1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
|
# pkg-rocm-tools
This builds a number of packages with utilities that facilitate working with
GPUs in environments isolated from the host system.
This currently builds three packages:
- **pkg-rocm-tools**
- **rocm-qemu-support**
- **rocm-podman-support**
## pkg-rocm-tools
These are utilities and files meant to assist Debian contributors in the
process of adding support for ROCm in their packages:
- **rocm-test-launcher**
Launches tests making use of AMD GPUs, and optionally exports system
information useful for debugging GPU issues. Use this to skip tests on
[ci.debian.net](https://ci.debian.net) (no GPUs), rather than fail them.
- **rocm-target-arch**
Package build helper for determining ROCm target architectures for a
particular distribution. Use this utility to get a list of target
architectures for the distribution you are building for, rather than
hard-coding a list in `debian/rules`.
During development or in Salsa CI, the `ROCM_TARGET_ARCH_FIXED` variable
can be used to override this, for example to just your local GPU.
The targets built for should also be recorded in a `X-ROCm-Target-GPUs`
variable in `debian/control`.
This is all very easy to implement, see rocrand
[26670b40](https://salsa.debian.org/rocm-team/rocrand/-/commit/26670b408f0658f080de1595e9f06dc885602994).
## rocm-qemu-support
These are utilities for working with ROCm inside QEMU VMs:
- **rocm-qemu-setup**
Prepare the system for GPU pass-through
- **rocm-qemu-create**
Build QEMU images with GPU pass-through support
- **rocm-qemu-run**
Boot a VM with GPU pass-through
- **autopkgtest-virt-qemu+rocm**
autopkgtest backend that extends the regular backend with automatic setup of
GPU pass-through
## rocm-podman-support
These are utilities for working with ROCm inside Podman containers:
- **rocm-podman-setup**
Prepare the system for GPU-in-container use
- **rocm-podman-create**
Build podman images with GPU support
- **rocm-podman-run**
Run a command in a container with GPU support
- **autopkgtest-virt-podman+rocm**
autopkgtest backend that extends the regular backend with automatic setup
of devices and subuid mappings.
## QEMU Examples
```shell
# Run setup to determine what changes your system needs
$ rocm-qemu-setup
# Alternatively, if you want to run as a user
$ rocm-qemu-setup -u somename
# Create default image for unstable
$ sudo rocm-qemu-create /var/tmp/unstable.img
# Live above, but include an SSH key
$ sudo rocm-qemu-create -a /path/to/authorized_keys
# Create image for bookworm, using a fast local mirror (eg APT cache)
$ sudo rocm-qemu-create -r bookworm -m http://10.1.2.3:9999/debian
# If you want a regular user to run the image
$ sudo chown <user>: /var/tmp/unstable.img
# Run rocrand's autopkgtests (from the official archive) in that image
$ autopkgtest -B rocrand -- qemu+rocm /var/tmp/unstable.img
# By default, the qemu+rocm backend will claim 75% of host cores and RAM. Use
# --cpus and --ram-size to adjust:
$ autopkgtest -B rocrand -- qemu+rocm --cpus 4 --ram-size 16384 /var/tmp/unstable.img
# Run rocrand's autopkgtest, passing in only the GPU at 2b:00.0
$ autopkgtest -B rocrand -- qemu+rocm --gpu 2b:00.0 /var/tmp/unstable.img
# Boot into the image if you need to, here with a shared directory
$ rocm-qemu-run -d /dir/on/host /var/tmp/unstable.img
```
## PODMAN Examples (running podman without sudo)
```shell
# Run setup to determine what changes your system needs
$ rocm-podman-setup
# Create default image for unstable (Will create rocm/debian:unstable podman image)
$ rocm-podman-create
# Run rocminfo's autopkgtest
$ autopkgtest -B rocminfo -- podman+rocm rocm/debian:unstable
```
|
