September 2012 - Devel - libvirt List Archives

[libvirt] Add a ./run script for running programs from the local directory.
by Richard W.M. Jones 18 Sep '12

18 Sep '12

In libguestfs we successfully use a ./run script which sets up the environment so you can run libguestfs programs (internal and external ones, in C and non-C) without needing to install anything. This adds a similar script for libvirt. Here is the documentation included in the script: # With this script you can run libvirt programs without needing to # install them first. You just have to do for example: # # ./run ./tools/virsh [args ...] # # If you are already in the tools/ subdirectory, then the following # command will also work: # # ../run ./virsh [...] # # You can also run the C programs under valgrind like this: # # ./run valgrind [valgrind opts...] ./program # # or under gdb: # # ./run gdb --args ./program # # This also works with sudo (eg. if you need root access for libvirt): # # sudo ./run ./tools/virsh list --all

2 8

[libvirt] [PATCH] virBitmap: fix build without HAVE_NUMACTL
by Ján Tomko 18 Sep '12

18 Sep '12

Commit 75b198b3e7b2c8b4106afbd42f8bb795c4773511 forgot to change arguments of dummy qemuProcessInitNumaMemoryPolicy from char* to virBitmapPtr. --- src/qemu/qemu_process.c | 2 +- 1 files changed, 1 insertions(+), 1 deletions(-) diff --git a/src/qemu/qemu_process.c b/src/qemu/qemu_process.c index 5095381..c28f5a5 100644 --- a/src/qemu/qemu_process.c +++ b/src/qemu/qemu_process.c @@ -1798,7 +1798,7 @@ cleanup: #else static int qemuProcessInitNumaMemoryPolicy(virDomainObjPtr vm, - const char *nodemask ATTRIBUTE_UNUSED) + virBitmapPtr nodemask ATTRIBUTE_UNUSED) { if (vm->def->numatune.memory.nodemask) { virReportError(VIR_ERR_INTERNAL_ERROR, "%s", -- 1.7.8.6

3 2

[libvirt] [PATCHv3 0/4] qemu: Add seccomp sandbox support
by Ján Tomko 18 Sep '12

18 Sep '12

This series adds support to run QEMU with seccomp sandbox enabled. It can be configured in qemu.conf to on, off, or the QEMU default, which is off in 1.2. Default value is the QEMU default. V1 of the patches supports tri-state configuration and includes a test for detecting the capability. V2 includes libvirtd_qemu.aug, renames vague 'sandbox' to 'seccompSandbox' and prints an error if the sandbox was requested but QEMU doesn't have the capability. V3 is a rebase, needed after the addition of usb-redir.filter and the change of virBitmapPtr to qemuCapsPtr. Ján Tomko (4): qemu: add capability flag for seccomp sandbox qemu: conf: add seccomp_sandbox option qemu: add -sandbox to command line if requested tests: add qemu-1.2.0 help data src/qemu/libvirtd_qemu.aug | 1 + src/qemu/qemu.conf | 8 + src/qemu/qemu_capabilities.c | 3 + src/qemu/qemu_capabilities.h | 1 + src/qemu/qemu_command.c | 11 ++ src/qemu/qemu_conf.c | 5 + src/qemu/qemu_conf.h | 1 + tests/qemuhelpdata/qemu-1.2.0 | 270 ++++++++++++++++++++++++++++++++++ tests/qemuhelpdata/qemu-1.2.0-device | 181 +++++++++++++++++++++++ tests/qemuhelptest.c | 82 ++++++++++ 10 files changed, 563 insertions(+), 0 deletions(-) create mode 100644 tests/qemuhelpdata/qemu-1.2.0 create mode 100644 tests/qemuhelpdata/qemu-1.2.0-device -- 1.7.8.6

2 5

[libvirt] [PATCH] qemu: Avoid deadlock on HandleAgentEOF
by Michal Privoznik 18 Sep '12

