- Devel - libvirt List Archives

[libvirt] libvirt-client leaks memory, Ubuntu and Debian-specific
by Igor Lukyanov 21 Feb '13

21 Feb '13

Hello. We faced a very strange leak while using libvirt library in long-running server application for cluster orchestration. Leak does not directly related to libvirt code and exposed only on specific build options (?) and/or system environment (?). Here are the key points: 1. Libvirt client leaks memory while making (RPC) calls to a server. I mean that RSS memory usage shown by ps, top, etc. indefinitely grows (plz check an attachment ps.log). Test app attached. 2. Leak detected on Debian and Ubuntu and absent on Mac OS and Gentoo, so it's exactly an environment or build problem. 3. Valgrind does not see the leak. From valgrind's point of view from start to finish application constantly consumes 110kb of memory (while ps shows multiple megabytes) and does not contain any leaks. 4. Logging activity of virMalloc, virRealloc, etc. functions does not show anything: as expected, all allocated memory is correctly freed, so it's definitely not a bug of code (we tested that before recognized that problem is distrib/platform specific). Some useful logs and test code attached. I think digging build options and legoing system libraries will help us to beat the problem but it would be nice if someone already had a working solution. Thank you for help.

3 5

[libvirt] [PATCH 0/6 v3] Shared disk table related fixes/improvements
by Osier Yang 21 Feb '13

21 Feb '13

Inspires by the crash when using CD-ROM or Floppy with empty source, except the fixes for the crashes, this also enrichs the hash table with a list of names of domain which use the shared disk. And to keep the hash table to be in consistent state, this also fixes the problems on adding/removing the hash entry when updateing/ejecting media of CD-ROM or Floppy. And updates the hash table when reconnecting domain process. v2 - v3: * Rebase on the top * Small changes after more testing. Osier Yang (6): qemu: Add checking in helpers for sgio setting qemu: Merge qemuCheckSharedDisk into qemuAddSharedDisk qemu: Record names of domain which uses the shared disk in hash table qemu: Update shared disk table when reconnecting qemu process qemu: Move the shared disk adding and sgio setting prior to attaching qemu: Remove the shared disk entry if the operation is ejecting or updating src/conf/domain_conf.c | 20 ++++ src/conf/domain_conf.h | 3 + src/libvirt_private.syms | 1 + src/qemu/qemu_conf.c | 243 ++++++++++++++++++++++++++++++++++++++++++--- src/qemu/qemu_conf.h | 26 ++++- src/qemu/qemu_driver.c | 108 ++++++++++++++++----- src/qemu/qemu_hotplug.c | 19 +---- src/qemu/qemu_hotplug.h | 1 + src/qemu/qemu_process.c | 81 ++++------------ src/qemu/qemu_process.h | 3 - 10 files changed, 377 insertions(+), 128 deletions(-) -- 1.7.7.6

3 19

[libvirt] [PATCH] maint: enforce private symbol section sorting
by Eric Blake 21 Feb '13

21 Feb '13

Automating a sorting check is the only way to ensure we don't regress. Suggested by Dan Berrange. * src/check-symsorting.pl (check_sorting): Add a parameter, validate that groups are in order, and that files exist. * src/Makefile.am (check-symsorting): Adjust caller. * src/libvirt_private.syms: Fix typo. * src/libvirt_linux.syms: Fix file name. * src/libvirt_vmx.syms: Likewise. * src/libvirt_xenxs.syms: Likewise. * src/libvirt_sasl.syms: Likewise. * src/libvirt_libssh2.syms: Likewise. * src/libvirt_esx.syms: Mention file name. * src/libvirt_openvz.syms: Likewise. --- While I can manage in perl, I probably don't have the best usage patterns; suggestions for improving this patch are quite welcome. src/Makefile.am | 3 ++- src/check-symsorting.pl | 35 +++++++++++++++++++++++++++++------ src/libvirt_esx.syms | 1 + src/libvirt_libssh2.syms | 3 +-- src/libvirt_linux.syms | 2 +- src/libvirt_openvz.syms | 2 ++ src/libvirt_private.syms | 2 +- src/libvirt_sasl.syms | 8 ++++---- src/libvirt_vmx.syms | 2 +- src/libvirt_xenxs.syms | 4 ++-- 10 files changed, 44 insertions(+), 18 deletions(-) diff --git a/src/Makefile.am b/src/Makefile.am index 780cd52..3ef9a9c 100644 --- a/src/Makefile.am +++ b/src/Makefile.am @@ -388,7 +388,8 @@ else check-symfile: endif check-symsorting: - $(AM_V_GEN)$(PERL) $(srcdir)/check-symsorting.pl $(USED_SYM_FILES) + $(AM_V_GEN)$(PERL) $(srcdir)/check-symsorting.pl \ + $(srcdir) $(USED_SYM_FILES) EXTRA_DIST += check-symfile.pl check-symsorting.pl PROTOCOL_STRUCTS = \ diff --git a/src/check-symsorting.pl b/src/check-symsorting.pl index 9c62246..cbcb737 100755 --- a/src/check-symsorting.pl +++ b/src/check-symsorting.pl @@ -3,22 +3,27 @@ use strict; use warnings; -die "syntax: $0 SYMFILE..." unless int(@ARGV) >= 1; +die "syntax: $0 SRCDIR SYMFILE..." unless int(@ARGV) >= 2; my $ret = 0; +my $srcdir = shift; +my $lastgroup = undef; foreach my $symfile (@ARGV) { open SYMFILE, $symfile or die "cannot read $symfile: $!"; my $line; + my $groupfile; my @group; while (<SYMFILE>) { chomp; - next if /^#/; - if (/^\s*$/) { + if (/^#/) { + $_ =~ s/^# //; + $groupfile = $_; + } elsif (/^\s*$/) { if (@group) { - &check_sorting(\@group, $symfile, $line); + &check_sorting(\@group, $symfile, $line, $groupfile); } @group = (); $line = $.; @@ -30,20 +35,38 @@ foreach my $symfile (@ARGV) { close SYMFILE; if (@group) { - &check_sorting(\@group, $symfile, $line); + &check_sorting(\@group, $symfile, $line, $groupfile); } + $lastgroup = undef; } sub check_sorting { my $group = shift; my $symfile = shift; my $line = shift; + my $groupfile = shift; my @group = @{$group}; my @sorted = sort { lc $a cmp lc $b } @group; my $sorted = 1; my $first; my $last; + + # Check that groups are in order and groupfile exists + if (defined $lastgroup && lc $lastgroup ge lc $groupfile) { + print "Symbol block at $symfile:$line: block not sorted\n"; + print "Move $groupfile block before $lastgroup block\n"; + print "\n"; + $ret = 1; + } + if (! -e "$srcdir/$groupfile") { + print "Symbol block at $symfile:$line: $groupfile not found\n"; + print "\n"; + $ret = 1; + } + $lastgroup = $groupfile; + + # Check that symbols within a group are in order for (my $i = 0 ; $i <= $#sorted ; $i++) { if ($sorted[$i] ne $group[$i]) { $first = $i unless defined $first; @@ -54,7 +77,7 @@ sub check_sorting { if (!$sorted) { @group = splice @group, $first, ($last-$first+1); @sorted = splice @sorted, $first, ($last-$first+1); - print "Symbol block at $symfile:$line symbols not sorted\n"; + print "Symbol block at $symfile:$line: symbols not sorted\n"; print map { " " . $_ . "\n" } @group; print "Correct ordering\n"; print map { " " . $_ . "\n" } @sorted; diff --git a/src/libvirt_esx.syms b/src/libvirt_esx.syms index ad3d60a..3c14b94 100644 --- a/src/libvirt_esx.syms +++ b/src/libvirt_esx.syms @@ -2,6 +2,7 @@ # These symbols are dependent upon --with-esx via WITH_ESX # +# esx/esx_util.h esxUtil_EscapeDatastoreItem; esxUtil_ParseDatastorePath; esxVI_DateTime_ConvertToCalendarTime; diff --git a/src/libvirt_libssh2.syms b/src/libvirt_libssh2.syms index 815195b..55022b5 100644 --- a/src/libvirt_libssh2.syms +++ b/src/libvirt_libssh2.syms @@ -2,8 +2,7 @@ # ssh session - specific symbols # -# virnetsshsession.h -# +# rpc/virnetsshsession.h virNetSSHChannelRead; virNetSSHChannelWrite; virNetSSHSessionAuthAddAgentAuth; diff --git a/src/libvirt_linux.syms b/src/libvirt_linux.syms index b5f7ee2..3500898 100644 --- a/src/libvirt_linux.syms +++ b/src/libvirt_linux.syms @@ -5,7 +5,7 @@ # nodeinfo.h linuxNodeInfoCPUPopulate; -# stats_linux.h +# util/virstatslinux.h linuxDomainInterfaceStats; # Let emacs know we want case-insensitive sorting diff --git a/src/libvirt_openvz.syms b/src/libvirt_openvz.syms index 58b93ae..ac0ed0d 100644 --- a/src/libvirt_openvz.syms +++ b/src/libvirt_openvz.syms @@ -1,6 +1,8 @@ # # These symbols are dependent upon --with-openvz via WITH_OPENVZ # + +# openvz/openvz_conf.h openvzLocateConfFile; openvzReadConfigParam; openvzReadNetworkConf; diff --git a/src/libvirt_private.syms b/src/libvirt_private.syms index 1686bb8..3c44cc3 100644 --- a/src/libvirt_private.syms +++ b/src/libvirt_private.syms @@ -420,7 +420,7 @@ virNetDevVlanFormat; virNetDevVlanParse; -# conf/netdev_vportprofile_conf.h +# conf/netdev_vport_profile_conf.h virNetDevVPortProfileFormat; virNetDevVPortProfileParse; virNetDevVPortTypeFromString; diff --git a/src/libvirt_sasl.syms b/src/libvirt_sasl.syms index 099f48a..beb8825 100644 --- a/src/libvirt_sasl.syms +++ b/src/libvirt_sasl.syms @@ -2,10 +2,10 @@ # SASL-specific symbols # -# virnetclient.h +# rpc/virnetclient.h virNetClientSetSASLSession; -# virnetsaslcontext.h +# rpc/virnetsaslcontext.h virNetSASLContextCheckIdentity; virNetSASLContextNewClient; virNetSASLContextNewServer; @@ -25,11 +25,11 @@ virNetSASLSessionServerStart; virNetSASLSessionServerStep; -# virnetserverclient.h +# rpc/virnetserverclient.h virNetServerClientSetSASLSession; -# virnetsocket.h +# rpc/virnetsocket.h virNetSocketSetSASLSession; # Let emacs know we want case-insensitive sorting diff --git a/src/libvirt_vmx.syms b/src/libvirt_vmx.syms index 99fe590..0b15f49 100644 --- a/src/libvirt_vmx.syms +++ b/src/libvirt_vmx.syms @@ -2,7 +2,7 @@ # These symbols are dependent upon --with-esx via WITH_ESX or --with-vmware via WITH_VMWARE. # -# vmx.h +# vmx/vmx.h virVMXConvertToUTF8; virVMXEscapeHex; virVMXFormatCDROM; diff --git a/src/libvirt_xenxs.syms b/src/libvirt_xenxs.syms index 75d5322..04b35c4 100644 --- a/src/libvirt_xenxs.syms +++ b/src/libvirt_xenxs.syms @@ -2,7 +2,7 @@ # These symbols are dependent upon --with-xen via WITH_XEN or --with-libxl via WITH_LIBXL. # -# xen_sxpr.h +# xenxs/xen_sxpr.h xenFormatSxpr; xenFormatSxprChr; xenFormatSxprDisk; @@ -16,7 +16,7 @@ xenParseSxprChar; xenParseSxprSound; xenParseSxprString; -# xen_xm.h +# xenxs/xen_xm.h xenFormatXM; xenParseXM; -- 1.8.1.2

