1.. SPDX-License-Identifier: GPL-2.0+ 2 3Colibri iMX7 4============ 5 6Quick Start 7----------- 8 9- Build U-Boot 10- NAND IMX image adjustments before flashing 11- Flashing manually U-Boot to eMMC 12- Flashing manually U-Boot to NAND 13- Using ``update_uboot`` script 14 15Build U-Boot 16------------ 17 18.. code-block:: bash 19 20 $ export CROSS_COMPILE=arm-linux-gnueabi- 21 $ make colibri_imx7_emmc_defconfig # For NAND: colibri_imx7_defconfig 22 $ make 23 24After build succeeds, you will obtain final ``u-boot-dtb.imx`` IMX specific 25image, ready for flashing (but check next section for additional 26adjustments). 27 28Final IMX program image includes (section ``6.6.7`` from `IMX7DRM 29<https://www.nxp.com/webapp/Download?colCode=IMX7DRM>`_): 30 31* **Image vector table** (IVT) for BootROM 32* **Boot data** -indicates the program image location, program image size 33 in bytes, and the plugin flag. 34* **Device configuration data** 35* **User image**: U-Boot image (``u-boot-dtb.bin``) 36 37 38IMX image adjustments prior to flashing 39--------------------------------------- 40 411. U-Boot for both Colibri iMX7 NAND and eMMC versions 42is built with HABv4 support (`AN4581.pdf 43<https://www.nxp.com/docs/en/application-note/AN4581.pdf>`_) 44enabled by default, which requires to generate a proper 45Command Sequence File (CSF) by srktool from NXP (not included in the 46U-Boot tree, check additional details in introduction_habv4.txt) 47and concatenate it to the final ``u-boot-dtb.imx``. 48 492. In case if you don't want to generate a proper ``CSF`` (for any reason), 50you still need to pad the IMX image so i has the same size as specified in 51in **Boot Data** section of IMX image. 52To obtain this value, run: 53 54.. code-block:: bash 55 56 $ od -X -N 0x30 u-boot-dtb.imx 57 0000000 402000d1 87800000 00000000 877ff42c 58 0000020 877ff420 877ff400 878a5000 00000000 59 ^^^^^^^^ 60 0000040 877ff000 000a8060 00000000 40b401d2 61 ^^^^^^^^ ^^^^^^^^ 62 63Where: 64 65* ``877ff400`` - IVT self address 66* ``877ff000`` - Program image address 67* ``000a8060`` - Program image size 68 69To calculate the padding: 70 71* IVT offset = ``0x877ff400`` - ``0x877ff000`` = ``0x400`` 72* Program image size = ``0xa8060`` - ``0x400`` = ``0xa7c60`` 73 74and then pad the image: 75 76.. code-block:: bash 77 78 $ objcopy -I binary -O binary --pad-to 0xa7c60 --gap-fill=0x00 \ 79 u-boot-dtb.imx u-boot-dtb.imx.zero-padded 80 813. Also, according to requirement from ``6.6.7.1``, the final image 82should have ``0x400`` offset for initial IVT table. 83 84For eMMC setup we handle this by flashing it to ``0x400``, howewer 85for NAND setup we adjust the image prior to flashing, adding padding in the 86beginning of the image. 87 88.. code-block:: bash 89 90 $ dd if=u-boot-dtb.imx.zero-padded of=u-boot-dtb.imx.ready bs=1024 seek=1 91 92Flash U-Boot IMX image to eMMC 93------------------------------ 94 95Flash the ``u-boot-dtb.imx.zero-padded`` binary to the primary eMMC hardware 96boot area partition: 97 98.. code-block:: bash 99 100 101 => load mmc 1:1 $loadaddr u-boot-dtb.imx.zero-padded 102 => setexpr blkcnt ${filesize} + 0x1ff && setexpr blkcnt ${blkcnt} / 0x200 103 => mmc dev 0 1 104 => mmc write ${loadaddr} 0x2 ${blkcnt} 105 106Flash U-Boot IMX image to NAND 107------------------------------ 108 109.. code-block:: bash 110 111 => load mmc 1:1 $loadaddr u-boot-dtb.imx.ready 112 => nand erase.part u-boot1 113 => nand write ${loadaddr} u-boot1 ${filesize} 114 => nand erase.part u-boot2 115 => nand write ${loadaddr} u-boot2 ${filesize} 116 117Using update_uboot script 118------------------------- 119 120You can also usb U-Boot env update_uboot script, 121which wraps all eMMC/NAND specific command invocation: 122 123.. code-block:: bash 124 125 => load mmc 1:1 $loadaddr u-boot-dtb.imx.ready 126 => run update_uboot 127