23 boots, 23 fail in some way
Posted: Thu Jan 15, 2015 11:35 amPlease try using this alternate devicetree from Fred:
http://forums.parallella.org/viewtopic. ... 037#p10037
Regards,
Andrew
Parallella Community
Supercomputing for Everyone
https://parallella.org/forums/
23 boots, 23 fail in some way
Page 1 of 1
Posted: Thu Jan 15, 2015 11:35 amRe: 23 boots, 23 fail in some way
Posted: Thu Jan 15, 2015 4:40 pmRe: 23 boots, 23 fail in some way
Posted: Fri Jan 16, 2015 12:01 amRe: 23 boots, 23 fail in some way
Posted: Fri Jan 16, 2015 5:02 pmThings get stranger. The board has been running Ephiphany tests for over 90 minutes. This is the first time since I unpacked it about 3 days ago that it has been able to do this for more than half a minute.
The only things that are different:
1. I have a serial cable connected. I'm pretty sure it's crashed a couple of times since I did that.
2. I configured my DHCP server to assign a static IP based on the ethernet address, rather than assign dynamically from a pool of IPs. Can't imagine why that would cause a board to crash, but I'm mentioning it for completeness.
3. I've been running 'top' in another window the whole time.
I think #3 may have something to do with it, because almost immediately after ^Cing out of 'top', the test suite became stuck on apps/matmul-16. The board has NOT crashed... I'm still logged in, and can use the shell interactively, but the the matmul-16 test is not completing.
I'm fairly confident that I have a faulty board, so this is really just curiosity, before I return it. I really hope it doesn't just "come good" for no apparent reason.
The only things that are different:
1. I have a serial cable connected. I'm pretty sure it's crashed a couple of times since I did that.
2. I configured my DHCP server to assign a static IP based on the ethernet address, rather than assign dynamically from a pool of IPs. Can't imagine why that would cause a board to crash, but I'm mentioning it for completeness.
3. I've been running 'top' in another window the whole time.
I think #3 may have something to do with it, because almost immediately after ^Cing out of 'top', the test suite became stuck on apps/matmul-16. The board has NOT crashed... I'm still logged in, and can use the shell interactively, but the the matmul-16 test is not completing.
I'm fairly confident that I have a faulty board, so this is really just curiosity, before I return it. I really hope it doesn't just "come good" for no apparent reason.
Re: 23 boots, 23 fail in some way
Posted: Sat Jan 17, 2015 5:00 amRe: 23 boots, 23 fail in some way
Posted: Sat Jan 17, 2015 7:35 pmWhat power supply are you using?
Re: 23 boots, 23 fail in some way
Posted: Sun Jan 18, 2015 1:09 am