Re: Fusion-ifying proto parse trees

Frank Birbacher wrote:
> Hi!
>
> Eric Niebler schrieb:
>
> > But what
> > Dave is saying is that the loop unrolling should be happening at a lower
> > level, and for the *all* Fusion sequences. It should be baked into
> > Fusion, so that nobody should have to hard-code their own unrolled loops
> > as you're doing above. They'll just call the appropriate Fusion algorithm.
>
>
> As I understand the discussion so far the goal (more or less) is:
>  - user writes "a >> b >> c >> d"
>  - proto: gives flattened et
>  - in spirit: "parse_sequence(S s) { fusion::for_each(s, &parse); }"
>  - in fusion: automated loop unrolling via for_each
>  - user gets efficient code
>
> right?


You're getting warmer. :-)

- user writes "a >> b >> c >> d"
- proto: gives ET tree
- in spirit: parse flattens proto ET tree using a proto 2nd
  phase transform
- in spirit: "parse_sequence(S s) { fusion::for_each(s, &parse); }"
- in fusion: automated loop unrolling via for_each
- user gets efficient code

Performance benchmarks showed it's faster for the ET tree to be 
flattened all at once when the entire ET is known then to do it 
piecemeal while building the ET.

--
Eric Niebler
Boost Consulting
www.boost-consulting.com


-------------------------------------------------------
Using Tomcat but need to do more? Need to support web services, security?
Get stuff done quickly with pre-integrated technology to make your job easier
Download IBM WebSphere Application Server v.1.0.1 based on Apache Geronimo
http://sel.as-us.falkag.net/sel?cmd=lnk&kid=120709&bid=263057&dat=121642