How the heck do these examples even compile?!?!?

Hi Steve,

Steve said:

I'm looking at a couple of IPC examples under ipc_3_40_01_08\examples\OMAPL138_linux_elf\ex02_messageQ and pc_3_40_01_08\examples\OMAPL138_linux_elf\ex11_ping and ... neither one of them has a .cfg file under the host subdir. The corresponding dsp dir does.  Looking at the source code in detail I see calls that would seem to require a .cfg file for the app to even build... What gives, or am I missing something basic???

This example has to be built using the command line. Able to built by following the readme.txt provided in the ex11_ping folder and generate the binaries.

I have attached the readme.txt here too. Follow this procedure and revert if have any questions on building. Before you build the IPC examples, you have to build the IPC once.

    #
    #  ======== readme.txt ========
    #

    ping - Send a message between all cores in the system

    Overview
    =========================================================================
    This example is used to exercise every communication path between all
    processors in the system (including local delivery on the current
    processor). The example is also organized in a suitable manner to develop
    an application with different compute units on each processor.

    Each executable will create two tasks: 1) the server task, and 2) the
    application task. The server task creates a message queue and then waits
    on that queue for incoming messages. When a message is received, the
    server task simply sends it back to the original sender.

    The application task creates its own message queue and then opens every
    server message queue in the system (including the server queue on the
    local processor). The task sends a message to a server and waits for the
    message to be returned. This is repeated for each server in the system
    (including the local server).


    Build Instructions
    =========================================================================

     1. Setup a development area. A typical setup might look like this. The
        Depot folder contains installed products. The work folder contains
        each of the examples (eg. hello, ping, etc.). Each example contains
        its own products.mak file which will include a parent products.mak
        file if it exists. The parent file must be created by you.

        testbench/
         |_ Depot/
         |   |_ bios_m_mm_pp_bb/
         |   |_ ipc_m_mm_pp_bb/
         |   |_ ti_c6x_m_m_p/
         |
         |_ work/
             |_ ping/
             |   |_ core0/
             |   |_ core1/
             |   |_ core2/
             |   |_ core3/
             |   |_ core4/
             |   |_ core5/
             |   |_ core6/
             |   |_ core7/
             |   |_ makefile
             |   |_ products.mak
             |   |_ readme.txt
             |
             |_ products.mak

        In the ping example, there is a top-level makefile which simply builds
        each of the lower directories. There is a directory for each processor
        named by the IPC processor name. The folder named 'shared' contains
        common files used by all executables, such as config.bld which defines
        the memory map.

     2. Unpack the ping.zip file into the work folder.

        cd work
        unzip <...>/ping.zip

     3. Setup the build environment. Edit products.mak and set the install paths
        as defined by your physical development area. Each example has its own
        products.mak file; you may also create a products.mak file in the parent
        directory which will be used by all examples.

        DEPOT = /testbench/Depot

        BIOS_INSTALL_DIR                = $(DEPOT)/bios_m_mm_pp_bb
        IPC_INSTALL_DIR                 = $(DEPOT)/ipc_m_mm_pp_bb
        XDC_INSTALL_DIR                 = $(DEPOT)/xdctools_m_mm_pp_bb
        ti.targets.elf.C66              = $(DEPOT)/ti_c6x_m_m_p

     4. Build the example. This will build only debug versions of the executables.
        Edit the lower makefiles and uncomment the release goals to build both
        debug and release executables.

        cd ping
        make

        Look in the following folders for the generated files. All executables
        are copied to these folders for convenience; it makes it easier when
        loading them into CCS.

        install/ping/debug/
        install/ping/release/

     5. Optional. Use the install goal to copy the executables to a specific
        folder of your choice.

        cd ping
        make install EXEC_DIR=/<full pathname>/<platform>

        The executables are copied to the following folders.

        /<full pathname>/<platform>/ping/debug
        /<full pathname>/<platform>/ping/release


    Build Configuration
    =========================================================================

    This example can be built for a subset of all available processors. For
    example, you might want to build and run this example for the following
    processors: CORE0 and CORE1. Use the following instructions to configure
    this example for a subset of processors.

    Note: you must always run the set of processors you configured and built.
    In this example, you must run the CORE0 and CORE1 executables; you
    cannot run just one of them. Because the IPC has been configured for
    ProcSync_ALL, all processors must participate in the call to Ipc_start().

     1. Specify the processor list in the top-level makefile.

        edit ping/makefile

        Edit the list of processors in the PROCLIST variable.

        PROCLIST = core0 core1

     2. Clean and rebuild the example.

        cd ping
        make clean
        make

        Look in the following folders for the generated files.

        ping/install/ping/debug
        ping/install/ping/release