classpath
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: JIT pluggability


From: Tom Tromey
Subject: Re: JIT pluggability
Date: 09 Jan 2004 13:57:16 -0700
User-agent: Gnus/5.09 (Gnus v5.9.0) Emacs/21.3.50

>>>>> "Brian" == C Brian Jones <address@hidden> writes:

Brian> Going off on a tangent, are there other interfaces where reuse
Brian> may be possible such as JVMPI, JPDA, and the wire protocols for
Brian> those... if the hooks into the free JVMs are somewhat standard?

That's a good question.  I haven't really looked too deeply into these
things.  gcj is a problem child for things like JDWP, JVMPI, etc --
they seem to be pretty hard to implement given our model :-(.  It
might still be interesting to try, or to write classes that handle the
protocol (e.g., we could write subclasses that use gdb to drive
libgcj, or some insane thing like that).


There are a couple other areas we can all cooperate in.  One we
already do -- we have a (more-or-less) shared jni.h.

Another area is the bytecode verifier.  The libgcj verifier can be
made to work with non-libgcj VMs fairly easily (I've already got a
patch to make it work inside gcj itself, a very different
environment).  It isn't perfect (still a known bug or two) and it is
written in C++, but it might still be suitable for many projects.  I'm
definitely interested in helping out projects that want to do this.

Tom




reply via email to

[Prev in Thread] Current Thread [Next in Thread]