- Devel - libvirt List Archives

Re: [libvirt] [Qemu-devel] [Patch V2 1/2] x86, mce: Basic support to add LMCE support to QEMU
by Eduardo Habkost 15 Dec '15

15 Dec '15

On Mon, Dec 14, 2015 at 07:17:27PM -0500, Raj, Ashok wrote: > On Mon, Dec 14, 2015 at 11:37:16PM +0100, Borislav Petkov wrote: > > On Mon, Dec 14, 2015 at 02:11:46PM -0500, Raj, Ashok wrote: > > > This is mostly harmless.. since the MCG_CAP space is shared and has no > > > conflict between vendors. Also just the CAP being set has no effect. > > > > Of course it does - we check SER_P in machine_check_poll() and when > > I emulate an AMD guest and inject errors into it, error handling is > > obviously wrong, see: > > > > https://lkml.kernel.org/r/20151123150355.GE5134@pd.tnic > > > > I can see how this hurts.. since the poller isn't doing cpu model specific > stuff..? > > in the LMCE case, even if you advertise MCG_LMCE_P in MCG_CAP, the guest kernel > wont call intel_init_lmce() only from mce_intel.c.. so the same problem > won't happen. > > but the issue Eduardo mentioned seems like the following. > > New QEMU_LMCE + New KVM_LMCE + New_GUEST_LMCE - No problem > > but if you were to migrage the Guest_LMCE to a non-LMCE supported KVM host > we could run into an issue.. > > is this the compatibility issue that you were looking to fix Eduardo? If I understood you correctly, yes. Also, note that currently kvm_arch_init_vcpu() simply warns about missing capabilities, instead of preventing the VM from running/migrating (as it should). We need to change that, and figure out a good way to report "feature FOO can't be enabled in this host" errors to management software[1]. The main problem is that we don't even have a QMP console available anymore if machine initialization is aborted. CCing libvir-list so they get in the loop. [1] This is similar to what we need for CPUID checks, but the new MCE feature means we need something more generic (that just reports QOM property names, probably?) -- Eduardo

1 0

[libvirt] [PATCHv2 0/3] security: misc cleanups
by Ján Tomko 15 Dec '15

15 Dec '15

v1: https://www.redhat.com/archives/libvir-list/2015-November/msg00781.html new in v2: * split by security driver * more functions renamed Ján Tomko (3): security_dac: remove extra Security from function names security_selinux: remove extra Security from function names security_stack: : remove extra security from function names src/security/security_dac.c | 242 +++++++++++++------------- src/security/security_selinux.c | 375 ++++++++++++++++++++-------------------- src/security/security_stack.c | 68 ++++---- 3 files changed, 338 insertions(+), 347 deletions(-) -- 2.4.6

2 5

[libvirt] volume download/upload but for snapshots
by Vasiliy Tolstov 15 Dec '15

15 Dec '15

Hi! I'm happy with libvirt volume download/upload ability, but also i need to download snapshot, what is the best way to do that? Define volume from snapshot? Or write libvirt function for this? -- Vasiliy Tolstov, e-mail: v.tolstov(a)selfip.ru

2 3

[libvirt] [PATCH v2 0/2] Add support for zero-write detection
by Martin Kletzander 15 Dec '15

15 Dec '15

QEMU supports detect-zeroes option since version 2.1, but we never added support for it in libvirt. If was requested by Vasiliy Tolstov in the list, so I just added it. There are two discussions to be had, optionally. One is to decide whether we should disable detect_zeros='unmap' if discard is not set to 'unmap', but this is getting very hypervisor-specific, so I just documented the behaviour. The other one is the naming. I described why I made the decision for "zeros" instead of "zeroes" the decision in the patch, but I have no problem changing it to what others like better. v2: - format detect_zeroes on the command line instead of detect_zeros v1: - https://www.redhat.com/archives/libvir-list/2015-December/msg00484.html Martin Kletzander (2): conf: Add support of zero-detection for disks qemu: Add support for zero-detection writes docs/formatdomain.html.in | 10 ++++++ docs/schemas/domaincommon.rng | 12 +++++++ src/conf/domain_conf.c | 23 +++++++++++++- src/conf/domain_conf.h | 11 +++++++ src/libvirt_private.syms | 2 ++ src/qemu/qemu_capabilities.c | 2 ++ src/qemu/qemu_capabilities.h | 1 + src/qemu/qemu_command.c | 11 +++++++ tests/qemucapabilitiesdata/caps_2.1.1-1.caps | 1 + tests/qemucapabilitiesdata/caps_2.4.0-1.caps | 1 + tests/qemucapabilitiesdata/caps_2.5.0-1.caps | 1 + .../qemuxml2argv-disk-drive-detect-zeros.args | 27 ++++++++++++++++ .../qemuxml2argv-disk-drive-detect-zeros.xml | 37 ++++++++++++++++++++++ tests/qemuxml2argvtest.c | 4 +++ tests/qemuxml2xmltest.c | 1 + 15 files changed, 143 insertions(+), 1 deletion(-) create mode 100644 tests/qemuxml2argvdata/qemuxml2argv-disk-drive-detect-zeros.args create mode 100644 tests/qemuxml2argvdata/qemuxml2argv-disk-drive-detect-zeros.xml -- 2.6.4

