- Devel - libvirt List Archives

[libvirt] [PATCH 0/2] Support PCI passthrough in libxl driver
by Chunyan Liu 24 Feb '13

24 Feb '13

This patch series is to add PCI passthrough support to libxenlight driver so that it can create VM with managed/non-managed host pci device defined. Chunyan Liu (2): libxl: add APIs to prepare/release host pci device libxl: support pci passthrough po/POTFILES.in | 1 + src/Makefile.am | 1 + src/libxl/libxl_conf.c | 44 ++++ src/libxl/libxl_conf.h | 5 +- src/libxl/libxl_driver.c | 16 ++ src/libxl/libxl_hostdev.c | 600 +++++++++++++++++++++++++++++++++++++++++++++ src/libxl/libxl_hostdev.h | 44 ++++ 7 files changed, 710 insertions(+), 1 deletions(-) create mode 100644 src/libxl/libxl_hostdev.c create mode 100644 src/libxl/libxl_hostdev.h -- 1.7.3.4

6 7

[libvirt] [PATCH] run: license as LGPL
by Eric Blake 24 Feb '13

24 Feb '13

It makes no sense to prohibit reuse of the wrapper in other LGPL projects, since most of libvirt is designed to be LGPL. Of course, when using the wrapper to wrap a GPL program, the combined result is still effectively GPL, but that shouldn't force us to license the wrapper as GPL in isolation. * run.in: Relicense to LGPLv2+. --- Any objections to this? run.in | 19 ++++++++++--------- 1 file changed, 10 insertions(+), 9 deletions(-) diff --git a/run.in b/run.in index 5063522..5d4a04c 100644 --- a/run.in +++ b/run.in @@ -2,18 +2,19 @@ # libvirt 'run' programs locally script # Copyright (C) 2012-2013 Red Hat, Inc. # -# This program is free software; you can redistribute it and/or modify -# it under the terms of the GNU General Public License as published by -# the Free Software Foundation; either version 2 of the License, or -# (at your option) any later version. +# This library is free software; you can redistribute it and/or +# modify it under the terms of the GNU Lesser General Public +# License as published by the Free Software Foundation; either +# version 2.1 of the License, or (at your option) any later version. # -# This program is distributed in the hope that it will be useful, +# This library is distributed in the hope that it will be useful, # but WITHOUT ANY WARRANTY; without even the implied warranty of -# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the -# GNU General Public License for more details. +# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU +# Lesser General Public License for more details. # -# You should have received a copy of the GNU General Public License -# along with this program; If not, see <http://www.gnu.org/licenses/>. +# You should have received a copy of the GNU Lesser General Public +# License along with this library; If not, see +# <http://www.gnu.org/licenses/>. #---------------------------------------------------------------------- # -- 1.8.1.2

2 3

[libvirt] [PATCH v4 0/9] Rework storage migration
by Michal Privoznik 23 Feb '13

23 Feb '13

Since there are many hidden problems with auto-creating storage for the domain in here, as previous rounds of reviewing the series has shown, I've decided to not do anything about it for now and the auto allocation is dropped completely. So we are back to the assumption we already have - users need to make sure files already exists and have the right size. Hawk. 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) diff to v4: -drop storage auto-allocation -include John's and Jirka's reviews Note, that most of the patches has been ACKed already. Michal Privoznik (9): 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_migration: Stop NBD server at Finish phase qemu_migration: Cancel running jobs on failed migration src/qemu/qemu_capabilities.c | 4 +- src/qemu/qemu_capabilities.h | 1 + src/qemu/qemu_domain.h | 1 + src/qemu/qemu_migration.c | 458 +++++++++++++++++++++++++++++++++++++++++-- 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 | 5 + 9 files changed, 631 insertions(+), 17 deletions(-) -- 1.8.1.2

3 21

[libvirt] [PATCH] run: use portable shell
by Eric Blake 23 Feb '13

23 Feb '13

Nothing in run required bash, except for the shebang. On systems where /bin/bash doesn't exist (I hit it on FreeBSD), using /bin/sh instead fixes a 'make check' failure: gmake[3]: Entering directory `/usr/home/dummy/libvirt/python' GEN check-local /usr/local/bin/bash: ../run: /bin/bash: bad interpreter: No such file or directory * run.in: Use /bin/sh, not bash. --- Pushing under the build-breaker rule. run.in | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/run.in b/run.in index 845eef5..5063522 100644 --- a/run.in +++ b/run.in @@ -1,6 +1,6 @@ -#!/bin/bash - +#!/bin/sh # libvirt 'run' programs locally script -# Copyright (C) 2012 Red Hat, Inc. +# Copyright (C) 2012-2013 Red Hat, Inc. # # This program is free software; you can redistribute it and/or modify # it under the terms of the GNU General Public License as published by -- 1.8.1.2

