[libvirt] [REPOST PATCH v2 0/9] Add group_name support for <iotune>
by John Ferlan
This is just a REPOST of the v2 series:
http://www.redhat.com/archives/libvir-list/2016-November/msg00363.html
The only difference being updating to the current top of tree
of commit id '0b4c3bd30'.
I did *not* add the NEWS change yet as that's newer than this, but will
update NEWS with this once/if this is ACK'd using whatever format becomes
agreed upon for the file contents/formatting options.
John Ferlan (9):
include: Add new "group_name" definition for iotune throttling
caps: Add new capability for the iotune group name
qemu: Adjust maxparams logic for qemuDomainGetBlockIoTune
qemu: Alter qemuMonitorJSONSetBlockIoThrottle command logic
qemu: Adjust various bool BlockIoTune set_ values into mask
qemu: Add support for parsing iotune group setting
conf: Add support for blkiotune group_name option
qemu: Add the group name option to the iotune command line
virsh: Add group name to blkdeviotune output
docs/formatdomain.html.in | 11 ++
docs/schemas/domaincommon.rng | 5 +
include/libvirt/libvirt-domain.h | 15 ++
src/conf/domain_conf.c | 10 ++
src/conf/domain_conf.h | 1 +
src/qemu/qemu_capabilities.c | 2 +
src/qemu/qemu_capabilities.h | 1 +
src/qemu/qemu_command.c | 13 ++
src/qemu/qemu_driver.c | 177 +++++++++++++--------
src/qemu/qemu_monitor.c | 2 +
src/qemu/qemu_monitor.h | 1 +
src/qemu/qemu_monitor_json.c | 97 ++++++-----
src/qemu/qemu_monitor_json.h | 1 +
tests/qemucapabilitiesdata/caps_2.4.0.x86_64.xml | 1 +
tests/qemucapabilitiesdata/caps_2.5.0.x86_64.xml | 1 +
.../caps_2.6.0-gicv2.aarch64.xml | 1 +
.../caps_2.6.0-gicv3.aarch64.xml | 1 +
tests/qemucapabilitiesdata/caps_2.6.0.ppc64le.xml | 1 +
tests/qemucapabilitiesdata/caps_2.6.0.x86_64.xml | 1 +
tests/qemucapabilitiesdata/caps_2.7.0.x86_64.xml | 1 +
tests/qemumonitorjsontest.c | 88 +++++++---
.../qemuxml2argv-blkdeviotune-group-num.args | 32 ++++
.../qemuxml2argv-blkdeviotune-group-num.xml | 61 +++++++
tests/qemuxml2argvtest.c | 4 +
.../qemuxml2xmlout-blkdeviotune-group-num.xml | 1 +
tests/qemuxml2xmltest.c | 1 +
tools/virsh-domain.c | 17 ++
tools/virsh.pod | 5 +-
28 files changed, 428 insertions(+), 124 deletions(-)
create mode 100644 tests/qemuxml2argvdata/qemuxml2argv-blkdeviotune-group-num.args
create mode 100644 tests/qemuxml2argvdata/qemuxml2argv-blkdeviotune-group-num.xml
create mode 120000 tests/qemuxml2xmloutdata/qemuxml2xmlout-blkdeviotune-group-num.xml
--
2.7.4
7 years, 11 months
[libvirt] [PATCH] cpu: Add support for pku and ospke Intel features for Memory Protection Keys
by Lin Ma
qemu commit: f74eefe0
https://lwn.net/Articles/667156/
Signed-off-by: Lin Ma <lma(a)suse.com>
---
src/cpu/cpu_map.xml | 7 +++++++
1 file changed, 7 insertions(+)
diff --git a/src/cpu/cpu_map.xml b/src/cpu/cpu_map.xml
index 6da8321..dca5720 100644
--- a/src/cpu/cpu_map.xml
+++ b/src/cpu/cpu_map.xml
@@ -255,6 +255,13 @@
<cpuid eax_in='0x07' ebx='0x10000000'/>
</feature>
+ <feature name='pku'>
+ <cpuid eax_in='0x07' ecx='0x00000008'/>
+ </feature>
+ <feature name='ospke'>
+ <cpuid eax_in='0x07' ecx='0x00000010'/>
+ </feature>
+
<!-- Processor Extended State Enumeration sub leaf 1 -->
<feature name='xsaveopt'>
<cpuid eax_in='0x0d' ecx_in='0x01' eax='0x00000001'/>
--
2.9.2
7 years, 11 months
[libvirt] [PATCH] cpu: Add support for more AVX512 Intel features
by Lin Ma
These features are included:
AVX512DQ, AVX512IFMA, AVX512BW, AVX512VL, AVX512VBMI, AVX512_4VNNIW and
AVX512_4FMAPS.
