- Devel - libvirt List Archives

[libvirt] [PATCH 2/3] Add a detailed message when tap device add fails
by Doug Goldstein 05 Aug '10

05 Aug '10

Added a more detailed error message when adding a tap devices fails and the kernel is missing tun support. Signed-off-by: Doug Goldstein <cardoe(a)gentoo.org> --- src/qemu/qemu_conf.c | 7 +++++++ 1 files changed, 7 insertions(+), 0 deletions(-) diff --git a/src/qemu/qemu_conf.c b/src/qemu/qemu_conf.c index 2ca3350..e92021a 100644 --- a/src/qemu/qemu_conf.c +++ b/src/qemu/qemu_conf.c @@ -1694,6 +1694,13 @@ qemudNetworkIfaceConnect(virConnectPtr conn, qemuReportError(VIR_ERR_INTERNAL_ERROR, _("Failed to add tap interface to bridge. " "%s is not a bridge device"), brname); + } else if (err == ENOENT) { + /* When the tun drive is missing, give a better message. */ + qemuReportError(VIR_ERR_INTERNAL_ERROR, "%s", + _("Failed to add tap interface to bridge. " + "Your kernel is missing the 'tun' module or " + "CONFIG_TUN or you need to add the " + "/dev/net/tun device node.")); } else if (template_ifname) { virReportSystemError(err, _("Failed to add tap interface to bridge '%s'"), -- 1.7.2

2 1

[libvirt] [PATCH] build: rerun bootstrap if po/Makevars got nuked
by Eric Blake 05 Aug '10

05 Aug '10

There has been a frequent complaint of: make[2]: Entering directory `/home/remote/eblake/libvirt/po' make[2]: *** No rule to make target `/config.status', needed by `Makefile'. Stop. It happens after nuking and regenerating the po directory, which is a common action after running anything like 'make dist' or 'make rpm' that dirties all the .po files. Teach autogen.sh that it must regenerate po/Makevars to avoid the missing variable declaration, and teach cfg.mk to recognize that a nuked po directory is cause to rerun autogen.sh. * cfg.mk (_update_required): Check for po/Makevars. * autogen.sh (bootstrap): Run bootstrap if it got lost. Diagnosed by Justin Clift. --- Tested by 'git clean -x -f po && make'. A bit drastic (I prefer 'git checkout po' as the fastest way to undo local changes to the po directory caused by 'make dist', since that won't erase po/Makevars), but using that sledgehammer to clean po/ proves that this patch works. This should even work in a VPATH build :) autogen.sh | 8 ++++++-- cfg.mk | 1 + 2 files changed, 7 insertions(+), 2 deletions(-) diff --git a/autogen.sh b/autogen.sh index 2f5b42d..c0a1c4a 100755 --- a/autogen.sh +++ b/autogen.sh @@ -74,10 +74,14 @@ bootstrap_hash() # Ensure that whenever we pull in a gnulib update or otherwise change to a # different version (i.e., when switching branches), we also rerun ./bootstrap. +# Also, running 'make rpm' tends to litter the po/ directory, and some people +# like to run 'git clean -x -f po' to fix it; but only ./bootstrap regenerates +# the required file po/Makevars. curr_status=.git-module-status t=$(bootstrap_hash; git diff .gnulib) -if test "$t" = "$(cat $curr_status 2>/dev/null)"; then - : # good, it's up to date, all we need is autoreconf +if test "$t" = "$(cat $curr_status 2>/dev/null)" \ + && test -f "$THEDIR/po/Makevars"; then + # good, it's up to date, all we need is autoreconf autoreconf -if else echo running bootstrap... diff --git a/cfg.mk b/cfg.mk index e12265e..7226828 100644 --- a/cfg.mk +++ b/cfg.mk @@ -459,6 +459,7 @@ ifeq (0,$(MAKELEVEL)) # b653eda3ac4864de205419d9f41eec267cb89eeb _submodule_hash = sed 's/^[ +-]//;s/ .*//' _update_required := $(shell \ + test -f po/Makevars || { echo 1; exit; }; \ cd '$(srcdir)'; \ actual=$$(git submodule status | $(_submodule_hash); \ git hash-object bootstrap.conf; \ -- 1.7.2

2 2

[libvirt] NUMA cell?
by adrian wyssen 05 Aug '10

05 Aug '10

Hi, I have a question regarding the documentation of libvirt. There is a structure called 'virNodeInfo' and in the doc, you speak about NUMA cell. I was wondering what that is. And what do you understand by 'the number of threads per core'? Thank you for a quick response, since I need to write an explanation in a report until this evening. Have a good day!