18 Sep '12

On agent EOF the qemuProcessHandleAgentEOF() callback is called which locks virDomainObjPtr. Then qemuAgentClose() is called (with domain object locked) which eventually calls qemuAgentFree() and qemuProcessHandleAgentDestroy(). This tries to lock the domain object again. Hence the deadlock. --- src/qemu/qemu_process.c | 6 +++--- 1 files changed, 3 insertions(+), 3 deletions(-) diff --git a/src/qemu/qemu_process.c b/src/qemu/qemu_process.c index 5ac1d2b..8f42c83 100644 --- a/src/qemu/qemu_process.c +++ b/src/qemu/qemu_process.c @@ -115,7 +115,7 @@ extern struct qemud_driver *qemu_driver; * performed */ static void -qemuProcessHandleAgentEOF(qemuAgentPtr agent ATTRIBUTE_UNUSED, +qemuProcessHandleAgentEOF(qemuAgentPtr agent, virDomainObjPtr vm) { struct qemud_driver *driver = qemu_driver; @@ -127,12 +127,12 @@ qemuProcessHandleAgentEOF(qemuAgentPtr agent ATTRIBUTE_UNUSED, virDomainObjLock(vm); priv = vm->privateData; - - qemuAgentClose(agent); priv->agent = NULL; virDomainObjUnlock(vm); qemuDriverUnlock(driver); + + qemuAgentClose(agent); } -- 1.7.8.6

2 2

[libvirt] [PATCH 0/4 v3] Support to set disk wwn
by Osier Yang 18 Sep '12

18 Sep '12

v2 - v3: * Just rebase on the top. v1 - v2: * Validate wwn while parsing * Error out earlier "if (disk->device == "lun" && disk->wwn)", no logic change. This introduces new element <wwn> for disk, to allow to set wwn (just like setting serial number) for the virtual disk (Only QEMU devices like ide-drive, ide-hd, ide-cd, scsi-disk, scsi-hd, and scsi-cd support it). Osier Yang (4): schema: Add schema for disk <wwn> conf: Parse and format disk <wwn> qemu: Add caps to indentify if setting wwn is supported by qemu qemu: Use disk wwn in qemu command line docs/formatdomain.html.in | 6 +++ docs/schemas/basictypes.rng | 6 +++ docs/schemas/domaincommon.rng | 5 +++ docs/schemas/nodedev.rng | 6 --- src/conf/domain_conf.c | 12 +++++++ src/conf/domain_conf.h | 1 + src/libvirt_private.syms | 1 + src/qemu/qemu_capabilities.c | 8 ++++ src/qemu/qemu_capabilities.h | 2 + src/qemu/qemu_command.c | 33 ++++++++++++++++++ src/util/util.c | 17 +++++++++ src/util/util.h | 2 + .../qemuxml2argv-disk-ide-wwn.args | 6 +++ .../qemuxml2argvdata/qemuxml2argv-disk-ide-wwn.xml | 28 ++++++++++++++++ .../qemuxml2argv-disk-scsi-disk-wwn.args | 10 ++++++ .../qemuxml2argv-disk-scsi-disk-wwn.xml | 35 ++++++++++++++++++++ tests/qemuxml2argvtest.c | 7 ++++ 17 files changed, 179 insertions(+), 6 deletions(-) create mode 100644 tests/qemuxml2argvdata/qemuxml2argv-disk-ide-wwn.args create mode 100644 tests/qemuxml2argvdata/qemuxml2argv-disk-ide-wwn.xml create mode 100644 tests/qemuxml2argvdata/qemuxml2argv-disk-scsi-disk-wwn.args create mode 100644 tests/qemuxml2argvdata/qemuxml2argv-disk-scsi-disk-wwn.xml -- 1.7.7.3

