ci: Add CI to the qemu-minimal project

Add two workflows to the GitHub CI flow for both spell-checking and
smoke-testing. The smoke-test does a simple VM creation via the
QEMU-based approach.

Fixes #20.

.github/workflows/smoke-test.yml

@@ -0,0 +1,25 @@
+name: qemu-minimal-smoke-test
+
+on:
+ pull_request:
+ branches: [ master ]
+
+jobs:
+ ubuntu-smoke-test:
+ name: ubuntu-smoke-test
+ runs-on: ubuntu-latest
+ steps:
+ - name: Check out code using action/checkout
+ uses: actions/checkout@v4.1.1
+ - name: Install a couple of packages via the apt package manager
+ run: sudo apt update && sudo apt install -y qemu-system-x86 cloud-image-utils
+ - name: Generate an ssh key-pair for this test
+ run: ssh-keygen -t rsa -q -f "$HOME/.ssh/id_rsa" -N ""
+ - name: Run a qemu-based ./gen-vm smoke-test
+ run: ./gen-vm
+ working-directory: qemu
+ env:
+ VM_NAME: qemu-minimal-smoke-test
+ KVM: none
+ USERNAME: ubuntu
+ PASS: password

.github/workflows/spell-check.yml

@@ -0,0 +1,19 @@
+name: qemu-minimal-spell-check
+
+on:
+ pull_request:
+ branches: [ master ]
+ paths:
+ - '**.md'
+ - '.wordlist.txt'
+ - '.spellcheck.yaml'
+
+jobs:
+ ubuntu-spell-check:
+ name: ubuntu-spell-check
+ runs-on: ubuntu-latest
+ steps:
+ - name: Check out code.
+ uses: actions/checkout@v4.1.1
+ - name: GitHub Spellcheck Action
+ uses: rojopolis/spellcheck-github-actions@v0

.spellcheck.yaml

@@ -0,0 +1,16 @@
+spellchecker: aspell
+matrix:
+- name: markdown
+ aspell:
+ lang: en
+ sources:
+ - '**/*.md'
+ pipeline:
+ - pyspelling.filters.markdown
+ dictionary:
+ wordlists:
+ - .wordlist.txt
+ output: wordlist.dic
+ encoding: utf-8
+ default_encoding: utf-8
+ expect_match: true

.wordlist.txt

@@ -0,0 +1,124 @@
+Ansible
+Bjorling's
+Busybox
+CMB
+CONFIG
+IPv
+Initramfs
+KVM
+LFS
+LightNVM
+Matias
+NVM
+NVMe
+NVMf
+OpenChannel
+OpenChannelSSD
+PMEM
+PRs
+QEMU
+QEMU's
+RoCE
+SSD
+SSDs
+Thie
+VM
+VMs
+VirtFS
+Virtfs
+addr
+alot
+ansible
+awhisten
+batesste
+bbs
+blogspot
+bss
+busybox
+busyboz
+bzImage
+bzimage
+cd
+chroot
+config
+cpio
+csortu
+currenly
+deboostrap
+dev
+differntly
+environemnt
+eth
+exe
+filesystem
+fs
+gdb
+github
+gz
+href
+html
+http
+https
+img
+infradead
+init
+initramfs
+initrd
+intel
+intresting
+jessie
+kbusch
+ko
+lfs
+li
+librxe
+libs
+libvirt
+linux
+localhost
+lt
+managable
+mnt
+modprobe
+mon
+multistrap
+nbd
+netdev
+nographic
+ntap
+nvme
+nvmef
+nvmf
+ol
+passthrough
+perfers
+powerdown
+qcow
+qemu
+queitly
+rc
+rdma
+repo
+roce
+rootfs
+runqemu
+rxe
+sbates
+seperate
+snapshotted
+stackoverflow
+sudo
+symlink
+sys
+targetted
+ttyS
+ubuntu
+usr
+virt
+virtfs
+virtio
+vm
+vmlinux
+wget
+www
+zerofree

README.md

@@ -14,7 +14,7 @@ be used as a starting point for building a suitable kernel.
 To run qemu using this image from the command should be:
 
 ```
-./runqemu <path_to_bzimage>
+./runqemu <path_to_bzImage>
 ```
 
 You can run
@@ -35,14 +35,14 @@ various stages of repair.
 
 A QEMU-only (no libvirt) way to generate new VMs. Uses the Ubuntu
 cloud image ISO and a cloud-init script. Note there is an issue that
-the files for this cannot reside on a guest's virtfs (i.e. running
+the files for this cannot reside on a guest's VirtFS (i.e. running
 this script inside a VM to generate a nested VM won't work due to file
 permission issues). This may be better with virtio-fs.
 
 ### ./libvirt/virt-install-ubuntu
 
 A libvirt-based script that uses cloud images and cloud-init and
-virt-install. Thie seems to work quite well on bare-metal.
+virt-install. This seems to work quite well on bare-metal.
 
 ### ./scripts/create
 
@@ -92,7 +92,7 @@ second drive has a Controller Memory Buffer (CMB) advertised on it.
 
 Note that since upstream has a different level of support for NVMe
 drive options than the other forks we use a -u option to handle that
-differntly.
+differently.
 
 The host's /home file is also passthrough mounted to the guests /home
 directory so test scripts, etc can be stored and run directly from the
@@ -181,7 +181,7 @@ following:
 For a more intresting initramfs example you can run busybox. To do
 this perform the following steps.
 
- 1. Download the busyboz source and build it to include all the tools
+ 1. Download the busybox source and build it to include all the tools
  that you want. Make sure this is a statically linked executable.
  2. cd scripts
  3. ./busybox <path to busybox exe>
@@ -194,7 +194,7 @@ This should boot into busybox shell and you can execute your installed
 command from there. Enjoy! Note this will not then pass on to a
 subsequent root filesystem (yet).
 
-## Virtfs
+## VirtFS
 
 By default we map the /home folder on the host to the /home folder on
 the guest using Plan 9 folder sharing over VirtFS. However this
@@ -229,7 +229,7 @@ folders and the jessie image needs updating with some of the libs
 
 For a few reasons it is easier to keep the rdma_rxe.ko module seperate
 to the kernel. Also you need the linux headers inside the VM root
-filesystem to compile librxe. You currenly need to copy (or symlink)
+filesystem to compile librxe. You currently need to copy (or symlink)
 /usr/lib64/* to /usr/lib. You also have to setup VM to VM networking
 for which I (for now) used -netdev socket. QEMU perfers ntap but
 that's alot more work. The IPv4 addresses on the eth0 interfaces of