2 2

[libvirt] [PATCH] qemu: Do not ignore mandatory features in migration cookie
by Jiri Denemark 20 Feb '13

20 Feb '13

Due to "feature"/"features" nasty typo, any features marked as mandatory by one side of a migration are silently considered optional by the other side. The following is the code that formats mandatory features in migration cookie: for (i = 0 ; i < QEMU_MIGRATION_COOKIE_FLAG_LAST ; i++) { if (mig->flagsMandatory & (1 << i)) virBufferAsprintf(buf, " <feature name='%s'/>\n", qemuMigrationCookieFlagTypeToString(i)); } --- src/qemu/qemu_migration.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/src/qemu/qemu_migration.c b/src/qemu/qemu_migration.c index 82c3f97..17c1714 100644 --- a/src/qemu/qemu_migration.c +++ b/src/qemu/qemu_migration.c @@ -863,7 +863,7 @@ qemuMigrationCookieXMLParse(qemuMigrationCookiePtr mig, /* Check to ensure all mandatory features from XML are also * present in 'flags' */ - if ((n = virXPathNodeSet("./features", ctxt, &nodes)) < 0) + if ((n = virXPathNodeSet("./feature", ctxt, &nodes)) < 0) goto error; for (i = 0 ; i < n ; i++) { -- 1.8.1.2

2 2

[libvirt] [PATCH v2] qemu: Run lzop with '--ignore-warn'
by Michal Privoznik 20 Feb '13

20 Feb '13

Currently, if lzop decompression binary produces a warning, it doesn't exit with zero status but 2 instead. Terrifying, but true. However, warnings may be ignored using '--ignore-warn' command line argument. Moreover, in which case, the exit status will be zero. --- src/qemu/qemu_driver.c | 62 ++++++++++++++++++++++++++++++++------------------ 1 file changed, 40 insertions(+), 22 deletions(-) diff --git a/src/qemu/qemu_driver.c b/src/qemu/qemu_driver.c index dc35b91..a0a1f04 100644 --- a/src/qemu/qemu_driver.c +++ b/src/qemu/qemu_driver.c @@ -2507,6 +2507,34 @@ qemuCompressProgramName(int compress) qemuSaveCompressionTypeToString(compress)); } +static virCommandPtr +qemuCompressGetCommand(virQEMUSaveFormat compression) +{ + virCommandPtr ret = NULL; + const char *prog = qemuSaveCompressionTypeToString(compression); + + if (!prog) { + virReportError(VIR_ERR_OPERATION_FAILED, + _("Invalid compressed save format %d"), + compression); + return NULL; + } + + ret = virCommandNew(prog); + virCommandAddArg(ret, "-dc"); + + switch (compression) { + case QEMU_SAVE_FORMAT_LZOP: + virCommandAddArg(ret, "--ignore-warn"); + break; + default: + /* Ain't no valid compressed save format */ + break; + } + + return ret; +} + /* Internal function to properly create or open existing files, with * ownership affected by qemu driver setup. */ static int @@ -4775,32 +4803,22 @@ qemuDomainSaveImageStartVM(virConnectPtr conn, if (!(caps = virQEMUDriverGetCapabilities(driver, false))) goto cleanup; - if (header->version == 2) { - const char *prog = qemuSaveCompressionTypeToString(header->compressed); - if (prog == NULL) { - virReportError(VIR_ERR_OPERATION_FAILED, - _("Invalid compressed save format %d"), - header->compressed); + if ((header->version == 2) && + (header->compressed != QEMU_SAVE_FORMAT_RAW)) { + if (!(cmd = qemuCompressGetCommand(header->compressed))) goto cleanup; - } - if (header->compressed != QEMU_SAVE_FORMAT_RAW) { - cmd = virCommandNewArgList(prog, "-dc", NULL); - intermediatefd = *fd; - *fd = -1; + intermediatefd = *fd; + *fd = -1; - virCommandSetInputFD(cmd, intermediatefd); - virCommandSetOutputFD(cmd, fd); - virCommandSetErrorBuffer(cmd, &errbuf); - virCommandDoAsyncIO(cmd); + virCommandSetInputFD(cmd, intermediatefd); + virCommandSetOutputFD(cmd, fd); + virCommandSetErrorBuffer(cmd, &errbuf); + virCommandDoAsyncIO(cmd); - if (virCommandRunAsync(cmd, NULL) < 0) { - virReportError(VIR_ERR_INTERNAL_ERROR, - _("Failed to start decompression binary %s"), - prog); - *fd = intermediatefd; - goto cleanup; - } + if (virCommandRunAsync(cmd, NULL) < 0) { + *fd = intermediatefd; + goto cleanup; } } -- 1.8.1.2

2 1

[libvirt] [PATCH v3 RESEND 00/12] Rework storage migration
by Michal Privoznik 20 Feb '13

20 Feb '13

This patch set re-implements migration with storage for enough new qemu. Currently, you can migrate a domain to a host without need for shared storage. This is done by setting 'blk' or 'inc' attribute (representing VIR_MIGRATE_NON_SHARED_DISK and VIR_MIGRATE_NON_SHARED_INC flags respectively) of 'migrate' monitor command. However, the qemu implementation is buggy and applications are advised to switch to new impementation which, moreover, offers some nice features, like migrating only explicitly specified disks. The new functionality is controlled via 'nbd-server-*' and 'drive-mirror' commands. The flow is meant to look like this: 1) User invokes libvirt's migrate functionality. 2) libvirt checks that no block jobs are active on the source. 3) libvirt starts the destination QEMU and sets up the NBD server using the nbd-server-start and nbd-server-add commands. 4) libvirt starts drive-mirror with a destination pointing to the remote NBD server, for example nbd:host:port:exportname=diskname (where diskname is the -drive id specified on the destination). 5) once all mirroring jobs reach steady state, libvirt invokes the migrate command. 6) once migration completed, libvirt invokes the nbd-server-stop command on the destination QEMU. If we just skip the 2nd step and there is an active block-job, qemu will fail in step 4. No big deal. Since we try to NOT break migration and keep things compatible, this feature is enabled iff both sides support it. Since there's obvious need for some data transfer between src and dst, I've put it into qemuCookieMigration: 1) src -> dest: (QEMU_MIGRATION_PHASE_BEGIN3 -> QEMU_MIGRATION_PHASE_PREPARE) <nbd> <disk target='vda' size='15032385536'/> <disk target='vdb' size='11534336'/> <disk target='vdc' size='13631488'/> </nbd> The source is telling the destination it supports the NBD feature. Moreover, which disks are to be transferred and what are their sizes. That's because disks needs to be fully allocated (even qcow) for successful transfer. 2) dst -> src: (QEMU_MIGRATION_PHASE_PREPARE -> QEMU_MIGRATION_PHASE_PERFORM3) <nbd port='5901'/> The destination is confirming it does support NBD as well. All disks were pre-created and NBD server is listening on given port (5901 in this case). If either src or dst doesn't support NBD, it is not used and whole process falls back to old implementation. diff to v1: -Eric's and Daniel's suggestions worked in. To point out the bigger ones: don't do NBD style when TUNNELLED requested, added 'b:writable' to 'nbd-server-add' -drop '/qemu-migration/nbd/disk/@src' attribute from migration cookie. As pointed out by Jirka, disk->src can be changed during migration (e.g. by migration hook or by passed xml). So I've tried (as suggested on the list) passing disk alias. However, since qemu hasn't been started on destination yet, the aliases hasn't been generated yet. So we have to rely on ordering completely. diff to v2: -rebase to reflect changes made by offline migration patch -send initial nbd cookie only if needed diff to v2.1: -nbd cookie reworked -don't rely on disk ordering in the cookie, but use disk target for that -adapt to virPortAllocator -unlink pre-created storage on migration fail -other of Jirka's suggestions worked in "diff" to v3: -just rebase & adapt to new qemu code after dropping QDL (Qemu Driver Lock) Michal Privoznik (12): qemu: Introduce NBD_SERVER capability Introduce NBD migration cookie qemu: Introduce nbd-server-start command qemu: Introduce nbd-server-add command qemu: Introduce nbd-server-stop command qemu_migration: Introduce qemuMigrationStartNBDServer() qemu_migration: Introduce qemuMigrationDriveMirror qemu_domain: Introduce qemuDomainGetDiskBlockInfo qemu_migration: Check size prerequisites qemu_migration: Stop NBD server at Finish phase qemu_migration: Cancel running jobs on failed migration qemu_migration: Unlink pre-created storage on error src/qemu/qemu_capabilities.c | 4 +- src/qemu/qemu_capabilities.h | 1 + src/qemu/qemu_domain.c | 157 ++++++++- src/qemu/qemu_domain.h | 7 + src/qemu/qemu_driver.c | 124 +------ src/qemu/qemu_migration.c | 767 ++++++++++++++++++++++++++++++++++++++++++- src/qemu/qemu_monitor.c | 63 ++++ src/qemu/qemu_monitor.h | 7 + src/qemu/qemu_monitor_json.c | 102 ++++++ src/qemu/qemu_monitor_json.h | 7 + src/qemu/qemu_process.c | 13 + 11 files changed, 1116 insertions(+), 136 deletions(-) -- 1.8.0.2

