Re: KASSERT triggered in pcu_load()

[Alexander Nasonov <alnsn%yandex.ru@localhost>]

Alexander Nasonov wrote:
> My aarch64 server had been running fine for a couple of weeks but
> it all suddenly changed yesterday when it crashed about tree times.
> 
> I captured a screenshot of one crash:
> 
> panic: kernel diagnostic assertion "l->l_pcu_cpu[id] == NULL", ...
> cpu7: Begin traceback...
> trace fp ffffffc0a4ee7cc0
> fp ffffffc0a4ee7cc0 vpanic() at ffffffc000408b68 netbsd:vpanic+0x190
> fp ffffffc0a4ee7d40 kern_assert() at ffffffc000529c60 netbsd:kern_assert+0x58
> fp ffffffc0a4ee7dd0 pcu_load() at ffffffc000401a2c netbsd:pcu_load +0x21c
> fp ffffffc0a4ee7e70 trap_el0_sync() at ffffffc0000656d0 netbsd:trap_el0_sync+0x1d0
> fp ffffffc0a4ee7ed0 el0_trap() at ffffffc00006330 netbsd:el0_trap
> 
> It's followed by registers dump (available upon request).
> 
> cpu6 was in sched_pstats, other cpus were idle.

It started panicking again with a newer version of the kernel.

-- 
Alex