Re: [RFC v1 00/25] Add Realm support to QEMU-VMM

[Gavin Shan <gshan@redhat.com>]

Hi Mathieu,

On 7/8/26 8:42 AM, Mathieu Poirier wrote:
> This patchset provides minimal functionality to start a Realm VM
> from an Arm RME cabable host using the following command line:
> 
> qemu-system-aarch64 -M confidential-guest-support=rme0 -object rme-guest,id=rme0 \
>          -cpu host -M virt -enable-kvm -M gic-version=3,its=on -nodefaults ..
> 
> This set is a refactoring of Jean-Philippe Brucker's initial work dating from
> a while back.  It is compatible with Steven Price's v14 revision [1] of his work
> adding CCA support to KVM.
> 
> The RME capable host it was tested on is the QEMU SBSA machine.  For convenicence
> a repository is hosted here [2], along with the TF-A [3], RMM [4] and Linux
> kernel [5] for the SBSA machine (all compatible with Steven's v14 patchset).
> 
> Device Assignment is not included.
> 
>  From hereon I will keep this set up to date with new releases from Steven.
> 
> Thanks,
> Mathieu
> 
> [1]. https://lore.kernel.org/kvm/20260513131757.116630-1-steven.price@arm.com/T/#m06dd14216aaf76acab65b0a76fb84653141ea64f
> [2]. https://gitlab.com/Linaro/cca-public/qemu/-/tree/upstream-v1?ref_type=heads
> [3]. https://gitlab.com/Linaro/cca-public/tf-a/trusted-firmware-a/-/tree/cca/v13?ref_type=heads
> [4]. https://gitlab.com/Linaro/cca-public/rmm/-/tree/cca/v14?ref_type=heads
> [5]. https://gitlab.com/Linaro/cca-public/linux/-/tree/cca/v14?ref_type=heads
> 
[...]

Gave this series + (v14) series a try and found the following issues so far. Note
some of the HMP commands like those relevant to migration have been skipped since
we know the migration isn't supported. I list the issues for record and we eventually
need to figure them out one by one.

(1) qemu is terminiated on exeute qmp command 'dump-guest-memory'

(qemu) dump-guest-memory -p "/tmp/mem" 0x40000000 0x10000000
**
ERROR:../system/memory_mapping.c:215:guest_phys_block_add_section: assertion failed: (predecessor->target_end <= target_start)
Bail out! ERROR:../system/memory_mapping.c:215:guest_phys_block_add_section: assertion failed: (predecessor->target_end <= target_start)
Aborted

(2) qemu emulation can't be stopped and resumed

(qemu) stop
(qemu) cont
Failed to set KVM_REG_ARM_TIMER_CNT

(3) qemu is terminiated on qmp command 'system_reset'

(qemu) system_reset
cpus are not resettable, terminating

(4) Incompatible to virtio-iommu-pci. The realm guest doesn't boot up with
     a virtio-iommu-pci device

