faulting memory access was to/from an implicit memory operand
(typically either an instruction fetch or stack access).
Rationalise use of macros for page fault error code flags.
This patch supercedes the fix in changeset 11242.
Signed-off-by: Keir Fraser <keir@xensource.com>
correct tool to use for hardware perfctr monitoring now.
Also remove unused xc_msr MSR accessor functions from libxenctrl.
Signed-off-by: Keir Fraser <keir@xensource.com>
ELF notes provide a cleaner interface for passing Xen specific
information from the kernel to the domain builder than the existing
__xen_guest section string. The __xen_guest string is retained in
kernels built with 3.0.2 compatability for domU however dom0 requires
the new interface.
Signed-off-by: Ian Campbell <ian.campbell@xensource.com>
(in fact flush_all_pending_eoi() was broken!).
Based on a patch from Steven Rostedt <srostedt@redhat.com>
Signed-off-by: Keir Fraser <keir@xensource.com>
This may allow guest kernels to be run outside ring 3 in future, and
also provides scope for optimisations today (e.g., using global bit on
user mappings).
Signed-off-by: Jun Nakajima <jun.nakajima@intel.com>
Signed-off-by: Keir Fraser <keir@xensource.com>
This makes the script more idempotent if run twice (e.g. from hotplug and udev), since the second time the device will be taken down and not brought up again, leaving the host without networking.
Signed-off-by: Anil Madhavapeddy <anil@xensource.com>
i386 version is taken from 2.6.18-rc4-mm1 and x86_64 version has been
submitted for inclusion.
Signed-off-by: Ian Campbell <ian.campbell@xensource.com>
with for_each_cpu() -- we want to ensure that per_cpu areas are
accessed only for cpus in cpu_possible_map.
Signed-off-by: Keir Fraser <keir@xensource.com>
When running 'xm info' it appears thus:
# xm info
host : 10.13.4.252
release : 2.6.17-1.2157_FC5xen0
version : #1 SMP Tue Jul 11 23:30:05 EDT 2006
machine : x86_64
...snipped rest of data...
cc_compile_domain : build.redhat.com
cc_compile_date : Tue Jul 11 22:50:42 EDT 2006
xend_config_format : 2
It will also appear in the SEXPR returned when doing a GET on the
/xend/node path in XenD's HTTP service:
(node
(system Linux)
(host 10.13.4.252)
(release 2.6.17-1.2157_FC5xen0)
(version '#1 SMP Tue Jul 11 23:30:05 EDT 2006')
(machine x86_64)
...snipped rest of data...
(cc_compile_domain build.redhat.com)
(cc_compile_date 'Tue Jul 11 22:50:42 EDT 2006')
(xend_config_format 2)
)
The patch sets the format to value '2' to indicate that we're on the version
with the new style HVM configuration for CDROM devices. The old style having
implicitly been version '1'.
Signed-off-by: Daniel P. Berrange <berrange@redhat.com>
qemu 0.8.2 has a flush callback to the storage backends, so now it is
possible to implement hdparm tunable IDE write cache enable/disable for
guest domains, allowing people to pick speed or data consistency on a
case by case basis.
As an added benefit, really large LBA48 IOs will now no longer be broken
up into smaller IOs on the host side.
From: Rik van Riel <riel@redhat.com>
Signed-off-by: Christian Limpach <Christian.Limpach@xensource.com>
The rombios supports trying to boot from more than one device and then
falling back. Set 'boot=dc' in your config file to try booting first
from the CD and then the hard drive.
Based on a patch from: Jeremy Katz <katzj@redhat.com>
Signed-off-by: Christian Limpach <Christian.Limpach@xensource.com>
fields into separate shadow2-specific structures.
Also rename shadow2_entry_points to shadow2_paging_mode.
Remove VCPUF_shadow2_translate_mode and replace with a better-named
field in vcpu.arch.shadow2
Signed-off-by: Keir Fraser <keir@xensource.com>
In particular: x86/64 *does* clobber its ragument registers
(just like i386). This happens in debug builds, and if a
continuation is created.
Signed-off-by: Keir Fraser <keir@xensource.com>
On debug register accesses, the EIP should not be updated.
Because of the way that lazy save/restore of the debug
registers is implemented, this initial debug register access
is lost.
Signed-off-by: Travis Betak <travis.betak@amd.com>
Again needed for domU kexec, so the next kernel instance can register
watches without running into -EBUSY.
Signed-off-by: Gerd Hoffmann <kraxel@suse.de>
and handle frontend state transition from "Closing" to "Initialising".
Instead of destroying the backend device then they wait for a
reconnect from the frontend. Needed for domU kexec support, to hand over
devices from one kernel instance to the next.
Signed-off-by: Gerd Hoffmann <kraxel@suse.de>
