Re: [ARMini-support] Occasional crashes
- From: Adrian Lees <a.lees@xxxxxxxxxxx>
- To: "armini-support@xxxxxxxxxxxxx" <armini-support@xxxxxxxxxxxxx>
- Date: Thu, 25 Dec 2014 20:09:58 +0000 (UTC)
My understanding was that the PandaBoard port was relatively immature -
although perhaps I am out of touch - and I would be surprised if, from what
I've read, it turned out to be an hardware problem. My expectation is that it
is a software incompatibility, perhaps related to reentrancy or cache cleaning
- something low-level and difficult to track down - that requires the attention
and analysis of a programmer with detailed knowledge of the OMAP4 and its RISC
OS port. As such, register/memory dumps at the point of failure are perhaps the
only way forwards.
The fact that multiple applications are affected does itself exonerate the
application code to some extent; remember that when the WindowManager reports
any exception it is the currently-running application that gets the 'blame,'
but there may be background code running, not on its behalf, and/or it may
innocently have induced a failure in other code by making a legitimate request
of it.
I do have a PandaBoard, although I have little knowledge of the port, so cannot
really help at the moment, sorry. I am happy to report that I have not seen any
unpredictable behaviour on my i.MX5 board. My focus at the moment is Aemulor.
Best wishes to all, have a great New Year,
A
From: Chris Bell (Lists) <chris.lists@xxxxxxxxxxxx>
To: armini-support@xxxxxxxxxxxxx
Sent: Thursday, 25 December 2014, 18:12
Subject: [ARMini-support] Occasional crashes
Chris Bell (Lists) <chris.lists@xxxxxxxxxxxx> wrote:
> As regards processor speed, I've reduced mine to 920 MHz, courtesy of
> Chris Johnson's CPUClock, and will leave the computer running 24/7 to
> see if there's any difference.
No, I've come back a few hours later to discover that LanMan98 has
crashed in my absence. 920MHz made no difference.
> I've now set 'fast' = 'slow' = 350MHz so there's no 'clocking
> back up' at all. I'll let you all know if that sorts it out.
No, no 'clocking back up', yet this time Temp-Bin had crashed.
Something else is wrong.
Chris.
--
---
To alter your preferences or leave the group,
visit //www.freelists.org/list/armini-support
List-related queries to info@xxxxxxxxxxxx
Other related posts:
