- Devel - libvirt List Archives

[libvirt] [PATCH 1/3] rpc: libssh: allow a NULL known_hosts file
by Pino Toscano 11 Jan '17

11 Jan '17

Make sure that virNetLibsshSessionSetHostKeyVerification accepts a NULL value for the path to the known_hosts file: - call ssh_options_set(SSH_OPTIONS_KNOWNHOSTS) anyway, using /dev/null, otherwise libssh will use its default path - do not call ssh_write_knownhost when no known hosts file was set Resolves: https://bugzilla.redhat.com/show_bug.cgi?id=1406457 --- src/rpc/virnetlibsshsession.c | 37 +++++++++++++++++++++++-------------- 1 file changed, 23 insertions(+), 14 deletions(-) diff --git a/src/rpc/virnetlibsshsession.c b/src/rpc/virnetlibsshsession.c index 5de6629..25f93ce 100644 --- a/src/rpc/virnetlibsshsession.c +++ b/src/rpc/virnetlibsshsession.c @@ -382,14 +382,16 @@ virNetLibsshCheckHostKey(virNetLibsshSessionPtr sess) VIR_FREE(askKey.result); } - /* write the host key file */ - if (ssh_write_knownhost(sess->session) < 0) { - errmsg = ssh_get_error(sess->session); - virReportError(VIR_ERR_LIBSSH, - _("failed to write known_host file '%s': %s"), - sess->knownHostsFile, - errmsg); - return -1; + /* write the host key file, if specified */ + if (sess->knownHostsFile) { + if (ssh_write_knownhost(sess->session) < 0) { + errmsg = ssh_get_error(sess->session); + virReportError(VIR_ERR_LIBSSH, + _("failed to write known_host file '%s': %s"), + sess->knownHostsFile, + errmsg); + return -1; + } } /* key was accepted and added */ return 0; @@ -1172,13 +1174,20 @@ virNetLibsshSessionSetHostKeyVerification(virNetLibsshSessionPtr sess, goto error; } - /* set the known hosts file */ - if (ssh_options_set(sess->session, SSH_OPTIONS_KNOWNHOSTS, hostsfile) < 0) - goto error; + /* set the known hosts file, if specified */ + if (hostsfile) { + if (ssh_options_set(sess->session, SSH_OPTIONS_KNOWNHOSTS, hostsfile) < 0) + goto error; - VIR_FREE(sess->knownHostsFile); - if (VIR_STRDUP(sess->knownHostsFile, hostsfile) < 0) - goto error; + VIR_FREE(sess->knownHostsFile); + if (VIR_STRDUP(sess->knownHostsFile, hostsfile) < 0) + goto error; + } else { + /* libssh does not support trying no known_host file at all: + * hence use /dev/null here, without storing it as file */ + if (ssh_options_set(sess->session, SSH_OPTIONS_KNOWNHOSTS, "/dev/null") < 0) + goto error; + } virObjectUnlock(sess); return 0; -- 2.7.4

