[rvm-research] Fwd: Re: Regarding the function call mechanism

[Eliot M. <mo...@cs...>]

Meant to send this to the list too -- E
-------- Original Message --------
Subject: Re: [rvm-research] Regarding the function call mechanism
Date: Mon, 25 Mar 2013 12:48:12 -0400
From: Eliot Moss <mo...@cs...>
Reply-To: mo...@cs...
To: Hari Krishnan <hk...@ti...>

Dear Hari -- I am not sure if this answers your
question, but, depending on the style of GC algorithm
requested, Jikes RVM / MMTk inserts write barriers
that note pointer field updates for use by the GC.
These are actually specified by MMTk, but inserted
by the Jikes RVM compilers on MMTk's behalf.

Regards -- Eliot Moss