⚠️ Warning.
The pre-loaded bootloaders (u-boot and ARM Trusted Firmware) on the RZ/V2L Evaluation Kit (SMARC board) are not up to date with the current BSP. It is strongly recommended that the bootloaders generated when building the BSP as described in the Startup Guide document that is distributed with the BSP Package, be programmed into SPI Flash on the board using Flash Writer.
Known issues with pre-loaded loaders.
Renesas has prepared pre-built images available for download from renesas.com.
⭐⭐ RZ/V2L AI SDK Applications Demo SD Image
RZ/V2L DRP-AI Object Detection and Pose Estimation DEMO
RZ/V2L DRP-AI USB Camera HTTP Demo SD Image version
RZ/V2L DRP-AI USB Camera HTTP Demo Pre-build version
Please see the DRP-AI Pretrained Models page for details about pretrained applications.
The RZV2L DRP-AI Demos can be found here.
These demos include pre-build applications, source code and documentation about the DRP-AI.
Information about the RZV2L Simple ISP can be found here.
These demos include pre-build applications, source code and documentation about the DRP-AI and ISP.
📄 Board User Manual
The User Manual for the boards can be found their website pages on renesas.com.
Check the 'General Information' section on this page for the links.
📦 What is included in the kit
🛒 What you will need to purchase
The Renesas RZV AI SDK provides board bring up of the EVK and AI Demos.
Info | ||
---|---|---|
| ||
Please read section V2L Boot Loaders Warning about the factory boot loaders programmed into the EVK boards. |
$ wget https://raw.githubusercontent.com/renesas-rz/rzg2_bsp_scripts/master/usb_sd_partition/usb_sd_partition.sh $ chmod +x usb_sd_partition.sh $ ./usb_sd_partition.sh
# Change to the Yocto output directory that contains the files $ cd build/tmp/deploy/images/smarc-rzv2l # Copy the Linux kernel and Device Tree to partition 1 $ sudo cp -v Image /media/$USER/RZ_FAT $ sudo cp -v r9a07g054l2-smarc.dtb /media/$USER/RZ_FAT # Copy and expand the Root File System to partition 2 $ sudo tar -xvf core-image-weston-smarc-rzv2l.tar.gz -C /media/$USER/RZ_ext (or if NOT using graphics) $ sudo tar -xvf core-image-minimal-smarc-rzv2l.tar.gz -C /media/$USER/RZ_ext # Make sure all files are finished writing before removing the USB card reader from the PC $ sync
SOM board uses SD Card socket SOM board uses eMMC Flash (recommended) SW1-1 = ON/OFF(JTAG) SW1-1 = ON/OFF(JTAG) SW1-2 = ON SW1-2 = OFF +-----+ +-----+ | ON | | ON | | = = | | = | | | | = | | 1 2 | | 1 2 | +-----+ +-----+
Info | ||
---|---|---|
| ||
Please see RZ/V2LBootLoadersWarning about the factory programmed evaluation boards |
HTML |
---|
<font size=2><B>Boot using SD Card on Carrier Board (<font color=green>Recommended</font>)</B></font> |
┌─────┬─────┬─────┐ │ │ SOM │ │ │ └─────┘ │ │ Carrier Board │ │ ┌─┐ │ └─────┴─┴─────────┘ ↑ SD Card Socket
RZ/V2L EVK Booting Commands:
# Create command macros and save them: => setenv sd_boot1 'mmc dev 1 ; fatload mmc 1:1 0x48080000 Image ; fatload mmc 1:1 0x48000000 Image-r9a07g054l2-smarc.dtb' => setenv sd_boot2 'setenv bootargs 'root=/dev/mmcblk1p2 rootwait' ; booti 0x48080000 - 0x48000000' => setenv bootcmd 'run sd_boot1 sd_boot2' => saveenv # Reset the board and it will automatically boot
HTML |
---|
<font size=2><B>Boot using eMMC Flash on SOM Board (Advanced)</B></font> |
HTML |
---|
<font size=2><B>Boot using SD Card on SOM Board (Advanced)</B></font> |
# Manually enter the following commands to boot => mmc dev 0 ; fatload mmc 0:1 0x48080000 Image ; fatload mmc 0:1 0x48000000 Image-r9a07g054l2-smarc.dtb => setenv bootargs 'root=/dev/mmcblk0p2 rootwait' ; booti 0x48080000 - 0x48000000
Instructions for RZV Cortex-M33 Multi-OS Package is located below.
Cortex-M33 Multi-OS Package Support
Additional Supported Camera Modules are listed below:
e-CAM20_CURZ is an AR0234-based Full HD Color Global Shutter Camera. It supports the UYVY image format and is provided with an M12 lens holder – which together make it compatible with Renesas ® RZ/V2L. Its key features make e-CAM20_CURZ a perfect fit for applications like AMR, smart shopping, smart traffic, etc. (Order number: e-CAM20_CURZ_1H01R2)
e-CAM21_CURZ is a Full HD Ultra low light camera for Renesas ® RZ/V2L based on the Sony® STARVIS™ IMX462 sensor. It comes with superior NIR performance and is equipped with an inbuilt ISP and an M12 lens holder – making it ideal for applications like people counting, animal tracking, smart agriculture, etc. (Order Number: 2. e-CAM21_CURZ_1H01R2 )
The instructions to enable the RZV2L to boot from eMMC Flash at on the RZG2L Page here
The instructions to enable the RZV2L to load the Linux kernel, DTB, and Filesystem RZG2L Page here
Fix rngd bug that randomly keeps all CPUs busy after startup
RZ MPUs (at least non-secure versions) do not have a hardware random number generation module (/dev/hwrng), but that's the default setting for the systemd rngd service.
With that, rngd occupies the CPU resources to build up entropies during start-up.
Randomly, this process does not end properly and slows down the entire system significantly.
But if an available pseudorandom number generator is selected, rngd finishes properly always.
The following modification is sufficient:
Before you build the BSP, modify
poky/meta/recipes-support/rng-tools/rng-tools/default
from
EXTRA_ARGS="-r /dev/hwrng"
to
EXTRA_ARGS="-r /dev/urandom -x jitter"
However, normally you do not want to modify the original recipe but instead create a new one to override the old one.
mkdir - p meta-renesas/recipes-rzg2l/recipes-support/rng-tools
cd meta-renesas/recipes-rzg2l/recipes-support/rng-tools
vi rng-tools_%.bbappend
FILESEXTRAPATHS:prepend := "${THISDIR}/${PN}:"
SRC_URI += "file://default"(1)
mkdir -p rng-tools
vi default
EXTRA_ARGS="-r /dev/urandom -x jitter"
bitbake core-image-weston
(1) Note that this is the new Yocto syntax. For older versions of Yocto you may need to replace : (colon) with _ (underscore): so FILESEXTRAPATHS_prepend
.
Alternatively, you can modify the
/etc/default/rng-tools
file in the current target rootfs accordingly.
This version to the DRP-AI Reserve Area has changed from 0x70000000 to 0x80000000, because of this older versions of the DRP-AI Implementation Source code will fail with the following errors.
[ERROR] Failed to run DRPAI_ASSIGN : errno=22
[ERROR] Failed to run load_data_to_mem : resnet50_cam/drp_desc.bin
[ERROR] Failed to load DRPAI Data
This is due to the generated DRP Translator files map to the older address. Please use the DRP Translator version 1.70 when generating the DRP files and use the latest sample code included in the DRP-AI package version 7.0.
HTML |
---|
Table of Contents | ||||||
---|---|---|---|---|---|---|
|