- Devel - libvirt List Archives

[libvirt] [PATCH 0/4] apparmor fixes triggered by multi disk snapshots
by Christian Ehrhardt 21 Nov '19

21 Nov '19

Hi, the bugs [1][2] that made me debug into this actually only need the last patch (one line), but while coming along I found several opportunities for minor improvements of the apparmor code in libvirt. But that way it became a 4 patch series around apparmor. [1]: https://bugzilla.redhat.com/show_bug.cgi?id=1746684 [2]: https://bugs.launchpad.net/libvirt/+bug/1845506 Christian Ehrhardt (4): virt-aa-helper: clarify command line options apparmor: drop useless call to get_profile_name apparmor: refactor AppArmorSetSecurityImageLabel apparmor: let AppArmorSetSecurityImageLabel append rules src/security/security_apparmor.c | 52 +++++++------------------------- src/security/virt-aa-helper.c | 14 +++++---- 2 files changed, 19 insertions(+), 47 deletions(-) -- 2.23.0

3 15

[libvirt] [PATCH 0/2] virt-aa-helper for shmem device
by Christian Ehrhardt 21 Nov '19

21 Nov '19

Cole was recently adding a few of the usual apparmor suspects to BZ 1761645 and I was taking a look at the low hanging fruits of it today. It isn't perfect, but would resolve the reported issue - so I'd appreciate a review. Limitations: - One could break the path creating elements in qemuBuildShmemBackendMemProps and qemuDomainPrepareShmemChardev into extra functions and then use those from virt-aa-helper. But I haven't done so yet and unless it is strictly required consider it too much for what we want/need to achieve here. - I haven't covered hotplug of shmem devices yet, it seems there is no infrastructure for their labels yet and I wasn't sure how important shmem-hotplug would even be to anyone. Christian Ehrhardt (2): virt-aa-helper: add rules for shmem devices virt-aa-helper: testcase for shmem devices src/security/virt-aa-helper.c | 35 +++++++++++++++++++++++++++++++++++ tests/virt-aa-helper-test | 15 +++++++++++++++ 2 files changed, 50 insertions(+) -- 2.23.0

3 18

[libvirt] [patch 1/1] virt-aa-helper: Add support for smartcard host-certificates
by Arnaud Patard 20 Nov '19

20 Nov '19

When emulating smartcard with host certificates, qemu needs to be able to read the certificates files, which is denied by apparmor. Add necessary code to add the smartcard certificates related directory to the apparmor profile. This code supports only this case smartcard 'host' and 'passthrough' settings are not supported, as I can't test them. Signed-off-by: Arnaud Patard <apatard(a)hupstream.com> Index: libvirt-5.0.0/src/security/virt-aa-helper.c =================================================================== --- libvirt-5.0.0.orig/src/security/virt-aa-helper.c +++ libvirt-5.0.0/src/security/virt-aa-helper.c @@ -1251,6 +1251,26 @@ get_files(vahControl * ctl) } } + for (i = 0; i < ctl->def->nsmartcards; i++) { + virDomainSmartcardDefPtr sc = ctl->def->smartcards[i]; + virDomainSmartcardType sc_type = sc->type; + char *sc_db = (char *)VIR_DOMAIN_SMARTCARD_DEFAULT_DATABASE; + if (sc->data.cert.database) + sc_db = sc->data.cert.database; + switch(sc_type) { + case VIR_DOMAIN_SMARTCARD_TYPE_HOST: + break; + case VIR_DOMAIN_SMARTCARD_TYPE_HOST_CERTIFICATES: + virBufferAsprintf(&buf, " \"%s/\" rk,\n", sc_db); + virBufferAsprintf(&buf, " \"%s/*\" rk,\n", sc_db); + break; + case VIR_DOMAIN_SMARTCARD_TYPE_PASSTHROUGH: + break; + case VIR_DOMAIN_SMARTCARD_TYPE_LAST: + break; + } + } + if (ctl->def->virtType == VIR_DOMAIN_VIRT_KVM) { for (i = 0; i < ctl->def->nnets; i++) { virDomainNetDefPtr net = ctl->def->nets[i];

3 3

[libvirt] [PATCH] virsh: limit completion of 'domhostname' to active domains
by Pino Toscano 20 Nov '19

20 Nov '19

Getting the hostname of guest usually requires a in-guest agent, or generally can be determined only on active domains. Signed-off-by: Pino Toscano <ptoscano(a)redhat.com> --- tools/virsh-domain.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/tools/virsh-domain.c b/tools/virsh-domain.c index b248a15c16..6be9780836 100644 --- a/tools/virsh-domain.c +++ b/tools/virsh-domain.c @@ -11668,7 +11668,7 @@ static const vshCmdInfo info_domhostname[] = { }; static const vshCmdOptDef opts_domhostname[] = { - VIRSH_COMMON_OPT_DOMAIN_FULL(0), + VIRSH_COMMON_OPT_DOMAIN_FULL(VIR_CONNECT_LIST_DOMAINS_ACTIVE), {.name = NULL} }; -- 2.21.0

3 2

[libvirt] FYI mail problems for libvirt lists
by Daniel P. Berrangé 20 Nov '19

20 Nov '19

It has come to our attention that many, possibly even all, people with non-redhat.com email addresses are unable to send mail to most libvirt mailing lists, receiving bounce messages saying the address doesn't exist eg Final-Recipient: rfc822; libvirt-users(a)redhat.com Action: failed Status: 5.0.0 Remote-MTA: dns; us-smtp-inbound-1.mimecast.com Diagnostic-Code: smtp; 550 Invalid Recipient - https://community.mimecast.com/docs/DOC-1369#550 [boadZMN8PrGodBBK6pzKWg.us309] In my testing, libvir-list(a)redhat.com is the only list that is currently accepting incoming mail from non-redhat.com addresses. I see bounces from libvirt-users(a)redhat.com, libvirt-ci(a)redhat.com, libvirt-announce(a)redhat.com and libvirt-security(a)redhat.com, as well as from the undocumented alias libvirt-list(a)redhat.com We believe outgoing mail delivery is still working normally in all cases. The problems appear to be caused by some changes that Red Hat administrators made to the email infrastructure recently. We have a severity 1 ticket open for this issue and are taking all possible steps to escalate it & get it resolved at the soonest opportunity. In the mean time if any users need assistance with libvirt questions, please feel free to ignore our normal guidance and use the main development list libvir-list(a)redhat.com for questions, as this appears to still be working. Alternatively you can use IRC #virt on irc.oftc.net If you need to report any security issues with libvirt, please try the libvirt-security(a)redhat.com list first, but if you get a bounce then email myself directly. Please accept our apologies for the disruption this is causing to the libvirt mailing lists. Regards, Daniel -- |: https://berrange.com -o- https://www.flickr.com/photos/dberrange :| |: https://libvirt.org -o- https://fstop138.berrange.com :| |: https://entangle-photo.org -o- https://www.instagram.com/dberrange :|

