use PVE::Exception qw(raise raise_param_exc);
use PVE::Format qw(render_duration render_bytes);
use PVE::GuestHelpers qw(safe_string_ne safe_num_ne safe_boolean_ne);
+use PVE::HA::Config;
+use PVE::Mapping::PCI;
+use PVE::Mapping::USB;
use PVE::INotify;
use PVE::JSONSchema qw(get_standard_option parse_property_string);
use PVE::ProcFSTools;
use PVE::QMPClient;
use PVE::QemuConfig;
-use PVE::QemuServer::Helpers qw(min_version config_aware_timeout windows_version);
+use PVE::QemuServer::Helpers qw(config_aware_timeout min_version windows_version);
use PVE::QemuServer::Cloudinit;
use PVE::QemuServer::CGroup;
-use PVE::QemuServer::CPUConfig qw(print_cpu_device get_cpu_options);
+use PVE::QemuServer::CPUConfig qw(print_cpu_device get_cpu_options get_cpu_bitness is_native_arch);
use PVE::QemuServer::Drive qw(is_valid_drivename drive_is_cloudinit drive_is_cdrom drive_is_read_only parse_drive print_drive);
use PVE::QemuServer::Machine;
-use PVE::QemuServer::Memory;
+use PVE::QemuServer::Memory qw(get_current_memory);
use PVE::QemuServer::Monitor qw(mon_cmd);
use PVE::QemuServer::PCI qw(print_pci_addr print_pcie_addr print_pcie_root_port parse_hostpci);
-use PVE::QemuServer::USB qw(parse_usb_device);
+use PVE::QemuServer::QMPHelpers qw(qemu_deviceadd qemu_devicedel qemu_objectadd qemu_objectdel);
+use PVE::QemuServer::USB;
my $have_sdn;
eval {
require PVE::Network::SDN::Zones;
+ require PVE::Network::SDN::Vnets;
$have_sdn = 1;
};
"$EDK2_FW_BASE/OVMF_CODE_4M.secboot.fd",
"$EDK2_FW_BASE/OVMF_VARS_4M.ms.fd",
],
+ # FIXME: These are legacy 2MB-sized images that modern OVMF doesn't supports to build
+ # anymore. how can we deperacate this sanely without breaking existing instances, or using
+ # older backups and snapshot?
default => [
"$EDK2_FW_BASE/OVMF_CODE.fd",
"$EDK2_FW_BASE/OVMF_VARS.fd",
optional => 1,
});
-PVE::JSONSchema::register_standard_option('pve-qemu-machine', {
- description => "Specifies the QEMU machine type.",
- type => 'string',
- pattern => '(pc|pc(-i440fx)?-\d+(\.\d+)+(\+pve\d+)?(\.pxe)?|q35|pc-q35-\d+(\.\d+)+(\+pve\d+)?(\.pxe)?|virt(?:-\d+(\.\d+)+)?(\+pve\d+)?)',
- maxLength => 40,
- optional => 1,
-});
-
# FIXME: remove in favor of just using the INotify one, it's cached there exactly the same way
my $nodename_cache;
sub nodename {
description => "Run fstrim after moving a disk or migrating the VM.",
type => 'boolean',
optional => 1,
- default => 0
+ default => 0,
+ },
+ 'freeze-fs-on-backup' => {
+ description => "Freeze/thaw guest filesystems on backup for consistency.",
+ type => 'boolean',
+ optional => 1,
+ default => 1,
},
type => {
description => "Select the agent type",
minimum => 4,
maximum => 512,
},
+ clipboard => {
+ description => 'Enable a specific clipboard. If not set, depending on the display type the'
+ .' SPICE one will be added. Migration with VNC clipboard is not yet supported!',
+ type => 'string',
+ enum => ['vnc'],
+ optional => 1,
+ },
};
my $ivshmem_fmt = {
},
memory => {
optional => 1,
- type => 'integer',
- description => "Amount of RAM for the VM in MiB. This is the maximum available memory when"
- ." you use the balloon device.",
- minimum => 16,
- default => 512,
+ type => 'string',
+ description => "Memory properties.",
+ format => $PVE::QemuServer::Memory::memory_fmt
},
balloon => {
optional => 1,
ostype => {
optional => 1,
type => 'string',
+ # NOTE: When extending, also consider extending `%guest_types` in `Import/ESXi.pm`.
enum => [qw(other wxp w2k w2k3 w2k8 wvista win7 win8 win10 win11 l24 l26 solaris)],
description => "Specify guest operating system.",
verbose_description => <<EODESC,
win7;; Microsoft Windows 7
win8;; Microsoft Windows 8/2012/2012r2
win10;; Microsoft Windows 10/2016/2019
-win11;; Microsoft Windows 11/2022
+win11;; Microsoft Windows 11/2022/2025
l24;; Linux 2.4 Kernel
-l26;; Linux 2.6 - 5.X Kernel
+l26;; Linux 2.6 - 6.X Kernel
solaris;; Solaris/OpenSolaris/OpenIndiania kernel
EODESC
},
verbose_description => <<EODESCR,
Arbitrary arguments passed to kvm, for example:
-args: -no-reboot -no-hpet
+args: -no-reboot -smbios 'type=0,vendor=FOO'
NOTE: this option is for experts only.
EODESCR
};
PVE::JSONSchema::register_format('pve-qm-cicustom', $cicustom_fmt);
+# any new option might need to be added to $cloudinitoptions in PVE::API2::Qemu
my $confdesc_cloudinit = {
citype => {
optional => 1,
.' recommended. Use ssh keys instead. Also note that older cloud-init versions do not'
.' support hashed passwords.',
},
+ ciupgrade => {
+ optional => 1,
+ type => 'boolean',
+ description => 'cloud-init: do an automatic package upgrade after the first boot.',
+ default => 1,
+ },
cicustom => {
optional => 1,
type => 'string',
PVE::JSONSchema::register_standard_option("pve-qm-$k", $v);
}
-my $MAX_USB_DEVICES = 14;
my $MAX_NETS = 32;
my $MAX_SERIAL_PORTS = 4;
my $MAX_PARALLEL_PORTS = 3;
-my $MAX_NUMA = 8;
-my $numa_fmt = {
- cpus => {
- type => "string",
- pattern => qr/\d+(?:-\d+)?(?:;\d+(?:-\d+)?)*/,
- description => "CPUs accessing this NUMA node.",
- format_description => "id[-id];...",
- },
- memory => {
- type => "number",
- description => "Amount of memory this NUMA node provides.",
- optional => 1,
- },
- hostnodes => {
- type => "string",
- pattern => qr/\d+(?:-\d+)?(?:;\d+(?:-\d+)?)*/,
- description => "Host NUMA nodes to use.",
- format_description => "id[-id];...",
- optional => 1,
- },
- policy => {
- type => 'string',
- enum => [qw(preferred bind interleave)],
- description => "NUMA allocation policy.",
- optional => 1,
- },
-};
-PVE::JSONSchema::register_format('pve-qm-numanode', $numa_fmt);
-my $numadesc = {
- optional => 1,
- type => 'string', format => $numa_fmt,
- description => "NUMA topology.",
-};
-PVE::JSONSchema::register_standard_option("pve-qm-numanode", $numadesc);
-
-for (my $i = 0; $i < $MAX_NUMA; $i++) {
- $confdesc->{"numa$i"} = $numadesc;
+for (my $i = 0; $i < $PVE::QemuServer::Memory::MAX_NUMA; $i++) {
+ $confdesc->{"numa$i"} = $PVE::QemuServer::Memory::numadesc;
}
my $nic_model_list = [
return $volid;
}
-my $usb_fmt = {
- host => {
- default_key => 1,
- type => 'string', format => 'pve-qm-usb-device',
- format_description => 'HOSTUSBDEVICE|spice',
- description => <<EODESCR,
-The Host USB device or port or the value 'spice'. HOSTUSBDEVICE syntax is:
-
- 'bus-port(.port)*' (decimal numbers) or
- 'vendor_id:product_id' (hexadeciaml numbers) or
- 'spice'
-
-You can use the 'lsusb -t' command to list existing usb devices.
-
-NOTE: This option allows direct access to host hardware. So it is no longer possible to migrate such
-machines - use with special care.
-
-The value 'spice' can be used to add a usb redirection devices for spice.
-EODESCR
- },
- usb3 => {
- optional => 1,
- type => 'boolean',
- description => "Specifies whether if given host option is a USB3 device or port."
