[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

[qemu-xen staging-4.13] pc: Don't make die-id mandatory unless necessary

commit 7ebcd375ade505358c1c45542de22f188c599bdd
Author:     Eduardo Habkost <ehabkost@xxxxxxxxxx>
AuthorDate: Fri Aug 16 14:07:50 2019 -0300
Commit:     Michael Roth <mdroth@xxxxxxxxxxxxxxxxxx>
CommitDate: Sun Oct 27 23:41:00 2019 -0500

    pc: Don't make die-id mandatory unless necessary
    We have this issue reported when using libvirt to hotplug CPUs:
    Basically, libvirt is not copying die-id from
    query-hotpluggable-cpus, but die-id is now mandatory.
    We could blame libvirt and say it is not following the documented
    interface, because we have this buried in the QAPI schema
    > Note: currently there are 5 properties that could be present
    > but management should be prepared to pass through other
    > properties with device_add command to allow for future
    > interface extension. This also requires the filed names to be kept in
    > sync with the properties passed to -device/device_add.
    But I don't think this would be reasonable from us.  We can just
    make QEMU more flexible and let die-id to be omitted when there's
    no ambiguity.  This will allow us to keep compatibility with
    existing libvirt versions.
    Test case included to ensure we don't break this again.
    Fixes: commit 176d2cda0dee ("i386/cpu: Consolidate die-id validity in smp 
    Signed-off-by: Eduardo Habkost <ehabkost@xxxxxxxxxx>
    Message-Id: <20190816170750.23910-1-ehabkost@xxxxxxxxxx>
    Signed-off-by: Eduardo Habkost <ehabkost@xxxxxxxxxx>
    (cherry picked from commit fea374e7c8079563bca7c8fac895c6a880f76adc)
    Signed-off-by: Michael Roth <mdroth@xxxxxxxxxxxxxxxxxx>
 hw/i386/pc.c                             |  8 ++++++++
 tests/acceptance/pc_cpu_hotplug_props.py | 35 ++++++++++++++++++++++++++++++++
 2 files changed, 43 insertions(+)

diff --git a/hw/i386/pc.c b/hw/i386/pc.c
index 549c437050..947f81070f 100644
--- a/hw/i386/pc.c
+++ b/hw/i386/pc.c
@@ -2403,6 +2403,14 @@ static void pc_cpu_pre_plug(HotplugHandler *hotplug_dev,
         int max_socket = (ms->smp.max_cpus - 1) /
                                 smp_threads / smp_cores / pcms->smp_dies;
+        /*
+         * die-id was optional in QEMU 4.0 and older, so keep it optional
+         * if there's only one die per socket.
+         */
+        if (cpu->die_id < 0 && pcms->smp_dies == 1) {
+            cpu->die_id = 0;
+        }
         if (cpu->socket_id < 0) {
             error_setg(errp, "CPU socket-id is not set");
diff --git a/tests/acceptance/pc_cpu_hotplug_props.py 
new file mode 100644
index 0000000000..08b7e632c6
--- /dev/null
+++ b/tests/acceptance/pc_cpu_hotplug_props.py
@@ -0,0 +1,35 @@
+# Ensure CPU die-id can be omitted on -device
+#  Copyright (c) 2019 Red Hat Inc
+# Author:
+#  Eduardo Habkost <ehabkost@xxxxxxxxxx>
+# This library is free software; you can redistribute it and/or
+# modify it under the terms of the GNU Lesser General Public
+# License as published by the Free Software Foundation; either
+# version 2 of the License, or (at your option) any later version.
+# This library is distributed in the hope that it will be useful,
+# but WITHOUT ANY WARRANTY; without even the implied warranty of
+# Lesser General Public License for more details.
+# You should have received a copy of the GNU Lesser General Public
+# License along with this library; if not, see <http://www.gnu.org/licenses/>.
+from avocado_qemu import Test
+class OmittedCPUProps(Test):
+    """
+    :avocado: tags=arch:x86_64
+    """
+    def test_no_die_id(self):
+        self.vm.add_args('-nodefaults', '-S')
+        self.vm.add_args('-smp', '1,sockets=2,cores=2,threads=2,maxcpus=8')
+        self.vm.add_args('-cpu', 'qemu64')
+        self.vm.add_args('-device', 
+        self.vm.launch()
+        self.assertEquals(len(self.vm.command('query-cpus')), 2)
generated by git-patchbot for /home/xen/git/qemu-xen.git#staging-4.13



Lists.xenproject.org is hosted with RackSpace, monitoring our
servers 24x7x365 and backed by RackSpace's Fanatical Support®.