1 1

[libvirt] [PATCH 00/13] drop usage of c-type gnulib module
by Pavel Hrdina 20 Nov '19

20 Nov '19

Pavel Hrdina (13): util: define IS_BLANK instead of using c_isblank from gnulib use g_ascii_isalnum instead of c_isalnum from gnulib use g_ascii_isalpha instead of c_isalpha from gnulib use g_ascii_iscntrl instead of c_iscntrl from gnulib use g_ascii_isdigit instead of c_isdigit frum gnulib use g_ascii_islower instead of c_islower from gnulib use g_ascii_isprint instead of c_isprint from gnulib use g_ascii_isspace instead of c_isspace from gnulib use g_ascii_isxdigit instead of c_isxdigit from gnulib use g_ascii_tolower instead of c_tolower from gnulib use g_ascii_toupper instead of c_toupper from gnulib syntax-check: update c-type checks to refer to Glib bootstrap.conf: drop usage of c-type gnulib module bootstrap.conf | 1 - build-aux/syntax-check.mk | 11 ++--------- src/bhyve/bhyve_parse_command.c | 3 +-- src/conf/network_conf.c | 3 +-- src/conf/nwfilter_conf.c | 3 +-- src/interface/interface_backend_udev.c | 5 ++--- src/libxl/libxl_conf.c | 3 +-- src/node_device/node_device_udev.c | 3 +-- src/qemu/qemu_agent.c | 3 +-- src/qemu/qemu_domain.c | 3 +-- src/qemu/qemu_monitor.c | 3 +-- src/qemu/qemu_qapi.c | 4 +--- src/remote/remote_driver.c | 3 +-- src/rpc/virnetsocket.c | 3 +-- src/storage/parthelper.c | 5 ++--- src/util/virbitmap.c | 5 ++--- src/util/virconf.c | 18 +++++++++--------- src/util/virfile.c | 8 +++----- src/util/virhostcpu.c | 7 +++---- src/util/virkeyfile.c | 9 +++++---- src/util/virmacaddr.c | 11 +++++------ src/util/virnetdevvportprofile.c | 3 +-- src/util/virpidfile.c | 3 +-- src/util/virstoragefile.c | 3 +-- src/util/virstring.c | 13 ++++++------- src/util/virutil.c | 11 +++++------ src/util/viruuid.c | 9 ++++----- src/vmx/vmx.c | 8 +++----- tools/virsh-console.c | 3 +-- tools/virsh-domain.c | 3 +-- tools/vsh-table.c | 5 ++--- tools/vsh.c | 5 ++--- 32 files changed, 71 insertions(+), 109 deletions(-) -- 2.23.0

1 14

[libvirt] [PATCH 0/9] Batch removal of gnulib modules
by Peter Krempa 20 Nov '19

20 Nov '19

This is a series of patches collecting gnulib module removals. Patches from Jano and Pavel have alreadny my R-bs but at least one of my patches does not. Jano, I couldn't find your patch for removal of the mkostemp functions so please provide it. I'll push it later today once I get Jano's patch so that we have only one point of disruption. Ján Tomko (3): gnulib: remove use of 'byteswap' module gnulib: remove use of 'vsnprintf' module bootstrap: remove regex module Pavel Hrdina (4): bootstrap.conf: remove usage of snprintf gnulib module syntax-check: update of sprintf rule to mention g_snpritnf syntax-check: forbid usage of snprintf bootstrap.conf: drop c-strcasestr gnulib module Peter Krempa (2): gnulib: remove 'areadlink' module gnulib: Remove use of 'strsep' module bootstrap.conf | 7 ------- build-aux/syntax-check.mk | 12 ++++++++++-- po/POTFILES.in | 1 - src/qemu/qemu_monitor_json.c | 1 - 4 files changed, 10 insertions(+), 11 deletions(-) -- 2.23.0

2 16

[libvirt] [PATCH] bootstrap.conf: drop c-strcasestr gnulib module
by Pavel Hrdina 20 Nov '19

20 Nov '19

Last usage was removed by commit <41f88886198e231285cc813f8c0687c8ec5c9488> and commit <0f4d31720430b4e3735064cc0d8f88a1a438e154> forgot to drop include. Signed-off-by: Pavel Hrdina <phrdina(a)redhat.com> --- bootstrap.conf | 1 - src/qemu/qemu_monitor_json.c | 1 - 2 files changed, 2 deletions(-) diff --git a/bootstrap.conf b/bootstrap.conf index 4c784487e2..3eab6e1f3c 100644 --- a/bootstrap.conf +++ b/bootstrap.conf @@ -24,7 +24,6 @@ bind byteswap c-ctype c-strcase -c-strcasestr canonicalize-lgpl chown clock-time diff --git a/src/qemu/qemu_monitor_json.c b/src/qemu/qemu_monitor_json.c index 10f6a4cadc..fb662cb18b 100644 --- a/src/qemu/qemu_monitor_json.c +++ b/src/qemu/qemu_monitor_json.c @@ -40,7 +40,6 @@ #include "virprobe.h" #include "virstring.h" #include "cpu/cpu_x86.h" -#include "c-strcasestr.h" #include "virenum.h" #ifdef WITH_DTRACE_PROBES -- 2.23.0

3 2

[libvirt] [PATCH] qemu: Forcibly mknod() even if it exists
by Michal Privoznik 20 Nov '19

