- Devel - libvirt List Archives

[libvirt] [PATCH] util: Fix uninitalized variable to avoid garbage value.
by Julio Faracco 29 Apr '19

29 Apr '19

This commit is similar with 692400f4. It fixes an uninitialized variable to avoid garbage value. This case, returns 0 jiffies if an error occurs with virNetDevBridgeGet. Signed-off-by: Julio Faracco <jcfaracco(a)gmail.com> --- src/util/virnetdevbridge.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/src/util/virnetdevbridge.c b/src/util/virnetdevbridge.c index 0f301d9216..4f21a82fb7 100644 --- a/src/util/virnetdevbridge.c +++ b/src/util/virnetdevbridge.c @@ -713,7 +713,7 @@ int virNetDevBridgeGetSTPDelay(const char *brname, int *delayms) { int ret = -1; - unsigned long val; + unsigned long val = 0; ret = virNetDevBridgeGet(brname, "forward_delay", &val); *delayms = JIFFIES_TO_MS(val); -- 2.20.1

2 1

[libvirt] [PATCH 00/21] qemu: qapi: Schema query improvements and feature detection (blockdev-add saga)
by Peter Krempa 29 Apr '19

29 Apr '19

Allow detecting dynamic auto-read-only which will be possible in qemu 4.1 by fixing and improving QAPI schema queries. Note that the last two patches should not be pushed and will require tweaking once qemu 4.0 is released. Peter Krempa (21): tests: qemumonitorjson: Rename schema validation test cases qemu: qapi: Use automatic memory cleanup tests: qemumonitorjson: Add tests for QAPI schema query qemu: qapi: Fix return value of impossible case in virQEMUQAPISchemaTraverse qemu: qapi: Return schema entry via argument in virQEMUQAPISchemaTraverse qemu: qapi: Modify values returned by virQEMUQAPISchemaPathGet qemu: qapi: Convert virQEMUQAPISchemaTraverse to recursive lookup qemu: qapi: Separate virQEMUQAPISchemaTraverse into functions by object type qemu: qapi: Optimize out some helper functions qemu: qapi: Convert arguments of QAPI traversal helpers to a struct qemu: qapi: Rename local vars in virQEMUQAPISchemaTraverseObject qemu: qapi: Add helpers for virQEMUQAPISchemaTraverseContext qemu: qapi: Use declarative approach for meta-type parsers in virQEMUQAPISchemaTraverse qemu: qapi: Report schema and user errors for QAPI queries qemu: qapi: Prepare for extension of virQEMUQAPISchemaPathGet docs qemu: qapi: Implement worker for introspecting enums qemu: qapi: Implement worker for introspecting builtin types qemu: qapi: Implement worker for introspecting alternate types qemu: qapi: Implement support for 'features' DO NOT PUSH: bump capabilities to 4.0.0 + patches qemu: caps: Add capability for dynamic 'auto-read-only' support for files src/qemu/qemu_capabilities.c | 4 + src/qemu/qemu_capabilities.h | 3 + src/qemu/qemu_qapi.c | 419 ++- .../qemu_4.0.0.x86_64.xml | 2 +- .../caps_4.0.0.x86_64.replies | 3167 +++++++++-------- .../caps_4.0.0.x86_64.xml | 27 +- tests/qemumonitorjsontest.c | 113 +- 7 files changed, 2081 insertions(+), 1654 deletions(-) -- 2.20.1

2 42

[libvirt] [PATCH] test: Use nodeInfo content for virNodeGetCPUMap
by Cole Robinson 29 Apr '19

