Getting Started with BPI-W3 - Revision history

Squy: /* nmcli */

2024-01-31T06:30:44Z

‎nmcli

Squy: /* nmtui */

2024-01-31T06:18:17Z

‎nmtui

Squy: /* Desktop connection */

2024-01-30T08:00:04Z

‎Desktop connection

Squy: /* Ethernet */

2024-01-30T03:17:12Z

‎Ethernet

Squy: /* Wlan&BT */

2024-01-30T01:55:09Z

‎Wlan&BT

Squy: /* Prepare */

2024-01-29T07:41:19Z

‎Prepare

Squy: /* Linux */

2024-01-29T07:35:53Z

‎Linux

Sinovoip: /* FAN */

2023-11-20T02:25:14Z

‎FAN

Sinovoip: /* NPU usage */

2023-09-07T06:37:22Z

‎NPU usage

Sinovoip: /* LED */

2023-09-07T06:35:05Z

‎LED

@@ Line 232: / Line 232: @@
 :Taking enP4p65s0 as an example, this step is similar to editing the network in nmtui, except that it changes from a graphical interface to a command line operation, with various names in the command line. This is only a partial introduction, and those interested can explore it on their own.
 :1.First, list the configuration of the connection. eth0 is currently connected to Wired connection 1
-   root@armsom-w3:/home/armsom# nmcli c s
+   root@w3:/home/armsom# nmcli c s
    NAME                UUID                                  TYPE      DEVICE
    Wired connection 1  e01f934d-7fae-344f-90bf-e2483db3f3e5  ethernet  enP4p65s0
@@ Line 252: / Line 252: @@
    sudo nmcli c up ifname eth0
 :4.After configuration, the IP changed
-   armsom@armsom-w3:~$ ip addr
+   root@w3:~$ ip addr
      lo: <LOOPBACK,UP,LOWER_UP> mtu 65536 qdisc noqueue state UNKNOWN group default qlen 1000
        link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00

@@ Line 162: / Line 162: @@
 ::2.Select the network you want to edit, here we take eth0 as an example
 ::Enter to see the default configuration of the network
+[[File:nmtui-network-ip.jpg]]
 ::3.Need to set IPV4 CONFIGURATION to Manual
@@ Line 228: / Line 228: @@
 bytes from 110.242.68.66 (110.242.68.66): icmp seg=2 ttl=50 time=54.2 ms
 bytes from 110.242.68.66 (110.242.68.66): icmp seg=3 ttl=50 time=45.8 ms
 ===nmcli===
 :Taking enP4p65s0 as an example, this step is similar to editing the network in nmtui, except that it changes from a graphical interface to a command line operation, with various names in the command line. This is only a partial introduction, and those interested can explore it on their own.

@@ Line 140: / Line 140: @@
 ===Desktop connection===
-:When using wired Internet access, please plug the network cable into the RJ45 interface. There will be a network connection icon in the upper right corner of the desktop. Select the network you want to connect to and enter the password to access the Internet.
+:Open WiFi, select the one you want to connect to, click and enter the password.
 ==Static network==
@@ Line 546: / Line 546: @@
   echo 1 > /sys/devices/platform/fd8b0010.pwm/pwm/pwmchip*/pwm0/enable
   #echo 0 > /sys/devices/platform/fd8b0010.pwm/pwm/pwmchip*/pwm0/enable
+==Overlays==
 ==Others==

@@ Line 141: / Line 141: @@
 ===Desktop connection===
 :When using wired Internet access, please plug the network cable into the RJ45 interface. There will be a network connection icon in the upper right corner of the desktop. Select the network you want to connect to and enter the password to access the Internet.
 ==Ethernet==

@@ Line 118: / Line 118: @@
 ==Wlan&BT==
 :J14 M.2 KEY E Slot is used for connect Wlan&BT adapter, default image support RTL8822CE(PCIE+USB) and RTL8822CS(SDIO+UART).
 ==Ethernet==

@@ Line 76: / Line 76: @@
 ==Prepare==
-:1. Good Type-C data cable for image download into EMMC
+:1. Usb to ttl cable if you need debug, default baudrate is 1500000.
-:2. Usb to ttl cable if you need debug, default baudrate is 1500000.
+:2. Download the  [https://wiki.banana-pi.org/Banana_Pi_BPI-W3#Linux Linux], and confirm that the md5 checksum is correct before flashing.
 ===Install Image to SDcard 1===

