- Devel - libvirt List Archives

[libvirt] [PATCH] storage: silently ignore missing files on pool refresh
by Ján Tomko 12 Jul '13

12 Jul '13

From: Wei Zhou <w.zhou(a)leaseweb.com> Make virStorageBackendVolOpenCheckMode return -2 instead of -1 if volume file is missing. https://bugzilla.redhat.com/show_bug.cgi?id=977706 --- src/storage/storage_backend.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/src/storage/storage_backend.c b/src/storage/storage_backend.c index e2527c9..f063601 100644 --- a/src/storage/storage_backend.c +++ b/src/storage/storage_backend.c @@ -1084,7 +1084,7 @@ virStorageBackendForType(int type) * Allows caller to silently ignore files with improper mode * * Returns -1 on error, -2 if file mode is unexpected or the - * volume is a dangling symbolic link. + * volume is a dangling symbolic link or file is missing. */ int virStorageBackendVolOpenCheckMode(const char *path, unsigned int flags) @@ -1097,7 +1097,7 @@ virStorageBackendVolOpenCheckMode(const char *path, unsigned int flags) virReportSystemError(errno, _("cannot stat file '%s'"), path); - return -1; + return -2; } if (S_ISFIFO(sb.st_mode)) { -- 1.8.1.5

3 3

[libvirt] [PATCH 0/3] libssh2: improve authentication options
by Peter Krempa 12 Jul '13

12 Jul '13

This series adds the ability to use tunelled password authentication and enables to store credentials for such authentication in libvirts credential storage configuration. Peter Krempa (3): virAuth: Don't require virConnectPtr to retrieve authentication creds libssh2: Improve password based authentication remote: Improve libssh2 password authentication src/remote/remote_driver.c | 3 +- src/rpc/virnetclient.c | 11 ++--- src/rpc/virnetclient.h | 4 +- src/rpc/virnetsocket.c | 8 ++-- src/rpc/virnetsocket.h | 3 +- src/rpc/virnetsshsession.c | 110 ++++++++++++++++++++++++++++++++------------- src/rpc/virnetsshsession.h | 5 ++- src/util/virauth.c | 107 ++++++++++++++++++++++++++++++------------- src/util/virauth.h | 17 ++++++- 9 files changed, 191 insertions(+), 77 deletions(-) -- 1.8.2.1

2 7

[libvirt] [PATCH] nodeinfo: Don't fail on non-contiguous NUMA topologies
by Peter Krempa 12 Jul '13

12 Jul '13

From: hejia hejia <jiakernel(a)gmail.com> nodeGetFreeMemory and nodeGetCellsFreeMemory assumed that the NUMA nodes are contiguous and starting from 0. Unfortunately there are machines that don't match this assumption: available: 1 nodes (1) node 1 cpus: 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 node 1 size: 16340 MB node 1 free: 11065 MB Before this patch: error: internal error Failed to query NUMA free memory error: internal error Failed to query NUMA free memory for node: 0 After this patch: Total: 15772580 KiB 0: 0 KiB Signed-off-by: Peter Krempa <pkrempa(a)redhat.com> --- src/nodeinfo.c | 18 ++++++------------ 1 file changed, 6 insertions(+), 12 deletions(-) diff --git a/src/nodeinfo.c b/src/nodeinfo.c index a2a2f73..4df4851 100644 --- a/src/nodeinfo.c +++ b/src/nodeinfo.c @@ -1686,12 +1686,9 @@ nodeGetCellsFreeMemory(unsigned long long *freeMems, for (numCells = 0, n = startCell; n <= lastCell; n++) { long long mem; - if (numa_node_size64(n, &mem) < 0) { - virReportError(VIR_ERR_INTERNAL_ERROR, - _("Failed to query NUMA free memory for node: %d"), - n); - goto cleanup; - } + if (numa_node_size64(n, &mem) < 0) + mem = 0; + freeMems[numCells++] = mem; } ret = numCells; @@ -1712,15 +1709,12 @@ nodeGetFreeMemory(void) for (n = 0; n <= numa_max_node(); n++) { long long mem; - if (numa_node_size64(n, &mem) < 0) { - virReportError(VIR_ERR_INTERNAL_ERROR, - "%s", _("Failed to query NUMA free memory")); - goto cleanup; - } + if (numa_node_size64(n, &mem) < 0) + continue; + freeMem += mem; } -cleanup: return freeMem; } -- 1.8.3.2

