Re: trouble with netbsd-8 on RPI3?

[Martin Husemann <martin%duskware.de@localhost>]

On Fri, May 15, 2020 at 02:31:25PM -0400, Greg Troxel wrote:
> 0x6dd97d00 in ?? () from /usr/lib/libpthread.so.1
> (gdb) bt
> #0  0x6dd97d00 in ?? () from /usr/lib/libpthread.so.1
> #1  0x6dd97d64 in ?? () from /usr/lib/libpthread.so.1
> #2  0x6dd980c8 in ?? () from /usr/lib/libpthread.so.1
> #3  0x6dd99b38 in pthread_cond_timedwait () from /usr/lib/libpthread.so.1
> #4  0x6de02448 in ?? () from /usr/lib/libisc.so.8
> #5  0x6dd9bb0c in ?? () from /usr/lib/libpthread.so.1
> #6  0x6d752980 in __mknod50 () from /usr/lib/libc.so.12
> Backtrace stopped: previous frame identical to this frame (corrupt stack?)
> (gdb) x/i 0x6dd97d00
> => 0x6dd97d00:  wfe

I have not found a table listing which processor types support "wait
for event", but our default for armv7 seems not to be among them (while
the default for aarch64 is):

> tools/bin/armv7eb--netbsdelf-eabihf-as test.s
test.s: Assembler messages:
test.s:3: Error: selected processor does not support `wfe' in ARM mode
> tools/bin/aarch64--netbsd-as test.s
>

So I would expect this to crash on a/some real armv7 machine(s) too - maybe
the old kernel just eroneously allowed it?

Can you verify it is present in the official builds of the armv7 libpthread
as well, and find out more details where it is in your build?

Martin