1 0

[libvirt] [PATCH] storage: use f_frsize, not f_bsize, for calculating available space
by Sage Weil 23 Feb '13

23 Feb '13

The bfree and blocks fields are supposed to be in units of frsize. We were calculating capacity correctly using those units, but the available calculation was using bsize instead. Most file systems report these as the same value specifically because many programs are buggy, but that is no reason to rely on that behavior, or to behave inconsistently. This bug has been present since e266ded (2008) and aa296e6c, when the code was originally introduced (the latter via cut and paste). Signed-off-by: Sage Weil <sage(a)newdream.net> --- src/parallels/parallels_storage.c | 2 +- src/storage/storage_backend_fs.c | 2 +- 2 files changed, 2 insertions(+), 2 deletions(-) diff --git a/src/parallels/parallels_storage.c b/src/parallels/parallels_storage.c index 1186296..f696fcb 100644 --- a/src/parallels/parallels_storage.c +++ b/src/parallels/parallels_storage.c @@ -709,7 +709,7 @@ parallelsStoragePoolGetAlloc(virStoragePoolDefPtr def) def->capacity = ((unsigned long long)sb.f_frsize * (unsigned long long)sb.f_blocks); def->available = ((unsigned long long)sb.f_bfree * - (unsigned long long)sb.f_bsize); + (unsigned long long)sb.f_frsize); def->allocation = def->capacity - def->available; return 0; diff --git a/src/storage/storage_backend_fs.c b/src/storage/storage_backend_fs.c index a582804..c1684f7 100644 --- a/src/storage/storage_backend_fs.c +++ b/src/storage/storage_backend_fs.c @@ -910,7 +910,7 @@ virStorageBackendFileSystemRefresh(virConnectPtr conn ATTRIBUTE_UNUSED, pool->def->capacity = ((unsigned long long)sb.f_frsize * (unsigned long long)sb.f_blocks); pool->def->available = ((unsigned long long)sb.f_bfree * - (unsigned long long)sb.f_bsize); + (unsigned long long)sb.f_frsize); pool->def->allocation = pool->def->capacity - pool->def->available; return 0; -- 1.7.2.5