- ." For modern guests (machine version >= 7.1 and ostype l26 and windows > 7), this flag"
- ." is irrelevant (all devices are plugged into a xhci controller).",
- default => 0,
- },
-};
-
-my $usbdesc = {
- optional => 1,
- type => 'string', format => $usb_fmt,
- description => "Configure an USB device (n is 0 to 4, for machine version >= 7.1 and ostype"
- ." l26 or windows > 7, n can be up to 14).",
-};
-PVE::JSONSchema::register_standard_option("pve-qm-usb", $usbdesc);
-
my $serialdesc = {
optional => 1,
type => 'string',
$confdesc->{$key} = $PVE::QemuServer::Drive::drivedesc_hash->{$key};
}
-for (my $i = 0; $i < $MAX_USB_DEVICES; $i++) {
- $confdesc->{"usb$i"} = $usbdesc;
+for (my $i = 0; $i < $PVE::QemuServer::USB::MAX_USB_DEVICES; $i++) {
+ $confdesc->{"usb$i"} = $PVE::QemuServer::USB::usbdesc;
}
my $boot_fmt = {
die "unable to parse hotplug option\n";
}
-sub scsi_inquiry {
- my($fh, $noerr) = @_;
-
- my $SG_IO = 0x2285;
- my $SG_GET_VERSION_NUM = 0x2282;
-
- my $versionbuf = "\x00" x 8;
- my $ret = ioctl($fh, $SG_GET_VERSION_NUM, $versionbuf);
- if (!$ret) {
- die "scsi ioctl SG_GET_VERSION_NUM failoed - $!\n" if !$noerr;
- return;
- }
- my $version = unpack("I", $versionbuf);
- if ($version < 30000) {
- die "scsi generic interface too old\n" if !$noerr;
- return;
- }
-
- my $buf = "\x00" x 36;
- my $sensebuf = "\x00" x 8;
- my $cmd = pack("C x3 C x1", 0x12, 36);
-
- # see /usr/include/scsi/sg.h
- my $sg_io_hdr_t = "i i C C s I P P P I I i P C C C C S S i I I";
+sub assert_clipboard_config {
+ my ($vga) = @_;
- my $packet = pack(
- $sg_io_hdr_t, ord('S'), -3, length($cmd), length($sensebuf), 0, length($buf), $buf, $cmd, $sensebuf, 6000
- );
+ my $clipboard_regex = qr/^(std|cirrus|vmware|virtio|qxl)/;
- $ret = ioctl($fh, $SG_IO, $packet);
- if (!$ret) {
- die "scsi ioctl SG_IO failed - $!\n" if !$noerr;
- return;
- }
-
- my @res = unpack($sg_io_hdr_t, $packet);
- if ($res[17] || $res[18]) {
- die "scsi ioctl SG_IO status error - $!\n" if !$noerr;
- return;
+ if (
+ $vga->{'clipboard'}
+ && $vga->{'clipboard'} eq 'vnc'
+ && $vga->{type}
+ && $vga->{type} !~ $clipboard_regex
+ ) {
+ die "vga type $vga->{type} is not compatible with VNC clipboard\n";
}
-
- my $res = {};
- $res->@{qw(type removable vendor product revision)} = unpack("C C x6 A8 A16 A4", $buf);
-
- $res->{removable} = $res->{removable} & 128 ? 1 : 0;
- $res->{type} &= 0x1F;
-
- return $res;
-}
-
-sub path_is_scsi {
- my ($path) = @_;
-
- my $fh = IO::File->new("+<$path") || return;
- my $res = scsi_inquiry($fh, 1);
- close($fh);
-
- return $res;
}
sub print_tabletdevice_full {
my ($maxdev, $controller, $controller_prefix) = scsihw_infos($conf, $drive);
my $unit = $drive->{index} % $maxdev;
- my $devicetype = 'hd';
- my $path = '';
- if (drive_is_cdrom($drive)) {
- $devicetype = 'cd';
- } else {
- if ($drive->{file} =~ m|^/|) {
- $path = $drive->{file};
- if (my $info = path_is_scsi($path)) {
- if ($info->{type} == 0 && $drive->{scsiblock}) {
- $devicetype = 'block';
- } elsif ($info->{type} == 1) { # tape
- $devicetype = 'generic';
- }
- }
- } else {
- $path = PVE::Storage::path($storecfg, $drive->{file});
- }
- # for compatibility only, we prefer scsi-hd (#2408, #2355, #2380)
- my $version = extract_version($machine_type, kvm_user_version());
- if ($path =~ m/^iscsi\:\/\// &&
- !min_version($version, 4, 1)) {
- $devicetype = 'generic';
- }
- }
+ my $machine_version = extract_version($machine_type, kvm_user_version());
+ my $device_type = PVE::QemuServer::Drive::get_scsi_device_type(
+ $drive, $storecfg, $machine_version);
if (!$conf->{scsihw} || $conf->{scsihw} =~ m/^lsi/ || $conf->{scsihw} eq 'pvscsi') {
- $device = "scsi-$devicetype,bus=$controller_prefix$controller.0,scsi-id=$unit";
+ $device = "scsi-$device_type,bus=$controller_prefix$controller.0,scsi-id=$unit";
} else {
- $device = "scsi-$devicetype,bus=$controller_prefix$controller.0,channel=0,scsi-id=0"
+ $device = "scsi-$device_type,bus=$controller_prefix$controller.0,channel=0,scsi-id=0"
.",lun=$drive->{index}";
}
$device .= ",drive=drive-$drive_id,id=$drive_id";
- if ($drive->{ssd} && ($devicetype eq 'block' || $devicetype eq 'hd')) {
+ if ($drive->{ssd} && ($device_type eq 'block' || $device_type eq 'hd')) {
$device .= ",rotation_rate=1";
}
$device .= ",wwn=$drive->{wwn}" if $drive->{wwn};
+ # only scsi-hd and scsi-cd support passing vendor and product information
+ if ($device_type eq 'hd' || $device_type eq 'cd') {
+ if (my $vendor = $drive->{vendor}) {
+ $device .= ",vendor=$vendor";
+ }
+ if (my $product = $drive->{product}) {
+ $device .= ",product=$product";
+ }
+ }
+
} elsif ($drive->{interface} eq 'ide' || $drive->{interface} eq 'sata') {
my $maxdev = ($drive->{interface} eq 'sata') ? $PVE::QemuServer::Drive::MAX_SATA_DISKS : 2;
my $controller = int($drive->{index} / $maxdev);
my $unit = $drive->{index} % $maxdev;
- my $devicetype = ($drive->{media} && $drive->{media} eq 'cdrom') ? "cd" : "hd";
- $device = "ide-$devicetype";
+ # machine type q35 only supports unit=0 for IDE rather than 2 units. This wasn't handled
+ # correctly before, so e.g. index=2 was mapped to controller=1,unit=0 rather than
+ # controller=2,unit=0. Note that odd indices never worked, as they would be mapped to
+ # unit=1, so to keep backwards compat for migration, it suffices to keep even ones as they
+ # were before. Move odd ones up by 2 where they don't clash.
+ if (PVE::QemuServer::Machine::machine_type_is_q35($conf) && $drive->{interface} eq 'ide') {
+ $controller += 2 * ($unit % 2);
+ $unit = 0;
+ }
+
+ my $device_type = ($drive->{media} && $drive->{media} eq 'cdrom') ? "cd" : "hd";
+
+ $device = "ide-$device_type";
if ($drive->{interface} eq 'ide') {
$device .= ",bus=ide.$controller,unit=$unit";
} else {
}
$device .= ",drive=drive-$drive_id,id=$drive_id";
- if ($devicetype eq 'hd') {
+ if ($device_type eq 'hd') {
if (my $model = $drive->{model}) {
$model = URI::Escape::uri_unescape($model);
$device .= ",model=$model";
return $initiator;
}
+my sub storage_allows_io_uring_default {
+ my ($scfg, $cache_direct) = @_;
+
+ # io_uring with cache mode writeback or writethrough on krbd will hang...
+ return if $scfg && $scfg->{type} eq 'rbd' && $scfg->{krbd} && !$cache_direct;
+
+ # io_uring with cache mode writeback or writethrough on LVM will hang, without cache only
+ # sometimes, just plain disable...
+ return if $scfg && $scfg->{type} eq 'lvm';
+
+ # io_uring causes problems when used with CIFS since kernel 5.15
+ # Some discussion: https://www.spinics.net/lists/linux-cifs/msg26734.html
+ return if $scfg && $scfg->{type} eq 'cifs';
+
+ return 1;
+}
+
+my sub drive_uses_cache_direct {
+ my ($drive, $scfg) = @_;
+
+ my $cache_direct = 0;
+
+ if (my $cache = $drive->{cache}) {
+ $cache_direct = $cache =~ /^(?:off|none|directsync)$/;
+ } elsif (!drive_is_cdrom($drive) && !($scfg && $scfg->{type} eq 'btrfs' && !$scfg->{nocow})) {
+ $cache_direct = 1;
+ }
+
+ return $cache_direct;
+}
+
sub print_drive_commandline_full {
- my ($storecfg, $vmid, $drive, $pbs_name, $io_uring) = @_;
+ my ($storecfg, $vmid, $drive, $live_restore_name, $io_uring) = @_;
my $path;
my $volid = $drive->{file};
if (drive_is_cdrom($drive)) {
$path = get_iso_path($storecfg, $vmid, $volid);
- die "$drive_id: cannot back cdrom drive with PBS snapshot\n" if $pbs_name;
+ die "$drive_id: cannot back cdrom drive with a live restore image\n" if $live_restore_name;
} else {
if ($storeid) {
$path = PVE::Storage::path($storecfg, $volid);
}
}
- if ($pbs_name) {
+ if ($live_restore_name) {
$format = "rbd" if $is_rbd;
die "$drive_id: Proxmox Backup Server backed drive cannot auto-detect the format\n"
if !$format;
$opts .= ",format=$format";
}
- my $cache_direct = 0;
+ my $cache_direct = drive_uses_cache_direct($drive, $scfg);
- if (my $cache = $drive->{cache}) {
- $cache_direct = $cache =~ /^(?:off|none|directsync)$/;
- } elsif (!drive_is_cdrom($drive) && !($scfg && $scfg->{type} eq 'btrfs' && !$scfg->{nocow})) {
- $opts .= ",cache=none";
- $cache_direct = 1;
- }
-
- # io_uring with cache mode writeback or writethrough on krbd will hang...
- my $rbd_no_io_uring = $scfg && $scfg->{type} eq 'rbd' && $scfg->{krbd} && !$cache_direct;
-
- # io_uring with cache mode writeback or writethrough on LVM will hang, without cache only
- # sometimes, just plain disable...
- my $lvm_no_io_uring = $scfg && $scfg->{type} eq 'lvm';
-
- # io_uring causes problems when used with CIFS since kernel 5.15
- # Some discussion: https://www.spinics.net/lists/linux-cifs/msg26734.html
- my $cifs_no_io_uring = $scfg && $scfg->{type} eq 'cifs';
+ $opts .= ",cache=none" if !$drive->{cache} && $cache_direct;
if (!$drive->{aio}) {
- if ($io_uring && !$rbd_no_io_uring && !$lvm_no_io_uring && !$cifs_no_io_uring) {
+ if ($io_uring && storage_allows_io_uring_default($scfg, $cache_direct)) {
# io_uring supports all cache modes
$opts .= ",aio=io_uring";
} else {
# note: 'detect-zeroes' works per blockdev and we want it to persist
# after the alloc-track is removed, so put it on 'file' directly
- my $dz_param = $pbs_name ? "file.detect-zeroes" : "detect-zeroes";
+ my $dz_param = $live_restore_name ? "file.detect-zeroes" : "detect-zeroes";
$opts .= ",$dz_param=$detectzeroes" if $detectzeroes;
}
- if ($pbs_name) {
- $opts .= ",backing=$pbs_name";
+ if ($live_restore_name) {
+ $opts .= ",backing=$live_restore_name";
$opts .= ",auto-remove=on";
}
- # my $file_param = $pbs_name ? "file.file.filename" : "file";
+ # my $file_param = $live_restore_name ? "file.file.filename" : "file";
my $file_param = "file";
- if ($pbs_name) {
+ if ($live_restore_name) {
# non-rbd drivers require the underlying file to be a seperate block
# node, so add a second .file indirection
$file_param .= ".file" if !$is_rbd;
}
if (min_version($machine_version, 7, 1) && $net->{model} eq 'virtio'){
- $tmpstr .= ",rx_queue_size=1024,tx_queue_size=1024";
+ $tmpstr .= ",rx_queue_size=1024,tx_queue_size=256";
}
$tmpstr .= ",bootindex=$net->{bootindex}" if $net->{bootindex} ;
if length($ifname) >= 16;
my $vhostparam = '';
- if (is_native($arch)) {
+ if (is_native_arch($arch)) {
