Re: libjava status on Tru64 UNIX V5.1

[Bryce McKinlay <bryce at waitaki dot otago dot ac dot nz>]

Tom Tromey wrote:

>>>>>>"Bryce" == Bryce McKinlay <bryce@waitaki.otago.ac.nz> writes:
>>>>>>
>
>Bryce> 1) make C++ able to use different EH models depending on
>Bryce> whether it is throwing a Java exception or not, so that Java
>Bryce> stuff can use SJLJ. In the short term this would most likely be
>Bryce> the best fix.
>
>If we could do that then maybe it would be good to eventually
>implement Jon Olson's cheap Java exception handling.  He referred to
>it several times a long time ago; I'm sure it is in the archives.
>

Here it is: http://gcc.gnu.org/ml/java/1999-q1/msg00016.html

Jon's scheme sounds good (certainly better than SJLJ, particularly in 
terms of code size) but a *major* disadvantage is that the exception 
stack is thread-local and thus requires synchronization (or at least a 
thread id lookup) at every entry to/exit from a "try" block.

It sounds like -fcheck-references may not be so terrible after all, 
compared to SJLJ. But we really should try to get dwarf2 EH working 
everywhere ASAP. Hopefully, libunwind will help us there.

>I don't expect any of this stuff to really change for 3.1.  I think it
>is too late for any major changes there.
>
Yeah.

Bryce.