qemu commits: cc728d14 and 95ea69fb
Signed-off-by: Lin Ma <lma(a)suse.com>
---
src/cpu/cpu_map.xml | 24 ++++++++++++++++++++++++
1 file changed, 24 insertions(+)
diff --git a/src/cpu/cpu_map.xml b/src/cpu/cpu_map.xml
index 6da8321..e9292e1 100644
--- a/src/cpu/cpu_map.xml
+++ b/src/cpu/cpu_map.xml
@@ -233,6 +233,9 @@
<feature name='avx512f'> <!-- AVX-512 Foundation -->
<cpuid eax_in='0x07' ebx='0x00010000'/>
</feature>
+ <feature name='avx512dq'> <!-- AVX-512 Doubleword & Quadword Instrs -->
+ <cpuid eax_in='0x07' ebx='0x00020000'/>
+ </feature>
<feature name='rdseed'>
<cpuid eax_in='0x07' ebx='0x00040000'/>
</feature>
@@ -242,6 +245,9 @@
<feature name='smap'>
<cpuid eax_in='0x07' ebx='0x00100000'/>
</feature>
+ <feature name='avx512ifma'> <!-- AVX-512 Integer Fused Multiply Add -->
+ <cpuid eax_in='0x07' ebx='0x00200000'/>
+ </feature>
<feature name='clflushopt'>
<cpuid eax_in='0x07' ebx='0x00800000'/>
</feature>
@@ -254,6 +260,24 @@
<feature name='avx512cd'> <!-- AVX-512 Conflict Detection -->
<cpuid eax_in='0x07' ebx='0x10000000'/>
</feature>
+ <feature name='avx512bw'> <!-- AVX-512 Byte and Word Instructions -->
+ <cpuid eax_in='0x07' ebx='0x40000000'/>
+ </feature>
+ <feature name='avx512vl'> <!-- AVX-512 Vector Length Extensions -->
+ <cpuid eax_in='0x07' ebx='0x80000000'/>
+ </feature>
+
+ <feature name='avx512vbmi'> <!-- AVX-512 Vector Byte Manipulation Instrs -->
+ <cpuid eax_in='0x07' ecx='0x00000002'/>
+ </feature>
+
+ <feature name='avx512-4vnniw'> <!-- AVX-512 Neural Network Instructions -->
+ <cpuid eax_in='0x07' edx='0x00000004'/>
+ </feature>
+ <!-- AVX-512 Multiply Accumulation Single Precision -->
+ <feature name='avx512-4fmaps'>
+ <cpuid eax_in='0x07' edx='0x00000008'/>
+ </feature>
<!-- Processor Extended State Enumeration sub leaf 1 -->
<feature name='xsaveopt'>
--
2.9.2
7 years, 11 months
[libvirt] [PATCH 0/5] Reject duplicate drive addresses
by Marc Hartmayer
Reject duplicate drive addresses at domain definition. Hot-plug for
disks and hostdevs is still to fix and this is why the old version of
drive address checking is still there. Additionally, it isn't that
easy to be sure that the changes won't break other drivers as these
are common code changes.
Marc Hartmayer (5):
conf: simplify functions virDomainSCSIDriveAddressIsUsedBy*()
conf: virDomainDriveAddressIsUsedByDisk: Rename type to bus_type
tests: don't use duplicate disk addresses
conf: add global check for duplicate drive addresses
tests: add test cases for address conflicts
src/conf/domain_conf.c | 184 +++++++++++++++++----
.../qemuxml2argv-disk-drive-address-conflict.xml | 27 +++
...xml2argv-disk-hostdev-scsi-address-conflict.xml | 30 ++++
...emuxml2argv-hostdevs-drive-address-conflict.xml | 33 ++++
.../qemuxml2argv-seclabel-dynamic-override.args | 4 +-
.../qemuxml2argv-seclabel-dynamic-override.xml | 2 +-
tests/qemuxml2argvtest.c | 8 +
.../qemuxml2xmlout-seclabel-dynamic-override.xml | 2 +-
8 files changed, 251 insertions(+), 39 deletions(-)
create mode 100644 tests/qemuxml2argvdata/qemuxml2argv-disk-drive-address-conflict.xml
create mode 100644 tests/qemuxml2argvdata/qemuxml2argv-disk-hostdev-scsi-address-conflict.xml
create mode 100644 tests/qemuxml2argvdata/qemuxml2argv-hostdevs-drive-address-conflict.xml
--
2.5.5
7 years, 11 months
[libvirt] [PATCH v2 1/1] gluster: cache glfs connection object per volume
by prasanna.kalever＠redhat.com
From: Prasanna Kumar Kalever <prasanna.kalever(a)redhat.com>
This patch optimizes calls to glfs_init() and friends
Currently, a start of a VM will call 2 glfs_new/glfs_init (which will create
glfs object, once for stat, read headers and next to chown) and then will fork
qemu process which will call once again (for actual read write IO).