$vhostparam = ',vhost=on' if kernel_has_vhost_net() && $net->{model} eq 'virtio';
}
}
}
- die "no devicetype for $vga->{type}\n" if !$type;
+ die "no device-type for $vga->{type}\n" if !$type;
my $memory = "";
if ($vgamem_mb) {
return "$type,id=${vgaid}${memory}${max_outputs}${pciaddr}${edidoff}";
}
-sub parse_number_sets {
- my ($set) = @_;
- my $res = [];
- foreach my $part (split(/;/, $set)) {
- if ($part =~ /^\s*(\d+)(?:-(\d+))?\s*$/) {
- die "invalid range: $part ($2 < $1)\n" if defined($2) && $2 < $1;
- push @$res, [ $1, $2 ];
- } else {
- die "invalid range: $part\n";
- }
- }
- return $res;
-}
-
-sub parse_numa {
- my ($data) = @_;
-
- my $res = parse_property_string($numa_fmt, $data);
- $res->{cpus} = parse_number_sets($res->{cpus}) if defined($res->{cpus});
- $res->{hostnodes} = parse_number_sets($res->{hostnodes}) if defined($res->{hostnodes});
- return $res;
-}
-
# netX: e1000=XX:XX:XX:XX:XX:XX,bridge=vmbr0,rate=<mbps>
sub parse_net {
my ($data, $disable_mac_autogen) = @_;
# check if we need to apply some handling for VMs that always use the latest machine version but
# had a machine version transition happen that affected HW such that, e.g., an OS config change
# would be required (we do not want to pin machine version for non-windows OS type)
+ my $machine_conf = PVE::QemuServer::Machine::parse_machine($conf->{machine});
if (
- (!defined($conf->{machine}) || $conf->{machine} =~ m/^(?:pc|q35|virt)$/) # non-versioned machine
+ (!defined($machine_conf->{type}) || $machine_conf->{type} =~ m/^(?:pc|q35|virt)$/) # non-versioned machine
&& (!defined($meta->{'creation-qemu'}) || !min_version($meta->{'creation-qemu'}, 6, 1)) # created before 6.1
&& (!$forced_vers || min_version($forced_vers, 6, 1)) # handle snapshot-rollback/migrations
&& min_version($kvmver, 6, 1) # only need to apply the change since 6.1
return;
}
-PVE::JSONSchema::register_format('pve-qm-usb-device', \&verify_usb_device);
-sub verify_usb_device {
- my ($value, $noerr) = @_;
-
- return $value if parse_usb_device($value);
-
- return if $noerr;
-
- die "unable to parse usb device\n";
-}
-
# add JSON properties for create and set function
sub json_config_properties {
my ($prop, $with_disk_alloc) = @_;
});
}
+ eval { delete_ifaces_ipams_ips($conf, $vmid)};
+ warn $@ if $@;
+
if (defined $replacement_conf) {
PVE::QemuConfig->write_config($vmid, $replacement_conf);
} else {
my ($conf, $noerr) = @_;
my @loc_res = ();
+ my $mapped_res = [];
+
+ my $nodelist = PVE::Cluster::get_nodelist();
+ my $pci_map = PVE::Mapping::PCI::config();
+ my $usb_map = PVE::Mapping::USB::config();
+
+ my $missing_mappings_by_node = { map { $_ => [] } @$nodelist };
+
+ my $add_missing_mapping = sub {
+ my ($type, $key, $id) = @_;
+ for my $node (@$nodelist) {
+ my $entry;
+ if ($type eq 'pci') {
+ $entry = PVE::Mapping::PCI::get_node_mapping($pci_map, $id, $node);
+ } elsif ($type eq 'usb') {
+ $entry = PVE::Mapping::USB::get_node_mapping($usb_map, $id, $node);
+ }
+ if (!scalar($entry->@*)) {
+ push @{$missing_mappings_by_node->{$node}}, $key;
+ }
+ }
+ };
push @loc_res, "hostusb" if $conf->{hostusb}; # old syntax
push @loc_res, "hostpci" if $conf->{hostpci}; # old syntax
push @loc_res, "ivshmem" if $conf->{ivshmem};
foreach my $k (keys %$conf) {
- next if $k =~ m/^usb/ && ($conf->{$k} =~ m/^spice(?![^,])/);
+ if ($k =~ m/^usb/) {
+ my $entry = parse_property_string('pve-qm-usb', $conf->{$k});
+ next if $entry->{host} && $entry->{host} =~ m/^spice$/i;
+ if ($entry->{mapping}) {
+ $add_missing_mapping->('usb', $k, $entry->{mapping});
+ push @$mapped_res, $k;
+ }
+ }
+ if ($k =~ m/^hostpci/) {
+ my $entry = parse_property_string('pve-qm-hostpci', $conf->{$k});
+ if ($entry->{mapping}) {
+ $add_missing_mapping->('pci', $k, $entry->{mapping});
+ push @$mapped_res, $k;
+ }
+ }
# sockets are safe: they will recreated be on the target side post-migrate
next if $k =~ m/^serial/ && ($conf->{$k} eq 'socket');
push @loc_res, $k if $k =~ m/^(usb|hostpci|serial|parallel)\d+$/;
die "VM uses local resources\n" if scalar @loc_res && !$noerr;
- return \@loc_res;
+ return wantarray ? (\@loc_res, $mapped_res, $missing_mappings_by_node) : \@loc_res;
}
# check if used storages are available on all nodes (use by migrate)
optional => 1,
},
qmpstatus => {
- description => "QEMU QMP agent status.",
+ description => "VM run state from the 'query-status' QMP monitor command.",
type => 'string',
optional => 1,
},
$d->{cpus} = $conf->{vcpus} if $conf->{vcpus};
$d->{name} = $conf->{name} || "VM $vmid";
- $d->{maxmem} = $conf->{memory} ? $conf->{memory}*(1024*1024)
- : $defaults->{memory}*(1024*1024);
+ $d->{maxmem} = get_current_memory($conf->{memory})*(1024*1024);
if ($conf->{balloon}) {
$d->{balloon_min} = $conf->{balloon}*(1024*1024);
});
}
+ # Used to distinguish different invocations in the log.
+ my $log_prefix = "[id=" . int(time()) . "] ";
+
my $emulator_cmd = [
"swtpm",
"socket",
"file=$paths->{pid}",
"--terminate", # terminate on QEMU disconnect
"--daemon",
+ "--log",
+ "file=/run/qemu-server/$vmid-swtpm.log,level=1,prefix=$log_prefix",
];
push @$emulator_cmd, "--tpm2" if $tpm->{version} eq 'v2.0';
run_command($emulator_cmd, outfunc => sub { print $1; });
return $1 || 1;
}
-sub is_native($) {
- my ($arch) = @_;
- return get_host_arch() eq $arch;
-}
-
sub get_vm_arch {
my ($conf) = @_;
return $conf->{arch} // get_host_arch();
sub get_vm_machine {
my ($conf, $forcemachine, $arch, $add_pve_version, $kvmversion) = @_;
- my $machine = $forcemachine || $conf->{machine};
+ my $machine_conf = PVE::QemuServer::Machine::parse_machine($conf->{machine});
+ my $machine = $forcemachine || $machine_conf->{type};
if (!$machine || $machine =~ m/^(?:pc|q35|virt)$/) {
$kvmversion //= kvm_user_version();
or die "no OVMF images known for architecture '$arch'\n";
my $type = 'default';
- if (defined($efidisk->{efitype}) && $efidisk->{efitype} eq '4m') {
- $type = $smm ? "4m" : "4m-no-smm";
- $type .= '-ms' if $efidisk->{'pre-enrolled-keys'};
+ if ($arch eq 'x86_64') {
+ if (defined($efidisk->{efitype}) && $efidisk->{efitype} eq '4m') {
+ $type = $smm ? "4m" : "4m-no-smm";
+ $type .= '-ms' if $efidisk->{'pre-enrolled-keys'};
+ } else {
+ # TODO: log_warn about use of legacy images for x86_64 with Promxox VE 9
+ }
}
my ($ovmf_code, $ovmf_vars) = $types->{$type}->@*;
};
sub get_command_for_arch($) {
my ($arch) = @_;
- return '/usr/bin/kvm' if is_native($arch);
+ return '/usr/bin/kvm' if is_native_arch($arch);
my $cmd = $Arch2Qemu->{$arch}
or die "don't know how to emulate architecture '$arch'\n";
# Since commit 277d33454f77ec1d1e0bc04e37621e4dd2424b67 in pve-qemu, smm is not off by default
# anymore. But smm=off seems to be required when using SeaBIOS and serial display.
my sub should_disable_smm {
- my ($conf, $vga) = @_;
+ my ($conf, $vga, $machine) = @_;
+
+ return if $machine =~ m/^virt/; # there is no smm flag that could be disabled
return (!defined($conf->{bios}) || $conf->{bios} eq 'seabios') &&
$vga->{type} && $vga->{type} =~ m/^(serial\d+|none)$/;
sub config_to_command {
my ($storecfg, $vmid, $conf, $defaults, $forcemachine, $forcecpu,
- $pbs_backing) = @_;
+ $live_restore_backing) = @_;
my ($globalFlags, $machineFlags, $rtcFlags) = ([], [], []);
my $devices = [];
my $kvm = $conf->{kvm};
my $nodename = nodename();
+ my $machine_conf = PVE::QemuServer::Machine::parse_machine($conf->{machine});
+
my $arch = get_vm_arch($conf);
my $kvm_binary = get_command_for_arch($arch);
my $kvmver = kvm_user_version($kvm_binary);
my $machine_type = get_vm_machine($conf, $forcemachine, $arch, $add_pve_version);
my $machine_version = extract_version($machine_type, $kvmver);
- $kvm //= 1 if is_native($arch);
+ $kvm //= 1 if is_native_arch($arch);
$machine_version =~ m/(\d+)\.(\d+)/;
my ($machine_major, $machine_minor) = ($1, $2);
}
if ($conf->{bios} && $conf->{bios} eq 'ovmf') {
+ die "OVMF (UEFI) BIOS is not supported on 32-bit CPU types\n"
+ if !$forcecpu && get_cpu_bitness($conf->{cpu}, $arch) == 32;
+
my ($code_drive_str, $var_drive_str) =
print_ovmf_drive_commandlines($conf, $storecfg, $vmid, $arch, $q35, $version_guard);
push $cmd->@*, '-drive', $code_drive_str;
# add usb controllers
my @usbcontrollers = PVE::QemuServer::USB::get_usb_controllers(
- $conf, $bridges, $arch, $machine_type, $usbdesc->{format}, $MAX_USB_DEVICES, $machine_version);
+ $conf, $bridges, $arch, $machine_type, $machine_version);
push @$devices, @usbcontrollers if @usbcontrollers;
my $vga = parse_vga($conf->{vga});
my $bootorder = device_bootorder($conf);
# host pci device passthrough
- my ($kvm_off, $gpu_passthrough, $legacy_igd) = PVE::QemuServer::PCI::print_hostpci_devices(
- $vmid, $conf, $devices, $vga, $winversion, $q35, $bridges, $arch, $machine_type, $bootorder);
+ my ($kvm_off, $gpu_passthrough, $legacy_igd, $pci_devices) = PVE::QemuServer::PCI::print_hostpci_devices(
+ $vmid, $conf, $devices, $vga, $winversion, $bridges, $arch, $machine_type, $bootorder);
# usb devices
my $usb_dev_features = {};
$usb_dev_features->{spice_usb3} = 1 if min_version($machine_version, 4, 0);
my @usbdevices = PVE::QemuServer::USB::get_usb_devices(
- $conf, $usbdesc->{format}, $MAX_USB_DEVICES, $usb_dev_features, $bootorder, $machine_version);
+ $conf, $usb_dev_features, $bootorder, $machine_version);
push @$devices, @usbdevices if @usbdevices;
# serial devices
}
} else {
die "no such serial device\n" if ! -c $path;
- push @$devices, '-chardev', "tty,id=serial$i,path=$path";
+ push @$devices, '-chardev', "serial,id=serial$i,path=$path";
push @$devices, '-device', "isa-serial,chardev=serial$i";
}
}
for (my $i = 0; $i < $MAX_PARALLEL_PORTS; $i++) {
if (my $path = $conf->{"parallel$i"}) {
die "no such parallel device\n" if ! -c $path;
- my $devtype = $path =~ m!^/dev/usb/lp! ? 'tty' : 'parport';
+ my $devtype = $path =~ m!^/dev/usb/lp! ? 'serial' : 'parallel';
push @$devices, '-chardev', "$devtype,id=parallel$i,path=$path";
push @$devices, '-device', "isa-parallel,chardev=parallel$i";
}
push @$devices, @$audio_devs;
}
- add_tpm_device($vmid, $devices, $conf);
+ # Add a TPM only if the VM is not a template,
+ # to support backing up template VMs even if the TPM disk is write-protected.
