The goal of bee is to establish an important framework for silicon validation.
Bee is a simple silicon validation framework at OS level, it has some advantages:
- It is small to run on a low-frequency chip;
- It is very easy to add a new case in Bee;
- You can run a single case, a case set, or all cases, which means that you can use Bee to verify Unit, IP, and Kit in silicon validation.
The Bee framework uses the recursive compilation build system, mainly involving bee_kernel_case, bee_user_case, bee_tools, etc. Users only need to edit the Makefile based on the template to add their test cases to Bee framework. Details can be found at bee_add_case;
Bee includes kernel cases and user space cases, kernel cases are executed asynchronously using the Munit module, and user-space cases are executed separately according to the shell script. Each case feature is a minimum execution unit.
enter the directory of Bee:
-
Find the help:
./build.sh -h
-
Compile the bee:
./build.sh -m /lib/modules/`uname -r`/build
-
Install the bee:
-
Install to the DEFAULT path in /Your path/Bee/build
./build.sh -i ""
-
Install the bee to another path eg. /Your path/
./build.sh -i "/Your path"
-
Clean the bee:
./build.sh -c
-
After compiling, you can run the cases by Bee;
-
Export Bee root path:
export BEETOPDIR=/Your Path/build
cd /Your Path && ls
you can see the files: bin ko lib64
-
Run the case by bt cmd:
cd /Your Path/bin/
-
Get help:
./bt -h
-
Insmod a kernel case driver module in /your path/ko
./bt -i modules.ko
-
You can get a cases list
./bt -l
-
You can run a case for kernel case:
./bt -k case_name_01
-
You can run all cases, now you don’t need to run bt -i to insmod the case ko:
./bt -a