2 5

[libvirt] [PATCH] maint: update to latest gnulib
by Eric Blake 11 Jan '17

11 Jan '17

Among other recent changes, this includes a workaround to avoid Clang compiler bug https://llvm.org/bugs/show_bug.cgi?id=16404 having spurious link failures. * .gnulib: Update. * bootstrap: Synchronize to upstream. Signed-off-by: Eric Blake <eblake(a)redhat.com> --- I'm debating about re-enabling test-lock (right now, we explicitly pass --avoid=lock-tests to gnulib-tool via our bootstrap.conf), now that upstream is finally starting to fix the bugs in the test that were causing extremely slow completion or outright hangs on multi-processor setups. But that's probably something better done right after a libvirt release, not right before a release candidate. Pushing under the gnulib maintenance rule; contains these gnulib patches: * .gnulib e210a3c...94386a1 (38): > maint.mk: enforce spelling of "timestamp" (i.e., no space) > dfa: minor simplification with emptyset > dfa: shrink constraints from 4 bits to 3 > dfa: omit unnecessary ptrdiff_t check > dfa: omit unnecessary allocation > dfa: omit unused local > maint: time stamp -> timestamp > maint: remove stray .texi files > getprogname: fix port to IRIX > dfa: melt down dfastate into build_state > dfa: simplify transition table allocation > dfa: fix reallocation bug when matching newlines > Avoid -Wundef warning about undefined WINDOWS_SOCKETS. > Avoid -Wundef warning about undefined __USE_FILE_OFFSET64. > stdioext: Port to Minix 3.2 and newer. > getprogname: port to IRIX > localename-tests: port to NetBSD 7 > glob, intprops, xalloc: work around Clang bug > dfa: fix 'return' typo > lock tests: Prefer semaphore over mutex. > parse-datetime: fix generated paths for coverage files > maint.mk: support parallel execution of coverage > lock: Provide guarantee to avoid writer starvation for rwlocks. > thread: Fix pth port. > parse-datetime: fix debug message on lone year number > parse-datetime: fix local timezone debug messages > parse-datetime: add debug warning about DST changes > parse-datetime: add debug warning about date arithmetic > parse-datetime: fix debug message of relative part after timezone > parse-datetime: fix incorrect debug message on lone number > exec > build-aux/mdate-sh > doc: fix typo in previous change > Revert copyright-year change to synced files > doc: modernize for C11 etc. > dfa: prefer functions to FETCH_WC macro > dfa: narrow more local var scopes > dfa: remove duplicate assignment .gnulib | 2 +- bootstrap | 4 ++-- 2 files changed, 3 insertions(+), 3 deletions(-) diff --git a/.gnulib b/.gnulib index e210a3c..94386a1 160000 --- a/.gnulib +++ b/.gnulib @@ -1 +1 @@ -Subproject commit e210a3cbaec0ee82a67ff8fc427e21bdd64dba1b +Subproject commit 94386a13667c645fd42544a7fd302cdddd39fcdf diff --git a/bootstrap b/bootstrap index 26066b2..932ff85 100755 --- a/bootstrap +++ b/bootstrap @@ -1,6 +1,6 @@ #! /bin/sh # Print a version string. -scriptversion=2016-11-03.18; # UTC +scriptversion=2017-01-09.19; # UTC # Bootstrap this package from checked-out sources. @@ -790,7 +790,7 @@ symlink_to_dir() # Leave any existing symlink alone, if it already points to the source, # so that broken build tools that care about symlink times # aren't confused into doing unnecessary builds. Conversely, if the - # existing symlink's time stamp is older than the source, make it afresh, + # existing symlink's timestamp is older than the source, make it afresh, # so that broken tools aren't confused into skipping needed builds. See # <http://lists.gnu.org/archive/html/bug-gnulib/2011-05/msg00326.html>. test -h "$dst" && -- 2.9.3

1 0

[libvirt] [PATCH v3 0/3] Improve release notes process
by Andrea Bolognani 11 Jan '17

11 Jan '17

Dan raised some concerns, but it doesn't look like they're critical enough for him to NACK the series; moreover, it's all stuff that we can easily address later on, so I'd like to move forward with this before the freeze. Details of changes can be found in the single patches. Andrea Bolognani (3): NEWS: Improve building pipeline NEWS: Reformat at generation time docs: Document the release notes process for contributors .gitignore | 1 + HACKING | 7 ++ Makefile.am | 26 ++++-- docs/Makefile.am | 28 +++++- docs/hacking.html.in | 10 ++ docs/news-ascii.xsl | 68 ++++++++++++++ docs/news-html.xsl | 96 +++++++++++++++++++ docs/news.html.in | 164 -------------------------------- docs/news.xml | 252 ++++++++++++++++++++++++++++++++++++++++++++++++++ docs/news.xsl | 50 ---------- docs/reformat-news.py | 104 +++++++++++++++++++++ 11 files changed, 583 insertions(+), 223 deletions(-) create mode 100644 docs/news-ascii.xsl create mode 100644 docs/news-html.xsl delete mode 100644 docs/news.html.in create mode 100644 docs/news.xml delete mode 100644 docs/news.xsl create mode 100755 docs/reformat-news.py -- 2.7.4

