NEMU启动虚拟机时遇到CPU卡顿事件

[PDSHan]

我在用NEMU启动虚拟机时，遇到guest os启动过程中信息打印不全和CPU卡顿事件，具体如下：

1、信息打印不全
run Guest Linux!

lkvm run -k ./apps/Image -m 100 -c 1 --name guest-36

Info: (riscv/kvm.c) kvm__arch_load_kernel_image:178: Placing fdt at 0x81c00000 - 0x863fffff
Info: (virtio/mmio.c) virtio_mmio_init:209: virtio-mmio.devices=0x200@0x10000000:5
Info: (virtio/mmio.c) virtio_mmio_init:209: virtio-mmio.devices=0x200@0x10000200:6
Info: (virtio/mmio.c) virtio_mmio_init:209: virtio-mmio.devices=0x200@0x10000400:7
Info: (virtio/mmio.c) virtio_mmio_init:209: virtio-mmio.devices=0x200@0x10000600:8
[ 2158.176326] loop: module loaded
没有打印出OF：reserved mem等其他信息

2、CPU卡顿
[ 2169.650693] NET: Registered PF_PACKET protocol family
[ 2169.910268] 9pnet: Installing 9P2000 support
[ 2170.670207] Key type dns_resolver registered
[ 2213.650294] debug_vm_pgtable: [debug_vm_pgtable ]: Validating architecture page table helpers
[ 2214.713116] Legacy PMU implementation is available
[ 2691.646370] watchdog: BUG: soft lockup - CPU#0 stuck for 2503s! [swapper/0:0]
[ 2691.799172] Modules linked in:
[ 2691.872538] CPU: 0 PID: 0 Comm: swapper/0 Not tainted 6.6.21-00030-g8b68c1917beb-dirty #117
[ 2691.955291] Hardware name: linux,dummy-virt (DT)
[ 2692.005375] epc : default_idle_call+0x26/0x34
[ 2692.064746] ra : default_idle_call+0x22/0x34
[ 2692.116287] epc : ffffffff808b0694 ra : ffffffff808b0690 sp : ffffffff81403dd0
[ 2692.242869] gp : ffffffff81501200 tp : ffffffff8140dac0 t0 : ffffffff80e0c07c
[ 2692.357977] t1 : 0000000000000005 t2 : 000000000000001f s0 : ffffffff81403de0
[ 2692.469498] s1 : 0000000000000000 a0 : 0000000000000004 a1 : 0000000000000000
[ 2692.588185] a2 : ffffaf80847c2000 a3 : 0000000000000034 a4 : ffffffff80c21290
[ 2692.709087] a5 : 0000000000000000 a6 : 4000000000000000 a7 : 0000000000000000
[ 2692.799212] s2 : ffffffff815031e8 s3 : 000000000000003f s4 : 0000000000000000
[ 2692.901505] s5 : ffffffff815033a0 s6 : 0000000000000000 s7 : 0000000000000000
[ 2692.973363] s8 : 0000000000000000 s9 : 0000000000000000 s10: 0000000000000000
[ 2693.028429] s11: 0000000000000000 t3 : 0000000000000394 t4 : 00000000e5b906e6
[ 2693.086247] t5 : 0000000500000000 t6 : 00000000e5b906e6
[ 2693.177123] status: 0000000200000120 badaddr: 0000000000000000 cause: 8000000000000005
[ 2693.343940] [] default_idle_call+0x26/0x34
[ 2693.472426] [] do_idle+0x1e4/0x226
[ 2693.563369] [] cpu_startup_entry+0x26/0x28
[ 2693.738762] [] kernel_init+0x0/0x10a
[ 2693.894235] [] arch_post_acpi_subsys_init+0x0/0x18
[ 2694.164809] [] console_on_rootfs+0x0/0x68
[ 2694.503788] rcu: INFO: rcu_sched self-detected stall on CPU
[ 2694.634310] rcu: 0-...!: (1 ticks this GP) idle=0034/1/0x4000000000000002 softirq=7/7 fqs=0
[ 2694.825619] rcu: (t=166688 jiffies g=-1195 q=126 ncpus=1)
[ 2694.942876] rcu: rcu_sched kthread timer wakeup didn't happen for 166687 jiffies! g-1195 f0x0 RCU_GP_WAIT_FQS(5) ->state=0x402
[ 2695.201617] rcu: Possible timer handling issue on cpu=0 timer-softirq=2
[ 2695.359216] rcu: rcu_sched kthread starved for 166688 jiffies! g-1195 f0x0 RCU_GP_WAIT_FQS(5) ->state=0x402 ->cpu=0
[ 2695.503693] rcu: Unless rcu_sched kthread gets sufficient CPU time, OOM is now expected behavior.
[ 2695.566814] rcu: RCU grace-period kthread stack dump:
[ 2695.611333] task:rcu_sched state:I stack:0 pid:15 ppid:2 flags:0x00000000
[ 2695.741114] Call Trace:
[ 2695.791679] [] __schedule+0x29a/0x806
[ 2695.928119] [] schedule+0x46/0xd0
[ 2696.063419] [] schedule_timeout+0x7a/0xee
[ 2696.207482] [] rcu_gp_fqs_loop+0x2de/0x3e2
[ 2696.339084] [] rcu_gp_kthread+0x106/0x126
[ 2696.512730] [] kthread+0xc4/0xe0
[ 2696.643659] [] ret_from_fork+0xe/0x1c
[ 2696.740392] rcu: Stack dump where RCU GP kthread last ran:
[ 2696.834812] CPU: 0 PID: 0 Comm: swapper/0 Tainted: G L 6.6.21-00030-g8b68c1917beb-dirty #117
[ 2697.121297] Hardware name: linux,dummy-virt (DT)
该事件一直无法得到解决，反复出现。但是最后guest os仍打印出了Run /virt/init as init process
[ 4083.726345] [] arch_post_acpi_subsys_init+0x0/0x18
[ 4083.790340] [] console_on_rootfs+0x0/0x68
[ 4084.129351] Run /virt/init as init process
Mounting...
/ # [ 22.415677] rcu: INFO: rcu_sched self-detected stall on CPU
[ 22.621987] rcu: 0-...!: (1 ticks this GP) idle=011c/1/0x4000000000000002 softirq=26/26 fqs=0
[ 22.834965] rcu: (t=84134 jiffies g=-1167 q=173 ncpus=1)
请问是哪里出了问题？

[NewPaulWalker]

请参考我们提供的 issue 模板 进行提问并提供详细的信息，比如你运行的二进制文件，反汇编文件，以及完整的 log 文件等。

[TRANSLATION]
Please refer to the issue template we provided when asking questions and include detailed information, such as the binary file you ran, the disassembled file, and the complete log file, etc.

[PDSHan]

Describe the bug
在NEMU上启动的os中，运行kvmtool启动虚拟机，但是遇到了CPU卡顿事件。

Error log or Screenshots
启动日志：
log.txt

Necessary information on versions
具体的软件版本：
1、NEMU使用master分支，riscv64-xs_defconfig。
2、guest os和host os用的都是linux-6.6.21_xiangshan版本。其中，guest os的config用的是默认config，并且开启CONFIG_DMA_RESTRICTED_POOL=y。host os用的xiangshan_defconfig，并且参照https://bosc.yuque.com/yny0gi/fzefxm/mypmesdadu6l7g7q 更改了config和根文件系统配置。
3、kvmtool用的版本号是e17d182ad3。
4、opensbi用的是202408_xiangshan a83569d platform: Add kunminghu support。
请问还有什么信息需要补充吗？