+ add_tpm_device($vmid, $devices, $conf) if (!PVE::QemuConfig->is_template($conf));
my $sockets = 1;
$sockets = $conf->{smp} if $conf->{smp}; # old style - no longer iused
if ($hotplug_features->{cpu} && min_version($machine_version, 2, 7)) {
push @$cmd, '-smp', "1,sockets=$sockets,cores=$cores,maxcpus=$maxcpus";
for (my $i = 2; $i <= $vcpus; $i++) {
- my $cpustr = print_cpu_device($conf,$i);
+ my $cpustr = print_cpu_device($conf, $arch, $i);
push @$cmd, '-device', $cpustr;
}
push @$cmd, '-boot', "menu=on,strict=on,reboot-timeout=1000,splash=/usr/share/qemu-server/bootsplash.jpg";
- push @$cmd, '-no-acpi' if defined($conf->{acpi}) && $conf->{acpi} == 0;
+ push $machineFlags->@*, 'acpi=off' if defined($conf->{acpi}) && $conf->{acpi} == 0;
push @$cmd, '-no-reboot' if defined($conf->{reboot}) && $conf->{reboot} == 0;
if ($winversion >= 6) {
push @$globalFlags, 'kvm-pit.lost_tick_policy=discard';
- push @$cmd, '-no-hpet';
+ push @$machineFlags, 'hpet=off';
}
push @$rtcFlags, 'driftfix=slew' if $tdf;
push @$cmd, get_cpu_options($conf, $arch, $kvm, $kvm_off, $machine_version, $winversion, $gpu_passthrough);
}
- PVE::QemuServer::Memory::config($conf, $vmid, $sockets, $cores, $defaults, $hotplug_features, $cmd);
+ PVE::QemuServer::Memory::config(
+ $conf, $vmid, $sockets, $cores, $hotplug_features->{memory}, $cmd);
push @$cmd, '-S' if $conf->{freeze};
my $spice_port;
- if ($qxlnum || $vga->{type} =~ /^virtio/) {
+ assert_clipboard_config($vga);
+ my $is_spice = $qxlnum || $vga->{type} =~ /^virtio/;
+
+ if ($is_spice || ($vga->{'clipboard'} && $vga->{'clipboard'} eq 'vnc')) {
if ($qxlnum > 1) {
if ($winversion){
for (my $i = 1; $i < $qxlnum; $i++){
my $pciaddr = print_pci_addr("spice", $bridges, $arch, $machine_type);
- my $pfamily = PVE::Tools::get_host_address_family($nodename);
- my @nodeaddrs = PVE::Tools::getaddrinfo_all('localhost', family => $pfamily);
- die "failed to get an ip address of type $pfamily for 'localhost'\n" if !@nodeaddrs;
-
push @$devices, '-device', "virtio-serial,id=spice$pciaddr";
- push @$devices, '-chardev', "spicevmc,id=vdagent,name=vdagent";
+ if ($vga->{'clipboard'} && $vga->{'clipboard'} eq 'vnc') {
+ push @$devices, '-chardev', 'qemu-vdagent,id=vdagent,name=vdagent,clipboard=on';
+ } else {
+ push @$devices, '-chardev', 'spicevmc,id=vdagent,name=vdagent';
+ }
push @$devices, '-device', "virtserialport,chardev=vdagent,name=com.redhat.spice.0";
- my $localhost = PVE::Network::addr_to_ip($nodeaddrs[0]->{addr});
- $spice_port = PVE::Tools::next_spice_port($pfamily, $localhost);
+ if ($is_spice) {
+ my $pfamily = PVE::Tools::get_host_address_family($nodename);
+ my @nodeaddrs = PVE::Tools::getaddrinfo_all('localhost', family => $pfamily);
+ die "failed to get an ip address of type $pfamily for 'localhost'\n" if !@nodeaddrs;
- my $spice_enhancement_str = $conf->{spice_enhancements} // '';
- my $spice_enhancement = parse_property_string($spice_enhancements_fmt, $spice_enhancement_str);
- if ($spice_enhancement->{foldersharing}) {
- push @$devices, '-chardev', "spiceport,id=foldershare,name=org.spice-space.webdav.0";
- push @$devices, '-device', "virtserialport,chardev=foldershare,name=org.spice-space.webdav.0";
- }
+ my $localhost = PVE::Network::addr_to_ip($nodeaddrs[0]->{addr});
+ $spice_port = PVE::Tools::next_spice_port($pfamily, $localhost);
- my $spice_opts = "tls-port=${spice_port},addr=$localhost,tls-ciphers=HIGH,seamless-migration=on";
- $spice_opts .= ",streaming-video=$spice_enhancement->{videostreaming}"
- if $spice_enhancement->{videostreaming};
+ my $spice_enhancement_str = $conf->{spice_enhancements} // '';
+ my $spice_enhancement = parse_property_string($spice_enhancements_fmt, $spice_enhancement_str);
+ if ($spice_enhancement->{foldersharing}) {
+ push @$devices, '-chardev', "spiceport,id=foldershare,name=org.spice-space.webdav.0";
+ push @$devices, '-device', "virtserialport,chardev=foldershare,name=org.spice-space.webdav.0";
+ }
- push @$devices, '-spice', "$spice_opts";
+ my $spice_opts = "tls-port=${spice_port},addr=$localhost,tls-ciphers=HIGH,seamless-migration=on";
+ $spice_opts .= ",streaming-video=$spice_enhancement->{videostreaming}"
+ if $spice_enhancement->{videostreaming};
+ push @$devices, '-spice', "$spice_opts";
+ }
}
# enable balloon by default, unless explicitly disabled
$ahcicontroller->{$controller}=1;
}
- my $pbs_conf = $pbs_backing->{$ds};
- my $pbs_name = undef;
- if ($pbs_conf) {
- $pbs_name = "drive-$ds-pbs";
- push @$devices, '-blockdev', print_pbs_blockdev($pbs_conf, $pbs_name);
+ my $live_restore = $live_restore_backing->{$ds};
+ my $live_blockdev_name = undef;
+ if ($live_restore) {
+ $live_blockdev_name = $live_restore->{name};
+ push @$devices, '-blockdev', $live_restore->{blockdev};
}
my $drive_cmd = print_drive_commandline_full(
- $storecfg, $vmid, $drive, $pbs_name, min_version($kvmver, 6, 0));
+ $storecfg, $vmid, $drive, $live_blockdev_name, min_version($kvmver, 6, 0));
# extra protection for templates, but SATA and IDE don't support it..
$drive_cmd .= ',readonly=on' if drive_is_read_only($conf, $drive);
push @$machineFlags, 'accel=tcg';
}
- push @$machineFlags, 'smm=off' if should_disable_smm($conf, $vga);
+ push @$machineFlags, 'smm=off' if should_disable_smm($conf, $vga, $machine_type);
my $machine_type_min = $machine_type;
if ($add_pve_version) {
}
push @$machineFlags, "type=${machine_type_min}";
+ PVE::QemuServer::Machine::assert_valid_machine_property($conf, $machine_conf);
+
+ if (my $viommu = $machine_conf->{viommu}) {
+ if ($viommu eq 'intel') {
+ unshift @$devices, '-device', 'intel-iommu,intremap=on,caching-mode=on';
+ push @$machineFlags, 'kernel-irqchip=split';
+ } elsif ($viommu eq 'virtio') {
+ push @$devices, '-device', 'virtio-iommu-pci';
+ }
+ }
+
push @$cmd, @$devices;
push @$cmd, '-rtc', join(',', @$rtcFlags) if scalar(@$rtcFlags);
push @$cmd, '-machine', join(',', @$machineFlags) if scalar(@$machineFlags);
push @$cmd, @$aa;
}
- return wantarray ? ($cmd, $vollist, $spice_port) : $cmd;
+ return wantarray ? ($cmd, $vollist, $spice_port, $pci_devices) : $cmd;
}
sub check_rng_source {
));
}
-sub qemu_deviceadd {
- my ($vmid, $devicefull) = @_;
-
- $devicefull = "driver=".$devicefull;
- my %options = split(/[=,]/, $devicefull);
-
- mon_cmd($vmid, "device_add" , %options);
-}
-
-sub qemu_devicedel {
- my ($vmid, $deviceid) = @_;
-
- my $ret = mon_cmd($vmid, "device_del", id => $deviceid);
-}
-
sub qemu_iothread_add {
my ($vmid, $deviceid, $device) = @_;
}
}
-sub qemu_objectadd {
- my ($vmid, $objectid, $qomtype) = @_;
-
- mon_cmd($vmid, "object-add", id => $objectid, "qom-type" => $qomtype);
-
- return 1;
-}
-
-sub qemu_objectdel {
- my ($vmid, $objectid) = @_;
-
- mon_cmd($vmid, "object-del", id => $objectid);
-
- return 1;
-}
-
sub qemu_driveadd {
my ($storecfg, $vmid, $device) = @_;
qemu_deviceadd($vmid, PVE::QemuServer::USB::print_qemu_xhci_controller($pciaddr));
}
- # print_usbdevice_full expects the parsed device
- my $d = parse_usb_device($device->{host});
- $d->{usb3} = $device->{usb3};
-
# add the new one
- vm_deviceplug($storecfg, $conf, $vmid, $deviceid, $d, $arch, $machine_type);
+ vm_deviceplug($storecfg, $conf, $vmid, $deviceid, $device, $arch, $machine_type);
}
sub qemu_cpu_hotplug {
if scalar(@{$currentrunningvcpus}) != $currentvcpus;
if (PVE::QemuServer::Machine::machine_version($machine_type, 2, 7)) {
+ my $arch = get_vm_arch($conf);
for (my $i = $currentvcpus+1; $i <= $vcpus; $i++) {
- my $cpustr = print_cpu_device($conf, $i);
+ my $cpustr = print_cpu_device($conf, $arch, $i);
qemu_deviceadd($vmid, $cpustr);
my $retry = 0;
my $running = check_running($vmid);
- $size = 0 if !PVE::Storage::volume_resize($storecfg, $volid, $size, $running);
+ PVE::Storage::volume_resize($storecfg, $volid, $size, $running);
return if !$running;
}
sub qemu_volume_snapshot_delete {
- my ($vmid, $deviceid, $storecfg, $volid, $snap) = @_;
+ my ($vmid, $storecfg, $volid, $snap) = @_;
my $running = check_running($vmid);
+ my $attached_deviceid;
- if($running) {
-
- $running = undef;
+ if ($running) {
my $conf = PVE::QemuConfig->load_config($vmid);
PVE::QemuConfig->foreach_volume($conf, sub {
my ($ds, $drive) = @_;
- $running = 1 if $drive->{file} eq $volid;
+ $attached_deviceid = "drive-$ds" if $drive->{file} eq $volid;
});
}
- if ($running && do_snapshots_with_qemu($storecfg, $volid, $deviceid)) {
- mon_cmd($vmid, 'blockdev-snapshot-delete-internal-sync', device => $deviceid, name => $snap);
+ if ($attached_deviceid && do_snapshots_with_qemu($storecfg, $volid, $attached_deviceid)) {
+ mon_cmd(
+ $vmid,
+ 'blockdev-snapshot-delete-internal-sync',
+ device => $attached_deviceid,
+ name => $snap,
+ );
} else {
- PVE::Storage::volume_snapshot_delete($storecfg, $volid, $snap, $running);
+ PVE::Storage::volume_snapshot_delete(
