Re: SJLJ exceptions are broken

[Mark Mitchell <mark at codesourcery dot com>]

>>>>> "Richard" == Richard Henderson <rth@redhat.com> writes:

    Richard> Oops.  I should add stub versions to unwind-sjlj.c.  They
    Richard> needn't do anything, since they won't be used, but
    Richard> clearly they need to be present.

Great.  Please let me know when that's done, so that I can fire off
some tests.

I don't understand the logic well enough to know why these functions
won't get used; it looks like parse_lsda_header will call
read_encoded_value, and that could end up in base_of_encoded_value.  I
guess you're saying that somehow the encoding will never be of the
type that requires a call to these functions?  If so, you could also 
just #ifdef out the calls, I suppose.

Thanks,

--
Mark Mitchell                   mark@codesourcery.com
CodeSourcery, LLC               http://www.codesourcery.com