29 Apr '19

Right now the test driver implementation of virNodeGetCPUMap uses hardcoded data. Instead use cpu data from our nodeInfo, which can be overridden with input test XML. This helps to emulate some virsh commands (vcpupin, hostcpumap) with a large number of host cpus. [1] The content reported for test:///default is now slightly different: previously it would report 8 total cpus with 0,2,4 online. Now it reports 16 cpus with all online, which is the pre-existing nodeInfo default. [1] https://www.redhat.com/archives/libvir-list/2019-March/msg02042.html Signed-off-by: Cole Robinson <crobinso(a)redhat.com> --- John, in trying to understand and test your bitmap patch, I added this to the test driver. I think there's a bigger issue at play but I'm still trying to work it out, I'll send a follow up email when I've collected my thoughts src/test/test_driver.c | 35 +++++++++++++++++++++++++++++------ 1 file changed, 29 insertions(+), 6 deletions(-) diff --git a/src/test/test_driver.c b/src/test/test_driver.c index d5eecf4b7f..f00b6eb1c8 100644 --- a/src/test/test_driver.c +++ b/src/test/test_driver.c @@ -5811,23 +5811,46 @@ static int testConnectListAllDomains(virConnectPtr conn, } static int -testNodeGetCPUMap(virConnectPtr conn ATTRIBUTE_UNUSED, +testNodeGetCPUMap(virConnectPtr conn, unsigned char **cpumap, unsigned int *online, unsigned int flags) { + testDriverPtr privconn = conn->privateData; + int maxcpus = VIR_NODEINFO_MAXCPUS(privconn->nodeInfo); + virBitmapPtr cpus = NULL; + int ret = -1; + int dummy; + size_t i; + virCheckFlags(0, -1); + if (!cpumap && !online) + return maxcpus; + + if (!(cpus = virBitmapNew(maxcpus))) + goto cleanup; + if (cpumap) { - if (VIR_ALLOC_N(*cpumap, 1) < 0) - return -1; - *cpumap[0] = 0x15; + for (i = 0; i < privconn->nodeInfo.cpus; i++) { + if (virBitmapSetBit(cpus, i) < 0) + goto cleanup; + } + + if (virBitmapToData(cpus, cpumap, &dummy) < 0) + goto cleanup; } if (online) - *online = 3; + *online = privconn->nodeInfo.cpus; - return 8; + ret = maxcpus; + + cleanup: + if (ret < 0 && cpumap) + VIR_FREE(*cpumap); + VIR_FREE(cpus); + return ret; } static char * -- 2.21.0

2 1

[libvirt] [PATCH] docs: Fix typo in the firmware value
by Fabiano Fidêncio 29 Apr '19

29 Apr '19

firmware attribute from <os/> takes either 'efi' or 'bios' as its allowed values. However, the current documentation mistakenly mentions 'uefi' instead of 'efi'. Signed-off-by: Fabiano Fidêncio <fidencio(a)redhat.com> --- docs/formatdomain.html.in | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/docs/formatdomain.html.in b/docs/formatdomain.html.in index d27ab07cbb..e1da878fcc 100644 --- a/docs/formatdomain.html.in +++ b/docs/formatdomain.html.in @@ -128,7 +128,7 @@ <pre> ... -<os firmware='uefi'> +<os firmware='efi'> <type>hvm</type> <loader readonly='yes' secure='no' type='rom'>/usr/lib/xen/boot/hvmloader</loader> <nvram template='/usr/share/OVMF/OVMF_VARS.fd'>/var/lib/libvirt/nvram/guest_VARS.fd</nvram> -- 2.21.0

2 1

[libvirt] [PATCH 0/2] Fix a couple of minor doc issues
by John Ferlan 29 Apr '19

29 Apr '19

The QE processing of the feature found a couple of minor doc type issues so just fix an error message and add the command to the man page. Both are rather trivial, but I can wait for reviews rather than push directly as trivial.. John Ferlan (2): conf: Fix typo in error message virsh: Add source-protocol-ver to pool-define-as docs src/conf/storage_conf.c | 2 +- tools/virsh.pod | 1 + 2 files changed, 2 insertions(+), 1 deletion(-) -- 2.20.1

2 3

[libvirt] [PATCH for 5.3.0] storage_backend_rbd: Don't fail iterating over snapshots
by Michal Privoznik 29 Apr '19

