- Devel - libvirt List Archives

[libvirt] [PATCH] libvirt_private.syms: add virProcessGetStartTime
by Ján Tomko 12 Jun '13

12 Jun '13

https://bugzilla.redhat.com/show_bug.cgi?id=973543 --- src/libvirt_private.syms | 1 + 1 file changed, 1 insertion(+) diff --git a/src/libvirt_private.syms b/src/libvirt_private.syms index 1ea7467..042081f 100644 --- a/src/libvirt_private.syms +++ b/src/libvirt_private.syms @@ -1711,6 +1711,7 @@ virPortAllocatorRelease; virProcessAbort; virProcessGetAffinity; virProcessGetNamespaces; +virProcessGetStartTime; virProcessKill; virProcessKillPainfully; virProcessSetAffinity; -- 1.8.1.5

2 2

[libvirt] VirtualBox 4.2 Support
by Dead Horse 12 Jun '13

12 Jun '13

Is there any active development to support VirtualBox 4.2 under libvirt? - DHC

2 2

[libvirt] Can we allow users in 'qemu' group to skip polkit auth?
by Cole Robinson 12 Jun '13

12 Jun '13

There's a bug report filed against Fedora libvirt requesting a polkit rule be installed that grants read/write libvirt access to all users in the 'qemu' group: https://bugzilla.redhat.com/show_bug.cgi?id=957300 I'm inclined to agree with the reporter, and time has shown that many users install custom polkit rules to grant their user passwordless access to libvirt so this would definitely fill a need. I'm sure there's plenty to consider here since we are talking about security. Thoughts? Thanks, Cole

2 1

[libvirt] Disabling bootmenu causes boot to skip over cdrom
by Keno Fischer 12 Jun '13

12 Jun '13

Hello, I have domain with the following setup: - 1 VirtIO Disk - 1 VirtIO CDROM - 2 IDE CDROMs Now, depending on whether or not I have the bootmenu flag enabled, the boot sequence skips over booting from CDROM entirely (it does work if the bootmenu is enabled). This started happening after I added the VirtIO CDROM. Is this a bug or am I doing something wrong? Thanks, Keno Fischer

2 1

[libvirt] [PATCH] doc: Explicitly document migration limitations on errors
by Peter Krempa 12 Jun '13

12 Jun '13

Migration of guests that are in error state for example because of I/O errors may not be possible and encountering of such errors may lead to cancellation of such migration. Document this behavior explicitly. --- src/libvirt.c | 20 ++++++++++++++++---- 1 file changed, 16 insertions(+), 4 deletions(-) diff --git a/src/libvirt.c b/src/libvirt.c index 620dbdd..1bdd193 100644 --- a/src/libvirt.c +++ b/src/libvirt.c @@ -5196,7 +5196,10 @@ virDomainMigrateDirect(virDomainPtr domain, * There are many limitations on migration imposed by the underlying * technology - for example it may not be possible to migrate between * different processors even with the same architecture, or between - * different types of hypervisor. + * different types of hypervisor. A domain may not be migratable + * because of configuration or error states. A domain may become + * unmigratable during migration. This will cancel the ongoing + * migration. * * Returns the new domain object if the migration was successful, * or NULL in case of error. Note that the new domain object @@ -5423,7 +5426,10 @@ error: * There are many limitations on migration imposed by the underlying * technology - for example it may not be possible to migrate between * different processors even with the same architecture, or between - * different types of hypervisor. + * different types of hypervisor. A domain may not be migratable + * because of configuration or error states. A domain may become + * unmigratable during migration. This will cancel the ongoing + * migration. * * If the hypervisor supports it, @dxml can be used to alter * host-specific portions of the domain XML that will be used on @@ -5664,7 +5670,10 @@ error: * There are many limitations on migration imposed by the underlying * technology - for example it may not be possible to migrate between * different processors even with the same architecture, or between - * different types of hypervisor. + * different types of hypervisor. A domain may not be migratable + * because of configuration or error states. A domain may become + * unmigratable during migration. This will cancel the ongoing + * migration. * * Returns 0 if the migration succeeded, -1 upon error. */ @@ -5826,7 +5835,10 @@ error: * There are many limitations on migration imposed by the underlying * technology - for example it may not be possible to migrate between * different processors even with the same architecture, or between - * different types of hypervisor. + * different types of hypervisor. A domain may not be migratable + * because of configuration or error states. A domain may become + * unmigratable during migration. This will cancel the ongoing + * migration. * * Returns 0 if the migration succeeded, -1 upon error. */ -- 1.8.2.1

