UniqAI/nvidia-container-toolkit
1
0
Fork 0
mirror of https://github.com/NVIDIA/nvidia-container-toolkit synced 2024-11-29 07:21:46 +00:00
Code Issues Actions Packages Projects Releases Wiki Activity
571 Commits 21 Branches 73 Tags 29 MiB
5c39cf4deb
Commit Graph

94 Commits

Author SHA1 Message Date
Evan Lezar
9191074666 Rename discover.NewList to discover.Merge 
Signed-off-by: Evan Lezar <elezar@nvidia.com>
 2022-07-05 10:28:40 +02:00
Evan Lezar
89824849d3 Merge branch 'refactor-envvar-devices' into 'main' 
Add DevicesFromEnvvars function to CUDA image abstraction

See merge request nvidia/container-toolkit/container-toolkit!178
 2022-07-04 08:47:28 +00:00
Evan Lezar
fd135f1a8b Add Relative function to Locator interface 
This adds a Relative function to the Locator interface and uses
this to determine the host and container paths for located files
(and devices). This ensures that the root (e.g. the nvidia driver
root) is stripped from the container path.

Signed-off-by: Evan Lezar <elezar@nvidia.com>
 2022-07-01 16:23:50 +02:00
Evan Lezar
4e08ec2405 Use CUDA.DevicesFromEnvvar to check if modifications are required 
Signed-off-by: Evan Lezar <elezar@nvidia.com>
 2022-07-01 16:14:36 +02:00
Evan Lezar
925c348565 Add DevicesFromEnvvars function to CUDA image 
Signed-off-by: Evan Lezar <elezar@nvidia.com>
 2022-07-01 16:12:13 +02:00
Evan Lezar
a1c2f07b6e Add /etc/cufile.json to list of required mounts 
Signed-off-by: Evan Lezar <elezar@nvidia.com>
 2022-07-01 14:54:58 +02:00
Evan Lezar
7f7bec0668 Create GDS and MOFED modifiers 
This change creates GDS and MOFED modifiers and adds them to the
modifer created for the selected runtime mode if the NVIDIA_GDS
and NVIDIA_MOFED envvars are set to "enabled", respectively.

Signed-off-by: Evan Lezar <elezar@nvidia.com>
 2022-07-01 14:54:05 +02:00
Evan Lezar
cb34f7c6d1 Add discovery of GDS and MOFED devices 
Signed-off-by: Evan Lezar <elezar@nvidia.com>
 2022-07-01 14:40:55 +02:00
Evan Lezar
7f47a61986 Allow globs in filenames for locators 
Signed-off-by: Evan Lezar <elezar@nvidia.com>
 2022-07-01 14:30:33 +02:00
Evan Lezar
e8843c38f2 Move cmd/nvidia-container-runtime/modifier package to internal/modifier 
Signed-off-by: Evan Lezar <elezar@nvidia.com>
 2022-07-01 14:28:40 +02:00
Evan Lezar
55ac8628c8 Add lists of modifiers to allow for modifier compositioning 
Signed-off-by: Evan Lezar <elezar@nvidia.com>
 2022-07-01 14:25:18 +02:00
Evan Lezar
73a5b70a02 Return default config if config path is not found 
Signed-off-by: Evan Lezar <elezar@nvidia.com>
 2022-05-25 13:22:45 +02:00
Evan Lezar
e07c7f0fa2 Ignore NVIDIA_REQUIRE_JETPACK* for image requirements 
Signed-off-by: Evan Lezar <elezar@nvidia.com>
 2022-05-24 09:53:37 +02:00
Evan Lezar
084eae6e0d Fix bug in tegra detection 
Signed-off-by: Evan Lezar <elezar@nvidia.com>
 2022-05-20 14:39:36 +02:00
Evan Lezar
55c1d7c256 Fix assertCharDevice matching on all files 
Signed-off-by: Evan Lezar <elezar@nvidia.com>
 2022-05-20 10:08:00 +02:00
Evan Lezar
c77e86137e Add version output to CLIs 
This change adds version output to the nvidia-continer-runtime,
nvidia-container-toolkit, and nvidia-ctk CLIs. The same version
is used in all cases and includes a version string and a git
revision if set.

The construction of the version string mirrors what is done in runc.

Signed-off-by: Evan Lezar <elezar@nvidia.com>
 2022-05-13 07:31:11 +02:00
Evan Lezar
ff86ecb2a5 Include HasNVML check in ResolveAutoMode 
Signed-off-by: Evan Lezar <elezar@nvidia.com>
 2022-05-12 10:55:58 +02:00
Evan Lezar
ad9ec1efae Add HasNVML function to check if NVML is supported 
Signed-off-by: Evan Lezar <elezar@nvidia.com>
 2022-05-12 10:55:13 +02:00
Evan Lezar
9db5f9c9e8 Remove unneeded legacy discovery 
Signed-off-by: Evan Lezar <elezar@nvidia.com>
 2022-05-12 10:53:52 +02:00
Evan Lezar
e591f3f26b Replace experimental and discover-mode 
These changes replace the nvidia-container-runtime config options
experimental and discover-mode with a single mode config option.

Note that mode is now a string with a default value of "auto"
and a mode value of "legacy" is equivalent to experimental == false.

