[PATCH v2] virt-host-validate: Allow longer list of CPU flags
by Michal Privoznik
On various occasions, virt-host-validate parses /proc/cpuinfo to
learn about CPU flags (see virHostValidateGetCPUFlags()). It does
so, by reading the file line by line until the line with CPU
flags is reached. Then the line is split into individual flags
(using space as a delimiter) and the list of flags is then
iterated over.
This works, except for cases when the line with CPU flags is too
long. Problem is - the line is capped at 1024 bytes and on newer
CPUs (and newer kernels), the line can be significantly longer.
I've seen a line that's ~1200 characters long (with 164 flags
reported).
Switch to unbounded read from the file (getline()).
Resolves: https://issues.redhat.com/browse/RHEL-39969
Signed-off-by: Michal Privoznik <mprivozn(a)redhat.com>
---
v2 of:
https://lists.libvirt.org/archives/list/devel@lists.libvirt.org/thread/LK...
diff to v2:
- Keep trimming of the optional newline
tools/virt-host-validate-common.c | 18 +++++++-----------
1 file changed, 7 insertions(+), 11 deletions(-)
diff --git a/tools/virt-host-validate-common.c b/tools/virt-host-validate-common.c
index 591143c24d..63cc3dbe7b 100644
--- a/tools/virt-host-validate-common.c
+++ b/tools/virt-host-validate-common.c
@@ -106,21 +106,19 @@ virBitmap *virHostValidateGetCPUFlags(void)
{
FILE *fp;
virBitmap *flags = NULL;
+ g_autofree char *line = NULL;
+ size_t linelen = 0;
if (!(fp = fopen("/proc/cpuinfo", "r")))
return NULL;
flags = virBitmapNew(VIR_HOST_VALIDATE_CPU_FLAG_LAST);
- do {
- char line[1024];
+ while (getline(&line, &linelen, fp) > 0) {
char *start;
g_auto(GStrv) tokens = NULL;
GStrv next;
- if (!fgets(line, sizeof(line), fp))
- break;
-
/* The line we're interested in is marked differently depending
* on the architecture, so check possible prefixes */
if (!STRPREFIX(line, "flags") &&
@@ -129,11 +127,9 @@ virBitmap *virHostValidateGetCPUFlags(void)
!STRPREFIX(line, "facilities"))
continue;
- /* fgets() includes the trailing newline in the output buffer,
- * so we need to clean that up ourselves. We can safely access
- * line[strlen(line) - 1] because the checks above would cause
- * us to skip empty strings */
- line[strlen(line) - 1] = '\0';
+ /* getline() may include the trailing newline in the output
+ * buffer, so we need to clean that up ourselves. */
+ virStringTrimOptionalNewline(line);
/* Skip to the separator */
if (!(start = strchr(line, ':')))
@@ -153,7 +149,7 @@ virBitmap *virHostValidateGetCPUFlags(void)
if ((value = virHostValidateCPUFlagTypeFromString(*next)) >= 0)
ignore_value(virBitmapSetBit(flags, value));
}
- } while (1);
+ }
VIR_FORCE_FCLOSE(fp);
--
2.44.2
1 month, 3 weeks
[PATCH] virt-host-validate: Allow longer list of CPU flags
by Michal Privoznik
On various occasions, virt-host-validate parses /proc/cpuinfo to
learn about CPU flags (see virHostValidateGetCPUFlags()). It does
so, by reading the file line by line until the line with CPU
flags is reached. Then the line is split into individual flags
(using space as a delimiter) and the list of flags is then
iterated over.
This works, except for cases when the line with CPU flags is too
long. Problem is - the line is capped at 1024 bytes and on newer
CPUs (and newer kernels), the line can be significantly longer.
I've seen a line that's ~1200 characters long (with 164 flags
reported).
Switch to unbounded read from the file (getline()).
