Commit Graph

72 Commits

Author SHA1 Message Date
Evan Lezar
5145b0a4b6 Revert "Merge pull request #694 from elezar/add-opt-in-to-sockets"
This reverts commit b061446694, reversing
changes made to c490baab63.

Signed-off-by: Evan Lezar <elezar@nvidia.com>
2024-09-20 20:26:45 +02:00
Evan Lezar
a4bfccc3fe Use include-persistenced-socket feature for CDI mode
This change ensures that the internal CDI representation includes
the persistenced socket if the include-persistenced-socket feature
flag is enabled.

Signed-off-by: Evan Lezar <elezar@nvidia.com>
2024-09-18 22:30:27 +02:00
Evan Lezar
ba1ed3232f Skip injection of nvidia-persistenced socket by default
This changes skips the injection of the nvidia-persistenced socket by
default.

An include-persistenced-socket feature flag is added to allow the
injection of this socket to be explicitly requested.

Signed-off-by: Evan Lezar <elezar@nvidia.com>
2024-09-18 22:10:09 +02:00
Avi Deitcher
179d8655f9 Move nvidia-ctk hook command into own binary
This change creates an nvidia-cdi-hook binary for implementing
CDI hooks. This allows for these hooks to be separated from the
nvidia-ctk command which may, for example, require libnvidia-ml
to support other functionality.

The nvidia-ctk hook subcommand is maintained as an alias for the
time being to allow for existing CDI specifications referring to
this path to work as expected.

Signed-off-by: Avi Deitcher <avi@deitcher.net>
2024-05-21 12:19:44 +02:00
Evan Lezar
011c658945 Create root.Driver instance at first usage
This allows for testing through injection of the driver root.

Signed-off-by: Evan Lezar <elezar@nvidia.com>
2024-04-03 15:03:33 +02:00
Evan Lezar
09341a0934 Add support for feature flags
This change adds a features config that allows
individual features to be toggled at a global level. Each feature can (by default)
be controlled by an environment variable.

The GDS, MOFED, NVSWITCH, and GDRCOPY features are examples of such features.

Signed-off-by: Evan Lezar <elezar@nvidia.com>
2024-04-03 11:58:37 +02:00
Evan Lezar
93763d25f0 Use functional options to construct driver root
Signed-off-by: Evan Lezar <elezar@nvidia.com>
2024-03-25 20:13:25 +02:00
Christopher Desiniotis
55097b3d7d Add a new gated modifier for GDRCopy which injects the gdrdrv device node
Signed-off-by: Christopher Desiniotis <cdesiniotis@nvidia.com>
2024-01-24 14:25:58 -08:00
Christopher Desiniotis
32c3bd1ded Fallback to standard CDI modifier when creation of automatic CDI modifier fails
Signed-off-by: Christopher Desiniotis <cdesiniotis@nvidia.com>
2023-12-06 09:02:19 -08:00
Christopher Desiniotis
b9ac54b922 Add GetDeviceSpecsByID() API to the nvcdi Interface
Signed-off-by: Christopher Desiniotis <cdesiniotis@nvidia.com>
2023-12-06 09:02:19 -08:00
Christopher Desiniotis
ae1b7e126c Extend the 'runtime.nvidia.com/gpu' CDI device kind to support full-GPUs specified by index or UUID
Signed-off-by: Christopher Desiniotis <cdesiniotis@nvidia.com>
2023-12-06 09:02:19 -08:00
Tariq Ibrahim
7627d48a5c run goimports -local against the entire codebase
Signed-off-by: Tariq Ibrahim <tibrahim@nvidia.com>
Signed-off-by: Evan Lezar <elezar@nvidia.com>
2023-12-01 11:13:17 +01:00
Evan Lezar
efae501834 Add support for injecting NVSWITCH devices
This change adds support for an NVIDIA_NVSWITCH environment variable.
When set to `enabled` this striggers the injection of all available
/dev/nvidia-nvswitch* device nodes.

Signed-off-by: Evan Lezar <elezar@nvidia.com>
2023-11-22 21:59:39 +01:00
Evan Lezar
3045954cd9 Consolidate GDS and MOFED modifiers
Signed-off-by: Evan Lezar <elezar@nvidia.com>
2023-11-22 21:59:17 +01:00
Evan Lezar
bbd9222206 Add driver root abstraction
This change adds a driver root abstraction that defines how
libraries are located relative to the root. This allows for
this driver root to be constructed once and passed to discovery
code.