2 2

[libvirt] [PATCHv2 0/3] avoid deadlock in child setting supplemental groups
by Eric Blake 12 Jul '13

12 Jul '13

v1 was here: https://www.redhat.com/archives/libvir-list/2013-May/msg01600.html changes since then: new patch 1, patch 2 (old patch 1/2) is greatly simplified by gnulib, and patch 3 (old patch 2/2) is rebased to pick up an additional call added in lxc. Eric Blake (3): util: grab primary group when doing user lookup util: add virGetGroupList util: make virSetUIDGID async-signal-safe .gnulib | 2 +- bootstrap.conf | 1 + configure.ac | 4 +- src/libvirt_private.syms | 1 + src/lxc/lxc_container.c | 9 ++- src/security/security_dac.c | 16 +++-- src/util/vircommand.c | 10 ++- src/util/virfile.c | 30 ++++++++- src/util/virutil.c | 149 +++++++++++++++++++++----------------------- src/util/virutil.h | 7 ++- 10 files changed, 135 insertions(+), 94 deletions(-) -- 1.8.1.4

2 11

[libvirt] [PATCH v2 0/7] Resolve Coverity issues
by John Ferlan 12 Jul '13

12 Jul '13

v1: https://www.redhat.com/archives/libvir-list/2013-July/msg00691.html Changes over v1: * Add additional files to patch #2 to include other places were sysconf(_SC_OPEN_MAX) was used within libvirt sources. * Change for loop end conditions to use "<variable> != -1" (should have looked deeper at other changes - wasn't even thinking of that option) John Ferlan (7): hellolibvirt: Resolve Coverity issues testutils: Resolve Coverity issues virsh-domain-monitor: Resolve Coverity issues virsh-interface: Resolve Coverity issues virsh-nodedev: Resolve Coverity issues storage_backend: Resolve Coverity issue qemu_hostdev: Resolve Coverity issue examples/hellolibvirt/hellolibvirt.c | 7 +++++++ src/lxc/lxc_container.c | 5 +++++ src/qemu/qemu_hostdev.c | 3 ++- src/storage/storage_backend.c | 1 + src/util/vircommand.c | 5 +++++ tests/commandhelper.c | 6 +++++- tests/testutils.c | 3 +++ tools/virsh-domain-monitor.c | 2 +- tools/virsh-interface.c | 4 ++-- tools/virsh-nodedev.c | 2 +- 10 files changed, 32 insertions(+), 6 deletions(-) -- 1.8.1.4

2 9

[libvirt] [PATCH] util: add rebase fix that was accidentally omitted from previous patch
by Laine Stump 12 Jul '13

12 Jul '13

I had made the change locally, so make check and make syntax-check were successful, but forgot to add/commit. Unfortunately, git allows a push when the local directory is dirty, so it didn't catch my mistake. --- Pushed under the build breaker rule. git *was* very informative when I tried to switch branches - *then* it complained about the uncommitted local changes. No problem to push with outstanding local changes though. That seems like a bug to me; is there any situation where that would be a desired thing? src/util/virpci.c | 2 +- src/util/virusb.c | 2 +- 2 files changed, 2 insertions(+), 2 deletions(-) diff --git a/src/util/virpci.c b/src/util/virpci.c index ed55b80..3fa4a35 100644 --- a/src/util/virpci.c +++ b/src/util/virpci.c @@ -1669,7 +1669,7 @@ virPCIDeviceListAdd(virPCIDeviceListPtr list, _("Device %s is already in use"), dev->name); return -1; } - return VIR_APPEND_ELEMENT(list->devs, list->count, dev, true); + return VIR_APPEND_ELEMENT(list->devs, list->count, dev); } diff --git a/src/util/virusb.c b/src/util/virusb.c index 60222f4..e901618 100644 --- a/src/util/virusb.c +++ b/src/util/virusb.c @@ -450,7 +450,7 @@ virUSBDeviceListAdd(virUSBDeviceListPtr list, dev->name); return -1; } - return VIR_APPEND_ELEMENT(list->devs, list->count, dev, true); + return VIR_APPEND_ELEMENT(list->devs, list->count, dev); } virUSBDevicePtr -- 1.7.11.7