2 1

[libvirt] Libvirt synchronous hooks
by Radek Hladik 05 Aug '10

05 Aug '10

I am trying Libvirt synchronous hooks and I would like to ask a question. I would like to use the machine start (qemu+kvm) hook to set up the storage for the machine. I already mentioned my setup in this mailing list but for now it is only important that VM storage is a md raid constructed from iSCSI disks. The VM's config is using simple file device: <disk type='file' device='disk'> <driver name='qemu' type=''/> <source file='/dev/disk/by-id/md-uuid-e464a51e:f61e98b4:bfe78010:bc810f04'/> <target dev='vda' bus='virtio'/> <address type='pci' domain='0x0000' bus='0x00' slot='0x05' function='0x0'/> </disk> Stopping the array from stop hook works perfectly however it seems that on the startup the storage is checked before the hook is executed. At least I do get errors that the storage file (/dev/disk/by-id/md-uuid-....) does not exist and my hook is not called. I would say that I am hooking on wrong places that I should hook on storage startup/shutdown but as far as I know there are no hooks for storage drivers. Are there any plans in this area? Or is there any other way how to do it that I do not see? Radek

3 2

[libvirt] [PATCH v4 0/8]: Add arbitrary qemu command-line and monitor commands
by Chris Lalancette 05 Aug '10

05 Aug '10

As we discussed previously, here is the patch series to add the ability to specify arbitrary qemu command-line parameters and environment variables, and also give arbitrary monitor commands to a guest. Because these extra arguments have a good shot at confusing libvirt, the use of them is not supported, but left available for advanced users and developers. They are also in a separate library and have a separate on-the-wire protocol. There is one bug left that I have not yet been able to fix. Because of the complicated way that virsh parses command-line arguments, it is not possible to pass through spaces and quotes when using the qemu-monitor-command. Unfortunately, the qemu monitor commands (and in particular when using QMP) depend heavily on quoting and spacing, so using virsh to send through command-lines is difficult. I'll have to think about how to better resolve this issue, but it should not hold up the rest of the series. I will point out one particular patch that could use some careful review, namely PATCH 6/8. In there, I change the remote_message_hdr to use an int proc instead of a enum remote_procedure proc. Now, as far as I can tell the two sizes should be the same, so the wire protocol should be the same. Indeed, testing seems to show that older virsh can talk just fine to a libvirtd with these patches applied. However, the regenerated remote_protocol.c file for parsing remote_message_header has some strange bits in it that I don't quite understand. I'm hoping someone else can look at it and confirm that it is OK. Thanks to Dan Berrange and Eric Blake for their reviews already, and to DV for the Relax NG schema changes. Changes since v3 are listed in the individual patches.

7 26

[libvirt] [PATCH] Fix a memory leak in the qemudBuildCommandLine.
by Chris Lalancette 05 Aug '10

05 Aug '10

ADD_ARG_LIT should only be used for literal arguments, since it duplicates the memory. Since virBufferContentAndReset is already allocating memory, we should only use ADD_ARG. Signed-off-by: Chris Lalancette <clalance(a)redhat.com> --- src/qemu/qemu_conf.c | 2 +- 1 files changed, 1 insertions(+), 1 deletions(-) diff --git a/src/qemu/qemu_conf.c b/src/qemu/qemu_conf.c index 172986e..d0655fd 100644 --- a/src/qemu/qemu_conf.c +++ b/src/qemu/qemu_conf.c @@ -4110,7 +4110,7 @@ int qemudBuildCommandLine(virConnectPtr conn, goto error; } - ADD_ARG_LIT(virBufferContentAndReset(&boot_buf)); + ADD_ARG(virBufferContentAndReset(&boot_buf)); } if (def->os.kernel) { -- 1.7.1.1

3 4

Re: [libvirt] [PATCH] Do not use boot=on on IDE device
by Antoine Mercadal 04 Aug '10

04 Aug '10

Hello, Just a feedback about this patch https://www.redhat.com/archives/libvir-list/2010-July/msg00691.html submitted by Daniel Veillard : It works like a charm. Before applying it, I wasn't able to start some windows XP guest on KVM using IDE drives. Cheers -- Antoine Mercadal (+33) 06 23 00 84 04 « Le mot liberté n'admet, par définition, aucune restriction. »