Signed-off-by: Evan Lezar <elezar@nvidia.com>
2023-11-22 13:27:48 +01:00
Evan Lezar
039d7fd324 Merge branch 'remove-config-import-from-discover' into 'main'
Remove NewGraphicsDiscoverer API simplification

See merge request nvidia/container-toolkit/container-toolkit!498
2023-11-20 22:52:02 +00:00
Evan Lezar
255181a5ff Rename NewGraphicsDiscoverer as NewDRMNodesDiscoverer
This change renames NewGraphicsDiscoverer to NewDRMNodesDiscoverer and
instead calls NewGraphicsMountsDiscoverer explicitly when constructing
a graphics modifier.

This avoids the import of config.Config into the discover package
which leads to a transitive dependency on toml-specifics and
requires that the vendor/github.com/pelletier/ package
be vendored in to consumers.

Signed-off-by: Evan Lezar <elezar@nvidia.com>
2023-11-20 23:10:57 +01:00
Christopher Desiniotis
dc36ea76e8 Automatically generate CDI spec for the runtime.nvidia.com/gpu=all device
Signed-off-by: Christopher Desiniotis <cdesiniotis@nvidia.com>
2023-11-20 13:35:07 -08:00
Evan Lezar
d4e21fdd10 Add devRoot option to CDI api
A driverRoot defines both the driver library root and the
root for device nodes. In the case of preinstalled drivers or
the driver container, these are equal, but in cases such as GKE
they do not match. In this case, drivers are extracted to a folder
and devices exist at the root /.

The changes here add a devRoot option to the nvcdi API that allows the
parent of /dev to be specified explicitly.

Signed-off-by: Evan Lezar <elezar@nvidia.com>
2023-11-20 21:29:35 +01:00
Evan Lezar
e56bb09889 Use tags.cncf.io for CDI imports
Signed-off-by: Evan Lezar <elezar@nvidia.com>
2023-11-01 12:40:51 +01:00
Evan Lezar
833254fa59 Support CDI devices as mounts
This change allows CDI devices to be requested as mounts in the
container. This enables their use in environments such as kind
where environment variables or annotations cannot be used.

Signed-off-by: Evan Lezar <elezar@nvidia.com>
2023-10-27 21:24:53 +02:00
Evan Lezar
709e27bf4b Fix implicit memory aliasing in for loop
Signed-off-by: Evan Lezar <elezar@nvidia.com>
2023-10-24 20:11:34 +02:00
Evan Lezar
e0df157f70 Remove unnecessary assignment to the blank identifier
Signed-off-by: Evan Lezar <elezar@nvidia.com>
2023-10-24 20:00:24 +02:00
Evan Lezar
12dc12ce09 Fix misspellings
Signed-off-by: Evan Lezar <elezar@nvidia.com>
2023-10-24 20:00:24 +02:00
Evan Lezar
73749285d5 Remove unused loadSaver interface
Signed-off-by: Evan Lezar <elezar@nvidia.com>
2023-10-24 20:00:24 +02:00
Tariq Ibrahim
6d3b29f3ca add a warning statement listing unresolved CDI devices 2023-08-10 08:38:33 -07:00
Evan Lezar
918bd03488 Move tegra-specifics to new package
Signed-off-by: Evan Lezar <elezar@nvidia.com>
2023-08-04 16:49:30 +02:00
Evan Lezar
01a7f7bb8e Explicitly generate CDI spec for CSV mode
This change explicitly generates a CDI specification from
the supplied CSV files when cdi mode is detected. This
ensures consistency between the behaviour on Tegra-based
systems.

Signed-off-by: Evan Lezar <elezar@nvidia.com>
2023-08-04 16:49:30 +02:00
Evan Lezar
6b48cbd1dc Move CDI modifier to separate package
Signed-off-by: Evan Lezar <elezar@nvidia.com>
2023-08-04 16:49:30 +02:00
Evan Lezar
cca343abb0 Pass image when constructing CSV modifier
Since the incoming OCI spec has already been parsed and used to
construct a CUDA image representation, pass this to the CSV
modifier constructor instead of re-creating an image representation.

Signed-off-by: Evan Lezar <elezar@nvidia.com>
2023-07-18 15:27:16 +02:00
Evan Lezar
e2f8d2a15f Set default spec dirs at config level
This change sets the default CDI spec dirs at a config level instead
of when a CDI runtime modifier is constructed. This makes this setting
consistent with other options such as the nvidia-ctk path.

Signed-off-by: Evan Lezar <elezar@nvidia.com>
2023-07-18 15:23:09 +02:00
Evan Lezar
083b789102 Use cdi parser package for IsQualiedName
Signed-off-by: Evan Lezar <elezar@nvidia.com>
2023-07-18 15:16:25 +02:00
Evan Lezar
d92300506c Construct CUDA image object once
Signed-off-by: Evan Lezar <elezar@nvidia.com>
2023-06-13 10:36:02 +02:00
Evan Lezar
1d0a733487 Replace logger.Warn(f) with logger.Warning(f)
This aligns better with klog used in other projects.