@@ Line 22: / Line 22: @@
 =specifications=
-=Linux=
+=Android=
 ==Prepare==
 :1. Good Type-C data cable for image download into EMMC
 :2. Usb to ttl cable if you need debug, default baudrate is 1500000.
 :3. Windows or Linux PC.
-:4. Download the Linux Release Image [https://wiki.banana-pi.org/Banana_Pi_BPI-W3#Image Release Linux Release Image], and confirm that the md5 checksum is correct before flashing.
+:4. Download the  [https://wiki.banana-pi.org/Banana_Pi_BPI-W3#Android Android], and confirm that the md5 checksum is correct before flashing.
 Note: The below picture is for BPI-W3's Debug Connecting, the red usb cable is Type-C data cable for image download,  the three colors usb to ttl cable is linux serial console, please install CH340E's USB convert serial driver.
 :[[File:linux_download.jpg|820px]]
-===Install Whole Image with Usb Download Tool===
+==Install Whole Image with Usb Download Tool==
-====Windows Driver install====
+===Windows Driver install===
 . Download and install [https://download.banana-pi.dev/d/ca025d76afd448aabc63/files/?p=%2FTools%2Fimage_download_tools%2FDriverAssitant_v5.11.zip Rockchip USB driver]
 :[[File:driver_install_2.PNG]]
@@ Line 54: / Line 51: @@
 :[[File:Bpi-r2pro-Firmware.png]]
-. Disconnect power adapter, connect dual male usb cable from the top usb host port to PC.Use the Type-C data cable to connect one end to the PC and the other end to the BPI-W3 board.
+. Disconnect the power adapter , connect the type-c cable to the bpi-w3 and PC.
 . There have two usb download mode for image upgrade.
@@ Line 76: / Line 73: @@
 :[[File:erase.PNG]]
-===Install Separate Images with Usb Download Tool===
+=Linux=
-. Download [https://download.banana-pi.dev/d/ca025d76afd448aabc63/files/?p=%2FTools%2Fimage_download_tools%2FUpdate-EMMC-Tools.zip Rockchip USB Download Tool for EMMC upgrade]
-. Open RKDevTool.exe, Switch to the “Download Image” page. The upgrade tool displays detailed all firmware image information. Then all image file is ready, you may push "run", the upgrade process begin.
+==Prepare==
-   Note: Because Linux's release package, Example: Ubuntu-22.04-Desktop-EMMCBoot-rockdev-MT7531SW-DSA.zip, unzip it into the same folder "Update-EMMC-Tools" inside.
+:1. Good Type-C data cable for image download into EMMC
-:[[File:Bpi-r2pro-emmc-flash.png]]
+:2. Usb to ttl cable if you need debug, default baudrate is 1500000.
-===Install Image with SDcard===
+::[[File:etcher.jpg]]
-. Download Rockchip [https://download.banana-pi.dev/d/ca025d76afd448aabc63/files/?p=%2FTools%2Fimage_download_tools%2FUpdate-SD-Tools.zip SDDiskTool].
+===Install Image to SDcard 2===
-. Insert microSD card to Windows PC, 8GB sdcard size at least.
+'''NOTE: Debian images use this method.'''
-. Run SD_Firmware_Tool, check the “Upgrade Firmware” box and select the correct removable disk device, Choose firmware image, then Click Create button to make it and wait until it is finshed
+: 1. Download Rockchip [https://download.banana-pi.dev/d/ca025d76afd448aabc63/files/?p=%2FTools%2Fimage_download_tools%2FUpdate-SD-Tools.zip SDDiskTool].
-:[[File:sd_download.png]]
+: 2. Insert card reader to Windows PC, 8GB sdcard size at least.
-. Safely remove the microSD card, insert it into the R2pro board, power on the board, it will start Linux System.
+:[[File:w3_debian.jpg]]
 ==Build Linux BSP Source Code==