Resolves: https://issues.redhat.com/browse/RHEL-39969
Signed-off-by: Michal Privoznik <mprivozn(a)redhat.com>
---
tools/virt-host-validate-common.c | 16 ++++------------
1 file changed, 4 insertions(+), 12 deletions(-)
diff --git a/tools/virt-host-validate-common.c b/tools/virt-host-validate-common.c
index 591143c24d..e5fa6606d2 100644
--- a/tools/virt-host-validate-common.c
+++ b/tools/virt-host-validate-common.c
@@ -106,21 +106,19 @@ virBitmap *virHostValidateGetCPUFlags(void)
{
FILE *fp;
virBitmap *flags = NULL;
+ g_autofree char *line = NULL;
+ size_t linelen = 0;
if (!(fp = fopen("/proc/cpuinfo", "r")))
return NULL;
flags = virBitmapNew(VIR_HOST_VALIDATE_CPU_FLAG_LAST);
- do {
- char line[1024];
+ while (getline(&line, &linelen, fp) > 0) {
char *start;
g_auto(GStrv) tokens = NULL;
GStrv next;
- if (!fgets(line, sizeof(line), fp))
- break;
-
/* The line we're interested in is marked differently depending
* on the architecture, so check possible prefixes */
if (!STRPREFIX(line, "flags") &&
@@ -129,12 +127,6 @@ virBitmap *virHostValidateGetCPUFlags(void)
!STRPREFIX(line, "facilities"))
continue;
- /* fgets() includes the trailing newline in the output buffer,
- * so we need to clean that up ourselves. We can safely access
- * line[strlen(line) - 1] because the checks above would cause
- * us to skip empty strings */
- line[strlen(line) - 1] = '\0';
-
/* Skip to the separator */
if (!(start = strchr(line, ':')))
continue;
@@ -153,7 +145,7 @@ virBitmap *virHostValidateGetCPUFlags(void)
if ((value = virHostValidateCPUFlagTypeFromString(*next)) >= 0)
ignore_value(virBitmapSetBit(flags, value));
}
- } while (1);
+ }
VIR_FORCE_FCLOSE(fp);
--
2.44.2
1 month, 3 weeks
[PATCH] rpc: report error from filing to add timer
by Daniel P. Berrangé
Signed-off-by: Daniel P. Berrangé <berrange(a)redhat.com>
---
src/rpc/virnetclientstream.c | 2 ++
1 file changed, 2 insertions(+)
diff --git a/src/rpc/virnetclientstream.c b/src/rpc/virnetclientstream.c
index 98034d737d..380b785869 100644
--- a/src/rpc/virnetclientstream.c
+++ b/src/rpc/virnetclientstream.c
@@ -725,6 +725,8 @@ int virNetClientStreamEventAddCallback(virNetClientStream *st,
virNetClientStreamEventTimer,
st,
virObjectUnref)) < 0) {
+ virReportError(VIR_ERR_INTERNAL_ERROR, "%s",
+ _("Unable to add timer to event loop"));
virObjectUnref(st);
goto cleanup;
}
--
2.45.2
1 month, 3 weeks
cpugroups for hyperv hypervisor
by Praveen K Paladugu
Hey folks,
My team is working on exposing `cpugroups` to Libvirt while using
'hyperv' hypervisor with cloud-hypervisor(VMM). cpugroups are relevant
in a specific configuration of hyperv called 'minroot'. In Minroot
configuration, hypervisor artificially restricts Dom0 to run on a subset
of cpus (Logical Processors). The rest of the cpus can be assigned to
guests.
cpugroups manage the CPUs assigned to guests and their scheduling
properties. Initially this looks similar to `cpuset` (in cgroups), but
the controls available with cpugroups don't map easily to those in
cgroups. For example:
* "IdleLPs" are the number of Logical Processors in a cpugroup, that
should be reserved to a guest even if they are idle
* "SchedulingPriority", the priority(values between 0..7) with which to
schedule CPUs in a cpugroup.
As controls like above don't easily map to anything in cgroups, using a
driver specific element in Domain xml, to configure cpugroups seems like
a right approach. For example:
<ch:cpugroups>
<idle_lps value='4'/>
<scheduling_priority value='6'/>
</ch:cpugroups>
As cpugroups is only relevant while using minroot configuration on
hyperv, I don't see any value in generalizing this setting. So, having
some "ch" driver specific settings seems like a good approach to
implement this feature.
Question1: Do you see any concerns with this approach?
The cpugroup settings can be applied/modified using sysfs interface or
using a cmdline tool on the host. I see Libvirt uses both these
mechanisms for various use cases. But, given a choice, sysfs based
interface seems like a simpler approach to me. With sysfs interface
Libvirt does not have to take install time dependencies on new tools.
Question2: Of "sysfs" vs "cmdline tool" which is preferred, given a choice?
Early feedback from the community will help us invest in the preferred
choices sooner than later. Thanks for your consideration.
References:
*
https://learn.microsoft.com/en-us/windows-server/virtualization/hyper-v/m...
--
Regards,
Praveen
1 month, 3 weeks
[PATCH] conf: virtiofs: add rlimit_nofile element
by Adam Julis
Add an element to configure the rlimit nofile size:
...
