LS1043 Memory ERR_detect (ERR_DETECT, E40h)

[BruceH]

Hi 

A DDR error was found during the LS1043 test, Register ERR_DETECT(E40h) ==> read not equal to 0, because the log is not recorded, we are not sure which bit is pulled up.

But two other tests at the same time did not see any problem
1. E58h:Single-Bit ECC memory error management (ERR_SBE) no detect to error.
2. Memory test pass.

==========================================================
1. We want to confirm that ERR_DETECT(E40h) is just a warning massege means there is a data transfer error or we should be aware of other issues with DDR?
2. In addition, when the E40h detects an error, it will continue to pull high under power on, is it because we need to clear it? Or will it automatically clear without error?
3. The value read by E40h after restarting is 0, and the abnormal phenomenon disappears

Thanks

Bruce

[yipingwang]

Please refer to the following update from the AE team.

Please provide more information:
what is the context for this inquiry?
is this ERR_DETECT <> 0 observed during a QCVS (DDRV) tool test? if yes please provide the test log.
is this a new board bring up activity and customer is observing some test fails in QCVS tool?
is this a TFA/uboot failure? board fails to get to the uboot prompt?

is customer connected with a CWTAP to the board? if yes they can get the full DDR register dump after failure (uboot or qcvs) via ccs.

How to get register dump via ccs
This is for debug use only.
1) Open a CCS window (C:\Freescale\CW4NET_vxxxxxx\Common\CCS\bin\ccs.exe)
2) Physical connection: USB to PC, JTAG to the customer board.
3) SW connection: in the ccs window type:


delete all
config cc cwtap
ccs::config_chain {ls1043a dap sap2}
display ccs::read_mem 32 0x1080000 4 0 1024
ccs::write_mem 32 0x1080FB0 4 0 0x10000000
display ccs::read_mem 32 0x1080000 4 0 1024

[BruceH]

Hi yiping

what is the context for this inquiry? ==>
The reason for asking this question is that when we did the memory test of LS1043, we also monitored "ERR_DETECT, E40h", and found that one or more bits of the E40h register were pulled up during a read and write test. Since the test program does not record the read value, we only know that ERR_DETECT has detected an error but not sure what kind of error it is.

This register is cleared after rebooting, and no error is detected by continuing the memory test.

We just want to make sure that this register only reflects a data error, not hardware damage to the LS1043 or the memory.

Thank you.

Bruce

[yipingwang]

the ERR_DETEC registers is describe in reference manual. what is the value of ERR_DETECT when the test fails? the ERR_DETECT after reporting an error will remain till it is cleared by SW by writing a 1's to it. or cleared after a reset.