- Devel - libvirt List Archives

[Libvir] [PATCH] Add bus attribute to disk target definition
by Soren Hansen 29 Jul '08

29 Jul '08

Hi! I'd like to propose that the following patch gets applied against libvirt. It adds the option of putting a bus attribute on a disk target. To acommodate this, it also changes the way drives are defined for kvm from the old "-hda /path/to/file -boot c" style to the new "-drive file=/path/to/file,if=ide,boot=on". This makes it possible to specify virtio, scsi, and ide disks. === modified file 'src/qemu_conf.c' --- src/qemu_conf.c 2008-04-28 15:14:59 +0000 +++ src/qemu_conf.c 2008-04-29 07:43:11 +0000 @@ -568,6 +568,7 @@ xmlChar *source = NULL; xmlChar *target = NULL; xmlChar *type = NULL; + xmlChar *bus = NULL; int typ = 0; type = xmlGetProp(node, BAD_CAST "type"); @@ -598,6 +599,7 @@ } else if ((target == NULL) && (xmlStrEqual(cur->name, BAD_CAST "target"))) { target = xmlGetProp(cur, BAD_CAST "dev"); + bus = xmlGetProp(cur, BAD_CAST "bus"); } else if (xmlStrEqual(cur->name, BAD_CAST "readonly")) { disk->readonly = 1; } @@ -646,7 +648,9 @@ strcmp((const char *)target, "hda") && strcmp((const char *)target, "hdb") && strcmp((const char *)target, "hdc") && - strcmp((const char *)target, "hdd")) { + strcmp((const char *)target, "hdd") && + strcmp((const char *)target, "hdd") && + strncmp((const char *)target, "vd", 2)) { qemudReportError(conn, NULL, NULL, VIR_ERR_INTERNAL_ERROR, _("Invalid harddisk device name: %s"), target); goto error; @@ -673,6 +677,20 @@ goto error; } + if (!bus) + disk->bus = QEMUD_DISK_BUS_IDE; + else if (!strcmp((const char *)bus, "ide")) + disk->bus = QEMUD_DISK_BUS_IDE; + else if (!strcmp((const char *)bus, "scsi")) + disk->bus = QEMUD_DISK_BUS_SCSI; + else if (!strcmp((const char *)bus, "virtio")) + disk->bus = QEMUD_DISK_BUS_VIRTIO; + else { + qemudReportError(conn, NULL, NULL, VIR_ERR_INTERNAL_ERROR, "Invalid bus type: %s", bus); + goto error; + } + + xmlFree(bus); xmlFree(device); xmlFree(target); xmlFree(source); @@ -688,6 +706,8 @@ xmlFree(source); if (device) xmlFree(device); + if (bus) + xmlFree(bus); return -1; } @@ -1350,6 +1370,68 @@ return -1; } +static int qemudDiskCompare(const void *aptr, const void *bptr) { + struct qemud_vm_disk_def *a = (struct qemud_vm_disk_def *) aptr; + struct qemud_vm_disk_def *b = (struct qemud_vm_disk_def *) bptr; + if (a->device == b->device) + return virDiskNameToIndex(a->dst) - virDiskNameToIndex(b->dst); + else + return a->device - b->device; +} + +static const char *qemudBusIdToName(int busId) { + const char *busnames[] = { "ide", + "scsi", + "virtio" }; + + if (busId >= 0 && busId < 3) + return busnames[busId]; + else + return 0; +} + +static char *qemudDriveOpt(struct qemud_vm_disk_def *disk, int boot) +{ + char opt[PATH_MAX]; + + switch (disk->device) { + case QEMUD_DISK_CDROM: + snprintf(opt, PATH_MAX, "file=%s,if=ide,media=cdrom%s", + disk->src, boot ? ",boot=on" : ""); + break; + case QEMUD_DISK_FLOPPY: + snprintf(opt, PATH_MAX, "file=%s,if=floppy%s", + disk->src, boot ? ",boot=on" : ""); + break; + case QEMUD_DISK_DISK: + snprintf(opt, PATH_MAX, "file=%s,if=%s%s", + disk->src, qemudBusIdToName(disk->bus), boot ? ",boot=on" : ""); + break; + default: + return 0; + } + return strdup(opt); +} + +static