29 Apr '19

When cloning an RBD volume we try to find a snapshot which is not different to the image we're trying to clone. This boils down to calling rbd_diff_iterate() or rbd_diff_iterate2() on systems with newer ceph. These two are passed a callback - virStorageBackendRBDIterateCb() which mistakenly states that if a negative value is returned from the callback, the iteration is stopped but no error is reported. Well, that is not the case. I've looked into librbd sources and found that a negative value does mean stop of iteration but it also means an error. That's why any attempt to clone an RBD image results in the following error: virsh # vol-clone --pool rbd_image_root coreos_2023 coreos00.disk error: Failed to clone vol from coreos_2023 error: failed to iterate RBD snapshot coreos_2023@base: Operation not permitted Reported on libvirt-users list: https://www.redhat.com/archives/libvirt-users/2019-April/msg00060.html Signed-off-by: Michal Privoznik <mprivozn(a)redhat.com> --- src/storage/storage_backend_rbd.c | 7 +------ 1 file changed, 1 insertion(+), 6 deletions(-) diff --git a/src/storage/storage_backend_rbd.c b/src/storage/storage_backend_rbd.c index f8c968e682..3fe188551c 100644 --- a/src/storage/storage_backend_rbd.c +++ b/src/storage/storage_backend_rbd.c @@ -1031,14 +1031,9 @@ virStorageBackendRBDIterateCb(uint64_t offset ATTRIBUTE_UNUSED, { /* * Just set that there is a diff for this snapshot, we do not care where - * - * When it returns a negative number the rbd_diff_iterate() function will stop - * - * That's why we return -1, meaning that there is a difference and we can stop - * searching any further. */ *(int*) arg = 1; - return -1; + return 0; } static int -- 2.21.0

2 2

[libvirt] [python PATCH] Allow virConnect.getDomainCapabilities() to have no arguments
by Michal Privoznik 29 Apr '19

29 Apr '19

Our C API allows no values to be passed (well, it accepts all NULLs). There's no reason that python binding should require all arguments. Signed-off-by: Michal Privoznik <mprivozn(a)redhat.com> --- libvirt-override-api.xml | 10 ++++++++++ 1 file changed, 10 insertions(+) diff --git a/libvirt-override-api.xml b/libvirt-override-api.xml index 7f578e0..53b15c5 100644 --- a/libvirt-override-api.xml +++ b/libvirt-override-api.xml @@ -754,5 +754,15 @@ <arg name='conn' type='virConnectPtr' info='pointer to hypervisor connection'/> <arg name='flags' type='int' info='unused, always pass 0'/> </function> + <function name='virConnectGetDomainCapabilities' file='libvirt-domain' module='libvirt-domain'> + <info>Query domain capabilities</info> + <return type='char *' info='NULL in case of error or an XML string defining the capabilities.'/> + <arg name='conn' type='virConnectPtr' info='pointer to the hypervisor connection'/> + <arg name='emulatorbin' type='const char *' info='optional path to emulator'/> + <arg name='arch' type='const char *' info='optional domain architecture'/> + <arg name='machine' type='const char *' info='optional machine type'/> + <arg name='virttype' type='const char *' info='optional virtualization type'/> + <arg name='flags' type='int' info='unused, always pass 0'/> + </function> </symbols> </api> -- 2.21.0

2 1

[libvirt] [PATCH] CONF: bug, release 'alloc' object in cleanup
by Wang Huaqiang 28 Apr '19

