forked from kata-containers/kata-containers
-
Notifications
You must be signed in to change notification settings - Fork 0
/
qemu_arch_base.go
865 lines (704 loc) · 25 KB
/
qemu_arch_base.go
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
//go:build linux
// +build linux
// Copyright (c) 2018 Intel Corporation
//
// SPDX-License-Identifier: Apache-2.0
//
package virtcontainers
import (
"context"
"encoding/hex"
"errors"
"fmt"
"os"
"runtime"
"strings"
govmmQemu "github.com/kata-containers/kata-containers/src/runtime/pkg/govmm/qemu"
"gitlab.com/nvidia/cloud-native/go-nvlib/pkg/nvpci"
"github.com/kata-containers/kata-containers/src/runtime/pkg/device/config"
"github.com/kata-containers/kata-containers/src/runtime/virtcontainers/types"
"github.com/kata-containers/kata-containers/src/runtime/virtcontainers/utils"
)
type qemuArch interface {
// enableNestingChecks nesting checks will be honoured
enableNestingChecks()
// disableNestingChecks nesting checks will be ignored
disableNestingChecks()
// runNested indicates if the hypervisor runs in a nested environment
runNested() bool
// enableVhostNet vhost will be enabled
enableVhostNet()
// disableVhostNet vhost will be disabled
disableVhostNet()
// machine returns the machine type
machine() govmmQemu.Machine
// qemuPath returns the path to the QEMU binary
qemuPath() string
// kernelParameters returns the kernel parameters
// if debug is true then kernel debug parameters are included
kernelParameters(debug bool) []Param
//capabilities returns the capabilities supported by QEMU
capabilities() types.Capabilities
// bridges sets the number bridges for the machine type
bridges(number uint32)
// cpuTopology returns the CPU topology for the given amount of vcpus
cpuTopology(vcpus, maxvcpus uint32) govmmQemu.SMP
// cpuModel returns the CPU model for the machine type
cpuModel() string
// memoryTopology returns the memory topology using the given amount of memoryMb and hostMemoryMb
memoryTopology(memoryMb, hostMemoryMb uint64, slots uint8) govmmQemu.Memory
// appendConsole appends a console to devices
appendConsole(ctx context.Context, devices []govmmQemu.Device, path string) ([]govmmQemu.Device, error)
// appendImage appends an image to devices
appendImage(ctx context.Context, devices []govmmQemu.Device, path string) ([]govmmQemu.Device, error)
// appendBlockImage appends an image as block device
appendBlockImage(ctx context.Context, devices []govmmQemu.Device, path string) ([]govmmQemu.Device, error)
// appendNvdimmImage appends an image as nvdimm device
appendNvdimmImage(devices []govmmQemu.Device, path string) ([]govmmQemu.Device, error)
// appendSCSIController appens a SCSI controller to devices
appendSCSIController(context context.Context, devices []govmmQemu.Device, enableIOThreads bool) ([]govmmQemu.Device, *govmmQemu.IOThread, error)
// appendBridges appends bridges to devices
appendBridges(devices []govmmQemu.Device) []govmmQemu.Device
// append9PVolume appends a 9P volume to devices
append9PVolume(ctx context.Context, devices []govmmQemu.Device, volume types.Volume) ([]govmmQemu.Device, error)
// appendSocket appends a socket to devices
appendSocket(devices []govmmQemu.Device, socket types.Socket) []govmmQemu.Device
// appendVSock appends a vsock PCI to devices
appendVSock(ctx context.Context, devices []govmmQemu.Device, vsock types.VSock) ([]govmmQemu.Device, error)
// appendNetwork appends a endpoint device to devices
appendNetwork(ctx context.Context, devices []govmmQemu.Device, endpoint Endpoint) ([]govmmQemu.Device, error)
// appendBlockDevice appends a block drive to devices
appendBlockDevice(ctx context.Context, devices []govmmQemu.Device, drive config.BlockDrive) ([]govmmQemu.Device, error)
// appendVhostUserDevice appends a vhost user device to devices
