On 21 August 2012 08:26, Andreas Fritiofson
<[email protected]> wrote:
>
> Actually, the endianness seems fine, it's just that only the last of
> the accesses in arm7_9_read_memory() actually produces any data for
> size 1 or 2. At most 14 accesses are performed each round so a long
> dump of halfwords or bytes has every 14th element correct, the rest
> are zero.
>
> Every round, all of the accesses are loaded into separate registers
> and then all used registers are dumped to host in one operation. I
> didn't have time to examine whether the registers are loaded
> incorrectly or if it's the register dump that doesn't work for less
> than size 4.
>
> I have found a workaround that can be used as a quick-fix if we don't
> want to delay the release process. It's a quick-fix in the sense that
> it doesn't solve the underlying problem and it has a performance
> penalty (limits to one access per round).
>
> I'll push the workaround this evening after work and I'll also try to
> debug the issue further.
>

I did a quick test and reverting the following commit's fix all the
issues i see:

991ed5a2b657e660f744eefddb084724e52938ea
cb90d32e386a7489d31136997209c61e9559ff5e
c3074f377c1da33ca8ba8493826e1b52351eebc6

we can always use this as a fallback.

Spen

------------------------------------------------------------------------------
Live Security Virtual Conference
Exclusive live event will cover all the ways today's security and 
threat landscape has changed and how IT managers can respond. Discussions 
will include endpoint security, mobile security and the latest in malware 
threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/
_______________________________________________
OpenOCD-devel mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/openocd-devel

Reply via email to