2 1

[libvirt] [PATCH] maint: check all symfiles for sorting
by Eric Blake 23 Feb '13

23 Feb '13

On FreeBSD, I got a 'make check' failure: GEN check-symsorting Symbol block at ./libvirt_atomic.syms:4: viratomic.h not found * src/Makefile.am (SYM_FILES): New define. (check-symsorting): Check on all symfiles, even when not used. * src/libvirt_atomic.syms: Fix offender. --- Pushing under the build-breaker rule. src/Makefile.am | 23 ++++++++++++++++++++++- src/libvirt_atomic.syms | 2 +- 2 files changed, 23 insertions(+), 2 deletions(-) diff --git a/src/Makefile.am b/src/Makefile.am index 3ef9a9c..c1659a4 100644 --- a/src/Makefile.am +++ b/src/Makefile.am @@ -389,7 +389,7 @@ check-symfile: endif check-symsorting: $(AM_V_GEN)$(PERL) $(srcdir)/check-symsorting.pl \ - $(srcdir) $(USED_SYM_FILES) + $(srcdir) $(SYM_FILES) EXTRA_DIST += check-symfile.pl check-symsorting.pl PROTOCOL_STRUCTS = \ @@ -809,6 +809,7 @@ libvirt_driver_la_CFLAGS = \ libvirt_driver_la_LIBADD = \ $(NUMACTL_LIBS) $(GNUTLS_LIBS) $(CURL_LIBS) $(DLOPEN_LIBS) +SYM_FILES = $(USED_SYM_FILES) USED_SYM_FILES = $(srcdir)/libvirt_private.syms if WITH_TEST @@ -1410,42 +1411,62 @@ check-augeas-lockd: test_libvirt_lockd.aug if WITH_DRIVER_MODULES USED_SYM_FILES += $(srcdir)/libvirt_driver_modules.syms +else +SYM_FILES += $(srcdir)/libvirt_driver_modules.syms endif if WITH_LINUX USED_SYM_FILES += $(srcdir)/libvirt_linux.syms +else +SYM_FILES += $(srcdir)/libvirt_linux.syms endif if WITH_ESX USED_SYM_FILES += $(srcdir)/libvirt_esx.syms +else +SYM_FILES += $(srcdir)/libvirt_esx.syms endif if WITH_LIBVIRTD USED_SYM_FILES += $(srcdir)/libvirt_daemon.syms +else +SYM_FILES += $(srcdir)/libvirt_daemon.syms endif if WITH_OPENVZ USED_SYM_FILES += $(srcdir)/libvirt_openvz.syms +else +SYM_FILES += $(srcdir)/libvirt_openvz.syms endif if WITH_VMX USED_SYM_FILES += $(srcdir)/libvirt_vmx.syms +else +SYM_FILES += $(srcdir)/libvirt_vmx.syms endif if WITH_XENXS USED_SYM_FILES += $(srcdir)/libvirt_xenxs.syms +else +SYM_FILES += $(srcdir)/libvirt_xenxs.syms endif if WITH_SASL USED_SYM_FILES += $(srcdir)/libvirt_sasl.syms +else +SYM_FILES += $(srcdir)/libvirt_sasl.syms endif if WITH_SSH2 USED_SYM_FILES += $(srcdir)/libvirt_libssh2.syms +else +SYM_FILES += $(srcdir)/libvirt_libssh2.syms endif if WITH_ATOMIC_OPS_PTHREAD USED_SYM_FILES += $(srcdir)/libvirt_atomic.syms +else +SYM_FILES += $(srcdir)/libvirt_atomic.syms endif EXTRA_DIST += \ diff --git a/src/libvirt_atomic.syms b/src/libvirt_atomic.syms index db80ce8..e2c2363 100644 --- a/src/libvirt_atomic.syms +++ b/src/libvirt_atomic.syms @@ -2,7 +2,7 @@ # These symbols are dependent upon !VIR_ATOMIC_OPS_GCC. # -# viratomic.h +# util/viratomic.h virAtomicLock; # Let emacs know we want case-insensitive sorting -- 1.8.1.2

1 0

[libvirt] [RFC PATCH 1/4] libvirt: add VIR_DOMAIN_START_PERSISTENT flag
by Doug Goldstein 22 Feb '13

22 Feb '13

Added a VIR_DOMAIN_START_PERSISTENT flag for virDomainCreateXML() so that the guest remains defined after it is destroyed. The result of using this flag is equivilent to calling virDomainDefineXML() followed by virDomainCreate() or virDomainCreateWithFlags(). --- Not sure if this is the correct place to add "Since 1.0.3 (likely 1.0.4)" --- include/libvirt/libvirt.h.in | 1 + src/libvirt.c | 4 ++++ 2 files changed, 5 insertions(+) diff --git a/include/libvirt/libvirt.h.in b/include/libvirt/libvirt.h.in index ad30cd0..e119215 100644 --- a/include/libvirt/libvirt.h.in +++ b/include/libvirt/libvirt.h.in @@ -336,6 +336,7 @@ typedef enum { VIR_DOMAIN_START_AUTODESTROY = 1 << 1, /* Automatically kill guest when virConnectPtr is closed */ VIR_DOMAIN_START_BYPASS_CACHE = 1 << 2, /* Avoid file system cache pollution */ VIR_DOMAIN_START_FORCE_BOOT = 1 << 3, /* Boot, discarding any managed save */ + VIR_DOMAIN_START_PERSISTENT = 1 << 4, /* Define guest to exist after it is destroyed */ } virDomainCreateFlags; diff --git a/src/libvirt.c b/src/libvirt.c index 1e78500..abe59cd 100644 --- a/src/libvirt.c +++ b/src/libvirt.c @@ -1951,6 +1951,10 @@ virDomainGetConnect(virDomainPtr dom) * is destroyed, or if the host is restarted (see virDomainDefineXML() to * define persistent domains). * + * If the VIR_DOMAIN_START_PERSISTENT flag is set, the guest domain + * will remain defined even after the guest is destroyed. This is + * similar to if you had called virDomainDefineXML() and virDomainCreate(). + * * If the VIR_DOMAIN_START_PAUSED flag is set, the guest domain * will be started, but its CPUs will remain paused. The CPUs * can later be manually started using virDomainResume. -- 1.7.12.4

2 6

[libvirt] [PATCH] Add support for qxl.revision in domain XML
by Christophe Fergeau 22 Feb '13

