Our IPF machines are the following configurations.
- Total logical cpus = 16
-- 4 sockets
-- 2 cores
-- 2 threads
I changed the setting of BIOS, and invalidated the hyperthread, and I built Xen with NR_CPUS=8.
As a result, it became like attached file xendmesg1.txt.
- Available cpus 3
- Total cpus 8
Next, I built Xen with NR_CPUS=16.
As a result, it became like attached file xendmesg2.txt.
- Available cpus 8
- Total cpus 16
I thought not to match the analysis of the lsapic entry to the value of NR_CPUS.
It is an outline of patch as follows.
1. Count up the lsapic entry by using acpi_table_count_madt()/acpi_table_count_madt_family().
2. Call acpi_parse_lsapic() by using the number of lsapic entries.
3. Count up the available_cpus by using acpi_parse_lsapic(), however NR_CPUS is not exceeded.
Signed-off-by: Masaki Kanno <kanno.masaki@jp.fujitsu.com>
Best Regards,
Kan
the exact one updated by scheduler. Or else easy to think how things get
messed under SMP environment. One phenomenon observed is about softirq.
Exit path on CPU1 checks pending indicator by local_cpu_data, and then
goto do_softirq if pending true. However do_softirq uses cpu_data(cpu)
to query pending bit again. Cpu index is aquired from smp_processor_id,
with thread_info->cpu not cared by scheduler. Finally do_softirq on CPU1
tries to operate percpu data on CPU0 then.
Also disable several prints within critical path, like sending IPI. Due to
slow serial speed, this will slow down overall performance heavily since
event channels among CPUs are in traffic.
This patch is necessary for host SMP.
Signed-off-by Kevin Tian <kevin.tian@intel.com>
Signed-off-by Anthony Xu <anthony.xu@intel.com>
Signed-off-by Kevin Tian <kevin.tian@intel.com>
The reason is that Xen/ia64 will try to pend guest timer interrupt to dom0
within each machine timer interrupt handler. To avoid duplicated delivery,
domain_itm_last recorded domain_itm of last injection. If two are identical,
it means interrupt injected but guest has not set new itm value. Or else
corresponding pending irr bit will be turned on. That works in most cases.
However currently guest linux may try to set itm multiple times within one
handler before turn on psr.i again. Among first few settings, new guest timer
interrupt may be pended. Then once guest linux enables interrupt, a new
timer interrupt will be injected immediately. However this injection may have
itc still smaller than the domain_itm set at the last round of last handle.
In this case, guest linux will set same domain_itm as last again. However
since domain_itm_last already equals to domain_itm at last read ivr, later
no guest timer interrupt can be injected any more since Xen always thinks
an instance already injected without guest's response.
Attahced patch fixed this issue by adding sanity check upon guest timer
vector when deciding to inject interrupt into dom0. We always compare current
itc with latest domain_itm that guest really wants. So if itc < domain_itm
at this point, we simply clear the pending bit and no injection.
There's also a small fix to vcpu_read_ivr, where domain_itm_last should
be updated before clearing irr bit. Or els a small window still remains
to add a duplicate interrupt.
There is a case where HOSTCC doesn't accept -Wdeclaration-after-statement,
while CC does. It results in a compilation error.
This patch fixes it
Signed-off-by: Isaku Yamahata <yamahata@valinux.co.jp>
that it is clear they are fully GPL-compatible and satisfy the free
software definitions in include/linux/module.h.
Signed-off-by: Keir Fraser <keir@xensource.com>
Signed-off-by: Stephen Tweedie <sct@redhat.com>
This reverts part of the previous changeset, which prevented
console=tty0 from doing the obvious thing in a domU guest
(i.e., send output on the virtual console). This is because
the tty index ends up being 1.
I don't think setting the index to other than -1 really
buys us anything anyway. Any other driver that specifies
a name that conflicts with us will almost certainly set their
index to -1, so we cannot avoid conflicts by choosing a
more restrictive (singleton) value.
Signed-off-by: Keir Fraser <keir@xensource.com>
index is identical to the device file index. For us, the tty index is
always zero regardless of the device file index.
Signed-off-by: Alex Williamson <alex.williamson@hp.com>
happens when Xend crashes, so retrying once allows it to restart and recover.
If we see Xend crash twice, we give up, but throw a more polite exception than
the BadStatusLine splat that we got previously.
Signed-off-by: Ewan Mellor <ewan@xensource.com>
versions of Xen, and allows one to name a vif when it is created. This is
useful for integration with IP accounting packages, for example.
Closes bug #439.
Signed-off-by: Ewan Mellor <ewan@xensource.com>
abstract base class, which was never going to work. (This is debugging code
only, so I presume that no-one has used it for a long time, but it's doing no
harm hanging around.)
Signed-off-by: Ewan Mellor <ewan@xensource.com>
This will simplify the way we differentiate between HV and PV domains,
as well as make it easier to run normal tests in either HV or PV mode.
This patch has been modified by Ewan Mellor, to match his recent changes to
remove the nics= configuration option.
Signed-off-by: Dan Smith <danms@us.ibm.com>
Signed-off-by: Dan Stekloff <dsteklof@us.ibm.com>
grant_table.c. We access "uop->frame_list" (a pointer) directly, rather
than using "op.frame_list".
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
Signed-off-by: Tony Breeds <tony@bakeyournoodle.com>
creation/generation of the XML security policy files for the Xen ACM
security architecture. It is a python- based, web-based tool named
xensec_gen that allows users to create or modify XML policy files
through a browser. The resulting XML policy files can then be copied
or moved to the appropriate location in the /etc/xen/acm-security
directory structure in order to be translated into binary and used
within the Xen system.
Signed-off-by: Tom Lendacky <toml@us.ibm.com>
On slow network, when creating a vmx guest remotely,
the key repeat feature may make it hard to even type
the user name and password to login in to guest OS.
Signed-off-by: Eddie Dong <eddie.dong@intel.com>
Signed-off-by: Xiaofeng Ling <xiaofeng.ling@intel.com>
entries in vif to guide the configuration. This is much less confusing.
Closes bug #440.
Change the example configuration files to match. Also change them to use the
XenSource OUI in the MAC addresses. Also change xm-test to match.
Remove the obsolete ipaddr configuration entry, and the backend_mac vif config
option.
Remove the preprocess_vifs function, folding it into the configure_vifs
function, and creating a simple comma_sep_kv_to_dict helper.
Remove the configure_vfr method, as it is unused.
Signed-off-by: Ewan Mellor <ewan@xensource.com>