+ $storecfg, $volid, $snap, $attached_deviceid ? 1 : undef);
}
}
my $volhash = {};
my $test_volid = sub {
- my ($key, $drive, $snapname) = @_;
+ my ($key, $drive, $snapname, $pending) = @_;
my $volid = $drive->{file};
return if !$volid;
$volhash->{$volid}->{shared} //= 0;
$volhash->{$volid}->{shared} = 1 if $drive->{shared};
- $volhash->{$volid}->{referenced_in_config} //= 0;
- $volhash->{$volid}->{referenced_in_config} = 1 if !defined($snapname);
+ $volhash->{$volid}->{is_unused} //= 0;
+ $volhash->{$volid}->{is_unused} = 1 if $key =~ /^unused\d+$/;
+
+ $volhash->{$volid}->{is_attached} //= 0;
+ $volhash->{$volid}->{is_attached} = 1
+ if !$volhash->{$volid}->{is_unused} && !defined($snapname) && !$pending;
$volhash->{$volid}->{referenced_in_snapshot}->{$snapname} = 1
if defined($snapname);
+ $volhash->{$volid}->{referenced_in_pending} = 1 if $pending;
+
my $size = $drive->{size};
$volhash->{$volid}->{size} //= $size if $size;
$volhash->{$volid}->{is_tpmstate} //= 0;
$volhash->{$volid}->{is_tpmstate} = 1 if $key eq 'tpmstate0';
- $volhash->{$volid}->{is_unused} //= 0;
- $volhash->{$volid}->{is_unused} = 1 if $key =~ /^unused\d+$/;
-
$volhash->{$volid}->{drivename} = $key if is_valid_drivename($key);
};
};
PVE::QemuConfig->foreach_volume_full($conf, $include_opts, $test_volid);
+
+ PVE::QemuConfig->foreach_volume_full($conf->{pending}, $include_opts, $test_volid, undef, 1)
+ if defined($conf->{pending}) && $conf->{pending}->%*;
+
foreach my $snapname (keys %{$conf->{snapshots}}) {
my $snap = $conf->{snapshots}->{$snapname};
PVE::QemuConfig->foreach_volume_full($snap, $include_opts, $test_volid, $snapname);
}
my $fast_plug_option = {
+ 'description' => 1,
+ 'hookscript' => 1,
'lock' => 1,
+ 'migrate_downtime' => 1,
+ 'migrate_speed' => 1,
'name' => 1,
'onboot' => 1,
+ 'protection' => 1,
'shares' => 1,
'startup' => 1,
- 'description' => 1,
- 'protection' => 1,
- 'vmstatestorage' => 1,
- 'hookscript' => 1,
'tags' => 1,
+ 'vmstatestorage' => 1,
};
for my $opt (keys %$confdesc_cloudinit) {
my $force = $pending_delete_hash->{$opt}->{force};
eval {
if ($opt eq 'hotplug') {
- die "skip\n" if ($conf->{hotplug} =~ /memory/);
+ die "skip\n" if ($conf->{hotplug} =~ /(cpu|memory)/);
} elsif ($opt eq 'tablet') {
die "skip\n" if !$hotplug_features->{usb};
if ($defaults->{tablet}) {
# enable balloon device is not hotpluggable
die "skip\n" if defined($conf->{balloon}) && $conf->{balloon} == 0;
# here we reset the ballooning value to memory
- my $balloon = $conf->{memory} || $defaults->{memory};
+ my $balloon = get_current_memory($conf->{memory});
mon_cmd($vmid, "balloon", value => $balloon*1024*1024);
} elsif ($fast_plug_option->{$opt}) {
# do nothing
} elsif ($opt =~ m/^net(\d+)$/) {
die "skip\n" if !$hotplug_features->{network};
vm_deviceunplug($vmid, $conf, $opt);
+ if($have_sdn) {
+ my $net = PVE::QemuServer::parse_net($conf->{$opt});
+ PVE::Network::SDN::Vnets::del_ips_from_mac($net->{bridge}, $net->{macaddr}, $conf->{name});
+ }
} elsif (is_valid_drivename($opt)) {
die "skip\n" if !$hotplug_features->{disk} || $opt =~ m/(ide|sata)(\d+)/;
vm_deviceunplug($vmid, $conf, $opt);
vmconfig_delete_or_detach_drive($vmid, $storecfg, $conf, $opt, $force);
} elsif ($opt =~ m/^memory$/) {
die "skip\n" if !$hotplug_features->{memory};
- PVE::QemuServer::Memory::qemu_memory_hotplug($vmid, $conf, $defaults, $opt);
+ PVE::QemuServer::Memory::qemu_memory_hotplug($vmid, $conf);
} elsif ($opt eq 'cpuunits') {
$cgroup->change_cpu_shares(undef);
} elsif ($opt eq 'cpulimit') {
eval {
if ($opt eq 'hotplug') {
die "skip\n" if ($value =~ /memory/) || ($value !~ /memory/ && $conf->{hotplug} =~ /memory/);
+ die "skip\n" if ($value =~ /cpu/) || ($value !~ /cpu/ && $conf->{hotplug} =~ /cpu/);
} elsif ($opt eq 'tablet') {
die "skip\n" if !$hotplug_features->{usb};
if ($value == 1) {
} elsif ($opt =~ m/^usb(\d+)$/) {
my $index = $1;
die "skip\n" if !$usb_hotplug;
- my $d = eval { parse_property_string($usbdesc->{format}, $value) };
+ my $d = eval { parse_property_string('pve-qm-usb', $value) };
my $id = $opt;
- if ($d->{host} eq 'spice') {
+ if ($d->{host} =~ m/^spice$/i) {
$id = "usbredirdev$index";
}
qemu_usb_hotplug($storecfg, $conf, $vmid, $id, $d, $arch, $machine_type);
# allow manual ballooning if shares is set to zero
if ((defined($conf->{shares}) && ($conf->{shares} == 0))) {
- my $balloon = $conf->{pending}->{balloon} || $conf->{memory} || $defaults->{memory};
+ my $memory = get_current_memory($conf->{memory});
+ my $balloon = $conf->{pending}->{balloon} || $memory;
mon_cmd($vmid, "balloon", value => $balloon*1024*1024);
}
} elsif ($opt =~ m/^net(\d+)$/) {
$vmid, $opt, $value, $arch, $machine_type);
} elsif ($opt =~ m/^memory$/) { #dimms
die "skip\n" if !$hotplug_features->{memory};
- $value = PVE::QemuServer::Memory::qemu_memory_hotplug($vmid, $conf, $defaults, $opt, $value);
+ $value = PVE::QemuServer::Memory::qemu_memory_hotplug($vmid, $conf, $value);
} elsif ($opt eq 'cpuunits') {
my $new_cpuunits = PVE::CGroup::clamp_cpu_shares($conf->{pending}->{$opt}); #clamp
$cgroup->change_cpu_shares($new_cpuunits);
# unplug xhci controller if no usb device is left
if ($usb_hotplug) {
my $has_usb = 0;
- for (my $i = 0; $i < $MAX_USB_DEVICES; $i++) {
+ for (my $i = 0; $i < $PVE::QemuServer::USB::MAX_USB_DEVICES; $i++) {
next if !defined($conf->{"usb$i"});
$has_usb = 1;
last;
die "internal error";
} elsif (defined($conf->{$opt}) && is_valid_drivename($opt)) {
vmconfig_delete_or_detach_drive($vmid, $storecfg, $conf, $opt, $force);
+ } elsif (defined($conf->{$opt}) && $opt =~ m/^net\d+$/) {
+ if($have_sdn) {
+ my $net = PVE::QemuServer::parse_net($conf->{$opt});
+ eval { PVE::Network::SDN::Vnets::del_ips_from_mac($net->{bridge}, $net->{macaddr}, $conf->{name}) };
+ warn if $@;
+ }
}
};
if (my $err = $@) {
eval {
if (defined($conf->{$opt}) && is_valid_drivename($opt)) {
vmconfig_register_unused_drive($storecfg, $vmid, $conf, parse_drive($opt, $conf->{$opt}))
+ } elsif (defined($conf->{pending}->{$opt}) && $opt =~ m/^net\d+$/) {
+ return if !$have_sdn; # return from eval if SDN is not available
+
+ my $new_net = PVE::QemuServer::parse_net($conf->{pending}->{$opt});
+ if ($conf->{$opt}) {
+ my $old_net = PVE::QemuServer::parse_net($conf->{$opt});
+
+ if (defined($old_net->{bridge}) && defined($old_net->{macaddr}) && (
+ safe_string_ne($old_net->{bridge}, $new_net->{bridge}) ||
+ safe_string_ne($old_net->{macaddr}, $new_net->{macaddr})
+ )) {
+ PVE::Network::SDN::Vnets::del_ips_from_mac($old_net->{bridge}, $old_net->{macaddr}, $conf->{name});
+ }
+ }
+ #fixme: reuse ip if mac change && same bridge
+ PVE::Network::SDN::Vnets::add_next_free_cidr($new_net->{bridge}, $conf->{name}, $new_net->{macaddr}, $vmid, undef, 1);
}
};
if (my $err = $@) {
if (safe_string_ne($oldnet->{model}, $newnet->{model}) ||
safe_string_ne($oldnet->{macaddr}, $newnet->{macaddr}) ||
safe_num_ne($oldnet->{queues}, $newnet->{queues}) ||
- !($newnet->{bridge} && $oldnet->{bridge})) { # bridge/nat mode change
+ safe_num_ne($oldnet->{mtu}, $newnet->{mtu}) ||
+ !($newnet->{bridge} && $oldnet->{bridge})
+ ) { # bridge/nat mode change
# for non online change, we try to hot-unplug
die "skip\n" if !$hotplug;
vm_deviceunplug($vmid, $conf, $opt);
+
+ if ($have_sdn) {
+ PVE::Network::SDN::Vnets::del_ips_from_mac($oldnet->{bridge}, $oldnet->{macaddr}, $conf->{name});
+ }
+
} else {
die "internal error" if $opt !~ m/net(\d+)/;
if (safe_string_ne($oldnet->{bridge}, $newnet->{bridge}) ||
safe_num_ne($oldnet->{tag}, $newnet->{tag}) ||
safe_string_ne($oldnet->{trunks}, $newnet->{trunks}) ||
- safe_num_ne($oldnet->{firewall}, $newnet->{firewall})) {
+ safe_num_ne($oldnet->{firewall}, $newnet->{firewall})
+ ) {
PVE::Network::tap_unplug($iface);
+ #set link_down in guest if bridge or vlan change to notify guest (dhcp renew for example)
+ if (safe_string_ne($oldnet->{bridge}, $newnet->{bridge}) ||
+ safe_num_ne($oldnet->{tag}, $newnet->{tag})
+ ) {
+ qemu_set_link_status($vmid, $opt, 0);
+ }
+
+ if (safe_string_ne($oldnet->{bridge}, $newnet->{bridge})) {
+ if ($have_sdn) {
+ PVE::Network::SDN::Vnets::del_ips_from_mac($oldnet->{bridge}, $oldnet->{macaddr}, $conf->{name});
+ PVE::Network::SDN::Vnets::add_next_free_cidr($newnet->{bridge}, $conf->{name}, $newnet->{macaddr}, $vmid, undef, 1);
+ }
+ }
+
if ($have_sdn) {
PVE::Network::SDN::Zones::tap_plug($iface, $newnet->{bridge}, $newnet->{tag}, $newnet->{firewall}, $newnet->{trunks}, $newnet->{rate});
} else {
PVE::Network::tap_plug($iface, $newnet->{bridge}, $newnet->{tag}, $newnet->{firewall}, $newnet->{trunks}, $newnet->{rate});
}
+
+ #set link_up in guest if bridge or vlan change to notify guest (dhcp renew for example)
+ if (safe_string_ne($oldnet->{bridge}, $newnet->{bridge}) ||
+ safe_num_ne($oldnet->{tag}, $newnet->{tag})
+ ) {
+ qemu_set_link_status($vmid, $opt, 1);