20 Nov '19

Another weird bug appeared concerning qemu namespaces. Basically the problem is as follows: 1) Issue an API that causes libvirt to create a node in domain's namespace, say /dev/sda with 8:0 as major:minor (the API can be attach-disk for instance). Or simply create the node from a console by hand. 2) Detach the disk from qemu. 3) Do something that makes /dev/sda change it's minor number. 4) Try to attach the disk again. The problem is, in a few cases - like disk-detach - we don't remove the corresponding /dev node from the mount namespace (because it may be used by some other disk's backing chain). But this creates a problem, because if the node changes its MAJ:MIN numbers we don't propagate the change into the domain's namespace. We do plain mknod() and ignore EEXIST which obviously is not enough because it doesn't guarantee that the node has updated MAJ:MIN pair. Resolves: https://bugzilla.redhat.com/show_bug.cgi?id=1752978 Signed-off-by: Michal Privoznik <mprivozn(a)redhat.com> --- src/qemu/qemu_domain.c | 25 +++++++++---------------- 1 file changed, 9 insertions(+), 16 deletions(-) diff --git a/src/qemu/qemu_domain.c b/src/qemu/qemu_domain.c index 262b74d1ab..f54b9b21ff 100644 --- a/src/qemu/qemu_domain.c +++ b/src/qemu/qemu_domain.c @@ -13203,16 +13203,14 @@ qemuDomainCreateDeviceRecursive(const char *device, allow_noent, ttl - 1) < 0) goto cleanup; } else if (isDev) { - if (create && - mknod(devicePath, sb.st_mode, sb.st_rdev) < 0) { - if (errno == EEXIST) { - ret = 0; - } else { + if (create) { + unlink(devicePath); + if (mknod(devicePath, sb.st_mode, sb.st_rdev) < 0) { virReportSystemError(errno, _("Failed to make device %s"), devicePath); + goto cleanup; } - goto cleanup; } } else if (isReg) { if (create && @@ -13996,17 +13994,12 @@ qemuDomainAttachDeviceMknodHelper(pid_t pid G_GNUC_UNUSED, } else if (isDev) { VIR_DEBUG("Creating dev %s (%d,%d)", data->file, major(data->sb.st_rdev), minor(data->sb.st_rdev)); + unlink(data->file); if (mknod(data->file, data->sb.st_mode, data->sb.st_rdev) < 0) { - /* Because we are not removing devices on hotunplug, or - * we might be creating part of backing chain that - * already exist due to a different disk plugged to - * domain, accept EEXIST. */ - if (errno != EEXIST) { - virReportSystemError(errno, - _("Unable to create device %s"), - data->file); - goto cleanup; - } + virReportSystemError(errno, + _("Unable to create device %s"), + data->file); + goto cleanup; } else { delDevice = true; } -- 2.23.0

2 4

[libvirt] [PATCH 0/4] qemu: block: implement optional removal of committed snapshot images
by Pavel Mores 20 Nov '19

20 Nov '19

Adds handling of the VIR_DOMAIN_BLOCK_COMMIT_DELETE flag to the qemu driver which previously just ignored it. This should fix a problem with the virsh blockcommit command where --delete did nothing. https://bugzilla.redhat.com/show_bug.cgi?id=1008350 Pavel Mores (4): qemu: block: propagate the delete flag to where it can actually be used qemu: block: use the delete flag to delete snapshot images if requested qemu: block: store the delete flag in libvirtd's status XML qemu: block: add unit test of storing the delete flag in status XML src/qemu/qemu_blockjob.c | 35 ++++++++++++++++++- src/qemu/qemu_blockjob.h | 4 ++- src/qemu/qemu_domain.c | 4 +++ src/qemu/qemu_driver.c | 5 +-- .../blockjob-blockdev-in.xml | 1 + 5 files changed, 45 insertions(+), 4 deletions(-) -- 2.21.0

1 5

[libvirt] [PATCH] lib: fix documentation typo in virDomainGetGuestInfo()
by Jonathon Jongsma 20 Nov '19

20 Nov '19

Signed-off-by: Jonathon Jongsma <jjongsma(a)redhat.com> --- src/libvirt-domain.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/src/libvirt-domain.c b/src/libvirt-domain.c index 02622cb2ca..2d9c4061e2 100644 --- a/src/libvirt-domain.c +++ b/src/libvirt-domain.c @@ -12272,7 +12272,7 @@ virDomainSetVcpu(virDomainPtr domain, * "timezone.offset" - the offset to UTC in seconds as an int * * VIR_DOMAIN_GUEST_INFO_FILESYSTEM: - * Returns inforamtion about the filesystems within the domain. The typed + * Returns information about the filesystems within the domain. The typed * parameter keys are in this format: * * "fs.count" - the number of filesystems defined on this domain -- 2.21.0

3 2

Re: [libvirt] [PATCH 0/6] VFIO mdev aggregated resources handling
by Alex Williamson 20 Nov '19

20 Nov '19

On Thu, 24 Oct 2019 13:08:23 +0800 Zhenyu Wang <zhenyuw(a)linux.intel.com> wrote: > Hi, > > This is a refresh for previous send of this series. I got impression that > some SIOV drivers would still deploy their own create and config method so > stopped effort on this. But seems this would still be useful for some other > SIOV driver which may simply want capability to aggregate resources. So here's > refreshed series. > > Current mdev device create interface depends on fixed mdev type, which get uuid > from user to create instance of mdev device. If user wants to use customized > number of resource for mdev device, then only can create new mdev type for that > which may not be flexible. This requirement comes not only from to be able to > allocate flexible resources for KVMGT, but also from Intel scalable IO > virtualization which would use vfio/mdev to be able to allocate arbitrary > resources on mdev instance. More info on [1] [2] [3]. > > To allow to create user defined resources for mdev, it trys to extend mdev > create interface by adding new "aggregate=xxx" parameter following UUID, for > target mdev type if aggregation is supported, it can create new mdev device > which contains resources combined by number of instances, e.g > > echo "<uuid>,aggregate=10" > create > > VM manager e.g libvirt can check mdev type with "aggregation" attribute which > can support this setting. If no "aggregation" attribute found for mdev type, > previous behavior is still kept for one instance allocation. And new sysfs > attribute "aggregated_instances" is created for each mdev device to show allocated number. Given discussions we've had recently around libvirt interacting with mdev, I think that libvirt would rather have an abstract interface via mdevctl[1]. Therefore can you evaluate how mdevctl would support this creation extension? It seems like it would fit within the existing mdev and mdevctl framework if aggregation were simply a sysfs attribute for the device. For example, the mdevctl steps might look like this: mdevctl define -u UUID -p PARENT -t TYPE mdevctl modify -u UUID --addattr=mdev/aggregation --value=2 mdevctl start -u UUID When mdevctl starts the mdev, it will first create it using the existing mechanism, then apply aggregation attribute, which can consume the necessary additional instances from the parent device, or return an error, which would unwind and return a failure code to the caller (libvirt). I think the vendor driver would then have freedom to decide when the attribute could be modified, for instance it would be entirely reasonable to return -EBUSY if the user attempts to modify the attribute while the mdev device is in-use. Effectively aggregation simply becomes a standardized attribute with common meaning. Thoughts? [cc libvirt folks for their impression] Thanks, Alex [1] https://github.com/mdevctl/mdevctl > References: > [1] https://software.intel.com/en-us/download/intel-virtualization-technology-f… > [2] https://software.intel.com/en-us/download/intel-scalable-io-virtualization-… > [3] https://schd.ws/hosted_files/lc32018/00/LC3-SIOV-final.pdf > > Zhenyu Wang (6): > vfio/mdev: Add new "aggregate" parameter for mdev create > vfio/mdev: Add "aggregation" attribute for supported mdev type > vfio/mdev: Add "aggregated_instances" attribute for supported mdev > device > Documentation/driver-api/vfio-mediated-device.rst: Update for > vfio/mdev aggregation support > Documentation/ABI/testing/sysfs-bus-vfio-mdev: Update for vfio/mdev > aggregation support > drm/i915/gvt: Add new type with aggregation support > > Documentation/ABI/testing/sysfs-bus-vfio-mdev | 24 ++++++ > .../driver-api/vfio-mediated-device.rst | 23 ++++++ > drivers/gpu/drm/i915/gvt/gvt.c | 4 +- > drivers/gpu/drm/i915/gvt/gvt.h | 11 ++- > drivers/gpu/drm/i915/gvt/kvmgt.c | 53 ++++++++++++- > drivers/gpu/drm/i915/gvt/vgpu.c | 56 ++++++++++++- > drivers/vfio/mdev/mdev_core.c | 36 ++++++++- > drivers/vfio/mdev/mdev_private.h | 6 +- > drivers/vfio/mdev/mdev_sysfs.c | 79 ++++++++++++++++++- > include/linux/mdev.h | 19 +++++ > 10 files changed, 294 insertions(+), 17 deletions(-) >