2 1

[libvirt] [PATCH] Fix F_DUPFD_CLOEXEC operation args
by Daniel P. Berrange 12 Jun '13

12 Jun '13

From: "Daniel P. Berrange" <berrange(a)redhat.com> The F_DUPFD_CLOEXEC operation with fcntl() expects a single int argument, specifying the minimum FD number for the newly dup'd file descriptor. We were not specifying that causing random stack data to be accessed as the FD number. Sometimes that worked, sometimes it didn't. Signed-off-by: Daniel P. Berrange <berrange(a)redhat.com> --- src/rpc/virnetsocket.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/src/rpc/virnetsocket.c b/src/rpc/virnetsocket.c index e950d7f..dcf98b1 100644 --- a/src/rpc/virnetsocket.c +++ b/src/rpc/virnetsocket.c @@ -1055,7 +1055,7 @@ int virNetSocketDupFD(virNetSocketPtr sock, bool cloexec) int fd; if (cloexec) - fd = fcntl(sock->fd, F_DUPFD_CLOEXEC); + fd = fcntl(sock->fd, F_DUPFD_CLOEXEC, 0); else fd = dup(sock->fd); if (fd < 0) { -- 1.8.2.1

4 4

[libvirt] [PATCH 0/2] qemu: Cancel/forbid migration on I/O error
by Peter Krempa 12 Jun '13

12 Jun '13

When qemu has paused itself due to a I/O error it's not safe to migrate it as it still might contain state in kernel buffers. These patches forbid and cancel ongoing migration if a I/O error happens. Peter Krempa (2): qemu: Cancel migration if guest encoutners I/O error while migrating qemu: Forbid migration of machines with I/O errors src/qemu/qemu_migration.c | 31 +++++++++++++++++++++++++++++++ 1 file changed, 31 insertions(+) -- 1.8.2.1

3 7

[libvirt] [PATCH] qemu: Fix memory leak in Prepare phase
by Jiri Denemark 11 Jun '13

