... just to keep you updated, I'm finishing testing the below, appears to work
well. Still missing: comments about ABI (no throw()) vs our specific libsupc++
implementation (suggestions very welcome ;)
Paolo.
//
Index: except.c
===
> Does that help?
Totally ;) And, for what is worth, I totally agree with your analysis.
Essentially, as far as I can see, only a TREE_NOTHROW (fn): line in
except.c:do_begin_catch is missing to complete the "baseline project" ;) Patch
forthcoming (I hope)
Thanks!
Paolo.
Paolo Carlini wrote:
Now, however, I'm not sure to understand which kind of way forward
you can see... Should the warning machinery be smart enough to not
warn at all in this case? Or? Sorry, probably it's just my ignorance
about those issues at the interface between core and library, I would
re
Hi Mark,
> This is not an ABI issue; the presence of absence of an
> exception-specifier does not affect the ABI.
Yes, I realized in the meanwhile, via my usual layman investigations using nm.
[snip explanation]
Ok. Actually I can clearly remember when you advertised the use of throw()
whenev
Paolo Carlini wrote:
... I see this comment in except.c:do_begin_catch:
/* Declare void* __cxa_begin_catch (void *). */
Instead, in libstdc++/libsup++/unwind-cxx.h:
extern "C" void *__cxa_begin_catch (void *) throw();
So, who is right? Note, the symbol is exported, I'm not suer w
... I see this comment in except.c:do_begin_catch:
/* Declare void* __cxa_begin_catch (void *). */
Instead, in libstdc++/libsup++/unwind-cxx.h:
extern "C" void *__cxa_begin_catch (void *) throw();
So, who is right? Note, the symbol is exported, I'm not suer which headroom for
chan
Hi,
> However, a svn log shows that neither unwind-cxx.h nor except.c
> in the C++ frontend changed this year in a relevant way. So my guess
> is that the warning is what changed. Either the warning settings during
> compilation, or a new warning was introduced/activated in the
> compiler itself.
Paolo Carlini wrote:
Hi all, hi Joseph,
on x86_64-linux I'm seeing many new warnings which didn't exist until a few
days ago, see below.
Is that known? Joseph, can it have something to do with your recent libsupc++
changes?
I noticed them too. These warnings say that the unwind-cxx.h
dec