2 25

[libvirt] [PATCHv2] qemu: enable direct migration over IPv6
by Ján Tomko 20 Feb '13

20 Feb '13

Use virURIParse in qemuMigrationPrepareDirect to allow parsing IPv6 addresses, which would cause an 'incorrect :port' error message before. To be able to migrate over IPv6, QEMU needs to listen on [::] instead of 0.0.0.0. This patch adds a call to getaddrinfo and sets the listen address based on the result. This will break migration if a hostname that can only be resolved on the source machine is passed in the migration URI, or if it does not resolve to the same address family on both sides. Bug: https://bugzilla.redhat.com/show_bug.cgi?id=846013 --- Diff to V1: * initialize uri_str * reuse STRSKIP("tcp:") result instead of doing strlen on it * print a warning instead of failing when the hostname can't be resolved src/qemu/qemu_migration.c | 64 +++++++++++++++++++++++++++++++++++++---------- 1 file changed, 51 insertions(+), 13 deletions(-) diff --git a/src/qemu/qemu_migration.c b/src/qemu/qemu_migration.c index 36e55d2..62e9260 100644 --- a/src/qemu/qemu_migration.c +++ b/src/qemu/qemu_migration.c @@ -22,7 +22,10 @@ #include <config.h> +#include <netdb.h> +#include <sys/socket.h> #include <sys/time.h> +#include <sys/types.h> #ifdef WITH_GNUTLS # include <gnutls/gnutls.h> # include <gnutls/x509.h> @@ -1836,7 +1839,11 @@ qemuMigrationPrepareDirect(virQEMUDriverPtr driver, char *hostname = NULL; char migrateFrom [64]; const char *p; + char *uri_str = NULL; int ret = -1; + bool ipv6 = false; + struct addrinfo *info; + virURIPtr uri; VIR_DEBUG("driver=%p, dconn=%p, cookiein=%s, cookieinlen=%d, " "cookieout=%p, cookieoutlen=%p, uri_in=%s, uri_out=%p, " @@ -1885,16 +1892,39 @@ qemuMigrationPrepareDirect(virQEMUDriverPtr driver, * URI when passing it to the qemu monitor, so bad * characters in hostname part don't matter. */ - if (!STRPREFIX(uri_in, "tcp:")) { + if (!(p = STRSKIP(uri_in, "tcp:"))) { virReportError(VIR_ERR_INVALID_ARG, "%s", _("only tcp URIs are supported for KVM/QEMU" " migrations")); goto cleanup; } - /* Get the port number. */ - p = strrchr(uri_in, ':'); - if (p == strchr(uri_in, ':')) { + /* Convert uri_in to well-formed URI with // after tcp: */ + if (!(STRPREFIX(uri_in, "tcp://"))) { + if (virAsprintf(&uri_str, "tcp://%s", p) < 0) { + virReportOOMError(); + goto cleanup; + } + } + + uri = virURIParse(uri_str ? uri_str : uri_in); + VIR_FREE(uri_str); + + if (uri == NULL) { + virReportError(VIR_ERR_INVALID_ARG, _("unable to parse URI: %s"), + uri_in); + goto cleanup; + } + + if (uri->server == NULL) { + virReportError(VIR_ERR_INVALID_ARG, _("missing host in migration" + " URI: %s"), uri_in); + goto cleanup; + } else { + hostname = uri->server; + } + + if (uri->port == 0) { /* Generate a port */ this_port = QEMUD_MIGRATION_FIRST_PORT + port++; if (port == QEMUD_MIGRATION_NUM_PORTS) @@ -1907,21 +1937,29 @@ qemuMigrationPrepareDirect(virQEMUDriverPtr driver, } } else { - p++; /* definitely has a ':' in it, see above */ - this_port = virParseNumber(&p); - if (this_port == -1 || p-uri_in != strlen(uri_in)) { - virReportError(VIR_ERR_INVALID_ARG, - "%s", _("URI ended with incorrect ':port'")); - goto cleanup; - } + this_port = uri->port; } } + if (getaddrinfo(hostname, NULL, NULL, &info)) { + VIR_WARN("unable to get address info for %s, defaulting to IPv4", + hostname); + } else { + ipv6 = info->ai_family == AF_INET6; + } + if (*uri_out) VIR_DEBUG("Generated uri_out=%s", *uri_out); - /* QEMU will be started with -incoming tcp:0.0.0.0:port */ - snprintf(migrateFrom, sizeof(migrateFrom), "tcp:0.0.0.0:%d", this_port); + /* QEMU will be started with -incoming tcp:0.0.0.0:port + * or -incoming tcp:[::]:port for IPv6 */ + if (ipv6) { + snprintf(migrateFrom, sizeof(migrateFrom), + "tcp:[::]:%d", this_port); + } else { + snprintf(migrateFrom, sizeof(migrateFrom), + "tcp:0.0.0.0:%d", this_port); + } ret = qemuMigrationPrepareAny(driver, dconn, cookiein, cookieinlen, cookieout, cookieoutlen, dname, dom_xml, -- 1.7.12.4

2 3

[libvirt] [PATCH v2 RESEND 0/2] cgroup refactor
by Hu Tao 20 Feb '13

20 Feb '13

This series is for early review. This series refactors cgroup code to: - provide lazy creation of cgroup directories, despite of what level they are. - remove cgroup directories if no one is using the corresponding virCgroup. changes from v2: - rebase to latest tree. - squash patches 2-5 in v2 into one. Hu Tao (2): refactor virCgroupDetectMounts and virCgroupDetectPlacement cgroup: refactor virCgroup src/conf/domain_conf.h | 5 + src/libvirt_private.syms | 7 +- src/lxc/lxc_cgroup.c | 40 +- src/lxc/lxc_cgroup.h | 2 +- src/lxc/lxc_controller.c | 31 +- src/lxc/lxc_driver.c | 177 +++---- src/lxc/lxc_process.c | 19 +- src/qemu/qemu_cgroup.c | 162 +++--- src/qemu/qemu_cgroup.h | 3 +- src/qemu/qemu_driver.c | 350 +++++-------- src/qemu/qemu_hotplug.c | 21 +- src/qemu/qemu_migration.c | 20 +- src/qemu/qemu_process.c | 7 +- src/util/vircgroup.c | 1263 +++++++++++++++++++++------------------------ src/util/vircgroup.h | 19 +- 15 files changed, 920 insertions(+), 1206 deletions(-) -- 1.8.0.1.240.ge8a1f5a

1 3

[libvirt] [PATCH] virsh: distinguish errors between missing argument and wrong option
by Guannan Ren 20 Feb '13

20 Feb '13

Specifying ':' to suppress the error messages printed by getopt(). Then, distinguish the two types of errors. Before: virsh: option requires an argument -- 'c' error: unsupported option '-?'. See --help. After: error: option '-c' requires an argument error: unsupported option '-x'. See --help. --- tools/virsh.c | 10 ++++++++-- 1 file changed, 8 insertions(+), 2 deletions(-) diff --git a/tools/virsh.c b/tools/virsh.c index f5a01b3..c0e62eb 100644 --- a/tools/virsh.c +++ b/tools/virsh.c @@ -2919,7 +2919,7 @@ vshParseArgv(vshControl *ctl, int argc, char **argv) /* Standard (non-command) options. The leading + ensures that no * argument reordering takes place, so that command options are * not confused with top-level virsh options. */ - while ((arg = getopt_long(argc, argv, "+d:hqtc:vVrl:e:", opt, NULL)) != -1) { + while ((arg = getopt_long(argc, argv, "+:d:hqtc:vVrl:e:", opt, NULL)) != -1) { switch (arg) { case 'd': if (virStrToLong_i(optarg, NULL, 10, &debug) < 0) { @@ -2973,8 +2973,14 @@ vshParseArgv(vshControl *ctl, int argc, char **argv) exit(EXIT_FAILURE); } break; + case ':': + vshError(ctl, _("option '-%c' requires an argument"), optopt); + exit(EXIT_FAILURE); + case '?': + vshError(ctl, _("unsupported option '-%c'. See --help."), optopt); + exit(EXIT_FAILURE); default: - vshError(ctl, _("unsupported option '-%c'. See --help."), arg); + vshError(ctl, _("unknown option")); exit(EXIT_FAILURE); } } -- 1.7.11.2

2 3

[libvirt] [PATCH 0/2] re-sort libvirt_private.syms
by Eric Blake 20 Feb '13

20 Feb '13

After Dan's recent round of file renames, our symbols file was not up-to-date. This fixes it. It looks big, but the second patch is mechanical (took me about 2 minutes); most of my time was spent on the first patch getting ready for the sort. Eric Blake (2): maint: fix header file owners of private symbols maint: sort private syms to reflect recent header renames src/libvirt_private.syms | 2046 +++++++++++++++++++++++----------------------- 1 file changed, 1025 insertions(+), 1021 deletions(-) -- 1.8.1.2

2 4

[libvirt] [PATCH v5 0/3] net: support set source address(es) and ports for NAT
by Natanael Copa 20 Feb '13