char *qemudAddBootDrive(virConnectPtr conn, + struct qemud_vm_def *def, + char *handledDisks, + int type) { + int j = 0; + struct qemud_vm_disk_def *disk = def->disks; + + while (disk) { + if (!handledDisks[j] && disk->device == type) { + handledDisks[j] = 1; + return qemudDriveOpt(disk, 1); + } + j++; + disk = disk->next; + } + qemudReportError(conn, NULL, NULL, VIR_ERR_INTERNAL_ERROR, + "Requested boot device type %d, but no such device defined.", type); + return 0; +} /* * Parses a libvirt XML definition of a guest, and populates the @@ -1739,7 +1821,6 @@ obj = xmlXPathEval(BAD_CAST "/domain/devices/disk", ctxt); if ((obj != NULL) && (obj->type == XPATH_NODESET) && (obj->nodesetval != NULL) && (obj->nodesetval->nodeNr >= 0)) { - struct qemud_vm_disk_def *prev = NULL; for (i = 0; i < obj->nodesetval->nodeNr; i++) { struct qemud_vm_disk_def *disk = calloc(1, sizeof(*disk)); if (!disk) { @@ -1752,13 +1833,20 @@ goto error; } def->ndisks++; - disk->next = NULL; if (i == 0) { + disk->next = NULL; def->disks = disk; } else { - prev->next = disk; + struct qemud_vm_disk_def *ptr = def->disks; + while (ptr) { + if (!ptr->next || qemudDiskCompare(ptr->next, disk) < 0) { + disk->next = ptr->next; + ptr->next = disk; + break; + } + ptr = ptr->next; + } } - prev = disk; } } xmlXPathFreeObject(obj); @@ -2207,30 +2295,32 @@ goto no_memory; } - for (i = 0 ; i < vm->def->os.nBootDevs ; i++) { - switch (vm->def->os.bootDevs[i]) { - case QEMUD_BOOT_CDROM: - boot[i] = 'd'; - break; - case QEMUD_BOOT_FLOPPY: - boot[i] = 'a'; - break; - case QEMUD_BOOT_DISK: - boot[i] = 'c'; - break; - case QEMUD_BOOT_NET: - boot[i] = 'n'; - break; - default: - boot[i] = 'c'; - break; + if (vm->def->virtType != QEMUD_VIRT_KVM) { + for (i = 0 ; i < vm->def->os.nBootDevs ; i++) { + switch (vm->def->os.bootDevs[i]) { + case QEMUD_BOOT_CDROM: + boot[i] = 'd'; + break; + case QEMUD_BOOT_FLOPPY: + boot[i] = 'a'; + break; + case QEMUD_BOOT_DISK: + boot[i] = 'c'; + break; + case QEMUD_BOOT_NET: + boot[i] = 'n'; + break; + default: + boot[i] = 'c'; + break; + } } + boot[vm->def->os.nBootDevs] = '\0'; + if (!((*argv)[++n] = strdup("-boot"))) + goto no_memory; + if (!((*argv)[++n] = strdup(boot))) + goto no_memory; } - boot[vm->def->os.nBootDevs] = '\0'; - if (!((*argv)[++n] = strdup("-boot"))) - goto no_memory; - if (!((*argv)[++n] = strdup(boot))) - goto no_memory; if (vm->def->os.kernel[0]) { if (!((*argv)[++n] = strdup("-kernel"))) @@ -2251,28 +2341,74 @@ goto no_memory; } - while (disk) { - char dev[NAME_MAX]; - char file[PATH_MAX]; - if (!strcmp(disk->dst, "hdc") && - disk->device == QEMUD_DISK_CDROM) { - if (disk->src[0]) - snprintf(dev, NAME_MAX, "-%s", "cdrom"); - else { - /* Don't put anything on the cmdline for an empty cdrom*/ - disk = disk->next; - continue; - } - } else - snprintf(dev, NAME_MAX, "-%s", disk->dst); - snprintf(file, PATH_MAX, "%s", disk->src); - - if (!((*argv)[++n] = strdup(dev))) - goto no_memory; - if (!((*argv)[++n] = strdup(file))) - goto no_memory; - - disk = disk->next; + if (vm->def->virtType == QEMUD_VIRT_KVM) { + char *handledDisks = NULL; + int j; + + handledDisks = calloc(sizeof(*handledDisks), vm->def->ndisks); + + if (!handledDisks) + goto no_memory; + + /* When using -drive notation, we need to provide the devices in boot + * preference order. */ + for (i = 0 ; i < vm->def->os.nBootDevs ; i++) { + if (!