Not that all, in case if we are have 4 extra attached disks, then the total
calls to glfs_init() and friends will be (4+1)*2 in libvirt and (4+1)*1 in
qemu space i.e 15 calls. Since we don't have control over qemu process as that
executes in a different process environment, lets do not bother much about it.
This patch shrinks these 10 calls (i.e objects from above example) to just
one, by maintaining a cache of glfs objects.
Additionally snapshot(external) scenario will further complex the situation ...
The glfs object is shared across other only if volume name and all the
volfile servers match (includes hostname, transport and port number).
In case of hit glfs object takes a ref and on close unref happens.
Thanks to 'Peter Krempa' for all the inputs.
Signed-off-by: Prasanna Kumar Kalever <prasanna.kalever(a)redhat.com>
---
v2: Address review comments from Peter on v1
* Rebased on latest master
* Changes to commit msg
* Introduce storage API's for Register and Unregister of volume
* During qemu process Start/Stop and snapshot create
* Check Transport and Port type
* Atomic element add/del to list and ref counting
Pending: Treating IP and FQDN belong to same host
v1: Initial patch
---
src/qemu/qemu_domain.c | 2 +-
src/qemu/qemu_domain.h | 5 +
src/qemu/qemu_driver.c | 29 ++++
src/qemu/qemu_process.c | 25 +++
src/storage/storage_backend_fs.c | 2 +
src/storage/storage_backend_gluster.c | 295 +++++++++++++++++++++++++++++++---
src/storage/storage_driver.c | 23 ++-
src/storage/storage_driver.h | 3 +
8 files changed, 357 insertions(+), 27 deletions(-)
diff --git a/src/qemu/qemu_domain.c b/src/qemu/qemu_domain.c
index 47332a8..35914c5 100644
--- a/src/qemu/qemu_domain.c
+++ b/src/qemu/qemu_domain.c
@@ -4837,7 +4837,7 @@ qemuDomainCleanupRun(virQEMUDriverPtr driver,
priv->ncleanupCallbacks_max = 0;
}
-static void
+void
qemuDomainGetImageIds(virQEMUDriverConfigPtr cfg,
virDomainObjPtr vm,
virStorageSourcePtr src,
diff --git a/src/qemu/qemu_domain.h b/src/qemu/qemu_domain.h
index 7650ff3..a9e38bd 100644
--- a/src/qemu/qemu_domain.h
+++ b/src/qemu/qemu_domain.h
@@ -591,6 +591,11 @@ bool qemuDomainDiskSourceDiffers(virDomainDiskDefPtr disk,
bool qemuDomainDiskChangeSupported(virDomainDiskDefPtr disk,
virDomainDiskDefPtr orig_disk);
+void qemuDomainGetImageIds(virQEMUDriverConfigPtr cfg,
+ virDomainObjPtr vm,
+ virStorageSourcePtr src,
+ uid_t *uid, gid_t *gid);
+
int qemuDomainStorageFileInit(virQEMUDriverPtr driver,
virDomainObjPtr vm,
virStorageSourcePtr src);
diff --git a/src/qemu/qemu_driver.c b/src/qemu/qemu_driver.c
index 3517aa2..7cae094 100644
--- a/src/qemu/qemu_driver.c
+++ b/src/qemu/qemu_driver.c
@@ -14211,6 +14211,7 @@ qemuDomainSnapshotCreateSingleDiskActive(virQEMUDriverPtr driver,
cleanup:
if (need_unlink && virStorageFileUnlink(newDiskSrc))
VIR_WARN("unable to unlink just-created %s", source);
+ virStorageFileDeinit(disk->src);
virStorageFileDeinit(newDiskSrc);
virStorageSourceFree(newDiskSrc);
virStorageSourceFree(persistDiskSrc);
@@ -14566,6 +14567,7 @@ qemuDomainSnapshotCreateXML(virDomainPtr domain,
virDomainSnapshotObjPtr snap = NULL;
virDomainSnapshotPtr snapshot = NULL;
virDomainSnapshotDefPtr def = NULL;
+ virDomainSnapshotDefPtr refDef = NULL;
bool update_current = true;
bool redefine = flags & VIR_DOMAIN_SNAPSHOT_CREATE_REDEFINE;
unsigned int parse_flags = VIR_DOMAIN_SNAPSHOT_PARSE_DISKS;