<binary>
<rlimit_nofile size='122333'/>
</binary>
...
Non-positive values are forbidden in 'domaincommon.rng'. Added separate
test file, created by modifying the 'vhost-user-fs-fd-memory.xml'.
Resolves: https://gitlab.com/libvirt/libvirt/-/issues/485
Signed-off-by: Adam Julis <ajulis(a)redhat.com>
---
docs/formatdomain.rst | 4 ++
src/conf/domain_conf.c | 12 +++++
src/conf/domain_conf.h | 1 +
src/conf/schemas/domaincommon.rng | 9 ++++
...vhost-user-fs-fd-rlimit.x86_64-latest.args | 34 +++++++++++++
.../vhost-user-fs-fd-rlimit.x86_64-latest.xml | 50 +++++++++++++++++++
.../vhost-user-fs-fd-rlimit.xml | 50 +++++++++++++++++++
tests/qemuxmlconftest.c | 1 +
8 files changed, 161 insertions(+)
create mode 100644 tests/qemuxmlconfdata/vhost-user-fs-fd-rlimit.x86_64-latest.args
create mode 100644 tests/qemuxmlconfdata/vhost-user-fs-fd-rlimit.x86_64-latest.xml
create mode 100644 tests/qemuxmlconfdata/vhost-user-fs-fd-rlimit.xml
diff --git a/docs/formatdomain.rst b/docs/formatdomain.rst
index ca33a2a570..10584dfe83 100644
--- a/docs/formatdomain.rst
+++ b/docs/formatdomain.rst
@@ -3734,6 +3734,10 @@ A directory on the host that can be accessed directly from the guest.
The thread pool helps increase the number of requests in flight when used with
storage that has a higher latency. However, it has an overhead, and so for
fast, low latency filesystems, it may be best to turn it off. ( :since:`Since 8.5.0` )
+ Element ``rlimit_profile`` accepts one attribute ``size`` which defines the
+ maximum number of file descriptors. Non-positive values are forbidden.
+ Although numbers greater than 1M are allowed, the virtiofsd documentation
+ states that in this case its set by virtiofsd to the 1M. ( :since:`Since 10.6.0` )
``source``
The resource on the host that is being accessed in the guest. The ``name``
attribute must be used with ``type='template'``, and the ``dir`` attribute
diff --git a/src/conf/domain_conf.c b/src/conf/domain_conf.c
index 32df908d95..6733857a3a 100644
--- a/src/conf/domain_conf.c
+++ b/src/conf/domain_conf.c
@@ -8868,6 +8868,7 @@ virDomainFSDefParseXML(virDomainXMLOption *xmlopt,
g_autofree char *queue_size = virXPathString("string(./driver/@queue)", ctxt);
g_autofree char *binary = virXPathString("string(./binary/@path)", ctxt);
g_autofree char *thread_pool_size = virXPathString("string(./binary/thread_pool/@size)", ctxt);
+ g_autofree char *rlimit_nofile = virXPathString("string(./binary/rlimit_nofile/@size)", ctxt);
xmlNodePtr binary_node = virXPathNode("./binary", ctxt);
xmlNodePtr binary_lock_node = virXPathNode("./binary/lock", ctxt);
xmlNodePtr binary_cache_node = virXPathNode("./binary/cache", ctxt);
@@ -8891,6 +8892,14 @@ virDomainFSDefParseXML(virDomainXMLOption *xmlopt,
goto error;
}
+ if (rlimit_nofile &&
+ virStrToLong_ull(rlimit_nofile, NULL, 10, &def->rlimit_nofile) < 0) {
+ virReportError(VIR_ERR_XML_ERROR,
+ _("cannot parse rlimit_nofile '%1$s' for virtiofs"),
+ rlimit_nofile);
+ goto error;
+ }
+
if (binary)
def->binary = virFileSanitizePath(binary);
@@ -23415,6 +23424,9 @@ virDomainFSDefFormat(virBuffer *buf,
if (def->thread_pool_size >= 0)
virBufferAsprintf(&binaryBuf, "<thread_pool size='%d'/>\n", def->thread_pool_size);
+ if (def->rlimit_nofile > 0)
+ virBufferAsprintf(&binaryBuf, "<rlimit_nofile size='%lld'/>\n", def->rlimit_nofile);
+
}
virDomainVirtioOptionsFormat(&driverAttrBuf, def->virtio);
diff --git a/src/conf/domain_conf.h b/src/conf/domain_conf.h
index 0fcc4f1f9b..8283493dfc 100644
--- a/src/conf/domain_conf.h