Signed-off-by: Evan Lezar <elezar@nvidia.com>
2023-06-12 10:48:04 +02:00
Evan Lezar
a02bc27c3e Define a basic logger interface
Signed-off-by: Evan Lezar <elezar@nvidia.com>
2023-06-12 10:46:10 +02:00
Evan Lezar
ac11727ec5 Add nvidia-contianer-runtime-hook.path config option
This change adds an nvidia-container-runtime-hook.path config option
to allow the path used for the prestart hook to be overridden. This
is useful in cases where multiple NVIDIA Container Toolkit installations
are present.

Signed-off-by: Evan Lezar <elezar@nvidia.com>
2023-05-25 12:05:33 +02:00
Evan Lezar
013a1b413b Fix ineffectual assignment
Signed-off-by: Evan Lezar <elezar@nvidia.com>
2023-05-23 21:14:02 +02:00
Evan Lezar
540dbcbc03 Move tegra system mounts to tegra-specific discoverer
Signed-off-by: Evan Lezar <elezar@nvidia.com>
2023-05-22 13:55:22 +02:00
Evan Lezar
a8265f8846 Add tegra discoverer
Signed-off-by: Evan Lezar <elezar@nvidia.com>
2023-05-22 13:55:22 +02:00
Evan Lezar
8bb0235c92 Remove discover.Config
These changes remove the use of discover.Config which was used
to pass the driver root and the nvidiaCTK path in some cases.

Instead, the nvidiaCTKPath is resolved at the begining of runtime
invocation to ensure that this is valid at all points where it is
used.

Signed-off-by: Evan Lezar <elezar@nvidia.com>
2023-05-10 15:03:37 +02:00
Evan Lezar
c46b118f37 Add nvidia-container-runtime.modes.cdi.annotation-prefixes config option.
This change adds an nvidia-container-runtime.modes.cdi.annotation-prefixes config
option that defaults to cdi.k8s.io/. This allows the annotation prefixes parsed
for CDI devices to be overridden in cases where CDI support in container engines such
as containerd or crio need to be overridden.

Signed-off-by: Evan Lezar <elezar@nvidia.com>
2023-03-27 16:36:54 +02:00
Christopher Desiniotis
48414e97bb Return empty list of devices for unprivileged containers when 'accept-nvidia-visible-devices-envvar-unprivileged=false'
Signed-off-by: Christopher Desiniotis <cdesiniotis@nvidia.com>
2023-03-10 13:11:29 -08:00
Evan Lezar
973e7bda5e Check accept-nvidia-visible-devices-envvar-when-unprivileged option for CDI
Signed-off-by: Evan Lezar <elezar@nvidia.com>
2023-03-09 11:15:53 +02:00
Evan Lezar
6d220ed9a2 Rework selection of devices in CDI mode
The following changes are made:
* The default-cdi-kind config option is used to convert an envvar entry to a fully-qualified device name
* If annotation devices exist, these are used instead of the envvar devices.
* The `all` device is no longer treated as a special case and MUST exist in the CDI spec.

Signed-off-by: Evan Lezar <elezar@nvidia.com>
2023-03-07 16:18:53 +02:00
Evan Lezar
daceac9117 Rename discover.Config.Root to discover.Config.DriverRoot
Signed-off-by: Evan Lezar <elezar@nvidia.com>
2023-02-02 15:57:15 +01:00
Evan Lezar
09d42f0ad9 Remove 'Executable' from config struct member
Signed-off-by: Evan Lezar <elezar@nvidia.com>
2023-01-18 17:02:42 +01:00
Evan Lezar
046d761f4c Ensure that an empty discoverer returns valid edits
Signed-off-by: Evan Lezar <elezar@nvidia.com>
2022-12-06 14:01:35 +01:00
Evan Lezar
8604c255c4 Use Options to set FileLocator options
Signed-off-by: Evan Lezar <elezar@nvidia.com>
2022-12-02 13:57:33 +01:00
Evan Lezar
76b69f45de Add discovery of DRM devices
This change adds the discovery of DRM devices associated with requested
devices. This means that the /dev/dri/card* and /dev/dri/renderD*
devices associated with each requested NVIDIA GPU are injected into
the container and that the /dev/dri/by-path symlinks associated with
these devices are created in the container.

Signed-off-by: Evan Lezar <elezar@nvidia.com>
2022-11-02 14:49:08 +01:00
Evan Lezar
73e65edaa9 Also trigger graphics modifier for display capability
Signed-off-by: Evan Lezar <elezar@nvidia.com>
2022-11-02 14:42:51 +01:00