- Devel - libvirt List Archives

[libvirt] [PATCH] udevEnumerateDevices: remove dead code
by Jim Meyering 01 Mar '10

01 Mar '10

Clang reported the dead store. >From 0557e23ca2f5eb78ed9f260040895011785f670f Mon Sep 17 00:00:00 2001 From: Jim Meyering <meyering(a)redhat.com> Date: Mon, 1 Mar 2010 16:14:35 +0100 Subject: [PATCH] udevEnumerateDevices: remove dead code * src/node_device/node_device_udev.c (udevEnumerateDevices): Remove unnecessary call to udev_list_entry_get_name. --- src/node_device/node_device_udev.c | 4 +--- 1 files changed, 1 insertions(+), 3 deletions(-) diff --git a/src/node_device/node_device_udev.c b/src/node_device/node_device_udev.c index 68509f4..eee44c4 100644 --- a/src/node_device/node_device_udev.c +++ b/src/node_device/node_device_udev.c @@ -1,11 +1,11 @@ /* * node_device_udev.c: node device enumeration - libudev implementation * - * Copyright (C) 2009 Red Hat + * Copyright (C) 2009-2010 Red Hat * * 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 library is distributed in the hope that it will be useful, @@ -1333,30 +1333,28 @@ static int udevProcessDeviceListEntry(struct udev *udev, } static int udevEnumerateDevices(struct udev *udev) { struct udev_enumerate *udev_enumerate = NULL; struct udev_list_entry *list_entry = NULL; - const char *name = NULL; int ret = 0; udev_enumerate = udev_enumerate_new(udev); ret = udev_enumerate_scan_devices(udev_enumerate); if (0 != ret) { VIR_ERROR("udev scan devices returned %d", ret); goto out; } udev_list_entry_foreach(list_entry, udev_enumerate_get_list_entry(udev_enumerate)) { udevProcessDeviceListEntry(udev, list_entry); - name = udev_list_entry_get_name(list_entry); } out: udev_enumerate_unref(udev_enumerate); return ret; } -- 1.7.0.1.414.g89213d

2 1

[libvirt] [PATCH] cmdPoolDiscoverSources: initialize earlier to avoid FP from clang
by Jim Meyering 01 Mar '10

01 Mar '10

This change hoists the initialization of srcSpec to the top, to keep clang from reporting a false-positive used-uninitialized problem. >From f487655124c887310be3bc82d7beae3cd47e7d21 Mon Sep 17 00:00:00 2001 From: Jim Meyering <meyering(a)redhat.com> Date: Mon, 1 Mar 2010 15:41:15 +0100 Subject: [PATCH] cmdPoolDiscoverSources: initialize earlier to avoid FP from clang * tools/virsh.c (cmdPoolDiscoverSources): Always initialize srcSpec. Otherwise, clang would report that srcSpec could be used uninitialized in the call to virConnectFindStoragePoolSources. --- tools/virsh.c | 7 +++---- 1 files changed, 3 insertions(+), 4 deletions(-) diff --git a/tools/virsh.c b/tools/virsh.c index 3b1011c..191e7ba 100644 --- a/tools/virsh.c +++ b/tools/virsh.c @@ -4431,17 +4431,16 @@ static const vshCmdOptDef opts_find_storage_pool_sources[] = { static int cmdPoolDiscoverSources(vshControl * ctl, const vshCmd * cmd ATTRIBUTE_UNUSED) { - char *type, *srcSpec, *srcSpecFile, *srcList; + char *type, *srcSpecFile, *srcList; + char *srcSpec = NULL; int found; type = vshCommandOptString(cmd, "type", &found); if (!found) return FALSE; srcSpecFile = vshCommandOptString(cmd, "srcSpec", &found); - if (!found) { + if (!found) srcSpecFile = NULL; - srcSpec = NULL; - } if (!vshConnectionUsability(ctl, ctl->conn, TRUE)) return FALSE; -- 1.7.0.1.414.g89213d

