Stop coercing drivers/disktype code into the tool stack. Make both
blktapctrl and tap-ctl transfer type/path pairs as "<type>:<path>"
strings. Remove the message.disktype integer altogether.
Signed-off-by: Daniel Stodden <daniel.stodden@citrix.com>
Signed-off-by: Jake Wires <jake.wires@citrix.com>
Implement xen_platform_pci support in xl/libxl to allow users to
disable the xen platform pci device from the VM config file.
Signed-off-by: Stefano Stabellini <stefano.stabellini@eu.citrix.com>
Now the init script is called xencommons and, as the name suggests, it
is a common script between xl and xend because it is used to start
xenconsoled, xenstored and xenbackendd and the initialization of these
three daemons has been removed from xend. The global
xen_toolstack=xl/xend variable has been removed. Regarding the
network setup, I made the vif scripts follow the same pattern as the
other scripts in xen-backend: a new script called vif-setup is
executed unconditionally. vif-setup is going to do the right thing
depending on the value of the environmental variable "script" (same
technique used before), defaulting to vif-bridge. In the common
scenario the toolstack doesn't need to set the variable "script"
because vif-bridge is going to be called anyway. There is no global
network script to setup the network bridges with xl, so if you are
using xl without xend, you need to manually configure the bridges
using your distro network setup, or, if you are lazy, you can just add
something like:
/etc/xen/scripts/network-bridge start
to your rc.local.
Signed-off-by: Stefano Stabellini <stefano.stabellini@eu.citrix.com>
Several changes to get the extended MCA information:
a) Use the x86_mcinfo_reserve in mcinfo_extended to reserve buffer
from
mc_info, instead of using the stack
b) For intel's extended MSR, we don't need write them one
by one as the MSR are continous
c) We don't need enum mca_extinfo, since we can consider
the extended MSR as either per bank, or global. Currently
we add a hook in global data collection, and didn't call
register intel_get_extended_msrs as callback. Later that
hook can be replaced by cleaner way
Signed-off-by: Jiang, Yunhong <yunhong.jiang@inte.com>
xen-hptool cannot be built on ia64, because xen-hptool depends on
xc_offline_page.c and xc_offline_page.c does not support ia64.
This patch disables it on ia64.
Signed-off-by: KUWAMURA Shin'ya <kuwa@jp.fujitsu.com>
xl help 'command' should give the help of the 'command',
but current it does not, it just gives a full list of xl
supported commands. This patch fix it.
Signed-off-by: Yang Hongyang <yanghy@cn.fujitsu.com>
1. Bug fix for error: "Error: Device /dev/xvdp (51952, tap2) is
already connected." (xenstore does not clean after DomU stoped)
2. Bug fix for error: "File 'vhd:/path/.../disk.img' doesn't exist."
(not correct parsing)
3. Bug fix for error: "Error: Device 51952 not connected" (in config
file for DomU we should be use prefix "tap2:tapdisk:xxx" for devices
from (aio, ram, qcow, vhd, remus) or "tap:tapdisk:xxx" for devices
from (sync, vmdk, qcow2, ioemu))
4. Bug fix for error: "Disk is not accessible" (if use 'tap2'-device
type, then '/dev/xpvd' may not be accessible immediately after its
creation)
Signed-off-by: eXeC001er <execooler@gmail.com>
this patch adds an init script for xl, so that it can be used without
xend installed on the system. In cases in which both toolstacks are
present, only one init script will be run, dependening on the value of
xen_toolstack in /etc/sysconfig/xen_toolstack. Please note that a
correct value of xen_toolstack (either xl or xend), is required now.
The network script and the vif script to be used with xl can be
specified in /etc/sysconfig/xenlightdaemons. Xl sets the vif script
to xl-vif-script by default, that reads /etc/sysconfig/xenlightdaemons
and calls the vif script set there. A port to a debian system should
be straightforward, just replacing /etc/sysconfig with /etc/default
everywhere.
Signed-off-by: Stefano Stabellini <stefano.stabellini@eu.citrix.com>
Each domain is allowed to set, reset and disable its timers; when any
timer runs out the domain is killed.
Patch from Christian Limpach <Christian.Limpach@citrix.com>
Signed-off-by: Tim Deegan <Tim.Deegan@citrix.com>
This simple patch enables hap by default in xl and also allows xl to
parse the hap option in the VM config file.
Signed-off-by: Stefano Stabellini <stefano.stabellini@eu.citrix.com>
...because the default timesource (TSC) drifts under load.
The flag is only defined in ACPI 3.0, and we provide ACPI 2.0 tables,
but Windows seems happy enough to obey it anyway.
Signed-off-by: Tim Deegan <Tim.Deegan@citrix.com>
If xen's time leaps forward by a large amount, the RTC will try to
model a tick for every second that it thinks has passed. This can
livelock a CPU with a series of timer requests each of which fires
immediately and requests the next one.
This patch treats a delay of more than a day between ticks as a
special case, abandoning the attempt to catch up. That should be good
enough to avoid livelock but doesn't fix the underlying time problem.
Signed-off-by: Tim Deegan <Tim.Deegan@citrix.com>
RHEL4 guests take a fair number of spurious pagefaults on Nehalem; no
need to report each one of them at XENLOG_WARNING.
Signed-off-by: Tim Deegan <Tim.Deegan@citrix.com>
Xapi tends to trigger this a lot by looking up the shadow memory
allocation of domains that aren't quite built yet.
Signed-off-by: Tim Deegan <Tim.Deegan@citrix.com>
Older (<4.6) RHEL kernels clear PAE pagetable entries without taking
care to write the bottom half first, often enough to make it annoying.
Signed-off-by: Tim Deegan <Tim.Deegan@citrix.com>
Its cpu field might be stale and refer to an offlined cpu.
Furthermore, this avoids performing invalid operations on killed
timers.
Signed-off-by: Keir Fraser <keir.fraser@citrix.com>
Useful for Windows guests, since the PV drivers are notified that
the domain is about to crash just before the crash dump gets written.
Signed-off-by: Tim Deegan <Tim.Deegan@citrix.com>
Signed-off-by: Keir Fraser <keir.fraser@citrix.com>
Otherwise timer_lock() can race a timer's ->cpu field being updated by
migrate_timers_from_cpu().
The rest of this patch is all debug paranoia.
Signed-off-by: Keir Fraser <keir.fraser@citrix.com>
Keir Fraser