Less usable USB on -current

[john%ziaspace.com@localhost]

Hi,

I've had a Raspberry Pi 4 server that has run well since about a year ago. It 
has a raidframe mirrored pair of 8TB spinning rust hard drives.

About two months ago, one of the drives started to fail. Because of Chia 
mining, finding a replacement hasn't been easy, but I finally found one and 
wanted to simplify the partition scheme, so I created a new raidframe instance 
with a missing disk on the new 8TB drive and I'm trying to migrate the data on 
the old raidframe instance.

However, within a half hour of rsyncing, the system invariably panics (or 
sometimes locks up):

[  3795.398611] panic: Trap: Data Abort (EL1): Translation Fault L0 with read 
access for 8000000100000008: pc ffffc000000ada78: ldr x20, [x19,#8]

[  3795.408609] cpu2: Begin traceback...
[  3795.418608] trace fp ffffc000ee74f600
[  3795.418608] fp ffffc000ee74f630 vpanic() at ffffc0000052eecc 
netbsd:vpanic+0x14c
[  3795.428608] fp ffffc000ee74f690 panic() at ffffc0000052efc4 
netbsd:panic+0x44
[  3795.438608] fp ffffc000ee74f720 data_abort_handler() at ffffc000000a837c 
netbsd:data_abort_handler+0x1ec
[  3795.448608] tf ffffc000ee74f790 el1_trap() at ffffc000000a9784 
netbsd:el1_trap
[  3795.448608] ---- trapframe 0xffffc000ee74f790 (304 bytes) ----
[  3795.458608]     pc=ffffc000000ada78,   spsr=0000000060000005
[  3795.458608]    esr=0000000096000004,    far=8000000100000008
[  3795.468608]     x0=ffff00000bb05ff8,     x1=ffff0001fd7d4080
[  3795.478609]     x2=0000000000000000,     x3=0000000000000000
[  3795.478609]     x4=0000000000000010,     x5=ffffc00000926000
[  3795.488608]     x6=ffffc00000926000,     x7=00000000ffffffff
[  3795.488608]     x8=0000000000000f00,     x9=0000000000088a68
[  3795.498608]    x10=0000000000000004,    x11=9e3779b185ebca87
[  3795.498608]    x12=00000000000016c0,    x13=000003ffffff3330
[  3795.508608]    x14=0000000071a50339,    x15=ffffc00104edbd38
[  3795.518608]    x16=ffffc0000009b87c,    x17=0000f573a6c9b7d4
[  3795.518608]    x18=0000000079622d32,    x19=8000000100000000
[  3795.528608]    x20=0000000000000000,    x21=ffff00000bb06000
[  3795.528608]    x22=ffff00000bb05ff8,    x23=0000000000000005
[  3795.538608]    x24=ffffc00000f5b580,    x25=0020000000000000
[  3795.538608]    x26=ffffc0000094a150,    x27=7ffffffffffff000
[  3795.548608]    x28=ffff0001aa475400, fp=x29=ffffc000ee74fac0
[  3795.558608] lr=x30=ffffc000000ada78,     sp=ffffc000ee74fac0
[  3795.558608] ------------------------------------------------
[  3795.568608] fp ffffc000ee74fac0 pmap_page_protect() at ffffc000000ada78 
netbsd:pmap_page_protect+0x78
[  3795.578609] fp ffffc000ee74fb30 genfs_do_putpages() at ffffc000005a5780 
netbsd:genfs_do_putpages+0xc20
[  3795.588608] fp ffffc000ee74fd50 VOP_PUTPAGES() at ffffc000005a1a00 
netbsd:VOP_PUTPAGES+0x3c
[  3795.598609] fp ffffc000ee74fda0 ffs_full_fsync() at ffffc000004713d8 
netbsd:ffs_full_fsync+0x158
[  3795.598609] fp ffffc000ee74fdf0 VOP_FSYNC() at ffffc000005a0ac0 
netbsd:VOP_FSYNC+0x40
[  3795.608609] fp ffffc000ee74fe40 sched_sync() at ffffc0000058897c 
netbsd:sched_sync+0x1b8
[  3795.618608] cpu2: End traceback...

[  3795.628609] dumping to dev 168,3 offset 305183
[  3795.628609] dump i/o error


[  3800.638730] rebooting...

or:

https://www.klos.com/~john/pi_panic.jpg


Another Pi 4 I have also seems to have stability issues when rooted on a USB 
disk, but I've had to power cycle it to get it to respond, so I haven't gotten 
any messages from it.

Is anyone else heavily using USB on their Pi?

John