appendVhostUserDevice(ctx context.Context, devices []govmmQemu.Device, drive config.VhostUserDeviceAttrs) ([]govmmQemu.Device, error)
// appendVFIODevice appends a VFIO device to devices
appendVFIODevice(devices []govmmQemu.Device, vfioDevice config.VFIODev) []govmmQemu.Device
// appendRNGDevice appends a RNG device to devices
appendRNGDevice(ctx context.Context, devices []govmmQemu.Device, rngDevice config.RNGDev) ([]govmmQemu.Device, error)
// addDeviceToBridge adds devices to the bus
addDeviceToBridge(ctx context.Context, ID string, t types.Type) (string, types.Bridge, error)
// removeDeviceFromBridge removes devices to the bus
removeDeviceFromBridge(ID string) error
// getBridges grants access to Bridges
getBridges() []types.Bridge
// setBridges grants access to Bridges
setBridges(bridges []types.Bridge)
// addBridge adds a new Bridge to the list of Bridges
addBridge(types.Bridge)
// getPFlash() get pflash from configuration
getPFlash() ([]string, error)
// setPFlash() grants access to pflash
setPFlash([]string)
// handleImagePath handles the Hypervisor Config image path
handleImagePath(config HypervisorConfig)
// supportGuestMemoryHotplug returns if the guest supports memory hotplug
supportGuestMemoryHotplug() bool
// setIgnoreSharedMemoryMigrationCaps set bypass-shared-memory capability for migration
setIgnoreSharedMemoryMigrationCaps(context.Context, *govmmQemu.QMP) error
// appendPCIeRootPortDevice appends a pcie-root-port device to pcie.0 bus
appendPCIeRootPortDevice(devices []govmmQemu.Device, number uint32, memSize32bit uint64, memSize64bit uint64) []govmmQemu.Device
// append vIOMMU device
appendIOMMU(devices []govmmQemu.Device) ([]govmmQemu.Device, error)
// append pvpanic device
appendPVPanicDevice(devices []govmmQemu.Device) ([]govmmQemu.Device, error)
// append protection device.
// This implementation is architecture specific, some archs may need
// a firmware, returns a string containing the path to the firmware that should
// be used with the -bios option, ommit -bios option if the path is empty.
appendProtectionDevice(devices []govmmQemu.Device, firmware, firmwareVolume string) ([]govmmQemu.Device, string, error)
// scans the PCIe space and returns the biggest BAR sizes for 32-bit
// and 64-bit addressable memory
getBARsMaxAddressableMemory() (uint64, uint64)
}
type qemuArchBase struct {
qemuExePath string
qemuMachine govmmQemu.Machine
PFlash []string
kernelParamsNonDebug []Param
kernelParamsDebug []Param
kernelParams []Param
Bridges []types.Bridge
memoryOffset uint64
networkIndex int
// Exclude from lint checking for it is ultimately only used in architecture-specific code
protection guestProtection //nolint:structcheck
nestedRun bool
vhost bool
disableNvdimm bool
dax bool
legacySerial bool
}
const (
defaultCores uint32 = 1
defaultThreads uint32 = 1
defaultCPUModel = "host"
defaultBridgeBus = "pcie.0"
defaultPCBridgeBus = "pci.0"
maxDevIDSize = 31
pcieRootPortPrefix = "rp"
)
// This is the PCI start address assigned to the first bridge that
// is added on the qemu command line. In case of x86_64, the first two PCI
// addresses (0 and 1) are used by the platform while in case of ARM, address
// 0 is reserved.
const bridgePCIStartAddr = 2
const (
// QemuQ35 is the QEMU Q35 machine type for amd64
QemuQ35 = "q35"
// QemuMicrovm is the QEMU microvm machine type for amd64
QemuMicrovm = "microvm"
// QemuVirt is the QEMU virt machine type for aarch64 or amd64
QemuVirt = "virt"
// QemuPseries is a QEMU virt machine type for ppc64le
QemuPseries = "pseries"
// QemuCCWVirtio is a QEMU virt machine type for for s390x
QemuCCWVirtio = "s390-ccw-virtio"
qmpCapMigrationIgnoreShared = "x-ignore-shared"
qemuNvdimmOption = "nvdimm=on"
)
// kernelParamsNonDebug is a list of the default kernel
// parameters that will be used in standard (non-debug) mode.