host$ qemu-system-aarch64 -enable-kvm -object rme-guest,id=rme0                \
-machine virt,gic-version=3,confidential-guest-support=rme0                    \
-m 2G -cpu host -smp maxcpus=4,cpus=4,sockets=1,clusters=1,cores=2,threads=2   \
-object memory-backend-ram,id=mem0,size=2G,share=yes                           \
-numa node,nodeid=0,cpus=0-3,memdev=mem0                                       \
-serial mon:stdio -monitor none -nographic -nodefaults                         \
-bios /mnt/edk2/Build/ArmVirtQemu-AARCH64/RELEASE_GCC5/FV/QEMU_EFI.fd          \
-kernel /mnt/linux/arch/arm64/boot/Image                                       \
-initrd /mnt/buildroot/output/images/rootfs.cpio.xz                            \
-append earlycon=pl011,mmio,0x10009000000                                      \
-device pcie-root-port,bus=pcie.0,chassis=1,id=pcie.1                          \
-device pcie-root-port,bus=pcie.0,chassis=2,id=pcie.2                          \
-device virtio-iommu-pci,bus=pcie.0                                            \
-netdev tap,id=tap1,vhost=on,script=/etc/qemu-ifup,downscript=/etc/qemu-ifdown \
-device virtio-net-pci,bus=pcie.2,netdev=tap1,mac=b8:3f:d2:1d:3e:c0
         :
UEFI firmware (version  built at 00:09:11 on Jun 30 2026)
         :
[    0.000000] Booting Linux on physical CPU 0x0000000000 [0x000f0510]
[    0.000000] Linux version 7.2.0-rc4-gavin-g1590cf032971 (gshan@nvidia-grace-hopper-01.khw.eng.bos2.dc.redhat.com) (gcc (GCC) 14.3.1 20251022 (Red Hat 14.3.1-4), GNU ld version 2.41-65.el10) #46 SMP PREEMPT Mon Jul 20 06:56:20 EDT 2026
[    0.000000] KASLR enabled
[    0.000000] earlycon: pl11 at MMIO 0x0000010009000000 (options '')
[    0.000000] printk: legacy bootconsole [pl11] enabled
[    0.000000] efi: EFI v2.7 by EDK II
[    0.000000] efi: SMBIOS 3.0=0xbff60000 ACPI 2.0=0xbdee2018 INITRD=0xbdd94f18 RNG=0xbdee2a18 MEMRESERVE=0xbdd94f98
         :
[   18.374507] virtio-pci 0000:00:03.0: enabling device (0000 -> 0002)
[   18.594890] Serial: 8250/16550 driver, 4 ports, IRQ sharing enabled
[   18.651840] Serial: AMBA driver
[   19.669576] software IO TLB: Memory encryption is active and system is using DMA bounce buffers
[   35.021315] Freeing initrd memory: 5772K
[   40.984997] rcu: INFO: rcu_preempt detected stalls on CPUs/tasks:
[   41.076752] rcu: 	3-...0: (77 ticks this GP) idle=aed4/1/0x4000000000000000 softirq=83/97 fqs=2365
[   41.210687] rcu: 	(detected by 2, t=5307 jiffies, g=-1079, q=352 ncpus=4)
[   41.311646] Sending NMI from CPU 2 to CPUs 3:
[   51.631564] After 10 seconds, these CPUS still haven't responded to the NMI: 3
[   51.739789] rcu: rcu_preempt kthread timer wakeup didn't happen for 2685 jiffies! g-1079 f0x0 RCU_GP_WAIT_FQS(5) ->state=0x402
[   51.910690] rcu: 	Possible timer handling issue on cpu=2 timer-softirq=2168
[   52.014654] rcu: rcu_preempt kthread starved for 2756 jiffies! g-1079 f0x0 RCU_GP_WAIT_FQS(5) ->state=0x402 ->cpu=2
[   52.169468] rcu: 	Unless rcu_preempt kthread gets sufficient CPU time, OOM is now expected behavior.
[   52.304725] rcu: RCU grace-period kthread stack dump:
[   52.379337] task:rcu_preempt     state:I stack:0     pid:15    tgid:15    ppid:2      task_flags:0x208040 flags:0x00000010
[   52.542813] Call trace:
[   52.578516]  __switch_to+0xfc/0x248 (T)
[   52.636293]  __schedule+0x3a4/0xe68
[   52.686511]  schedule+0x3c/0xe8
[   52.731765]  schedule_timeout+0x80/0x120
[   52.788489]  rcu_gp_fqs_loop+0x110/0x440
[   52.846921]  rcu_gp_kthread+0x10c/0x120
[   52.906472]  kthread+0x138/0x150
[   52.955082]  ret_from_fork+0x10/0x20

(5) As reported previously, 'reboot' on the realm guest causes the termination of qemu process.

root@guest:~# reboot
     :
The system is going down NOW!
Sent SIGTERM to all processes
Sent SIGKILL to all processes
Requesting system reboot
[   61.874435] reboot: Restarting system
qemu-system-aarch64: cpus are not resettable, terminating
qemu-system-aarch64: cpus are not resettable, terminating
qemu-system-aarch64: cpus are not resettable, terminating
qemu-system-aarch64: cpus are not resettable, terminating

Thanks,
Gavin