Kernel Dump

[Peter Wemm]

On Thursday 23 June 2005 06:53 am, David O'Brien wrote:
> On Wed, Jun 22, 2005 at 11:23:25PM -0700, Peter Wemm wrote:
> > IA64 and Sparc systems usually had sparse memory configurations and
> > the old raw format didn't have any facility to avoid storing holes.
>
> ..
>
> > I've just rewritten the AMD64 crashdump support to use ELF like
> > ia64. In fact, I reused most of the ia64 code. We ran into serious
> > problems at work, first on the amd64 platform and now also the i386
> > platform. The problem is that x86 machines are increasinly having
> > memory holes. The simplistic dump code has no way to skip the
> > memory hole and tries to dump things like the AGP frame buffer, pci
> > card MMIO space, PCI-Express configuration space (this means
> > accessing hardware!!) and so on.
>
> Hi Peter,
>
> This is great news!  I'm really happy to see this work.  Will we be
> able to switch to this for 6.0R?

I'd very much like for this to happen.  I've been testing primarily with 
amd64, but Paul Saab is doing a quick port to i386 (should be trivial).

The good news is that the libkvm impact is pretty much trivial.  I was 
pleasantly surprised.

There'll be patches to look at shortly.

BTW: It turns out that sparc64 uses its own custom dump format.  It uses 
a custom region descriptor instead of elf's core dump chunk 
descriptors.  I believe it also passes a pointer to its kernel page 
table.
-- 
Peter Wemm - peter at wemm.org; peter at FreeBSD.org; peter at yahoo-inc.com
"All of this is for nothing if we don't go to the stars" - JMS/B5