README.md

Hey!

This script will install gpgpusim on your system.
Currently it can only install gpgpusim with cuda 9 but maybe if i can find some spare time in future i will add cuda 11.x to it as well.

What it does

1. Bootstrap debian 7.
2. Install GCC 5 on it.
3. Install CUDA 9 on it.
4. Install gpgpusim on it.
5. Generate some scripts for ease of use.

Requirements

Debian/Ubuntu

sudo apt install debootstrap make gcc git wget

Arch

sudo pacman -S debootstrap make gcc git wget

Fedora

dnf install debootstrap make gcc git wget

Build

git clone https://github.com/thegraveyardwarden/gpgpusim-install.git && cd gpgpusim-install && make

How to install

usage: bin/gpgpusim-install COMMAND [OPTIONS]

COMMANDS:
         reinstall: restart the operation and ignore checkpoints
         install: start installation

OPTIONS
         -g: path to gcc 5 tar file (if not present it will be downloaded)
         -c: path to cuda 9 run file (if not present it will be downloaded)
         -d: path to install debian (if not present /opt/wheezy)

So if it's your first time trying to install you probably don't have cuda 9 runfile or gcc 5 tar file. just run:

bin/gpgpusim-install install

It will download them for you.

When installer wants to install cuda it will ask you some questions:

1. Installing cuda 9...Do you accept the previously read EULA? -> y
2. You are attempting to install on an unsupported configuration. Do you wish to continue? -> y
3. Install NVIDIA Accelerated Graphics Driver for Linux-x86_64 384.81? -> n
4. Install the CUDA 9.0 Toolkit? -> y
5. Enter Toolkit Location -> leave blank for default
6. Do you want to install a symbolic link at /usr/local/cuda? -> y
7. Install the CUDA 9.0 Samples? -> y
8. Enter CUDA Samples Location -> /opt

This is my prefrence. Change them if you want.

How to use

Now you can run this command when you want to chroot to debian and use gpgpusim:

gpgpusim-use

Ctrl+D pops you out into your own os.
gpgpusim files are under root directory.