@@ -14574,6 +14576,9 @@ qemuDomainSnapshotCreateXML(virDomainPtr domain,
bool align_match = true;
virQEMUDriverConfigPtr cfg = NULL;
virCapsPtr caps = NULL;
+ unsigned int dIndex;
+ uid_t uid;
+ gid_t gid;
virCheckFlags(VIR_DOMAIN_SNAPSHOT_CREATE_REDEFINE |
VIR_DOMAIN_SNAPSHOT_CREATE_CURRENT |
@@ -14690,6 +14695,19 @@ qemuDomainSnapshotCreateXML(virDomainPtr domain,
qemuDomainObjSetAsyncJobMask(vm, QEMU_JOB_NONE);
+ for (dIndex = 0; dIndex < def->ndisks; dIndex++) {
+ virDomainSnapshotDiskDef disk = def->disks[dIndex];
+
+ if (virStorageSourceIsEmpty(disk.src))
+ continue;
+
+ qemuDomainGetImageIds(cfg, vm, disk.src, &uid, &gid);
+
+ if (virStorageVolumeRegister(disk.src, uid, gid) <0)
+ goto cleanup;
+ }
+
+
if (redefine) {
if (virDomainSnapshotRedefinePrep(domain, vm, &def, &snap,
&update_current, flags) < 0)
@@ -14800,6 +14818,17 @@ qemuDomainSnapshotCreateXML(virDomainPtr domain,
snapshot = virGetDomainSnapshot(domain, snap->def->name);
endjob:
+ refDef = (!snap) ? def : snap->def;
+
+ for (dIndex = 0; dIndex < refDef->ndisks; dIndex++) {
+ virDomainSnapshotDiskDef disk = refDef->disks[dIndex];
+
+ if (virStorageSourceIsEmpty(disk.src))
+ continue;
+
+ virStorageVolumeUnRegister(disk.src);
+ }
+
if (snapshot && !(flags & VIR_DOMAIN_SNAPSHOT_CREATE_NO_METADATA)) {
if (qemuDomainSnapshotWriteMetadata(vm, snap, driver->caps,
cfg->snapshotDir) < 0) {
diff --git a/src/qemu/qemu_process.c b/src/qemu/qemu_process.c
index ecd7ded..20793cf 100644
--- a/src/qemu/qemu_process.c
+++ b/src/qemu/qemu_process.c
@@ -4612,6 +4612,9 @@ qemuProcessInit(virQEMUDriverPtr driver,
qemuDomainObjPrivatePtr priv = vm->privateData;
int stopFlags;
int ret = -1;
+ unsigned int dIndex;
+ uid_t uid;
+ gid_t gid;
VIR_DEBUG("vm=%p name=%s id=%d migration=%d",
vm, vm->def->name, vm->def->id, migration);
@@ -4664,6 +4667,18 @@ qemuProcessInit(virQEMUDriverPtr driver,
if (qemuDomainSetPrivatePaths(driver, vm) < 0)
goto cleanup;
+ for (dIndex = 0; dIndex < vm->def->ndisks; dIndex++) {
+ virDomainDiskDefPtr disk = vm->def->disks[dIndex];
+
+ if (virStorageSourceIsEmpty(disk->src))
+ continue;
+
+ qemuDomainGetImageIds(cfg, vm, disk->src, &uid, &gid);
+
+ if (virStorageVolumeRegister(disk->src, uid, gid) < 0)
+ goto cleanup;
+ }
+
ret = 0;
cleanup:
@@ -5671,6 +5686,7 @@ qemuProcessFinishStartup(virConnectPtr conn,
{
virQEMUDriverConfigPtr cfg = virQEMUDriverGetConfig(driver);
int ret = -1;
+ unsigned int dIndex;
if (startCPUs) {
VIR_DEBUG("Starting domain CPUs");
@@ -5695,6 +5711,15 @@ qemuProcessFinishStartup(virConnectPtr conn,
VIR_HOOK_SUBOP_BEGIN) < 0)
goto cleanup;
+ for (dIndex = 0; dIndex < vm->def->ndisks; dIndex++) {
+ virDomainDiskDefPtr disk = vm->def->disks[dIndex];
+
+ if (virStorageSourceIsEmpty(disk->src))
+ continue;
+
+ virStorageVolumeUnRegister(disk->src);
+ }
+
ret = 0;
cleanup:
diff --git a/src/storage/storage_backend_fs.c b/src/storage/storage_backend_fs.c
index de0e8d5..0e03e06 100644
--- a/src/storage/storage_backend_fs.c
+++ b/src/storage/storage_backend_fs.c
@@ -1488,6 +1488,8 @@ virStorageFileBackendFileDeinit(virStorageSourcePtr src)
VIR_FREE(priv->canonpath);
VIR_FREE(priv);
+
+ VIR_FREE(src->drv);
}
diff --git a/src/storage/storage_backend_gluster.c b/src/storage/storage_backend_gluster.c
index 8e86704..18b7fc3 100644
--- a/src/storage/storage_backend_gluster.c