22 Feb '13

QXL devices have an associated 'revision' which is raised when new features have been introduced which would break migration to older versions. This commit makes it possible to set this revision as QEMU sometimes support newer QXL revisions than what it defaults to. --- docs/formatdomain.html.in | 5 ++++- docs/schemas/domaincommon.rng | 5 +++++ src/conf/domain_conf.c | 17 +++++++++++++++++ src/conf/domain_conf.h | 1 + src/qemu/qemu_command.c | 8 ++++++++ .../qemuxml2argv-graphics-spice-compression.args | 3 ++- .../qemuxml2argv-graphics-spice-compression.xml | 4 ++-- .../qemuxml2argv-graphics-spice-qxl-vga.args | 3 ++- .../qemuxml2argv-graphics-spice-qxl-vga.xml | 4 ++-- tests/qemuxml2argvdata/qemuxml2argv-graphics-spice.args | 3 ++- tests/qemuxml2argvdata/qemuxml2argv-graphics-spice.xml | 4 ++-- 11 files changed, 47 insertions(+), 10 deletions(-) diff --git a/docs/formatdomain.html.in b/docs/formatdomain.html.in index 7ad8aea..4b269c8 100644 --- a/docs/formatdomain.html.in +++ b/docs/formatdomain.html.in @@ -3569,7 +3569,10 @@ qemu-kvm -net nic,model=? /dev/null attribute <code>ram</code> (<span class="since">since 1.0.2</span>) is allowed for "qxl" type only and specifies the size of the primary bar, while <code>vram</code> specifies the - secondary bar size. If "ram" is not supplied a default value is used. + secondary bar size. If "ram" or "vram" are not supplied a default + value is used. The optional attribute <code>revision</code> (<span + class="since">since 1.0.3</span>) specifies the revision of + the QXL device, newer revisions provides more functionality. </dd> <dt><code>model</code></dt> diff --git a/docs/schemas/domaincommon.rng b/docs/schemas/domaincommon.rng index 049f232..fc78e2d 100644 --- a/docs/schemas/domaincommon.rng +++ b/docs/schemas/domaincommon.rng @@ -2280,6 +2280,11 @@ <ref name="unsignedInt"/> </attribute> </optional> + <optional> + <attribute name="revision"> + <ref name="unsignedInt"/> + </attribute> + </optional> </group> </choice> <optional> diff --git a/src/conf/domain_conf.c b/src/conf/domain_conf.c index 5782abb..83be711 100644 --- a/src/conf/domain_conf.c +++ b/src/conf/domain_conf.c @@ -7391,6 +7391,7 @@ virDomainVideoDefParseXML(const xmlNodePtr node, char *vram = NULL; char *ram = NULL; char *primary = NULL; + char *revision = NULL; if (VIR_ALLOC(def) < 0) { virReportOOMError(); @@ -7406,6 +7407,7 @@ virDomainVideoDefParseXML(const xmlNodePtr node, ram = virXMLPropString(cur, "ram"); vram = virXMLPropString(cur, "vram"); heads = virXMLPropString(cur, "heads"); + revision = virXMLPropString(cur, "revision"); if ((primary = virXMLPropString(cur, "primary")) != NULL) if (STREQ(primary, "yes")) @@ -7456,6 +7458,19 @@ virDomainVideoDefParseXML(const xmlNodePtr node, def->vram = virDomainVideoDefaultRAM(dom, def->type); } + if (revision) { + if (def->type != VIR_DOMAIN_VIDEO_TYPE_QXL) { + virReportError(VIR_ERR_XML_ERROR, "%s", + _("revision attribute only supported for type of qxl")); + goto error; + } + if (virStrToLong_ui(revision, NULL, 10, &def->revision) < 0) { + virReportError(VIR_ERR_XML_ERROR, + _("cannot parse video revision '%s'"), revision); + goto