README.docker

To run the example in docker, do the following

Download the docker install file for your system

For the pre-built docker image, run (future build)

   docker run -p 4000:80 -it --entrypoint /bin/bash essentialsofparallelcomputing/chapter13

Alternatively, build an image locally and run it 

   docker build -t chapter13 .

   docker run -it --entrypoint /bin/bash chapter13

To run the docker image with a graphical user interface (GUI) on Mac OSx

install XQuartz -- brew cask install xQuartz

start up xQuartz (may have to start up an app with a right click) and then in the menu bar at the top of the screen, 
select XQuartz->Preferences. Go to Security tab and add allow connections from network clients

reboot Mac system

Start up xQuartz

Start up docker image with

docker run -it -e DISPLAY=host.docker.internal:0 -v /tmp/.X11-unix:/tmp/.X11-unix --entrypoint /bin/bash chapter13

Now run the GUI with

nvvp 

Other applications such as CodeXL may still have difficulties

Getting access to the Nvidia GPUs is done with the --gpus option to Docker. This feature was added with 
the 19.0.3 version of Docker. For Intel GPUs, try --device=/dev/dri. For AMD GPUs, try

docker run -it --device=/dev/kfd --device=/dev/dri --security-opt seccomp=unconfined

For Intel Advisor, Inspector or VTune, add --cap-add=SYS_ADMIN --cap-add=SYS_PTRACE to the docker run command.