On Sun, 2008-04-20 at 23:58 -0400, Brian A. Seklecki (Mobile) wrote:
On Sat, 2008-04-19 at 12:45 +0200, Christoph Egger wrote:
Hi!
I need someone who can reproduce this crash:
http://mail-index.netbsd.org/port-xen/2008/04/10/msg003555.html
Sorry for the delay. I'll be able to to test it in the AM.
Unfortunately, that machine is getting rebuilt immediately there-after
with FreeBSD (don't ask)
At 128MB for the Dom0, the pmap problem with mfi(4) is gone, but now
manifests itself in bnx(4), 192MB seems to be a good consensus.
See attached for panic @ 128MB Dom0
~BAS
P.S. Maybe I'm greedy, but I would think that the Xen kernel + NetBSD
DOM0 + MFS could use ~128MB or less in a Veso-like deployment,
maximizing physical RAM available to DomUs. Dom0 existing to manage
iSCSI targets, provide VLAN bridging, environmental data, and RAID
monitoring optionally (just musing)
~BAS
I want to know if this patch fixes it.
http://www.netbsd.org/~cegger/mfi.diff
Christoph