+++ b/src/conf/domain_conf.h
@@ -890,6 +890,7 @@ struct _virDomainFSDef {
bool symlinksResolved;
char *binary;
unsigned long long queue_size;
+ unsigned long long rlimit_nofile;
virTristateSwitch xattr;
virDomainFSCacheMode cache;
virTristateSwitch posix_lock;
diff --git a/src/conf/schemas/domaincommon.rng b/src/conf/schemas/domaincommon.rng
index 2d23fcf123..ab5374d5f0 100644
--- a/src/conf/schemas/domaincommon.rng
+++ b/src/conf/schemas/domaincommon.rng
@@ -3380,6 +3380,15 @@
</optional>
</element>
</optional>
+ <optional>
+ <element name="rlimit_nofile">
+ <optional>
+ <attribute name="size">
+ <data type="positiveInteger"/>
+ </attribute>
+ </optional>
+ </element>
+ </optional>
</interleave>
</element>
</define>
diff --git a/tests/qemuxmlconfdata/vhost-user-fs-fd-rlimit.x86_64-latest.args b/tests/qemuxmlconfdata/vhost-user-fs-fd-rlimit.x86_64-latest.args
new file mode 100644
index 0000000000..b4c2e3fe98
--- /dev/null
+++ b/tests/qemuxmlconfdata/vhost-user-fs-fd-rlimit.x86_64-latest.args
@@ -0,0 +1,34 @@
+LC_ALL=C \
+PATH=/bin \
+HOME=/var/lib/libvirt/qemu/domain--1-guest \
+USER=test \
+LOGNAME=test \
+XDG_DATA_HOME=/var/lib/libvirt/qemu/domain--1-guest/.local/share \
+XDG_CACHE_HOME=/var/lib/libvirt/qemu/domain--1-guest/.cache \
+XDG_CONFIG_HOME=/var/lib/libvirt/qemu/domain--1-guest/.config \
+/usr/bin/qemu-system-x86_64 \
+-name guest=guest,debug-threads=on \
+-S \
+-object '{"qom-type":"secret","id":"masterKey0","format":"raw","file":"/var/lib/libvirt/qemu/domain--1-guest/master-key.aes"}' \
+-machine pc,usb=off,dump-guest-core=off,acpi=off \
+-accel kvm \
+-cpu qemu64 \
+-m size=14680064k \
+-overcommit mem-lock=off \
+-smp 2,sockets=2,cores=1,threads=1 \
+-object '{"qom-type":"memory-backend-file","id":"ram-node0","mem-path":"/var/lib/libvirt/qemu/ram/-1-guest/ram-node0","share":true,"size":15032385536}' \
+-numa node,nodeid=0,cpus=0-1,memdev=ram-node0 \
+-uuid 126f2720-6f8e-45ab-a886-ec9277079a67 \
+-display none \
+-no-user-config \
+-nodefaults \
+-chardev socket,id=charmonitor,fd=1729,server=on,wait=off \
+-mon chardev=charmonitor,id=monitor,mode=control \
+-rtc base=utc \
+-no-shutdown \
+-boot strict=on \
+-chardev socket,id=chr-vu-fs0,path=/var/lib/libvirt/qemu/domain--1-guest/fs0-fs.sock \
+-device '{"driver":"vhost-user-fs-pci","id":"fs0","chardev":"chr-vu-fs0","queue-size":1027,"tag":"mount_tag","bus":"pci.0","addr":"0x2"}' \
+-audiodev '{"id":"audio1","driver":"none"}' \
+-sandbox on,obsolete=deny,elevateprivileges=deny,spawn=deny,resourcecontrol=deny \
+-msg timestamp=on
diff --git a/tests/qemuxmlconfdata/vhost-user-fs-fd-rlimit.x86_64-latest.xml b/tests/qemuxmlconfdata/vhost-user-fs-fd-rlimit.x86_64-latest.xml
new file mode 100644
index 0000000000..2983d3f275
--- /dev/null
+++ b/tests/qemuxmlconfdata/vhost-user-fs-fd-rlimit.x86_64-latest.xml
@@ -0,0 +1,50 @@
+<domain type='kvm'>
+ <name>guest</name>
+ <uuid>126f2720-6f8e-45ab-a886-ec9277079a67</uuid>
+ <memory unit='KiB'>14680064</memory>
+ <currentMemory unit='KiB'>14680064</currentMemory>
+ <memoryBacking>
+ <source type='file'/>
+ <access mode='shared'/>
+ </memoryBacking>
+ <vcpu placement='static'>2</vcpu>
+ <os>
+ <type arch='x86_64' machine='pc'>hvm</type>
+ <boot dev='hd'/>
+ </os>
+ <cpu mode='custom' match='exact' check='none'>
+ <model fallback='forbid'>qemu64</model>
+ <numa>
+ <cell id='0' cpus='0-1' memory='14680064' unit='KiB' memAccess='shared'/>
+ </numa>
+ </cpu>
+ <clock offset='utc'/>
+ <on_poweroff>destroy</on_poweroff>
+ <on_reboot>restart</on_reboot>