2 3

[libvirt] [PATCH] maint: update to latest gnulib
by Eric Blake 12 Jul '13

12 Jul '13

Future patches need LGPLv2+ versions of some modules that had recent license changes; but separating the gnulib update from the actual use of the modules makes it easier to backport to an older version while avoiding a submodule update (assuming, of course, that the backport is to a system where glibc provides adequate functionaliy without needing the gnulib module). * .gnulib: Update to latest, for modules needed in later patches. Signed-off-by: Eric Blake <eblake(a)redhat.com> --- Pushing under the trivial gnulib rule. * .gnulib b72ff2a...da8d59e (9): > regex: port to --with-included-regex --enable-gcc-warnings non-threaded > accept4, dup3, pipe2: port to Cygwin > mountlist: add support for deallocating returned list entries > stdalign, verify: port to FreeBSD 9.1, to C11, and to C++11 > fnmatch: don't goto over declaration > pipe2: decouple from binary-io a bit > mgetgroups: relax license to LGPLv2+ > xalloc-oversized: relax license to LGPLv2+ > nproc: relax license to LGPLv2+ .gnulib | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/.gnulib b/.gnulib index b72ff2a..da8d59e 160000 --- a/.gnulib +++ b/.gnulib @@ -1 +1 @@ -Subproject commit b72ff2a45efde544c406804186d37a3254728571 +Subproject commit da8d59ee79138b85daee1ad5b22ea12e4fb6ce47 -- 1.8.1.4

1 0

[libvirt] [PATCH v2 00/10] Re-enable memballoon driver statistics reporting
by John Ferlan 12 Jul '13

12 Jul '13

This patchset replaces: https://www.redhat.com/archives/libvir-list/2013-July/msg00108.html Changes since V1: * Remove the external (eg, qemu_monitor.{c,h}) API's to the qom-list, qom-get, qom-set interfaces * Handle errors in virXPathInt() return checking domain_conf.c * Move the 'period' from _qemuDomainObjPrivate into _qemuMonitor * Move/rename qemuProcessFindBalloonObjectPath from qemu_process.c into qemu_monitor.c as qemuMonitorFindBalloonObjectPath() * Create qemuMonitorJSONSetMemoryStatsPeriod() in qemu_monitor_json.c from parts of prior change to add qemuProcessUpdateBalloonStatsPeriod() in qemu_process.c and hav * Add qemuMonitorSetMemoryStatsPeriod() in qemu_monitor.c to set the collection period. Adjust the code to take parts of prior change and make check for balloon path in this path (set if necessary). * Adjust code in qemuProcessReconnect(), qemuProcessStart(), and qemuProcessAttach() to call qemuMonitorSetMemoryStatsPeriod() * Adjust qemuMonitorJSONGetMemoryStats() to get the balloon stats once the "actual" stat is collected from qemuMonitorJSONGetBalloonInfo(). Reviewer note: The prior code to get "actual" was a duplicate of the BalloonInfo code. Since "balloon-info" is where the actual memory is still stored it still must be called as that data is not present in the balloon stats qom-get "stats" fetch implemented as the new API qemuMonitorJSONGetBalloonStats(). * Avoid overloading virDomainSetMemoryFlags()... * Create/use virDomainSetMemoryStatsPeriodFlags() to set the collection period dynamically from "virsh dommemstats" using the --period qualifier. John Ferlan (10): Add qemuMonitorJSONGetObjectListPaths() method for QMP qom-list command Add qemuMonitorJSONGetObjectProperty() method for QMP qom-get command Add qemuMonitorJSONSetObjectProperty() method for QMP qom-set command Add 'period' for Memballoon statistics gathering capability Determine whether to start balloon memory stats gathering. Add capability to fetch balloon stats Add new public API virDomainSetMemoryStatsPeriodFlags Specify remote protocol for virDomainSetMemoryStatsPeriodFlags Implement the virDomainSetMemoryStatsPeriodFlags for QEMU driver Allow balloon driver collection to be adjusted dynamically docs/formatdomain.html.in | 19 ++ docs/schemas/domaincommon.rng | 7 + include/libvirt/libvirt.h.in | 3 + src/conf/domain_conf.c | 44 +++- src/conf/domain_conf.h | 1 + src/driver.h | 6 + src/libvirt.c | 65 ++++++ src/libvirt_public.syms | 5 + src/qemu/qemu_driver.c | 66 ++++++ src/qemu/qemu_monitor.c | 133 ++++++++++- src/qemu/qemu_monitor.h | 2 + src/qemu/qemu_monitor_json.c | 498 ++++++++++++++++++++++++++++++++---------- src/qemu/qemu_monitor_json.h | 64 ++++++ src/qemu/qemu_process.c | 21 +- src/remote/remote_driver.c | 1 + src/remote/remote_protocol.x | 15 +- src/remote_protocol-structs | 6 + tests/qemumonitorjsontest.c | 186 ++++++++++++++++ tools/virsh-domain-monitor.c | 66 +++++- 19 files changed, 1085 insertions(+), 123 deletions(-) -- 1.8.1.4

