→ Check out the preceding part of this tutorial series before starting on this: nRF Connect SDK Tutorial series - Part 0
This part of the NCS Tutorial series will be short and concise and will get you up and running with NCS and your board as quickly as possible. The first section will demonstrate how to build and run a “hello world” sample on your board, and the second section will provide you with the most vital information for the beginning phase for your particular board.
As mentioned in Part 0, you should choose a board and set <board> and <board_variant> accordingly. This applies to this part as well, and when you see one of these variables use your chosen value instead.
Contents
In this setup, the sample will be put outside the nrf folder. This is a cleaner and more modular approach than building it directly in the nrf folder since you keep the project detached from the nRF Connect SDK.
It should look like this:
cmake_minimum_required(VERSION 3.8.2)
find_package(Zephyr REQUIRED HINTS $ENV{ZEPHYR_BASE})
project(NONE)
target_sources(app PRIVATE src/main.c)
CONFIG_SERIAL=y
This config is required in order to see the logging from a serial terminal and will enable the serial driver. However, it is actually not necessary to enable it, as it's are already enabled by default in NCS for the listed boards. E.g. in the board folder for the nRF9160 DK (non-secure) this configuration is set inside nrf9160dk_nrf9160ns_defconfig. I included it anyway to make you familiar with the prj.conf file.
The main.c file should look like this:
#include <zephyr.h>
#include <sys/printk.h>
void main(void)
{
printk("Hello World!\n");
}
There is a guide in the Toolchain Manager on how to build and run a project, Open Toolchain Manager (TM) app and click on "First step to build" to see it. Choose the one with <version> equal to v1.4.0
For the record, I will go through the process here as well.
Two important environment variables to set are the Zephyr base and the GNU ARM Embedded Toolchain directory, however, if you've installed NCS through the TM, these will be set automatically. In SES, click on Tools→Options→nRF Connect to see the values (see image below). E.g. if you're using NCS v1.4.0, the GNU Toolchain and the Zephyr base may be equal to respectively "C:\Users\<user name>\ncs\v1.4.0\toolchain\opt" and "C:\Users\<user name>\ncs\v1.4.0\zephyr".
E.g. if you are building with the board variant nrf9160dk_nrf9160ns, it should look like this:
If you are using the nRF9160 DK, make sure SW5 is set to nRF91
After resetting the board, you should see the following log inside one of the terminals:
*** Booting Zephyr OS build v2.1.99-ncs1 ***
Hello World!
If you've installed NCS through the Toolchain Manager it is not necessary to set any environmental variables since everything is set automatically upon installation and runs out of the box. Please skip section 1.3.1 and go straight to section 1.3.2 Build and flash if this is the case for you.
In order to build the application using west, some environment variables need to be set, specifically the toolchain and the Zephyr base. Even though you're not using the TM and are building any of the samples or applications directly in the nrf folder it is not necessary to set the Zephyr base, since the CMake tool will locate it automatically when building your application through west build
. However, in this tutorial, we keep the project detached from the SDK and the environmental variable ZEPHYR_BASE
needs to be set (if NCS is installed manually).
If you follow the NCS documentation you'll be told to set environmental variables through the cmd file zephyr-env.cmd. This is a nice way of doing it if you are working with different projects, and don't want the env. variables to interfere, but you'll have to run the cmd file every time you open the command line.
For this tutorial, let's add the toolchain environmental variables to the whole system instead. Follow these steps:
setx -m ZEPHYR_TOOLCHAIN_VARIANT "gnuarmemb"
setx -m GNUARMEMB_TOOLCHAIN_PATH c:\gnuarmemb
C:\WINDOWS\system32>setx -m ZEPHYR_TOOLCHAIN_VARIANT "gnuarmemb" && setx -m GNUARMEMB_TOOLCHAIN_PATH c:\gnuarmemb
SUCCESS: Specified value was saved.
SUCCESS: Specified value was saved.
C:\my_projects\hello_world>echo %GNUARMEMB_TOOLCHAIN_PATH% & echo %ZEPHYR_TOOLCHAIN_VARIANT%
c:\gnuarmemb
gnuarmemb
Let's set the Zephyr base.
set ZEPHYR_BASE=C:\Nordic_SDK\ncs\zephyr
(change the path if you have put the nRF Connect SDK somewhere else)C:\my_projects\hello_world>set ZEPHYR_BASE=C:\Nordic_SDK\ncs\zephyr
echo %ZEPHYR_BASE%
C:\my_projects\hello_world>echo %ZEPHYR_BASE%
C:\Nordic_SDK\ncs\zephyr
There are mainly two different ways of setting ZEPHYR BASE
, by using set
or setx
. The former one (set
), which is used above, will set it temporary and immediately in the current shell and the latter (setx
), which was used when setting the toolchain env. variables, will set it permanently with the need to reopen the shell. Pick the option that suits you best.
Let's get the sample running.
It is important to open the shell (either bash or the command prompt) from the Toolchain Manager when building your project, since all the tools from C:\Users\<user_name>\ncs\v1.4.0\toolchain will only be accessible then.
cd C:/my_projects/hello_world
. Use another path if you've placed your sample somewhere else.west build -b <board_variant>
to build the application with west:... MINGW64 /c/my_projects/hello_world $ west build -b <board_variant> -- west build: build configuration: source directory: C:\my_projects\hello_world build directory: C:\my_projects\hello_world\build (created) . . .
