- Devel - libvirt List Archives

[libvirt] [PREPOST] XL parser
by Kiarie Kahurani 27 Aug '14

27 Aug '14

Kiarie Kahurani (1): src/xenconfig: XL parser po/POTFILES.in | 1 + src/Makefile.am | 3 +- src/libvirt_xenconfig.syms | 5 + src/libxl/libxl_driver.c | 44 ++- src/xenconfig/xen_common.c | 140 ++++---- src/xenconfig/xen_common.h | 24 +- src/xenconfig/xen_xl.c | 461 ++++++++++++++++++++++++++ src/xenconfig/xen_xl.h | 29 ++ tests/Makefile.am | 9 +- tests/testutilsxen.c | 50 +++ tests/testutilsxen.h | 9 +- tests/xlconfigdata/test-fullvirt-new-disk.cfg | 27 ++ tests/xlconfigdata/test-fullvirt-new-disk.xml | 46 +++ tests/xlconfigtest.c | 222 +++++++++++++ 14 files changed, 981 insertions(+), 89 deletions(-) create mode 100644 src/xenconfig/xen_xl.c create mode 100644 src/xenconfig/xen_xl.h create mode 100644 tests/xlconfigdata/test-fullvirt-new-disk.cfg create mode 100644 tests/xlconfigdata/test-fullvirt-new-disk.xml create mode 100644 tests/xlconfigtest.c After *toooo* long the xen-xl parser has arrived with support for xl disk format and spice graphics(with admittedly very kludgy code) quick overview 1)coverts xl disk config such as disk = ["/dev/sda,raw,hda,rw,"] into xml config <disk type='block' device='disk'> <driver name='phy'/> <source dev='/dev/sda/'> <target dev= 'hda' bus='ide'/> </disk> and vice-versa You notice that in the xl disk config there is no driver type but it appears on the xml; I use the format to intelligently guess the driver but this does not look very good again you notice I do not support disk backend config which is because there is no member of _virDomainDiskDef which I could relate to the backend 2)converts spice graphics config spice = 1 spicehost = "127.0.0.1" spiceport = 500 spicetls_port = 590 spicedisable_ticketing = 1 #with this set you require a passwd spicepasswd = "thebeast" spiceagent_mouse = 1 to <graphics type='spice' port='500' tlsPort='590' autoport='no' listen='127.0.0.1'> <listen type='address' address='127.0.0.1'/> <mouse mode='server'/> and vice-versa You notice that the passwd is not in the xml, am still trying to figure out why Again you will notice the current tests are a bit dependent on xen code(when initializing the xmlOpt) which would *better* be done away with but that would require some code in src/libxl/ Into the bargain, am not *sure* how exactly to initialize _virCapsPtr which is reqiured for xml formatting because I *assume* that tests should be able to run even with drivers not loaded. Am also not *sure* that how to map disks to device type, the existing code(xm) code maps all the disks using another driver other than 'phy' to 'block'.Am not sure this is correct. With the current patch $virsh domxml-from/to-native xen-xl works perfectly TODO Fix tests -- 1.8.4.5

2 3

[libvirt] [PATCH 0/9] qemu: Introduce support for new the block_set_io_throttle parameters add in the version 1.7 of qemu.
by Matthias Gatto 27 Aug '14

