Re: [Libunwind-devel] Libunwind support for CFI from .debug_frame in DWARF

[Andrew Cagney]

David,

A standard "gcc -g" includes .debug_info but not .eh_frame (and often in 
a separate debug-info).  By ignoring .eh_frame you're precluding any 
possibility of libunwind working effectively with that compiler.  Can we 
revist this?

Andrew

David Mosberger-Tang wrote:
> Hi Corey,
>
> On 2/6/07, Corey Ashford <address@hidden> wrote:
>
> > Are there any plans to add support to libunwind for the case when only
> > .debug_frame sections are present?
>
> Can't speak for others, but I myself do not have plans for this.  As I
> recall, I considered it briefly, but decided against it.  I think part
> of it was that the info in .debug_frame isn't exactly the same as in
> the .eh_frame sections.
>
> Regards,
>
>  --david
>
> > I ask this knowing that the default switches for building Linux these
> > days includes -fexceptions, however many user apps will not have this
> > switch set.
> >
> > ---
> >
> > As an aside, for the Frysk developers out there:
> >
> > Is there some reliance on the -fasynchronous-unwind-tables switch being
> > used to compile the app-under-test besides getting libunwind to work?
> > In other words, are there other parts of Frysk that rely on .eh_frame
> > sections?
> >
> > Thanks for your consideration,
> >
> > - Corey
> >
> >
> >
> > --
> > Corey Ashford
> > Software Engineer
> > IBM Linux Technology Center, Linux Toolchain
> > Beaverton, OR
> > 503-578-3507
> > address@hidden
> >
> >
> > _______________________________________________
> > Libunwind-devel mailing list
> > address@hidden
> > http://lists.nongnu.org/mailman/listinfo/libunwind-devel