EPICS Home EPICS Controls Argonne National Laboratory

Experimental Physics and
Industrial Control System

Format page
for printing


Search Tech-talk
1994  1995  1996  1997  1998  1999  2000  2001  2002  2003  2004  2005  2006  2007  2008  2009  2010  2011  2012  2013  <2014>  2015  2016  2017  2018  2019  2020  2021  2022  2023  2024  Index 1994  1995  1996  1997  1998  1999  2000  2001  2002  2003  2004  2005  2006  2007  2008  2009  2010  2011  2012  2013  <2014>  2015  2016  2017  2018  2019  2020  2021  2022  2023  2024 
<== Date ==> <== Thread ==>
Subject: Re: Newport XPS-Q8 and Motor Record - armv5teb architecture
From: Stephen Beckwith <[email protected]>
To: "J. Lewis Muir" <[email protected]>
Cc: Tonia Batten <[email protected]>, "[email protected]" <[email protected]>
Date: Tue, 6 May 2014 11:15:00 -0400
Tonia,
   Yes, as J. Lewis Muir said:  if you move the order the fields in the structure, might shed some light on things.
Remember, ARM is MUCH more restrictive in alignment than Intel. 

From the ARM Reference Manual:  (A2.8)
Prior to ARMv6, doubleword (
LDRD
/
STRD
) accesses to memory, where the a
ddress is not doubleword-aligned,
are
UNPREDICTABLE
. Also, data accesses to non-aligned word an
d halfword data are treated as aligned from
the memory interface perspective. That is:
• the address is treated as truncat
ed, with address bits[1:0] treate
d as zero for word accesses, and
address bit[0] treated as
zero for halfword accesses.
• load single word ARM instructions
are architecturally defined to ro
tate right the wo
rd-aligned data
transferred by a non word-aligned address one, two
or three bytes depending on the value of the two
least significant address bits.
• alignment checking is defined fo
r implementations supporting a Syst
em Control coprocessor using
the A bit in CP15 register 1

As previously suggested, printout the "thing" being pointed to (message) at point A (where correct) then again at point B (where incorrect) and look for differences.
The Linux 2.6.10 is a very old kernel.  Might be some library/kernel issue as well. . .

Is this structure "statically" allocated (i.e. not on the stack)?
Also, check the compiler directives w.r.t. alignment issues.

Your idea of a test program is also very good. 

Regards,
Stephen Beckwith



On Tue, May 6, 2014 at 10:53 AM, J. Lewis Muir <[email protected]> wrote:
On 5/6/14, 9:41 AM, Tonia Batten wrote:
> What do you mean by "What happens if you move the double and int?"  Do
> you mean casting or copying the value?  I am wondering about setting
> up a simple c program just to verify that the value is correct when
> casting to different types.

I would guess he means change the order of the fields in the struct.

Lewis

References:
RE: Newport XPS-Q8 and Motor Record - armv5teb architecture Mark Rivers
Re: Newport XPS-Q8 and Motor Record - armv5teb architecture J. Lewis Muir
Re: Newport XPS-Q8 and Motor Record - armv5teb architecture Stephen Beckwith
RE: Newport XPS-Q8 and Motor Record - armv5teb architecture Tonia Batten
Re: Newport XPS-Q8 and Motor Record - armv5teb architecture J. Lewis Muir
Navigate by Date:
Prev: Re: Newport XPS-Q8 and Motor Record - armv5teb architecture J. Lewis Muir
Next: Trouble with OID name using devsnmp. Nicoletta Petrella
Index: 1994  1995  1996  1997  1998  1999  2000  2001  2002  2003  2004  2005  2006  2007  2008  2009  2010  2011  2012  2013  <2014>  2015  2016  2017  2018  2019  2020  2021  2022  2023  2024 
Navigate by Thread:
Prev: Re: Newport XPS-Q8 and Motor Record - armv5teb architecture J. Lewis Muir
Next: Re: Newport XPS-Q8 and Motor Record - armv5teb architecture Torsten Bögershausen
Index: 1994  1995  1996  1997  1998  1999  2000  2001  2002  2003  2004  2005  2006  2007  2008  2009  2010  2011  2012  2013  <2014>  2015  2016  2017  2018  2019  2020  2021  2022  2023  2024 
ANJ, 17 Dec 2015 Valid HTML 4.01! · Home · News · About · Base · Modules · Extensions · Distributions · Download ·
· Search · EPICS V4 · IRMIS · Talk · Bugs · Documents · Links · Licensing ·