August 2014 - Devel - libvirt List Archives

[libvirt] [PATCH] Fix missing attribute dscp in supported protocols in nwfilter docs
by Jianwei Hu 19 Aug '14

19 Aug '14

Added attribute dscp to below supported protocols table in nwfilter docs. IPV4 (ip) TCP/UDP/SCTP (tcp/udp/sctp) ICMP (icmp) IGMP,ESP,AH,UDPLITE,'ALL' (igmp,esp,ah,udplite,all) TCP/UDP/SCTP over IPV6 (tcp-ipv6,udp-ipv6,sctp-ipv6) ICMPV6 (icmpv6) IGMP,ESP,AH,UDPLITE,'ALL' over IPv6 (igmp-ipv6,esp-ipv6,ah-ipv6,udplite-ipv6,all-ipv6) Here is a simple example: [root@localhost ~]# virsh nwfilter-dumpxml myself <filter name='myself' chain='root'> <uuid>7192ef51-cd50-4f14-ad7b-fa5c69ea19e3</uuid> <rule action='accept' direction='in' priority='500'> <ip dscp='1'/> </rule> <rule action='accept' direction='in' priority='500'> <tcp dscp='1'/> </rule> <rule action='accept' direction='in' priority='500'> <tcp-ipv6 dscp='2'/> </rule> <rule action='accept' direction='in' priority='500'> <icmp dscp='55'/> </rule> <rule action='accept' direction='in' priority='500'> <icmpv6 dscp='55'/> </rule> <rule action='accept' direction='in' priority='500'> <udp dscp='3'/> </rule> <rule action='accept' direction='in' priority='500'> <udp-ipv6 dscp='4'/> </rule> <rule action='accept' direction='in' priority='500'> <udplite dscp='5'/> </rule> <rule action='accept' direction='in' priority='500'> <udplite-ipv6 dscp='6'/> </rule> <rule action='accept' direction='in' priority='500'> <esp dscp='7'/> </rule> <rule action='accept' direction='in' priority='500'> <esp-ipv6 dscp='8'/> </rule> <rule action='accept' direction='in' priority='500'> <ah dscp='9'/> </rule> <rule action='accept' direction='in' priority='500'> <ah-ipv6 dscp='10'/> </rule> <rule action='accept' direction='in' priority='500'> <sctp dscp='11'/> </rule> <rule action='accept' direction='in' priority='500'> <sctp-ipv6 dscp='55'/> </rule> <rule action='accept' direction='in' priority='500'> <igmp dscp='55'/> </rule> <rule action='accept' direction='in' priority='500'> <all-ipv6 dscp='55'/> </rule> <rule action='accept' direction='in' priority='500'> <all dscp='55'/> </rule> </filter> --- docs/formatnwfilter.html.in | 35 +++++++++++++++++++++++++++++++++++ 1 file changed, 35 insertions(+) diff --git a/docs/formatnwfilter.html.in b/docs/formatnwfilter.html.in index 7169fa9..46b0bfa 100644 --- a/docs/formatnwfilter.html.in +++ b/docs/formatnwfilter.html.in @@ -1101,6 +1101,11 @@ <td>End of range of valid destination ports; requires <code>protocol</code></td> </tr> <tr> + <td>dscp</td> + <td>UINT8 (0x0-0x3f, 0 - 63)</td> + <td>codepoint within the DS field in the IP header</td> + </tr> + <tr> <td>comment (Since 0.8.5)</td> <td>STRING</td> <td>text with max. 256 characters</td> @@ -1284,6 +1289,11 @@ <td>End of range of valid destination ports</td> </tr> <tr> + <td>dscp</td> + <td>UINT8 (0x0-0x3f, 0 - 63)</td> + <td>codepoint within the DS field in the IP header</td> + </tr> + <tr> <td>comment (Since 0.8.5)</td> <td>STRING</td> <td>text with max. 256 characters</td> @@ -1399,6 +1409,11 @@ <td>ICMP code</td> </tr> <tr> + <td>dscp</td> + <td>UINT8 (0x0-0x3f, 0 - 63)</td> + <td>codepoint within the DS field in the IP header</td> + </tr> + <tr> <td>comment (Since 0.8.5)</td> <td>STRING</td> <td>text with max. 256 characters</td> @@ -1498,6 +1513,11 @@ <td>End of range of destination IP address</td> </tr> <tr> + <td>dscp</td> + <td>UINT8 (0x0-0x3f, 0 - 63)</td> + <td>codepoint within the DS field in the IP header</td> + </tr> + <tr> <td>comment (Since 0.8.5)</td> <td>STRING</td> <td>text with max. 256 characters</td> @@ -1604,6 +1624,11 @@ <td>End of range of valid destination ports</td> </tr> <tr> + <td>dscp</td> + <td>UINT8 (0x0-0x3f, 0 - 63)</td> + <td>codepoint within the DS field in the IP header</td> + </tr> + <tr> <td>comment (Since 0.8.5)</td> <td>STRING</td> <td>text with max. 256 characters</td> @@ -1705,6 +1730,11 @@ <td>ICMPv6 code</td> </tr> <tr> + <td>dscp</td> + <td>UINT8 (0x0-0x3f, 0 - 63)</td> + <td>codepoint within the DS field in the IP header</td> + </tr> + <tr> <td>comment (Since 0.8.5)</td> <td>STRING</td> <td>text with max. 256 characters</td> @@ -1789,6 +1819,11 @@ <td>End of range of destination IP address</td> </tr> <tr> + <td>dscp</td> + <td>UINT8 (0x0-0x3f, 0 - 63)</td> + <td>codepoint within the DS field in the IP header</td> + </tr> + <tr> <td>comment (Since 0.8.5)</td> <td>STRING</td> <td>text with max. 256 characters</td> -- 1.8.3.1