20 Feb '13

Fixes the commented nitpicks[1]. I also provide an optional 3rd patch, that uses structs for address and port ranges. There was a struct for DHCP address range that could be used after a rename. I put the structs in virsocketaddr.h so they are accessible from util/viriptables. I don't think it is worth its own headerfile and sockaddr was what I thought was most relevant. Skip the third patch if unsure. Please CC me as i don't subscribe to the list. Changes v5: - Remove unused "longdef" leftover from rebase - UseCamelCaseVariableNames - Use virAsprintf instead of snprintf and avoid strings on stack - Add an optional 3rd patch that uses structs for addr and port ranges. Natanael Copa (3): net: support set public ip range for forward mode nat net: add support for specifying port range for forward mode nat net: use structs for address and port ranges docs/formatnetwork.html.in | 33 +++++++ src/conf/network_conf.c | 206 +++++++++++++++++++++++++++++++++++++++++--- src/conf/network_conf.h | 13 ++- src/network/bridge_driver.c | 16 ++++ src/util/viriptables.c | 83 ++++++++++++++++-- src/util/viriptables.h | 4 + src/util/virsocketaddr.h | 14 +++ 7 files changed, 341 insertions(+), 28 deletions(-) [1] http://www.redhat.com/archives/libvir-list/2013-February/msg00817.html -- 1.8.1.3

2 6

[libvirt] Online resize of virtio-blk device does not emit udev event
by Milos Vyletel 20 Feb '13

20 Feb '13

Hi, I hope this is the right list to ask. If not please tell me what will be more appropriate place to ask this question. I'm doing some work with online resizing of guest's block device on LVM. What I do is call lvresize to expand logical volume itself and virsh blockresize to notify guest about this change. Besides the fact that I needed to patch kernel to get this working on mounted partitions(BZ 906050) everything works just fine. Now that I've had this part successfully tested I wanted to make a script that will be triggered by udev whenever particular device is changed. I assumed that after capacity change virtio_blk driver would emit udev event but that's not the case. I'm able to see one on host but not on guest. [root@host ~]# lvresize -f -L +100M /dev/vgguests/evd2 && \ > virsh blockresize resize vdb 1 Rounding size to boundary between physical extents: 128.00 MiB Extending logical volume evd2 to 15.00 GiB Logical volume evd2 successfully resized Block device 'vdb' is resized [root@host ~]# udevadm monitor ... KERNEL[1361292279.619233] change /devices/virtual/block/dm-9 (block) UDEV [1361292279.736012] change /devices/virtual/block/dm-9 (block) ... I was looking at the virtblk_config_changed_work function in RHEL6.3 kernel's drivers/block/virtio_blk.c which I believe is the function handling blockresize and it does not look like it tries to emit any kobject uevent. Before I jump into patching kernel my question is whether it makes sense to have such uevent? I surely can use a way how to detect capacity change from userspace. Thanks, Milos

2 2

[libvirt] [PATCH] qemu: Use atomic ops for driver->nactive
by Jiri Denemark 20 Feb '13

20 Feb '13

--- src/qemu/qemu_conf.h | 2 +- src/qemu/qemu_process.c | 13 +++++-------- 2 files changed, 6 insertions(+), 9 deletions(-) diff --git a/src/qemu/qemu_conf.h b/src/qemu/qemu_conf.h index b5a3281..a6bd527 100644 --- a/src/qemu/qemu_conf.h +++ b/src/qemu/qemu_conf.h @@ -168,7 +168,7 @@ struct _virQEMUDriver { virCgroupPtr cgroup; /* Atomic inc/dec only */ - size_t nactive; + unsigned int nactive; /* Immutable pointers. Caller must provide locking */ virStateInhibitCallback inhibitCallback; diff --git a/src/qemu/qemu_process.c b/src/qemu/qemu_process.c index 74406a6..b612d72 100644 --- a/src/qemu/qemu_process.c +++ b/src/qemu/qemu_process.c @@ -69,6 +69,7 @@ #include "virtime.h" #include "virnetdevtap.h" #include "virbitmap.h" +#include "viratomic.h" #define VIR_FROM_THIS VIR_FROM_QEMU @@ -3300,9 +3301,8 @@ qemuProcessReconnect(void *opaque) goto error; } - if (!driver->nactive && driver->inhibitCallback) + if (virAtomicIntInc(&driver->nactive) == 1 && driver->inhibitCallback) driver->inhibitCallback(true, driver->inhibitOpaque); - driver->nactive++; endjob: if (!qemuDomainObjEndJob(driver, obj)) @@ -3589,9 +3589,8 @@ int qemuProcessStart(virConnectPtr conn, qemuDomainSetFakeReboot(driver, vm, false); virDomainObjSetState(vm, VIR_DOMAIN_SHUTOFF, VIR_DOMAIN_SHUTOFF_UNKNOWN); - if (!driver->nactive && driver->inhibitCallback) + if (virAtomicIntInc(&driver->nactive) == 1 && driver->inhibitCallback) driver->inhibitCallback(true, driver->inhibitOpaque); - driver->nactive++; /* Run an early hook to set-up missing devices */ if (virHookPresent(VIR_HOOK_DRIVER_QEMU)) { @@ -4188,8 +4187,7 @@ void qemuProcessStop(virQEMUDriverPtr driver, */ vm->def->id = -1; - driver->nactive--; - if (!driver->nactive && driver->inhibitCallback) + if (virAtomicIntDecAndTest(&driver->nactive) && driver->inhibitCallback) driver->inhibitCallback(false, driver->inhibitOpaque); if ((logfile = qemuDomainCreateLog(driver, vm, true)) < 0) { @@ -4443,9 +4441,8 @@ int qemuProcessAttach(virConnectPtr conn ATTRIBUTE_UNUSED, vm->def->id = qemuDriverAllocateID(driver); - if (!driver->nactive && driver->inhibitCallback) + if (virAtomicIntInc(&driver->nactive) == 1 && driver->inhibitCallback) driver->inhibitCallback(true, driver->inhibitOpaque); - driver->nactive++; if (virFileMakePath(cfg->logDir) < 0) { virReportSystemError(errno, -- 1.8.1.2

3 4

[libvirt] [PATCH] rpc: Avoid deadlock when closing client connection
by Jiri Denemark 20 Feb '13

20 Feb '13

We need to drop the server lock before calling virObjectUnlock(client) since in case we had the last reference to the client, its dispose callback would be called and that could possibly try to lock the server and cause a deadlock. This is exactly what happens when there is only one QEMU domain running and it is marked to be autodestroyed when the connection dies. This results in qemuProcessAutoDestroy -> qemuProcessStop -> virNetServerRemoveShutdownInhibition call sequence, where the last function locks the server. --- src/rpc/virnetserver.c | 6 +++++- 1 file changed, 5 insertions(+), 1 deletion(-) diff --git a/src/rpc/virnetserver.c b/src/rpc/virnetserver.c index 95333d0..e536cc3 100644 --- a/src/rpc/virnetserver.c +++ b/src/rpc/virnetserver.c @@ -1120,7 +1120,7 @@ void virNetServerRun(virNetServerPtr srv) if (virNetServerClientWantClose(srv->clients[i])) virNetServerClientClose(srv->clients[i]); if (virNetServerClientIsClosed(srv->clients[i])) { - virObjectUnref(srv->clients[i]); + virNetServerClientPtr client = srv->clients[i]; if (srv->nclients > 1) { memmove(srv->clients + i, srv->clients + i + 1, @@ -1131,6 +1131,10 @@ void virNetServerRun(virNetServerPtr srv) srv->nclients = 0; } + virObjectUnlock(srv); + virObjectUnref(client); + virObjectLock(srv); + goto reprocess; } } -- 1.8.1.2

2 2

[libvirt] [PATCH] Avoid resetting errors in virTypedParamsFree
by Jiri Denemark 20 Feb '13

20 Feb '13

The function does not report any errors so there should be no need too reset an existing error first. Moreover, virTypedParamsFree is mostly called in cleanup phase where it has the potential to reset any useful reported earlier. --- src/util/virtypedparam.c | 1 - 1 file changed, 1 deletion(-) diff --git a/src/util/virtypedparam.c b/src/util/virtypedparam.c index ae2855a..c196321 100644 --- a/src/util/virtypedparam.c +++ b/src/util/virtypedparam.c @@ -1026,7 +1026,6 @@ void virTypedParamsFree(virTypedParameterPtr params, int nparams) { - virResetLastError(); virTypedParamsClear(params, nparams); VIR_FREE(params); } -- 1.8.1.2

2 2

[libvirt] [PATCH] virsh: Always print capacity unit
by Jiri Denemark 20 Feb '13

20 Feb '13

It doesn't make sense to print the unit (B) only with Ki, Mi, ... prefixes. Even those poor bytes under 1 KiB are still bytes :-) --- tools/virsh.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/tools/virsh.c b/tools/virsh.c index f5a01b3..d64e539 100644 --- a/tools/virsh.c +++ b/tools/virsh.c @@ -148,7 +148,7 @@ double vshPrettyCapacity(unsigned long long val, const char **unit) { if (val < 1024) { - *unit = ""; + *unit = "B"; return (double)val; } else if (val < (1024.0l * 1024.0l)) { *unit = "KiB"; -- 1.8.1.2

2 2

[libvirt] [PATCH] qemu: Run lzop with '--ignore-warn'
by Michal Privoznik 20 Feb '13