2 7

[libvirt] [PATCH 0/2] Couple of build fixes
by Michal Privoznik 11 Jan '17

11 Jan '17

Spotted by our CI. Michal Privoznik (2): qemu_domain: Move qemuDomainGetPreservedMounts virSecuritySELinuxSetFileconHelper: Fix build with broken selinux.h src/qemu/qemu_domain.c | 140 ++++++++++++++++++++-------------------- src/security/security_selinux.c | 2 +- 2 files changed, 71 insertions(+), 71 deletions(-) -- 2.11.0

2 3

[libvirt] [PATCH] libxl: implement virDomainGetMaxVcpus
by Jim Fehlig 11 Jan '17

11 Jan '17

The libxl driver already supports getting maximum vcpu count via libxlDomainGetVcpusFlags, allowing to trivially implement virDomainGetMaxVcpus. Signed-off-by: Jim Fehlig <jfehlig(a)suse.com> --- docs/news.html.in | 2 ++ src/libxl/libxl_driver.c | 8 ++++++++ 2 files changed, 10 insertions(+) diff --git a/docs/news.html.in b/docs/news.html.in index c222075..2365c39 100644 --- a/docs/news.html.in +++ b/docs/news.html.in @@ -39,6 +39,8 @@ _virStorageVolInfo structure. The --physical flag has been added to the virsh vol-info command to access the data </li> + <li>libxl: Implement virDomainGetMaxVcpus API + </li> </ul> </li> <li><strong>Improvements</strong> diff --git a/src/libxl/libxl_driver.c b/src/libxl/libxl_driver.c index 7e5d9b6..51fe3d8 100644 --- a/src/libxl/libxl_driver.c +++ b/src/libxl/libxl_driver.c @@ -2363,6 +2363,13 @@ libxlDomainGetVcpusFlags(virDomainPtr dom, unsigned int flags) } static int +libxlDomainGetMaxVcpus(virDomainPtr dom) +{ + return libxlDomainGetVcpusFlags(dom, (VIR_DOMAIN_AFFECT_LIVE | + VIR_DOMAIN_VCPU_MAXIMUM)); +} + +static int libxlDomainPinVcpuFlags(virDomainPtr dom, unsigned int vcpu, unsigned char *cpumap, int maplen, unsigned int flags) @@ -6446,6 +6453,7 @@ static virHypervisorDriver libxlHypervisorDriver = { .domainSetVcpus = libxlDomainSetVcpus, /* 0.9.0 */ .domainSetVcpusFlags = libxlDomainSetVcpusFlags, /* 0.9.0 */ .domainGetVcpusFlags = libxlDomainGetVcpusFlags, /* 0.9.0 */ + .domainGetMaxVcpus = libxlDomainGetMaxVcpus, /* 3.0.0 */ .domainPinVcpu = libxlDomainPinVcpu, /* 0.9.0 */ .domainPinVcpuFlags = libxlDomainPinVcpuFlags, /* 1.2.1 */ .domainGetVcpus = libxlDomainGetVcpus, /* 0.9.0 */ -- 2.9.2

2 2

[libvirt] [PATCH 0/2] replacement for Patches 5/8 & 6/8 of "aggregate multiple pcie-root-ports onto a single slot"
by Laine Stump 11 Jan '17

