Peter Mamonov wrote:
> On Tue, 04 Aug 2015 08:45:49 -0400
> Sinan Akman <[email protected]> wrote:
> 
>>   Hi Peter
>>
>> Peter Mamonov wrote:
>>> Hi Sinan,
>>>
>>> I was able to halt while u-boot was running. The situation becomes
>>> more complicated  after linux start: linux runs on both cores and
>>> after one core halts, it reboots the system.
>>>
>>> The crucial configuration option is "reset_config trst_and_srst":
>>> scan chain won't work until openocd issues tap reset using hardware
>>> TRST signal. Here is the config, that works for me:
>>    Thanks much for this. I realized there is no JTAG communication
>> possible with CN70xx while reset is asserted. Were you able
>> to RESET HALT and stop soon after reset ?
> 
> Didn't try "reset halt". Guess it should work.

   For CN70xx, reset halt does not work as there is no debug
communication possible when RESET is asserted. I assume this
is not the case for CN610xx and it should halt at the reset
vector.

> By the way, you can "press any key" when u-boot prompts for it, to
> interrupt linux autoboot.

   Sure this is known but I was more referring to openocd being
able to halt at the reset vector or soon after if a rom code is
running.

> 
>>    Issuing HALT specifically while u-boot is coming up works
>> but resume generates an "ERROR wait_for_pracc_rw", I am not
>> sure if you were seeing this with cn61xx.
> 
> Just ignore this error message after "resume" command. Actually it
> should be a debug-level message (invisible for a user), as it is in the 
> mips32 target.

   So I guess you are suggesting to change this to debug info
instead of an error. I am running with debug_level 0 and
as it is an error it is still coming up.

   What is generating this by the way. I didn't take a look
at the source in detail yet but if it is a result of a an early
poll we can perhaps have the error only after few attempts.

> 
>>    I'll debug this further and also take a look at halting
>> cores while Linux is running.
> 
> Thanks for testing! Could you, please, report results of the testing
> here: http://openocd.zylin.com/#/c/2321/ ?

   Sure I will definitely do so. I actually tested with Linux and
I observe same reboot as you did. To halt all the cores I believe
we will have to set Root.MultiCoreDebug[] to prevent further
interrupts but I have to read up on that some more.

   Thanks for your patches for adding mips64 support.

   Regards
   Sinan Akman

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

Reply via email to