4 6

[libvirt] [PATCH] lxc: Fix 'domblkstat' error with attached disk devices.
by jcfaracco＠gmail.com 19 Nov '19

19 Nov '19

From: Julio Faracco <jcfaracco(a)gmail.com> LXC was not working when you attached a new disk that points to block device. See: https://bugzilla.redhat.com/show_bug.cgi?id=1697115. Command line from virsh was showing problems with alias first (this feature is not supported) and after, problems to query block device. It was happening because extra disks were not being included into cgroup blkio.throttle properties and libvirt could not query this info. Applying those devices into 'allowed' list is not enough, libvirt should reset blkio.throttle as a workaround to include disks into container's cgroup directory. Before: virsh # domblkstat CentOS hda error: Failed to get block stats for domain 'CentOS' device 'hda' error: internal error: domain stats query failed Now: virsh # domblkstat CentOS hda hda rd_req 0 hda rd_bytes 0 hda wr_req 0 hda wr_bytes 0 Signed-off-by: Julio Faracco <jcfaracco(a)gmail.com> --- src/lxc/lxc_cgroup.c | 29 ++++++++++++++++++++++++++++- src/lxc/lxc_cgroup.h | 4 ++++ src/lxc/lxc_driver.c | 8 ++++---- 3 files changed, 36 insertions(+), 5 deletions(-) diff --git a/src/lxc/lxc_cgroup.c b/src/lxc/lxc_cgroup.c index 5efb495b56..de6d892521 100644 --- a/src/lxc/lxc_cgroup.c +++ b/src/lxc/lxc_cgroup.c @@ -302,6 +302,24 @@ virLXCTeardownHostUSBDeviceCgroup(virUSBDevicePtr dev G_GNUC_UNUSED, return 0; } +int +virLXCCgroupResetBlkioDeviceThrottle(virCgroupPtr cgroup, + const char *path) +{ + if (virCgroupSetBlkioDeviceReadBps(cgroup, path, 0) < 0) + return -1; + + if (virCgroupSetBlkioDeviceWriteBps(cgroup, path, 0) < 0) + return -1; + + if (virCgroupSetBlkioDeviceReadIops(cgroup, path, 0) < 0) + return -1; + + if (virCgroupSetBlkioDeviceWriteIops(cgroup, path, 0) < 0) + return -1; + + return 0; +} static int virLXCCgroupSetupDeviceACL(virDomainDefPtr def, virCgroupPtr cgroup) @@ -309,6 +327,7 @@ static int virLXCCgroupSetupDeviceACL(virDomainDefPtr def, int capMknod = def->caps_features[VIR_DOMAIN_CAPS_FEATURE_MKNOD]; int ret = -1; size_t i; + const char *src_path = NULL; static virLXCCgroupDevicePolicy devices[] = { {'c', LXC_DEV_MAJ_MEMORY, LXC_DEV_MIN_NULL}, {'c', LXC_DEV_MAJ_MEMORY, LXC_DEV_MIN_ZERO}, @@ -346,8 +365,16 @@ static int virLXCCgroupSetupDeviceACL(virDomainDefPtr def, !virStorageSourceIsBlockLocal(def->disks[i]->src)) continue; + /* Workaround to include disks into blkio.throttle. + * To include it, we need to reset any feature of + * blkio.throttle.* */ + src_path = virDomainDiskGetSource(def->disks[i]); + if (virLXCCgroupResetBlkioDeviceThrottle(cgroup, + src_path) < 0) + goto cleanup; + if (virCgroupAllowDevicePath(cgroup, - virDomainDiskGetSource(def->disks[i]), + src_path, (def->disks[i]->src->readonly ? VIR_CGROUP_DEVICE_READ : VIR_CGROUP_DEVICE_RW) | diff --git a/src/lxc/lxc_cgroup.h b/src/lxc/lxc_cgroup.h index 63e9e837b0..3d643a4fea 100644 --- a/src/lxc/lxc_cgroup.h +++ b/src/lxc/lxc_cgroup.h @@ -46,3 +46,7 @@ int virLXCTeardownHostUSBDeviceCgroup(virUSBDevicePtr dev, const char *path, void *opaque); + +int +virLXCCgroupResetBlkioDeviceThrottle(virCgroupPtr cgroup, + const char *path); diff --git a/src/lxc/lxc_driver.c b/src/lxc/lxc_driver.c index 204a3ed522..87da55f308 100644 --- a/src/lxc/lxc_driver.c +++ b/src/lxc/lxc_driver.c @@ -2339,14 +2339,14 @@ lxcDomainBlockStats(virDomainPtr dom, goto endjob; } - if (!disk->info.alias) { + if (!disk->src->path) { virReportError(VIR_ERR_INTERNAL_ERROR, _("missing disk device alias name for %s"), disk->dst); goto endjob; } ret = virCgroupGetBlkioIoDeviceServiced(priv->cgroup, - disk->info.alias, + disk->src->path, &stats->rd_bytes, &stats->wr_bytes, &stats->rd_req, @@ -2424,14 +2424,14 @@ lxcDomainBlockStatsFlags(virDomainPtr dom, goto endjob; } - if (!disk->info.alias) { + if (!disk->src->path) { virReportError(VIR_ERR_INTERNAL_ERROR, _("missing disk device alias name for %s"), disk->dst); goto endjob; } if (virCgroupGetBlkioIoDeviceServiced(priv->cgroup, - disk->info.alias, + disk->src->path, &rd_bytes, &wr_bytes, &rd_req, -- 2.20.1

3 2

[libvirt] [PATCH v2 0/3] remove 'cleanup: return' labels
by Daniel Henrique Barboza 19 Nov '19

