[XenPPC] Slab corruption

2007-02-01 Thread Ryan Harper
I noticed the following output from the kernel running XenPPC on JS20, Slab corruption: start=c6d9f800, len=2048 680: a5 0c 00 00 6b 6b 6b 6b 6b 6b 6b 6b 6b 6b 6b 6b Prev obj: start=c6d9f000, len=2048 000: 30 30 2c 0a 30 78 30 30 30 30 30 30 30 30 2c 30 010: 78 30 30 30 30 30 30

Re: [XenPPC] basic Xen achitecture question

2007-02-01 Thread Jimi Xenidis
On Feb 1, 2007, at 4:07 PM, Yoder Stuart-B08248 wrote: I'm trying to understand the paravirtualization interface between the OS and Xen. Here is one thing that's confusing me-- In Linux, the xen probe() routine calls hpte_init_lpar() which hooks up some pseries hypervisor MMU routines.

Re: [XenPPC] Slab corruption

2007-02-01 Thread Jimi Xenidis
Thanks Ryan. This is dom0, and the last time I saw this was because we had the flat-tree reservations wrong in domu. I wonder if there is any howto's on debugging a slab corruption? -JX On Feb 1, 2007, at 4:18 PM, Ryan Harper wrote: I noticed the following output from the kernel running