((*argv)[++n] = strdup("-drive"))) + goto no_memory; + + switch (vm->def->os.bootDevs[i]) { + case QEMUD_BOOT_CDROM: + if (!((*argv)[++n] = qemudAddBootDrive(conn, vm->def, handledDisks, QEMUD_DISK_CDROM))) + goto error; + break; + case QEMUD_BOOT_FLOPPY: + if (!((*argv)[++n] = qemudAddBootDrive(conn, vm->def, handledDisks, QEMUD_DISK_FLOPPY))) + goto error; + break; + case QEMUD_BOOT_DISK: + if (!((*argv)[++n] = qemudAddBootDrive(conn, vm->def, handledDisks, QEMUD_DISK_DISK))) + goto error; + break; + } + } + + /* Pick up the rest of the devices */ + j=0; + while (disk) { + if (!handledDisks[j]) { + handledDisks[j] = 1; + if (!((*argv)[++n] = strdup("-drive"))) + goto no_memory; + if (!((*argv)[++n] = qemudDriveOpt(disk, 0))) + goto no_memory; + } + disk = disk->next; + j++; + } + } else { + while (disk) { + char dev[NAME_MAX]; + char file[PATH_MAX]; + + if (!strcmp(disk->dst, "hdc") && + disk->device == QEMUD_DISK_CDROM) + if (disk->src[0]) + snprintf(dev, NAME_MAX, "-%s", "cdrom"); + else { + disk = disk->next; + continue; + } + else + snprintf(dev, NAME_MAX, "-%s", disk->dst); + snprintf(file, PATH_MAX, "%s", disk->src); + + if (!((*argv)[++n] = strdup(dev))) + goto no_memory; + if (!((*argv)[++n] = strdup(file))) + goto no_memory; + + disk = disk->next; + } } if (!net) { @@ -3565,6 +3701,7 @@ virBufferVSprintf(&buf, " <source %s='%s'/>\n", typeAttrs[disk->type], disk->src); + virBufferVSprintf(&buf, " <target dev='%s' bus='%s'/>\n", disk->dst, qemudBusIdToName(disk->bus)); virBufferVSprintf(&buf, " <target dev='%s'/>\n", disk->dst); if (disk->readonly) === modified file 'src/qemu_conf.h' --- src/qemu_conf.h 2008-04-25 20:46:13 +0000 +++ src/qemu_conf.h 2008-04-29 07:13:16 +0000 @@ -56,10 +56,17 @@ QEMUD_DISK_FLOPPY, }; +enum qemud_vm_disk_bus { + QEMUD_DISK_BUS_IDE, + QEMUD_DISK_BUS_SCSI, + QEMUD_DISK_BUS_VIRTIO +}; + /* Stores the virtual disk configuration */ struct qemud_vm_disk_def { int type; int device; + int bus; char src[PATH_MAX]; char dst[NAME_MAX]; int readonly; === modified file 'src/util.c' --- src/util.c 2008-04-25 14:53:05 +0000 +++ src/util.c 2008-04-29 06:59:49 +0000 @@ -771,3 +771,43 @@ return -1; } + +/* Translates a device name of the form (regex) "[fhv]d[a-z]+" into + * the corresponding index (e.g. sda => 1, hdz => 26, vdaa => 27) + * @param name The name of the device + * @return name's index, or 0 on failure + */ +int virDiskNameToIndex(const char *name) { + const char *ptr = NULL; + int idx = 0; + + if (strlen(name) < 3) + return 0; + + switch (*name) { + case 'f': + case 'h': + case 'v': + break; + default: + return 0; + } + + if (*(name + 1) != 'd') + return 0; + + ptr = name+2; + + while (*ptr) { + idx = idx * 26; + + if ('a' > *ptr || 'z' < *ptr) + return 0; + + idx += *ptr - 'a' + 1; + ptr++; + } + + return idx; +} + === modified file 'src/util.h' --- src/util.h 2008-04-25 14:53:05 +0000 +++ src/util.h 2008-04-29 06:59:57 +0000 @@ -92,4 +92,6 @@ int virParseMacAddr(const char* str, unsigned char *addr); +int virDiskNameToIndex(const char* str); + #endif /* __VIR_UTIL_H__ */ -- Soren Hansen | Virtualisation specialist | Ubuntu Server Team Canonical Ltd. | http://www.ubuntu.com/

