LS1043A - Flashing SD Card

cancel
Showing results for 
Show  only  | Search instead for 
Did you mean: 

LS1043A - Flashing SD Card

741 Views
tmorrison
Contributor III

Hi,

Currently I have a script that flashes the SD Card as follows:

   dd if=u.bin of=/dev/mmcblk0 bs=512 seek=8 && sync
   dd if=fsl_fman_ucode_t2080_r1.1_108_4_5.bin of=/dev/mmcblk0 bs=512 seek=1500 && sync
   dd if=iram_Type_A_LS1021a_r1.0.bin of=/dev/mmcblk0 bs=512 seek=1580 && sync

I was having problems when the SPL was loading that part of the data was being corrupted.

Well, I had a thought. What if the U-Boot was extending beyond the 1500 sector where I write

the fsl_fman_ucodexxx.bin?

Well, if I changed these lines to:

   dd if=fsl_fman_ucode_t2080_r1.1_108_4_5.bin of=/dev/mmcblk0 bs=512 seek=1800 && sync
   dd if=iram_Type_A_LS1021a_r1.0.bin of=/dev/mmcblk0 bs=512 seek=1880 && sync

U-Boot data wasn't being corrupted.

QUESTION:

  1. What module does the loading of the fsl_fma_xxx.bin from sector 1500
  2. What module does the loading of the iram_Type_xxxx.bin from sector 1580

Obviously, I need to modify these locations in software!

Tom

Labels (1)
0 Kudos
3 Replies

634 Views
yipingwang
NXP TechSupport
NXP TechSupport

Hello Thomas Morrison,

Please refer to section "4.2 LSDK memory layout and Userland" in LSDK 19.09 user manual which could be downloaded from Layerscape SDK | NXP .

DPAA1 FMAN ucode should be deploy at block no. 0x04800(18432)
QE firmware should be deployed at block no. 0x04A00(18944)

Thanks,

Yiping

0 Kudos

634 Views
tmorrison
Contributor III

Hi Yiping,

I do understand that - I am asking for some guidance if our u-boot is larger than the allocated area.

As indicated from the original post, we were given by NXP a script that wrote these 3 sections:

   dd if=u.bin of=/dev/mmcblk0 bs=512 seek=8 && sync
   dd if=fsl_fman_ucode_t2080_r1.1_108_4_5.bin of=/dev/mmcblk0 bs=512 seek=1500 && sync
   dd if=iram_Type_A_LS1021a_r1.0.bin of=/dev/mmcblk0 bs=512 seek=1580 && sync

But, when writing u-boot (dd if=u.bin of=/dev/mmcblk0 bs=512 seek=8 && sync) indicates:

   "1621+1 records out"

Unfortunately, the fma code starts being written at sector 1500... thus overwriting our code.

Are you saying that the script we have is wrong?

Thanks in advance.

Tom Morrison

0 Kudos

634 Views
tmorrison
Contributor III

Never mind - I figured it out. 

I have many more questions though

0 Kudos