updated libunwind kernel patch

updated libunwind kernel patch

[david mosberger]

[-- Attachment #1: Type: text/plain, Size: 1056 bytes --]

Attached below is an updated libunwind kernel patch relative to Tony's
linux-2.6 git repository.  On the ia64-side, I just updated it
slightly to account for minor changes in Tony's tree.  There is also
the beginning of some x86-64 support (to demonstrate the portability
of the API), but that code isn't ready for prime-time.  I haven't made
a final decision on what to do about that.  On the one hand, I'd like
to have libunwind support for more than one platform but on the other,
there are some GCC issues which are standing in the way for libunwind
being used as the primary unwinder on DWARF-based platforms (at least
as of last time I checked) so I may still decide to push for an
initial ia64-only version, since that's where we'd get the biggest
benefit (single source tree which can be shared 95% with the
user-level libunwind & lots of bug-fixes compared to the old kernel
unwinder).

  --david
-- 
Mosberger Consulting LLC, voice/fax: 510-744-9372,
http://www.mosberger-consulting.com/
35706 Runckel Lane, Fremont, CA 94536

[-- Attachment #2: libunwind-050815.diff.gz --]
[-- Type: application/x-gzip, Size: 0 bytes --]

Re: updated libunwind kernel patch

[Bob Picco]

david mosberger wrote:	[Mon Aug 15 2005, 05:52:14PM EDT]
> Attached below is an updated libunwind kernel patch relative to Tony's
> linux-2.6 git repository.  On the ia64-side, I just updated it
> slightly to account for minor changes in Tony's tree.  There is also
> the beginning of some x86-64 support (to demonstrate the portability
> of the API), but that code isn't ready for prime-time.  I haven't made
> a final decision on what to do about that.  On the one hand, I'd like
> to have libunwind support for more than one platform but on the other,
> there are some GCC issues which are standing in the way for libunwind
> being used as the primary unwinder on DWARF-based platforms (at least
> as of last time I checked) so I may still decide to push for an
> initial ia64-only version, since that's where we'd get the biggest
> benefit (single source tree which can be shared 95% with the
> user-level libunwind & lots of bug-fixes compared to the old kernel
> unwinder).
> 
>   --david
Hi David:

Any chance you can post the attachment?  It appears to be empty.

thanks,

bob