2 1

[libvirt] Release of libvirt-0.8.3
by Daniel Veillard 04 Aug '10

04 Aug '10

Nearly on time :-) It's available as usual from: ftp://libvirt.org/libvirt/ That's mostly a bug fix release exept for ESX 4.1 and vSphere support and the new APIs to allow easier hacking at the QEmu level. Features: - esx: Support vSphere 4.1 (Matthias Bolte) - Qemu arbitrary monitor commands. (Chris Lalancette) - Qemu Monitor API entry point. (Chris Lalancette) Documentation: - docs: Link wiki FAQ to main page (Cole Robinson) - Document the memory balloon device (Daniel P. Berrange) - man pages: update authors and copyright notice for libvirtd and virsh (Justin Clift) - Add openauth example to demonstrate a custom auth callback (Matthias Bolte) - docs: fix so generated .html files are removed with make clean (Justin Clift) - virsh: Fix man page syntax (Jiri Denemark) - html docs: added firewall explanation page by daniel berrange (Justin Clift) - libvirtd: add man page for libvirtd (Justin Clift) Portability: - Fix compile on i686. (Chris Lalancette) - daemon: dispatch.c should include stdio.h (and stdarg.h) (Ryota Ozaki) Bug fixes: - qemu: Fix PCI address allocation (Jiri Denemark) - Don't leak delay string when freeing virInterfaceBridgeDefs (Laine Stump) - qemu: don't lose error on setting monitor capabilities (Eric Blake) - Add iptables rule to fixup DHCP response checksum. (Laine Stump) - Fix the ACS checking in the PCI code. (Chris Lalancette) - Free up memballoon def. (Chris Lalancette) - Fix a bogus warning when parsing <hostdev> (Chris Lalancette) - Update ID after stopping a domain (Matthias Bolte) - openvzDomainCreateWithFlags: set domain id to the correct value (Jean-Baptiste Rouault) - xenapi: Update ID after starting a domain (Matthias Bolte) - esx: Update ID after starting a domain (Matthias Bolte) - Fix DMI uuid parsing. (Chris Lalancette) - Do not activate boot=on on devices when not using KVM (Daniel Veillard) - Fix a memory leak in the qemudBuildCommandLine. (Chris Lalancette) - esx: Fix freeing of heterogeneous lists (Matthias Bolte) - xen: fix logic bug (Eric Blake) - Eliminate memory leak in xenUnifiedDomainInfoListFree (Laine Stump) - lxc: Fix 'autostart' doesn't take effect actually (Ryota Ozaki) - Fix --with-xen-proxy related compile error (Matthias Bolte) - Fix a potential race in pciInitDevice. (Chris Lalancette) - Invert logic for checking for QEMU disk cache options (Daniel P. Berrange) - libvirt-guests: Don't throw errors if libvirtd is not installed (Jiri Denemark) - pciResetDevice: use inactive devices to determine safe reset (Chris Wright) - secaatest: Fix compilation (Matthias Bolte) - virt-aa-helper-test: Fix failure due to the new disk format probing option (Matthias Bolte) - virt-aa-helper: Make getopt accept the p option (Matthias Bolte) - virt-aa-helper: Fix return value of add_file_path (Matthias Bolte) - Fix SEGV on exit after domainEventDeregister() (Philipp Hahn) - pciSharesBusWithActive fails to find multiple devices on bus (Chris Wright) - Fix incorrect use of private data in remote driver (Daniel P. Berrange) - Set a stable & high MAC addr for guest TAP devices on host (Daniel P. Berrange) - Fix PCI address assignment if no IDE controller is present (Daniel P. Berrange) - lxc: force kill of init process by sending SIGKILL if needed (Ryota Ozaki) - Fix a NULL dereference in the case that the arg in question didn't exist. (Chris Lalancette) - Remove bogus free of static strings (Daniel P. Berrange) - Fix a deadlock in bi-directional p2p concurrent migration. (Chris Lalancette) - Make