error; + } + } + if (heads) { if (virStrToLong_ui(heads, NULL, 10, &def->heads) < 0) { virReportError(VIR_ERR_INTERNAL_ERROR, @@ -13406,6 +13421,8 @@ virDomainVideoDefFormat(virBufferPtr buf, virBufferAsprintf(buf, " heads='%u'", def->heads); if (def->primary) virBufferAddLit(buf, " primary='yes'"); + if (def->revision) + virBufferAsprintf(buf, " revision='%u'", def->revision); if (def->accel) { virBufferAddLit(buf, ">\n"); virDomainVideoAccelDefFormat(buf, def->accel); diff --git a/src/conf/domain_conf.h b/src/conf/domain_conf.h index 9a9e0b1..81925b1 100644 --- a/src/conf/domain_conf.h +++ b/src/conf/domain_conf.h @@ -1160,6 +1160,7 @@ struct _virDomainVideoDef { unsigned int ram; /* kibibytes (multiples of 1024) */ unsigned int vram; /* kibibytes (multiples of 1024) */ unsigned int heads; + unsigned int revision; bool primary; virDomainVideoAccelDefPtr accel; virDomainDeviceInfo info; diff --git a/src/qemu/qemu_command.c b/src/qemu/qemu_command.c index f6273c1..e45c808 100644 --- a/src/qemu/qemu_command.c +++ b/src/qemu/qemu_command.c @@ -3598,6 +3598,9 @@ qemuBuildDeviceVideoStr(virDomainVideoDefPtr video, /* QEMU accepts bytes for vram_size. */ virBufferAsprintf(&buf, ",vram_size=%u", video->vram * 1024); + + if (video->revision != 0) + virBufferAsprintf(&buf, ",revision=%u", video->revision); } if (qemuBuildDeviceAddressStr(&buf, &video->info, caps) < 0) @@ -6631,6 +6634,11 @@ qemuBuildCommandLine(virConnectPtr conn, virCommandAddArgFormat(cmd, "%s.vram_size=%u", dev, vram * 1024); } + if (def->videos[0]->revision) { + virCommandAddArg(cmd, "-global"); + virCommandAddArgFormat(cmd, "%s.revision=%u", + dev, def->videos[0]->revision); + } } } diff --git a/tests/qemuxml2argvdata/qemuxml2argv-graphics-spice-compression.args b/tests/qemuxml2argvdata/qemuxml2argv-graphics-spice-compression.args index 59f064b..05f5579 100644 --- a/tests/qemuxml2argvdata/qemuxml2argv-graphics-spice-compression.args +++ b/tests/qemuxml2argvdata/qemuxml2argv-graphics-spice-compression.args @@ -7,5 +7,6 @@ image-compression=auto_glz,jpeg-wan-compression=auto,\ zlib-glz-wan-compression=auto,\ playback-compression=on,streaming-video=filter -vga \ qxl -global qxl.ram_size=67108864 -global qxl.vram_size=18874368 \ --device qxl,id=video1,ram_size=67108864,vram_size=33554432,bus=pci.0,addr=0x4 \ +-global qxl.revision=4 \ +-device qxl,id=video1,ram_size=67108864,vram_size=33554432,revision=4,bus=pci.0,addr=0x4 \ -device virtio-balloon-pci,id=balloon0,bus=pci.0,addr=0x3 diff --git a/tests/qemuxml2argvdata/qemuxml2argv-graphics-spice-compression.xml b/tests/qemuxml2argvdata/qemuxml2argv-graphics-spice-compression.xml index a8c4ad8..2dc5776 100644 --- a/tests/qemuxml2argvdata/qemuxml2argv-graphics-spice-compression.xml +++ b/tests/qemuxml2argvdata/qemuxml2argv-graphics-spice-compression.xml @@ -31,10 +31,10 @@ <streaming mode='filter'/> </graphics> <video> - <model type='qxl' ram='65536' vram='18432' heads='1'/> + <model type='qxl' ram='65536' vram='18432' heads='1' revision='4'/> </video> <video> - <model type='qxl' ram='65536' vram='32768' heads='1'/> + <model type='qxl' ram='65536' vram='32768' heads='1' revision='4'/> </video> <memballoon