Re: [libffi] Patch 14922: fix libgcj EH breakage

[Matt Austern <austern at apple dot com>]

On Apr 17, 2004, at 1:31 AM, Andrew Haley wrote:

> Matt Austern writes:
>
> > There are still 15 unexpected failures in the libjava test suite:
> > FAIL: cxxtest.c compilation
> > FAIL: Throw_2 execution - source compiled test
> > FAIL: Throw_2 execution - gij test
> > FAIL: Throw_2 execution - bytecode->native test
> > FAIL: Throw_2 -O3 execution - source compiled test
> > FAIL: Throw_2 execution - gij test
> > FAIL: Throw_2 -O3 execution - bytecode->native test
> > FAIL: anfi execution - gij test
> > FAIL: anfi execution - gij test
> > FAIL: indirect execution - gij test
> > FAIL: indirect execution - gij test
> > FAIL: invoke_from_inner execution - gij test
> > FAIL: invoke_from_inner execution - gij test
> > FAIL: update_outer execution - gij test
> > FAIL: update_outer execution - gij test
> >
> > These sound scary, but they aren't.
>
> Well, they scare me.
>
> The Throw_2 failures are surely due to the fact that we aren't
> catching null pointer traps.
>
> The gij failures may be due to something broken in libffi.

Possibly, but if so it was already broken on March 8---that is, it 
cannot be something that was broken by libffi changes that I made on 
March 13.  I'm attaching the libjava summary from the run I made 
yesterday, which was from a compiler that I checked out with -D "8 Mar 
2004".

I claim that this patch restores libjava to the state it was in before 
my linkonce changes.  That's as far as I can go.

			--Matt

Attachment: libjava.sum
Description: Binary data