19 Nov '19

changes from v1 [1], all suggested by Ján Tomko: - qemu patch (former patch 10): * qemu_driver.c: using 'return -1' instead of 'return ret' in qemuDumpToFd() and qemuDomainBlockPivot() * qemu_hotplug.c: removed the use of 'ret' variable in qemuDomainChangeGraphics(), returning -1 in all error conditions and returning 0 in the end - vbox patch (former patch 13): * vbox_common.c: changed 'return -2' to 'return -1' in vboxDomainGetInfo() - tests patch (former patch 20): * domainconftest.c: rolled back the change since the 'goto cleanup' is more readable than copying EXIT_SUCCESS multiple times. [1] https://www.redhat.com/archives/libvir-list/2019-October/msg01420.html Daniel Henrique Barboza (3): qemu: remove unneeded cleanup labels vbox: remove unneeded cleanup labels tests: remove unneeded cleanup labels src/qemu/qemu_cgroup.c | 15 ++-- src/qemu/qemu_domain.c | 84 ++++++------------ src/qemu/qemu_domain_address.c | 31 +++---- src/qemu/qemu_driver.c | 145 +++++++++++-------------------- src/qemu/qemu_hotplug.c | 75 +++++++--------- src/qemu/qemu_interface.c | 26 +++--- src/qemu/qemu_migration.c | 38 +++----- src/qemu/qemu_process.c | 69 +++++---------- src/qemu/qemu_tpm.c | 18 ++-- src/vbox/vbox_common.c | 12 ++- src/vbox/vbox_snapshot_conf.c | 75 +++++++--------- tests/commandtest.c | 17 ++-- tests/networkxml2firewalltest.c | 10 +-- tests/nsstest.c | 27 +++--- tests/nwfilterebiptablestest.c | 4 +- tests/nwfilterxml2firewalltest.c | 16 ++-- tests/qemuhotplugtest.c | 19 ++-- tests/qemuxml2argvtest.c | 11 +-- tests/virauthconfigtest.c | 11 +-- tests/vircgroupmock.c | 11 +-- tests/virendiantest.c | 58 ++++++------- tests/virkeycodetest.c | 14 +-- tests/virmacmaptest.c | 5 +- tests/virnetdevtest.c | 11 +-- tests/virpcimock.c | 31 +++---- tests/virpcitest.c | 3 +- tests/virpolkittest.c | 65 +++++--------- tests/virstringtest.c | 18 ++-- 28 files changed, 333 insertions(+), 586 deletions(-) -- 2.21.0

2 5

[libvirt] [PATCH 0/3] tests: schema: Try to improve debugability
by Peter Krempa 19 Nov '19

19 Nov '19

Peter Krempa (3): tests: schema: Simplify memory handling using g_autofree tests: utils: Introduce helper for dispatching libvirt errors tests: virschema: Propagate errors from directory traversal in testSchemaDir tests/testutils.c | 29 +++++++++++++++++++++++++---- tests/testutils.h | 1 + tests/virschematest.c | 34 +++++++++++++--------------------- 3 files changed, 39 insertions(+), 25 deletions(-) -- 2.23.0

2 4

[libvirt] [PATCH 1/3] qemuProcessStop: Set @def early
by Michal Privoznik 19 Nov '19

19 Nov '19

The @def variable holds pointer to the domain defintion, but is set only somewhere in the middle of the function. This is suboptimal. Signed-off-by: Michal Privoznik <mprivozn(a)redhat.com> --- src/qemu/qemu_process.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/src/qemu/qemu_process.c b/src/qemu/qemu_process.c index 90ebdb43a0..8cb12ac9a6 100644 --- a/src/qemu/qemu_process.c +++ b/src/qemu/qemu_process.c @@ -7365,6 +7365,8 @@ void qemuProcessStop(virQEMUDriverPtr driver, goto endjob; } + def = vm->def; + qemuProcessBuildDestroyMemoryPaths(driver, vm, NULL, false); if (virAtomicIntDecAndTest(&driver->nactive) && driver->inhibitCallback) @@ -7386,7 +7388,6 @@ void qemuProcessStop(virQEMUDriverPtr driver, virDomainConfVMNWFilterTeardown(vm); if (cfg->macFilter) { - def = vm->def; for (i = 0; i < def->nnets; i++) { virDomainNetDefPtr net = def->nets[i]; if (net->ifname == NULL) @@ -7483,7 +7484,6 @@ void qemuProcessStop(virQEMUDriverPtr driver, qemuHostdevReAttachDomainDevices(driver, vm->def); - def = vm->def; for (i = 0; i < def->nnets; i++) { virDomainNetDefPtr net = def->nets[i]; vport = virDomainNetGetActualVirtPortProfile(net); -- 2.23.0

2 2

[libvirt] [PATCH] Add news for virDomainAgentSetResponseTimeout()
by Jonathon Jongsma 19 Nov '19

19 Nov '19

Signed-off-by: Jonathon Jongsma <jjongsma(a)redhat.com> --- docs/news.xml | 13 +++++++++++++ 1 file changed, 13 insertions(+) diff --git a/docs/news.xml b/docs/news.xml index 240d288c37..72c0edb317 100644 --- a/docs/news.xml +++ b/docs/news.xml @@ -66,6 +66,19 @@ initialized the vGPU. </description> </change> + <change> + <summary> + Add API to change the response timeout for guest agent commands + </summary> + <description> + By default, when a command is sent to the guest agent, libvirt waits + forever for a response from the guest agent. If the guest is + unresponsive for any reason, this can block the calling thread + indefinitely. By setting a custom timeout using + <code>virDomainAgentSetResponseTimeout()</code>, API users can change + this behavior. + </description> + </change> </section> <section title="Improvements"> </section> -- 2.21.0

2 1

[libvirt] [libvirt-dbus][PATCH v4 0/2] Implement snapshots
by Simon Kobyda 18 Nov '19

18 Nov '19

Changes since v2: Implement and use virtDBusDomainSnapshotGetVirDomainSnapshot Add domain snapshot xml file to data/meson.build Fix flake8 style errors Simon Kobyda (2): Introduce Domain Snapshot Interface Implement snapshots APIs data/meson.build | 1 + data/org.libvirt.Domain.xml | 26 +++ data/org.libvirt.DomainSnapshot.xml | 41 ++++ src/connect.c | 6 + src/connect.h | 1 + src/domain.c | 154 +++++++++++++++ src/domainsnapshot.c | 291 ++++++++++++++++++++++++++++ src/domainsnapshot.h | 9 + src/meson.build | 1 + src/util.c | 53 +++++ src/util.h | 16 ++ tests/libvirttest.py | 12 ++ tests/meson.build | 1 + tests/test_domain.py | 8 + tests/test_snapshot.py | 45 +++++ tests/xmldata.py | 6 + 16 files changed, 671 insertions(+) create mode 100644 data/org.libvirt.DomainSnapshot.xml create mode 100644 src/domainsnapshot.c create mode 100644 src/domainsnapshot.h create mode 100755 tests/test_snapshot.py -- 2.21.0