← Older revision		Revision as of 02:25, 20 November 2023
Line 391:		Line 391:
	echo 1 > /sys/devices/platform/fd8b0010.pwm/pwm/pwmchip*/pwm0/enable		echo 1 > /sys/devices/platform/fd8b0010.pwm/pwm/pwmchip*/pwm0/enable
	#echo 0 > /sys/devices/platform/fd8b0010.pwm/pwm/pwmchip*/pwm0/enable		#echo 0 > /sys/devices/platform/fd8b0010.pwm/pwm/pwmchip*/pwm0/enable
		+
		+
		+	==Others==
		+	===NPU usage===
		+	‎<youtube width="1920" height="1080">y7mYxn3rq0U</youtube>
		+	====Prepare tools====
		+
		+	1. Use the Ubuntu18.04 / Ubuntu20.04 operating system (OS).
		+
		+	2. An W3-PRO board
		+
		+	====Preparation procedure====
		+	First make sure you have docker installed on your Ubuntu system,If not, refer to the Internet installation tutorial
		+
		+	We provide the source code and the docker image of the installed environment:[https://pan.baidu.com/s/1qm7qPQ-6Cau7kVhf-T-hzg?pwd=1234 docker image]
		+
		+	Create a rknpu folder on the PC server and copy the firmware to the folder
		+
		+	rknpu/rknn-toolkit2-1.4.0/docker$ ls
		+	md5sum.txt rknn-toolkit2-1.4.0-cp36-docker.tar.gz rknn-toolkit2-1.4.0-cp38-docker.tar.gz
		+
		+	Run the following command to run the docker image. After the Docker image is run, the bash environment of the image is displayed
		+
		+	docker run -t -i --privileged -v /dev/bus/usb:/dev/bus/usb rknn-toolkit2:1.4.0-cp38 /bin/bash
		+
		+	Map examples code into a Docker environment by attaching "-v <host src folder>:<image dst folder>"Parameters, such as:
		+
		+	docker run -t -i --privileged -v /dev/bus/usb:/dev/bus/usb -v /your/rknn-toolkit2-1.x.x/examples:/examples rknn-toolkit2:1.x.x /bin/bash
		+
		+	The code is synchronized after mapping
		+
		+	The rknn service needs to run on the development board
		+
		+	BOARD ARCH corresponds to the aarch64 directory on 64-bit Linux systems and to the armhf directory on 32-bit systems
		+
		+	1. adb push all files in Linux/rknn server/${B0ARD_ ARCH}/usr/bin/ to /usr/bin
		+	2. adb push Linux/librknn api/${BOARD ARCH}/ librknrnt. so to /usr/1ib
		+	3. Access the serial port terminal of the board and run the following command
		+	chmod +x /usr/bin/rknn server
		+	chmod +X /usr/bin/start_ rknn.sh
		+	chmod +X /usr/bin/restart rknn.sh
		+	restart_ rknn. sh
		+
		+	====run program====
		+	Execute adb devices in the docker image first, remembering the adb ID number
		+
		+	Go to /examples/onnx/yolov5 and change test.py
		+
		+	ret = rknn.init_runtime(target='rk3588', device_id=DEVICE_ID, perf_debug=True,eval_mem=True)
		+	outputs = rknn.inference(inputs=[img])
		+	ret = rknn.eval_perf(inputs=[img], is_print=True)
		+	cv2.imwrite("result.jpg", img_1)
		+
		+	The above four functions are not added
		+
		+	Run python3 test.py

