Re: ia64 c++ abi exception handling

[Richard Henderson <rth at redhat dot com>]

On Mon, Mar 12, 2001 at 12:46:29PM -0800, Benjamin Kosnik wrote:
> size_t should always be qualified in c++ code with std::, now that the
> headers are properly constructed.

Fixed.

> I'm a bit confused by this: seems kind of wonky. Since we're taking
> pains to generate the C++ header file <limits> correctly for each
> architecture, I'd think you could do:
> 
>  numeric_limits<int>::max() //largest value
>  numeric_limits<int>::digits() //bit size

I can't use these to choose between int and long for typedefs, surely?

> eh_bad.cc seems kind of weak, eh_aux_runtime.cc seems better, but whatever.

No problem.

> Similarly, all that's in here is std::exception::what definition, so
> you might want to consider using the C++ iso 14882 tags:
> or exception.cc.

exception.cc it is.

> Also, kill the pragma implementation: we shouldn't be using this
> pragma anymore.

??

$ grep pragma *
exception:#pragma interface "exception"
exception_support.cc:#pragma implementation "exception"
new:#pragma interface "new"
new_handler.cc:#pragma implementation "new"
tinfo.cc:#pragma implementation "typeinfo"
typeinfo:#pragma interface "typeinfo"

At present I've done nothing but preserve existing behaviour.
Should I remove them all then?

> 4) Index: libstdc++-v3/libsupc++/eh_term.cc
> 
> blah blah blah: you are being consistent but the naming is terse. 

eh_terminate.cc?

> Remove exception_support.h from headers.

Err, I did.  Looking at the wrong side of the diff?


r~