20 Feb '13

Currently, if lzop decompression binary produces a warning, it doesn't exit with zero status but 2 instead. Terrifying, but true. However, warnings may be ignored using '--ignore-warn' command line argument. Moreover, in which case, the exit status will be zero. --- src/qemu/qemu_driver.c | 68 ++++++++++++++++++++++++++++++++++---------------- 1 file changed, 46 insertions(+), 22 deletions(-) diff --git a/src/qemu/qemu_driver.c b/src/qemu/qemu_driver.c index dc35b91..41c6168 100644 --- a/src/qemu/qemu_driver.c +++ b/src/qemu/qemu_driver.c @@ -2507,6 +2507,36 @@ qemuCompressProgramName(int compress) qemuSaveCompressionTypeToString(compress)); } +static virCommandPtr +qemuCompressGetCommand(virQEMUSaveFormat compression) +{ + virCommandPtr ret = NULL; + const char *prog = qemuSaveCompressionTypeToString(compression); + + if (!prog) { + virReportError(VIR_ERR_OPERATION_FAILED, + _("Invalid compressed save format %d"), + compression); + return NULL; + } + + if (!(ret = virCommandNew(prog))) { + virReportOOMError(); + return NULL; + } + + switch (compression) { + case QEMU_SAVE_FORMAT_LZOP: + virCommandAddArg(ret, "--ignore-warn"); + break; + default: + /* Don't add nothing for now */ + break; + } + + return ret; +} + /* Internal function to properly create or open existing files, with * ownership affected by qemu driver setup. */ static int @@ -4775,32 +4805,26 @@ qemuDomainSaveImageStartVM(virConnectPtr conn, if (!(caps = virQEMUDriverGetCapabilities(driver, false))) goto cleanup; - if (header->version == 2) { - const char *prog = qemuSaveCompressionTypeToString(header->compressed); - if (prog == NULL) { - virReportError(VIR_ERR_OPERATION_FAILED, - _("Invalid compressed save format %d"), - header->compressed); + if ((header->version == 2) && + (header->compressed != QEMU_SAVE_FORMAT_RAW)) { + if (!(cmd = qemuCompressGetCommand(header->compressed))) goto cleanup; - } - if (header->compressed != QEMU_SAVE_FORMAT_RAW) { - cmd = virCommandNewArgList(prog, "-dc", NULL); - intermediatefd = *fd; - *fd = -1; + intermediatefd = *fd; + *fd = -1; - virCommandSetInputFD(cmd, intermediatefd); - virCommandSetOutputFD(cmd, fd); - virCommandSetErrorBuffer(cmd, &errbuf); - virCommandDoAsyncIO(cmd); + virCommandAddArg(cmd, "-dc"); + virCommandSetInputFD(cmd, intermediatefd); + virCommandSetOutputFD(cmd, fd); + virCommandSetErrorBuffer(cmd, &errbuf); + virCommandDoAsyncIO(cmd); - if (virCommandRunAsync(cmd, NULL) < 0) { - virReportError(VIR_ERR_INTERNAL_ERROR, - _("Failed to start decompression binary %s"), - prog); - *fd = intermediatefd; - goto cleanup; - } + if (virCommandRunAsync(cmd, NULL) < 0) { + virReportError(VIR_ERR_INTERNAL_ERROR, + _("Failed to start decompression binary %s"), + qemuCompressProgramName(header->compressed)); + *fd = intermediatefd; + goto cleanup; } } -- 1.8.1.2

2 1

[libvirt] [PATCH 0/4] Finish QEMU locking refactoring
by Daniel P. Berrange 20 Feb '13

20 Feb '13

This is the last set of patches for the QEMU driver locking refactor. The interesting one here is the last patch which actually deletes the QEMU driver lock usage from almost everywhere, except a few helper APIs in qemu_conf.c And there was much rejoicing (and hopefully no race conditions and crashes) across the world ;-P

4 11

[libvirt] libvirt support for LXC
by Kunal Kushwaha 19 Feb '13

19 Feb '13

Hi, I am new to this community and am not much aware of current status libvirt work for LXC. Currently I am evaluating LXC with libvirt. and I found still many features like checkpoint/restore/live migration , Attach storage device etc are not implemented. Is there any work going on these feature? or any plan or Roadmap? Following is a table showing libvirt functionality support for LXC. Please feel free to comment in case, I am wrong or tested not propely the libvirt and marked some working feature as not working for LXC. Sr. No. Command Name Description (brief) Status (working or not) 01 autostart Enable and disable the automatic starting of a guest domain when the libvirt daemon starts. ₒ 02 Console connect to the virtual serial console for the guest. ₒ 03 create creates a guest domain from an xml file # does not work. gives error. - 04 Define defines but doesnt start a guest domain from an xml file ○ 05 destroy immediately terminates a running guest domain and releases any resources in use by it. Not to be used lightly ○ 06 attach-disk / detach-disk attach or detach a disk device Note : Doesnt work. maybe i am doing something wrong. - 07 domjobinfo info about running guest domain jobs - 08 domjobabort aborts current running guest domain job - 09 Dump Core dump a guest domain - 10 managedsave / managedsave-remove Save and destroy a running guest domain, so it can be restarted from the same state at a later time. When the virsh start command is next run for the guest domain, it will automatically be started from this saved state - 11 Migrate ** migrates a guest domain to another host 12 Reboot Runs a reboot in the guest domain - 13 Restore restores a guest domain - 14 Resume resumes a guest domain ○ 15 Save saves the running state of a guest domain to a file. - 16 shutdown runs shutdown in a guest domain - 17 Start starts a guest domain either form a last managedsave or via a fresh boot if no managed save is present ○ 18 Update-device update device from an XML file - 19 virt-viewer display a vnc type interface to running domain ○ 20 maxvcpus show maximum number of virtual CPUs for guest domains on this connection - 21 memtune gets or sets the current memory parameters for a guest domain ○ 22 schedinfo show or set scheduler parameters ○ 23 setmaxmem change the maximum memory allocation limit in the guest domain ○ 24 setmem change the current memory allocation in the guest domain ○ 25 setvcpus change the number of virtual cpus in the guest domain - 26 vcpuinfo returns basic info about the guest domains virtual CPU's - 27 dominfo displays some basic information about the domain specified ○ 28 dommemstat displays info about the memory statistics of the domain - 29 domblkinfo gets the block device size or a guest domain ○ 30 domblkstat gets device block stats for a running guest domain ○ 31 domifstat gets network interface stat for a running guest domain ○ 32 dominfo gets basic information about a guest domain ○ 33 dommemstat gets memory statistics for a running guest domain ○ 34 domstate returns the state of a guest domain i.e. whether running etc ○ 35 list returns a list of guest domains ○ 36 Capabilities returns the capabilities of the host and guest in xml format ○ 37 freecell Display available free memory for a NUMA cell ○ 38 hostname Display the name of the hypervisor host ○ 39 nodeinfo Returns basic information about the node ○ 40 iface-define define a physical host network interface. The file expected is of the type xml. We have an example of how it is executed. ○ 41 iface-destroy shutsdown and siables a physical host network device ○ 42 iface-dumpxml Output information for a physical host network interface, as an XML dump to stdout ○ 43 iface-list Returns a list of physical host network interfaces ○ 44 iface-start Enables and starts a physical host network interface ○ 45 iface-undefine Removes the configuration information for a physical host network interface ○ 46 net-autostart Enable or disable the automatic starting of a virtual network, when the libvirt daemon starts ○ 47 net-define Adds a new permanent virtual network from an XML file, without starting it ○ 48 net-dumpxml Displays the XML configuration for a virtual network (to stdout) ○ 49 net-edit opens the network config xml for editing by user. ○ 50 net-info displays basic information about a network ○ 51 net-start starts a network ○ 52 net-undefine undefines a network from the configuration file. ○ 53 Domain snapshot commands (multiple) Does not work with LXC. - Thanks & Regards, Kunal Kushwaha

3 4

[libvirt] [PATCHv3 0/3] S390: Finish support for native CCW bus
by Viktor Mihajlovski 19 Feb '13

19 Feb '13

Originally, QEMU did not implement a native I/O bus for s390. The initial implementation had a machine type 's390-virtio' featuring a fully paravirtualized I/O system with an artificial bus type 'virtio-s390'. This bus had a number of short-comings, like the need for a non-standard device discovery mechanism, a limited number of devices, limited hotplugging capabilites and the lack of persistent device addresses. To resolve these issues a new machine type 's390-ccw-virtio' has been recently added to QEMU which implementa the native s390 CCW I/O bus. Guests with arch='s390x' and machine='s390-ccw-virtio' can now be defined with up to 262144 virtio devices. This series adds the support for the s390-ccw-virtio machine type and the CCW bus to libvirt. V2 Changes - use an attribute triple cssid, ssid, schid instead of devno to better represent the libvirt CCW address structure - rebase to current upstream, mainly address qemuCapsXXX rename V3 Changes - skip already acked patches 1/5 and 2/5 - revert machine-based capabilities - use machine definition from domain object J.B. Joret (1): S390: Add hotplug support for s390 virtio devices Viktor Mihajlovski (2): S390: QEMU driver support for CCW addresses S390: Testcases for virtio-ccw machines src/qemu/qemu_capabilities.c | 7 +- src/qemu/qemu_capabilities.h | 1 + src/qemu/qemu_command.c | 280 ++++++++++++++++++-- src/qemu/qemu_command.h | 6 + src/qemu/qemu_domain.c | 1 + src/qemu/qemu_domain.h | 3 + src/qemu/qemu_driver.c | 6 +- src/qemu/qemu_hotplug.c | 154 +++++++---- src/qemu/qemu_hotplug.h | 14 +- src/qemu/qemu_process.c | 3 + .../qemuxml2argv-console-virtio-ccw.args | 10 + .../qemuxml2argv-console-virtio-ccw.xml | 27 ++ .../qemuxml2argv-console-virtio-s390.args | 4 +- .../qemuxml2argv-console-virtio-s390.xml | 2 +- .../qemuxml2argv-disk-virtio-ccw-many.args | 12 + .../qemuxml2argv-disk-virtio-ccw-many.xml | 40 +++ .../qemuxml2argv-disk-virtio-ccw.args | 8 + .../qemuxml2argv-disk-virtio-ccw.xml | 30 +++ .../qemuxml2argv-net-virtio-ccw.args | 8 + .../qemuxml2argv-net-virtio-ccw.xml | 30 +++ tests/qemuxml2argvtest.c | 12 +- tests/testutilsqemu.c | 3 +- 22 files changed, 578 insertions(+), 83 deletions(-) create mode 100644 tests/qemuxml2argvdata/qemuxml2argv-console-virtio-ccw.args create mode 100644 tests/qemuxml2argvdata/qemuxml2argv-console-virtio-ccw.xml create mode 100644 tests/qemuxml2argvdata/qemuxml2argv-disk-virtio-ccw-many.args create mode 100644 tests/qemuxml2argvdata/qemuxml2argv-disk-virtio-ccw-many.xml create mode 100644 tests/qemuxml2argvdata/qemuxml2argv-disk-virtio-ccw.args create mode 100644 tests/qemuxml2argvdata/qemuxml2argv-disk-virtio-ccw.xml create mode 100644 tests/qemuxml2argvdata/qemuxml2argv-net-virtio-ccw.args create mode 100644 tests/qemuxml2argvdata/qemuxml2argv-net-virtio-ccw.xml -- 1.7.9.5

1 4

[libvirt] [test-API][PATCH v2] Add 7 memory API related cases
by Wayne Sun 19 Feb '13

