- Devel - libvirt List Archives

[libvirt] [PATCH] Unify most error reporting (ver 2)
by Cole Robinson 10 Oct '08

10 Oct '08

The attached patch is a more complete cut at unifying error reporting among the different source files. Most files have their own custom Error function, each with varying degrees of functionality, and lots of code duplication. The attached patch adds a helper function to virterror.c that centralizes all this logic, and redefines each local error function as a macro that calls out to the helper. The fixed files now offer printf style reporting, and the macros pass off function name, file name, and line number of the reported error (currently not used, but handy to have). This change centralizes probably 90% of the error calls. Some files have pretty custom error functions that don't map easily to the helper, or call __virRaiseError directly in a number of places, so I skipped these for now. Eventually we should move all these edge cases over so any src/ wide error changes will be pretty easy to make. The one thing we lose here is that some some places were logging info in the err{1,2,3} and int{1,2} fields of raised error. I don't consider this a loss: even where it was used, it was used inconsistently and rarely for anything that would be useful to a user. In the places where the data had some value, I included it in the error string. One side question: is src/xmlrpc.{c,h} even used? And if not, can it be removed? Thanks, Cole domain_conf.c | 27 +----------- hash.c | 26 +----------- internal.h | 6 ++ lxc_conf.c | 21 --------- lxc_conf.h | 7 +-- network_conf.c | 23 +--------- openvz_conf.c | 22 ---------- openvz_conf.h | 7 ++- proxy_internal.c | 23 +--------- qemu_conf.c | 21 --------- qemu_conf.h | 8 +-- qparams.c | 10 +--- sexpr.c | 25 +---------- storage_conf.c | 19 --------- storage_conf.h | 7 +-- test.c | 28 ++++--------- util.c | 27 +----------- virterror.c | 40 ++++++++++++++++++ xen_internal.c | 116 ++++++++++++++++++------------------------------------- xen_unified.c | 24 ++--------- xend_internal.c | 59 ++------------------------- xm_internal.c | 28 ++----------- xml.c | 43 +++++--------------- xs_internal.c | 23 +--------- 24 files changed, 164 insertions(+), 476 deletions(-)

4 12

[libvirt] libvirt - Configure fails with error message
by Ricardo Tiago 10 Oct '08

10 Oct '08

Hi, When i try to run configure it fails asking me to install the gnutls library, which is already installed. Output from configure: checking for struct _xmlURI.query_raw... no checking gnutls/gnutls.h usability... yes checking gnutls/gnutls.h presence... yes checking for gnutls/gnutls.h... yes checking for gnutls_handshake in -lgnutls... no configure: error: You must install the GnuTLS library in order to compile and run libvirt [user1@xxx libvirt-0.4.6]# rpm -qa | grep gnutls gnutls-devel-1.4.1-3.el5_1 gnutls-1.4.1-3.el5_1 After some checking in configure.in I verified that what was really missing is libgcrypt-devel. I believe the error message should be updated and provide more details or at least a check for libgcrypt should be done before the gnutls check. Ricardo

2 1

[libvirt] [PATCH 0/2] Update MinGW specfile and add NSIS support to nightly builds
by Richard W.M. Jones 10 Oct '08

10 Oct '08

This patch is in two parts. The first part just updates the MinGW specfile to the latest version. Note that it is also renamed (mingw-* -> mingw32-*) to match the approved Fedora MinGW packaging guidelines. The second part adds NSIS support. You will need mingw32-nsis and mingw32-nsiswrapper installed, and then you'll get a nightly Windows installer containing (at the moment) just libvirt-0.dll, virsh.exe and some documentation and dependencies. Rich. -- Richard Jones, Emerging Technologies, Red Hat http://et.redhat.com/~rjones virt-top is 'top' for virtual machines. Tiny program with many powerful monitoring features, net stats, disk stats, logging, etc. http://et.redhat.com/~rjones/virt-top

