Post #2178016
2026-04-18 15:38 UTC
@CanLehmann@mastodon.social @smarr@mastodon.acm.org @lindsey@recurse.social @verijit@mastodon.social verijit is actually doing the second futamura projection! The result (a residual cpu, specialized to a concrete program) is the same, but the compilation process itself is faster. (Graal is doing the first, rpython too). And yes, partial evaluation and meta tracing are not exactly the same, but have a lot of similarities (there's a chapter about the similarities/differences in my diss)
Replies (1)
-
@CanLehmann@mastodon.social 2026-04-18 15:48
@smarr@mastodon.acm.org @lindsey@recurse.social @verijit@mastodon.social Ok, so I just had a little discussion with @cfbolz@mastodon.social and we actually kind of do both. So: 1. The specialization of the CPU on the instruction stream is the 1st projection 2. We use a generating extension to generate the traces. So yes, we actually also perform the 2nd projection. As @cfbolz@mastodon.social mentioned this is actually pretty important for tracing performance.