var kernelParamsNonDebug = []Param{
{"quiet", ""},
}
// kernelParamsSystemdNonDebug is a list of the default systemd related
// kernel parameters that will be used in standard (non-debug) mode.
var kernelParamsSystemdNonDebug = []Param{
{"systemd.show_status", "false"},
}
// kernelParamsDebug is a list of the default kernel
// parameters that will be used in debug mode (as much boot output as
// possible).
var kernelParamsDebug = []Param{
{"debug", ""},
}
// kernelParamsSystemdDebug is a list of the default systemd related kernel
// parameters that will be used in debug mode (as much boot output as
// possible).
var kernelParamsSystemdDebug = []Param{
{"systemd.show_status", "true"},
{"systemd.log_level", "debug"},
}
func (q *qemuArchBase) enableNestingChecks() {
q.nestedRun = true
}
func (q *qemuArchBase) disableNestingChecks() {
q.nestedRun = false
}
func (q *qemuArchBase) runNested() bool {
return q.nestedRun
}
func (q *qemuArchBase) enableVhostNet() {
q.vhost = true
}
func (q *qemuArchBase) disableVhostNet() {
q.vhost = false
}
func (q *qemuArchBase) machine() govmmQemu.Machine {
return q.qemuMachine
}
func (q *qemuArchBase) qemuPath() string {
return q.qemuExePath
}
func (q *qemuArchBase) kernelParameters(debug bool) []Param {
params := q.kernelParams
if debug {
params = append(params, q.kernelParamsDebug...)
} else {
params = append(params, q.kernelParamsNonDebug...)
}
return params
}
func (q *qemuArchBase) capabilities() types.Capabilities {
var caps types.Capabilities
caps.SetBlockDeviceHotplugSupport()
caps.SetMultiQueueSupport()
caps.SetFsSharingSupport()
return caps
}
func (q *qemuArchBase) bridges(number uint32) {
for i := uint32(0); i < number; i++ {
q.Bridges = append(q.Bridges, types.NewBridge(types.PCI, fmt.Sprintf("%s-bridge-%d", types.PCI, i), make(map[uint32]string), 0))
}
}
func (q *qemuArchBase) cpuTopology(vcpus, maxvcpus uint32) govmmQemu.SMP {
smp := govmmQemu.SMP{
CPUs: vcpus,
Sockets: maxvcpus,
Cores: defaultCores,
Threads: defaultThreads,
MaxCPUs: maxvcpus,
}
return smp
}
func (q *qemuArchBase) cpuModel() string {
return defaultCPUModel
}
func (q *qemuArchBase) memoryTopology(memoryMb, hostMemoryMb uint64, slots uint8) govmmQemu.Memory {
memMax := fmt.Sprintf("%dM", hostMemoryMb)
mem := fmt.Sprintf("%dM", memoryMb)
memory := govmmQemu.Memory{
Size: mem,
Slots: slots,
MaxMem: memMax,
}
return memory
}
func (q *qemuArchBase) appendConsole(_ context.Context, devices []govmmQemu.Device, path string) ([]govmmQemu.Device, error) {
var serial, console govmmQemu.Device
var consoleKernelParams []Param
if q.legacySerial {
serial = govmmQemu.LegacySerialDevice{
Chardev: "charconsole0",
}
console = govmmQemu.CharDevice{
Driver: govmmQemu.LegacySerial,
Backend: govmmQemu.Socket,
DeviceID: "console0",
ID: "charconsole0",
Path: path,
}
consoleKernelParams = []Param{
{"console", "ttyS0"},
}
} else {
serial = govmmQemu.SerialDevice{
Driver: govmmQemu.VirtioSerial,
ID: "serial0",
DisableModern: q.nestedRun,
MaxPorts: uint(2),
}
console = govmmQemu.CharDevice{
Driver: govmmQemu.Console,
Backend: govmmQemu.Socket,
DeviceID: "console0",
ID: "charconsole0",
Path: path,
}
consoleKernelParams = []Param{
{"console", "hvc0"},
{"console", "hvc1"},
}
}
devices = append(devices, serial)
devices = append(devices, console)
q.kernelParams = append(q.kernelParams, consoleKernelParams...)