+++ b/src/storage/storage_backend_gluster.c
@@ -31,11 +31,34 @@
#include "virstoragefile.h"
#include "virstring.h"
#include "viruri.h"
+#include "viratomic.h"
#define VIR_FROM_THIS VIR_FROM_STORAGE
VIR_LOG_INIT("storage.storage_backend_gluster");
+static bool virGlusterGlobalError;
+static virOnceControl glusterConnOnce = VIR_ONCE_CONTROL_INITIALIZER;
+
+typedef struct _virStorageBackendGlusterState virStorageBackendGlusterState;
+typedef virStorageBackendGlusterState *virStorageBackendGlusterStatePtr;
+
+typedef struct _virStorageFileBackendGlusterPriv virStorageFileBackendGlusterPriv;
+typedef virStorageFileBackendGlusterPriv *virStorageFileBackendGlusterPrivPtr;
+
+typedef struct _unixSocketAddress unixSocketAddress;
+
+typedef struct _inetSocketAddress inetSocketAddress;
+
+typedef struct _glusterServer glusterServer;
+typedef struct _glusterServer *glusterServerPtr;
+
+typedef struct _virStorageBackendGlusterStatePreopened virStorageBackendGlusterStatePreopened;
+typedef virStorageBackendGlusterStatePreopened *virStorageBackendGlusterStatePtrPreopened;
+
+typedef struct _virStorageBackendGlusterconnCache virStorageBackendGlusterconnCache;
+typedef virStorageBackendGlusterconnCache *virStorageBackendGlusterconnCachePtr;
+
struct _virStorageBackendGlusterState {
glfs_t *vol;
@@ -47,8 +70,241 @@ struct _virStorageBackendGlusterState {
char *dir; /* dir from URI, or "/"; always starts and ends in '/' */
};
-typedef struct _virStorageBackendGlusterState virStorageBackendGlusterState;
-typedef virStorageBackendGlusterState *virStorageBackendGlusterStatePtr;
+struct _virStorageFileBackendGlusterPriv {
+ glfs_t *vol;
+ char *canonpath;
+};
+
+struct _unixSocketAddress {
+ char *path;
+};
+
+struct _inetSocketAddress {
+ char *host;
+ char *port;
+};
+
+struct _glusterServer {
+ virStorageNetHostTransport type;
+ union { /* union tag is @type */
+ unixSocketAddress uds;
+ inetSocketAddress tcp;
+ } u;
+};
+
+struct _virStorageBackendGlusterStatePreopened {
+ virStorageFileBackendGlusterPrivPtr priv;
+ unsigned int nservers;
+ glusterServerPtr *hosts;
+ char *volname;
+ volatile int ref;
+};
+
+struct _virStorageBackendGlusterconnCache {
+ virMutex lock;
+ size_t nConn;
+ virStorageBackendGlusterStatePtrPreopened *conn;
+};
+
+virStorageBackendGlusterconnCachePtr connCache = {0,};
+
+
+static void
+virGlusterConnAlloc(void)
+{
+ if ((VIR_ALLOC(connCache) < 0))
+ virGlusterGlobalError = false;
+}
+
+static int
+virStorageBackendGlusterSetPreopened(virStorageSourcePtr src,
+ virStorageFileBackendGlusterPrivPtr priv)
+{
+ unsigned int idx;
+ virStorageBackendGlusterStatePtrPreopened entry = NULL;
+
+ if (connCache == NULL && (virOnce(&glusterConnOnce,
+ virGlusterConnAlloc) < 0))
+ return -1;
+
+ if (virGlusterGlobalError)
+ return -1;
+
+ if (!connCache->nConn) {
+ if (virMutexInit(&connCache->lock) < 0) {
+ virReportError(VIR_ERR_INTERNAL_ERROR, "%s",
+ _("Unable to initialize mutex"));
+ return -1;
+ }
+ }
+
+ for (idx = 0; idx < connCache->nConn; idx++) {
+ if (STREQ(connCache->conn[idx]->volname, src->volume))
+ return 0;
+ }
+
+ if (VIR_ALLOC(entry) < 0)
+ return -1;
+
+ if (VIR_STRDUP(entry->volname, src->volume) < 0)
+ goto error;
+
+ if (VIR_ALLOC_N(entry->hosts, entry->nservers) < 0)
+ goto error;
+
+ entry->nservers = src->nhosts;
+
+ for (idx = 0; idx < src->nhosts; idx++) {
+ if (VIR_ALLOC(entry->hosts[idx]) < 0)