27 Aug '14

This series of patches add suport for bps_max, bps_rd_max, bps_wr_max, bps_max, bps_rd_max, bps_wr_max, and iops_size in the functions qemuDomainSetBlockIoTune, and qemuDomainGetBlockIoTune. The last patch add suport for therse parameters to the virsh blkdeviotune commande. Matthias Gatto (9): qemu: Add defines for the new throttle options qemu: Add new throttle options to the structure _virDomainBlockIoTuneInfo. qemu: Add the capabilitie to detect if the qemu binary have the capability to use bps_max and friends qemu: Add bps_max and friends qemu driver qemu: Add bps_max and friend to domain_conf qemu: Add bps_max and friends QMP suport qemu: Add bps_max and friends "text" suport qemu: add bps_max and friends to qemu command genaration qemu: Add bps_max and friends to virsh include/libvirt/libvirt.h.in | 54 +++++++++++++++ src/conf/domain_conf.c | 89 +++++++++++++++++++++++- src/conf/domain_conf.h | 8 +++ src/qemu/qemu_capabilities.c | 6 ++ src/qemu/qemu_capabilities.h | 2 +- src/qemu/qemu_command.c | 49 ++++++++++++++ src/qemu/qemu_driver.c | 158 +++++++++++++++++++++++++++++++++++++++++-- src/qemu/qemu_monitor_json.c | 58 +++++++++++++--- src/qemu/qemu_monitor_text.c | 75 ++++++++++++++++++-- tests/qemumonitorjsontest.c | 2 +- tools/virsh-domain.c | 119 ++++++++++++++++++++++++++++++++ 11 files changed, 595 insertions(+), 25 deletions(-) -- 1.8.3.1

1 9

[libvirt] [RFC] VIR_ERROR in detect_scsi_host_caps into warnings
by Cedric Bosdonnat 27 Aug '14

27 Aug '14

Hi all, detect_scsi_host_caps function is called without caring of the result in most cases. Wouldn't it be more logical to have VIR_ERRORs in this function transformed into VIR_WARN as those problems aren't blocking libvirt? Note: I can do the (trivial) patch if the idea is ACKed. -- Cedric

1 0

[libvirt] [PATCH] virsh: fix keepalive error msg, man page update
by Erik Skultety 27 Aug '14

27 Aug '14

resolves https://bugzilla.redhat.com/show_bug.cgi?id=1132305 --- tools/virsh.c | 14 ++++++++++---- tools/virsh.pod | 6 ++++-- 2 files changed, 14 insertions(+), 6 deletions(-) diff --git a/tools/virsh.c b/tools/virsh.c index 30a84c1..f9b3991 100644 --- a/tools/virsh.c +++ b/tools/virsh.c @@ -3472,8 +3472,11 @@ vshParseArgv(vshControl *ctl, int argc, char **argv) case 'k': if (virStrToLong_i(optarg, NULL, 0, &keepalive) < 0 || keepalive < 0) { - vshError(ctl, _("option %s requires a positive numeric argument"), - longindex == -1 ? "-k" : "--keepalive-interval"); + vshError(ctl, + _("option %s requires a positive integer argument " + "within range <0,%d>"), + longindex == -1 ? "-k" : "--keepalive-interval", + INT_MAX); exit(EXIT_FAILURE); } ctl->keepalive_interval = keepalive; @@ -3481,8 +3484,11 @@ vshParseArgv(vshControl *ctl, int argc, char **argv) case 'K': if (virStrToLong_i(optarg, NULL, 0, &keepalive) < 0 || keepalive < 0) { - vshError(ctl, _("option %s requires a positive numeric argument"), - longindex == -1 ? "-K" : "--keepalive-count"); + vshError(ctl, + _("option %s requires a positive integer argument " + "within range <0,%d>"), + longindex == -1 ? "-K" : "--keepalive-count", + INT_MAX); exit(EXIT_FAILURE); } ctl->keepalive_count = keepalive; diff --git a/tools/virsh.pod b/tools/virsh.pod index e0dfd13..9bee664 100644 --- a/tools/virsh.pod +++ b/tools/virsh.pod @@ -77,13 +77,15 @@ given instead. =item B<-k>, B<--keepalive-interval> I<INTERVAL> Set an I<INTERVAL> (in seconds) for sending keepalive messages to -check whether connection to the server is still alive. Setting the +check whether connection to the server is still alive. I<INTERVAL> +must be an integer value within range <0,INT_MAX>. Setting the interval to 0 disables client keepalive mechanism. =item B<-K>, B<--keepalive-count> I<COUNT> Set a number of times keepalive message can be sent without getting an -answer from the server without marking the connection dead. There is +answer from the server without marking the connection dead. I<COUNT> +must be an integer value within range <0,INT_MAX>. There is no effect to this setting in case the I<INTERVAL> is set to 0. =item B<-l>, B<--log> I<FILE> -- 1.9.3