28 Apr '19

In case error happens before 'alloc' is transferred to 'resctrl', the 'alloc' object should be released. Signed-off-by: Wang Huaqiang <huaqiang.wang(a)intel.com> --- src/conf/domain_conf.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/src/conf/domain_conf.c b/src/conf/domain_conf.c index 6331424..7d20276 100644 --- a/src/conf/domain_conf.c +++ b/src/conf/domain_conf.c @@ -19396,6 +19396,7 @@ virDomainCachetuneDefParse(virDomainDefPtr def, ret = 0; cleanup: + virObjectUnref(alloc); virDomainResctrlDefFree(resctrl); return ret; } @@ -19597,6 +19598,7 @@ virDomainMemorytuneDefParse(virDomainDefPtr def, ret = 0; cleanup: + virObjectUnref(alloc); virDomainResctrlDefFree(resctrl); return ret; } -- 2.7.4

3 2

[libvirt] [PATCH 0/4] Couple of almost trivial disk cold-/hot-plug fixes
by Michal Privoznik 28 Apr '19

28 Apr '19

*** BLURB HERE *** Michal Prívozník (4): src: Check for virDomainDiskInsert() retval properly qemuDomainAttachDeviceLiveAndConfig: Don't overwrite @ret qemu: On attach to live XML check for user alias collision only live XML qemu: Check for user alias collisions in coldplug src/libxl/libxl_driver.c | 2 +- src/lxc/lxc_driver.c | 2 +- src/qemu/qemu_driver.c | 32 ++++++++++++++++++-------------- 3 files changed, 20 insertions(+), 16 deletions(-) -- 2.21.0

3 10

[libvirt] [PATCH] qemu: Add entry for balloon stats stat-htlb-pgalloc and stat-htlb-pgfail
by Han Han 27 Apr '19