2 9

[libvirt] [PATCH 0/3] block commit support
by Eric Blake 18 Sep '12

18 Sep '12

I'm still working on the qemu implementation of this API, but want to get the API reviewed and included in 0.10.2 before adding piecewise support for more and more of the API. My biggest issues for current implementations are: online qemu: Jeff's block-commit API is still not part of qemu.git, and might be subject to change, in particular based on review from Paolo's drive-mirror changes (since both patch series will be sharing related code on how to handle error reporting). offline qemu: qemu-img is a blocking command, but this API promises to be asynchronous. I'll have to run qemu-img in a worker thread, and I still need to figure out whether qemu-img can display progress for virDomainBlockJobInfo (can you send it a SIGUSR1, like for dd?), otherwise, the job will look like it is making no progress. But as neither of those issues invalidate the API, I'm comfortable with committing the API now (after all, that's what we did back in 0.9.12 with the virDomainBlockRebase API - we committed code in preparation for qemu to implement drive-mirror, and we are STILL waiting on qemu to follow through with that promise...) Obviously, there will be more patches to come as I get more things working at the qemu layer (I'm not sure whether online or offline qemu will be done first, as I have branches working on both approaches while I am testing against Jeff's qemu patch proposals). Eric Blake (3): blockjob: add virDomainBlockCommit blockjob: add virsh blockcommit blockjob: add blockcommit support to rpc docs/apibuild.py | 1 + include/libvirt/libvirt.h.in | 20 ++++++ src/driver.h | 5 ++ src/libvirt.c | 105 +++++++++++++++++++++++++++++ src/libvirt_public.syms | 1 + src/remote/remote_driver.c | 1 + src/remote/remote_protocol.x | 12 +++- src/remote_protocol-structs | 9 +++ src/rpc/gendispatch.pl | 1 + tools/virsh-domain.c | 153 ++++++++++++++++++++++++++++++++++++++++++- tools/virsh.pod | 30 +++++++++ 11 files changed, 336 insertions(+), 2 deletions(-) -- 1.7.11.4

2 9

[libvirt] [PATCH] qemu: drop unused arguments for dump-guest-memory
by Eric Blake 18 Sep '12

