Open staticfloat opened 2 years ago
Hi, I don't know if it is related, but currently, i'm trying to run debian on NaxRiscv, and got kind of a similar issue with the liteuart, it goes well for 60 secondes of boots, and then, whe iti try to bind liteuart0, things turn wrong, generaly, the timer in linux get called endlessly while the time base do not advance.
Not sure if my issue is releated to a NaxRiscv bug (as it is quite fresh), or a driver bug.
(note, from memory my stack trace is quite different
Yes, I think it's likely different. In my case, I think we may have tracked it down to old liteuart
module versions being used with newer gateware; because the register mapping might change with a new gateware, we believe it could cause this kind of an issue.
When using the
liteuart
driver, we have run into a few kernel hangs that are reproducible. Opening multiplelitex_term
processes to talk to a single device usually triggers it, which is more understandable, however we have also found that the same issue occurs when userA
attempts to communicate to the device, whereas userB
can communicate just fine, which is much more puzzling.The kernel hang is surfaced in
dmesg
as the following message:This is using the
2022.04
release oflitex
, communicating with the board defined in the xtrx_julia repository.