27 Apr '19

Qemu added reporting of virtio balloon new statistics stat-htlb-pgalloc and stat-htlb-pgfail since qemu-3.0 commit b7b12644297. The value of stat-htlb-pgalloc represents the number of successful hugetlb page allocations while stat-htlb-pgfail represents the number of failed ones. Add this statistics reporting to libvirt. Signed-off-by: Han Han <hhan(a)redhat.com> --- include/libvirt/libvirt-domain.h | 14 +++++++++++++- src/libvirt-domain.c | 5 ++++- src/qemu/qemu_driver.c | 2 ++ src/qemu/qemu_monitor_json.c | 5 +++++ tools/virsh-domain-monitor.c | 4 ++++ tools/virsh.pod | 2 ++ 6 files changed, 30 insertions(+), 2 deletions(-) diff --git a/include/libvirt/libvirt-domain.h b/include/libvirt/libvirt-domain.h index 7d36820b5a..192d25c1db 100644 --- a/include/libvirt/libvirt-domain.h +++ b/include/libvirt/libvirt-domain.h @@ -636,11 +636,23 @@ typedef enum { */ VIR_DOMAIN_MEMORY_STAT_DISK_CACHES = 10, + /* + * The amount of successful hugetlb(Huge Page Tables) allocations via + * virtio balloon. + */ + VIR_DOMAIN_MEMORY_STAT_HUGETLB_PGALLOC = 11, + + /* + * The amount of failed hugetlb(Huge Page Tables) allocations via + * virtio balloon. + */ + VIR_DOMAIN_MEMORY_STAT_HUGETLB_PGFAIL = 12, + /* * The number of statistics supported by this version of the interface. * To add new statistics, add them to the enum and increase this value. */ - VIR_DOMAIN_MEMORY_STAT_NR = 11, + VIR_DOMAIN_MEMORY_STAT_NR = 13, # ifdef VIR_ENUM_SENTINELS VIR_DOMAIN_MEMORY_STAT_LAST = VIR_DOMAIN_MEMORY_STAT_NR diff --git a/src/libvirt-domain.c b/src/libvirt-domain.c index 80284a99f0..a95690e8a4 100644 --- a/src/libvirt-domain.c +++ b/src/libvirt-domain.c @@ -5734,7 +5734,10 @@ virDomainGetInterfaceParameters(virDomainPtr domain, * VIR_DOMAIN_MEMORY_STAT_DISK_CACHES * Memory that can be reclaimed without additional I/O, typically disk * caches (in kb). - * + * VIR_DOMAIN_MEMORY_STAT_HUGETLB_PGALLOC + * The amount of successful hugetlb(Huge Page Tables) allocations + * VIR_DOMAIN_MEMORY_STAT_HUGETLB_PGFAIL + * The amount of failed hugetlb(Huge Page Tables) allocations * Returns: The number of stats provided or -1 in case of failure. */ int diff --git a/src/qemu/qemu_driver.c b/src/qemu/qemu_driver.c index f48d9256e4..2d849beac8 100644 --- a/src/qemu/qemu_driver.c +++ b/src/qemu/qemu_driver.c @@ -20245,6 +20245,8 @@ qemuDomainGetStatsBalloon(virQEMUDriverPtr driver, STORE_MEM_RECORD(LAST_UPDATE, "last-update") STORE_MEM_RECORD(USABLE, "usable") STORE_MEM_RECORD(DISK_CACHES, "disk_caches") + STORE_MEM_RECORD(HUGETLB_PGALLOC, "hugetlb_pgalloc") + STORE_MEM_RECORD(HUGETLB_PGFAIL, "hugetlb_pgfail") } #undef STORE_MEM_RECORD diff --git a/src/qemu/qemu_monitor_json.c b/src/qemu/qemu_monitor_json.c index 908967f46c..5e3df5e759 100644 --- a/src/qemu/qemu_monitor_json.c +++ b/src/qemu/qemu_monitor_json.c @@ -2150,6 +2150,11 @@ int qemuMonitorJSONGetMemoryStats(qemuMonitorPtr mon, VIR_DOMAIN_MEMORY_STAT_LAST_UPDATE, 1); GET_BALLOON_STATS(statsdata, "stat-disk-caches", VIR_DOMAIN_MEMORY_STAT_DISK_CACHES, 1024); + GET_BALLOON_STATS(statsdata, "stat-htlb-pgalloc", + VIR_DOMAIN_MEMORY_STAT_HUGETLB_PGALLOC, 1); + GET_BALLOON_STATS(statsdata, "stat-htlb-pgfail", + VIR_DOMAIN_MEMORY_STAT_HUGETLB_PGFAIL, 1); + ret = got; cleanup: virJSONValueFree(cmd); diff --git a/tools/virsh-domain-monitor.c b/tools/virsh-domain-monitor.c index d87475f6f6..0092a9786e 100644 --- a/tools/virsh-domain-monitor.c +++ b/tools/virsh-domain-monitor.c @@ -376,6 +376,10 @@ cmdDomMemStat(vshControl *ctl, const vshCmd *cmd) vshPrint(ctl, "last_update %llu\n", stats[i].val); if (stats[i].tag == VIR_DOMAIN_MEMORY_STAT_DISK_CACHES) vshPrint(ctl, "disk_caches %llu\n", stats[i].val); + if (stats[i].tag == VIR_DOMAIN_MEMORY_STAT_HUGETLB_PGALLOC) + vshPrint(ctl, "hugetlb_pgalloc %llu\n", stats[i].val); + if (stats[i].tag == VIR_DOMAIN_MEMORY_STAT_HUGETLB_PGFAIL) + vshPrint(ctl, "hugetlb_pgfail %llu\n", stats[i].val); } ret = true; diff --git a/tools/virsh.pod b/tools/virsh.pod index afc1684db0..ef27c527d6 100644 --- a/tools/virsh.pod +++ b/tools/virsh.pod @@ -938,6 +938,8 @@ without causing host swapping (in KiB) last-update - Timestamp of the last update of statistics (in seconds) disk_caches - The amount of memory that can be reclaimed without additional I/O, typically disk caches (in KiB) + hugetlb_pgalloc - The number of successful huge page table allocations + hugetlb_pgfail - The number of failed huge page table allocations For QEMU/KVM with a memory balloon, setting the optional I<--period> to a value larger than 0 in seconds will allow the balloon driver to return -- 2.20.1

1 0