Dear team,
My customer is facing the issue that i.MX6DualLite can hang when executing an instruction which disables CACHE after resuming from suspend mode.
This issue can occur on the customer’s board.
They found the hang-up occurs at when the following instruction was executed.
1573 /*
1574 * Restore control register but don't enable Data caches here.
1575 * Caches will be enabled after restoring MMU table entry.
1576 */
1577 ldmea r0!, {r4}
1578 mov r11, r4
1579 ldr r2, =CACHE_DISABLE_MASK
1580 and r4, r4, r2
1581 mcr p15, 0, r4, c1, c0, 0
1582 isb
The above is the source code of Android which is executed when resuming from suspended mode. And it seems to handle CACHE mode.
When i.MX6DL execute the instruction of line-1581, i.MX6DL hangs.
Could you show me why i.MX6DL can hang when manipulating CACHE mode?
And could you please show me the workaround?
Thanks,
Miyamoto
Solved! Go to Solution.
Dear Miyamoto,
As far as I understand, the issue is the same as discussed in the original case #00105430 in Salesforce. So, I continue working on the case there, so, further, please refer to this original case #00105430.
Also, I noticed that the customer uses very early 13.4.1 Android BSP. Ask them to try the latest Marshmallow 6.0.1_2.1.0 BSP, whose source code, pre-built images and corresponding documentation are available on the i.MX6/i.MX7 Software Center web page:
on their affected boards. It should improve the situation. If not, I will escalate the issue to R&D with all the details, found so far.
Best Regards,
Artur
Dear Miyamoto,
As far as I understand, the issue is the same as discussed in the original case #00105430 in Salesforce. So, I continue working on the case there, so, further, please refer to this original case #00105430.
Also, I noticed that the customer uses very early 13.4.1 Android BSP. Ask them to try the latest Marshmallow 6.0.1_2.1.0 BSP, whose source code, pre-built images and corresponding documentation are available on the i.MX6/i.MX7 Software Center web page:
on their affected boards. It should improve the situation. If not, I will escalate the issue to R&D with all the details, found so far.
Best Regards,
Artur
Hello Artur,
Thanks.
Yes, this is same as case #00105430.
Best Regards,
Miyamoto