return devices, nil
}
func genericImage(path string) (config.BlockDrive, error) {
if _, err := os.Stat(path); os.IsNotExist(err) {
return config.BlockDrive{}, err
}
randBytes, err := utils.GenerateRandomBytes(8)
if err != nil {
return config.BlockDrive{}, err
}
id := utils.MakeNameID("image", hex.EncodeToString(randBytes), maxDevIDSize)
drive := config.BlockDrive{
File: path,
Format: "raw",
ID: id,
ShareRW: true,
ReadOnly: true,
}
return drive, nil
}
func (q *qemuArchBase) appendNvdimmImage(devices []govmmQemu.Device, path string) ([]govmmQemu.Device, error) {
imageFile, err := os.Open(path)
if err != nil {
return nil, err
}
defer imageFile.Close()
imageStat, err := imageFile.Stat()
if err != nil {
return nil, err
}
object := govmmQemu.Object{
Driver: govmmQemu.NVDIMM,
Type: govmmQemu.MemoryBackendFile,
DeviceID: "nv0",
ID: "mem0",
MemPath: path,
Size: (uint64)(imageStat.Size()),
ReadOnly: true,
}
devices = append(devices, object)
return devices, nil
}
func (q *qemuArchBase) appendImage(ctx context.Context, devices []govmmQemu.Device, path string) ([]govmmQemu.Device, error) {
return q.appendBlockImage(ctx, devices, path)
}
func (q *qemuArchBase) appendBlockImage(ctx context.Context, devices []govmmQemu.Device, path string) ([]govmmQemu.Device, error) {
drive, err := genericImage(path)
if err != nil {
return nil, err
}
devices, err = q.appendBlockDevice(ctx, devices, drive)
if err != nil {
return nil, err
}
return devices, nil
}
func genericSCSIController(enableIOThreads, nestedRun bool) (govmmQemu.SCSIController, *govmmQemu.IOThread) {
scsiController := govmmQemu.SCSIController{
ID: scsiControllerID,
DisableModern: nestedRun,
}
var t *govmmQemu.IOThread
if enableIOThreads {
randBytes, _ := utils.GenerateRandomBytes(8)
t = &govmmQemu.IOThread{
ID: fmt.Sprintf("%s-%s", "iothread", hex.EncodeToString(randBytes)),
}
scsiController.IOThread = t.ID
}
return scsiController, t
}
func (q *qemuArchBase) appendSCSIController(_ context.Context, devices []govmmQemu.Device, enableIOThreads bool) ([]govmmQemu.Device, *govmmQemu.IOThread, error) {
d, t := genericSCSIController(enableIOThreads, q.nestedRun)
devices = append(devices, d)
return devices, t, nil
}
// appendBridges appends to devices the given bridges
func (q *qemuArchBase) appendBridges(devices []govmmQemu.Device) []govmmQemu.Device {
return genericAppendBridges(devices, q.Bridges, q.qemuMachine.Type)
}
func generic9PVolume(volume types.Volume, nestedRun bool) govmmQemu.FSDevice {
devID := fmt.Sprintf("extra-9p-%s", volume.MountTag)
if len(devID) > maxDevIDSize {
devID = devID[:maxDevIDSize]
}
return govmmQemu.FSDevice{
Driver: govmmQemu.Virtio9P,
FSDriver: govmmQemu.Local,
ID: devID,
Path: volume.HostPath,
MountTag: volume.MountTag,
SecurityModel: govmmQemu.None,
DisableModern: nestedRun,
Multidev: govmmQemu.Remap,
}
}
func genericAppend9PVolume(devices []govmmQemu.Device, volume types.Volume, nestedRun bool) (govmmQemu.FSDevice, error) {
d := generic9PVolume(volume, nestedRun)
return d, nil
}
func (q *qemuArchBase) append9PVolume(_ context.Context, devices []govmmQemu.Device, volume types.Volume) ([]govmmQemu.Device, error) {