3 7

[libvirt] Exclusive VM lock
by Dmitry Andreev 15 Dec '15

15 Dec '15

Hi, I'm looking for a mechanism to do a sequence of API calls as atomic operation. Is there any way for libvirt's API client to acquire an exclusive VM lock to prevent other client from changing VM state through libvirt?

3 5

[libvirt] [PATCH] qemu: Warn when using vhost-user without shared memory
by Martin Kletzander 15 Dec '15

15 Dec '15

When user configures vhost-user interface and forgets to also configure any shared memory, the search for the root cause of non-operational interface might take unpleasantly long time. Let's enhance user experience by emitting a warning in the logs. Resolves: https://bugzilla.redhat.com/show_bug.cgi?id=1266982 Signed-off-by: Martin Kletzander <mkletzan(a)redhat.com> --- src/qemu/qemu_process.c | 44 ++++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 44 insertions(+) diff --git a/src/qemu/qemu_process.c b/src/qemu/qemu_process.c index 420196264685..fb471342e790 100644 --- a/src/qemu/qemu_process.c +++ b/src/qemu/qemu_process.c @@ -4542,6 +4542,7 @@ qemuProcessLaunch(virConnectPtr conn, unsigned int hostdev_flags = 0; size_t nnicindexes = 0; int *nicindexes = NULL; + bool check_shmem = false; VIR_DEBUG("vm=%p name=%s id=%d asyncJob=%d " "incoming.launchURI=%s incoming.deferredURI=%s " @@ -4749,6 +4750,49 @@ qemuProcessLaunch(virConnectPtr conn, goto cleanup; } + VIR_DEBUG("Checking for any possible (non-fatal) issues"); + + /* + * For vhost-user to work, the domain has to have some type of + * shared memory configured. We're not the proper once to judge + * whether shared hugepages or shm are enough and will be in the + * future, so we'll just warn in case there is none of that + * configured. Moreover failing would give the false illusion + * that libvirt is really checking that everything works before + * running the domain and not only we are unable to do that, but + * it's also not our aim to do so. + */ + for (i = 0; i < vm->def->nnets; i++) { + if (virDomainNetGetActualType(vm->def->nets[i]) == + VIR_DOMAIN_NET_TYPE_VHOSTUSER) { + check_shmem = true; + break; + } + } + + if (check_shmem) { + bool shmem = vm->def->nshmems; + + /* + * This check is by no means complete. We merely check + * whetere there are *some* hugepages enabled and *some* NUMA + * nodes with shared memory access. + */ + if (!shmem && vm->def->mem.nhugepages) { + for (i = 0; i < virDomainNumaGetNodeCount(vm->def->numa); i++) { + if (virDomainNumaGetNodeMemoryAccessMode(vm->def->numa, i) == + VIR_NUMA_MEM_ACCESS_SHARED) + shmem = true; + break; + } + } + + if (!shmem) { + VIR_WARN("Detected vhost-user interface without any shared memory. " + "The interface might not be operational"); + } + } + VIR_DEBUG("Building emulator command line"); if (!(cmd = qemuBuildCommandLine(conn, driver, vm->def, priv->monConfig, priv->monJSON, priv->qemuCaps, -- 2.6.3

3 7

[libvirt] [PATCH] qemuMonitorJSONEjectMedia: don't stringify the replay at all
by Pavel Hrdina 15 Dec '15

15 Dec '15

Commit 256496e1 introduced a detection if "is locked" in error replay from qemu monitor. Commit c4073657 fixed a memory leak, but it was pointed out by Peter, that this could be done cleaner without stringifing the replay. Signed-off-by: Pavel Hrdina <phrdina(a)redhat.com> --- src/qemu/qemu_monitor_json.c | 11 ++++++----- 1 file changed, 6 insertions(+), 5 deletions(-) diff --git a/src/qemu/qemu_monitor_json.c b/src/qemu/qemu_monitor_json.c index 4ae63cf..d4b6514 100644 --- a/src/qemu/qemu_monitor_json.c +++ b/src/qemu/qemu_monitor_json.c @@ -2223,11 +2223,12 @@ int qemuMonitorJSONEjectMedia(qemuMonitorPtr mon, ret = qemuMonitorJSONCheckError(cmd, reply); if (ret < 0) { - char *replyStr = virJSONValueToString(reply, false); - - if (c_strcasestr(replyStr, "is locked")) - ret = -2; - VIR_FREE(replyStr); + virJSONValuePtr error = virJSONValueObjectGet(reply, "error"); + if (error) { + const char *errorStr = virJSONValueObjectGetString(error, "desc"); + if (errorStr && c_strcasestr(errorStr, "is locked")) + ret = -2; + } } virJSONValueFree(cmd); -- 2.6.4