model='virtio'/> </devices> diff --git a/tests/qemuxml2argvdata/qemuxml2argv-graphics-spice-qxl-vga.args b/tests/qemuxml2argvdata/qemuxml2argv-graphics-spice-qxl-vga.args index ef499e6..0b08038 100644 --- a/tests/qemuxml2argvdata/qemuxml2argv-graphics-spice-qxl-vga.args +++ b/tests/qemuxml2argvdata/qemuxml2argv-graphics-spice-qxl-vga.args @@ -4,5 +4,6 @@ unix:/tmp/test-monitor,server,nowait -no-acpi -boot c -usb -hda \ /dev/HostVG/QEMUGuest1 -spice port=5903,tls-port=5904,addr=127.0.0.1,\ x509-dir=/etc/pki/libvirt-spice,tls-channel=main,plaintext-channel=inputs -vga \ qxl -global qxl-vga.ram_size=67108864 -global qxl-vga.vram_size=33554432 \ --device qxl,id=video1,ram_size=67108864,vram_size=67108864,bus=pci.0,addr=0x4 \ +-global qxl-vga.revision=4 \ +-device qxl,id=video1,ram_size=67108864,vram_size=67108864,revision=4,bus=pci.0,addr=0x4 \ -device virtio-balloon-pci,id=balloon0,bus=pci.0,addr=0x3 diff --git a/tests/qemuxml2argvdata/qemuxml2argv-graphics-spice-qxl-vga.xml b/tests/qemuxml2argvdata/qemuxml2argv-graphics-spice-qxl-vga.xml index 563d371..3cd0c42 100644 --- a/tests/qemuxml2argvdata/qemuxml2argv-graphics-spice-qxl-vga.xml +++ b/tests/qemuxml2argvdata/qemuxml2argv-graphics-spice-qxl-vga.xml @@ -28,10 +28,10 @@ <channel name='inputs' mode='insecure'/> </graphics> <video> - <model type='qxl' ram='65536' vram='32768' heads='1'/> + <model type='qxl' ram='65536' vram='32768' heads='1' revision='4'/> </video> <video> - <model type='qxl' ram='65536' vram='65536' heads='1'/> + <model type='qxl' ram='65536' vram='65536' heads='1' revision='4'/> </video> <memballoon model='virtio'/> </devices> diff --git a/tests/qemuxml2argvdata/qemuxml2argv-graphics-spice.args b/tests/qemuxml2argvdata/qemuxml2argv-graphics-spice.args index d7cfae0..082eaf7 100644 --- a/tests/qemuxml2argvdata/qemuxml2argv-graphics-spice.args +++ b/tests/qemuxml2argvdata/qemuxml2argv-graphics-spice.args @@ -6,5 +6,6 @@ x509-dir=/etc/pki/libvirt-spice,tls-channel=default,tls-channel=main,plaintext-c image-compression=auto_glz,jpeg-wan-compression=auto,zlib-glz-wan-compression=auto,\ playback-compression=on,streaming-video=filter,disable-copy-paste -vga \ qxl -global qxl.ram_size=67108864 -global qxl.vram_size=18874368 \ --device qxl,id=video1,ram_size=67108864,vram_size=33554432,bus=pci.0,addr=0x4 \ +-global qxl.revision=4 \ +-device qxl,id=video1,ram_size=67108864,vram_size=33554432,revision=4,bus=pci.0,addr=0x4 \ -device virtio-balloon-pci,id=balloon0,bus=pci.0,addr=0x3 diff --git a/tests/qemuxml2argvdata/qemuxml2argv-graphics-spice.xml b/tests/qemuxml2argvdata/qemuxml2argv-graphics-spice.xml index 9a36660..e99dbc8 100644 --- a/tests/qemuxml2argvdata/qemuxml2argv-graphics-spice.xml +++ b/tests/qemuxml2argvdata/qemuxml2argv-graphics-spice.xml @@ -34,10 +34,10 @@ <clipboard copypaste='no'/> </graphics> <video> - <model type='qxl' ram='65536' vram='18432' heads='1'/> + <model type='qxl' ram='65536' vram='18432' heads='1' revision='4'/> </video> <video> - <model type='qxl' ram='65536' vram='32768' heads='1'/> + <model type='qxl' ram='65536' vram='32768' heads='1' revision='4'/> </video> <memballoon model='virtio'/> </devices> -- 1.8.1