11 Jan '17

During review of the aforementioned series, Andrea expressed misgivings about patch 6, which automatically set the multifunction=on option transiently at the time the qemu commandline was generated. Dan agreed. These two patches replace patches 5 and 6 of that series with two patches that will set multi='on' in the domain's config rather than just temporarily in the commandline. This way it will remain set even if all the devices on the non-0 function are later removed (unless it's explicitly changed by the user/management app, that is). The rest of the series is ACKed and ready to push, which I'd like to do before the freeze on Wednesday; these two patche are all that's holding it up. Laine Stump (2): conf: new function virDomainPCIAddressSetAllMulti() qemu: use virDomainPCIAddressSetAllMulti() to set multi when needed src/conf/domain_addr.c | 75 +++++++++++++ src/conf/domain_addr.h | 4 + src/libvirt_private.syms | 1 + src/qemu/qemu_domain_address.c | 5 + .../qemuxml2argv-q35-multifunction.args | 42 ++++++++ .../qemuxml2argv-q35-multifunction.xml | 51 +++++++++ tests/qemuxml2argvtest.c | 23 ++++ .../qemuxml2xmlout-q35-multifunction.xml | 120 +++++++++++++++++++++ tests/qemuxml2xmltest.c | 23 ++++ 9 files changed, 344 insertions(+) create mode 100644 tests/qemuxml2argvdata/qemuxml2argv-q35-multifunction.args create mode 100644 tests/qemuxml2argvdata/qemuxml2argv-q35-multifunction.xml create mode 100644 tests/qemuxml2xmloutdata/qemuxml2xmlout-q35-multifunction.xml -- 2.7.4

2 6

[libvirt] qemu-kvm blocked for more than 120 seconds when "libvirt-guests" is enabled
by Oscar Segarra 11 Jan '17

11 Jan '17

Hi, I just have two virtual machines in my environment, I want them to gracefully stop when host is powered off gracefully. Nevertheless, system hangs on shutdown: [root@vdicnode01 ~]# virsh list Id Name State ---------------------------------------------------- 1 vdicdb01 running 2 vdicone01 running The configuration: /usr/libexec/libvirt-guests.sh URIS=default ON_BOOT=ignore ON_SHUTDOWN=shutdown SHUTDOWN_TIMEOUT=60 PARALLEL_SHUTDOWN=5 START_DELAY=0 BYPASS_CACHE=0 CONNECT_RETRIES=10 RETRIES_SLEEP=1 SYNC_TIME=0 [root@vdicnode01 ~]# cat /etc/libvirt/qemu.conf user = "oneadmin" group = "oneadmin" dynamic_ownership = 0 spice_tls = 0 With libvirt-guests service started: [root@vdicnode01 ~]# service libvirt-guests status Redirecting to /bin/systemctl status libvirt-guests.service ● libvirt-guests.service - Suspend Active Libvirt Guests Loaded: loaded (/usr/lib/systemd/system/libvirt-guests.service; disabled; vendor preset: disabled) Active: active (exited) since Sat 2016-12-31 17:13:34 CET; 5s ago Docs: man:libvirtd(8) http://libvirt.org Process: 6619 ExecStart=/usr/libexec/libvirt-guests.sh start (code=exited, status=0/SUCCESS) Main PID: 6619 (code=exited, status=0/SUCCESS) Dec 31 17:13:34 vdicnode01.vdicube.com systemd[1]: Starting Suspend Active Libvirt Guests... Dec 31 17:13:34 vdicnode01.vdicube.com libvirt-guests.sh[6619]: libvirt-guests is configured not to start any guests on boot Dec 31 17:13:34 vdicnode01.vdicube.com systemd[1]: Started Suspend Active Libvirt Guests. [root@vdicnode01 ~]# If I stop locally libvirt-guests (as root) Looks work perfectly: [root@vdicnode01 ~]# virsh list Id Name State ---------------------------------------------------- [root@vdicnode01 ~]# --> Now I start again the virtual machines and the libvirt-guests service <-- [root@vdicnode01 ~]# service libvirtd restart Redirecting to /bin/systemctl restart libvirtd.service [root@vdicnode01 ~]# service libvirt-guests start Redirecting to /bin/systemctl start libvirt-guests.service [root@vdicnode01 ~]# virsh list Id Name State ---------------------------------------------------- 1 vdicone01 running 2 vdicdb01 running [root@vdicnode01 ~]# But If I shutdown the host it look not work properly (I attack screenshots of the shutdown process). Any help will be welcome. Thanks a lot.