2 1

[libvirt] [PATCH] qemu: avoid null dereference on failed migration
by Eric Blake 01 Mar '10

01 Mar '10

From: Eric Blake <eblake(a)redhat.com> * src/qemu/qemu_monitor_text.c (qemuMonitorTextGetMigrationStatus): Check for failed strchr, to silence a coverity warning. Signed-off-by: Eric Blake <eblake(a)redhat.com> --- src/qemu/qemu_monitor_text.c | 5 +++++ 1 files changed, 5 insertions(+), 0 deletions(-) diff --git a/src/qemu/qemu_monitor_text.c b/src/qemu/qemu_monitor_text.c index 62ffcc6..e7b4b1f 100644 --- a/src/qemu/qemu_monitor_text.c +++ b/src/qemu/qemu_monitor_text.c @@ -989,6 +989,11 @@ int qemuMonitorTextGetMigrationStatus(qemuMonitorPtr mon, if ((tmp = strstr(reply, MIGRATION_PREFIX)) != NULL) { tmp += strlen(MIGRATION_PREFIX); end = strchr(tmp, '\r'); + if (end == NULL) { + qemuReportError(VIR_ERR_INTERNAL_ERROR, + _("unexpected migration status in %s"), reply); + goto cleanup; + } *end = '\0'; if ((*status = qemuMonitorMigrationStatusTypeFromString(tmp)) < 0) { -- 1.6.6.1

3 2

[libvirt] [PATCH] Do not search xenstore for disk device IDs
by Jim Fehlig 01 Mar '10

01 Mar '10

Disk devices can be referenced by name in Xen, e.g. when modifying their configuration or remvoving them. As such, don't search xenstore for a device ID corresponding to the disk device. Instead, search the disks contained in the domain definition and use the disk's target name if found. This approach allows removing a disk when domain is inactive. We obviously can't search xenstore when the domain is inactive. --- src/xen/xend_internal.c | 38 +++++++++++++++++++++++--------------- 1 files changed, 23 insertions(+), 15 deletions(-) diff --git a/src/xen/xend_internal.c b/src/xen/xend_internal.c index 9d95291..d0d32e2 100644 --- a/src/xen/xend_internal.c +++ b/src/xen/xend_internal.c @@ -93,6 +93,7 @@ xenDaemonFormatSxprOnePCI(virConnectPtr conn, static int virDomainXMLDevID(virDomainPtr domain, + virDomainDefPtr domDef, virDomainDeviceDefPtr dev, char *class, char *ref, @@ -4212,7 +4213,7 @@ xenDaemonAttachDeviceFlags(virDomainPtr domain, const char *xml, sexpr = virBufferContentAndReset(&buf); - if (virDomainXMLDevID(domain, dev, class, ref, sizeof(ref))) { + if (virDomainXMLDevID(domain, def, dev, class, ref, sizeof(ref))) { /* device doesn't exist, define it */ ret = xend_op(domain->conn, domain->name, "op", "device_create", "config", sexpr, NULL); @@ -4306,7 +4307,7 @@ xenDaemonDetachDeviceFlags(virDomainPtr domain, const char *xml, def, xml, VIR_DOMAIN_XML_INACTIVE))) goto cleanup; - if (virDomainXMLDevID(domain, dev, class, ref, sizeof(ref))) + if (virDomainXMLDevID(domain, def, dev, class, ref, sizeof(ref))) goto cleanup; if (dev->type == VIR_DOMAIN_DEVICE_HOSTDEV) { @@ -6048,6 +6049,7 @@ error: */ static int virDomainXMLDevID(virDomainPtr domain, + virDomainDefPtr domDef, virDomainDeviceDefPtr dev, char *class, char *ref, @@ -6056,27 +6058,33 @@ virDomainXMLDevID(virDomainPtr domain, xenUnifiedPrivatePtr priv = domain->conn->privateData; char *xref; char *tmp; + unsigned int i; + virDomainDiskDefPtr disk; if (dev->type == VIR_DOMAIN_DEVICE_DISK) { + if (dev->data.disk->dst == NULL) + return -1; if (dev->data.disk->driverName && STREQ(dev->data.disk->driverName, "tap")) strcpy(class, "tap"); else strcpy(class, "vbd"); - if (dev->data.disk->dst == NULL) - return -1; - xenUnifiedLock(priv); - xref = xenStoreDomainGetDiskID(domain->conn, domain->id, - dev->data.disk->dst); - xenUnifiedUnlock(priv); - if (xref == NULL) - return -1; - - tmp = virStrcpy(ref, xref, ref_len); - VIR_FREE(xref); - if (tmp == NULL) - return -1; + /* For disks, the device name can be used directly. + * If disk device exists in domain definintion, + * copy it to ref and return success. + */ + for (i = 0; i < domDef->ndisks; i++) { + disk = domDef->disks[i]; + if (STREQ(dev->data.disk->dst, disk->dst)) { + tmp = virStrcpy(ref, disk->dst, ref_len); + if (tmp == NULL) + return -1; + else + return 0; + } + } + return -1; } else if (dev->type == VIR_DOMAIN_DEVICE_NET) { char mac[30]; virDomainNetDefPtr def = dev->data.net; -- 1.6.0.2