2 3

[libvirt] [PATCH] g_mkostemp_full: pass O_RDWR
by Ján Tomko 18 Nov '19

18 Nov '19

This flag is not implied by g_mkostemp_full, only by g_mkostemp. Signed-off-by: Ján Tomko <jtomko(a)redhat.com> Reported-by: Bjoern Walk <bwalk(a)linux.ibm.com> Fixes: 4ac47730408eaf91683f6502ec10541f4f711a5c --- src/qemu/qemu_driver.c | 4 ++-- src/storage/storage_util.c | 2 +- src/util/virlog.c | 2 +- src/vbox/vbox_common.c | 2 +- tests/virfiletest.c | 2 +- tools/vsh.c | 2 +- 6 files changed, 7 insertions(+), 7 deletions(-) diff --git a/src/qemu/qemu_driver.c b/src/qemu/qemu_driver.c index 60f5732708..7fd87a9d76 100644 --- a/src/qemu/qemu_driver.c +++ b/src/qemu/qemu_driver.c @@ -4024,7 +4024,7 @@ qemuDomainScreenshot(virDomainPtr dom, if (!(tmp = g_strdup_printf("%s/qemu.screendump.XXXXXX", cfg->cacheDir))) goto endjob; - if ((tmp_fd = g_mkstemp_full(tmp, O_CLOEXEC, S_IRUSR | S_IWUSR)) == -1) { + if ((tmp_fd = g_mkstemp_full(tmp, O_RDWR | O_CLOEXEC, S_IRUSR | S_IWUSR)) == -1) { virReportSystemError(errno, _("g_mkstemp(\"%s\") failed"), tmp); goto endjob; } @@ -11970,7 +11970,7 @@ qemuDomainMemoryPeek(virDomainPtr dom, goto endjob; /* Create a temporary filename. */ - if ((fd = g_mkstemp_full(tmp, O_CLOEXEC, S_IRUSR | S_IWUSR)) == -1) { + if ((fd = g_mkstemp_full(tmp, O_RDWR | O_CLOEXEC, S_IRUSR | S_IWUSR)) == -1) { virReportSystemError(errno, _("g_mkstemp(\"%s\") failed"), tmp); goto endjob; diff --git a/src/storage/storage_util.c b/src/storage/storage_util.c index 6d41df8d7e..f2d8810813 100644 --- a/src/storage/storage_util.c +++ b/src/storage/storage_util.c @@ -1215,7 +1215,7 @@ storageBackendCreateQemuImgSecretPath(virStoragePoolObjPtr pool, if (!(secretPath = virStoragePoolObjBuildTempFilePath(pool, vol))) goto cleanup; - if ((fd = g_mkstemp_full(secretPath, O_CLOEXEC, S_IRUSR | S_IWUSR)) < 0) { + if ((fd = g_mkstemp_full(secretPath, O_RDWR | O_CLOEXEC, S_IRUSR | S_IWUSR)) < 0) { virReportSystemError(errno, "%s", _("failed to open secret file for write")); goto error; diff --git a/src/util/virlog.c b/src/util/virlog.c index 05052e9d09..d45e2dd316 100644 --- a/src/util/virlog.c +++ b/src/util/virlog.c @@ -992,7 +992,7 @@ virLogOutputToJournald(virLogSourcePtr source, * and pass an FD to the journal */ - if ((buffd = g_mkstemp_full(path, O_CLOEXEC, S_IRUSR | S_IWUSR)) < 0) + if ((buffd = g_mkstemp_full(path, O_RDWR | O_CLOEXEC, S_IRUSR | S_IWUSR)) < 0) return; if (unlink(path) < 0) diff --git a/src/vbox/vbox_common.c b/src/vbox/vbox_common.c index 88ef41b5df..9e41b6767a 100644 --- a/src/vbox/vbox_common.c +++ b/src/vbox/vbox_common.c @@ -7385,7 +7385,7 @@ vboxDomainScreenshot(virDomainPtr dom, tmp = g_strdup_printf("%s/vbox.screendump.XXXXXX", cacheDir); - if ((tmp_fd = g_mkstemp_full(tmp, O_CLOEXEC, S_IRUSR | S_IWUSR)) == -1) { + if ((tmp_fd = g_mkstemp_full(tmp, O_RDWR | O_CLOEXEC, S_IRUSR | S_IWUSR)) == -1) { virReportSystemError(errno, _("g_mkstemp(\"%s\") failed"), tmp); VIR_FREE(tmp); VBOX_RELEASE(machine); diff --git a/tests/virfiletest.c b/tests/virfiletest.c index 193c5bedd4..781c640e2b 100644 --- a/tests/virfiletest.c +++ b/tests/virfiletest.c @@ -133,7 +133,7 @@ makeSparseFile(const off_t offsets[], off_t len = 0; size_t i; - if ((fd = g_mkstemp_full(path, O_CLOEXEC, S_IRUSR | S_IWUSR)) < 0) + if ((fd = g_mkstemp_full(path, O_RDWR | O_CLOEXEC, S_IRUSR | S_IWUSR)) < 0) goto error; if (unlink(path) < 0) diff --git a/tools/vsh.c b/tools/vsh.c index b5de06f26b..1076c8254b 100644 --- a/tools/vsh.c +++ b/tools/vsh.c @@ -2400,7 +2400,7 @@ vshEditWriteToTempFile(vshControl *ctl, const char *doc) tmpdir = getenv("TMPDIR"); if (!tmpdir) tmpdir = "/tmp"; ret = g_strdup_printf("%s/virshXXXXXX.xml", tmpdir); - fd = g_mkstemp_full(ret, O_CLOEXEC, S_IRUSR | S_IWUSR); + fd = g_mkstemp_full(ret, O_RDWR | O_CLOEXEC, S_IRUSR | S_IWUSR); if (fd == -1) { vshError(ctl, _("g_mkstemp_full: failed to create temporary file: %s"), virStrerror(errno, ebuf, sizeof(ebuf))); -- 2.21.0

2 1

[libvirt] Bug: virsh edit broken on master
by Bjoern Walk 18 Nov '19

18 Nov '19

Hey all, commit 4ac47730: Use g_mkstemp_full instead of mkostemp(s) broke `virsh edit` (at least on S390): virsh # edit test error: write: /tmp/virshQ7SXB0.xml: failed to write to temporary file: Bad file descriptorx I didn't find out why, everything looks normal: # ls -l /proc/$(pgrep virsh)/fd/ total 0 [...] lr-x------. 1 root root 64 Nov 18 16:58 10 -> /tmp/virshAEKYB0.xml # ls -l /tmp/virshAEKYB0.xml -rw-------. 1 root root 0 Nov 18 16:59 /tmp/virshAEKYB0.xml Best, Bjoern -- IBM Systems Linux on Z & Virtualization Development -------------------------------------------------- IBM Deutschland Research & Development GmbH Schönaicher Str. 220, 71032 Böblingen Phone: +49 7031 16 1819 -------------------------------------------------- Vorsitzende des Aufsichtsrats: Matthias Hartmann Geschäftsführung: Dirk Wittkopp Sitz der Gesellschaft: Böblingen Registergericht: Amtsgericht Stuttgart, HRB 243294