@@ Line 238: / Line 238: @@
 ==NPU usage==
 Prepare tools
 . Use the Ubuntu18.04 / Ubuntu20.04 operating system (OS).
@@ Line 244: / Line 248: @@
 Preparation procedure
 First make sure you have docker installed on your Ubuntu system,If not, refer to the Internet installation tutorial
@@ Line 250: / Line 255: @@
 Create a rknpu folder on the PC server and copy the firmware to the folder
-rknpu/rknn-toolkit2-1.4.0/docker$ ls
+ rknpu/rknn-toolkit2-1.4.0/docker$ ls
-md5sum.txt  rknn-toolkit2-1.4.0-cp36-docker.tar.gz  rknn-toolkit2-1.4.0-cp38-docker.tar.gz
+ md5sum.txt  rknn-toolkit2-1.4.0-cp36-docker.tar.gz  rknn-toolkit2-1.4.0-cp38-docker.tar.gz
 Run the following command to run the docker image. After the Docker image is run, the bash environment of the image is displayed
-docker run -t -i --privileged -v /dev/bus/usb:/dev/bus/usb rknn-toolkit2:1.4.0-cp38 /bin/bash
+ docker run -t -i --privileged -v /dev/bus/usb:/dev/bus/usb rknn-toolkit2:1.4.0-cp38 /bin/bash
 Map examples code into a Docker environment by attaching "-v <host src folder>:<image dst folder>"Parameters, such as:
-docker run -t -i --privileged -v /dev/bus/usb:/dev/bus/usb -v /your/rknn-toolkit2-1.x.x/examples:/examples rknn-toolkit2:1.x.x /bin/bash
+ docker run -t -i --privileged -v /dev/bus/usb:/dev/bus/usb -v /your/rknn-toolkit2-1.x.x/examples:/examples rknn-toolkit2:1.x.x /bin/bash
 The code is synchronized after mapping
@@ Line 264: / Line 272: @@
 BOARD ARCH corresponds to the aarch64 directory on 64-bit Linux systems and to the armhf directory on 32-bit systems
 . adb push all files in Linux/rknn server/${B0ARD_ ARCH}/usr/bin/ to /usr/bin
 . adb push Linux/librknn api/${BOARD ARCH}/ librknrnt. so to /usr/1ib
 . Access the serial port terminal of the board and run the following command
   chmod +x /usr/bin/rknn server
   chmod +X /usr/bin/start_ rknn.sh
   chmod +X /usr/bin/restart rknn.sh
   restart_ rknn. sh
 run program
 Execute adb devices in the docker image first, remembering the adb ID number
 Go to /examples/onnx/yolov5 and change test.py
-ret = rknn.init_runtime(target='rk3588', device_id=DEVICE_ID, perf_debug=True,eval_mem=True)
+ ret = rknn.init_runtime(target='rk3588', device_id=DEVICE_ID, perf_debug=True,eval_mem=True)
-outputs = rknn.inference(inputs=[img])
+ outputs = rknn.inference(inputs=[img])
-ret = rknn.eval_perf(inputs=[img], is_print=True)
+ ret = rknn.eval_perf(inputs=[img], is_print=True)
-cv2.imwrite("result.jpg", img_1)
+ cv2.imwrite("result.jpg", img_1)
 The above four functions are not added

← Older revision		Revision as of 06:35, 7 September 2023
Line 236:		Line 236:

	[[File:1200px-Capture.png]]		[[File:1200px-Capture.png]]
		+
		+	==NPU usage==
		+	Prepare tools
		+	1. Use the Ubuntu18.04 / Ubuntu20.04 operating system (OS).
		+
		+	2. An W3-PRO board
		+
		+	Preparation procedure
		+	First make sure you have docker installed on your Ubuntu system,If not, refer to the Internet installation tutorial
		+
		+	We provide the source code and the docker image of the installed environment:docker image
		+
		+	Create a rknpu folder on the PC server and copy the firmware to the folder
		+
		+	rknpu/rknn-toolkit2-1.4.0/docker$ ls
		+	md5sum.txt rknn-toolkit2-1.4.0-cp36-docker.tar.gz rknn-toolkit2-1.4.0-cp38-docker.tar.gz
		+	Run the following command to run the docker image. After the Docker image is run, the bash environment of the image is displayed
		+
		+	docker run -t -i --privileged -v /dev/bus/usb:/dev/bus/usb rknn-toolkit2:1.4.0-cp38 /bin/bash
		+	Map examples code into a Docker environment by attaching "-v <host src folder>:<image dst folder>"Parameters, such as:
		+
		+	docker run -t -i --privileged -v /dev/bus/usb:/dev/bus/usb -v /your/rknn-toolkit2-1.x.x/examples:/examples rknn-toolkit2:1.x.x /bin/bash
		+	The code is synchronized after mapping
		+
		+	The rknn service needs to run on the development board
		+
		+	BOARD ARCH corresponds to the aarch64 directory on 64-bit Linux systems and to the armhf directory on 32-bit systems
		+
		+	1. adb push all files in Linux/rknn server/${B0ARD_ ARCH}/usr/bin/ to /usr/bin
		+	2. adb push Linux/librknn api/${BOARD ARCH}/ librknrnt. so to /usr/1ib
		+	3. Access the serial port terminal of the board and run the following command
		+	chmod +x /usr/bin/rknn server
		+	chmod +X /usr/bin/start_ rknn.sh
		+	chmod +X /usr/bin/restart rknn.sh
		+	restart_ rknn. sh
		+	run program
		+	Execute adb devices in the docker image first, remembering the adb ID number
		+
		+	Go to /examples/onnx/yolov5 and change test.py
		+
		+	ret = rknn.init_runtime(target='rk3588', device_id=DEVICE_ID, perf_debug=True,eval_mem=True)
		+	outputs = rknn.inference(inputs=[img])
		+	ret = rknn.eval_perf(inputs=[img], is_print=True)
		+	cv2.imwrite("result.jpg", img_1)
		+	The above four functions are not added
		+
		+	Run python3 test.py

	==LED==		==LED==