+ <on_crash>destroy</on_crash>
+ <devices>
+ <emulator>/usr/bin/qemu-system-x86_64</emulator>
+ <controller type='usb' index='0' model='none'/>
+ <controller type='pci' index='0' model='pci-root'/>
+ <filesystem type='mount' accessmode='passthrough'>
+ <driver type='virtiofs' queue='1027'/>
+ <binary xattr='on'>
+ <cache mode='always'/>
+ <sandbox mode='chroot'/>
+ <thread_pool size='16'/>
+ <rlimit_nofile size='122333'/>
+ </binary>
+ <idmap>
+ <uid start='0' target='100000' count='65535'/>
+ <gid start='0' target='100000' count='65535'/>
+ </idmap>
+ <source dir='/path'/>
+ <target dir='mount_tag'/>
+ <address type='pci' domain='0x0000' bus='0x00' slot='0x02' function='0x0'/>
+ </filesystem>
+ <input type='mouse' bus='ps2'/>
+ <input type='keyboard' bus='ps2'/>
+ <audio id='1' type='none'/>
+ <memballoon model='none'/>
+ </devices>
+</domain>
diff --git a/tests/qemuxmlconfdata/vhost-user-fs-fd-rlimit.xml b/tests/qemuxmlconfdata/vhost-user-fs-fd-rlimit.xml
new file mode 100644
index 0000000000..2983d3f275
--- /dev/null
+++ b/tests/qemuxmlconfdata/vhost-user-fs-fd-rlimit.xml
@@ -0,0 +1,50 @@
+<domain type='kvm'>
+ <name>guest</name>
+ <uuid>126f2720-6f8e-45ab-a886-ec9277079a67</uuid>
+ <memory unit='KiB'>14680064</memory>
+ <currentMemory unit='KiB'>14680064</currentMemory>
+ <memoryBacking>
+ <source type='file'/>
+ <access mode='shared'/>
+ </memoryBacking>
+ <vcpu placement='static'>2</vcpu>
+ <os>
+ <type arch='x86_64' machine='pc'>hvm</type>
+ <boot dev='hd'/>
+ </os>
+ <cpu mode='custom' match='exact' check='none'>
+ <model fallback='forbid'>qemu64</model>
+ <numa>
+ <cell id='0' cpus='0-1' memory='14680064' unit='KiB' memAccess='shared'/>
+ </numa>
+ </cpu>
+ <clock offset='utc'/>
+ <on_poweroff>destroy</on_poweroff>
+ <on_reboot>restart</on_reboot>
+ <on_crash>destroy</on_crash>
+ <devices>
+ <emulator>/usr/bin/qemu-system-x86_64</emulator>
+ <controller type='usb' index='0' model='none'/>
+ <controller type='pci' index='0' model='pci-root'/>
+ <filesystem type='mount' accessmode='passthrough'>
+ <driver type='virtiofs' queue='1027'/>
+ <binary xattr='on'>
+ <cache mode='always'/>
+ <sandbox mode='chroot'/>
+ <thread_pool size='16'/>
+ <rlimit_nofile size='122333'/>
+ </binary>
+ <idmap>
+ <uid start='0' target='100000' count='65535'/>
+ <gid start='0' target='100000' count='65535'/>
+ </idmap>
+ <source dir='/path'/>
+ <target dir='mount_tag'/>
+ <address type='pci' domain='0x0000' bus='0x00' slot='0x02' function='0x0'/>
+ </filesystem>
+ <input type='mouse' bus='ps2'/>
+ <input type='keyboard' bus='ps2'/>
+ <audio id='1' type='none'/>
+ <memballoon model='none'/>
+ </devices>
+</domain>
diff --git a/tests/qemuxmlconftest.c b/tests/qemuxmlconftest.c
index 389d31800b..2aaf6103db 100644
--- a/tests/qemuxmlconftest.c
+++ b/tests/qemuxmlconftest.c
@@ -2814,6 +2814,7 @@ mymain(void)
DO_TEST_CAPS_ARCH_LATEST("launch-security-s390-pv", "s390x");
DO_TEST_CAPS_LATEST("vhost-user-fs-fd-memory");
+ DO_TEST_CAPS_LATEST("vhost-user-fs-fd-rlimit");
DO_TEST_CAPS_LATEST("vhost-user-fs-hugepages");
DO_TEST_CAPS_LATEST_PARSE_ERROR("vhost-user-fs-readonly");
--
2.45.2
1 month, 3 weeks
[PATCH v2 0/1] Expose availability of SEV-ES
by Takashi Kajinami
This introduces the new "model" field in sev elements so that clients can
check whether SEV-ES, the 2nd generation of AMD SEV, is available in
the taget hyprvisor. There is the maxESGuests field (along with the maxGuests
field) but this field does not explain whether SEV-ES is actually
enabled in KVM.