3 6

[libvirt] [PATCHv2] storage: fix uid_t|gid_t handling on 32 bit Linux
by Philipp Hahn 22 Feb '13

22 Feb '13

uid_t and gid_t are opaque types, ranging from s32 to u32 to u64. Unfortunately libvirt uses the value -1 to mean "current process", which on Linux32 gets converted to UINT_MAX := +(2^32-1) = 4294967295. Different libvirt versions used different formatting in the past, which break one or the other parsing function: virXPathLong(): (signed long)-1 != (double)UINT_MAX virXPathULong(): (unsigned long)-1 != (double)-1 Roll our own version of virXPathULong(), which also accepts -1, which is silently converted to UINT_MAX. For output: do the reverse and print -1 instead of UINT_MAX. Allow -1 for owner and group in schema for storage volumes. Change UINT_MAX in one test case to -1 to follow the changes from above. Signed-off-by: Philipp Hahn <hahn(a)univention.de> --- [v2] - change schema for storage volume to accept -1 for owner and group - Use UINT_MAX instead of ALLONE - Windows64 uses s64, mention s16 and u16 as well - Fix Sheepdog test case to use -1 instead of UINT_MAX --- docs/schemas/storagevol.rng | 10 +++- src/conf/storage_conf.c | 76 ++++++++++++++++++++++----- tests/storagevolxml2xmlout/vol-sheepdog.xml | 4 +- 3 files changed, 72 insertions(+), 18 deletions(-) diff --git a/docs/schemas/storagevol.rng b/docs/schemas/storagevol.rng index d4a29c7..ba95c55 100644 --- a/docs/schemas/storagevol.rng +++ b/docs/schemas/storagevol.rng @@ -49,10 +49,16 @@ <ref name='octalMode'/> </element> <element name='owner'> - <ref name='unsignedInt'/> + <choice> + <ref name='unsignedInt'/> + <value>-1</value> + </choice> </element> <element name='group'> - <ref name='unsignedInt'/> + <choice> + <ref name='unsignedInt'/> + <value>-1</value> + </choice> </element> <optional> <element name='label'> diff --git a/src/conf/storage_conf.c b/src/conf/storage_conf.c index 9134a22..617b19f 100644 --- a/src/conf/storage_conf.c +++ b/src/conf/storage_conf.c @@ -659,13 +659,14 @@ cleanup: return ret; } + static int virStorageDefParsePerms(xmlXPathContextPtr ctxt, virStoragePermsPtr perms, const char *permxpath, int defaultmode) { char *mode; - long v; + double d; int ret = -1; xmlNodePtr relnode; xmlNodePtr node; @@ -699,26 +700,58 @@ virStorageDefParsePerms(xmlXPathContextPtr ctxt, VIR_FREE(mode); } + /* uid_t and gid_t are *opaque* types: on Linux they're u32, on Windows64 + * they're s64, on Solaris they were s32 in the past, and u16 and s16 have + * been used as well. + * + * Unfortunately libvirt uses the value -1 to mean "current process", which + * on Linux32 gets converted to UINT_MAX := +(2^32-1). + * + * Different libvirt versions used different formatting in the past, which + * break one or the other parsing function: + * virXPathLong(): (signed long)-1 != (double)UINT_MAX + * virXPathULong(): (unsigned long)-1 != (double)-1 + */ if (virXPathNode("./owner", ctxt) == NULL) { perms->uid = (uid_t) -1; } else { - if (virXPathLong("number(./owner)", ctxt, &v) < 0) { + ret = virXPathNumber("number(./owner)", ctxt, &d); + if (ret == 0) { + if (d == (double) -1) { + perms->uid = (uid_t) -1; + } else { + perms->uid = (uid_t) (unsigned long) d; + if (perms->uid != d) { + ret = -2; + } + } + } + if (ret < 0) { virReportError(VIR_ERR_XML_ERROR, "%s", _("malformed owner element")); goto error; } - perms->uid = (int)v; } if (virXPathNode("./group", ctxt) == NULL) { perms->gid = (gid_t) -1; } else { - if (virXPathLong("number(./group)", ctxt, &v) < 0) { + ret = virXPathNumber("number(./group)", ctxt, &d); + if (ret == 0) { + if (d == (double) -1) { + perms->gid = (gid_t)-1; + } else { + perms->gid = (gid_t) (unsigned long) d; + if (perms->gid != d) { + ret = -2; + } + } + } + if (ret < 0) { virReportError(VIR_ERR_XML_ERROR, "%s", _("malformed group element")); goto error; } - perms->gid = (int)v; } /* NB, we're ignoring missing labels here - they'll simply inherit */ @@ -1060,10 +1093,18 @@ virStoragePoolDefFormat(virStoragePoolDefPtr def) { virBufferAddLit(&buf," <permissions>\n"); virBufferAsprintf(&buf," <mode>0%o</mode>\n", def->target.perms.mode); - virBufferAsprintf(&buf," <owner>%d</owner>\n", - (int) def->target.perms.uid); - virBufferAsprintf(&buf," <group>%d</group>\n", - (int) def->target.perms.gid); + if (def->target.perms.uid == (uid_t) -1) { + virBufferAddLit(&buf, " <owner>-1</owner>\n"); + } else { + virBufferAsprintf(&buf, " <owner>%u</owner>\n", + (unsigned int) def->target.perms.uid); + } + if (def->target.perms.gid == (gid_t) -1) { + virBufferAddLit(&buf, " <group>-1</group>\n"); + } else { + virBufferAsprintf(&buf, " <group>%u</group>\n", + (unsigned int) def->target.perms.gid); + } if (def->target.perms.label) virBufferAsprintf(&buf," <label>%s</label>\n", @@ -1313,11 +1354,18 @@ virStorageVolTargetDefFormat(virStorageVolOptionsPtr options, virBufferAddLit(buf," <permissions>\n"); virBufferAsprintf(buf," <mode>0%o</mode>\n", def->perms.mode); - virBufferAsprintf(buf," <owner>%u</owner>\n", - (unsigned int) def->perms.uid); - virBufferAsprintf(buf," <group>%u</group>\n", - (unsigned int) def->perms.gid); - + if (def->perms.uid == (uid_t) -1) { + virBufferAddLit(buf, " <owner>-1</owner>\n"); + } else { + virBufferAsprintf(buf, " <owner>%u</owner>\n", + (unsigned int) def->perms.uid); + } + if (def->perms.gid == (gid_t) -1) { + virBufferAddLit(buf, " <group>-1</group>\n"); + } else { + virBufferAsprintf(buf, " <group>%u</group>\n", + (unsigned int) def->perms.gid); + } if (def->perms.label) virBufferAsprintf(buf," <label>%s</label>\n", diff --git a/tests/storagevolxml2xmlout/vol-sheepdog.xml b/tests/storagevolxml2xmlout/vol-sheepdog.xml index 2f19af8..fc04f19 100644 --- a/tests/storagevolxml2xmlout/vol-sheepdog.xml +++ b/tests/storagevolxml2xmlout/vol-sheepdog.xml @@ -10,8 +10,8 @@ <format type='unknown'/> <permissions> <mode>0600</mode> - <owner>4294967295</owner> - <group>4294967295</group> + <owner>-1</owner> + <group>-1</group> </permissions> </target> </volume> -- 1.7.10.4

