Skip to content
This repository has been archived by the owner on Jun 24, 2022. It is now read-only.

Latest commit

 

History

History
93 lines (81 loc) · 4.41 KB

install.creole

File metadata and controls

93 lines (81 loc) · 4.41 KB
Raw

Installation

GNU Tools for ARM Embedded Processors

You need to first install the GNU toolchain from the developer.arm.com site. You should select the version that matches the gcc4mbed project you are trying to use. The installation directory containing the arm-none-eabi-* tools should be added to your operating system's PATH. NOTE: Using the install scripts in the root of the gcc4mbed project take care of these downloads for you and this is the supported method of acquiring the necessary GCC toolchain.

Where is GCC4MBED?

The gcc4mbed code is available on github at https://github.com/adamgreen/gcc4mbed. You can click the Clone or download button (found at the top of the source file list on the right hand side) on this page to popup a dialog allowing you to download a compressed archive of the gcc4mbed source code to your computer.

Decompress this archive on your computer. For the rest of the steps in this tutorial, we will assume that it was placed in /gcc4mbed on a *nix based system or c:\gcc4mbed on a Windows system.

Build and deploy

You are now ready to try building and deploying GCC built mbed samples.

Build

*nix:
    cd /gcc4mbed
    make
Windows:
    cd c:\gcc4mbed
    external\win32\make

This will have built various programs in the samples directory off of the root of the gcc4mbed project.

./gcc4mbed/samples/BLEHeartRate/
./gcc4mbed/samples/Blink/
./gcc4mbed/samples/FileTest/
./gcc4mbed/samples/HelloWorld/
./gcc4mbed/samples/SdPerf/
./gcc4mbed/samples/StdIO/
./gcc4mbed/samples/TCPSocket_HelloWorld/
./gcc4mbed/samples/Ticker/
./gcc4mbed/samples/USBMouse/

Deploy

While we could manually copy the binary up to the mbed device, it would be nice if it could be done for us at the end of the build process. The only issue is that different environments will require different deployment steps. To abstract away this environment dependency, the makefiles use an environment variable called GCC4MBED_DEPLOY for specifying how the deployment should take place. The user should set this GCC4MBED_DEPLOY environment variable to something appropriate for their environment. On *nix it could be exported from something like the ~/.profile shell script and on Windows it could be set in a batch file used for development environment configuration or added to the System Properties.

GCC4MBED_DEPLOY Examples

The author uses this in the ~/.profile script on his Mac where the MBED device mounts as /Volumes/MBED:

export GCC4MBED_DEPLOY='cp PROJECT.bin /Volumes/MBED/'

The author uses this in a batch file on his Windows XP machine where the MBED device mounts as the F: drive:

set GCC4MBED_DEPLOY=copy PROJECT.bin f:\

The following example might be useful if you were deploying to a LPCXpresso using the lpc21isp tool:

export GCC4MBED_DEPLOY='lpc21isp PROJECT.hex /dev/ttyUSB0 115200 14746'

It should be noted that the PROJECT identifier in the above examples has a special meaning. This identifier will be replaced by the makefile to match the name of the program currently being built.

At this point, you want to set the GCC4MBED_DEPLOY variable to something appropriate for your environment.

Deploy HelloWorld

After all of the that work, wouldn't it be great to see something actually run on the mbed device? Let's start with the HelloWorld sample which is probably the same program you first tried with the cloud compiler.

*nix:
    cd /gcc4mbed/samples/HelloWorld
    make clean all deploy
Windows:
    cd c:\gcc4mbed\samples\HelloWorld
    external\win32\make clean all deploy

Those build rules (clean all deploy) will force all of the output files to be deleted, then rebuilt from a clean start, and finally copied up to your mbed-LPC1768 device. Pressing the reset button on the mbed device should result in the blinking LED.

If you just want to deploy a binary that has already been built, you can just run "make deploy"

Note: If you are using a device other than a mbed-LPC1768 then you need to use a deploy rule specific to that target. For example if you were using a K64F board, then you would want to run "make K64F-deploy" to have the K64F binaries built and deployed to the device.

Uninstall

All of the files required by gcc4mbed, including the GNU Tools for ARM Embedded Processors, are installed under the same directory. To uninstall you just need to delete the gcc4mbed directory.