4 32

[libvirt] [PATCH 0/7] Resolve Coverity issues
by John Ferlan 12 Jul '13

12 Jul '13

The recent changes to change 'int i' to 'size_t i' seem to have uncovered a few new Coverity issues - resolve those. Also add a [dead_error_code] tag to a Coverity issue that's been there for a bit. After these patches the result is just back to 1 Coverity error in 'virLXCProcessReboot' regarding the usage of 'conn' which still needs to be eventually resolved, see: https://www.redhat.com/archives/libvir-list/2013-May/msg01218.html and subsequent followups John Ferlan (7): hellolibvirt: Resolve Coverity issues testutils: Resolve Coverity issues virsh-domain-monitor: Resolve Coverity issues virsh-interface: Resolve Coverity issues virsh-nodedev: Resolve Coverity issues storage_backend: Resolve Coverity issue qemu_hostdev: Resolve Coverity issue examples/hellolibvirt/hellolibvirt.c | 7 +++++++ src/qemu/qemu_hostdev.c | 12 +++++++----- src/storage/storage_backend.c | 1 + tests/testutils.c | 3 +++ tools/virsh-domain-monitor.c | 10 ++++++---- tools/virsh-interface.c | 24 +++++++++++++----------- tools/virsh-nodedev.c | 12 +++++++----- 7 files changed, 44 insertions(+), 25 deletions(-) -- 1.8.1.4

4 18

[libvirt] [PATCH RFC]:tolerate numa_node_size64 < 0 because nodeid might start from 1 instead of 0
by hejia hejia 11 Jul '13

11 Jul '13

In nodeGetFreeMemory/nodeGetCellsFreeMemory, they will calculate the free memory of every nodes. They assumed that nodeid of NUMA machine must be continuous and start from 0. But here is a counter-example: # numactl -H available: 1 nodes (1) node 1 cpus: 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 node 1 size: 16340 MB node 1 free: 11065 MB test results before this patch: #virsh freecell error: internal error Failed to query NUMA free memory #virsh freecell 0 error: internal error Failed to query NUMA free memory for node: 0 after this patch: # virsh freecell Total: 15772580 KiB # virsh freecell 0 0: 0 KiB --- libvirt/src/nodeinfo.c.orig 2013-07-08 04:25:11.970351101 -0500 +++ libvirt/src/nodeinfo.c 2013-07-08 09:00:30.834471495 -0500 @@ -1717,10 +1717,6 @@ nodeGetCellsFreeMemory(unsigned long lon for (numCells = 0, n = startCell; n <= lastCell; n++) { long long mem; if (numa_node_size64(n, &mem) < 0) { - virReportError(VIR_ERR_INTERNAL_ERROR, - _("Failed to query NUMA free memory for node: %d"), - n); - goto cleanup; } freeMems[numCells++] = mem; } @@ -1743,9 +1739,6 @@ nodeGetFreeMemory(void) for (n = 0; n <= numa_max_node(); n++) { long long mem; if (numa_node_size64(n, &mem) < 0) { - virReportError(VIR_ERR_INTERNAL_ERROR, - "%s", _("Failed to query NUMA free memory")); - goto cleanup; } freeMem += mem; }

3 4