3 2

[libvirt] [PATCH for 1.2.8] conf: fix leak by freeing def->mem.hugepages
by Martin Kletzander 27 Aug '14

27 Aug '14

Signed-off-by: Martin Kletzander <mkletzan(a)redhat.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 3b295ab..d6fd1ab 100644 --- a/src/conf/domain_conf.c +++ b/src/conf/domain_conf.c @@ -2136,6 +2136,8 @@ void virDomainDefFree(virDomainDefPtr def) virDomainMemballoonDefFree(def->memballoon); virDomainNVRAMDefFree(def->nvram); + VIR_FREE(def->mem.hugepages); + for (i = 0; i < def->nseclabels; i++) virSecurityLabelDefFree(def->seclabels[i]); VIR_FREE(def->seclabels); -- 2.1.0

2 2

[libvirt] [PATCH] API: Tweak virDomainOpenGraphics to return fd directly
by Eric Blake 27 Aug '14

27 Aug '14

Let's fix this before we bake in a painful API. Since we know that we have exactly one non-negative fd on success, we might as well return the fd directly instead of forcing the user to pass in a pointer. Fix a memory leak I found while reviewing. * include/libvirt/libvirt.h.in (virDomainOpenGraphicsFD): Drop unneeded parameter. * src/driver.h (virDrvDomainOpenGraphicsFD): Likewise. * src/libvirt.c (virDomainOpenGraphicsFD): Adjust interface to return fd directly. * daemon/remote.c (remoteDispatchDomainOpenGraphicsFd): Adjust semantics. * src/qemu/qemu_driver.c (qemuDomainOpenGraphicsFD): Likewise. * src/remote/remote_driver.c (remoteDomainOpenGraphicsFD): Likewise, and plug memory leak. Signed-off-by: Eric Blake <eblake(a)redhat.com> --- daemon/remote.c | 9 +++++---- include/libvirt/libvirt.h.in | 1 - src/driver.h | 1 - src/libvirt.c | 14 ++++++-------- src/qemu/qemu_driver.c | 3 +-- src/remote/remote_driver.c | 17 +++++++++++------ 6 files changed, 23 insertions(+), 22 deletions(-) diff --git a/daemon/remote.c b/daemon/remote.c index 27d1aa8..940dcfa 100644 --- a/daemon/remote.c +++ b/daemon/remote.c @@ -4419,10 +4419,9 @@ remoteDispatchDomainOpenGraphicsFd(virNetServerPtr server ATTRIBUTE_UNUSED, if (!(dom = get_nonnull_domain(priv->conn, args->dom))) goto cleanup; - if (virDomainOpenGraphicsFD(dom, - args->idx, - &fd, - args->flags) < 0) + if ((fd = virDomainOpenGraphicsFD(dom, + args->idx, + args->flags)) < 0) goto cleanup; if (virNetMessageAddFD(msg, fd) < 0) @@ -4442,6 +4441,8 @@ remoteDispatchDomainOpenGraphicsFd(virNetServerPtr server ATTRIBUTE_UNUSED, virDomainFree(dom); return rv; } + + static int remoteDispatchDomainGetInterfaceParameters(virNetServerPtr server ATTRIBUTE_UNUSED, virNetServerClientPtr client ATTRIBUTE_UNUSED, diff --git a/include/libvirt/libvirt.h.in b/include/libvirt/libvirt.h.in index 2a108b8..e79c9ad 100644 --- a/include/libvirt/libvirt.h.in +++ b/include/libvirt/libvirt.h.in @@ -5401,7 +5401,6 @@ int virDomainOpenGraphics(virDomainPtr dom, int virDomainOpenGraphicsFD(virDomainPtr dom, unsigned int idx, - int *fd, unsigned int flags); int virDomainInjectNMI(virDomainPtr domain, unsigned int flags); diff --git a/src/driver.h b/src/driver.h index cd136ba..a7366e4 100644 --- a/src/driver.h +++ b/src/driver.h @@ -890,7 +890,6 @@ typedef int typedef int (*virDrvDomainOpenGraphicsFD)(virDomainPtr dom, unsigned int idx, - int *fd, unsigned int flags); typedef int diff --git a/src/libvirt.c b/src/libvirt.c index 