Takashi Kajinami (1):
Expose available AMD SEV models in domain capabilities
Changes since v1:
* Fixed one code path where available models are not added
* Fixed missing update of "report" flag
* Updated the documentation to explain the new model field in addition
to the existing but undocumanted cpu0Id field
Takashi Kajinami (1):
Expose available AMD SEV models in domain capabilities
docs/formatdomaincaps.rst | 5 ++
src/conf/domain_capabilities.c | 2 +
src/conf/domain_capabilities.h | 1 +
src/conf/domain_conf.c | 7 +++
src/conf/domain_conf.h | 8 ++++
src/qemu/qemu_capabilities.c | 84 +++++++++++++++++++++++++---------
6 files changed, 85 insertions(+), 22 deletions(-)
--
2.43.0
1 month, 3 weeks
[PATCH 0/2] Add Granite Rapids cpu model
by Tim Wiederhake
Also, update the sync tool with new features introduced in qemu.
Tim Wiederhake (2):
sync_qemu_models_i386.py: Add missing features
cpu_map: Add GraniteRapids CPU model
src/cpu_map/index.xml | 1 +
src/cpu_map/meson.build | 1 +
src/cpu_map/sync_qemu_models_i386.py | 5 +
src/cpu_map/x86_GraniteRapids.xml | 200 ++++++++++++++++++
.../domaincapsdata/qemu_8.1.0-q35.x86_64.xml | 1 +
.../domaincapsdata/qemu_8.1.0-tcg.x86_64.xml | 1 +
tests/domaincapsdata/qemu_8.1.0.x86_64.xml | 1 +
.../domaincapsdata/qemu_8.2.0-q35.x86_64.xml | 1 +
.../domaincapsdata/qemu_8.2.0-tcg.x86_64.xml | 1 +
tests/domaincapsdata/qemu_8.2.0.x86_64.xml | 1 +
.../domaincapsdata/qemu_9.0.0-q35.x86_64.xml | 1 +
.../domaincapsdata/qemu_9.0.0-tcg.x86_64.xml | 1 +
tests/domaincapsdata/qemu_9.0.0.x86_64.xml | 1 +
.../domaincapsdata/qemu_9.1.0-q35.x86_64.xml | 1 +
.../domaincapsdata/qemu_9.1.0-tcg.x86_64.xml | 1 +
tests/domaincapsdata/qemu_9.1.0.x86_64.xml | 1 +
16 files changed, 219 insertions(+)
create mode 100644 src/cpu_map/x86_GraniteRapids.xml
--
2.43.0
1 month, 3 weeks
[PATCH] cpu_map: Add libcpuinfo as optional data source
by Tim Wiederhake
This adds an option to use libcpuinfo [1] as data source for
libvirt's list of x86 cpu features. This is purely optional and
does not change the script's behavior if libcpuinfo is not
installed.
libcpuinfo is a cross-vendor, cross-architecture source for CPU
related information that has the capability to replace libvirt's
dependence on qemu's cpu feature list.
[1] https://gitlab.com/twiederh/libcpuinfo
Signed-off-by: Tim Wiederhake <twiederh(a)redhat.com>
---
src/cpu_map/libcpuinfo_aliases.xml | 75 +++++++++++++++++++++++++
src/cpu_map/sync_qemu_features_i386.py | 77 +++++++++++++++++++++++---
2 files changed, 145 insertions(+), 7 deletions(-)
create mode 100644 src/cpu_map/libcpuinfo_aliases.xml
diff --git a/src/cpu_map/libcpuinfo_aliases.xml b/src/cpu_map/libcpuinfo_aliases.xml
new file mode 100644
index 0000000000..75d243fead
--- /dev/null
+++ b/src/cpu_map/libcpuinfo_aliases.xml
@@ -0,0 +1,75 @@
+<!--
+ libvirt uses slightly different names for some cpu features than other
+ software does. Install this file into libcpuinfo's alias directory
+ (e.g. /usr/share/libcpuinfo/aliases/libvirt.xml) to have libcpuinfo
+ automatically translate feature names into the names libvirt uses.