+ }
+
} elsif (safe_num_ne($oldnet->{rate}, $newnet->{rate})) {
# Rate can be applied on its own but any change above needs to
# include the rate in tap_plug since OVS resets everything.
}
if ($hotplug) {
+ if ($have_sdn) {
+ PVE::Network::SDN::Vnets::add_next_free_cidr($newnet->{bridge}, $conf->{name}, $newnet->{macaddr}, $vmid, undef, 1);
+ PVE::Network::SDN::Vnets::add_dhcp_mapping($newnet->{bridge}, $newnet->{macaddr}, $vmid, $conf->{name});
+ }
vm_deviceplug($storecfg, $conf, $vmid, $opt, $newnet, $arch, $machine_type);
} else {
die "skip\n";
# update existing disk
# skip non hotpluggable value
- if (safe_string_ne($drive->{discard}, $old_drive->{discard}) ||
+ if (safe_string_ne($drive->{aio}, $old_drive->{aio}) ||
+ safe_string_ne($drive->{discard}, $old_drive->{discard}) ||
safe_string_ne($drive->{iothread}, $old_drive->{iothread}) ||
safe_string_ne($drive->{queues}, $old_drive->{queues}) ||
+ safe_string_ne($drive->{product}, $old_drive->{product}) ||
safe_string_ne($drive->{cache}, $old_drive->{cache}) ||
safe_string_ne($drive->{ssd}, $old_drive->{ssd}) ||
+ safe_string_ne($drive->{vendor}, $old_drive->{vendor}) ||
safe_string_ne($drive->{ro}, $old_drive->{ro})) {
die "skip\n";
}
my $scfg = PVE::Storage::storage_config($storecfg, $storeid);
return if $scfg->{shared};
+ my $format = qemu_img_format($scfg, $volname);
+
# replicated disks re-use existing state via bitmap
my $use_existing = $replicated_volumes->{$volid} ? 1 : 0;
- $local_volumes->{$ds} = [$volid, $storeid, $volname, $drive, $use_existing];
+ $local_volumes->{$ds} = [$volid, $storeid, $drive, $use_existing, $format];
});
return $local_volumes;
}
my $nbd = {};
foreach my $opt (sort keys %$source_volumes) {
- my ($volid, $storeid, $volname, $drive, $use_existing, $format) = @{$source_volumes->{$opt}};
+ my ($volid, $storeid, $drive, $use_existing, $format) = @{$source_volumes->{$opt}};
if ($use_existing) {
$nbd->{$opt}->{drivestr} = print_drive($drive);
next;
}
- # storage mapping + volname = regular migration
- # storage mapping + format = remote migration
+ $storeid = PVE::JSONSchema::map_id($storagemap, $storeid);
+
# order of precedence, filtered by whether storage supports it:
# 1. explicit requested format
- # 2. format of current volume
- # 3. default format of storage
- if (!$storagemap->{identity}) {
- $storeid = PVE::JSONSchema::map_id($storagemap, $storeid);
- my ($defFormat, $validFormats) = PVE::Storage::storage_default_format($storecfg, $storeid);
- if (!$format || !grep { $format eq $_ } @$validFormats) {
- if ($volname) {
- my $scfg = PVE::Storage::storage_config($storecfg, $storeid);
- my $fileFormat = qemu_img_format($scfg, $volname);
- $format = $fileFormat
- if grep { $fileFormat eq $_ } @$validFormats;
- }
- $format //= $defFormat;
- }
- } else {
- # can't happen for remote migration, so $volname is always defined
- my $scfg = PVE::Storage::storage_config($storecfg, $storeid);
- $format = qemu_img_format($scfg, $volname);
- }
+ # 2. default format of storage
+ my ($defFormat, $validFormats) = PVE::Storage::storage_default_format($storecfg, $storeid);
+ $format = $defFormat if !$format || !grep { $format eq $_ } $validFormats->@*;
my $size = $drive->{size} / 1024;
my $newvolid = PVE::Storage::vdisk_alloc($storecfg, $storeid, $vmid, $format, undef, $size);
# timeout => in seconds
# paused => start VM in paused state (backup)
# resume => resume from hibernation
-# pbs-backing => {
+# live-restore-backing => {
# sata0 => {
-# repository
-# snapshot
-# keyfile
-# archive
+# name => blockdev-name,
+# blockdev => "arg to the -blockdev command instantiating device named 'name'",
# },
# virtio2 => ...
# }
print "Resuming suspended VM\n";
}
- my ($cmd, $vollist, $spice_port) = config_to_command($storecfg, $vmid,
- $conf, $defaults, $forcemachine, $forcecpu, $params->{'pbs-backing'});
+ my ($cmd, $vollist, $spice_port, $pci_devices) = config_to_command($storecfg, $vmid,
+ $conf, $defaults, $forcemachine, $forcecpu, $params->{'live-restore-backing'});
my $migration_ip;
my $get_migration_ip = sub {
$migrate->{addr} = "[$migrate->{addr}]" if Net::IP::ip_is_ipv6($migrate->{addr});
}
- my $pfamily = PVE::Tools::get_host_address_family($nodename);
- $migrate->{port} = PVE::Tools::next_migrate_port($pfamily);
- $migrate->{uri} = "tcp:$migrate->{addr}:$migrate->{port}";
- push @$cmd, '-incoming', $migrate->{uri};
+ # see #4501: port reservation should be done close to usage - tell QEMU where to listen
+ # via QMP later
+ push @$cmd, '-incoming', 'defer';
push @$cmd, '-S';
} elsif ($statefile eq 'unix') {
push @$cmd, '-S';
}
- my $start_timeout = $params->{timeout} // config_aware_timeout($conf, $resume);
+ my $memory = get_current_memory($conf->{memory});
+ my $start_timeout = $params->{timeout} // config_aware_timeout($conf, $memory, $resume);
- my $pci_devices = {}; # host pci devices
- for (my $i = 0; $i < $PVE::QemuServer::PCI::MAX_HOSTPCI_DEVICES; $i++) {
- my $dev = $conf->{"hostpci$i"} or next;
- $pci_devices->{$i} = parse_hostpci($dev);
+ my $pci_reserve_list = [];
+ for my $device (values $pci_devices->%*) {
+ next if $device->{mdev}; # we don't reserve for mdev devices
+ push $pci_reserve_list->@*, map { $_->{id} } $device->{ids}->@*;
}
- # do not reserve pciid for mediated devices, sysfs will error out for duplicate assignment
- my $real_pci_devices = [ grep { !(defined($_->{mdev}) && scalar($_->{pciid}->@*) == 1) } values $pci_devices->%* ];
-
- # map to a flat list of pci ids
- my $pci_id_list = [ map { $_->{id} } map { $_->{pciid}->@* } $real_pci_devices->@* ];
-
# reserve all PCI IDs before actually doing anything with them
- PVE::QemuServer::PCI::reserve_pci_usage($pci_id_list, $vmid, $start_timeout);
+ PVE::QemuServer::PCI::reserve_pci_usage($pci_reserve_list, $vmid, $start_timeout);
eval {
my $uuid;
for my $id (sort keys %$pci_devices) {
my $d = $pci_devices->{$id};
- for my $dev ($d->{pciid}->@*) {
- my $info = PVE::QemuServer::PCI::prepare_pci_device($vmid, $dev->{id}, $id, $d->{mdev});
+ my ($index) = ($id =~ m/^hostpci(\d+)$/);
+
+ my $chosen_mdev;
+ for my $dev ($d->{ids}->@*) {
+ my $info = eval { PVE::QemuServer::PCI::prepare_pci_device($vmid, $dev->{id}, $index, $d->{mdev}) };
+ if ($d->{mdev}) {
+ warn $@ if $@;
+ $chosen_mdev = $info;
+ last if $chosen_mdev; # if successful, we're done
+ } else {
+ die $@ if $@;
+ }
+ }
+
+ next if !$d->{mdev};
+ die "could not create mediated device\n" if !defined($chosen_mdev);
- # nvidia grid needs the uuid of the mdev as qemu parameter
- if ($d->{mdev} && !defined($uuid) && $info->{vendor} eq '10de') {
- $uuid = PVE::QemuServer::PCI::generate_mdev_uuid($vmid, $id);
+ # nvidia grid needs the uuid of the mdev as qemu parameter
+ if (!defined($uuid) && $chosen_mdev->{vendor} =~ m/^(0x)?10de$/) {
+ if (defined($conf->{smbios1})) {
+ my $smbios_conf = parse_smbios1($conf->{smbios1});
+ $uuid = $smbios_conf->{uuid} if defined($smbios_conf->{uuid});
}
+ $uuid = PVE::QemuServer::PCI::generate_mdev_uuid($vmid, $index) if !defined($uuid);
}
}
push @$cmd, '-uuid', $uuid if defined($uuid);
PVE::Storage::activate_volumes($storecfg, $vollist);
- eval {
- run_command(['/bin/systemctl', 'stop', "$vmid.scope"], outfunc => sub{}, errfunc => sub{});
- };
+
+ my %silence_std_outs = (outfunc => sub {}, errfunc => sub {});
+ eval { run_command(['/bin/systemctl', 'reset-failed', "$vmid.scope"], %silence_std_outs) };
+ eval { run_command(['/bin/systemctl', 'stop', "$vmid.scope"], %silence_std_outs) };