772cc0d..7a3f01a 100644 --- a/src/libvirt.c +++ b/src/libvirt.c @@ -20305,11 +20305,10 @@ virDomainOpenGraphics(virDomainPtr dom, * virDomainOpenGraphicsFD: * @dom: pointer to domain object * @idx: index of graphics config to open - * @fd: returned file descriptor * @flags: bitwise-OR of virDomainOpenGraphicsFlags * * This will create a socket pair connected to the graphics backend of @dom. - * One socket will be returned as @fd. + * One end of the socket will be returned on success. * If @dom has multiple graphics backends configured, then @idx will determine * which one is opened, starting from @idx 0. * @@ -20320,21 +20319,18 @@ virDomainOpenGraphics(virDomainPtr dom, * libvirt hypervisor, over a UNIX domain socket. Attempts * to use this method over a TCP connection will always fail * - * Returns 0 on success, -1 on failure + * Returns an fd on success, -1 on failure */ int virDomainOpenGraphicsFD(virDomainPtr dom, unsigned int idx, - int *fd, unsigned int flags) { - VIR_DOMAIN_DEBUG(dom, "idx=%u, fd=%p, flags=%x", - idx, fd, flags); + VIR_DOMAIN_DEBUG(dom, "idx=%u, flags=%x", idx, flags); virResetLastError(); virCheckDomainReturn(dom, -1); - virCheckNonNullArgGoto(fd, error); virCheckReadOnlyGoto(dom->conn->flags, error); @@ -20347,7 +20343,7 @@ virDomainOpenGraphicsFD(virDomainPtr dom, if (dom->conn->driver->domainOpenGraphicsFD) { int ret; - ret = dom->conn->driver->domainOpenGraphicsFD(dom, idx, fd, flags); + ret = dom->conn->driver->domainOpenGraphicsFD(dom, idx, flags); if (ret < 0) goto error; return ret; @@ -20359,6 +20355,8 @@ virDomainOpenGraphicsFD(virDomainPtr dom, virDispatchError(dom->conn); return -1; } + + /** * virConnectSetKeepAlive: * @conn: pointer to a hypervisor connection diff --git a/src/qemu/qemu_driver.c b/src/qemu/qemu_driver.c index 5ff2059..57c999c 100644 --- a/src/qemu/qemu_driver.c +++ b/src/qemu/qemu_driver.c @@ -15808,7 +15808,6 @@ qemuDomainOpenGraphics(virDomainPtr dom, static int qemuDomainOpenGraphicsFD(virDomainPtr dom, unsigned int idx, - int *fd, unsigned int flags) { virQEMUDriverPtr driver = dom->conn->privateData; @@ -15871,7 +15870,7 @@ qemuDomainOpenGraphicsFD(virDomainPtr dom, if (!qemuDomainObjEndJob(driver, vm)) vm = NULL; - *fd = pair[0]; + ret = pair[0]; cleanup: if (ret < 0) { diff --git a/src/remote/remote_driver.c b/src/remote/remote_driver.c index e01216a..e949223 100644 --- a/src/remote/remote_driver.c +++ b/src/remote/remote_driver.c @@ -6448,7 +6448,6 @@ remoteDomainOpenGraphics(virDomainPtr dom, static int remoteDomainOpenGraphicsFD(virDomainPtr dom, unsigned int idx, - int *fd, unsigned int flags) { int rv = -1; @@ -6472,15 +6471,21 @@ remoteDomainOpenGraphicsFD(virDomainPtr dom, goto done; if (fdoutlen != 1) { - virReportError(VIR_ERR_INTERNAL_ERROR, "%s", - _("no file descriptor received")); + if (fdoutlen) { + virReportError(VIR_ERR_INTERNAL_ERROR, "%s", + _("too many file descriptors received")); + while (fdoutlen) + VIR_FORCE_CLOSE(fdout[--fdoutlen]); + } else { + virReportError(VIR_ERR_INTERNAL_ERROR, "%s", + _("no file descriptor received")); + } goto done; } - *fd = fdout[0]; - - rv = 0; + rv = fdout[0]; done: + VIR_FREE(fdout); remoteDriverUnlock(priv); return rv; -- 1.9.3