2 1

[libvirt] [PATCH] cmdMigrate: move vshConnect before vshWatchJob
by Chunyan Liu 19 Aug '14

19 Aug '14

A possible fix to issue: http://www.redhat.com/archives/libvir-list/2014-August/thread.html#00227 While doing migration on KVM host, found problem sometimes: VM is already running on the target host and disappears from source host, but 'virsh migrate' command line hangs, cannot exit normally. If pressing "ENTER" key, it will exit. The code hangs at tools/virsh-domain.c: cmdMigrate ->vshWatchJob->poll(): poll() is trying to select pipe_fd, which is used to receive message from doMigrate thread. In debugging, found that doMigrate finishes and at the end it does call safewrite() to write the retval ('0' or '1') to pipe_fd, and the write is completed. But cmdMigrate poll() cannot get the event. If pressing "ENTER" key, poll() can get the event and select pipe_fd, then command line can exit. In current code, authentication thread which is called by vshConnect will use stdin, and at the same time, in cmdMigrate main process, poll() is listening to stdin, that probably affect poll() to get pipe_fd event. Better to move authentication before vshWatchJob. With this change, above problem does not exist. Signed-off-by: Chunyan Liu <cyliu(a)suse.com> --- tools/virsh-domain.c | 26 ++++++++++++++++++++------ tools/virsh.h | 1 + 2 files changed, 21 insertions(+), 6 deletions(-) diff --git a/tools/virsh-domain.c b/tools/virsh-domain.c index f7193cb..2562326 100644 --- a/tools/virsh-domain.c +++ b/tools/virsh-domain.c @@ -8965,6 +8965,7 @@ doMigrate(void *opaque) virTypedParameterPtr params = NULL; int nparams = 0; int maxparams = 0; + virConnectPtr dconn = data->dconn; sigemptyset(&sigmask); sigaddset(&sigmask, SIGINT); @@ -9079,18 +9080,12 @@ doMigrate(void *opaque) ret = '0'; } else { /* For traditional live migration, connect to the destination host directly. */ - virConnectPtr dconn = NULL; virDomainPtr ddom = NULL; - dconn = vshConnect(ctl, desturi, false); - if (!dconn) - goto out; - if ((ddom = virDomainMigrate3(dom, dconn, params, nparams, flags))) { virDomainFree(ddom); ret = '0'; } - virConnectClose(dconn); } out: @@ -9152,6 +9147,23 @@ cmdMigrate(vshControl *ctl, const vshCmd *cmd) data.cmd = cmd; data.writefd = p[1]; + if (vshCommandOptBool(cmd, "p2p") || vshCommandOptBool(cmd, "direct")) { + data.dconn = NULL; + } else { + /* For traditional live migration, connect to the destination host. */ + virConnectPtr dconn = NULL; + const char *desturi = NULL; + + if (vshCommandOptStringReq(ctl, cmd, "desturi", &desturi) < 0) + goto cleanup; + + dconn = vshConnect(ctl, desturi, false); + if (!dconn) + goto cleanup; + + data.dconn = dconn; + } + if (virThreadCreate(&workerThread, true, doMigrate, @@ -9163,6 +9175,8 @@ cmdMigrate(vshControl *ctl, const vshCmd *cmd) virThreadJoin(&workerThread); cleanup: + if (data.dconn) + virConnectClose(data.dconn); virDomainFree(dom); VIR_FORCE_CLOSE(p[0]); VIR_FORCE_CLOSE(p[1]); diff --git a/tools/virsh.h b/tools/virsh.h index 7656407..b4df24b 100644 --- a/tools/virsh.h +++ b/tools/virsh.h @@ -371,6 +371,7 @@ struct _vshCtrlData { vshControl *ctl; const vshCmd *cmd; int writefd; + virConnectPtr dconn; }; /* error handling */ -- 1.8.4.5