3 13

[libvirt] PATCH: Disable QEMU drive caching
by Daniel P. Berrange 10 Oct '08

10 Oct '08

QEMU defaults to allowing the host OS to cache all disk I/O. THis has a couple of problems - It is a waste of memory because the guest already caches I/O ops - It is unsafe on host OS crash - all unflushed guest I/O will be lost, and there's no ordering guarentees, so metadata updates could be flushe to disk, while the journal updates were not. Say goodbye to your filesystem. - It makes benchmarking more or less impossible / worthless because what the benchmark things are disk writes just sit around in memory so guest disk performance appears to exceed host diskperformance. This patch disables caching on all QEMU guests. NB, Xen has long done this for both PV & HVM guests - QEMU only gained this ability when -drive was introduced, and sadly kept the default to unsafe cache=on settings. Daniel diff -r 4a0ccc9dc530 src/qemu_conf.c --- a/src/qemu_conf.c Wed Oct 08 11:53:45 2008 +0100 +++ b/src/qemu_conf.c Wed Oct 08 11:59:33 2008 +0100 @@ -460,6 +460,8 @@ flags |= QEMUD_CMD_FLAG_DRIVE; if (strstr(help, "boot=on")) flags |= QEMUD_CMD_FLAG_DRIVE_BOOT; + if (strstr(help, "cache=on")) + flags |= QEMUD_CMD_FLAG_DRIVE_CACHE; if (version >= 9000) flags |= QEMUD_CMD_FLAG_VNC_COLON; @@ -959,13 +961,15 @@ break; } - snprintf(opt, PATH_MAX, "file=%s,if=%s,%sindex=%d%s", + snprintf(opt, PATH_MAX, "file=%s,if=%s,%sindex=%d%s%s", disk->src ? disk->src : "", bus, media ? media : "", idx, bootable && disk->device == VIR_DOMAIN_DISK_DEVICE_DISK - ? ",boot=on" : ""); + ? ",boot=on" : "", + qemuCmdFlags & QEMUD_CMD_FLAG_DRIVE_BOOT + ? ",cache=off" : ""); ADD_ARG_LIT("-drive"); ADD_ARG_LIT(opt); diff -r 4a0ccc9dc530 src/qemu_conf.h --- a/src/qemu_conf.h Wed Oct 08 11:53:45 2008 +0100 +++ b/src/qemu_conf.h Wed Oct 08 11:59:33 2008 +0100 @@ -44,7 +44,8 @@ QEMUD_CMD_FLAG_NO_REBOOT = (1 << 2), QEMUD_CMD_FLAG_DRIVE = (1 << 3), QEMUD_CMD_FLAG_DRIVE_BOOT = (1 << 4), - QEMUD_CMD_FLAG_NAME = (1 << 5), + QEMUD_CMD_FLAG_DRIVE_CACHE = (1 << 5), + QEMUD_CMD_FLAG_NAME = (1 << 6), }; /* Main driver state */ -- |: Red Hat, Engineering, London -o- http://people.redhat.com/berrange/ :| |: http://libvirt.org -o- http://virt-manager.org -o- http://ovirt.org :| |: http://autobuild.org -o- http://search.cpan.org/~danberr/ :| |: GnuPG: 7D3B9505 -o- F3C9 553F A1DA 4AC2 5648 23C1 B3DF F742 7D3B 9505 :|

7 25

[libvirt] [PATCH] cleanup virDomainCreateLinux into virDomainDefineXML
by Daniel Veillard 10 Oct '08

10 Oct '08

As promised in the libvirt-qpid thread, virDomainCreateLinux() call name makes no sense (anymore), and it should be renamed, i guess since virDomainCreate() already exist and by similarity with virDomainDefineXML() the best name is virDomainCreateXML(). The associated patch rename virDomainCreateLinux to virDomainCreateXML create a small function virDomainCreateLinux calling the former, document it as deprecated. To help comprehension of the source code it's also best to rename the internal driver method in a similar way, which inflates the patch a bit but is IMHO worth it. The patch also fixes a few #elif define(__sun__) and turn them into the correct #elif defined(__sun__) cpp instructions, and changes include/libvirt/virterror.h to improve the generated HTML page about the deprected fileds in the structure, since we need to regenerate the API xml and the docs, it was a good opportunity for that small change. I removed the docs/ subdir part from the patch as it's generated and mostly unreadable, that keeps it smaller too. Daniel -- Daniel Veillard | libxml Gnome XML XSLT toolkit http://xmlsoft.org/ daniel(a)veillard.com | Rpmfind RPM search engine http://rpmfind.net/ http://veillard.com/ | virtualization library http://libvirt.org/

2 2

[libvirt] [PATCH] fix index creation for disks > {sd,hd,xvd,vd}z
by Chris Wright 10 Oct '08