3 2

[libvirt] [PATCH 0/3] A few more Coverity patches
by John Ferlan 18 Nov '19

18 Nov '19

Finally got Coverity to build with the recent build env changes. With doing that Coverity re-examines everything and bridge_driver generated a couple of longer term things. The virFileRewrite is from more recent changes. John Ferlan (3): util: Remove unnecessary check in virFileRewrite network: Use local variables in networkUpdatePortBandwidth network: Check for QOS before blindly using it src/network/bridge_driver.c | 31 ++++++++++++++++++++++--------- src/util/virfile.c | 3 +-- 2 files changed, 23 insertions(+), 11 deletions(-) -- 2.20.1

4 10

[libvirt] [PATCH 0/3] Fix remote dispatch code trying to allocate 0-sized return buffers
by Erik Skultety 18 Nov '19

18 Nov '19

This happens because of the switch to glib whose method g_malloc0 actually correctly returns NULL on size 0 which doesn't make sense to do. The outcome of that is that because virAllocN always returns 0, the dispatch code never fails allocation and passes the NULL pointer to the server-side public API which actually checks that and fails. https://bugzilla.redhat.com/show_bug.cgi?id=1772842 Erik Skultety (3): rpc: gendispatch: Fix a couple of places adding trailing spaces rpc: gendispatch: Add a check for zero size client-side buffers libvirt-domain: virConnectListDomains: Return 0 on zero-size buffer src/libvirt-domain.c | 3 +++ src/rpc/gendispatch.pl | 12 ++++++++++-- 2 files changed, 13 insertions(+), 2 deletions(-) -- 2.23.0

2 8

[libvirt] [PATCH] qemu: Fix NULL ptr dereference caused by qemuDomainDefFormatBufInternal
by Erik Skultety 18 Nov '19

18 Nov '19

qemuDomainDefFormatBufInternal function wasn't testing whether the CPU was actually defined in the XML and saving such a domain resulted in the following backtrace: 0 in qemuDomainMakeCPUMigratable (cpu=0x0) 1 in qemuDomainDefFormatBufInternal() 2 in qemuDomainDefFormatXMLInternal() 3 in qemuDomainDefFormatLive() 4 in qemuDomainSaveInternal() 5 in qemuDomainSaveFlags() 6 in qemuDomainSave() 7 in virDomainSave() Signed-off-by: Erik Skultety <eskultet(a)redhat.com> --- src/qemu/qemu_domain.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/src/qemu/qemu_domain.c b/src/qemu/qemu_domain.c index e14b414518..ed2a5cc7b6 100644 --- a/src/qemu/qemu_domain.c +++ b/src/qemu/qemu_domain.c @@ -9105,7 +9105,7 @@ qemuDomainDefFormatBufInternal(virQEMUDriverPtr driver, goto cleanup; } - if (qemuDomainMakeCPUMigratable(def->cpu) < 0) + if (def->cpu && qemuDomainMakeCPUMigratable(def->cpu) < 0) goto cleanup; } -- 2.23.0

2 1

[libvirt] [PATCH v2 00/24] scripts: convert most perl scripts to python
by Daniel P. Berrangé 18 Nov '19