19 Feb '13

v1: add 7 new cases using domain memory related API add 1 conf for domain memory testing v2: fix the file not close problem and a typo 7 new cases are: memory_params_config: test set memory params with config flag memory_params_live: test set memory params with live flag memory_peek: test memory peek memory_stats: test get memory stats set_maxmem_config: test set maximum memory with config flag set_memory_config: test set current memory with config flag set_memory_live: test set current memory with live flag memory hotplug is not supported yet, so live set max memory case is not added. Signed-off-by: Wayne Sun <gsun(a)redhat.com> --- cases/domain_memory_test.conf | 99 +++++++++++++++++++++++++++++++ repos/domain/memory_params_config.py | 96 ++++++++++++++++++++++++++++++ repos/domain/memory_params_live.py | 112 +++++++++++++++++++++++++++++++++++ repos/domain/memory_peek.py | 48 +++++++++++++++ repos/domain/memory_stats.py | 65 ++++++++++++++++++++ repos/domain/set_maxmem_config.py | 59 ++++++++++++++++++ repos/domain/set_memory_config.py | 94 +++++++++++++++++++++++++++++ repos/domain/set_memory_live.py | 86 +++++++++++++++++++++++++++ 8 files changed, 659 insertions(+) create mode 100644 cases/domain_memory_test.conf create mode 100644 repos/domain/memory_params_config.py create mode 100644 repos/domain/memory_params_live.py create mode 100644 repos/domain/memory_peek.py create mode 100644 repos/domain/memory_stats.py create mode 100644 repos/domain/set_maxmem_config.py create mode 100644 repos/domain/set_memory_config.py create mode 100644 repos/domain/set_memory_live.py diff --git a/cases/domain_memory_test.conf b/cases/domain_memory_test.conf new file mode 100644 index 0000000..90879ab --- /dev/null +++ b/cases/domain_memory_test.conf @@ -0,0 +1,99 @@ +domain:install_linux_cdrom + guestname + $defaultname + guestos + $defaultos + guestarch + $defaultarch + vcpu + $defaultvcpu + memory + $defaultmem + hddriver + $defaulthd + nicdriver + $defaultnic + macaddr + 54:52:00:4a:c1:22 + +domain:balloon_memory + guestname + $defaultname + memorypair + 1024,2048 + +domain:destroy + guestname + $defaultname + +domain:memory_params_config + guestname + $defaultname + hard_limit + 0 + soft_limit + 9007199254740991 + swap_hard_limit + -1 + +domain:set_maxmem_config + guestname + $defaultname + memory + 16777216 + +domain:set_memory_config + guestname + $defaultname + memory + 1048576 + maxmem + 4194304 + +domain:start + guestname + $defaultname + +domain:memory_stats + guestname + $defaultname + +domain:memory_peek + guestname + $defaultname + +domain:memory_params_live + guestname + $defaultname + hard_limit + 25417224 + soft_limit + 9007199254740900 + swap_hard_limit + -1 + +domain:set_memory_live + guestname + $defaultname + memory + 2097152 + username + $username + password + $password + +domain:set_memory_config + guestname + $defaultname + memory + 4194304 + +domain:destroy + guestname + $defaultname + +domain:undefine + guestname + $defaultname + +options cleanup=enable diff --git a/repos/domain/memory_params_config.py b/repos/domain/memory_params_config.py new file mode 100644 index 0000000..af9781b --- /dev/null +++ b/repos/domain/memory_params_config.py @@ -0,0 +1,96 @@ +#!/usr/bin/env python +# Test set domain memory parameters with flag +# VIR_DOMAIN_AFFECT_CONFIG + +from xml.dom import minidom + +import libvirt +from libvirt import libvirtError + +from src import sharedmod + +required_params = ('guestname', 'hard_limit', 'soft_limit', 'swap_hard_limit', ) +optional_params = {} + +UNLIMITED = 9007199254740991 + +def get_memory_config(domobj, param_dict): + """get domain config memory parameters + """ + new_dict = {} + try: + guestxml = domobj.XMLDesc(2) + logger.debug("domain %s xml is :\n%s" %(domobj.name(), guestxml)) + xml = minidom.parseString(guestxml) + logger.info("get domain memory parameters in config xml") + for i in param_dict.keys(): + if xml.getElementsByTagName(i): + limit_element = xml.getElementsByTagName(i)[0] + limit = int(limit_element.childNodes[0].data) + logger.info("%s in config xml is: %s" % (i, limit)) + new_dict[i] = limit + else: + logger.info("%s is not in config xml" % i) + new_dict[i] = 0 + + except libvirtError, e: + logger.error("libvirt call failed: " + str(e)) + return False + + return new_dict + +def memory_params_config(params): + """set domain memory parameters with config flag and check + """ + global logger + logger = params['logger'] + guestname = params['guestname'] + hard_limit = int(params['hard_limit']) + soft_limit = int(params['soft_limit']) + swap_hard_limit = int(params['swap_hard_limit']) + + logger.info("the name of virtual machine is %s" % guestname) + param_dict = {'hard_limit': hard_limit, + 'soft_limit': soft_limit, + 'swap_hard_limit': swap_hard_limit + } + + for i in param_dict.keys(): + if param_dict[i] == -1: + param_dict[i] = UNLIMITED + + logger.info("the param dict for setting is %s" % param_dict) + + conn = sharedmod.libvirtobj['conn'] + + try: + domobj = conn.lookupByName(guestname) + flags = libvirt.VIR_DOMAIN_AFFECT_CONFIG + logger.info("set %s memory parameters with flag: %s" % + (guestname, flags)) + domobj.setMemoryParameters(param_dict, flags) + logger.info("set memory parameters done") + + logger.info("get %s memory parameters with flag: %s" % + (guestname, flags)) + ret = domobj.memoryParameters(flags) + logger.info("%s memory parameters is %s" % (guestname, ret)) + + if ret == param_dict: + logger.info("memory parameters is as expected") + else: + logger.error("memory parameters is not as expected") + return 1 + + ret = get_memory_config(domobj, param_dict) + if ret == param_dict: + logger.info("memory parameters is as expected in config xml") + else: + logger.error("memory parameters is not as expected in config xml") + return 1 + + except libvirtError, e: + logger.error("libvirt call failed: " + str(e)) + return 1 + + return 0 diff --git a/repos/domain/memory_params_live.py b/repos/domain/memory_params_live.py new file mode 100644 index 0000000..44fb8b4 --- /dev/null +++ b/repos/domain/memory_params_live.py @@ -0,0 +1,112 @@ +#!/usr/bin/env python +# Test set domain memory parameters with flag +# VIR_DOMAIN_AFFECT_LIVE + +import os +import math +from xml.dom import minidom + +import libvirt +from libvirt import libvirtError + +from src import sharedmod + +required_params = ('guestname', 'hard_limit', 'soft_limit', 'swap_hard_limit', ) +optional_params = {} + +UNLIMITED = 9007199254740991 +CGROUP_PATH = "/cgroup/memory/libvirt/qemu" + +def get_cgroup_setting(guestname): + """get domain memory parameters in cgroup + """ + if os.path.exists(CGROUP_PATH): + cgroup_path = "%s/%s" % (CGROUP_PATH, guestname) + else: + cgroup_path = "/sys/fs%s/%s" % (CGROUP_PATH, guestname) + + f = open("%s/memory.limit_in_bytes" % cgroup_path) + hard = int(f.read()) + logger.info("memory.limit_in_bytes value is %s" % hard) + f.close() + + f = open("%s/memory.soft_limit_in_bytes" % cgroup_path) + soft = int(f.read()) + logger.info("memory.soft_limit_in_bytes value is %s" % soft) + f.close() + + f = open("%s/memory.memsw.limit_in_bytes" % cgroup_path) + swap = int(f.read()) + logger.info("memory.memsw.limit_in_bytes value is %s" % swap) + f.close() + + new_dict = {'hard_limit': hard/1024, + 'soft_limit': soft/1024, + 'swap_hard_limit': swap/1024 + } + logger.debug("memory parameters dict get from cgroup is %s" % new_dict) + + return new_dict + +def memory_params_live(params): + """set domain memory parameters with live flag and check + """ + global logger + logger = params['logger'] + guestname = params['guestname'] + hard_limit = int(params['hard_limit']) + soft_limit = int(params['soft_limit']) + swap_hard_limit = int(params['swap_hard_limit']) + + logger.info("the name of virtual machine is %s" % guestname) + param_dict = {'hard_limit': hard_limit, + 'soft_limit': soft_limit, + 'swap_hard_limit': swap_hard_limit + } + + for i in param_dict.keys(): + if param_dict[i] == -1: + param_dict[i] = UNLIMITED + + logger.info("the param dict for setting is %s" % param_dict) + + conn = sharedmod.libvirtobj['conn'] + + try: + domobj = conn.lookupByName(guestname) + flags = libvirt.VIR_DOMAIN_AFFECT_LIVE + logger.info("get %s memory parameters with flag: %s" % + (guestname, flags)) + ret_pre = domobj.memoryParameters(flags) + logger.info("%s memory parameters is %s" % (guestname, ret_pre)) + + logger.info("set %s memory parameters with flag: %s" % + (guestname, flags)) + domobj.setMemoryParameters(param_dict, flags) + logger.info("set memory parameters done") + + logger.info("get %s memory parameters with flag: %s" % + (guestname, flags)) + ret_pos = domobj.memoryParameters(flags) + logger.info("%s memory parameters is %s" % (guestname, ret_pos)) + + if ret_pos == param_dict: + logger.info("memory parameters is as expected") + else: + logger.error("memory parameters is not as expected") + return 1 + + logger.info("check memory parameters in cgroup") + ret = get_cgroup_setting(guestname) + for i in param_dict.keys(): + if math.fabs(param_dict[i] - ret[i]) > 1: + logger.error("%s value not match with cgroup setting" % i) + return 1 + + logger.info("memory parameters is as expected in cgroup setting") + + except libvirtError, e: + logger.error("libvirt call failed: " + str(e)) + return 1 + + return 0 diff --git a/repos/domain/memory_peek.py b/repos/domain/memory_peek.py new file mode 100644 index 0000000..de1ff87 --- /dev/null +++ b/repos/domain/memory_peek.py @@ -0,0 +1,48 @@ +#!