if volume.MountTag == "" || volume.HostPath == "" {
return devices, nil
}
d, err := genericAppend9PVolume(devices, volume, q.nestedRun)
if err != nil {
return nil, err
}
devices = append(devices, d)
return devices, nil
}
func (q *qemuArchBase) appendSocket(devices []govmmQemu.Device, socket types.Socket) []govmmQemu.Device {
devID := socket.ID
if len(devID) > maxDevIDSize {
devID = devID[:maxDevIDSize]
}
devices = append(devices,
govmmQemu.CharDevice{
Driver: govmmQemu.VirtioSerialPort,
Backend: govmmQemu.Socket,
DeviceID: socket.DeviceID,
ID: devID,
Path: socket.HostPath,
Name: socket.Name,
},
)
return devices
}
func (q *qemuArchBase) appendVSock(_ context.Context, devices []govmmQemu.Device, vsock types.VSock) ([]govmmQemu.Device, error) {
devices = append(devices,
govmmQemu.VSOCKDevice{
ID: fmt.Sprintf("vsock-%d", vsock.ContextID),
ContextID: vsock.ContextID,
VHostFD: vsock.VhostFd,
DisableModern: q.nestedRun,
},
)
return devices, nil
}
func networkModelToQemuType(model NetInterworkingModel) govmmQemu.NetDeviceType {
switch model {
case NetXConnectMacVtapModel:
return govmmQemu.MACVTAP
default:
//TAP should work for most other cases
return govmmQemu.TAP
}
}
func genericNetwork(endpoint Endpoint, vhost, nestedRun bool, index int) (govmmQemu.NetDevice, error) {
var d govmmQemu.NetDevice
switch ep := endpoint.(type) {
case *VethEndpoint, *MacvlanEndpoint, *IPVlanEndpoint:
netPair := ep.NetworkPair()
d = govmmQemu.NetDevice{
Type: networkModelToQemuType(netPair.NetInterworkingModel),
Driver: govmmQemu.VirtioNet,
ID: fmt.Sprintf("network-%d", index),
IFName: netPair.TAPIface.Name,
MACAddress: netPair.TAPIface.HardAddr,
DownScript: "no",
Script: "no",
VHost: vhost,
DisableModern: nestedRun,
FDs: netPair.VMFds,
VhostFDs: netPair.VhostFds,
}
case *MacvtapEndpoint:
d = govmmQemu.NetDevice{
Type: govmmQemu.MACVTAP,
Driver: govmmQemu.VirtioNet,
ID: fmt.Sprintf("network-%d", index),
IFName: ep.Name(),
MACAddress: ep.HardwareAddr(),
DownScript: "no",
Script: "no",
VHost: vhost,
DisableModern: nestedRun,
FDs: ep.VMFds,
VhostFDs: ep.VhostFds,
}
case *TuntapEndpoint:
netPair := ep.NetworkPair()
d = govmmQemu.NetDevice{
Type: govmmQemu.NetDeviceType("tap"),
Driver: govmmQemu.VirtioNet,
ID: fmt.Sprintf("network-%d", index),
IFName: netPair.TAPIface.Name,
MACAddress: netPair.TAPIface.HardAddr,
DownScript: "no",
Script: "no",
VHost: vhost,
DisableModern: nestedRun,
FDs: netPair.VMFds,
VhostFDs: netPair.VhostFds,
}
default:
return govmmQemu.NetDevice{}, fmt.Errorf("Unknown type for endpoint")
}
return d, nil
}
func (q *qemuArchBase) appendNetwork(_ context.Context, devices []govmmQemu.Device, endpoint Endpoint) ([]govmmQemu.Device, error) {
d, err := genericNetwork(endpoint, q.vhost, q.nestedRun, q.networkIndex)
if err != nil {
return devices, fmt.Errorf("Failed to append network %v", err)
}
q.networkIndex++
devices = append(devices, d)
return devices, nil
}
func genericBlockDevice(drive config.BlockDrive, nestedRun bool) (govmmQemu.BlockDevice, error) {
if drive.File == "" || drive.ID == "" || drive.Format == "" {