virsh setmaxmem balloon only when successful. (Chris Lalancette) - fsync new storage volumes even if new volume was copied. (Laine Stump) - Don't skip zero'ing end of volume file when inputvol is shorter than newvol (Laine Stump) - Always clear out the last_error in virshReportError. (Chris Lalancette) - CVE-2010-2242 Apply a source port mapping to virtual network masquerading (Daniel P. Berrange) - uml_driver: correct logic error in umlMonitorCommand (Jim Meyering) - qemuConnectMonitor: fix a bug that would have masked SELinux failure (Jim Meyering) - python: Fix IOErrorReasonCallback bindings (Cole Robinson) - cpuCompare: Fix crash on unexpected CPU XML (Jiri Denemark) - cpu: Fail when CPU type cannot be detected from XML (Jiri Denemark) - cpuCompare: Fix comparison of two host CPUs (Jiri Denemark) - Fix potential crash in QEMU monitor JSON impl (Daniel P. Berrange) Improvements: - OpenVZ: implement suspend/resume driver APIs (Jean-Baptiste Rouault) - esx: Set storage pool target path to host.mountInfo.path (Matthias Bolte) - esx: Make storage pool lookup by name and UUID more robust (Matthias Bolte) - esx: Restrict vpx:// to handle a single host in a vCenter (Matthias Bolte) - esx: Map some managed object types (Matthias Bolte) - esx: Parse the path of the URI (Matthias Bolte) - Make virsh -d check its input (Daniel Veillard) - esx: Switch from name to number checks in the subdrivers (Matthias Bolte) - esx: Improve blocked task detection and fix race condition (Matthias Bolte) - build: distribute libvirt_qemu.syms (Eric Blake) - build: restore operation of bit-rotted 'make cov' (Eric Blake) - qemu: virtio console support (Cole Robinson) - domain conf: Track <console> target type (Cole Robinson) - domain conf: char: Add an explicit targetType field (Cole Robinson) - domain conf: Rename character prop targetType -> deviceType (Cole Robinson) - docs: domain: Document virtio <channel> (Cole Robinson) - tests: Test qemuxml2xml when expected xml changes (Cole Robinson) - fix handling of PORT_PROFILE_RESPONSE_INPROGRESS netlink message (Gerhard Stenzel) - maint: turn on gcc logical-op checking (Eric Blake) - libvirt-guests: add reload, condrestart (Eric Blake) - libvirt-guests: enhance status (Eric Blake) - libvirt-guests: detect invalid arguments (Eric Blake) - qemu: Allow setting boot menu on/off (Cole Robinson) - qemu: Error on unsupported graphics config (Cole Robinson) - Force FLR on for buggy SR-IOV devices. (Chris Lalancette) - qemudDomainAttachHostPciDevice refactor to use new helpers (Chris Wright) - Add helpers qemuPrepareHostdevPCIDevice and qemuDomainReAttachHostdevDevices (Chris Wright) - qemuGetPciHostDeviceList take hostdev list directly (Chris Wright) - esx: Add vpx:// scheme to allow direct connection to a vCenter (Matthias Bolte) - esx: Don't ignore the vcenter query parameter (Matthias Bolte) - esx: Add autodetection for the SCSI controller model (Matthias Bolte) - esx: Allow 'vmpvscsi' as SCSI controller model (Matthias Bolte) - Add tests for the new Qemu namespace XML. (Chris Lalancette) - Qemu remote protocol. (Chris Lalancette) - Handle arbitrary qemu command-lines in qemuParseCommandLine. (Chris Lalancette) - Qemu arbitrary command-line arguments. (Chris Lalancette) - Add namespace callback hooks to domain_conf. (Chris Lalancette) - Remove erroneous setting of return value to errno. (Laine Stump) - Change virDirCreate to return -errno on failure. (Laine Stump) - Make virStorageBackendCopyToFD return -errno. (Laine Stump) - Change virFileOperation to return -errno (ie < 0) on