/usr/bin/env python +# Test domain memory peek + +import libvirt +from libvirt import libvirtError + +from src import sharedmod + +required_params = ('guestname', ) +optional_params = {} + +def memory_peek(params): + """domain memory peek + """ + logger = params['logger'] + guestname = params['guestname'] + + flag_dict = {'1':"VIR_MEMORY_VIRTUAL", '2':"VIR_MEMORY_PHYSICAL"} + + logger.info("the name of virtual machine is %s" % guestname) + + conn = sharedmod.libvirtobj['conn'] + + try: + domobj = conn.lookupByName(guestname) + logger.info("test memory peek API") + for flag in flag_dict.keys(): + logger.info("using flag: %s" % flag_dict[flag]) + mem = domobj.memoryPeek(0, 0, int(flag)) + if mem: + return 1 + logger.info("memory peek API works fine with flag: %s" % + flag_dict[flag]) + + logger.info("peek 8 bytes from domain memory") + for flag in flag_dict.keys(): + logger.info("using flag: %s" % flag_dict[flag]) + mem = domobj.memoryPeek(0, 8, int(flag)) + if not mem: + return 1 + logger.info("8 bytes start with 0 with flag %s is: %s" % + (flag_dict[flag], mem)) + + except libvirtError, e: + logger.error("libvirt call failed: " + str(e)) + return 1 + + return 0 diff --git a/repos/domain/memory_stats.py b/repos/domain/memory_stats.py new file mode 100644 index 0000000..a82adad --- /dev/null +++ b/repos/domain/memory_stats.py @@ -0,0 +1,65 @@ +#!/usr/bin/env python +# Test get domain memory stats + +import libvirt +from libvirt import libvirtError + +from src import sharedmod +from utils import utils + +required_params = ('guestname', ) +optional_params = {} + +VIRSH = "virsh qemu-monitor-command" + +def get_memory_actual(guestname): + """get memory stats with virsh commands + """ + qmp_actual = -1 + cmd ="%s %s '{ \"execute\": \"query-balloon\" }'" % (VIRSH, guestname) + logger.info("check memory stats with virsh command: %s" % cmd) + ret, out = utils.exec_cmd(cmd, shell=True) + out_dict = eval(out[0]) + if out_dict.has_key('return'): + if out_dict['return'].has_key('actual'): + qmp_actual = out_dict['return']['actual'] + else: + return False + + if qmp_actual == -1: + return False + + logger.info("the memory actual is: %s" % qmp_actual) + return qmp_actual + +def memory_stats(params): + """get domain memory stats + """ + global logger + logger = params['logger'] + guestname = params['guestname'] + + logger.info("the name of virtual machine is %s" % guestname) + + conn = sharedmod.libvirtobj['conn'] + + try: + domobj = conn.lookupByName(guestname) + mem = domobj.memoryStats() + logger.info("%s memory stats is: %s" % (guestname, mem)) + ret = get_memory_actual(guestname) + if not ret: + logger.error("get memory actual with qmp command failed") + return 1 + + if ret == mem['actual']*1024: + logger.info("actual memory is equal to output of qmp command") + else: + logger.error("actual memory is not equal to output of qmp command") + return 1 + + except libvirtError, e: + logger.error("libvirt call failed: " + str(e)) + return 1 + + return 0 diff --git a/repos/domain/set_maxmem_config.py b/repos/domain/set_maxmem_config.py new file mode 100644 index 0000000..262d7b1 --- /dev/null +++ b/repos/domain/set_maxmem_config.py @@ -0,0 +1,59 @@ +#!/usr/bin/env python +# Test set domain max memory with API setMaxMemory. + +from xml.dom import minidom + +import libvirt +from libvirt import libvirtError + +from src import sharedmod + +required_params = ('guestname', 'memory', ) +optional_params = {} + +def set_maxmem_config(params): + """set domain max memory, check with config xml and + maxMemory API + """ + global logger + logger = params['logger'] + guestname = params['guestname'] + memory = int(params['memory']) + + logger.info("the name of virtual machine is %s" % guestname) + logger.info("the given max memory value is %s" % memory) + + conn = sharedmod.libvirtobj['conn'] + + try: + domobj = conn.lookupByName(guestname) + logger.info("set domain max memory as %s" % memory) + domobj.setMaxMemory(memory) + + guestxml = domobj.XMLDesc(2) + logger.debug("domain %s xml is :\n%s" %(guestname, guestxml)) + xml = minidom.parseString(guestxml) + mem = xml.getElementsByTagName('memory')[0] + maxmem = int(mem.childNodes[0].data) + logger.info("domain max memory in config xml is: %s" % maxmem) + if maxmem == memory: + logger.info("max memory in domain config xml is equal to set") + else: + logger.error("set max memory failed") + return 1 + + maxmem = domobj.maxMemory() + logger.info("max memory got by maxMemory API is: %s" % maxmem) + if maxmem == memory: + logger.info("max memory got by maxMemory API is equal to set") + else: + logger.error("set max memory failed") + return 1 + + logger.info("set max memory succeed") + + except libvirtError, e: + logger.error("libvirt call failed: " + str(e)) + return 1 + + return 0 diff --git a/repos/domain/set_memory_config.py b/repos/domain/set_memory_config.py new file mode 100644 index 0000000..c8ef6c3 --- /dev/null +++ b/repos/domain/set_memory_config.py @@ -0,0 +1,94 @@ +#!/usr/bin/env python +# Test set domain balloon memory with flag VIR_DOMAIN_AFFECT_CONFIG +# or VIR_DOMAIN_VCPU_MAXIMUM, depend on which optional param is +# given. + +from xml.dom import minidom + +import libvirt +from libvirt import libvirtError + +from src import sharedmod + +required_params = ('guestname', ) +optional_params = {'memory': 1048576, + 'maxmem': 4194304, + } +def get_memory_config(domobj): + """get domain config current memory and max memory + """ + try: + guestxml = domobj.XMLDesc(2) + logger.debug("domain %s xml is :\n%s" %(domobj.name(), guestxml)) + xml = minidom.parseString(guestxml) + + logger.info("get domain memory info in config xml") + mem = xml.getElementsByTagName('currentMemory')[0] + current = int(mem.childNodes[0].data) + logger.info("current memory in config xml is: %s" % current) + + mem = xml.getElementsByTagName('memory')[0] + max_memory = int(mem.childNodes[0].data) + logger.info("max memory in config xml is: %s" % max_memory) + + except libvirtError, e: + logger.error("libvirt call failed: " + str(e)) + return False + + return current, max_memory + +def set_memory_config(params): + """set domain memory with live flag and check + """ + global logger + logger = params['logger'] + guestname = params['guestname'] + memory = params.get('memory', None) + maxmem = params.get('maxmem', None) + + logger.info("the name of virtual machine is %s" % guestname) + if memory == None and maxmem == None: + logger.error("at least one of memory or maxmem should be provided") + return 1 + + conn = sharedmod.libvirtobj['conn'] + + try: + domobj = conn.lookupByName(guestname) + if memory: + memory = int(memory) + flags = libvirt.VIR_DOMAIN_AFFECT_CONFIG + logger.info("set domain memory as %s with flag: %s" % + (memory, flags)) + domobj.setMemoryFlags(memory, flags) + ret = get_memory_config(domobj) + if not ret: + return 1 + + if ret[0] == memory: + logger.info("set current memory succeed") + else: + logger.error("set current memory failed") + return 1 + + if maxmem: + maxmem = int(maxmem) + flags = libvirt.VIR_DOMAIN_MEM_MAXIMUM + logger.info("set domain max memory as %s with flag: %s" % + (maxmem, flags)) + domobj.setMemoryFlags(maxmem, flags) + ret = get_memory_config(domobj) + if not ret: + return 1 + + if ret[1] == maxmem: + logger.info("set max memory succeed") + else: + logger.error("set max memory failed") + return 1 + + except libvirtError, e: + logger.error("libvirt call failed: " + str(e)) + return 1 + + return 0 diff --git a/repos/domain/set_memory_live.py b/repos/domain/set_memory_live.py new file mode 100644 index 0000000..f7c77f9 --- /dev/null +++ b/repos/domain/set_memory_live.py @@ -0,0 +1,86 @@ +#!/usr/bin/env python +# Test set domain balloon memory with flag VIR_DOMAIN_AFFECT_LIVE. +# Check domain info and inside domain to get current memory value. +# The live flag only work on running domain, so test on shutoff +# domain will fail. + +import time +import math + +import libvirt +from libvirt import libvirtError + +from src import sharedmod +from utils import utils + +required_params = ('guestname', 'memory', 'username', 'password', ) +optional_params = {} + +def compare_memory(expect_memory, actual_memory): + """ comparing expected memory size with actual memory size """ + + logger.info("expected memory size is %s" % expect_memory) + logger.info("actual memory size is %s" % actual_memory) + diff = int(expect_memory) - int(actual_memory) + + if math.fabs(diff)/expect_memory < 0.05: + return 0 + else: + return 1 + +def get_current_memory(guestname, username, password): + """get domain current memory inside domain + """ + logger.debug("get the mac address of vm %s" % guestname) + mac = utils.get_dom_mac_addr(guestname) + logger.debug("the mac address of vm %s is %s" % (guestname, mac)) + ip = utils.mac_to_ip(mac, 180) + current = utils.get_remote_memory(ip, username, password) + + return current + +def set_memory_live(params): + """set domain memory with live flag and check + """ + global logger + logger = params['logger'] + params.pop('logger') + guestname = params['guestname'] + memory = int(params['memory']) + username = params['username'] + password = params['password'] + + logger.info("the name of virtual machine is %s" % guestname) + logger.info("the given memory value is %s" % memory) + + conn = sharedmod.libvirtobj['conn'] + + try: + domobj = conn.lookupByName(guestname) + logger.info("set domain memory as %s with flag: %s" % + (memory, libvirt.VIR_DOMAIN_AFFECT_LIVE)) + domobj.setMemoryFlags(memory, libvirt.VIR_DOMAIN_AFFECT_LIVE) + logger.info("get domain current memory") + time.sleep(3) + dominfo = domobj.info() + logger.debug("domain info list is: %s" % dominfo) + logger.info("domain current memory value is: %s KiB" % dominfo[2]) + if memory == dominfo[2]: + logger.info("set memory match with domain info") + else: + logger.error("set memory not match with domain info") + return 1 + + logger.info("check domain memory inside domain") + ret = get_current_memory(guestname, username, password) + if not compare_memory(memory, ret): + logger.info("set domain memory succeed") + else: + logger.error("set domain memory failed") + return 1 + + except libvirtError, e: + logger.error("libvirt call failed: " + str(e)) + return 1 + + return 0 -- 1.8.1