2 2

[libvirt] qemu logging/tracing option in domain xml
by Dmitry Andreev 15 Dec '15

15 Dec '15

Hi all, QEMU has the '-d' command line option to configure logs/traces. As I can see, <qemu:commandline> is the only way to set it in domain xml. <qemu:commandline> makes a configuration tainted and doesn't allow to update tracing/logging configuration in runtime. The ability to update configuration in runtime will be very useful for us in Virtuozzo for QEMU problem investigation and development. So I want to ask you will you mind if I prepare a patch-set that introduces a new section for domain xml with next scheme: <element name="debug"> <zeroOrMore> <element name="log"> <attribute name="name"> <ref name="filter-param-value"/> </attribute> </element> </zeroOrMore> </element> Is this acceptable?

3 2

[libvirt] [PATCH] pci: Use 'addr' instead of 'dev' for virPCIDeviceAddressPtr
by Andrea Bolognani 15 Dec '15

15 Dec '15

The name 'dev' is more appropriate for virPCIDevicePtr. --- src/util/virpci.c | 7 ++++--- src/util/virpci.h | 4 ++-- 2 files changed, 6 insertions(+), 5 deletions(-) diff --git a/src/util/virpci.c b/src/util/virpci.c index bb874ac..9cc4613 100644 --- a/src/util/virpci.c +++ b/src/util/virpci.c @@ -2655,12 +2655,13 @@ virPCIGetSysfsFile(char *virPCIDeviceName, char **pci_sysfs_device_link) } int -virPCIDeviceAddressGetSysfsFile(virPCIDeviceAddressPtr dev, +virPCIDeviceAddressGetSysfsFile(virPCIDeviceAddressPtr addr, char **pci_sysfs_device_link) { if (virAsprintf(pci_sysfs_device_link, - PCI_SYSFS "devices/%04x:%02x:%02x.%x", dev->domain, - dev->bus, dev->slot, dev->function) < 0) + PCI_SYSFS "devices/%04x:%02x:%02x.%x", + addr->domain, addr->bus, + addr->slot, addr->function) < 0) return -1; return 0; } diff --git a/src/util/virpci.h b/src/util/virpci.h index 6516f05..f3d5676 100644 --- a/src/util/virpci.h +++ b/src/util/virpci.h @@ -159,7 +159,7 @@ virPCIDeviceListPtr virPCIDeviceGetIOMMUGroupList(virPCIDevicePtr dev); int virPCIDeviceAddressGetIOMMUGroupAddresses(virPCIDeviceAddressPtr devAddr, virPCIDeviceAddressPtr **iommuGroupDevices, size_t *nIommuGroupDevices); -int virPCIDeviceAddressGetIOMMUGroupNum(virPCIDeviceAddressPtr dev); +int virPCIDeviceAddressGetIOMMUGroupNum(virPCIDeviceAddressPtr addr); char *virPCIDeviceGetIOMMUGroupDev(virPCIDevicePtr dev); int virPCIDeviceIsAssignable(virPCIDevicePtr dev, @@ -180,7 +180,7 @@ int virPCIGetVirtualFunctionIndex(const char *pf_sysfs_device_link, const char *vf_sysfs_device_link, int *vf_index); -int virPCIDeviceAddressGetSysfsFile(virPCIDeviceAddressPtr dev, +int virPCIDeviceAddressGetSysfsFile(virPCIDeviceAddressPtr addr, char **pci_sysfs_device_link); int virPCIGetNetName(char *device_link_sysfs_path, char **netname); -- 2.5.0

2 2

[libvirt] [PATCH 0/3] vz: allow to create CT from template and boot VM from cdrom
by Mikhail Feoktistov 15 Dec '15

15 Dec '15

Mikhail Feoktistov (3): vz: allow to create container based on template vz: allow to boot VM from cdrom vz: implementation of boot device order src/vz/vz_sdk.c | 174 ++++++++++++++++++++++++++++++++++++++++---------------- 1 file changed, 125 insertions(+), 49 deletions(-) -- 1.8.3.1

1 3