return govmmQemu.BlockDevice{}, fmt.Errorf("Empty File, ID or Format for drive %v", drive)
}
if len(drive.ID) > maxDevIDSize {
drive.ID = drive.ID[:maxDevIDSize]
}
return govmmQemu.BlockDevice{
Driver: govmmQemu.VirtioBlock,
ID: drive.ID,
File: drive.File,
AIO: govmmQemu.Threads,
Format: govmmQemu.BlockDeviceFormat(drive.Format),
Interface: "none",
DisableModern: nestedRun,
ShareRW: drive.ShareRW,
ReadOnly: drive.ReadOnly,
}, nil
}
func (q *qemuArchBase) appendBlockDevice(_ context.Context, devices []govmmQemu.Device, drive config.BlockDrive) ([]govmmQemu.Device, error) {
d, err := genericBlockDevice(drive, q.nestedRun)
if err != nil {
return devices, fmt.Errorf("Failed to append block device %v", err)
}
devices = append(devices, d)
return devices, nil
}
func (q *qemuArchBase) appendVhostUserDevice(ctx context.Context, devices []govmmQemu.Device, attr config.VhostUserDeviceAttrs) ([]govmmQemu.Device, error) {
qemuVhostUserDevice := govmmQemu.VhostUserDevice{}
switch attr.Type {
case config.VhostUserNet:
qemuVhostUserDevice.TypeDevID = utils.MakeNameID("net", attr.DevID, maxDevIDSize)
qemuVhostUserDevice.Address = attr.MacAddress
qemuVhostUserDevice.VhostUserType = govmmQemu.VhostUserNet
case config.VhostUserSCSI:
qemuVhostUserDevice.TypeDevID = utils.MakeNameID("scsi", attr.DevID, maxDevIDSize)
qemuVhostUserDevice.VhostUserType = govmmQemu.VhostUserSCSI
case config.VhostUserBlk:
qemuVhostUserDevice.VhostUserType = govmmQemu.VhostUserBlk
case config.VhostUserFS:
qemuVhostUserDevice.TypeDevID = utils.MakeNameID("fs", attr.DevID, maxDevIDSize)
qemuVhostUserDevice.Tag = attr.Tag
qemuVhostUserDevice.CacheSize = attr.CacheSize
qemuVhostUserDevice.VhostUserType = govmmQemu.VhostUserFS
}
qemuVhostUserDevice.SocketPath = attr.SocketPath
qemuVhostUserDevice.CharDevID = utils.MakeNameID("char", attr.DevID, maxDevIDSize)
devices = append(devices, qemuVhostUserDevice)
return devices, nil
}
func (q *qemuArchBase) appendVFIODevice(devices []govmmQemu.Device, vfioDev config.VFIODev) []govmmQemu.Device {
if vfioDev.BDF == "" {
return devices
}
devices = append(devices,
govmmQemu.VFIODevice{
BDF: vfioDev.BDF,
VendorID: vfioDev.VendorID,
DeviceID: vfioDev.DeviceID,
Bus: vfioDev.Bus,
},
)
return devices
}
func (q *qemuArchBase) appendRNGDevice(_ context.Context, devices []govmmQemu.Device, rngDev config.RNGDev) ([]govmmQemu.Device, error) {
devices = append(devices,
govmmQemu.RngDevice{
ID: rngDev.ID,
Filename: rngDev.Filename,
},
)
return devices, nil
}
func (q *qemuArchBase) handleImagePath(config HypervisorConfig) {
if config.ImagePath != "" {
kernelRootParams := commonVirtioblkKernelRootParams
if !q.disableNvdimm {
q.qemuMachine.Options = strings.Join([]string{
q.qemuMachine.Options, qemuNvdimmOption,
}, ",")
if q.dax {
kernelRootParams = commonNvdimmKernelRootParams
} else {
kernelRootParams = commonNvdimmNoDAXKernelRootParams
}
}
q.kernelParams = append(q.kernelParams, kernelRootParams...)
q.kernelParamsNonDebug = append(q.kernelParamsNonDebug, kernelParamsSystemdNonDebug...)
q.kernelParamsDebug = append(q.kernelParamsDebug, kernelParamsSystemdDebug...)