11 Jun '13

Avoid leaking virDomainDef if Prepare phase fails before it gets to qemuMigrationPrepareAny. --- src/qemu/qemu_driver.c | 12 ++++-------- src/qemu/qemu_migration.c | 29 ++++++++++++++--------------- src/qemu/qemu_migration.h | 4 ++-- 3 files changed, 20 insertions(+), 25 deletions(-) diff --git a/src/qemu/qemu_driver.c b/src/qemu/qemu_driver.c index 3bad629..621e184 100644 --- a/src/qemu/qemu_driver.c +++ b/src/qemu/qemu_driver.c @@ -9552,8 +9552,7 @@ qemuDomainMigratePrepareTunnel(virConnectPtr dconn, ret = qemuMigrationPrepareTunnel(driver, dconn, NULL, 0, NULL, NULL, /* No cookies in v2 */ - st, def, flags); - def = NULL; + st, &def, flags); cleanup: virDomainDefFree(def); @@ -9628,8 +9627,7 @@ qemuDomainMigratePrepare2(virConnectPtr dconn, ret = qemuMigrationPrepareDirect(driver, dconn, NULL, 0, NULL, NULL, /* No cookies */ uri_in, uri_out, - def, flags); - def = NULL; + &def, flags); cleanup: virDomainDefFree(def); @@ -9864,8 +9862,7 @@ qemuDomainMigratePrepare3(virConnectPtr dconn, cookiein, cookieinlen, cookieout, cookieoutlen, uri_in, uri_out, - def, flags); - def = NULL; + &def, flags); cleanup: virDomainDefFree(def); @@ -9926,8 +9923,7 @@ qemuDomainMigratePrepareTunnel3(virConnectPtr dconn, ret = qemuMigrationPrepareTunnel(driver, dconn, cookiein, cookieinlen, cookieout, cookieoutlen, - st, def, flags); - def = NULL; + st, &def, flags); cleanup: virDomainDefFree(def); diff --git a/src/qemu/qemu_migration.c b/src/qemu/qemu_migration.c index dec04a5..0a54a30 100644 --- a/src/qemu/qemu_migration.c +++ b/src/qemu/qemu_migration.c @@ -2016,7 +2016,7 @@ qemuMigrationPrepareAny(virQEMUDriverPtr driver, int cookieinlen, char **cookieout, int *cookieoutlen, - virDomainDefPtr def, + virDomainDefPtr *def, virStreamPtr st, unsigned int port, unsigned long flags) @@ -2064,7 +2064,7 @@ qemuMigrationPrepareAny(virQEMUDriverPtr driver, if (!(caps = virQEMUDriverGetCapabilities(driver, false))) goto cleanup; - if (!qemuMigrationIsAllowed(driver, NULL, def, true)) + if (!qemuMigrationIsAllowed(driver, NULL, *def, true)) goto cleanup; /* Let migration hook filter domain XML */ @@ -2072,12 +2072,12 @@ qemuMigrationPrepareAny(virQEMUDriverPtr driver, char *xml; int hookret; - if (!(xml = qemuDomainDefFormatXML(driver, def, + if (!(xml = qemuDomainDefFormatXML(driver, *def, VIR_DOMAIN_XML_SECURE | VIR_DOMAIN_XML_MIGRATABLE))) goto cleanup; - hookret = virHookCall(VIR_HOOK_DRIVER_QEMU, def->name, + hookret = virHookCall(VIR_HOOK_DRIVER_QEMU, (*def)->name, VIR_HOOK_QEMU_OP_MIGRATE, VIR_HOOK_SUBOP_BEGIN, NULL, xml, &xmlout); VIR_FREE(xml); @@ -2098,13 +2098,13 @@ qemuMigrationPrepareAny(virQEMUDriverPtr driver, if (!newdef) goto cleanup; - if (!virDomainDefCheckABIStability(def, newdef)) { + if (!virDomainDefCheckABIStability(*def, newdef)) { virDomainDefFree(newdef); goto cleanup; } - virDomainDefFree(def); - def = newdef; + virDomainDefFree(*def); + *def = newdef; } } } @@ -2122,7 +2122,7 @@ qemuMigrationPrepareAny(virQEMUDriverPtr driver, .ai_socktype = SOCK_STREAM }; if (!(qemuCaps = virQEMUCapsCacheLookupCopy(driver->qemuCapsCache, - def->emulator))) + (*def)->emulator))) goto cleanup; /* Listen on :: instead of 0.0.0.0 if QEMU understands it @@ -2146,14 +2146,14 @@ qemuMigrationPrepareAny(virQEMUDriverPtr driver, } } - if (!(vm = virDomainObjListAdd(driver->domains, def, + if (!(vm = virDomainObjListAdd(driver->domains, *def, driver->xmlopt, VIR_DOMAIN_OBJ_LIST_ADD_LIVE | VIR_DOMAIN_OBJ_LIST_ADD_CHECK_LIVE, NULL))) goto cleanup; - def = NULL; + *def = NULL; priv = vm->privateData; priv->origname = origname; origname = NULL; @@ -2272,7 +2272,6 @@ cleanup: VIR_FREE(migrateFrom); VIR_FREE(origname); VIR_FREE(xmlout); - virDomainDefFree(def); VIR_FORCE_CLOSE(dataFD[0]); VIR_FORCE_CLOSE(dataFD[1]); if (vm) { @@ -2315,7 +2314,7 @@ qemuMigrationPrepareTunnel(virQEMUDriverPtr driver, char **cookieout, int *cookieoutlen, virStreamPtr st, - virDomainDefPtr def, + virDomainDefPtr *def, unsigned long flags) { int ret; @@ -2324,7 +2323,7 @@ qemuMigrationPrepareTunnel(virQEMUDriverPtr driver, "cookieout=%p, cookieoutlen=%p, st=%p, def=%p, " "flags=%lx", driver, dconn, NULLSTR(cookiein), cookieinlen, - cookieout, cookieoutlen, st, def, flags); + cookieout, cookieoutlen, st, *def, flags); ret = qemuMigrationPrepareAny(driver, dconn, cookiein, cookieinlen, cookieout, cookieoutlen, def, @@ -2342,7 +2341,7 @@ qemuMigrationPrepareDirect(virQEMUDriverPtr driver, int *cookieoutlen, const char *uri_in, char **uri_out, - virDomainDefPtr def, + virDomainDefPtr *def, unsigned long flags) { static int port = 0; @@ -2358,7 +2357,7 @@ qemuMigrationPrepareDirect(virQEMUDriverPtr driver, "def=%p, flags=%lx", driver, dconn, NULLSTR(cookiein), cookieinlen, cookieout, cookieoutlen, NULLSTR(uri_in), uri_out, - def, flags); + *def, flags); /* The URI passed in may be NULL or a string "tcp://somehostname:port". * diff --git a/src/qemu/qemu_migration.h b/src/qemu/qemu_migration.h index b42fe4e..22b04b4 100644 --- a/src/qemu/qemu_migration.h +++ b/src/qemu/qemu_migration.h @@ -98,7 +98,7 @@ int qemuMigrationPrepareTunnel(virQEMUDriverPtr driver, char **cookieout, int *cookieoutlen, virStreamPtr st, - virDomainDefPtr def, + virDomainDefPtr *def, unsigned long flags); int qemuMigrationPrepareDirect(virQEMUDriverPtr driver, @@ -109,7 +109,7 @@ int qemuMigrationPrepareDirect(virQEMUDriverPtr driver, int *cookieoutlen, const char *uri_in, char **uri_out, - virDomainDefPtr def, + virDomainDefPtr *def, unsigned long flags); int qemuMigrationPerform(virQEMUDriverPtr driver, -- 1.8.2.1

