lpmake is a command-line tool for generating a “super” partition image. It can currently generate two types of images:
The following command-line arguments are required:
-d,--device-size- The size of the “super” partition on the device. It must match exactly, and it must be evenly divisible by the sector size (512 bytes).
-m,--metadata-size- The maximum size that partition metadata may consume. A partition entry uses 64 bytes and an extent entry uses 16 bytes. The minimum size is 512 bytes.
-s,--metadata-slots- The number of slots available for storing metadata. This should match the number of update slots on the device, 2 for A/B devices and 1 for non-A/B.
-p,--partition=DATA- Add a partition to the metadata. At least one partition must be defined. The format for the data string is
<name>:<GUID>:<attributes>:<size>. The attributes must be either
readonly. The size will be rounded up to the nearest aligned block (see below).
-o,--output=FILE- The output file for writing the image.
--alignment=N- By default, lpmake will align partitions to 1MiB boundaries. However, an alternate alignment can be specified if desired. This is useful for devices with a minimum I/O request size where mis-aligned partition boundaries could be a performance issue.
--alignment-offset=N- In some cases, the “super” partition is misaligned within its parent block device. This offset can be used to correct for that.
--sparse- If set, the output image will be in sparse format for flashing with fastboot. Otherwise, by default, the image will be a minimal format usable with lpdump and lpflash.
-b,--block-size=N- When writing a sparse image, the device may require a specific block size. That block size can be specified here. The alignment must be a multiple of the block size. By default the block size is 4096.
-i,--image=[NAME=FILE]- When writing a sparse image, include the contents of FILE as the data for the partition named NAME. The file can be a normal file or a sparse image, but the destination size must be less than or equal to the partition size. This option is only available when creating sparse images.
Example usage. This specifies a 10GB super partition for an A/B device, with a single 64MiB “cache” partition.
lpmake --device-size 10240000000 \ --metadata-size 65536 \ --metadata-slots 2 \ -o /tmp/super.img \ -p "cache:2da85788-f0e1-4fda-9ee7-e5177eab184b:none:67108864" \ -i "cache=out/target/hikey960/cache.img"
lpdump displays pretty-printed partition metadata. It accepts a single argument, which can be:
It also accepts an optional argument
-s,--slot=N which can dump a specific metadata slot (the default is 0).
lpdump [-s,--slot=N] PATH
lpadd is a command-line tool for adding images to a super.img file, or a partition to a super_empty.img file. This is useful for mixed or split builds involving dynamic partitions. The syntax is:
lpadd [options] SUPER_FILE PART_NAME GROUP_NAME [IMAGE_FILE]
The parameters are:
--readonly- The partition should be mapped as read-only.
super_empty.imgfile. If the image is sparsed, it will be temporarily unsparsed, and re-sparsed at the end.
PART_NAME- The partition name. It must not already exist.
GROUP_NAME- The updateable group name for the partition.
IMAGE_FILE- If specified, the contents of the image will be embedded in the given super.img. This does not work for a
super_empty.imgfile. If the source image is sparsed, the unsparsed content will be embedded. The new partition size will be the smallest block-aligned size capable of holding the entire image.
Note that when interacting with sparsed images,
lpadd can consume a great deal of space in
TMPDIR does not have enough free space, it can be set in the environment, eg:
TMPDIR=/path/to/temp lpadd ...
lpflash writes a non-sparse image from lpmake to a block device. It is intended to be run on the device itself.
lpflash /dev/block/sdX /path/to/image/file