18 Sep '12

Upstream qemu has raised a concern about whether dumping guest memory by reading guest paging tables is a security hole: https://lists.gnu.org/archive/html/qemu-devel/2012-09/msg02607.html While auditing libvirt to see if we would be impacted, I noticed that we had some dead code. It is simpler to nuke the dead code and limit our monitor code to just the subset we make use of. * src/qemu/qemu_monitor.h (QEMU_MONITOR_DUMP): Drop poorly named and mostly-unused enum. * src/qemu/qemu_monitor.c (qemuMonitorDumpToFd): Drop arguments. * src/qemu/qemu_monitor_json.h (qemuMonitorJSONDump): Likewise. * src/qemu/qemu_monitor_json.c (qemuMonitorJSONDump): Likewise. * src/qemu/qemu_driver.c (qemuDumpToFd): Update caller. --- src/qemu/qemu_driver.c | 2 +- src/qemu/qemu_monitor.c | 12 ++++-------- src/qemu/qemu_monitor.h | 11 +---------- src/qemu/qemu_monitor_json.c | 24 +++++++----------------- src/qemu/qemu_monitor_json.h | 5 +---- 5 files changed, 14 insertions(+), 40 deletions(-) diff --git a/src/qemu/qemu_driver.c b/src/qemu/qemu_driver.c index f14c220..e6e5d02 100644 --- a/src/qemu/qemu_driver.c +++ b/src/qemu/qemu_driver.c @@ -3136,7 +3136,7 @@ static int qemuDumpToFd(struct qemud_driver *driver, virDomainObjPtr vm, if (qemuDomainObjEnterMonitorAsync(driver, vm, asyncJob) < 0) return -1; - ret = qemuMonitorDumpToFd(priv->mon, 0, fd, 0, 0); + ret = qemuMonitorDumpToFd(priv->mon, fd); qemuDomainObjExitMonitorWithDriver(driver, vm); return ret; diff --git a/src/qemu/qemu_monitor.c b/src/qemu/qemu_monitor.c index b7730fd..b772b28 100644 --- a/src/qemu/qemu_monitor.c +++ b/src/qemu/qemu_monitor.c @@ -2045,15 +2045,11 @@ int qemuMonitorMigrateCancel(qemuMonitorPtr mon) return ret; } -int qemuMonitorDumpToFd(qemuMonitorPtr mon, - unsigned int flags, - int fd, - unsigned long long begin, - unsigned long long length) +int +qemuMonitorDumpToFd(qemuMonitorPtr mon, int fd) { int ret; - VIR_DEBUG("mon=%p fd=%d flags=%x begin=%llx length=%llx", - mon, fd, flags, begin, length); + VIR_DEBUG("mon=%p fd=%d", mon, fd); if (!mon) { virReportError(VIR_ERR_INVALID_ARG, "%s", @@ -2073,7 +2069,7 @@ int qemuMonitorDumpToFd(qemuMonitorPtr mon, if (qemuMonitorSendFileHandle(mon, "dump", fd) < 0) return -1; - ret = qemuMonitorJSONDump(mon, flags, "fd:dump", begin, length); + ret = qemuMonitorJSONDump(mon, "fd:dump"); if (ret < 0) { if (qemuMonitorCloseFileHandle(mon, "dump") < 0) diff --git a/src/qemu/qemu_monitor.h b/src/qemu/qemu_monitor.h index e37dac8..232ae96 100644 --- a/src/qemu/qemu_monitor.h +++ b/src/qemu/qemu_monitor.h @@ -387,17 +387,8 @@ int qemuMonitorMigrateToUnix(qemuMonitorPtr mon, int qemuMonitorMigrateCancel(qemuMonitorPtr mon); -typedef enum { - QEMU_MONITOR_DUMP_HAVE_FILTER = 1 << 0, - QEMU_MONITOR_DUMP_PAGING = 1 << 1, - QEMU_MONITOR_DUMP_FLAGS_LAST -} QEMU_MONITOR_DUMP; - int qemuMonitorDumpToFd(qemuMonitorPtr mon, - unsigned int flags, - int fd, - unsigned long long begin, - unsigned long long length); + int fd); int qemuMonitorGraphicsRelocate(qemuMonitorPtr mon, int type, diff --git a/src/qemu/qemu_monitor_json.c b/src/qemu/qemu_monitor_json.c index b23be1e..a5198b0 100644 --- a/src/qemu/qemu_monitor_json.c +++ b/src/qemu/qemu_monitor_json.c @@ -2552,28 +2552,18 @@ int qemuMonitorJSONMigrateCancel(qemuMonitorPtr mon) return ret; } -int qemuMonitorJSONDump(qemuMonitorPtr mon, - unsigned int flags, - const char *protocol, - unsigned long long begin, - unsigned long long length) +int +qemuMonitorJSONDump(qemuMonitorPtr mon, + const char *protocol) { int ret; virJSONValuePtr cmd = NULL; virJSONValuePtr reply = NULL; - if (flags & QEMU_MONITOR_DUMP_HAVE_FILTER) - cmd = qemuMonitorJSONMakeCommand("dump-guest-memory", - "b:paging", flags & QEMU_MONITOR_DUMP_PAGING ? 1 : 0, - "s:protocol", protocol, - "U:begin", begin, - "U:length", length, - NULL); - else - cmd = qemuMonitorJSONMakeCommand("dump-guest-memory", - "b:paging", flags & QEMU_MONITOR_DUMP_PAGING ? 1 : 0, - "s:protocol", protocol, - NULL); + cmd = qemuMonitorJSONMakeCommand("dump-guest-memory", + "b:paging", false, + "s:protocol", protocol, + NULL); if (!cmd) return -1; diff --git a/src/qemu/qemu_monitor_json.h b/src/qemu/qemu_monitor_json.h index bdcf819..b592d12 100644 --- a/src/qemu/qemu_monitor_json.h +++ b/src/qemu/qemu_monitor_json.h @@ -138,10 +138,7 @@ int qemuMonitorJSONMigrate(qemuMonitorPtr mon, int qemuMonitorJSONMigrateCancel(qemuMonitorPtr mon); int qemuMonitorJSONDump(qemuMonitorPtr mon, - unsigned int flags, - const char *protocol, - unsigned long long begin, - unsigned long long length); + const char *protocol); int qemuMonitorJSONGraphicsRelocate(qemuMonitorPtr mon, int type, -- 1.7.11.4

2 2

[libvirt] [PATCH 0/3]generate man pages to describe valid keycodes
by Guannan Ren 18 Sep '12

18 Sep '12

This patchset will try to generate manpages to describe valid keycodes. It does this job via a helper perl script passing file src/util/keymaps.csv. we have 1 man page per keycode set, that is, 11 manpages to be generated in total. Each of them corresponds to codeset defined in virsh manpage. libvirt-keycode-linux.5 libvirt-keycode-xt.5 libvirt-keycode-atset1.5 libvirt-keycode-atset2.5 libvirt-keycode-atset3.5 libvirt-keycode-os_x.5 libvirt-keycode-kbd.5 libvirt-keycode-win32.5 libvirt-keycode-usb.5 libvirt-keycode-rfb.5 I really think that the words in these manpages still be polished again. If anyone could help, that would be appreciated. Guannan Ren(3) doc: Add a perl script to generate keycodes mapping automake: generate libvirt keycodes manpages doc: fix some typoes on virsh manpage .gitignore | 2 +- tools/Makefile.am | 31 +++++++++++++++++++++++++--- tools/keymap-gen.pl | 202 +++++++++++++++++++++++++++++++++++++++++++++ tools/virsh.pod | 27 +++++++++++++------------ 4 files changed, 245 insertions(+), 17 deletions(-)

