Ah, ok. Well I think David addressed that somewhat in his response. I'm not sure where the JIT lives in those, but I'm sure at least a few of them expose bits of it.
Cheers, Joel On Wed, Aug 5, 2015 at 2:25 PM Chris Marshall <[email protected]> wrote: > Hi Joel- > > My question is related to developing a new JIT > capability for PDL3. > > Since one could use libtcc itself to build some > XS code at runtime to load to load and run from > perl, it is clear that JIT compiling does not have > any specific dependency on the perl keywords > API itself. > > I'm not trying to change C::Blocks, just want to > determine how much I'll be able to leverage from > David's efforts. > > One model for JIT support for PDL3 could be to use > roles which could allow for pluggable JIT support. > The older perls could use a "clunky JIT" using libtcc > or tcc as above. Newer perls would support the improved > performance and *much* lower latency possible > with C::Blocks. > > --Chris > > On Tue, Aug 4, 2015 at 9:29 PM, Joel Berger <[email protected]> > wrote: > >> Chris I disagree. For new code I see little reason to limit the api to >> functionality available that far back. The keyword API is a great new >> feature to show off, why limit things? >> >> On Tue, Aug 4, 2015 at 7:40 PM Chris Marshall <[email protected]> >> wrote: >> >>> Does C::Blocks really require the perl keyword API to work? >>> It would be nice if there was a version that could work with >>> perl 5.10.x. >>> >>> >>> --Chris >>> >>> >>> On 8/4/2015 17:33, David Mertens wrote: >>> >>> Hey Chris, >>> >>> I wish I could say something more precise and causal about why PDL is >>> slowing down, but it's hard without measurements. I think the JIT has some >>> promise both in producing context-optimized code, but more importantly for >>> playing around with the internals. >>> >>> The keyword API was introduced in v5.12. Also, I just found a way to get >>> lexical variable insertion working for Perls earlier than 5.18, although I >>> have not yet pushed it to CPAN. >>> >>> David >>> >>> >
------------------------------------------------------------------------------
_______________________________________________ pdl-general mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/pdl-general