+-->
+
+<external_aliases>
+ <external_alias>
+ <type>feature</type>
+ <canonical>pclmulqdq</canonical>
+ <name>pclmuldq</name>
+ <domain>libvirt</domain>
+ </external_alias>
+ <external_alias>
+ <type>feature</type>
+ <canonical>ds-cpl</canonical>
+ <name>ds_cpl</name>
+ <domain>libvirt</domain>
+ </external_alias>
+ <external_alias>
+ <type>feature</type>
+ <canonical>sse4-1</canonical>
+ <name>sse4.1</name>
+ <domain>libvirt</domain>
+ </external_alias>
+ <external_alias>
+ <type>feature</type>
+ <canonical>sse4-2</canonical>
+ <name>sse4.2</name>
+ <domain>libvirt</domain>
+ </external_alias>
+ <external_alias>
+ <type>feature</type>
+ <canonical>tsc-adjust</canonical>
+ <name>tsc_adjust</name>
+ <domain>libvirt</domain>
+ </external_alias>
+ <external_alias>
+ <type>feature</type>
+ <canonical>lahf-lm</canonical>
+ <name>lahf_lm</name>
+ <domain>libvirt</domain>
+ </external_alias>
+ <external_alias>
+ <type>feature</type>
+ <canonical>cmp-legacy</canonical>
+ <name>cmp_legacy</name>
+ <domain>libvirt</domain>
+ </external_alias>
+ <external_alias>
+ <type>feature</type>
+ <canonical>nodeid-msr</canonical>
+ <name>nodeid_msr</name>
+ <domain>libvirt</domain>
+ </external_alias>
+ <external_alias>
+ <type>feature</type>
+ <canonical>perfctr-core</canonical>
+ <name>perfctr_core</name>
+ <domain>libvirt</domain>
+ </external_alias>
+ <external_alias>
+ <type>feature</type>
+ <canonical>perfctr-nb</canonical>
+ <name>perfctr_nb</name>
+ <domain>libvirt</domain>
+ </external_alias>
+ <external_alias>
+ <type>feature</type>
+ <canonical>fxsr-opt</canonical>
+ <name>fxsr_opt</name>
+ <domain>libvirt</domain>
+ </external_alias>
+</external_aliases>
diff --git a/src/cpu_map/sync_qemu_features_i386.py b/src/cpu_map/sync_qemu_features_i386.py
index e4b1f7275a..3b3ad5a643 100755
--- a/src/cpu_map/sync_qemu_features_i386.py
+++ b/src/cpu_map/sync_qemu_features_i386.py
@@ -4,6 +4,11 @@ import argparse
import os
import re
+try:
+ import pycpuinfo
+except ImportError:
+ pycpuinfo = None
+
# features in qemu that we do not want in libvirt
FEATURES_IGNORE = (
@@ -22,6 +27,7 @@ FEATURES_IGNORE = (
"kvm-steal-time",
"kvmclock",
"kvmclock-stable-bit",
+ "kvmclock2",
"xstore",
"xstore-en",
@@ -295,6 +301,53 @@ def add_feature_qemu(query, data):
add_feature_cpuid(eax, ecx, reg, bit, name)
+def add_features_cpuinfo():
+ def decode_bit(value):
+ for i in range(0, 64):
+ if value == (1 << i):
+ return i
+
+ def decode_cpuid(v):
+ if v[0] != 0 and v[1] == 0 and v[2] == 0 and v[3] == 0:
+ reg, val = "eax", v[0]
+ if v[0] == 0 and v[1] != 0 and v[2] == 0 and v[3] == 0:
+ reg, val = "ebx", v[1]
+ if v[0] == 0 and v[1] == 0 and v[2] != 0 and v[3] == 0:
+ reg, val = "ecx", v[2]
+ if v[0] == 0 and v[1] == 0 and v[2] == 0 and v[3] != 0:
+ reg, val = "edx", v[3]
+
+ return reg, decode_bit(val)
+
+ x86 = pycpuinfo.Family.find("x86", "")
+
+ for feature in pycpuinfo.features():
+ if feature.family() != x86:
+ continue
+
+ if list(feature.features()):
+ continue
+
+ name = feature.name("libvirt")
+ if name in FEATURES_IGNORE:
+ continue
+
+ cpuid = feature.extra_x86_cpuid()
+ if cpuid:
+ eax = cpuid[0]
+ ecx = cpuid[1]
+ if ecx == pycpuinfo.x86.CPUINFO_X86_CPUID_ECX_NONE:
+ ecx = None
+ reg, bit = decode_cpuid(cpuid[2:])
+ add_feature_cpuid(eax, ecx, reg, bit, name)
+
+ msr = feature.extra_x86_msr()
+ if msr:
+ index = msr[0]
+ bit = decode_bit(msr[1] | (msr[2] << 32))