error. (Laine Stump) - Re-arrange PCI device address assignment to match QEMU's default (Daniel P. Berrange) - Explicitly represent balloon device in XML and handle PCI address (Daniel P. Berrange) - Rearrange VGA/IDE controller address reservation (Daniel P. Berrange) - Use unsigned long in cmdSetmem. (Chris Lalancette) - Fix up inconsistent virsh option error reporting. (Chris Lalancette) - Use the extract backing store format in storage volume lookup (Daniel P. Berrange) - Rewrite qemu-img backing store format handling (Daniel P. Berrange) - Add ability to set a default driver name/type when parsing disks (Daniel P. Berrange) - Disable all disk probing in QEMU driver & add config option to re-enable (Daniel P. Berrange) - Pass security driver object into all security driver callbacks (Daniel P. Berrange) - Convert all disk backing store loops to shared helper API (Daniel P. Berrange) - Add an API for iterating over disk paths (Daniel P. Berrange) - Require format to be passed into virStorageFileGetMetadata (Daniel P. Berrange) - Refactor virStorageFileGetMetadataFromFD to separate functionality (Daniel P. Berrange) - Remove 'type' field from FileTypeInfo struct (Daniel P. Berrange) - Extract the backing store format as well as name, if available (Daniel P. Berrange) - RFC: Canonicalize block device paths (David Allan) - .gitignore: Ignore generated libvirtd docs (Cole Robinson) - esx: Make esxVI_*_Deserialize dynamically dispatched (Matthias Bolte) - qemu: Use -nodefconfig when probing for CPU models (Jiri Denemark) - Ensure we return the callback ID in python events binding (Daniel P. Berrange) - virsh: add new --details option to vol-list (Justin Clift) - Implement virsh managedsave-remove command. (Chris Lalancette) - cpu: Add new models from qemu's target-x86_64.conf (Jiri Denemark) - cpu: Add support for CPU vendor (Jiri Denemark) - cpuBaseline: Detect empty set of common features (Jiri Denemark) - cpuBaseline: Don't mess with the CPU returned by arch driver (Jiri Denemark) - Make html docs in non-srcdir build (Jiri Denemark) Cleanups: - Fix build error in virsh.c (Laine Stump) - Fix virsh error message when -d arg is not numeric (Eric Blake) - Fix a couple of typo in iSCSI backend (Aurelien ROUGEMONT) - Don't put a semicolon on the end of a VIR_ENUM_IMPL. (Chris Lalancette) - Remove duplicate </p> from downloads.html.in (Matthias Bolte) - storage: kill dead stores (Eric Blake) - qemu: kill some dead stores (Eric Blake) - network: kill dead store (Eric Blake) - esx: silence spurious compiler warning (Eric Blake) - build: fix 'make syntax-check' failure (Eric Blake) - lxc: Fix return values of veth.c functions (Ryota Ozaki) - maint: fix comment typos (Eric Blake) - Fix up confusing indentation in qemudDomainAttachHostPciDevice. (Chris Lalancette) - build: fix VPATH builds (Eric Blake) - virt-aa-helper: Ignore open errors again (Matthias Bolte) - qemu-api: avoid build failure (Eric Blake) - Fix .mailmap after accidental wrong committer address (Daniel P. Berrange) - Remove inappropriate use of VIR_ERR_NO_SUPPORT (Daniel P. Berrange) - Remove unused and bitrotting vshCommandOptStringList (Chris Lalancette) - Remove error checking after using vshMalloc. (Chris Lalancette) - Remove the "showerror" parameter from vshConnectionUsability. (Chris Lalancette) - Eliminate compiler warning due to gettext string with no format args (Laine Stump) - Fix build by removing unknown pod2man flag (Daniel P. Berrange) Thanks everybody for the help on this release ! 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/