2 6

[libvirt] [PATCH] build: Fix build failure on mingw32 platform
by Osier Yang 18 Sep '12

18 Sep '12

--- I'm not sure if we prefer indentions inside the block between "#ifdef __linux__" ... "#else" clause. nodeinfo.c uses both (indentions/no indentions). --- src/nodeinfo.c | 12 ++++++++---- 1 files changed, 8 insertions(+), 4 deletions(-) diff --git a/src/nodeinfo.c b/src/nodeinfo.c index 7af6512..1ec3ada 100644 --- a/src/nodeinfo.c +++ b/src/nodeinfo.c @@ -971,13 +971,14 @@ cleanup: int nodeSetMemoryParameters(virConnectPtr conn ATTRIBUTE_UNUSED, - virTypedParameterPtr params, - int nparams, + virTypedParameterPtr params ATTRIBUTE_UNUSED, + int nparams ATTRIBUTE_UNUSED, unsigned int flags) { virCheckFlags(0, -1); #ifdef __linux__ + { int ret = 0; int i; @@ -1010,6 +1011,7 @@ nodeSetMemoryParameters(virConnectPtr conn ATTRIBUTE_UNUSED, } return ret; + } #else virReportError(VIR_ERR_NO_SUPPORT, "%s", _("node set memory parameters not implemented" @@ -1066,13 +1068,14 @@ cleanup: #define NODE_MEMORY_PARAMETERS_NUM 7 int nodeGetMemoryParameters(virConnectPtr conn ATTRIBUTE_UNUSED, - virTypedParameterPtr params, - int *nparams, + virTypedParameterPtr params ATTRIBUTE_UNUSED, + int *nparams ATTRIBUTE_UNUSED, unsigned int flags) { virCheckFlags(VIR_TYPED_PARAM_STRING_OKAY, -1); #ifdef __linux__ + { unsigned int pages_to_scan; unsigned int sleep_millisecs; unsigned long long pages_shared; @@ -1174,6 +1177,7 @@ nodeGetMemoryParameters(virConnectPtr conn ATTRIBUTE_UNUSED, } return 0; + } #else virReportError(VIR_ERR_NO_SUPPORT, "%s", _("node get memory parameters not implemented" -- 1.7.7.3

2 3

[libvirt] [PATCH] virsh: Clarify behavior of domain list filtering.
by Peter Krempa 17 Sep '12

17 Sep '12

Some combinations of filtering flags produce no result. This patch tries to clarify this. Also fix unintentional double space. --- tools/virsh.pod | 11 ++++++----- 1 file changed, 6 insertions(+), 5 deletions(-) diff --git a/tools/virsh.pod b/tools/virsh.pod index c6f205a..2efb277 100644 --- a/tools/virsh.pod +++ b/tools/virsh.pod @@ -395,10 +395,11 @@ into s3 state. Normally only active domains are listed. To list inactive domains specify I<--inactive> or I<--all> to list both active and inactive domains. -To filter the list of domains present on the hypervisor you may specify one or -more of filtering flags supported by the B<list> command. These flags are -grouped by function. Specifying one or more flags from a group enables the -filter group. Supported filtering flags and groups: +To further filter the list of domains you may specify one or more of filtering +flags supported by the B<list> command. These flags are grouped by function. +Specifying one or more flags from a group enables the filter group. Note that +some combinations of flags may yield no results. Supported filtering flags and +groups: =over 4 @@ -435,7 +436,7 @@ domains without a snapshot I<--without-snapshot>. When talking to older servers, this command is forced to use a series of API calls with an inherent race, where a domain might not be listed or might appear more than once if it changed state between calls while the list was being -collected. Newer servers do not have this problem. +collected. Newer servers do not have this problem. If I<--managed-save> is specified, then domains that have managed save state (only possible if they are in the B<shut off> state, so you need to specify -- 1.7.12

3 4