On 09/07/2015 08:42 PM, Andreas Färber wrote:
> Hi Marc,
> 
> Am 07.09.2015 um 17:42 schrieb Marc Kleine-Budde:
>> I'm on a dual-core imx6, trying to debug some strange problems in the
>> Linux kernel. The JTAG is a SEGGER J-Link ARM.
>>
>> I'm using gdb to get backtraces while the CPU is in the kernel, this
>> however often fails. Failing backtraces are only a symptom of what's
>> wrong, which is the way the gdb sees the memory.
>>
>> I think the gdb doesn't read via the CPU->L1->L2->Memory, it probably
>> skips the caches altogether. Is there a way to configure gdb to access
>> the memory the same way the CPU does?
> 
> You may be interested in this thread:
> https://www.mail-archive.com/openocd-devel%40lists.sourceforge.net/msg07268.html

Thanks for pointing me to this thread, I agree 100% with Paul here, both
gdb and openocd should see the memory the same way as the firmware does.
openocd does, but gdb doesn't.

> Oleksij has posted several patches to Gerrit that you may want to review
> and test. It seems that on the Cortex-A side things are generally moving
> slower than on the Cortex-M side...

I hope I've all necessary patches, I'm using Oleksij's
olerem/l2x-2015.07.29 branch.

Marc

-- 
Pengutronix e.K.                  | Marc Kleine-Budde           |
Industrial Linux Solutions        | Phone: +49-231-2826-924     |
Vertretung West/Dortmund          | Fax:   +49-5121-206917-5555 |
Amtsgericht Hildesheim, HRA 2686  | http://www.pengutronix.de   |

Attachment: signature.asc
Description: OpenPGP digital signature

------------------------------------------------------------------------------
_______________________________________________
OpenOCD-devel mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/openocd-devel

Reply via email to