1 0

Re: [libvirt] [RHEL6 PATCH] Correct cpuid flags and "model" fields, V2
by john cooper 04 Aug '10

04 Aug '10

A question arose in today's kvm meeting concerning any impact to libvirt from this change. I've discussed it with Cole and it seems to be a non-issue. But just to err on the side of caution, here's a summary: The current cpu model definition of "qemu64" upstream is problematic from kvm's perspective such that we need to modify it slightly (BZ justifications listed below). Doing so we've left the qemu64 definition as-is, but added "cpu64-rhel6" and "cpu64-rhel5" models which are now selected by default via "-M <machine>", for "RHEL 6.0.0 PC" and "RHEL 5.x.y PC" respectively. So the only issue would be libvirt invoking qemu with neither a "-cpu" nor "-M" argument (which defaults to qemu64) or explicitly requesting "-cpu qemu64". >From my discussion with Cole it appears the use cases where this may happen fall outside of routine/expected usage and would need to be explicitly requested by the user. However I wanted to call this out here in the event we're overlooking something. Thanks, -john http://post-office.corp.redhat.com/archives/rhvirt-patches/2010-July/msg011… http://post-office.corp.redhat.com/archives/rhvirt-patches/2010-July/msg008… john cooper wrote: > Addresses BZs: > > #618332 - CPUID_EXT_POPCNT enabled in qemu64 and qemu32 built-in models. > #613892 - [SR-IOV]VF device can not start on 32bit Windows2008 SP2 > > Summary: > > CPU feature flags for several processor definitions require correction > to accurately reflect the corresponding shipped silicon. In particular > overly conservative values for cpuid "model" fields cause disable of > MSI support in windows guests (BZ #613892). Also recent upstream changes > of qemu64's built-in definition enables POPCNT (for the benefit of TCG) > but risks kvm migration breakage (BZ #618332). The following patch > address these issues collectively. > > Changes relative to previous version: > > - drop of the "qemu32-rhel5" model as it appears to be unneeded. > > - rename of "qemu64-rhel?" to "cpu64-rhel?" as we're diverging from > upstream qemu64's definition but haven't migrated to kvm64 quite yet. > > - Correction of several fields for the (now) cpu64-rhel5 model. > > Further detail may be found in the associated mail thread common to > both BZ cases starting about here: > > http://post-office.corp.redhat.com/archives/rhvirt-patches/2010-July/msg008… > > Brew Build: > > https://brewweb.devel.redhat.com/taskinfo?taskID=2643552 > > Upstream status: > > BZ #618332 is a local, interim change and doesn't relate to upstream > concerns. Longer-term this qemu64 derived model would be displaced > by kvm64. The update of cpu definitions (BZ #613892) however needs > to be pushed upstream upon validation in rhel6. We're currently the > sole users of the new cpu models which fosters this inverted process. -- john.cooper(a)redhat.com

2 1

[libvirt] [PATCH] qemu: Fix PCI address allocation
by Jiri Denemark 04 Aug '10

04 Aug '10

When attaching a PCI device which doesn't explicitly set its PCI address, libvirt allocates the address automatically. The problem is that when checking which PCI address is unused, we only check for those with slot number higher than the highest slot number ever used. Thus attaching/detaching such device several times in a row (31 is the theoretical limit, less then 30 tries are enough in practise) makes any further device attachment fail. Furthermore, attaching a device with predefined PCI address to 0:0:31 immediately forbids attachment of any PCI device without explicit address. This patch changes the logic so that we always check all PCI addresses before we say there is no PCI address available. --- src/qemu/qemu_conf.c | 14 ++++---------- 1 files changed, 4 insertions(+), 10 deletions(-) diff --git a/src/qemu/qemu_conf.c b/src/qemu/qemu_conf.c index 57bc02f..eaebcc1 100644 --- a/src/qemu/qemu_conf.c +++ b/src/qemu/qemu_conf.c @@ -2055,8 +2055,6 @@ qemuAssignDeviceAliases(virDomainDefPtr def, unsigned long long qemuCmdFlags) #define QEMU_PCI_ADDRESS_LAST_SLOT 31 struct _qemuDomainPCIAddressSet { virHashTablePtr used; - int nextslot; - /* XXX add domain, bus later when QEMU allows > 1 */ }; @@ -2148,9 +2146,6 @@ int qemuDomainPCIAddressReserveAddr(qemuDomainPCIAddressSetPtr addrs, return -1; } - if (dev->addr.pci.slot > addrs->nextslot) - addrs->nextslot = dev->addr.pci.slot + 1; - return 0; } @@ -2217,7 +2212,7 @@ int qemuDomainPCIAddressSetNextAddr(qemuDomainPCIAddressSetPtr addrs, { int i; - for (i = addrs->nextslot ; i <= QEMU_PCI_ADDRESS_LAST_SLOT ; i++) { + for (i = 0 ; i <= QEMU_PCI_ADDRESS_LAST_SLOT ; i++) { virDomainDeviceInfo maybe; char *addr; @@ -2228,13 +2223,14 @@ int qemuDomainPCIAddressSetNextAddr(qemuDomainPCIAddressSetPtr addrs, addr = qemuPCIAddressAsString(&maybe); - VIR_DEBUG("Allocating PCI addr %s", addr); - if (virHashLookup(addrs->used, addr)) { + VIR_DEBUG("PCI addr %s already in use", addr); VIR_FREE(addr); continue; } + VIR_DEBUG("Allocating PCI addr %s", addr); + if (virHashAddEntry(addrs->used, addr, addr) < 0) { VIR_FREE(addr); return -1; @@ -2245,8 +2241,6 @@ int qemuDomainPCIAddressSetNextAddr(qemuDomainPCIAddressSetPtr addrs, dev->addr.pci.bus = 0; dev->addr.pci.slot = i; - addrs->nextslot = i + 1; - return 0; } -- 1.7.2

4 6