3 3

[libvirt] [PATCHv2 00/16] LXC network configuration support
by Cédric Bosdonnat 27 Aug '14

27 Aug '14

Hi all, Here is the whole series resent with a major addition: the functions used to set the IP and add a route now use libnl when possible. The idea behind this is to avoid requiring iproute2 or ifconfig installed in the container rootfs. Otherwise nothing changed since v1. Cédric Bosdonnat (16): Forgot to cleanup ifname_guest* in domain network def parsing Domain conf: allow more than one IP address for net devices LXC: set IP addresses to veth devices in the container lxc conf2xml: convert IP addresses Allow network capabilities hostdev to configure IP addresses lxc conf2xml: convert ip addresses for hostdev NICs Domain network devices can now have a <gateway> element lxc conf2xml: convert lxc.network.ipv[46].gateway LXC: use the new net devices gateway definition LXC: honour network devices link state Wrong place for virDomainNetIpsFormat virNetDevSetIPv4Address: libnl implementation Renamed virNetDevSetIPv4Address to virNetDevSetIPAddress virNetDevAddRoute: implementation using netlink virNetDevClearIPv4Address: netlink implementation Renamed virNetDevClearIPv4Address to virNetDevClearIPAddress docs/formatdomain.html.in | 39 +++ docs/schemas/domaincommon.rng | 55 +++- src/conf/domain_conf.c | 214 +++++++++++++-- src/conf/domain_conf.h | 22 +- src/libvirt_private.syms | 7 +- src/lxc/lxc_container.c | 74 ++++- src/lxc/lxc_native.c | 173 ++++++++---- src/network/bridge_driver.c | 4 +- src/openvz/openvz_conf.c | 2 +- src/openvz/openvz_driver.c | 6 +- src/qemu/qemu_driver.c | 25 +- src/qemu/qemu_hotplug.c | 6 +- src/uml/uml_conf.c | 2 +- src/util/virnetdev.c | 305 ++++++++++++++++++--- src/util/virnetdev.h | 12 +- src/util/virnetlink.c | 38 +++ src/util/virnetlink.h | 2 + src/vbox/vbox_common.c | 3 +- src/xenconfig/xen_common.c | 15 +- src/xenconfig/xen_sxpr.c | 12 +- .../lxcconf2xmldata/lxcconf2xml-physnetwork.config | 4 + tests/lxcconf2xmldata/lxcconf2xml-physnetwork.xml | 3 + tests/lxcconf2xmldata/lxcconf2xml-simple.config | 4 + tests/lxcconf2xmldata/lxcconf2xml-simple.xml | 3 + tests/lxcxml2xmldata/lxc-hostdev.xml | 3 + tests/lxcxml2xmldata/lxc-idmap.xml | 3 + 26 files changed, 880 insertions(+), 156 deletions(-) -- 1.8.4.5