6 17

[libvirt] KVM migration patch (2008-07-29)
by Richard W.M. Jones 29 Jul '08

29 Jul '08

This is just a rebase of the KVM migration patch to the latest libvirt CVS: http://www.annexia.org/tmp/libvirt-kvm-migrate-20080729.patch Rich. -- Richard Jones, Emerging Technologies, Red Hat http://et.redhat.com/~rjones Read my OCaml programming blog: http://camltastic.blogspot.com/ Fedora now supports 59 OCaml packages (the OPEN alternative to F#) http://cocan.org/getting_started_with_ocaml_on_red_hat_and_fedora

1 0

[libvirt] [PATCH]: Fix Qemu CD-ROM with no source
by Chris Lalancette 29 Jul '08

29 Jul '08

The new generic domain re-factor introduced a small regression into the drive handling code. In particular, if you had a section of XML like: <disk type='file' device='cdrom'> <target dev='hdc' bus='ide'/> <readonly/> </disk> This used to work with older libvirt, but now fails. This is because we are actually passing the literal string (null) to the qemu command-line, which qemu barfs on. This patch fixes it up by making it blank, which allows qemu to continue on it's merry way. Signed-off-by: Chris Lalancette <clalance(a)redhat.com>

3 3

[libvirt] Now ESXi is 'free'
by Stefan de Konink 29 Jul '08

29 Jul '08

Now ESXi is 'free as in beer' would it be appropriate to hack a libvirtd in it? Stefan

2 1

[libvirt] [PATCH] OpenVZ xml refactoring
by Evgeniy Sokolov 29 Jul '08

29 Jul '08

Patch switch OpenVZ XML format to generic. main changes: - I used generic virDomainNetDef to define network in container. And wrote function to apply virDomainNetDef for container. Method virDomainNetDefParseXML is public now. - tag "container" is changed to "devices" - changed format of tag "filesystem" - added processing "vcpu" tag.

3 7

[libvirt] [PATCH]: qemu_driver: convert uuid to string in error messages
by Guido Günther 28 Jul '08

28 Jul '08

Hi, qemu_driver misses some conversions to a char* before printing the uuid. Possible fix attached. Cheers, -- Guido