2 2

[libvirt] [PATCH v3] selinux: assume 's0' if the range is empty
by Ján Tomko 11 Jun '13

11 Jun '13

This fixes a crash: https://bugzilla.redhat.com/show_bug.cgi?id=969878 --- v3: rewrote the patch to assume 's0' if the range is empty, instead of reporting an error v2: https://www.redhat.com/archives/libvir-list/2013-June/msg00082.html src/security/security_selinux.c | 9 +++++++-- tests/securityselinuxtest.c | 6 ++++++ 2 files changed, 13 insertions(+), 2 deletions(-) diff --git a/src/security/security_selinux.c b/src/security/security_selinux.c index b862fbf..6fe063e 100644 --- a/src/security/security_selinux.c +++ b/src/security/security_selinux.c @@ -170,11 +170,13 @@ virSecuritySELinuxMCSFind(virSecurityManagerPtr mgr, /* * This needs to cope with several styles of range * + * system_u:system_r:virtd_t * system_u:system_r:virtd_t:s0 * system_u:system_r:virtd_t:s0-s0 * system_u:system_r:virtd_t:s0-s0:c0.c1023 * - * In the first two cases, we'll assume c0.c1023 for + * In the first case we'll assume s0:c0.c1023 and + * in the next two cases, we'll assume c0.c1023 for * the category part, since that's what we're really * interested in. This won't work in Enforcing mode, * but will prevent libvirtd breaking in Permissive @@ -189,6 +191,7 @@ virSecuritySELinuxMCSGetProcessRange(char **sens, context_t ourContext = NULL; char *cat = NULL; char *tmp; + const char *contextRange; int ret = -1; if (getcon_raw(&ourSecContext) < 0) { @@ -202,8 +205,10 @@ virSecuritySELinuxMCSGetProcessRange(char **sens, ourSecContext); goto cleanup; } + if (!(contextRange = context_range_get(ourContext))) + contextRange = "s0"; - if (VIR_STRDUP(*sens, context_range_get(ourContext)) < 0) + if (VIR_STRDUP(*sens, contextRange) < 0) goto cleanup; /* Find and blank out the category part (if any) */ diff --git a/tests/securityselinuxtest.c b/tests/securityselinuxtest.c index bdf248b..8d85c68 100644 --- a/tests/securityselinuxtest.c +++ b/tests/securityselinuxtest.c @@ -333,6 +333,12 @@ mymain(void) "system_u", "system_r", "object_r", "svirt_t", "svirt_image_t", 2, 3, 0, 1023); + DO_TEST_GEN_LABEL("dynamic virtd, missing range", + "system_u:system_r:virtd_t", + true, NULL, NULL, + "system_u", "system_r", "object_r", + "svirt_t", "svirt_image_t", + 0, 0, 0, 1023); return (ret == 0) ? EXIT_SUCCESS : EXIT_FAILURE; } -- 1.8.1.5

2 2

[libvirt] [PATCH] qemu_migrate: Dispose listen address if set from config
by Michal Privoznik 11 Jun '13

