eMMC

An embedded Multi-Media Card (eMMC) flash device is populated on the phyCORE-AM57x SOM as a programmable nonvolatile storage. This guide will show you how to mount, read from, and write to the phyCORE-AM57x onboard eMMC. To learn more information about the phyCORE-AM57x eMMC flash memory, please see section 6.2 in the Hardware Manual.

Note

In order to follow this guide your phyCORE-AM62x development kit must be booting from SD Card.

Viewing Available eMMC Partition Information

Note

Be careful with the fdisk command. You can accidentally delete the contents of other memory devices connected to the system so it is recommended to copy and paste the following commands.

You can verify the eMMC partitions by using the following command to list the partition information of known MMC devices:

Target (Linux)

 fdisk -l

Expected Output

root@phycore-am57xx-1:~# fdisk -l
Disk /dev/mmcblk1: 7264 MB, 7616856064 bytes, 14876672 sectors
922 cylinders, 256 heads, 63 sectors/track
Units: sectors of 1 * 512 = 512 bytes

Device       Boot StartCHS    EndCHS        StartLBA     EndLBA    Sectors  Size Id Type
/dev/mmcblk1p1    0,0,2       1023,255,63          1   14876671   14876671 7263M ee EFI GPT
Partition 1 has different physical/logical end:
     phys=(1023,255,63) logical=(922,105,41)
Disk /dev/mmcblk0: 15 GB, 15931539456 bytes, 31116288 sectors
243096 cylinders, 4 heads, 32 sectors/track
Units: sectors of 1 * 512 = 512 bytes

Device       Boot StartCHS    EndCHS        StartLBA     EndLBA    Sectors  Size Id Type
/dev/mmcblk0p1 *  16,0,1      1023,3,32         2048     133119     131072 64.0M  c Win95 FAT32 (LBA)
/dev/mmcblk0p2    1023,3,32   1023,3,32       133120    2036277    1903158  929M 83 Linux

As can be seen in the above Example Output, this eMMC has already been formatted with a data partition, /dev/mmcblk1p1. To follow the rest of this guide, one data partition will be required.

If you did not see a /dev/mmcblk1p* partition refer to the following commands in order to create one:

Target (Linux)

  fdisk /dev/mmcblk0

  # Enter into fdisk interactive session:
  # Use the following commands in order to create a Linux partition.

  n                         # create new partition
  p                         # partition type
  1                         # partition number
  <Enter>                   # first sector
  <Enter>                   # last sector
  t                         # change partition type
  83                        # linux filesystem
  w                         # write changes

Now reboot the system and you should see the eMMC device mounted automatically in the following steps.
Target (Linux)
```
 reboot
```
Verify the eMMC partitions.
Target (Linux)
```
 fdisk -l
```

Setup a Root Filesystem on the eMMC

In order to interact with the eMMC, the eMMC has to be partitioned and a rootfilesystem needs to be setup.

Create a ext4 type root filesystem

Target (Linux)

mkfs.ext4 /dev/mmcblk1p1

Expected Output

root@phycore-am57xx-1:~# mkfs.ext4 /dev/mmcblk1p1
mke2fs 1.45.7 (28-Jan-2021)
Discarding device blocks: done
Creating filesystem with 1859576 4k blocks and 465120 inodes
Filesystem UUID: 87f11df6-48ec-4ac9-bbe9-174ba7d1ff7a
Superblock backups stored on blocks:
       32768, 98304, 163840, 229376, 294912, 819200, 884736, 1605632

Allocating group tables: done
Writing inode tables: done
Creating journal (16384 blocks): done
Writing superblocks and filesystem accounting information: done

Mounting the eMMC

Typically, if the filesystem is already setup then it should mount automatically during boot at /run/media/. These steps will help you manually mount the filesystem.

In order to read and write to the eMMC you will need to create a directory and then mount the partition you want to read from to that directory. Follow the steps below to mount partition 1 (“p1”) of the eMMC (you should replace this with the “Linux” type partition if your eMMC was pre-flashed):
Target (Linux)
```
mkdir temp
mount /dev/mmcblk1p1 temp/
```
If you mounted a newly created partition it will likely be empty. If you had a pre-flashed eMMC then you will notice that the contents of the mounted directory look like the root directory!
Target (Linux)
```
ls temp
```

Writing to the eMMC

You can write to the eMMC by using the copy or move commands just like in Linux.
Target (Linux)
```
echo "Hello World" > test.txt
cp test.txt temp/
```
An example of what the contents look like following the copy is provided below:
Example Output
```
root@am57xx-phycore-kit:~# ls temp/
test.txt
```

Make sure the file was not corrupted during the transfer using md5sum.

Target (Linux)

md5sum test.txt ~/temp/test.txt

Expected Output

root@phycore-am57xx-1:~# md5sum test.txt ~/temp/test.txt
e59ff97941044f85df5297e1c302d260  test.txt
e59ff97941044f85df5297e1c302d260  /root/temp/test.txt

Reading from the eMMC

Use the copy (cp) or move (mv) command to put this file back onto your SD card.
Target (Linux)
```
cp ~/temp/test.txt ~/test-READ.txt
```

Make sure the file was not corrupted during the transfer using md5sum.

Target (Linux)

md5sum ~/temp/test.txt test-READ.txt

Expected Output

root@phycore-am57xx-1:~# md5sum ~/temp/test.txt test-READ.txt
e59ff97941044f85df5297e1c302d260  /root/temp/test.txt
e59ff97941044f85df5297e1c302d260  test-READ.txt