2 1

[libvirt] [PATCH v4] add pci-bridge support
by liguang 19 Feb '13

19 Feb '13

Now, it's impossible to arrange devices into multi-pci-bus, for example: <sound model='ac97'> <address type='pci' domain='0x0000' bus='0x00' slot='0x04' function='0x0'/> </sound> <video> <model type='cirrus' vram='9216' heads='1'/> <address type='pci' domain='0x0000' bus='0x1' slot='0x02' function='0x0'/> </video> libvirt will complain about "bus != 0", fortunately, qemu supports pci-to-pci bridge, if we want to use multi-pci-bus, we can define 2 pci bridge controllers, then attach 1 to the other as a subordinate pci-bus, so, 2 pci-buses appear. for example: <controller type='pci-bridge' index='0'/> <controller type='pci-bridge' index='1'> <address type='pci' domain='0x0000' bus='0x00' slot='0x05' function='0x0'/> </controller> <sound model='ac97'> <address type='pci' domain='0x0000' bus='0x01' slot='0x02' function='0x0'/> </sound> <video> <model type='cirrus' vram='9216' heads='1'/> <address type='pci' domain='0x0000' bus='0x00' slot='0x02' function='0x0'/> </video> src/conf/domain_conf.c | 98 ++++- src/conf/domain_conf.h | 1 + docs/schemas/domaincommon.rng | 1 + src/qemu/qemu_capabilities.c | 4 + src/qemu/qemu_capabilities.h | 1 + src/qemu/qemu_command.c | 43 ++++++++++++++++++++----- 6 files changed, 126 insertions(+), 22 deletions(-)

1 4

[libvirt] [PATCH 0/6 v2] Fix problems on shared disk management
by Osier Yang 19 Feb '13

19 Feb '13

This fixes problems on shared disk management, inspired by the crash when changing the media of cdrom and floppy disk, and the source of cdrom and floppy disk is empty. v1 - v2: * v1 uses a bitmap to record whether the disk is added into the table when do qemuProcessStart. v2 added a array to the hash entry to record the names of domain which uses the shared disk. * Update the shared disk table when reconnecting qemu process. Osier Yang (6): qemu: Add checking in helpers for sgio setting qemu: Merge qemuCheckSharedDisk into qemuAddSharedDisk qemu: Record names of domain which uses the shared disk in hash table qemu: Update shared disk table when reconnecting qemu process qemu: Move the shared disk adding and sgio setting prior to attaching qemu: Remove the shared disk entry if the operation is ejecting src/qemu/qemu_conf.c | 187 ++++++++++++++++++++++++++++++++++++++++++---- src/qemu/qemu_conf.h | 22 +++++- src/qemu/qemu_driver.c | 84 ++++++++++++++------- src/qemu/qemu_hotplug.c | 10 ++- src/qemu/qemu_hotplug.h | 3 +- src/qemu/qemu_process.c | 81 +++++--------------- src/qemu/qemu_process.h | 3 - 7 files changed, 276 insertions(+), 114 deletions(-) -- 1.7.7.6

2 12

[libvirt] Fwd: There seems a deadlock in libvirt
by Chun-Hung Chen 19 Feb '13

19 Feb '13

Hi, all, We were running OpenStack with Ubuntu and libvirt 0.9.10. We found that libvirt monitor command not working well. There were a lot of error in libvirtd.log like this 2013-02-07 06:07:39.000+0000: 18112: error : qemuDomainObjBeginJobInternal:773 : Timed out during operation: cannot acquire state change lock We dig into libvirtd by strace and find one of the thread only have the following command futex(0x7f69ac0ec0ec, FUTEX_WAIT_PRIVATE, 2717, NULL It seems this thread waiting for reply but nothing came back thus other threads would wait for it. We also saw there is a function called virCondWaitUntil(). Is it safe for us to modify the code from virCondWait() to virCondWaitUntil() to prevent such deadlock scenario? Thanks. Following is the gdb -p 'libvirt.pid' and 'thread id' and 'bt full' #0 0x00007f69c8c1dd84 in pthread_cond_wait@@GLIBC_2.3.2 () from /lib/x86_64-linux-gnu/libpthread.so.0 No symbol table info available. #1 0x00007f69c9ee884a in virCondWait (c=<optimized out>, m=<optimized out>) at util/threads-pthread.c:117 ret = <optimized out> #2 0x000000000049c749 in qemuMonitorSend (mon=0x7f69ac0ec0c0, msg=<optimized out>) at qemu/qemu_monitor.c:826 ret = -1 __func__ = "qemuMonitorSend" __FUNCTION__ = "qemuMonitorSend" #3 0x00000000004ac8ed in qemuMonitorJSONCommandWithFd (mon=0x7f69ac0ec0c0, cmd=0x7f6998028280, scm_fd=-1, reply=0x7f69c57829f8) at qemu/qemu_monitor_json.c:230 ret = -1 msg = {txFD = -1, txBuffer = 0x7f69980e9b00 "{\"execute\":\"query-balloon\",\"id\":\"libvirt-1359\"}\r\n", txOffset = 49, txLength = 49, rxBuffer = 0x0, rxLength = 0, rxObject = 0x0, finished = false, passwordHandler = 0, passwordOpaque = 0x0} cmdstr = 0x7f69980ef2f0 "{\"execute\":\"query-balloon\",\"id\":\"libvirt-1359\"}" id = 0x7f69980b0a20 "libvirt-1359" exe = <optimized out> __FUNCTION__ = "qemuMonitorJSONCommandWithFd" __func__ = "qemuMonitorJSONCommandWithFd" #4 0x00000000004ae794 in qemuMonitorJSONGetBalloonInfo (mon=0x7f69ac0ec0c0, currmem=0x7f69c5782a48) at qemu/qemu_monitor_json.c:1190 ret = <optimized out> cmd = 0x7f6998028280 reply = 0x0 __FUNCTION__ = "qemuMonitorJSONGetBalloonInfo" #5 0x0000000000457451 in qemudDomainGetInfo (dom=<optimized out>, info=0x7f69c5782b50) at qemu/qemu_driver.c:2181 priv = 0x7f69a0093b00 driver = 0x7f69b80ca8e0 vm = 0x7f69a0093370 ret = -1 err = <optimized out> balloon = <optimized out> __FUNCTION__ = "qemudDomainGetInfo" #6 0x00007f69c9f63eda in virDomainGetInfo (domain=0x7f69980e3650, info=0x7f69c5782b50) at libvirt.c:4230 ret = <optimized out> conn = <optimized out> __func__ = "virDomainGetInfo" __FUNCTION__ = "virDomainGetInfo" #7 0x0000000000439bca in remoteDispatchDomainGetInfo (ret=0x7f6998000c20, args=<optimized out>, rerr=0x7f69c5782c50, client=0x157e730, server=<optimized out>, msg=<optimized out>) at remote_dispatch.h:1640 rv = -1 tmp = {state = 1 '\001', maxMem = 2097152, memory = 0, nrVirtCpu = 0, cpuTime = 5981880000000} dom = 0x7f69980e3650 priv = <optimized out> #8 remoteDispatchDomainGetInfoHelper (server=<optimized out>, client=0x157e730, msg=<optimized out>, rerr=0x7f69c5782c50, args=<optimized out>, ret=0x7f6998000c20) at remote_dispatch.h:1616 __func__ = "remoteDispatchDomainGetInfoHelper" #9 0x00007f69c9fbb915 in virNetServerProgramDispatchCall (msg=0x1689cc0, client=0x157e730, server=0x1577c90, prog=0x15825d0) at rpc/virnetserverprogram.c:416 ret = 0x7f6998000c20 "" rv = -1 i = <optimized out> arg = 0x7f6998027950 "\360e\n\230i\177" dispatcher = 0x73de40 rerr = {code = 0, domain = 0, message = 0x0, level = 0, dom = 0x0, str1 = 0x0, str2 = 0x0, str3 = 0x0, int1 = 0, int2 = 0, net = 0x0} #10 virNetServerProgramDispatch (prog=0x15825d0, server=0x1577c90, client=0x157e730, msg=0x1689cc0) at rpc/virnetserverprogram.c:289 ret = -1 rerr = {code = 0, domain = 0, message = 0x0, level = 0, dom = 0x0, str1 = 0x0, str2 = 0x0, str3 = 0x0, int1 = 0, int2 = 0, net = 0x0} __func__ = "virNetServerProgramDispatch" __FUNCTION__ = "virNetServerProgramDispatch" #11 0x00007f69c9fb6461 in virNetServerHandleJob (jobOpaque=<optimized out>, opaque=0x1577c90) at rpc/virnetserver.c:164 srv = 0x1577c90 job = 0x155dfa0 __func__ = "virNetServerHandleJob" #12 0x00007f69c9ee8e3e in virThreadPoolWorker (opaque=<optimized out>) at util/threadpool.c:144 data = 0x0 pool = 0x1577d80 cond = 0x1577de0 priority = false job = 0x162dd20 #13 0x00007f69c9ee84e6 in virThreadHelper (data=<optimized out>) at util/threads-pthread.c:161 args = 0x0 local = {func = 0x7f69c9ee8d00 <virThreadPoolWorker>, opaque = 0x1559f90} #14 0x00007f69c8c19e9a in start_thread () from /lib/x86_64-linux-gnu/libpthread.so.0 No symbol table info available. #15 0x00007f69c89474bd in clone () from /lib/x86_64-linux-gnu/libc.so.6 No symbol table info available. #16 0x0000000000000000 in ?? () No symbol table info available. Regards, Chun-Hung

2 1

[libvirt] [PATCH V6 0/3] Add support for QEMU file descriptor sets
by Stefan Berger 19 Feb '13

19 Feb '13

The following patch series adds initial support for QEMU file descriptor sets implementing support for creating the proper command line. Some devices are using the sets now. Regards, Stefan

3 15