# Issues with the above 'stop' not being fully completed are extremely rare, a very low
# timeout should be more than enough here...
PVE::Systemd::wait_for_unit_removed("$vmid.scope", 20);
PVE::Systemd::enter_systemd_scope($vmid, "Proxmox VE VM $vmid", %systemd_properties);
my $tpmpid;
- if (my $tpm = $conf->{tpmstate0}) {
+ if ((my $tpm = $conf->{tpmstate0}) && !PVE::QemuConfig->is_template($conf)) {
# start the TPM emulator so QEMU can connect on start
$tpmpid = start_swtpm($storecfg, $vmid, $tpm, $migratedfrom);
}
if ($conf->{hugepages}) {
my $code = sub {
- my $hugepages_topology = PVE::QemuServer::Memory::hugepages_topology($conf);
+ my $hotplug_features =
+ parse_hotplug_features(defined($conf->{hotplug}) ? $conf->{hotplug} : '1');
+ my $hugepages_topology =
+ PVE::QemuServer::Memory::hugepages_topology($conf, $hotplug_features->{memory});
+
my $hugepages_host_topology = PVE::QemuServer::Memory::hugepages_host_topology();
PVE::QemuServer::Memory::hugepages_mount();
# re-reserve all PCI IDs now that we can know the actual VM PID
my $pid = PVE::QemuServer::Helpers::vm_running_locally($vmid);
- eval { PVE::QemuServer::PCI::reserve_pci_usage($pci_id_list, $vmid, undef, $pid) };
+ eval { PVE::QemuServer::PCI::reserve_pci_usage($pci_reserve_list, $vmid, undef, $pid) };
warn $@ if $@;
- if (defined($res->{migrate})) {
- print "migration listens on $res->{migrate}->{uri}\n";
+ if (defined(my $migrate = $res->{migrate})) {
+ if ($migrate->{proto} eq 'tcp') {
+ my $nodename = nodename();
+ my $pfamily = PVE::Tools::get_host_address_family($nodename);
+ $migrate->{port} = PVE::Tools::next_migrate_port($pfamily);
+ $migrate->{uri} = "tcp:$migrate->{addr}:$migrate->{port}";
+ mon_cmd($vmid, "migrate-incoming", uri => $migrate->{uri});
+ }
+ print "migration listens on $migrate->{uri}\n";
} elsif ($statefile) {
eval { mon_cmd($vmid, "cont"); };
warn $@ if $@;
}
if (!defined($conf->{balloon}) || $conf->{balloon}) {
- mon_cmd(
- $vmid,
- 'qom-set',
- path => "machine/peripheral/balloon0",
- property => "guest-stats-polling-interval",
- value => 2
- );
+ eval {
+ mon_cmd(
+ $vmid,
+ 'qom-set',
+ path => "machine/peripheral/balloon0",
+ property => "guest-stats-polling-interval",
+ value => 2
+ );
+ };
+ log_warn("could not set polling interval for ballooning - $@") if $@;
}
if ($resume) {
PVE::GuestHelpers::exec_hookscript($conf, $vmid, 'post-start');
+ my ($current_machine, $is_deprecated) =
+ PVE::QemuServer::Machine::get_current_qemu_machine($vmid);
+ if ($is_deprecated) {
+ log_warn(
+ "current machine version '$current_machine' is deprecated - see the documentation and ".
+ "change to a newer one",
+ );
+ }
+
return $res;
}
# NOTE: avoid PVE::SysFSTools::pci_cleanup_mdev_device as it requires PCI ID and we
# don't want to break ABI just for this two liner
my $dev_sysfs_dir = "/sys/bus/mdev/devices/$uuid";
- PVE::SysFSTools::file_write("$dev_sysfs_dir/remove", "1") if -e $dev_sysfs_dir;
+
+ # some nvidia vgpu driver versions want to clean the mdevs up themselves, and error
+ # out when we do it first. so wait for up to 10 seconds and then try it manually
+ if ($d->{ids}->[0]->[0]->{vendor} =~ m/^(0x)?10de$/ && -e $dev_sysfs_dir) {
+ my $count = 0;
+ while (-e $dev_sysfs_dir && $count < 10) {
+ sleep 1;
+ $count++;
+ }
+ print "waited $count seconds for mediated device driver finishing clean up\n";
+ }
+
+ if (-e $dev_sysfs_dir) {
+ print "actively clean up mediated device with UUID $uuid\n";
+ PVE::SysFSTools::file_write("$dev_sysfs_dir/remove", "1");
+ }
}
}
PVE::QemuServer::PCI::remove_pci_reservation($vmid);
});
}
+sub check_bridge_access {
+ my ($rpcenv, $authuser, $conf) = @_;
+
+ return 1 if $authuser eq 'root@pam';
+
+ for my $opt (sort keys $conf->%*) {
+ next if $opt !~ m/^net\d+$/;
+ my $net = parse_net($conf->{$opt});
+ my ($bridge, $tag, $trunks) = $net->@{'bridge', 'tag', 'trunks'};
+ PVE::GuestHelpers::check_vnet_access($rpcenv, $authuser, $bridge, $tag, $trunks);
+ }
+ return 1;
+};
+
+sub check_mapping_access {
+ my ($rpcenv, $user, $conf) = @_;
+
+ for my $opt (keys $conf->%*) {
+ if ($opt =~ m/^usb\d+$/) {
+ my $device = PVE::JSONSchema::parse_property_string('pve-qm-usb', $conf->{$opt});
+ if (my $host = $device->{host}) {
+ die "only root can set '$opt' config for real devices\n"
+ if $host !~ m/^spice$/i && $user ne 'root@pam';
+ } elsif ($device->{mapping}) {
+ $rpcenv->check_full($user, "/mapping/usb/$device->{mapping}", ['Mapping.Use']);
+ } else {
+ die "either 'host' or 'mapping' must be set.\n";
+ }
+ } elsif ($opt =~ m/^hostpci\d+$/) {
+ my $device = PVE::JSONSchema::parse_property_string('pve-qm-hostpci', $conf->{$opt});
+ if ($device->{host}) {
+ die "only root can set '$opt' config for non-mapped devices\n" if $user ne 'root@pam';
+ } elsif ($device->{mapping}) {
+ $rpcenv->check_full($user, "/mapping/pci/$device->{mapping}", ['Mapping.Use']);
+ } else {
+ die "either 'host' or 'mapping' must be set.\n";
+ }
+ }
+ }
+};
+
+sub check_restore_permissions {
+ my ($rpcenv, $user, $conf) = @_;
+
+ check_bridge_access($rpcenv, $user, $conf);
+ check_mapping_access($rpcenv, $user, $conf);
+}
# vzdump restore implementaion
sub tar_archive_read_firstfile {
}
};
-my $restore_merge_config = sub {
+sub restore_merge_config {
my ($filename, $backup_conf_raw, $override_conf) = @_;
my $backup_conf = parse_vm_config($filename, $backup_conf_raw);
}
return $backup_conf;
-};
+}
sub scan_volids {
my ($cfg, $vmid) = @_;
$new_conf_raw .= "\nlock: create";
}
- my $new_conf = $restore_merge_config->($conffile, $new_conf_raw, $options->{override_conf});
+ my $new_conf = restore_merge_config($conffile, $new_conf_raw, $options->{override_conf});
+ check_restore_permissions($rpcenv, $user, $new_conf);
PVE::QemuConfig->write_config($vmid, $new_conf);
eval { rescan($vmid, 1); };
print "starting VM for live-restore\n";
print "repository: '$opts->{repo}', snapshot: '$opts->{snapshot}'\n";
- my $pbs_backing = {};
+ my $live_restore_backing = {};
for my $ds (keys %$restored_disks) {
$ds =~ m/^drive-(.*)$/;
my $confname = $1;
- $pbs_backing->{$confname} = {
+ my $pbs_conf = {};
+ $pbs_conf = {
repository => $opts->{repo},
snapshot => $opts->{snapshot},
archive => "$ds.img.fidx",
};
- $pbs_backing->{$confname}->{keyfile} = $opts->{keyfile} if -e $opts->{keyfile};
- $pbs_backing->{$confname}->{namespace} = $opts->{namespace} if defined($opts->{namespace});
+ $pbs_conf->{keyfile} = $opts->{keyfile} if -e $opts->{keyfile};
+ $pbs_conf->{namespace} = $opts->{namespace} if defined($opts->{namespace});
my $drive = parse_drive($confname, $conf->{$confname});
print "restoring '$ds' to '$drive->{file}'\n";
+
+ my $pbs_name = "drive-${confname}-pbs";
+ $live_restore_backing->{$confname} = {
+ name => $pbs_name,
+ blockdev => print_pbs_blockdev($pbs_conf, $pbs_name),
+ };
}
my $drives_streamed = 0;
# start VM with backing chain pointing to PBS backup, environment vars for PBS driver
# in QEMU (PBS_PASSWORD and PBS_FINGERPRINT) are already set by our caller
- vm_start_nolock($storecfg, $vmid, $conf, {paused => 1, 'pbs-backing' => $pbs_backing}, {});
+ vm_start_nolock($storecfg, $vmid, $conf, {paused => 1, 'live-restore-backing' => $live_restore_backing}, {});
my $qmeventd_fd = register_qmeventd_handle($vmid);
}
}
+# Inspired by pbs live-restore, this restores with the disks being available as files.
+# Theoretically this can also be used to quick-start a full-clone vm if the
+# disks are all available as files.
