Which SSD works good on starfive?

having issues with Samsung 980PRO

Debian GNU/Linux trixie/sid starfive ttyS0

starfive login: [ 36.394683] mipi_0p9: disabling
[ 55.104633] rcu: INFO: rcu_sched detected stalls on CPUs/tasks:
[ 55.110563] rcu: 1-…!: (1 GPs behind) idle=861c/1/0x4000000000000000 softirq=5135/5142 fqs=396
[ 55.119429] rcu: 3-…!: (1 GPs behind) idle=e36c/1/0x4000000000000000 softirq=4773/4774 fqs=396
[ 55.128293] rcu: (detected by 2, t=2102 jiffies, g=2257, q=774 ncpus=4)
[ 55.134988] Task dump for CPU 1:
[ 55.138214] task:systemd state:R running task stack:0 pid:902 ppid:1 flags:0x00000000
[ 55.148124] Call Trace:
[ 55.150568] [] __schedule+0x272/0x716
[ 55.155799] Task dump for CPU 3:
[ 55.159024] task:kworker/u12:0 state:R running task stack:0 pid:36 ppid:2 flags:0x00000000
[ 55.168933] Workqueue: events_unbound commit_work
[ 55.173642] Call Trace:
[ 55.176088] [] __schedule+0x272/0x716
[ 55.181311] rcu: rcu_sched kthread timer wakeup didn’t happen for 1308 jiffies! g2257 f0x0 RCU_GP_WAIT_FQS(5) ->state=0x402
[ 55.192427] rcu: Possible timer handling issue on cpu=0 timer-softirq=777
[ 55.199293] rcu: rcu_sched kthread starved for 1309 jiffies! g2257 f0x0 RCU_GP_WAIT_FQS(5) ->state=0x402 ->cpu=0
[ 55.209454] rcu: Unless rcu_sched kthread gets sufficient CPU time, OOM is now expected behavior.
[ 55.218400] rcu: RCU grace-period kthread stack dump:
[ 55.223445] task:rcu_sched state:I stack:0 pid:15 ppid:2 flags:0x00000000
[ 55.231789] Call Trace:
[ 55.234233] [] __schedule+0x272/0x716
[ 55.239456] [] schedule+0x44/0xca
[ 55.244331] [] schedule_timeout+0x6c/0xd2
[ 55.249902] [] rcu_gp_fqs_loop+0x2b2/0x3bc
[ 55.255562] [] rcu_gp_kthread+0xf2/0x112
[ 55.261045] [] kthread+0xae/0xca
[ 55.265835] [] ret_from_fork+0xe/0x1c
[ 55.271057] rcu: Stack dump where RCU GP kthread last ran:
[ 55.276535] Task dump for CPU 0:
[ 55.279760] task:swapper/0 state:R running task stack:0 pid:0 ppid:0 flags:0x00000000
[ 55.289666] Call Trace:
[ 55.292110] [] __schedule+0x272/0x716
[ 55.297332] [] default_idle_call+0x22/0x34
[ 80.294632] watchdog: BUG: soft lockup - CPU#2 stuck for 37s! [systemd-timesyn:245]
[ 80.302281] Modules linked in: jpu(O) venc(O) vdec(O) dm_mod
[ 80.307952] CPU: 2 PID: 245 Comm: systemd-timesyn Tainted: G O T 6.6.20-starfive #41SF
[ 80.317073] Hardware name: StarFive VisionFive 2 v1.3B (DT)
[ 80.322638] epc : smp_call_function_many_cond+0x186/0x40e
[ 80.328037] ra : smp_call_function_many_cond+0x37a/0x40e
[ 80.333433] epc : ffffffff800a23d6 ra : ffffffff800a25ca sp : ffffffc80047bb70
[ 80.340647] gp : ffffffff81915e98 tp : ffffffd8c53e1780 t0 : 0000000000a92325
[ 80.347860] t1 : 0000000000ff0100 t2 : 000000005c000000 s0 : ffffffc80047bc30
[ 80.355073] s1 : 0000000000000001 a0 : 000000000000ffff a1 : 0000000000000001
[ 80.362285] a2 : 0000000000000040 a3 : 0000000000000001 a4 : ffffffd9f8f82a60
[ 80.369498] a5 : ffffffffffffffff a6 : ffffffff80e284c0 a7 : 00000000ffffffff
[ 80.376710] s2 : 0000000000000000 s3 : ffffffff8194cb70 s4 : 0000000000000000
[ 80.383923] s5 : 0000000000000001 s6 : ffffffda78157000 s7 : 0000000000000001
[ 80.391136] s8 : 0000000000000008 s9 : 0000000000000040 s10: ffffffff80e2ba60
[ 80.398348] s11: ffffffd9f8f924c0 t3 : ffffffff81917228 t4 : 0000000000000000
[ 80.405562] t5 : 0000000001831330 t6 : 000000000c416615
[ 80.410867] status: 0000000200000120 badaddr: 0000000000000000 cause: 8000000000000005

Which power supply are you using ?
Does it output 12V2.5A ?

Given how slow the vf5 is (and it being PCIe 3.0), I just got the cheapest reputable-ish-brand one, so I landed on the $20 Patriot P300 256GB

I haven’t had any issues with it. Just got a second for another risc-v board I’m setting up (p550).