4 4

[libvirt] Virtual box storage and network sub-drivers
by Michal Privoznik 19 Aug '14

19 Aug '14

Dear list, Virtualbox driver has its own implementation of storage and network sub-drivers. But as of commit ba5f3c7c8ecc1037e44904916989a1c65777a9d5 (contained in the 1.0.6 release) when the VBox moved from client to daemon, the storage and network sub-drivers are indeed registered but in fact never called. It's due to our virConnectOpen function where the general network and storage drivers take precedence. So I guess my question is: should we drop the VBox sub-drivers or perhaps fix the virConnectOpen function? Reported-by: Taowei Luo Michal

3 3

[libvirt] [PATCH 0/5] rework layout of src/xenxs directory
by Jim Fehlig 19 Aug '14

19 Aug '14

This series reworks the layout of src/xenxs directory, changing it to src/xenconfig and moving common XM and XL config parsing/formatting functions to src/xenconfig/xen_common.[ch]. Jim Fehlig (3): xen: rename xenxs to xenconfig src/xenconfig: remove dead code src/xenconfig: move common parsing/formatting to xen_common Kiarie Kahurani (2): src/xenconfig: wrap common parsing code src/xenconfig: wrap common formatting code cfg.mk | 2 +- configure.ac | 2 +- po/POTFILES.in | 5 +- src/Makefile.am | 39 +- src/{libvirt_xenxs.syms => libvirt_xenconfig.syms} | 4 +- src/{xenxs/xen_xm.c => xenconfig/xen_common.c} | 874 +++++++-------------- src/xenconfig/xen_common.h | 46 ++ src/{xenxs => xenconfig}/xen_sxpr.c | 2 +- src/{xenxs => xenconfig}/xen_sxpr.h | 0 src/xenconfig/xen_xm.c | 421 ++++++++++ src/{xenxs => xenconfig}/xen_xm.h | 0 src/{xenxs => xenconfig}/xenxs_private.h | 0 tests/sexpr2xmltest.c | 2 +- tests/xmconfigdata/test-escape-paths.cfg | 2 +- tests/xmconfigdata/test-fullvirt-force-hpet.cfg | 2 +- tests/xmconfigdata/test-fullvirt-force-nohpet.cfg | 2 +- tests/xmconfigdata/test-fullvirt-localtime.cfg | 2 +- tests/xmconfigdata/test-fullvirt-net-ioemu.cfg | 2 +- tests/xmconfigdata/test-fullvirt-net-netfront.cfg | 2 +- tests/xmconfigdata/test-fullvirt-new-cdrom.cfg | 2 +- tests/xmconfigdata/test-fullvirt-old-cdrom.cfg | 2 +- tests/xmconfigdata/test-fullvirt-parallel-tcp.cfg | 2 +- .../test-fullvirt-serial-dev-2-ports.cfg | 2 +- .../test-fullvirt-serial-dev-2nd-port.cfg | 2 +- tests/xmconfigdata/test-fullvirt-serial-file.cfg | 2 +- tests/xmconfigdata/test-fullvirt-serial-null.cfg | 2 +- tests/xmconfigdata/test-fullvirt-serial-pipe.cfg | 2 +- tests/xmconfigdata/test-fullvirt-serial-pty.cfg | 2 +- tests/xmconfigdata/test-fullvirt-serial-stdio.cfg | 2 +- .../test-fullvirt-serial-tcp-telnet.cfg | 2 +- tests/xmconfigdata/test-fullvirt-serial-tcp.cfg | 2 +- tests/xmconfigdata/test-fullvirt-serial-udp.cfg | 2 +- tests/xmconfigdata/test-fullvirt-serial-unix.cfg | 2 +- tests/xmconfigdata/test-fullvirt-sound.cfg | 2 +- tests/xmconfigdata/test-fullvirt-usbmouse.cfg | 2 +- tests/xmconfigdata/test-fullvirt-usbtablet.cfg | 2 +- tests/xmconfigdata/test-fullvirt-utc.cfg | 2 +- tests/xmconfigdata/test-no-source-cdrom.cfg | 2 +- tests/xmconfigdata/test-paravirt-net-e1000.cfg | 2 +- tests/xmconfigdata/test-paravirt-net-vifname.cfg | 2 +- .../test-paravirt-new-pvfb-vncdisplay.cfg | 2 +- tests/xmconfigdata/test-paravirt-new-pvfb.cfg | 2 +- .../test-paravirt-old-pvfb-vncdisplay.cfg | 2 +- tests/xmconfigdata/test-paravirt-old-pvfb.cfg | 2 +- tests/xmconfigdata/test-paravirt-vcpu.cfg | 2 +- tests/xmconfigdata/test-pci-devs.cfg | 2 +- tests/xmconfigtest.c | 2 +- tests/xml2sexprtest.c | 2 +- 48 files changed, 796 insertions(+), 671 deletions(-) rename src/{libvirt_xenxs.syms => libvirt_xenconfig.syms} (91%) rename src/{xenxs/xen_xm.c => xenconfig/xen_common.c} (72%) create mode 100644 src/xenconfig/xen_common.h rename src/{xenxs => xenconfig}/xen_sxpr.c (99%) rename src/{xenxs => xenconfig}/xen_sxpr.h (100%) create mode 100644 src/xenconfig/xen_xm.c rename src/{xenxs => xenconfig}/xen_xm.h (100%) rename src/{xenxs => xenconfig}/xenxs_private.h (100%) -- 1.8.4.5