+#
+# The mapping should provide a path by config entry, such as
+# `{ scsi0 => { format => <qcow2|raw|...>, path => "/path/to/file", sata1 => ... } }`
+#
+# This is used when doing a `create` call with the `--live-import` parameter,
+# where the disks get an `import-from=` property. The non-live part is
+# therefore already handled in the `$create_disks()` call happening in the
+# `create` api call
+sub live_import_from_files {
+ my ($mapping, $vmid, $conf, $restore_options) = @_;
+
+ my $live_restore_backing = {};
+ for my $dev (keys %$mapping) {
+ die "disk not support for live-restoring: '$dev'\n"
+ if !is_valid_drivename($dev) || $dev =~ /^(?:efidisk|tpmstate)/;
+
+ die "mapping contains disk '$dev' which does not exist in the config\n"
+ if !exists($conf->{$dev});
+
+ my $info = $mapping->{$dev};
+ my ($format, $path) = $info->@{qw(format path)};
+ die "missing path for '$dev' mapping\n" if !$path;
+ die "missing format for '$dev' mapping\n" if !$format;
+ die "invalid format '$format' for '$dev' mapping\n"
+ if !grep { $format eq $_ } qw(raw qcow2 vmdk);
+
+ $live_restore_backing->{$dev} = {
+ name => "drive-$dev-restore",
+ blockdev => "driver=$format,node-name=drive-$dev-restore"
+ . ",read-only=on"
+ . ",file.driver=file,file.filename=$path"
+ };
+ };
+
+ my $storecfg = PVE::Storage::config();
+ eval {
+
+ # make sure HA doesn't interrupt our restore by stopping the VM
+ if (PVE::HA::Config::vm_is_ha_managed($vmid)) {
+ run_command(['ha-manager', 'set', "vm:$vmid", '--state', 'started']);
+ }
+
+ vm_start_nolock($storecfg, $vmid, $conf, {paused => 1, 'live-restore-backing' => $live_restore_backing}, {});
+
+ # prevent shutdowns from qmeventd when the VM powers off from the inside
+ my $qmeventd_fd = register_qmeventd_handle($vmid);
+
+ # begin streaming, i.e. data copy from PBS to target disk for every vol,
+ # this will effectively collapse the backing image chain consisting of
+ # [target <- alloc-track -> PBS snapshot] to just [target] (alloc-track
+ # removes itself once all backing images vanish with 'auto-remove=on')
+ my $jobs = {};
+ for my $ds (sort keys %$live_restore_backing) {
+ my $job_id = "restore-$ds";
+ mon_cmd($vmid, 'block-stream',
+ 'job-id' => $job_id,
+ device => "drive-$ds",
+ );
+ $jobs->{$job_id} = {};
+ }
+
+ mon_cmd($vmid, 'cont');
+ qemu_drive_mirror_monitor($vmid, undef, $jobs, 'auto', 0, 'stream');
+
+ print "restore-drive jobs finished successfully, removing all tracking block devices\n";
+
+ for my $ds (sort keys %$live_restore_backing) {
+ mon_cmd($vmid, 'blockdev-del', 'node-name' => "drive-$ds-restore");
+ }
+
+ close($qmeventd_fd);
+ };
+
+ my $err = $@;
+
+ if ($err) {
+ warn "An error occurred during live-restore: $err\n";
+ _do_vm_stop($storecfg, $vmid, 1, 1, 10, 0, 1);
+ die "live-restore failed\n";
+ }
+
+ PVE::QemuConfig->remove_lock($vmid, "import");
+}
+
sub restore_vma_archive {
my ($archive, $vmid, $user, $opts, $comp) = @_;
$add_pipe->(['vma', 'extract', '-v', '-r', $mapfifo, $readfrom, $tmpdir]);
- my $oldtimeout;
- my $timeout = 5;
-
my $devinfo = {}; # info about drives included in backup
my $virtdev_hash = {}; # info about allocated drives
$fh->close();
};
+ my $oldtimeout;
+
eval {
# enable interrupts
local $SIG{INT} =
local $SIG{PIPE} = sub { die "interrupted by signal\n"; };
local $SIG{ALRM} = sub { die "got timeout\n"; };
- $oldtimeout = alarm($timeout);
+ $oldtimeout = alarm(5); # for reading the VMA header - might hang with a corrupted one
my $parser = sub {
my $line = shift;
$devinfo->{$devname} = { size => $size, dev_id => $dev_id };
} elsif ($line =~ m/^CTIME: /) {
# we correctly received the vma config, so we can disable
- # the timeout now for disk allocation (set to 10 minutes, so
- # that we always timeout if something goes wrong)
- alarm(600);
+ # the timeout now for disk allocation
+ alarm($oldtimeout || 0);
+ $oldtimeout = undef;
&$print_devmap();
print $fifofh "done\n";
- my $tmp = $oldtimeout || 0;
- $oldtimeout = undef;
- alarm($tmp);
close($fifofh);
$fifofh = undef;
}
die $err;
}
- my $new_conf = $restore_merge_config->($conffile, $new_conf_raw, $opts->{override_conf});
+ my $new_conf = restore_merge_config($conffile, $new_conf_raw, $opts->{override_conf});
+ check_restore_permissions($rpcenv, $user, $new_conf);
PVE::QemuConfig->write_config($vmid, $new_conf);
eval { rescan($vmid, 1); };
}
sub qemu_img_convert {
- my ($src_volid, $dst_volid, $size, $snapname, $is_zero_initialized) = @_;
+ my ($src_volid, $dst_volid, $size, $snapname, $is_zero_initialized, $bwlimit) = @_;
my $storecfg = PVE::Storage::config();
my ($src_storeid, $src_volname) = PVE::Storage::parse_volume_id($src_volid, 1);
if $snapname && $src_format && $src_format eq "qcow2";
push @$cmd, '-t', 'none' if $dst_scfg->{type} eq 'zfspool';
push @$cmd, '-T', $cachemode if defined($cachemode);
+ push @$cmd, '-r', "${bwlimit}K" if defined($bwlimit);
if ($src_is_iscsi) {
push @$cmd, '--image-opts';
sub qemu_img_format {
my ($scfg, $volname) = @_;
- if ($scfg->{path} && $volname =~ m/\.($PVE::QemuServer::Drive::QEMU_FORMAT_RE)$/) {
+ # FIXME: this entire function is kind of weird given that `parse_volname`
+ # also already gives us a format?
+ my $is_path_storage = $scfg->{path} || $scfg->{type} eq 'esxi';
+
+ if ($is_path_storage && $volname =~ m/\.($PVE::QemuServer::Drive::QEMU_FORMAT_RE)$/) {
return $1;
} else {
return "raw";
}
}
+# Check for bug #4525: drive-mirror will open the target drive with the same aio setting as the
+# source, but some storages have problems with io_uring, sometimes even leading to crashes.
+my sub clone_disk_check_io_uring {
+ my ($src_drive, $storecfg, $src_storeid, $dst_storeid, $use_drive_mirror) = @_;
+
+ return if !$use_drive_mirror;
+
+ # Don't complain when not changing storage.
+ # Assume if it works for the source, it'll work for the target too.
+ return if $src_storeid eq $dst_storeid;
+
+ my $src_scfg = PVE::Storage::storage_config($storecfg, $src_storeid);
+ my $dst_scfg = PVE::Storage::storage_config($storecfg, $dst_storeid);
+
+ my $cache_direct = drive_uses_cache_direct($src_drive);
+
+ my $src_uses_io_uring;
+ if ($src_drive->{aio}) {
+ $src_uses_io_uring = $src_drive->{aio} eq 'io_uring';
+ } else {
+ $src_uses_io_uring = storage_allows_io_uring_default($src_scfg, $cache_direct);
+ }
+
+ die "target storage is known to cause issues with aio=io_uring (used by current drive)\n"
+ if $src_uses_io_uring && !storage_allows_io_uring_default($dst_scfg, $cache_direct);
+}
+
sub clone_disk {
my ($storecfg, $source, $dest, $full, $newvollist, $jobs, $completion, $qga, $bwlimit) = @_;
my ($newvmid, $dst_drivename, $efisize) = $dest->@{qw(vmid drivename efisize)};
my ($storage, $format) = $dest->@{qw(storage format)};
- my $use_drive_mirror = $full && $running && $src_drivename && !$snapname;
+ my $unused = defined($src_drivename) && $src_drivename =~ /^unused/;
+ my $use_drive_mirror = $full && $running && $src_drivename && !$snapname && !$unused;
if ($src_drivename && $dst_drivename && $src_drivename ne $dst_drivename) {
die "cloning from/to EFI disk requires EFI disk\n"
$newvolid = PVE::Storage::vdisk_clone($storecfg, $drive->{file}, $newvmid, $snapname);
push @$newvollist, $newvolid;
} else {
-
- my ($storeid, $volname) = PVE::Storage::parse_volume_id($drive->{file});
- $storeid = $storage if $storage;
+ my ($src_storeid, $volname) = PVE::Storage::parse_volume_id($drive->{file});
+ my $storeid = $storage || $src_storeid;
my $dst_format = resolve_dst_disk_format($storecfg, $storeid, $volname, $format);
$dst_format = 'raw';
$size = PVE::QemuServer::Drive::TPMSTATE_DISK_SIZE;
} else {
- ($size) = PVE::Storage::volume_size_info($storecfg, $drive->{file}, 10);
+ clone_disk_check_io_uring($drive, $storecfg, $src_storeid, $storeid, $use_drive_mirror);
+
+ $size = PVE::Storage::volume_size_info($storecfg, $drive->{file}, 10);
}
$newvolid = PVE::Storage::vdisk_alloc(
$storecfg, $storeid, $newvmid, $dst_format, $name, ($size/1024)
qemu_drive_mirror($vmid, $src_drivename, $newvolid, $newvmid, $sparseinit, $jobs,
$completion, $qga, $bwlimit);
} else {
- # TODO: handle bwlimits
if ($dst_drivename eq 'efidisk0') {
# the relevant data on the efidisk may be smaller than the source
# e.g. on RBD/ZFS, so we use dd to copy only the amount
push $cmd->@*, "bs=$bs", "osize=$size", "if=$src_path", "of=$dst_path";
run_command($cmd);
} else {
- qemu_img_convert($drive->{file}, $newvolid, $size, $snapname, $sparseinit);
+ qemu_img_convert($drive->{file}, $newvolid, $size, $snapname, $sparseinit, $bwlimit);
}
}
}
no_data_clone:
- my ($size) = eval { PVE::Storage::volume_size_info($storecfg, $newvolid, 10) };
+ my $size = eval { PVE::Storage::volume_size_info($storecfg, $newvolid, 10) };
my $disk = dclone($drive);
delete $disk->{format};
$disk->{file} = $newvolid;
- $disk->{size} = $size if defined($size);
+ $disk->{size} = $size if defined($size) && !$unused;
return $disk;
}
PVE::Storage::activate_volumes($storecfg, [$volid]);
qemu_img_convert($ovmf_vars, $volid, $vars_size_b, undef, 0);
- my ($size) = PVE::Storage::volume_size_info($storecfg, $volid, 3);
+ my $size = PVE::Storage::volume_size_info($storecfg, $volid, 3);
return ($volid, $size/1024);
}
sub nbd_stop {
my ($vmid) = @_;
- mon_cmd($vmid, 'nbd-server-stop');
+ mon_cmd($vmid, 'nbd-server-stop', timeout => 25);
}
sub create_reboot_request {
}
sub vm_is_paused {
- my ($vmid) = @_;
+ my ($vmid, $include_suspended) = @_;
my $qmpstatus = eval {
PVE::QemuConfig::assert_config_exists_on_node($vmid);
mon_cmd($vmid, "query-status");
};
warn "$@\n" if $@;
- return $qmpstatus && $qmpstatus->{status} eq "paused";
+ return $qmpstatus && (
+ $qmpstatus->{status} eq "paused" ||
+ $qmpstatus->{status} eq "prelaunch" ||
+ ($include_suspended && $qmpstatus->{status} eq "suspended")
+ );
}
sub check_volume_storage_type {
}
my $bridge = $net->{bridge};
+ if (!$bridge) {
+ log_warn("Interface '$iface' not attached to any bridge.");
+ next;
+ }
if ($have_sdn) {
- PVE::Network::SDN::Zones::add_bridge_fdb($iface, $mac, $bridge, $net->{firewall});
+ PVE::Network::SDN::Zones::add_bridge_fdb($iface, $mac, $bridge);
} elsif (-d "/sys/class/net/$bridge/bridge") { # avoid fdb management with OVS for now
- PVE::Network::add_bridge_fdb($iface, $mac, $net->{firewall});
+ PVE::Network::add_bridge_fdb($iface, $mac);
}
}
}
my $bridge = $net->{bridge};
if ($have_sdn) {
- PVE::Network::SDN::Zones::del_bridge_fdb($iface, $mac, $bridge, $net->{firewall});
+ PVE::Network::SDN::Zones::del_bridge_fdb($iface, $mac, $bridge);
} elsif (-d "/sys/class/net/$bridge/bridge") { # avoid fdb management with OVS for now
- PVE::Network::del_bridge_fdb($iface, $mac, $net->{firewall});
+ PVE::Network::del_bridge_fdb($iface, $mac);
+ }
+ }
+}
+
+sub create_ifaces_ipams_ips {
+ my ($conf, $vmid) = @_;
+
+ return if !$have_sdn;
+
+ foreach my $opt (keys %$conf) {
+ if ($opt =~ m/^net(\d+)$/) {
+ my $value = $conf->{$opt};
+ my $net = PVE::QemuServer::parse_net($value);
+ eval { PVE::Network::SDN::Vnets::add_next_free_cidr($net->{bridge}, $conf->{name}, $net->{macaddr}, $vmid, undef, 1) };
+ warn $@ if $@;
+ }
+ }
+}
+
+sub delete_ifaces_ipams_ips {
+ my ($conf, $vmid) = @_;
+
+ return if !$have_sdn;
+
+ foreach my $opt (keys %$conf) {
+ if ($opt =~ m/^net(\d+)$/) {
+ my $net = PVE::QemuServer::parse_net($conf->{$opt});
+ eval { PVE::Network::SDN::Vnets::del_ips_from_mac($net->{bridge}, $net->{macaddr}, $conf->{name}) };
+ warn $@ if $@;
}
}
}
projects / qemu-server.git / blobdiff