--- Dalibor Topic <[EMAIL PROTECTED]> wrote: > Hi Tim, > > --- Timothy Stack <[EMAIL PROTECTED]> wrote: > > > > A lame guess is that this is due to a cycle in the > > serialized objects and > > kaffe not being smart enough to notice. But, I > > don't know that much about > > the serialization code (pat touched it last if we > go > > the hot potato > > route). > > Sound like as good guess as any. I'll see what I can > find.
It works if I double the stack size, -ss 64K. I've talked to Godmar about setting the default stack size to 64K for jit3, but he thinks that it's bad enough as it is ;) And he's right. That would only buy us cases where the serialization stack fits into 64K. In order to be able to serialize more complicated objects, we'd have to bump up the stack size again. So I'm wondering if serialization could be made non-recursive using a List to store the entries in the order to be serialized, and then iterating over the list and writing them out on the stream. That should in theory handle as large objects as can fit into memory. What do you think? best regards, dalibor topic __________________________________________________ Do you Yahoo!? Yahoo! Mail Plus - Powerful. Affordable. Sign up now. http://mailplus.yahoo.com _______________________________________________ kaffe mailing list [EMAIL PROTECTED] http://kaffe.org/cgi-bin/mailman/listinfo/kaffe