+ goto error;
+ if (src->hosts[idx].transport == VIR_STORAGE_NET_HOST_TRANS_UNIX) {
+ if (VIR_STRDUP(entry->hosts[idx]->u.uds.path,
+ src->hosts[idx].socket) < 0)
+ goto error;
+ entry->hosts[idx]->type = VIR_STORAGE_NET_HOST_TRANS_UNIX;
+ } else if (src->hosts[idx].transport ==
+ VIR_STORAGE_NET_HOST_TRANS_TCP) {
+ if (VIR_STRDUP(entry->hosts[idx]->u.tcp.host,
+ src->hosts[idx].name) < 0)
+ goto error;
+ if (VIR_STRDUP(entry->hosts[idx]->u.tcp.port,
+ src->hosts[idx].port) < 0)
+ goto error;
+ entry->hosts[idx]->type = VIR_STORAGE_NET_HOST_TRANS_TCP;
+ } else {
+ entry->hosts[idx]->type = VIR_STORAGE_NET_HOST_TRANS_LAST;
+ }
+ }
+ entry->priv = priv;
+
+ virMutexLock(&connCache->lock);
+ virAtomicIntSet(&entry->ref, 1);
+ if (VIR_INSERT_ELEMENT(connCache->conn, -1, connCache->nConn, entry) < 0) {
+ virMutexUnlock(&connCache->lock);
+ goto error;
+ }
+ virMutexUnlock(&connCache->lock);
+
+ return 0;
+
+ error:
+ for (idx = 0; idx < entry->nservers; idx++) {
+ if (entry->hosts[idx]->type == VIR_STORAGE_NET_HOST_TRANS_UNIX) {
+ VIR_FREE(entry->hosts[idx]->u.uds.path);
+ } else {
+ VIR_FREE(entry->hosts[idx]->u.tcp.host);
+ VIR_FREE(entry->hosts[idx]->u.tcp.port);
+ }
+ VIR_FREE(entry->hosts[idx]);
+ }
+
+ VIR_FREE(entry->hosts);
+ VIR_FREE(entry->volname);
+ VIR_FREE(entry);
+
+ return -1;
+}
+
+static glfs_t *
+virStorageBackendGlusterFindPreopened(virStorageSourcePtr src)
+{
+ unsigned int cIdx, sIdx, nIdx; /* connectionIdx, savedIdx, newIdx */
+ bool flag = false;
+
+ if (connCache == NULL)
+ return NULL;
+
+ virStorageBackendGlusterStatePtrPreopened entry;
+
+ for (cIdx = 0; cIdx < connCache->nConn; cIdx++) {
+ entry = connCache->conn[cIdx];
+ if (STREQ(entry->volname, src->volume)) {
+ if (entry->nservers == src->nhosts) {
+ for (sIdx = 0; sIdx < entry->nservers; sIdx++) {
+ for (nIdx = 0; nIdx < src->nhosts; nIdx++) {
+ if (entry->hosts[sIdx]->type ==
+ src->hosts[nIdx].transport) {
+ if (entry->hosts[sIdx]->type
+ == VIR_STORAGE_NET_HOST_TRANS_UNIX) {
+ if (STREQ(entry->hosts[sIdx]->u.uds.path,
+ src->hosts[nIdx].socket)) {
+ flag = true;
+ break;
+ }
+ } else {
+ if (STREQ(entry->hosts[sIdx]->u.tcp.host,
+ src->hosts[nIdx].name) &&
+ STREQ(entry->hosts[sIdx]->u.tcp.port,
+ src->hosts[nIdx].port)) {
+ flag = true;
+ break;
+ }
+ }
+ }
+ }
+ if (!flag)
+ return NULL;
+ flag = false;
+ }
+ virAtomicIntInc(&entry->ref);
+ return entry->priv->vol;
+ } else {
+ return NULL;
+ }
+ }
+ }
+ return NULL;
+}
+
+static void
+virStorageBackendGlusterClosePreopened(virStorageSourcePtr src)
+{
+ virStorageFileBackendGlusterPrivPtr priv = src->drv->priv;
+ unsigned int cIdx, hIdx; /* connectionIdx, hostIdx */
+
+ for (cIdx = 0; cIdx < connCache->nConn; cIdx++) {
+ if (STREQ(connCache->conn[cIdx]->volname, src->volume)) {
+ virAtomicIntDecAndTest(&connCache->conn[cIdx]->ref);
+ if (virAtomicIntGet(&connCache->conn[cIdx]->ref) != 0) {
+ if (priv && priv->canonpath) {
+ VIR_FREE(priv->canonpath);
+ VIR_FREE(priv);
+ src->drv->priv = NULL;
+ }
+ return;
+ }
+
+ glfs_fini(connCache->conn[cIdx]->priv->vol);
+
+ VIR_FREE(connCache->conn[cIdx]->priv->canonpath);
+ VIR_FREE(connCache->conn[cIdx]->priv);
+
+ for (hIdx = 0; hIdx < connCache->conn[cIdx]->nservers; hIdx++) {
+ if (connCache->conn[cIdx]->hosts[hIdx]->type ==
+ VIR_STORAGE_NET_HOST_TRANS_UNIX) {
+ VIR_FREE(connCache->conn[cIdx]->hosts[hIdx]->u.uds.path);
