Hello All, I try port xen-unstable to illumos-based platform. I have made some changes on illumos and built xen, but have panic - take a look http://pastebin.com/fZSrExek I''d like to see some comments - what I need to check/update. --- Best regards, Igor Kozhukhov IRC# igork _______________________________________________ Xen-devel mailing list Xen-devel@lists.xen.org http://lists.xen.org/xen-devel
On Mon, Oct 01, 2012 at 11:24:06PM +0400, Igor Kozhukhov wrote:> Hello All, > I try port xen-unstable to illumos-based platform. > I have made some changes on illumos and built xen, but have panic - take a > look [1]http://pastebin.com/fZSrExek > I''d like to see some comments - what I need to check/update.kmdb: single-step stop on division by zero (#de) trap kmdb: target stopped at: sti+0x85: popq %r10 [0]> $C ffffff0003ed1c20 sti+0x85() ffffff0003e05970 switch_sp_and_call+0x13() ffffff0003e059d0 do_interrupt+0x14a() ffffff0003e05a80 xen_callback_handler+0x42d(ffffff0003e05a90, fffffffffbc1d778) ffffff0003e05a90 xen_callback+0x18f() ffffff0003e05ba0 __hypercall2+0xa() ffffff0003e05bb0 HYPERVISOR_block+0x10() ffffff0003e05bc0 mach_cpu_idle+0x1d() ffffff0003e05bf0 cpu_idle+0xfe() ffffff0003e05c00 cpu_idle_adaptive+0x13() ffffff0003e05c20 idle+0x41() ffffff0003e05c30 thread_start+8() So it stopped/crashed because of a division by zero in the dom0 kernel? You need to fix the kernel not to do a divizion by zero? -- Pasi
Yes - I try to find panic on dom0. I see differences in XEN_SYSCTL_cpu_hotplug implementation between xen-3.x and xen-4.x - was removed response on 4.x. --- Best regards, Igor Kozhukhov IRC# igork On 10/2/12 10:37 AM, "Pasi Kärkkäinen" <pasik@iki.fi> wrote:>On Mon, Oct 01, 2012 at 11:24:06PM +0400, Igor Kozhukhov wrote: >> Hello All, >> I try port xen-unstable to illumos-based platform. >> I have made some changes on illumos and built xen, but have panic - >>take a >> look [1]http://pastebin.com/fZSrExek >> I''d like to see some comments - what I need to check/update. > >kmdb: single-step stop on division by zero (#de) trap >kmdb: target stopped at: >sti+0x85: popq %r10 >[0]> $C >ffffff0003ed1c20 sti+0x85() >ffffff0003e05970 switch_sp_and_call+0x13() >ffffff0003e059d0 do_interrupt+0x14a() >ffffff0003e05a80 xen_callback_handler+0x42d(ffffff0003e05a90, >fffffffffbc1d778) >ffffff0003e05a90 xen_callback+0x18f() >ffffff0003e05ba0 __hypercall2+0xa() >ffffff0003e05bb0 HYPERVISOR_block+0x10() >ffffff0003e05bc0 mach_cpu_idle+0x1d() >ffffff0003e05bf0 cpu_idle+0xfe() >ffffff0003e05c00 cpu_idle_adaptive+0x13() >ffffff0003e05c20 idle+0x41() >ffffff0003e05c30 thread_start+8() > >So it stopped/crashed because of a division by zero in the dom0 kernel? >You need to fix the kernel not to do a divizion by zero? > >-- Pasi >