3 11

[libvirt] [PATCH] storage: Fix storage_backend probing when PARTED not installed.
by John Ferlan 11 Jan '17

11 Jan '17

Commit id 'a48c674f' caused problems for systems without PARTED installed. So move the PARTED probing code back to storage_backend_disk.c and create a shim within storage_backend.c to call it if WITH_STORAGE_DISK is true; otherwise, just return -1 with the error. Signed-off-by: John Ferlan <jferlan(a)redhat.com> --- Reported to me internally via Peter Krempa... Not pushing as a build breaker - I'll let Peter review and ACK. src/storage/storage_backend.c | 155 ++++--------------------------------- src/storage/storage_backend_disk.c | 152 ++++++++++++++++++++++++++++++++++++ src/storage/storage_backend_disk.h | 3 + 3 files changed, 169 insertions(+), 141 deletions(-) diff --git a/src/storage/storage_backend.c b/src/storage/storage_backend.c index 060b1c0..18433e9 100644 --- a/src/storage/storage_backend.c +++ b/src/storage/storage_backend.c @@ -2851,157 +2851,30 @@ virStorageBackendBLKIDFindEmpty(const char *device ATTRIBUTE_UNUSED, #endif /* #if WITH_BLKID */ -typedef enum { - VIR_STORAGE_PARTED_ERROR = -1, - VIR_STORAGE_PARTED_MATCH, /* Valid label found and matches format */ - VIR_STORAGE_PARTED_DIFFERENT, /* Valid label found but not match format */ - VIR_STORAGE_PARTED_UNKNOWN, /* No or unrecognized label */ - VIR_STORAGE_PARTED_NOPTTYPE, /* Did not find the Partition Table type */ - VIR_STORAGE_PARTED_PTTYPE_UNK, /* Partition Table type unknown*/ -} virStorageBackendPARTEDResult; - -/** - * Check for a valid disk label (partition table) on device using - * the PARTED command - * - * returns virStorageBackendPARTEDResult - */ -static virStorageBackendPARTEDResult -virStorageBackendPARTEDFindLabel(const char *device, - const char *format) -{ - const char *const args[] = { - device, "print", "--script", NULL, - }; - virCommandPtr cmd = virCommandNew(PARTED); - char *output = NULL; - char *error = NULL; - char *start, *end; - int ret = VIR_STORAGE_PARTED_ERROR; - - virCommandAddArgSet(cmd, args); - virCommandAddEnvString(cmd, "LC_ALL=C"); - virCommandSetOutputBuffer(cmd, &output); - virCommandSetErrorBuffer(cmd, &error); - - /* if parted succeeds we have a valid partition table */ - ret = virCommandRun(cmd, NULL); - if (ret < 0) { - if ((output && strstr(output, "unrecognised disk label")) || - (error && strstr(error, "unrecognised disk label"))) { - ret = VIR_STORAGE_PARTED_UNKNOWN; - } - goto cleanup; - } - - /* Search for "Partition Table:" in the output. If not present, - * then we cannot validate the partition table type. - */ - if (!(start = strstr(output, "Partition Table: ")) || - !(end = strstr(start, "\n"))) { - VIR_DEBUG("Unable to find tag in output: %s", output); - ret = VIR_STORAGE_PARTED_NOPTTYPE; - goto cleanup; - } - start += strlen("Partition Table: "); - *end = '\0'; - - /* on disk it's "msdos", but we document/use "dos" so deal with it here */ - if (STREQ(start, "msdos")) - start += 2; - - /* Make sure we know about this type */ - if (virStoragePoolFormatDiskTypeFromString(start) < 0) { - ret = VIR_STORAGE_PARTED_PTTYPE_UNK; - goto cleanup; - } - - /* Does the on disk match what the pool desired? */ - if (STREQ(start, format)) - ret = VIR_STORAGE_PARTED_MATCH; - - ret = VIR_STORAGE_PARTED_DIFFERENT; - - cleanup: - virCommandFree(cmd); - VIR_FREE(output); - VIR_FREE(error); - return ret; -} - +#if WITH_STORAGE_DISK -/** - * Determine whether the label on the disk is valid or in a known format - * for the purpose of rewriting the label during build or being able to - * start a pool on a device. - * - * When 'writelabel' is true, if we find a valid disk label on the device, - * then we shouldn't