1 0

[libvirt] [PATCH 00/14] Add support for XBZRLE migration compression
by Jiri Denemark 22 Feb '13

22 Feb '13

QEMU supports XBZRLE compression of repeatedly transferred pages during live migration and this series makes it usable through libvirt. The first two patches add support for VIR_MIGRATE_COMPRESSED flag that enables XBZRLE compression. Patches 3--9 add an enhanced and extensible version of virDomainGetJobInfo and make various migration statistics available through it. The rest of the patches (10--14) add APIs for querying and changing XBZRLE compression cache size. Jiri Denemark (14): Introduce VIR_MIGRATE_COMPRESSED flag qemu: Add support for compressed migration Introduce virDomainGetJobStats API python: Implement virDomainGetJobStats wrapper remote: Auto-allocate params in remoteDeserializeTypedParameters remote: Implement virDomainGetJobStats virsh: Use virDomainGetJobStats in domjobinfo if available qemu: Parse more fields from query-migrate QMP command qemu: Implement virDomainGetJobStats Introduce virDomainMigrate*CompressionCache APIs python: Implement virDomainMigrateGetCompressionCache wrapper remote: Implement virDomainMigrate*CompressionCache virsh: Add migrate-compcache command qemu: Implement virDomainMigrate*CompressionCache daemon/remote.c | 44 ++++++ include/libvirt/libvirt.h.in | 213 ++++++++++++++++++++++++++ python/generator.py | 2 + python/libvirt-override-api.xml | 13 ++ python/libvirt-override.c | 70 +++++++++ src/driver.h | 18 +++ src/libvirt.c | 158 +++++++++++++++++++ src/libvirt_public.syms | 3 + src/qemu/qemu_domain.c | 1 + src/qemu/qemu_domain.h | 3 +- src/qemu/qemu_driver.c | 235 ++++++++++++++++++++++++++++ src/qemu/qemu_migration.c | 103 ++++++++++--- src/qemu/qemu_migration.h | 3 +- src/qemu/qemu_monitor.c | 101 ++++++++++-- src/qemu/qemu_monitor.h | 52 ++++++- src/qemu/qemu_monitor_json.c | 331 ++++++++++++++++++++++++++++++++++------ src/qemu/qemu_monitor_json.h | 15 +- src/qemu/qemu_monitor_text.c | 47 +++--- src/qemu/qemu_monitor_text.h | 5 +- src/remote/remote_driver.c | 138 ++++++++++++----- src/remote/remote_protocol.x | 31 +++- src/remote_protocol-structs | 26 ++++ src/rpc/gendispatch.pl | 2 +- tools/virsh-domain.c | 291 +++++++++++++++++++++++++++++------ tools/virsh.pod | 18 ++- 25 files changed, 1718 insertions(+), 205 deletions(-) -- 1.8.1.2

4 36