2 12

[libvirt] sVirt adequate protection when guest hugepages option enabled?
by bancfc＠openmailbox.org 19 Aug '14

19 Aug '14

Hi I've been looking at enabling hugepages for guests and came across the note that SELinux was not able to handle their confinement properly back then: https://fedoraproject.org/wiki/Features/KVM_Huge_Page_Backed_Memory#Current… Was this ever resolved? Does sVirt provide adequate protection when this option is used for guests? Thanks.

1 0

[libvirt] [PATCH 0/5] Fix vcpu pinning info propagation on vcpu cold-unplug
by Peter Krempa 18 Aug '14

18 Aug '14

Remove the vcpupin info on cold cpuunplug. Also refactor a few unpleasant-to-look-at functions found while looking for the problem. Peter Krempa (5): qemu: process: Remove unnecessary argument and rename function conf: cpupin: Remove useless checking of vcpupin element count conf: Refactor virDomainVcpuPinDefParseXML qemu: cpu: unplug: Remove vcpu pinning on cold cpu unplug qemu: process: Pin on per-vcpu basis instead of per-vcpupin element src/conf/domain_conf.c | 64 +++++++++++++++++++++---------------------------- src/qemu/qemu_driver.c | 7 ++++++ src/qemu/qemu_process.c | 20 +++++++++------- 3 files changed, 46 insertions(+), 45 deletions(-) -- 2.0.2

2 7

[libvirt] [PATCH] virsh: man: Crosslink "desc" and "metadata" sections
by Peter Krempa 18 Aug '14

18 Aug '14

Those two commands work with a single API so cross-link them. --- tools/virsh.pod | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) diff --git a/tools/virsh.pod b/tools/virsh.pod index d072a7f..da68c08 100644 --- a/tools/virsh.pod +++ b/tools/virsh.pod @@ -656,6 +656,7 @@ effect on the next boot. Show or modify description and title of a domain. These values are user fields that allow to store arbitrary textual data to allow easy identification of domains. Title should be short, although it's not enforced. +(See also B<metadata> that works with XML based domain metadata.) Flags I<--live> or I<--config> select whether this command works on live or persistent definitions of the domain. If both I<--live> and I<--config> @@ -1261,7 +1262,8 @@ Show or modify custom XML metadata of a domain. The metadata is a user defined XML that allows to store arbitrary XML data in the domain definition. Multiple separate custom metadata pieces can be stored in the domain XML. The pieces are identified by a private XML namespace provided via the -I<uri> argument. +I<uri> argument. (See also B<desc> that works with textual metadata of +a domain.) Flags I<--live> or I<--config> select whether this command works on live or persistent definitions of the domain. If both I<--live> and I<--config> -- 2.0.2

2 2

[libvirt] [PATCH] man: virsh: Add man page for "virsh metadata"
by Peter Krempa 18 Aug '14