be attempting to write as the volume may contain - * data. Force the usage of the overwrite flag to the build command in - * order to be certain. When the disk label is unrecognized, then it - * should be safe to write. - * - * When 'writelabel' is false, only if we find a valid disk label on the - * device should we allow the start since for this path we won't be - * rewriting the label. - * - * Return: 0 if it's OK - * -1 if something's wrong - */ static int virStorageBackendPARTEDValidLabel(const char *device, const char *format, bool writelabel) { - int ret = -1; - virStorageBackendPARTEDResult check; - - check = virStorageBackendPARTEDFindLabel(device, format); - switch (check) { - case VIR_STORAGE_PARTED_ERROR: - virReportError(VIR_ERR_OPERATION_FAILED, "%s", - _("Error checking for disk label, failed to get " - "disk partition information")); - break; - - case VIR_STORAGE_PARTED_MATCH: - if (writelabel) - virReportError(VIR_ERR_OPERATION_INVALID, - _("Disk label already formatted using '%s'"), - format); - else - ret = 0; - break; - - case VIR_STORAGE_PARTED_DIFFERENT: - virReportError(VIR_ERR_OPERATION_FAILED, "%s", - _("Known, but different label format present, " - "requires build --overwrite")); - break; - - case VIR_STORAGE_PARTED_UNKNOWN: - if (writelabel) - ret = 0; - else - virReportError(VIR_ERR_OPERATION_FAILED, "%s", - _("Unrecognized disk label found, requires build")); - break; - - case VIR_STORAGE_PARTED_NOPTTYPE: - virReportError(VIR_ERR_OPERATION_FAILED, "%s", - _("Unable to determine Partition Type, " - "requires build --overwrite")); - break; + return virStorageBackendDiskValidLabel(device, format, writelabel); +} - case VIR_STORAGE_PARTED_PTTYPE_UNK: - virReportError(VIR_ERR_OPERATION_FAILED, "%s", - _("Unknown Partition Type, requires build --overwrite")); - break; - } +#else - return ret; +static int +virStorageBackendPARTEDValidLabel(const char *device ATTRIBUTE_UNUSED, + const char *format ATTRIBUTE_UNUSED, + bool writelabel ATTRIBUTE_UNUSED) +{ + virReportError(VIR_ERR_OPERATION_INVALID, "%s", + _("PARTED is unsupported by this build")); + return -1; } +#endif /* #if WITH_STORAGE_DISK */ + /* virStorageBackendDeviceIsEmpty: * @devpath: Path to the device to check diff --git a/src/storage/storage_backend_disk.c b/src/storage/storage_backend_disk.c index eae6c90..04ffc6c 100644 --- a/src/storage/storage_backend_disk.c +++ b/src/storage/storage_backend_disk.c @@ -418,6 +418,158 @@ virStorageBackendDiskRefreshPool(virConnectPtr conn ATTRIBUTE_UNUSED, } +typedef enum { + VIR_STORAGE_PARTED_ERROR = -1, + VIR_STORAGE_PARTED_MATCH, /* Valid label found and matches format */ + VIR_STORAGE_PARTED_DIFFERENT, /* Valid label found but not match format */ + VIR_STORAGE_PARTED_UNKNOWN, /* No or unrecognized label */ + VIR_STORAGE_PARTED_NOPTTYPE, /* Did not find the Partition Table type */ + VIR_STORAGE_PARTED_PTTYPE_UNK, /* Partition Table type unknown*/ +} virStorageBackendPARTEDResult; + +/** + * Check for a valid disk label (partition table) on device using + * the PARTED command + * + * returns virStorageBackendPARTEDResult + */ +static virStorageBackendPARTEDResult +virStorageBackendPARTEDFindLabel(const char *device, + const char *format) +{ + const char *const args[] = { + device, "print", "--script", NULL, + }; + virCommandPtr cmd = virCommandNew(PARTED); + char *output = NULL; + char *error = NULL; + char *start, *end; + int ret = VIR_STORAGE_PARTED_ERROR; + + virCommandAddArgSet(cmd, args); + virCommandAddEnvString(cmd, "LC_ALL=C"); + virCommandSetOutputBuffer(cmd, &output); + virCommandSetErrorBuffer(cmd, &error); + + /* if parted succeeds we have a valid partition table */ + ret = virCommandRun(cmd, NULL); + if (ret < 0) { + if ((output && strstr(output, "unrecognised disk label")) || + (error && strstr(error, "unrecognised disk label"))) { + ret = VIR_STORAGE_PARTED_UNKNOWN; + } + goto cleanup; + } + + /* Search for "Partition Table:" in the output. If not present, + * then we cannot validate the partition table type. + */ + if (!