Rémi and I confirmed that it is in fact dead code (JRuby's fault) which is causing a problem for ASM 4 (ASM's fault). He's looking into a fix, and I'll try to eliminate the dead code.
Hopefully it can be fixed in ASM soon, since the code in question is being generated by a particularly gnarly part of JRuby's compiler. - Charlie (mobile) On Apr 29, 2011, at 15:58, Mark Roos <[email protected]> wrote: > > I have seen a stack map issue with ASM 4 as well. In my case it was dead ( > unreachable ) code which had > a return instruction in it. Seemed to confuse the next stack map > computation. I am allowing ASM 4 to compute > the maps on its own. > > I replaced it (the return) with no ops and the problem went away. The other > dead instructions did not seem to > matter. > > mark > > > From: Charles Oliver Nutter <[email protected]> > To: Da Vinci Machine Project <[email protected]> > Date: 04/28/2011 11:17 AM > Subject: Re: Assembly output from JRuby 'fib' > Sent by: [email protected] > > > > > On Thu, Apr 28, 2011 at 11:03 AM, Rémi Forax <[email protected]> wrote: > > On 04/28/2011 03:56 PM, Charles Oliver Nutter wrote: > >> stack map is invalid. Could be an ASM bug? > > > > yes. it could :) > > > > Also it could be a bug on your side because to calculate stackmap > > ASM needs to be able to find the common supertype of two types. > > I don't know if you provide your own method for doing that but > > by default it relies on reflection (getSupperClass/isAssignableFrom) > > which doesn't work if your code use a Class that doesn't exist yet. > > If you're willing, I'll toss you 1.5 and 1.6 ASM-produced bytecode > off-list and perhaps you can help me figure out if I'm doing something > wrong or if ASM is choking on our peculiar code structure. I need this > working to start running tests with invokedynamic. > > - Charlie > _______________________________________________ > mlvm-dev mailing list > [email protected] > http://mail.openjdk.java.net/mailman/listinfo/mlvm-dev > > > _______________________________________________ > mlvm-dev mailing list > [email protected] > http://mail.openjdk.java.net/mailman/listinfo/mlvm-dev
_______________________________________________ mlvm-dev mailing list [email protected] http://mail.openjdk.java.net/mailman/listinfo/mlvm-dev