18 Aug '14

Patch adding the command forgot to add the man page entry. Resolves: https://bugzilla.redhat.com/show_bug.cgi?id=1130379 --- tools/virsh.pod | 30 ++++++++++++++++++++++++++++++ 1 file changed, 30 insertions(+) diff --git a/tools/virsh.pod b/tools/virsh.pod index f07deec..d072a7f 100644 --- a/tools/virsh.pod +++ b/tools/virsh.pod @@ -1254,6 +1254,36 @@ I<--total> for only the total stats, I<start> for only the per-cpu stats of the CPUs from I<start>, I<count> for only I<count> CPUs' stats. +=item B<metadata> I<domain> [[I<--live>] [I<--config>] | [I<--current>]] +[I<--edit>] [I<uri>] [I<key>] [I<set>] [I<--remove>] + +Show or modify custom XML metadata of a domain. The metadata is a user +defined XML that allows to store arbitrary XML data in the domain definition. +Multiple separate custom metadata pieces can be stored in the domain XML. +The pieces are identified by a private XML namespace provided via the +I<uri> argument. + +Flags I<--live> or I<--config> select whether this command works on live +or persistent definitions of the domain. If both I<--live> and I<--config> +are specified, the I<--config> option takes precedence on getting the current +description and both live configuration and config are updated while setting +the description. I<--current> is exclusive and implied if none of these was +specified. + +Flag I<--remove> specifies that the metadata element specified by the I<uri> +argument should be removed rather than updated. + +Flag I<--edit> specifies that an editor with the metadata identified by the +I<uri> argument should be opened and the contents saved back afterwards. +Otherwise the new contents can be provided via the I<set> argument. + +When setting metadata via I<--edit> or I<set> the I<key> argument must be +specified and is used to prefix the custom elements to bind them +to the private namespace. + +If neither of I<--edit> and I<set> are specified the XML metadata corresponding +to the I<uri> namespace is displayed instead of being modified. + =item B<migrate> [I<--live>] [I<--offline>] [I<--direct>] [I<--p2p> [I<--tunnelled>]] [I<--persistent>] [I<--undefinesource>] [I<--suspend>] [I<--copy-storage-all>] [I<--copy-storage-inc>] [I<--change-protection>] [I<--unsafe>] [I<--verbose>] -- 2.0.2

2 3

[libvirt] [PATCH] man: virsh: Add 'vcpu_period' and 'vcpu_quota' support info for LXC
by Li Yang 18 Aug '14

18 Aug '14

Signed-off-by: Li Yang <liyang.fnst(a)cn.fujitsu.com> --- tools/virsh.pod | 2 +- 1 files changed, 1 insertions(+), 1 deletions(-) diff --git a/tools/virsh.pod b/tools/virsh.pod index f07deec..a2d89b2 100644 --- a/tools/virsh.pod +++ b/tools/virsh.pod @@ -1570,7 +1570,7 @@ I<domain> Allows you to show (and set) the domain scheduler parameters. The parameters available for each hypervisor are: -LXC (posix scheduler) : cpu_shares +LXC (posix scheduler) : cpu_shares, vcpu_period, vcpu_quota QEMU/KVM (posix scheduler): cpu_shares, vcpu_period, vcpu_quota, emulator_period, emulator_quota -- 1.7.1

2 1

[libvirt] [libvirt-glib] [PATCH 0/4] Add some async versions of synchronous snapshot API
by Timm Bäder 18 Aug '14

18 Aug '14

This adds async versions of a few snapshot related functions and a patch that cleans up some doc comments I crewed up in earlier commits. All the sync versions don't seem cancellable to me since we can't just throw the result away like we do when fetching snapshots in case the operation gets cancelled, so none of them take a GCancellable. Timm Bäder (4): GVirDomain: Fix some doc comments of the snapshot API GVirDomainSnapshot: Add async version of _revert_to GVirDomain: Add async version of _create_snapshot GVirDomainSnapshot: Add async version of _delete libvirt-gobject/libvirt-gobject-domain-snapshot.c | 112 ++++++++++++++++++++++ libvirt-gobject/libvirt-gobject-domain-snapshot.h | 19 ++++ libvirt-gobject/libvirt-gobject-domain.c | 77 ++++++++++++++- libvirt-gobject/libvirt-gobject-domain.h | 11 +++ libvirt-gobject/libvirt-gobject.sym | 6 ++ 5 files changed, 220 insertions(+), 5 deletions(-) -- 2.0.4

2 9