3 4

[libvirt] running libvirt on Xen Cloud Platform 0.7.1
by Gadi Naor 01 Mar '10

01 Mar '10

Hi, I am trying to run libvirt daemon in Dom0 of Xen Cloud Platform 0.7.1. For that purpose I downloaded the libvirt source files 0.7.6 and trying to build the libvirt rpm with XEN driver using the XCP DDK. I am running into to many problems getting it done - so I was wondering whether this is supported or if anyone is using it in such way. Thank You, Gadi _________________________________________________________________ Hotmail: Free, trusted and rich email service. https://signup.live.com/signup.aspx?id=60969

1 0

[libvirt] and for LXC?
by Mihamina Rakotomandimby 01 Mar '10

01 Mar '10

Manao ahoana, Hello, Bonjour, Is there some plan/schedule to make libvirt LXC friendly? Or is it already? Misaotra, Thanks, Merci. -- Architecte Informatique chez Blueline/Gulfsat: Administration Systeme, Recherche & Developpement +261 34 29 155 34 / +261 33 11 207 36

3 2

[libvirt] (no subject)
by Daniel P. Berrange 01 Mar '10

01 Mar '10

Repost of http://www.redhat.com/archives/libvir-list/2010-February/msg00177.html The virsh output is now scaled to show KB/MB as needed, instead of always bytes. The public API gained extra status VIR_DOMAIN_JOB_COMPLETED = 3, /* Job has finished, but isn't cleaned up */ VIR_DOMAIN_JOB_FAILED = 4, /* Job hit error, but isn't cleaned up */ VIR_DOMAIN_JOB_CANCELLED = 6, /* Job was aborted, but isn't cleaned up */ to allow to more easily detecting end of a job, once we allow for truely async usage

2 21

[libvirt] [PATCH] [macvtap] interface type direct web docs addition
by Stefan Berger 01 Mar '10

01 Mar '10

This adds a description about the 'direct' type of interface recently added for macvtap device type support on the host.

3 2

[libvirt] [PATCH] Free the macvtap mode string
by Stefan Berger 01 Mar '10

01 Mar '10

I forgot to free the macvtap mode string. This fixes it. Signed-off-by: Stefan Berger <stefanb(a)us.ibm.com>

3 2

[libvirt] Hypervisor raw options
by Daniel Dehennin 01 Mar '10

01 Mar '10

Hello, I wonder if it's possible to add hypervisor pass-through options, I add some usage of kvm -startdate and need to use a script which call kvm with the desired option. Something like: <emulator>/usr/bin/kvm</emulator> <emulator-options> <option>-option1</option> <option>-option2</option> </emulator-options> The several <option/> will be passed to the <emulator/> command. Regards. -- Daniel Dehennin Récupérer ma clef GPG: gpg --keyserver pgp.mit.edu --recv-keys 0x6A2540D1

2 1