LS1088ARDB Boot Error

[aaronYR]

NOTICE: BL2: v2.4(release):LSDK-21.08-0-g340b20bcb
NOTICE: BL2: Built : 22:23:06, Aug 27 2021
ERROR: First slot cannot be empty.
ERROR: Parsing DIMM Error
ERROR: DDR init failed.
NOTICE: Incorrect DRAM0 size is defined in platform_def.h
ERROR: mmap_add_region_check() failed. error -22
ERROR: mmap_add_region_check() failed. error -22
NOTICE: BL2: v2.4(release):LSDK-21.08-0-g340b20bcb
NOTICE: BL2: Built : 22:23:06, Aug 27 2021
ERROR: First slot cannot be empty.
ERROR: Parsing DIMM Error
ERROR: DDR init failed.
NOTICE: Incorrect DRAM0 size is defined in platform_def.h
ERROR: mmap_add_region_check() failed. error -22
ERROR: mmap_add_region_check() failed. error -22
NOTICE: BL2: v2.4(release):LSDK-21.08-0-g340b20bcb
NOTICE: BL2: Built : 22:23:06, Aug 27 2021


I keep getting that error and there are no signs of physical damage and I swapped the RAM. 

[June_Lu]

Yes, it's possible a hardware issue. And SDA is also connected to U207, pin G5.

[June_Lu]

Could you please help check why the J33_SDA pin stays at 0 V?

According to the schematic, it is pulled up to 3V3_SB through R619 (10 kΩ). [SPF-28869_C | PDF]

1) Please measure 3V3_SB

Could you measure 3V3_SB (for example, at the R619 3V3_SB-side pad) to confirm whether 3V3_SB is 0 V?

2) If 3V3_SB is OK (not 0 V)

Then SDA being 0 V usually indicates that one device on the same net is pulling the line low. Could you help identify which device(s) share the same SDA net and check whether any of them could be holding SDA low?

3) To exclude software configuration impact

You may also program the default LS1088ARDB firmware image and boot from SD card to put the board into a known default state.

wget https://www.nxp.com/lgfiles/sdk/lsdk2108/firmware_ls1088ardb_pb_sdboot.img

If you cannot flash the firmware via U-Boot, please program the SD card offline and boot the LS1088ARDB from the SD card.

Thanks

[aaronYR]

I performed the requested measurements and have the following results:

1. 3V3_SB measurement
I measured the 3V3_SB rail at the R619 3V3_SB-side pad and confirmed it is present at approximately 3.3 V, so the pull-up supply appears to be OK.

2. R619 / SDA behavior
One side of R619 (3V3_SB side) measures ~3.3 V
The other side (connected to J33_SDA) remains at 0 V

3. SDA line investigation

J33 pin 3 (SDA) shows a very low resistance to GND (~10–20 Ω) with the board powered off
Devices on the SDA net
From the schematic, I identified that the SDA net is shared with the PCA9547 (U87)
I measured directly at U87 pin 23 (SDA) and observed the same low resistance to GND (~10–20 Ω). SCL does not exhibit this behavior

At this point, it appears that something on the SDA net is pulling the line low at a hardware level. Based on measurements, U87 (PCA9547) or something physically on the SDA trace/header area could be responsible.

[June_Lu]

To help narrow down whether this is a hardware or software issue, could you please check the I2C signals with an oscilloscope?

[aaronYR]

It is a hardware problem. The SDA pin on the J33 I2C interface is reading 0V. Every other interface reads correct voltage.

[aaronYR]

I even tried booting from SD Card however I am still getting the same issue. The card had firmware_ls1088ardb_pb_sdboot.img installed. 

I changed the boot switches to be for sd card boot up as well.

[June_Lu]

How is the SE050 connected to the LS1088ARDB?

Thanks

[aaronYR]

It was through I2C. 

However, booting without the HSM still spits out the same issue.

[aaronYR]

I tested both SO-DIMM modules that came with the board.

I also tested booting with no DIMM installed.

All three cases produce the same BL2 error sequence: First slot cannot be empty, Parsing DIMM Error, DDR init failed.

The issue started after probing/working around the I2C interface for an SE050 connection.

[June_Lu]

Just to make sure we’re on the same page, when you mention the RAM, do you mean the DIMM module?

Also, when you say it was “swapped,” could you clarify what was swapped exactly?

Could you also confirm whether the DIMM currently installed on the LS1088ARDB is the original one that came with the board?

If the DIMM is confirmed to be the original and looks fine, the issue may be related to DIMM SPD (I2C) access during DDR initialization.

At this point, it would be helpful to take a closer look at the SPD/I2C behavior using more detailed BL2 logs. If it’s convenient on your side, checking the I2C signals or using a debug build of the ATF firmware (with DEBUG enabled) could help.

Thanks