}
}
func (q *qemuArchBase) supportGuestMemoryHotplug() bool {
return q.protection == noneProtection
}
func (q *qemuArchBase) setIgnoreSharedMemoryMigrationCaps(ctx context.Context, qmp *govmmQemu.QMP) error {
err := qmp.ExecSetMigrationCaps(ctx, []map[string]interface{}{
{
"capability": qmpCapMigrationIgnoreShared,
"state": true,
},
})
return err
}
func (q *qemuArchBase) addDeviceToBridge(ctx context.Context, ID string, t types.Type) (string, types.Bridge, error) {
addr, b, err := genericAddDeviceToBridge(ctx, q.Bridges, ID, t)
if err != nil {
return "", b, err
}
return fmt.Sprintf("%02x", addr), b, nil
}
func genericAddDeviceToBridge(ctx context.Context, bridges []types.Bridge, ID string, t types.Type) (uint32, types.Bridge, error) {
var err error
var addr uint32
if len(bridges) == 0 {
return 0, types.Bridge{}, errors.New("failed to get available address from bridges")
}
// looking for an empty address in the bridges
for _, b := range bridges {
if t != b.Type {
continue
}
addr, err = b.AddDevice(ctx, ID)
if err == nil {
return addr, b, nil
}
}
return 0, types.Bridge{}, fmt.Errorf("no more bridge slots available")
}
func (q *qemuArchBase) removeDeviceFromBridge(ID string) error {
var err error
for _, b := range q.Bridges {
err = b.RemoveDevice(ID)
if err == nil {
// device was removed correctly
return nil
}
}
return err
}
func (q *qemuArchBase) getBridges() []types.Bridge {
return q.Bridges
}
func (q *qemuArchBase) setBridges(bridges []types.Bridge) {
q.Bridges = bridges
}
func (q *qemuArchBase) addBridge(b types.Bridge) {
q.Bridges = append(q.Bridges, b)
}
// appendPCIeRootPortDevice appends to devices the given pcie-root-port
func (q *qemuArchBase) appendPCIeRootPortDevice(devices []govmmQemu.Device, number uint32, memSize32bit uint64, memSize64bit uint64) []govmmQemu.Device {
return genericAppendPCIeRootPort(devices, number, q.qemuMachine.Type, memSize32bit, memSize64bit)
}
func (q *qemuArchBase) getBARsMaxAddressableMemory() (uint64, uint64) {
pci := nvpci.New()
devs, _ := pci.GetAllDevices()
// Since we do not know which devices are going to be hotplugged,
// we're going to use the GPU with the biggest BARs to initialize the
// root port, this should work for all other devices as well.
// defaults are 2MB for both, if no suitable devices found
max32bit := uint64(2 * 1024 * 1024)
max64bit := uint64(2 * 1024 * 1024)
for _, dev := range devs {
if !dev.IsGPU() {
continue
}
memSize32bit, memSize64bit := dev.Resources.GetTotalAddressableMemory(true)
if max32bit < memSize32bit {
max32bit = memSize32bit
}
if max64bit < memSize64bit {
max64bit = memSize64bit
}
}
// The actual 32bit is most of the time a power of 2 but we need some
// buffer so double that to leave space for other IO functions.
// The 64bit size is not a power of 2 and hence is already rounded up
// to the higher value.
return max32bit * 2, max64bit
}
// appendIOMMU appends a virtual IOMMU device
func (q *qemuArchBase) appendIOMMU(devices []govmmQemu.Device) ([]govmmQemu.Device, error) {
switch q.qemuMachine.Type {
case QemuQ35:
iommu := govmmQemu.IommuDev{
Intremap: true,
DeviceIotlb: true,
CachingMode: true,
}
devices = append(devices, iommu)
return devices, nil
default:
return devices, fmt.Errorf("Machine Type %s does not support vIOMMU", q.qemuMachine.Type)
}
}
// appendPVPanicDevice appends a pvpanic device
func (q *qemuArchBase) appendPVPanicDevice(devices []govmmQemu.Device) ([]govmmQemu.Device, error) {
devices = append(devices, govmmQemu.PVPanicDevice{NoShutdown: true})
return devices, nil
}
func (q *qemuArchBase) getPFlash() ([]string, error) {
return q.PFlash, nil
}
func (q *qemuArchBase) setPFlash(p []string) {
q.PFlash = p
}
// append protection device
func (q *qemuArchBase) appendProtectionDevice(devices []govmmQemu.Device, firmware, firmwareVolume string) ([]govmmQemu.Device, string, error) {
hvLogger.WithField("arch", runtime.GOARCH).Warnf("Confidential Computing has not been implemented for this architecture")
return devices, firmware, nil
}