If the build process was successful, you should see a folder named build inside the hello_world folder:
The build folder should contain the following files and folders (it may vary slightly depending on the choice of <board_variant>):
If you are using the nRF9160DK, make sure SW5 is set to nRF91
west flash
, either from the project folder or the build directory.
The nRF9160 SiP incorporates an LTE modem, which makes it possible to run cellular IoT applications. The modem always operates in a non-secure domain (read more about this in External domain access control on the infocenter), and the application must do the same in order to communicate with it. Therefore, all the samples in <..>/nrf/samples/nrf9160 as well as the asset_tracker and the serial_lte_modem, have to be built as nonsecure in order to work.
To build a sample as non-secure, the board nrf9160dk_nrf9160ns (corresponds to <board_variant>) must be used. When building an application using this board, the following configurations will be set:
→ CONFIG_TRUSTED_EXECUTION_NONSECURE
: This will make the application non-secure and make it run in the non-secure domain
→ CONFIG_SPM
: This will build the SPM sample.
To get a deeper understanding of secure/non-secure environments, jump forward to section 1.6 Secure vs. nonsecure in part 2 of this tutorial series.
After the build process is complete, the file merged.hex (<..>nrf\samples\nrf9160\<sample>\build\zephyr\merged.hex) is generated, which contains both the SPM sample in addition to the application. When running west flash
, this hex file will be chosen and get programmed onto the chip. Similarly in SES, if you program the chip through Build→Build Solution, the merged.hex file will get flashed.
Let's try to build the AT Client sample on the nRF9160 DK using SES. It is demonstrated how to build this sample in the NCS documentation, but for the record, let's do it again:
Your ncs location may differ, open the Toolchain Manager app and click on settings to see the location.
*** Booting Zephyr OS build v2.3.0-rc1-ncs1 ***
Flash regions Domain Permissions
00 00 0x00000 0x08000 Secure rwxl
01 31 0x08000 0x100000 Non-Secure rwxl
.
.
.
24 NRF_GPIOTE1 Non-Secure OK
25 NRF_REGULATORS Non-Secure OK
SPM: NS image at 0xc000
SPM: NS MSP at 0x20027aa8
SPM: NS reset vector at 0xd3d9
SPM: prepare to jump to Non-Secure image.
*** Booting Zephyr OS build v2.3.0-rc1-ncs1 ***
The AT host sample started
Check out the section AT Client→Building and running for more generic explanations and instructions on how to test the example to see if everything works as expected.
If you would like to use west instead, follow the steps below
west build -b nrf9160dk_nrf9160ns
west flash
→ If you are encountering any errors, you may have to update the modem firmware to the latest version, check out nRF9160 DK Getting Started→Updating the modem firmware for instructions on how to do this.
→ If you have problems connecting the nRF9160 to nRF Connect for Cloud, you may have to update the nRF for Cloud certificates, check out nRF9160 DK Getting Started→Updating the nRF Connect for Cloud certificate for guidance on how this can be achieved.
→ Be aware that PSM and/or eDRX is not supported on all networks yet, or that some networks do not allow roaming SIM cards to use those features. As a consequence, applications such as the Asset Tracker won't work correctly.
→https://webinars.nordicsemi.com/videos-1
The nRF5340 SoC incorporates two processors, the network core and the application core. The BLE controller has to run on the network core while the BLE host usually runs on the application core in addition to the application (Read more about the BLE Stack Architecture and the BLE Layers in the Zephyr documentation). As mentioned in the NCS documentation, Working with nRF5340, the network core has to be programmed with the Bluetooth: HCI RPMsg sample in order to run a BLE sample on the nRF5340. If that is done, the application core can be programmed with any of the samples in <..>\nrf\samples\bluetooth.
On the ncs master branch, it is not necessary to build and flash firmware for both of the cores. All you need to do is to build and flash a BLE application from nrf/samples/bluetooth onto the application core, and the network core will automatically get programmed with the required firmware.
Let's demonstrate how to build the BLE Peripheral LBS sample on the nRF5340 DK using SES:
Your ncs location may differ, open the Toolchain Manager app and click on settings to see the location.
Next we need to program the HCI RPMsg sample. When building BLE samples in NCS v1.4.0 with the nRF53, the hci_rpmsg sample will automatically get added as a child image. You can see the generated build output for the hci_rpmsg sample inside nrf\samples\bluetooth\peripheral_lbs\build_nrf5340dk_nrf5340_cpuapp\hci_rpmsg. In order for the BLE sample to function properly on the nRF53 the hex file <build folder>\zephyr\merged.hex has to get programmed onto the application core and the hex file <build folder>hci_rpmsg\zephyr\merged_CPUNET.hex has to get programmed onto the network core. However, SES V5.10d will only flash merged.hex onto the application core, and the other hex file needs to get flashed manually using nrfjprog. Follow the steps below to achieve this:
cd nrf\samples\bluetooth\peripheral_lbs\build_nrf5340dk_nrf5340_cpuapp\hci_rpmsg\zephyr
nrfjprog -f NRF53 --coprocessor CP_NETWORK --eraseall && nrfjprog -f NRF53 --coprocessor CP_NETWORK --program merged_CPUNET.hex
nrfjprog --pinreset
If you would like to use west instead, follow the steps below
west build -b nrf5340dk_nrf5340_cpuapp && west flash
→ NUS on NCS: The Nordic UART Service with the nRF Connect SDK
→ NCS BLE tutorial part 1: Custom Service in Peripheral role
If you have any questions regarding this tutorial, please create a ticket in DevZone.
→ Check out the next part of this tutorial series: nRF Connect SDK Tutorial - Part 2