+ } else {
+ VIR_FREE(connCache->conn[cIdx]->hosts[hIdx]->u.tcp.host);
+ VIR_FREE(connCache->conn[cIdx]->hosts[hIdx]->u.tcp.port);
+ }
+ VIR_FREE(connCache->conn[cIdx]->hosts[hIdx]);
+ }
+
+ VIR_FREE(connCache->conn[cIdx]->hosts);
+ VIR_FREE(connCache->conn[cIdx]->volname);
+ VIR_FREE(connCache->conn[cIdx]);
+
+ virMutexLock(&connCache->lock);
+ VIR_DELETE_ELEMENT(connCache->conn, cIdx, connCache->nConn);
+ virMutexUnlock(&connCache->lock);
+
+ VIR_FREE(src->drv);
+ }
+ }
+
+ if (!connCache->conn)
+ virMutexDestroy(&connCache->lock);
+}
static void
virStorageBackendGlusterClose(virStorageBackendGlusterStatePtr state)
@@ -538,30 +794,14 @@ virStorageBackend virStorageBackendGluster = {
};
-typedef struct _virStorageFileBackendGlusterPriv virStorageFileBackendGlusterPriv;
-typedef virStorageFileBackendGlusterPriv *virStorageFileBackendGlusterPrivPtr;
-
-struct _virStorageFileBackendGlusterPriv {
- glfs_t *vol;
- char *canonpath;
-};
-
-
static void
virStorageFileBackendGlusterDeinit(virStorageSourcePtr src)
{
- virStorageFileBackendGlusterPrivPtr priv = src->drv->priv;
-
VIR_DEBUG("deinitializing gluster storage file %p (gluster://%s:%s/%s%s)",
src, src->hosts->name, src->hosts->port ? src->hosts->port : "0",
src->volume, src->path);
- if (priv->vol)
- glfs_fini(priv->vol);
- VIR_FREE(priv->canonpath);
-
- VIR_FREE(priv);
- src->drv->priv = NULL;
+ virStorageBackendGlusterClosePreopened(src);
}
static int
@@ -612,6 +852,7 @@ virStorageFileBackendGlusterInitServer(virStorageFileBackendGlusterPrivPtr priv,
static int
virStorageFileBackendGlusterInit(virStorageSourcePtr src)
{
+ int ret = 0;
virStorageFileBackendGlusterPrivPtr priv = NULL;
size_t i;
@@ -625,6 +866,12 @@ virStorageFileBackendGlusterInit(virStorageSourcePtr src)
if (VIR_ALLOC(priv) < 0)
return -1;
+ priv->vol = virStorageBackendGlusterFindPreopened(src);
+ if (priv->vol) {
+ src->drv->priv = priv;
+ return ret;
+ }
+
VIR_DEBUG("initializing gluster storage file %p "
"(priv='%p' volume='%s' path='%s') as [%u:%u]",
src, priv, src->volume, src->path,
@@ -635,6 +882,10 @@ virStorageFileBackendGlusterInit(virStorageSourcePtr src)
goto error;
}
+ ret = virStorageBackendGlusterSetPreopened(src, priv);
+ if (ret < 0)
+ goto error;
+
for (i = 0; i < src->nhosts; i++) {
if (virStorageFileBackendGlusterInitServer(priv, src->hosts + i) < 0)
goto error;
@@ -648,13 +899,13 @@ virStorageFileBackendGlusterInit(virStorageSourcePtr src)
}
src->drv->priv = priv;
+ ret = 0;
- return 0;
+ return ret;
error:
- if (priv->vol)
- glfs_fini(priv->vol);
VIR_FREE(priv);
+ virStorageBackendGlusterClosePreopened(src);
return -1;
}
diff --git a/src/storage/storage_driver.c b/src/storage/storage_driver.c
index df65807..9c3bffb 100644
--- a/src/storage/storage_driver.c
+++ b/src/storage/storage_driver.c
@@ -2900,11 +2900,8 @@ virStorageFileDeinit(virStorageSourcePtr src)
if (!virStorageFileIsInitialized(src))
return;
- if (src->drv->backend &&
- src->drv->backend->backendDeinit)
+ if (src->drv->backend && src->drv->backend->backendDeinit)
src->drv->backend->backendDeinit(src);
-
- VIR_FREE(src->drv);
}
@@ -2967,6 +2964,24 @@ virStorageFileInit(virStorageSourcePtr src)
return virStorageFileInitAs(src, -1, -1);
}
+int
+virStorageVolumeRegister(virStorageSourcePtr src,
+ uid_t uid, gid_t gid)
+{
+ if (virStorageFileInitAs(src, uid, gid) < 0)