+ add_feature_msr(index, bit, name)
+
+
# read the `feature_word_info` struct from qemu's cpu.c into a list of strings
def read_cpu_c(path):
pattern_comment = re.compile("/\\*.*?\\*/")
@@ -450,6 +503,12 @@ def main():
nargs="?",
type=os.path.realpath,
)
+ if pycpuinfo:
+ parser.add_argument(
+ "--libcpuinfo",
+ help="Use libcpuinfo as data source instead",
+ action="store_true",
+ )
parser.add_argument(
"--output",
"-o",
@@ -459,14 +518,18 @@ def main():
)
args = parser.parse_args()
- if not os.path.isdir(args.qemu):
- parser.print_help()
- exit("qemu source directory not found")
+ if pycpuinfo and args.libcpuinfo:
+ add_features_cpuinfo()
+ else:
+ if not os.path.isdir(args.qemu):
+ parser.print_help()
+ exit("qemu source directory not found")
+
+ read_headers(args.qemu)
+ lines = read_cpu_c(args.qemu)
+ parse_feature_words(lines)
+ add_extra_features()
- read_headers(args.qemu)
- lines = read_cpu_c(args.qemu)
- parse_feature_words(lines)
- add_extra_features()
write_output(args.output)
print(
--
2.43.0
1 month, 3 weeks
[PATCH] formatstorage: Document qcow2 default version change
by Peter Krempa
Based on discussion after commit f432114d9c was pushed it was pointed
out that the documentation still mentions the older version.
Fix the documentation to state the new version and introduce ambiguity
for future updates.
Fixes: f432114d9cf507a4047aa9dc1344b1c13356db08
Signed-off-by: Peter Krempa <pkrempa(a)redhat.com>
---
Posting this documentation update to document what happened rather than
introduce (almost pointless) complication in adding a config file which
is unlikely to be ever used.
docs/formatstorage.rst | 13 +++++++++----
1 file changed, 9 insertions(+), 4 deletions(-)
diff --git a/docs/formatstorage.rst b/docs/formatstorage.rst
index 86e167d9cb..9d9a4143eb 100644
--- a/docs/formatstorage.rst
+++ b/docs/formatstorage.rst
@@ -700,10 +700,15 @@ host filesystem. It can contain the following child elements:
Encryption <formatstorageencryption.html>`__ page for more information.
``compat``
Specify compatibility level. So far, this is only used for ``type='qcow2'``
- volumes. Valid values are ``0.10`` and ``1.1`` so far, specifying QEMU
- version the images should be compatible with. If the ``feature`` element is
- present, 1.1 is used. :since:`Since 1.1.0` If omitted, 0.10 is used.
- :since:`Since 1.1.2`
+ volumes. Valid values are ``0.10`` (QCOW2 v2) and ``1.1`` (QCOW2 v3) so far.
+ The values were meant to specify QEMU version the images should be compatible
+ with.
+
+ The default, if the ``feature`` element is present is ``1.1``. :since:`Since 1.1.0`
+ If ``feature`` is not present, ``0.10`` was used :since:`Since 1.1.2` and
+ :since:`Since 10.2.0` ``1.1`` is used as it's the default of ``qemu-img``.
+
+ Any tool depending on a specific version should specify this field explicitly.
``nocow``
Turn off COW of the newly created volume. So far, this is only valid for a
file image in btrfs file system. It will improve performance when the file
--
2.44.0
1 month, 3 weeks
Plans for 10.6.0 release (freeze on Monday 29 Jul)
by Jiri Denemark
We are getting close to 10.6.0 release of libvirt. Since I'll be away
from computers next week, I suggest moving the "last week in a month"
freeze a bit and aim for the release on Monday 05 Aug after entering the
freeze on Monday 29 Jul. There will be no RC2 this time.
I hope this works for everyone.
Jirka
1 month, 3 weeks