(start = strstr(output, "Partition Table: ")) || + !(end = strstr(start, "\n"))) { + VIR_DEBUG("Unable to find tag in output: %s", output); + ret = VIR_STORAGE_PARTED_NOPTTYPE; + goto cleanup; + } + start += strlen("Partition Table: "); + *end = '\0'; + + /* on disk it's "msdos", but we document/use "dos" so deal with it here */ + if (STREQ(start, "msdos")) + start += 2; + + /* Make sure we know about this type */ + if (virStoragePoolFormatDiskTypeFromString(start) < 0) { + ret = VIR_STORAGE_PARTED_PTTYPE_UNK; + goto cleanup; + } + + /* Does the on disk match what the pool desired? */ + if (STREQ(start, format)) + ret = VIR_STORAGE_PARTED_MATCH; + + ret = VIR_STORAGE_PARTED_DIFFERENT; + + cleanup: + virCommandFree(cmd); + VIR_FREE(output); + VIR_FREE(error); + return ret; +} + + +/** + * Determine whether the label on the disk is valid or in a known format + * for the purpose of rewriting the label during build or being able to + * start a pool on a device. + * + * When 'writelabel' is true, if we find a valid disk label on the device, + * then we shouldn't be attempting to write as the volume may contain + * data. Force the usage of the overwrite flag to the build command in + * order to be certain. When the disk label is unrecognized, then it + * should be safe to write. + * + * When 'writelabel' is false, only if we find a valid disk label on the + * device should we allow the start since for this path we won't be + * rewriting the label. + * + * Return: 0 if it's OK + * -1 if something's wrong + */ +int +virStorageBackendDiskValidLabel(const char *device, + const char *format, + bool writelabel) +{ + int ret = -1; + virStorageBackendPARTEDResult check; + + check = virStorageBackendPARTEDFindLabel(device, format); + switch (check) { + case VIR_STORAGE_PARTED_ERROR: + virReportError(VIR_ERR_OPERATION_FAILED, "%s", + _("Error checking for disk label, failed to get " + "disk partition information")); + break; + + case VIR_STORAGE_PARTED_MATCH: + if (writelabel) + virReportError(VIR_ERR_OPERATION_INVALID, + _("Disk label already formatted using '%s'"), + format); + else + ret = 0; + break; + + case VIR_STORAGE_PARTED_DIFFERENT: + virReportError(VIR_ERR_OPERATION_FAILED, "%s", + _("Known, but different label format present, " + "requires build --overwrite")); + break; + + case VIR_STORAGE_PARTED_UNKNOWN: + if (writelabel) + ret = 0; + else + virReportError(VIR_ERR_OPERATION_FAILED, "%s", + _("Unrecognized disk label found, requires build")); + break; + + case VIR_STORAGE_PARTED_NOPTTYPE: + virReportError(VIR_ERR_OPERATION_FAILED, "%s", + _("Unable to determine Partition Type, " + "requires build --overwrite")); + break; + + case VIR_STORAGE_PARTED_PTTYPE_UNK: + virReportError(VIR_ERR_OPERATION_FAILED, "%s", + _("Unknown Partition Type, requires build --overwrite")); + break; + } + + return ret; +} + + static int virStorageBackendDiskStartPool(virConnectPtr conn ATTRIBUTE_UNUSED, virStoragePoolObjPtr pool) diff --git a/src/storage/storage_backend_disk.h b/src/storage/storage_backend_disk.h index aaabe62..79c0e6b 100644 --- a/src/storage/storage_backend_disk.h +++ b/src/storage/storage_backend_disk.h @@ -28,4 +28,7 @@ extern virStorageBackend virStorageBackendDisk; +int virStorageBackendDiskValidLabel(const char *device, + const char *format, + bool writelabel); #endif /* __VIR_STORAGE_BACKEND_DISK_H__ */ -- 2.7.4