3 6

[libvirt] Re: kexec/kdump of a kvm guest?
by Mike Snitzer 27 Jul '08

27 Jul '08

On Thu, Jul 24, 2008 at 9:15 AM, Vivek Goyal <vgoyal(a)redhat.com> wrote: > On Thu, Jul 24, 2008 at 07:49:59AM -0400, Mike Snitzer wrote: >> On Thu, Jul 24, 2008 at 4:39 AM, Alexander Graf <agraf(a)suse.de> wrote: >> > As you're stating that the host kernel breaks with kvm modules loaded, maybe >> > someone there could give a hint. >> >> OK, I can try using a newer kernel on the host too (e.g. 2.6.25.x) to >> see how kexec/kdump of the host fairs when kvm modules are loaded. >> >> On the guest side of things, as I mentioned in my original post, >> kexec/kdump wouldn't work within a 2.6.22.19 guest with the host >> running 2.6.25.4 (with kvm-70). >> > > Hi Mike, > > I have never tried kexec/kdump inside a kvm guest. So I don't know if > historically they have been working or not. Avi indicated he seems to remember that at least kexec worked last he tried (didn't provide when/what he tried though). > Having said that, Why do we need kdump to work inside the guest? In this > case qemu should be knowing about the memory of guest kernel and should > be able to capture a kernel crash dump? I am not sure if qemu already does > that. If not, then probably we should think about it? > > To me, kdump is a good solution for baremetal but not for virtualized > environment where we already have another piece of software running which > can do the job for us. We will end up wasting memory in every instance > of guest (memory reserved for kdump kernel in every guest). I haven't looked into what mechanics qemu provides for collecting the entire guest memory image; I'll dig deeper at some point. It seems the libvirt mid-layer ("virsh dump" - dump the core of a domain to a file for analysis) doesn't support saving a kvm guest core: # virsh dump guest10 guest10.dump libvir: error : this function is not supported by the hypervisor: virDomainCoreDump error: Failed to core dump domain guest10 to guest10.dump Seems that libvirt functionality isn't available yet with kvm (I'm using libvirt 0.4.2, I'll give libvirt 0.4.4 a try). cc'ing the libvirt-list to get their insight. That aside, having the crash dump collection be multi-phased really isn't workable (that is if it requires a crashed guest to be manually saved after the fact). The host system _could_ be rebooted; whereby losing the guest's core image. So automating qemu and/or libvirtd to trigger a dump would seem worthwhile (maybe its already done?). So while I agree with you its ideal to not have to waste memory in each guest for the purposes of kdump; if users want to model a guest image as closely as possible to what will be deployed on bare metal it really would be ideal to support a 1:1 functional equivalent with kvm. I work with people who refuse to use kvm because of the lack of kexec/kdump support. I can do further research but welcome others' insight: do others have advice on how best to collect a crashed kvm guest's core? > It will be interesting to look at your results with 2.6.25.x kernels with > kvm module inserted. Currently I can't think what can possibly be wrong. If the host's 2.6.25.4 kernel has both the kvm and kvm-intel modules loaded kexec/kdump does _not_ work (simply hangs the system). If I only have the kvm module loaded kexec/kdump works as expected (likewise if no kvm modules are loaded at all). So it would appear that kvm-intel and kexec are definitely mutually exclusive at the moment (at least on both 2.6.22.x and 2.6.25.x). Mike

4 4

[libvirt] virsh
by drew einhorn 25 Jul '08

25 Jul '08

Hi, I have virsh running in a remote ssh session in a dom0 xen VM, it is presenting a virtual console for a domU vm running a RHEL5 installer The RHEL5 is going astray setting up the network using DHCP, I want to do Ctrl-Alt-Fn to view the installer log files, and access a shell prompt. But the Ctrl-Alt-Fn is being consumed by the laptop running the remote ssh ssh session. How do I get the virsh to send Ctrl-Alt-Fn to the domU console? -- Drew Einhorn

1 0

[libvirt] [PATCH]: Utterly trivial comment cleanups in src/domain_conf.c
by Chris Lalancette 25 Jul '08

25 Jul '08

The subject pretty much says it all, except for the minor fact of using "domain.xml" in the xmlReadDoc(). This cleanup really just makes the code easier to read. Signed-off-by: Chris Lalancette <clalance(a)redhat.com>

3 2

[libvirt] Problems when using libvirt-cim provider
by SHAN, CHUN -HCHBJ 25 Jul '08

25 Jul '08

Hi,all: I am trying to use libvirt API and libvirt-cim now. After installed them successfully, I can see the "Virt_*****" providers using "cimprovider -l -s"commond. My cimom is pegasus. Then I used wbemcli as cim client. But when I used "wbemcli ei" to enumerate instances of a specified class,for example,Xen_DiskPool, error code was: wbemcli:Cim(1) CIM_ERR_FAILED: A general error occurred that is not covered by a more specific error code:"For provider Virt_DevicePool the library name was empty. Check provider registered location". Could you please help me to solve this problem? Did it mean that the providers were not registered correctly? But I "make postinstall" the provider and no errors occourred. Thank you! Regards, Chun Shan Disclaimer: The contents of this e-mail, and its attachments, if any, are confidential and may be protected by law against any unauthorized use. If you have received this e-mail by mistake or have reason to believe that you are not the intended recipient, please notify the sender by reply e-mail as soon as possible and delete it from your computer system immediately thereafter. If you are not the intended recipient, you must not copy this e-mail or attachment or disclose the contents to any other person. While we have made every effort to keep our network virus free, we take no responsibility for any computer virus which might be transferred by way of this e-mail.

3 2