+ return -1;
+
+ src->drv->priv = NULL;
+
+ return 0;
+}
+
+void
+virStorageVolumeUnRegister(virStorageSourcePtr src)
+{
+ virStorageFileDeinit(src);
+}
+
/**
* virStorageFileCreate: Creates an empty storage file via storage driver
diff --git a/src/storage/storage_driver.h b/src/storage/storage_driver.h
index 3f2549d..89a3d55 100644
--- a/src/storage/storage_driver.h
+++ b/src/storage/storage_driver.h
@@ -71,4 +71,7 @@ char *virStoragePoolObjBuildTempFilePath(virStoragePoolObjPtr pool,
int storageRegister(void);
+int virStorageVolumeRegister(virStorageSourcePtr src, uid_t uid, gid_t gid);
+void virStorageVolumeUnRegister(virStorageSourcePtr src);
+
#endif /* __VIR_STORAGE_DRIVER_H__ */
--
2.7.4
7 years, 11 months
[libvirt] [PATCH] storage: Fix type PLOOP type check for storageVolUpload
by John Ferlan
Commit id '03e750f3' added support for checking the PLOOP type; however,
it used 'target.type' which no storage code ever fills in, so it will
never be set. Change to just vol->type (could use vol->target.format
as well).
Signed-off-by: John Ferlan <jferlan(a)redhat.com>
---
Found this by chance while digging through the code to investigate
something else. Turns out the storage/volume code does not fill in
the target.type field.
src/storage/storage_driver.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/src/storage/storage_driver.c b/src/storage/storage_driver.c
index df65807..a79acc6 100644
--- a/src/storage/storage_driver.c
+++ b/src/storage/storage_driver.c
@@ -2427,7 +2427,7 @@ storageVolUpload(virStorageVolPtr obj,
if (VIR_ALLOC(cbdata) < 0 ||
VIR_STRDUP(cbdata->pool_name, pool->def->name) < 0)
goto cleanup;
- if (vol->target.type == VIR_STORAGE_VOL_PLOOP &&
+ if (vol->type == VIR_STORAGE_VOL_PLOOP &&
VIR_STRDUP(cbdata->vol_path, vol->target.path) < 0)
goto cleanup;
--
2.7.4
7 years, 11 months
[libvirt] [PATCH] NEWS: Remove end-of-sentence period
by Andrea Bolognani
None of the other sentences in the file ends with the period,
and we want to maintain a consistent style.
---
Pushed as trivial.
docs/news.html.in | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/docs/news.html.in b/docs/news.html.in
index 8a86dd4..27d3387 100644
--- a/docs/news.html.in
+++ b/docs/news.html.in
@@ -67,7 +67,7 @@
<ul>
<li>vz: Fix migration in P2P mode
</li>
- <li>Forbid newline character in names of some libvirt objects.
+ <li>Forbid newline character in names of some libvirt objects
</li>
<li>Fix compilation on macOS
</li>
--
2.7.4
7 years, 11 months
[libvirt] [PATCH 0/2] Post-release version bump to 3.0.0
by Andrea Bolognani
Pushed as "trivial enough that I got it wrong the first time
around" :/
Andrea Bolognani (2):
Post-release version bump to 2.6.0
Fix version number
configure.ac | 2 +-
docs/news.html.in | 10 ++++++++++
2 files changed, 11 insertions(+), 1 deletion(-)
--
2.7.4
7 years, 11 months
[libvirt] [PATCH] NEWS: Add release date for 2.5.0
by Andrea Bolognani
---
Pushed as trivial.
docs/news.html.in | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/docs/news.html.in b/docs/news.html.in
index f7bac14..3ede17c 100644
--- a/docs/news.html.in
+++ b/docs/news.html.in
@@ -12,7 +12,7 @@
<a href="http://libvirt.org/git/?p=libvirt.git;a=log">git log</a>.
</p>
- <h3>v2.5.0 (<i>unreleased</i>)</h3>
+ <h3>v2.5.0 (2016-12-04)</h3>
<ul>
<li><strong>New features</strong>
<ul>
--
2.7.4
7 years, 11 months