10 Oct '08

Calling virDiskNameToIndex with a disk name > {sd,hd,xvd,vd}z, such as vdaa, generates a bogus index. Account for iterations through the loop. Old behaviour: vda -> 0 vdz -> 25 vdaa -> 0 vdaz -> 25 New behaviour: vda -> 0 vdz -> 25 vdaa -> 26 vdaz -> 51 This was discovered by Sanjay Rao, thanks for the report. Signed-off-by: Chris Wright <chrisw(a)redhat.com> Cc: Mark Wagner <mwagner(a)redhat.com> Cc: Sanjay Rao <srao(a)redhat.com> --- src/domain_conf.c | 2 +- src/util.c | 6 +++--- 2 files changed, 4 insertions(+), 4 deletions(-) diff --git a/src/domain_conf.c b/src/domain_conf.c index 6a35064..b5897a7 100644 --- a/src/domain_conf.c +++ b/src/domain_conf.c @@ -3395,7 +3395,7 @@ char *virDomainConfigFile(virConnectPtr conn, /* Translates a device name of the form (regex) "[fhv]d[a-z]+" into * the corresponding bus,index combination (e.g. sda => (0,0), sdi (1,1), - * hdd => (1,1), vdaa => (0,27)) + * hdd => (1,1), vdaa => (0,26)) * @param disk The disk device * @param busIdx parsed bus number * @param devIdx parsed device number diff --git a/src/util.c b/src/util.c index ca14be1..56390d7 100644 --- a/src/util.c +++ b/src/util.c @@ -1055,7 +1055,7 @@ const char *virEnumToString(const char *const*types, } /* Translates a device name of the form (regex) "[fhv]d[a-z]+" into - * the corresponding index (e.g. sda => 1, hdz => 26, vdaa => 27) + * the corresponding index (e.g. sda => 0, hdz => 25, vdaa => 26) * @param name The name of the device * @return name's index, or -1 on failure */ @@ -1075,8 +1075,8 @@ int virDiskNameToIndex(const char *name) { if (!ptr) return -1; - while (*ptr) { - idx = idx * 26; + for (i = 0; *ptr; i++) { + idx = (idx + i) * 26; if (!c_islower(*ptr)) return -1;

3 2

[libvirt] Libvirt.org will be down in an hour till tomorrow morning
by Daniel Veillard 10 Oct '08

10 Oct '08

There is major electric change at the hosting place for libvirt.org and the server will need to be shutdown for the night (french time, i.e. beginning of the afternoon US EST time). Hopefully the services will be restarted tomorrow morning, sorry for the disruption and the late announcement, I just learned about it... Daniel -- Daniel Veillard | libxml Gnome XML XSLT toolkit http://xmlsoft.org/ daniel(a)veillard.com | Rpmfind RPM search engine http://rpmfind.net/ http://veillard.com/ | virtualization library http://libvirt.org/

2 2

[libvirt] Xen and out of memory stuff, blocks libvirtd
by Stefan de Konink 09 Oct '08

09 Oct '08

-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA512 [2008-10-08 19:14:05 13001] ERROR (XendCheckpoint:157) Save failed on domain klant1_monetdb (17) - resuming. Traceback (most recent call last): File "usr/lib64/python2.5/site-packages/xen/xend/XendCheckpoint.py", line 125, in save forkHelper(cmd, fd, saveInputHandler, False) File "usr/lib64/python2.5/site-packages/xen/xend/XendCheckpoint.py", line 358, in forkHelper child = xPopen3(cmd, True, -1, [fd, xc.handle()]) File "usr/lib64/python2.5/site-packages/xen/util/xpopen.py", line 100, in __init__ self.pid = os.fork() OSError: [Errno 12] Cannot allocate memory [2008-10-08 19:14:05 13001] DEBUG (XendDomainInfo:2456) XendDomainInfo.resumeDomain(17) After this point, I am unable to contact the libvird daemon. Stefan -----BEGIN PGP SIGNATURE----- Version: GnuPG v2.0.9 (GNU/Linux) Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org iEYEAREKAAYFAkjs7qQACgkQYH1+F2Rqwn0heQCeNJWDo2h3BR1IhcXmt+w9tpjD 304An1SlqWbig7nL8jgeElzJgu/KqgIa =dl3M -----END PGP SIGNATURE-----

2 7

[libvirt] [PATCH] Unify *ReportError logic
by Cole Robinson 08 Oct '08

08 Oct '08

Currently, most src/* files have their own ReportError function. Some support printf style arguments, others only allow reporting a single string message. The code for all of them does virtually the same thing, possibly passing a different constant off to another function. The attached patch adds a function to virterror.c which encapsulates the common ReportError logic. I used this to replace qemudReportError with a macro, which also allows passing off filename and line number info if we wanted to do something with it later. I did just the one function conversion to see what people think: if I'm missing something, or ideas for anything else to add. Seems to work as expected in my testing. Thanks, Cole

3 5

[libvirt] first argument in virDomainMigrate, is it free'ed?
by Stefan de Konink 08 Oct '08

08 Oct '08

-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA512 Hi, I saw that freeing the dom after a migration takes place results in: libvir: Domain error : invalid domain pointer in virDomainFree Now I wonder, if this is the case, could a normale migrate function 'just' be: domain = virDomainMigrate(domain, ....) Or is this unsafe; for example as a domain migration fails? Stefan -----BEGIN PGP SIGNATURE----- Version: GnuPG v2.0.9 (GNU/Linux) Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org iEYEAREKAAYFAkjqOYwACgkQYH1+F2Rqwn0NZACffFESznbeElKNVgt65YXPGzXX ausAmwZ7HScE0/RTtDWm2uug/dC1S21K =xpn6 -----END PGP SIGNATURE-----

3 3