Re: [libunwind] backtrace on EM64T

[David Mosberger-Tang]

On 11/21/05, Arun Sharma <address@hidden> wrote:
> Keith Owens wrote:

> > One problem with the ia64 unwind cache is that it uses kmalloc().  Code
> > expects to be able to call the unwinder when the system is running out
> > of memory or even from an NMI/MCA/INIT handler, at which point the slab
> > locks and chains are unreliable.  My KDB patches introduce a last ditch
> > allocator for the ia64 unwinder, so we can get backtraces even when the
> > entire system is dying.
>
> Having a malloc anywhere in the unw_step() path is a deal killer for us.
>
> http://gcc.gnu.org/bugzilla/show_bug.cgi?id=24724
>
> Do you see this issue in libunwind as well? If yes, could you be more
> specific? Thanks.

My libunwind kernel patches have a variable unwi_page_alloc_usable
which controls whether or not the page allocator is used by libunwind.
 The intention is for debuggers such as kdb to clear this flag while
they're running and restore it afterwards.

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