18 Nov '19

This series is an effort to reduce the number of different languages we use by eliminating most use of perl in favour of python. This aligns with fact that the likely future build system we'll use (meson) is written in python, and that python is much more commonly used/understood by developers these days than perl. With this applied we use perl in a handful of places only: - src/rpc/gendispatch.pl - this is a horrendously large script and very hard to understand/follow. A straight syntax conversion to Python would still leave a hgue and hard to understand/follow script. It really needs a clean room rewrite from scratch, with better structure. - src/rpc/genprotocol.pl - fairly easy to convert, but might be obsolete depending on approach for rewriting gendispatch.pl, so ignored for now - tests/oomtrace.pl - will be purge by the patches that drop OOM handling anyway - tools/wireshark/util/genxdrstub.pl - a very large script, which I haven't got the courage to tackle yet. - cfg.mk/maint.mk - many syntax rules involve regexes which are fed to perl. Decision on what to do with syntax-check rules punted to another time. - build-aux/gitlog-to-changelog - build-aux/useless-if-before-free - Both pulled in from gnulib. Could be rewritten quite easily if desired, but given that we aren't maintainers of them right now, they're ignored as they don't really impact our developers. In v2: - Pulled in patch to hacking file - Converted many more scripts - Forced UTF-8 character set to avoid ascii codec on py3 < 3.7 Daniel P. Berrangé (24): docs: document that C & Python are the preferred languages cfg.mk: fix comment detection for python semicolon check build: force a UTF-8 locale for python build-aux: rewrite augest test generator in Python build-aux: rewrite po file minimizer in Python build-aux: rewrite duplicate header checker in Python build-aux: rewrite whitespace checker in Python build-aux: rewrite mock inline checker in Python build-aux: rewrite header ifdef checker in Python src: rewrite ACL permissions checker in Python src: rewrite symfile sorting checker in Python src: rewrite symfile library checker in Python src: rewrite systemtap probe generator in Python src: rewrite systemtap function generator in Python src: rewrite driver name checker in Python src: rewrite driver impl checker in Python src: rewrite ACL rule checker in Python src: rewrite polkit ACL generator in Python src: rewrite remote protocol checker in Python tests: rewrite test argv line wrapper in Python tests: rewrite qemu capability grouper in Python tests: rewrite file access checker in Python docs: rewrite hvsupport.html page generator in python docs: rewrite polkit docs generator in Python Makefile.am | 14 +- build-aux/augeas-gentest.pl | 60 ---- build-aux/augeas-gentest.py | 72 ++++ build-aux/check-spacing.pl | 198 ---------- build-aux/check-spacing.py | 204 +++++++++++ build-aux/header-ifdef.pl | 182 ---------- build-aux/header-ifdef.py | 206 +++++++++++ build-aux/minimize-po.pl | 37 -- build-aux/minimize-po.py | 60 ++++ build-aux/mock-noinline.pl | 75 ---- build-aux/mock-noinline.py | 88 +++++ build-aux/prohibit-duplicate-header.pl | 26 -- build-aux/prohibit-duplicate-header.py | 54 +++ cfg.mk | 32 +- configure.ac | 8 + docs/Makefile.am | 17 +- docs/genaclperms.pl | 125 ------- docs/genaclperms.py | 122 +++++++ docs/hacking.html.in | 30 ++ docs/hvsupport.pl | 458 ----------------------- docs/hvsupport.py | 479 +++++++++++++++++++++++++ po/Makefile.am | 2 +- src/Makefile.am | 156 +++----- src/access/Makefile.inc.am | 6 +- src/access/genpolkit.pl | 119 ------ src/access/genpolkit.py | 119 ++++++ src/bhyve/Makefile.inc.am | 4 +- src/check-aclperms.pl | 73 ---- src/check-aclperms.py | 77 ++++ src/check-aclrules.pl | 252 ------------- src/check-aclrules.py | 244 +++++++++++++ src/check-driverimpls.pl | 80 ----- src/check-driverimpls.py | 96 +++++ src/check-drivername.pl | 83 ----- src/check-drivername.py | 112 ++++++ src/check-remote-protocol.py | 131 +++++++ src/check-symfile.pl | 70 ---- src/check-symfile.py | 80 +++++ src/check-symsorting.pl | 106 ------ src/check-symsorting.py | 112 ++++++ src/dtrace2systemtap.pl | 130 ------- src/dtrace2systemtap.py | 140 ++++++++ src/esx/Makefile.inc.am | 2 +- src/hyperv/Makefile.inc.am | 2 +- src/interface/Makefile.inc.am | 2 +- src/libxl/Makefile.inc.am | 4 +- src/locking/Makefile.inc.am | 6 +- src/logging/Makefile.inc.am | 2 +- src/lxc/Makefile.inc.am | 4 +- src/network/Makefile.inc.am | 2 +- src/node_device/Makefile.inc.am | 2 +- src/nwfilter/Makefile.inc.am | 2 +- src/qemu/Makefile.inc.am | 4 +- src/remote/Makefile.inc.am | 4 +- src/rpc/Makefile.inc.am | 2 +- src/rpc/gensystemtap.pl | 193 ---------- src/rpc/gensystemtap.py | 174 +++++++++ src/secret/Makefile.inc.am | 2 +- src/storage/Makefile.inc.am | 2 +- src/util/Makefile.inc.am | 8 +- src/vbox/Makefile.inc.am | 2 +- src/vz/Makefile.inc.am | 2 +- tests/Makefile.am | 4 +- tests/check-file-access.pl | 126 ------- tests/check-file-access.py | 121 +++++++ tests/file_access_whitelist.txt | 2 +- tests/group-qemu-caps.pl | 124 ------- tests/group-qemu-caps.py | 115 ++++++ tests/test-wrap-argv.pl | 174 --------- tests/test-wrap-argv.py | 162 +++++++++ tests/testutils.c | 16 +- 71 files changed, 3132 insertions(+), 2872 deletions(-) delete mode 100755 build-aux/augeas-gentest.pl create mode 100755 build-aux/augeas-gentest.py delete mode 100755 build-aux/check-spacing.pl create mode 100755 build-aux/check-spacing.py delete mode 100644 build-aux/header-ifdef.pl create mode 100644 build-aux/header-ifdef.py delete mode 100755 build-aux/minimize-po.pl create mode 100755 build-aux/minimize-po.py delete mode 100644 build-aux/mock-noinline.pl create mode 100644 build-aux/mock-noinline.py delete mode 100644 build-aux/prohibit-duplicate-header.pl create mode 100644 build-aux/prohibit-duplicate-header.py delete mode 100755 docs/genaclperms.pl create mode 100755 docs/genaclperms.py delete mode 100755 docs/hvsupport.pl create mode 100755 docs/hvsupport.py delete mode 100755 src/access/genpolkit.pl create mode 100755 src/access/genpolkit.py delete mode 100755 src/check-aclperms.pl create mode 100755 src/check-aclperms.py delete mode 100755 src/check-aclrules.pl create mode 100755 src/check-aclrules.py delete mode 100755 src/check-driverimpls.pl create mode 100755 src/check-driverimpls.py delete mode 100755 src/check-drivername.pl create mode 100644 src/check-drivername.py create mode 100644 src/check-remote-protocol.py delete mode 100755 src/check-symfile.pl create mode 100755 src/check-symfile.py delete mode 100755 src/check-symsorting.pl create mode 100755 src/check-symsorting.py delete mode 100755 src/dtrace2systemtap.pl create mode 100755 src/dtrace2systemtap.py delete mode 100755 src/rpc/gensystemtap.pl create mode 100755 src/rpc/gensystemtap.py delete mode 100755 tests/check-file-access.pl create mode 100755 tests/check-file-access.py delete mode 100755 tests/group-qemu-caps.pl create mode 100755 tests/group-qemu-caps.py delete mode 100755 tests/test-wrap-argv.pl create mode 100755 tests/test-wrap-argv.py -- 2.21.0

4 42

[libvirt] [PATCH] qemu_process: fix starting VMs if machine group has limited cpuset.cpus
by Pavel Hrdina 18 Nov '19

18 Nov '19

Commit <f136b83139c63f20de0df3285d9e82df2fb97bfc> reworked process affinity setting but did not take cgroups into account which introduced an issue when starting VM with custom cpuset.cpus for the whole machine group. If the machine group is limited to some pCPUs libvirt should not try to set a VM to run on all pCPUs as it will result in permission denied when writing to cpuset.cpus. To fix this the affinity has to be set separately from cgroups cpuset. Resolves: <https://bugzilla.redhat.com/show_bug.cgi?id=1746517> Signed-off-by: Pavel Hrdina <phrdina(a)redhat.com> --- src/qemu/qemu_process.c | 8 ++++++-- 1 file changed, 6 insertions(+), 2 deletions(-) diff --git a/src/qemu/qemu_process.c b/src/qemu/qemu_process.c index ed8666e9d1..355b740caf 100644 --- a/src/qemu/qemu_process.c +++ b/src/qemu/qemu_process.c @@ -2636,6 +2636,7 @@ qemuProcessSetupPid(virDomainObjPtr vm, virDomainNumatuneMemMode mem_mode; virCgroupPtr cgroup = NULL; virBitmapPtr use_cpumask = NULL; + virBitmapPtr afinity_cpumask = NULL; g_autoptr(virBitmap) hostcpumap = NULL; char *mem_mask = NULL; int ret = -1; @@ -2661,7 +2662,7 @@ qemuProcessSetupPid(virDomainObjPtr vm, * its config file */ if (qemuProcessGetAllCpuAffinity(&hostcpumap) < 0) goto cleanup; - use_cpumask = hostcpumap; + afinity_cpumask = hostcpumap; } /* @@ -2702,8 +2703,11 @@ qemuProcessSetupPid(virDomainObjPtr vm, } + if (!afinity_cpumask) + afinity_cpumask = use_cpumask; + /* Setup legacy affinity. */ - if (use_cpumask && virProcessSetAffinity(pid, use_cpumask) < 0) + if (afinity_cpumask && virProcessSetAffinity(pid, afinity_cpumask) < 0) goto cleanup; /* Set scheduler type and priority, but not for the main thread. */ -- 2.23.0

2 1