IPC Shared Memory Mutex using Spinlock

Table of Contents

• Introduction
• Supported Combinations
• Steps to Run the Example
• See Also
• Sample Output

Introduction

This example shows shared memory access between multiple cores in an atomic way using spinlock. Shared memory is allocated across two cores - using gUserSharedMem variable.

In this example, this memory is just a counter for illustration purpose. Each core does a read-modify write to this variable in a continuous manner.

Spinlocks are used to protect this shared memory access. This ensures that the access is atomic and the shared variable is incremented properly. If proper protection is done, then at the end of the iteration the counter value will be N times loop count where N is number of cores acting on the variable.

When iteration count reaches SPINLOCK_TEST_LOOPCNT, the example is completed.

Supported Combinations

Parameter	Value
CPU + OS	r5fss0-0 freertos
	r5fss0-1 nortos
Toolchain	ti-arm-clang
Board	am64x-evm
Example folder	examples/drivers/ipc/ipc_spinlock_sharedmem

Steps to Run the Example

Note

This is a system or multi-core project, so refer to system project build instructions for CCS project or makefiles when building the example.

• When using CCS projects to build, import the system CCS project and build it using the CCS project menu (see Using SDK with CCS Projects). This will build all the dependant CPU projects as well
• When using makefiles to build, build the system makefile using make command (see Using SDK with Makefiles). This will build all the dependant CPU makefiles as well.
• Launch a CCS debug session and run the executables, see CCS Launch, Load and Run
• This is a multi-core example. Hence the executables should be loaded and run for all the above mentioned cores
• The application has a sync mechanism at the start which waits for all cores to start before doing the test. Hence the cores can be loaded and run in any sequence.

See Also

SPINLOCK

Sample Output

Shown below is a sample output when the application is run,

[IPC Spinlock Sharedmem] Example started ...
Waiting for all cores to start ...
Waiting for all cores to complete ...
[r5f0-1]     0.000016s : [IPC Spinlock Sharedmem] Example started ...
All tests have passed!!
[r5f0-1]     0.000036s : Waiting for all cores to start ...
[r5f0-1]     0.199144s : Waiting for all cores to complete ...
[r5f0-1]     0.199157s : All tests have passed!!