Perl 6 - the future is here, just unevenly distributed

IRC log for #moarvm, 2015-02-04

| Channels | #moarvm index | Today | | Search | Google Search | Plain-Text | summary

All times shown according to UTC.

Time Nick Message
02:24 colomon joined #moarvm
04:14 kjs_ joined #moarvm
05:34 kjs_ joined #moarvm
07:04 FROGGS joined #moarvm
09:03 zakharyas joined #moarvm
10:44 kjs_ joined #moarvm
10:44 rurban joined #moarvm
10:45 pyrimidine joined #moarvm
12:05 kjs_ joined #moarvm
12:14 Util_ joined #moarvm
12:16 rurban_ joined #moarvm
12:17 bonsaikitten joined #moarvm
12:21 camelia joined #moarvm
13:10 kjs_ joined #moarvm
13:11 donaldh joined #moarvm
13:17 kjs_ joined #moarvm
13:37 kjs_ joined #moarvm
13:45 zakharyas joined #moarvm
14:04 FROGGS joined #moarvm
14:52 FROGGS joined #moarvm
17:09 woolfy1 joined #moarvm
17:10 lizmat_ joined #moarvm
18:34 colomon joined #moarvm
19:04 colomon joined #moarvm
19:05 rurban_ joined #moarvm
19:46 woolfy1 left #moarvm
20:20 zakharyas joined #moarvm
21:52 brrt joined #moarvm
21:52 brrt i'll just leave this here: http://cr.openjdk.java.net/~jrose/pres/201502-JVMChallenges.pdf
21:53 brrt it's interesting to consider how much the jvm is stealing from moarvm :-P
21:53 brrt or rather, how much overlap there really is between the awesome work that jnthn++ is doing and the jvm teams are doing
21:54 jnthn I think we have some things to steal from the JVM also :)
21:55 brrt certainly
21:58 jnthn This *is* an interesting slide deck
21:58 jnthn brrt++
21:59 jnthn "Problem: hot-path customization"...we're grappling with this in spesh too :)
21:59 brrt thanks :-)
21:59 brrt yeah... for my information... i have a theory on how tracing works and how it relates to inlining
22:00 brrt basically,  what you'd do is for each invocation, inline the path you think you'll take, and translate the alternative paths to guards
22:02 brrt but if that's correct, we could do that with relatively minor modifications on inlining
22:04 FROGGS_ joined #moarvm
22:05 jnthn Tracing is more like recording the path you *do* take
22:06 brrt right, but i mean compiling the trace :-)
22:06 brrt the path
22:06 jnthn Ah
22:06 jnthn I was wondering about building a spesh graph *as* we trace
22:07 brrt hmmm
22:07 brrt that's an interesting thought
22:07 jnthn Along with the deopt table.
22:07 jnthn And inserting guards.
22:07 brrt requires a bit of rethinking i think
22:07 jnthn Kinda. :)
22:08 brrt but i can see how it could be done
22:08 brrt you might not need to build the spesh graph during tracing, but maybe link graphs together
22:08 jnthn One paper I read (forget which, grmbl) noted that you can make it SSA as you construct it too.
22:08 jnthn "trivially"
22:08 brrt is that so
22:09 kjs_ joined #moarvm
22:09 brrt ok, i'm curious
22:09 jnthn Well, given a trace has no branches... :)
22:09 brrt that's true
22:09 brrt can a trace have loops?
22:09 brrt how are loops dealt with
22:09 jnthn Well, you normally are tracing a loop
22:10 jnthn Finding loop headers is step 1 for tracing.
22:10 jnthn The tricky one is nested loops. There are various strategies for that. I need to do more reading on this bit.
22:10 brrt we should have a 'papers' section on moarvm.org
22:11 jnthn Yeah
22:11 brrt as in, stuff we read that helped us along
22:11 jnthn http://6guts.wordpress.com/interesting-papers/ has a handful
22:12 jnthn I should make a better job of keeping that updated with new ones I discover, maybe :)
22:12 brrt yeah :-) thanks for that link
22:12 brrt those are nice
22:12 brrt i'll be off now
22:13 jnthn o/
22:14 brrt \o
22:14 brrt left #moarvm
22:45 colomon joined #moarvm
23:52 chrisseaton left #moarvm

| Channels | #moarvm index | Today | | Search | Google Search | Plain-Text | summary