2 1

[libvirt] [PATCH] rpc: libssh: don't crash when known_hosts file path is not provided
by Peter Krempa 11 Jan '17

11 Jan '17

When connecting as root, the "hostsfile" variable would be NULL due to the code leading to this point. This would result into a crash when attempting to set the known hosts file path. To avoid deviating from the approach taken in the libssh2 driver set the file to /dev/null so that all entries are discarded unless explicitly specified. Resolves: https://bugzilla.redhat.com/show_bug.cgi?id=1406457 --- Note that it would be much simpler just to skip ssh_options_set if 'hostsfile' is NULL. This would result in using /root/.ssh/known_hosts (according to the config) which would be different to the approach taken in libssh2. With libssh2 this can't be done (at least the last time I checked) as it happened to corrupt the file in some cases. src/rpc/virnetlibsshsession.c | 3 +++ 1 file changed, 3 insertions(+) diff --git a/src/rpc/virnetlibsshsession.c b/src/rpc/virnetlibsshsession.c index 5de6629d7..5fc16ba8a 100644 --- a/src/rpc/virnetlibsshsession.c +++ b/src/rpc/virnetlibsshsession.c @@ -1172,6 +1172,9 @@ virNetLibsshSessionSetHostKeyVerification(virNetLibsshSessionPtr sess, goto error; } + if (!hostsfile) + hostsfile = "/dev/null"; + /* set the known hosts file */ if (ssh_options_set(sess->session, SSH_OPTIONS_KNOWNHOSTS, hostsfile) < 0) goto error; -- 2.11.0

2 2

[libvirt] [PATCH 0/6] qemu: blockjob: make automatic disk locking less useless
by Peter Krempa 11 Jan '17

11 Jan '17

If automatic disk locking would be enabled with either lock manager the operations would most probably fail. This was mostly as libvirt did not bother to release images that were no longer written to by qemu, or tried re-lock the same image. This fixes the locking code for creation of external snapshots (both disk-only and with memory), block copy and active block commit. This series also kills the horrible "rollback" code in the external disk snapshot code. Regular block commit work will be tracked by: https://bugzilla.redhat.com/show_bug.cgi?id=1405537 Peter Krempa (6): qemu: blockcopy: Save monitor error prior to calling into lock manager locking: Fix documentation on how automatic sanlock leases are stored qemu: snapshot: Don't redetect backing chain after snapshot qemu: snapshot: Refactor snapshot rollback on failure qemu: snapshot: Properly handle image locking qemu: blockjob: Fix locking of block copy/active block commit src/locking/sanlock.conf | 2 +- src/qemu/qemu_blockjob.c | 26 ++-- src/qemu/qemu_driver.c | 389 +++++++++++++++++++++++++++-------------------- 3 files changed, 242 insertions(+), 175 deletions(-) -- 2.11.0

3 14