11 Jun '13

https://bugzilla.redhat.com/show_bug.cgi?id=971485 As of d7f9d827531bc843b7c5aa9d3e8c08738a1de248 we copy the listen address from the qemu.conf config file in case none has been provided via XML. But later, when migrating, we should not include such listen address in the migratable XML as it is something autogenerated, not requested by user. Moreover, the binding to the listen address will likely fail, unless the address is '0.0.0.0' or its IPv6 equivalent. This patch introduces a new boolean attribute to virDomainGraphicsListenDef to distinguish autofilled listen addresses. However, we must keep the attribute over libvirtd restarts, so it must be kept within status XML. --- src/conf/domain_conf.c | 28 ++++++++++++++++++++++++++++ src/conf/domain_conf.h | 1 + src/qemu/qemu_process.c | 1 + 3 files changed, 30 insertions(+) diff --git a/src/conf/domain_conf.c b/src/conf/domain_conf.c index 3e81013..2373397 100644 --- a/src/conf/domain_conf.c +++ b/src/conf/domain_conf.c @@ -7688,6 +7688,8 @@ virDomainGraphicsListenDefParseXML(virDomainGraphicsListenDefPtr def, char *type = virXMLPropString(node, "type"); char *address = virXMLPropString(node, "address"); char *network = virXMLPropString(node, "network"); + char *fromConfig = virXMLPropString(node, "fromConfig"); + int tmp; if (!type) { virReportError(VIR_ERR_XML_ERROR, "%s", @@ -7724,6 +7726,17 @@ virDomainGraphicsListenDefParseXML(virDomainGraphicsListenDefPtr def, network = NULL; } + if (fromConfig && + flags & VIR_DOMAIN_XML_INTERNAL_STATUS) { + if (virStrToLong_i(fromConfig, NULL, 10, &tmp) < 0) { + virReportError(VIR_ERR_XML_ERROR, + _("Invalid fromConfig value: %s"), + fromConfig); + goto error; + } + def->fromConfig = tmp != 0; + } + ret = 0; error: if (ret < 0) @@ -7731,6 +7744,7 @@ error: VIR_FREE(type); VIR_FREE(address); VIR_FREE(network); + VIR_FREE(fromConfig); return ret; } @@ -15365,6 +15379,11 @@ virDomainGraphicsListenDefFormat(virBufferPtr buf, virDomainGraphicsListenDefPtr def, unsigned int flags) { + /* If generating migratable XML, skip listen address + * dragged in from config file */ + if ((flags & VIR_DOMAIN_XML_MIGRATABLE) && def->fromConfig) + return; + virBufferAddLit(buf, " <listen"); if (def->type) { @@ -15386,6 +15405,9 @@ virDomainGraphicsListenDefFormat(virBufferPtr buf, virBufferEscapeString(buf, " network='%s'", def->network); } + if (flags & VIR_DOMAIN_XML_INTERNAL_STATUS) + virBufferAsprintf(buf, " fromConfig='%d'", def->fromConfig); + virBufferAddLit(buf, "/>\n"); } @@ -15412,6 +15434,9 @@ virDomainGraphicsDefFormat(virBufferPtr buf, for (i = 0; i < def->nListens; i++) { if (virDomainGraphicsListenGetType(def, i) == VIR_DOMAIN_GRAPHICS_LISTEN_TYPE_ADDRESS) { + if (flags & VIR_DOMAIN_XML_MIGRATABLE && + def->listens[i].fromConfig) + continue; listenAddr = virDomainGraphicsListenGetAddress(def, i); break; } @@ -15531,6 +15556,9 @@ virDomainGraphicsDefFormat(virBufferPtr buf, if (virDomainGraphicsListenGetType(def, i) == VIR_DOMAIN_GRAPHICS_LISTEN_TYPE_NONE) continue; + if (flags & VIR_DOMAIN_XML_MIGRATABLE && + def->listens[i].fromConfig) + continue; if (!children) { virBufferAddLit(buf, ">\n"); children = true; diff --git a/src/conf/domain_conf.h b/src/conf/domain_conf.h index b8edb1e..5b159ac 100644 --- a/src/conf/domain_conf.h +++ b/src/conf/domain_conf.h @@ -1420,6 +1420,7 @@ struct _virDomainGraphicsListenDef { int type; /* enum virDomainGraphicsListenType */ char *address; char *network; + bool fromConfig; /* true if the @address is config file originated */ }; struct _virDomainGraphicsDef { diff --git a/src/qemu/qemu_process.c b/src/qemu/qemu_process.c index c412ea2..5a0f18b 100644 --- a/src/qemu/qemu_process.c +++ b/src/qemu/qemu_process.c @@ -3486,6 +3486,7 @@ int qemuProcessStart(virConnectPtr conn, VIR_SHRINK_N(graphics->listens, graphics->nListens, 1); goto cleanup; } + graphics->listens[0].fromConfig = true; } } } -- 1.8.1.5

2 1