.200901/0000.html>

Julien.
--------------------------------------------------------------------------
mercury-users mailing list
Post messages to:       mercury-users at csse.unimelb.edu.au
Administrative Queries: owner-mercury-users at csse.unimelb.edu.au
Subscriptions:          mercury-users-request at csse.unimelb.edu.au
--------------------------------------------------------------------------