4 19

[libvirt] [PATCH 0/3] Fix VirtualBox registration
by Michal Privoznik 27 Aug '14

27 Aug '14

This is supposed to fix bug report which appeared on the list [1]. The driver registration ordering is back as it was, but the subdrivers are registered before. Unfortunately, if we don't want to dig vbox network and storage code out, we have to compile the sources couple of times. Michal Privoznik (3): virdrivermoduletest: Test all the modules virDriverLoadModule: Honor libvirt func name tranlsation vbox: Register per partes daemon/libvirtd.c | 16 +++++++++++++-- src/Makefile.am | 41 ++++++++++++++++++++++++++++++++++--- src/driver.c | 19 +++++++++++++++-- src/vbox/vbox_driver.c | 50 ++++++++++++++++++++++++++++++++++++++++----- src/vbox/vbox_driver.h | 10 +++++++++ tests/virdrivermoduletest.c | 18 ++++++++++------ 6 files changed, 136 insertions(+), 18 deletions(-) -- 1.8.5.5

2 5

[libvirt] [PATCH] blockcopy: allow larger buf-size
by Eric Blake 27 Aug '14

27 Aug '14

While qemu definitely caps granularity to 64 MiB, it places no limits on buf-size. On a machine beefy enough for lots of memory, a buf-size larger than 2 GiB is feasible, so we should pass a 64-bit parameter. * include/libvirt/libvirt.h.in (VIR_DOMAIN_BLOCK_COPY_BUF_SIZE): Allow 64 bits. Signed-off-by: Eric Blake <eblake(a)redhat.com> --- Based on Peter's review of my patch series; until we have an implementation, it is just a doc bug, but I'd rather document it correctly and avoid baking in a too-small API. include/libvirt/libvirt.h.in | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/include/libvirt/libvirt.h.in b/include/libvirt/libvirt.h.in index e79c9ad..e80634e 100644 --- a/include/libvirt/libvirt.h.in +++ b/include/libvirt/libvirt.h.in @@ -2663,8 +2663,8 @@ typedef enum { * VIR_DOMAIN_BLOCK_COPY_BUF_SIZE: * Macro for the virDomainBlockCopy buffer size tunable: it represents * how much data in bytes can be in flight between source and destination, - * as an unsigned int. Specifying 0 is the same as omitting this parameter, - * to request the hypervisor default. + * as an unsigned long long. Specifying 0 is the same as omitting this + * parameter, to request the hypervisor default. */ #define VIR_DOMAIN_BLOCK_COPY_BUF_SIZE "buf-size" -- 1.9.3

1 0

[libvirt] [PATCH v2 0/8] new virDomainBlockCopy API
by Eric Blake 27 Aug '14

27 Aug '14

v2: finish the series for local files, address Peter's doc suggestions on patch 1, and make several tweaks to many of the patches to fix issues that I found in testing the final product. Not included - patches to actually target a gluster destination. But those can be delayed until 1.2.9. Eric Blake (8): blockcopy: virDomainBlockCopy with XML destination, typed params blockcopy: split out virsh implementation blockcopy: expose new API in virsh blockcopy: remote implementation for new API blockcopy: tweak how rebase calls into copy blockcopy: add a way to parse disk source blockcopy: add qemu implementation of new API blockcopy: add qemu implementation of new tunables include/libvirt/libvirt.h.in | 61 +++++++++++++-- src/conf/domain_conf.c | 147 +++++++++++++++++++++++------------ src/conf/domain_conf.h | 4 + src/driver.h | 11 ++- src/libvirt.c | 120 ++++++++++++++++++++++++++++- src/libvirt_private.syms | 1 + src/libvirt_public.syms | 5 ++ src/qemu/qemu_driver.c | 177 ++++++++++++++++++++++++++++++++----------- src/qemu/qemu_migration.c | 2 +- src/qemu/qemu_monitor.c | 8 +- src/qemu/qemu_monitor.h | 2 + src/qemu/qemu_monitor_json.c | 3 + src/qemu/qemu_monitor_json.h | 2 + src/remote/remote_driver.c | 1 + src/remote/remote_protocol.x | 18 ++++- src/remote_protocol-structs | 11 +++ tests/qemumonitorjsontest.c | 2 +- tools/virsh-domain.c | 153 +++++++++++++++++++++++++++++++------ tools/virsh.pod | 35 +++++---- 19 files changed, 623 insertions(+), 140 deletions(-) -- 1.9.3

2 23