Signed-off-by: Evan Lezar <elezar@nvidia.com>
 2022-05-12 10:53:50 +02:00
Evan Lezar
e0ad82e467 Move ResolveAutoMode to info package 
Signed-off-by: Evan Lezar <elezar@nvidia.com>
 2022-05-12 10:28:56 +02:00
Evan Lezar
3a1404f2f4 Move isTegraSystem to internal info package 
Signed-off-by: Evan Lezar <elezar@nvidia.com>
 2022-05-12 10:28:56 +02:00
Evan Lezar
cf7bb91481 Update nvidia-container-runtime config options 
Signed-off-by: Evan Lezar <elezar@nvidia.com>
 2022-05-12 10:28:56 +02:00
Evan Lezar
ba0e606df2 Use toml unmarshal to read runtime config 
Signed-off-by: Evan Lezar <elezar@nvidia.com>
 2022-05-12 10:28:56 +02:00
Evan Lezar
ae57a2fc93 Merge branch 'CNT-2875/create-specific-symlinks' into 'main' 
Create specific symlinks for CSV mode

See merge request nvidia/container-toolkit/container-toolkit!150
 2022-05-12 05:27:43 +00:00
Evan Lezar
1eb0e3c8b3 Merge branch 'fix-executable-locator' into 'main' 
Fix location of executables in PATH

See merge request nvidia/container-toolkit/container-toolkit!148
 2022-05-12 05:26:22 +00:00
Evan Lezar
675fbace01 Add hook to create specific links 
This change updates the create-symlinks hook to also create symlinks for
libcuda.so, libGLX_indirect.so.0, and libnvidia-opticalflow.so

Signed-off-by: Evan Lezar <elezar@nvidia.com>
 2022-05-11 16:36:49 +02:00
Evan Lezar
d2516cb5d5 Merge branch 'fix-container-root' into 'main' 
Fix bug in update-ldcache hook when OCI spec contains a relative root

See merge request nvidia/container-toolkit/container-toolkit!147
 2022-05-10 22:01:14 +00:00
Evan Lezar
4696d7ee69 Merge branch 'fix-hook-flags' into 'main' 
Use singular instead of plural for hook arguments

See merge request nvidia/container-toolkit/container-toolkit!146
 2022-05-10 22:00:51 +00:00
Evan Lezar
ef6f48e9f7 Use singular instead of plural for hook arguments 
Signed-off-by: Evan Lezar <elezar@nvidia.com>
 2022-05-10 19:55:31 +02:00
Evan Lezar
088db09180 Use executable locator to find low-level runtime 
Signed-off-by: Evan Lezar <elezar@nvidia.com>
 2022-05-10 15:21:48 +02:00
Evan Lezar
1d2e1bd403 Add lookup.GetPath and lookup.GetPaths functions 
Signed-off-by: Evan Lezar <elezar@nvidia.com>
 2022-05-10 14:52:47 +02:00
Evan Lezar
395f6cecb2 Add GetContainerRoot to oci.State type 
This change adds a GetContainerRoot to the oci.State type to
encapsulate the logic around determining the container root.
This Fixes a bug where relative roots (e.g. as generated by contianerd)
are not supported.

Signed-off-by: Evan Lezar <elezar@nvidia.com>
 2022-05-10 11:48:43 +02:00
Evan Lezar
7574a0d7de Make output of bundle directory a debug message 
Signed-off-by: Evan Lezar <elezar@nvidia.com>
 2022-05-09 09:38:16 +02:00
Evan Lezar
335de5a352 Switch to debug logging when locating runtimes 
Signed-off-by: Evan Lezar <elezar@nvidia.com>
 2022-05-09 09:38:16 +02:00
Evan Lezar
c76946cbcc Add nvidia-container-runtime.runtimes config option 
Signed-off-by: Evan Lezar <elezar@nvidia.com>
 2022-05-09 09:38:12 +02:00
Evan Lezar
785f120c31 Fix form -> from in comment 
Signed-off-by: Evan Lezar <elezar@nvidia.com>
 2022-05-06 13:22:34 +02:00
Evan Lezar
9e46d41dbe Add debug logging when checking requirements 
Signed-off-by: Evan Lezar <elezar@nvidia.com>
 2022-05-05 14:14:01 +02:00
Evan Lezar
9f50ac95c4 Add CUDA ComputeCapability function 
Signed-off-by: Evan Lezar <elezar@nvidia.com>
 2022-05-05 14:09:28 +02:00
Evan Lezar
583793b7ae Add processing for requirements and constraints 
This change adds a Requirements abstraction that can be used to check
an images' NVIDIA_REQUIRE_* envvars against the host properties such
as CUDA version or architecture.

Signed-off-by: Evan Lezar <elezar@nvidia.com>
 2022-05-05 13:43:13 +02:00
Evan Lezar
5d7b3a4a96 Return raw spec from Spec.Load 
Signed-off-by: Evan Lezar <elezar@nvidia.com>
 2022-05-05 13:43:13 +02:00
Evan Lezar
a672713dba Add basic CUDA wrapper 
Signed-off-by: Evan Lezar <elezar@nvidia.com>
 2022-05-05 13:43:13 +02:00
Evan Lezar
8f0e1906c2 Add CUDA image abstraction 
This change adds a CUDA image abstraction that encapsulates
the queries performed on a container image (e.g. envvars) to
check certain CUDA properties.

Signed-off-by: Evan Lezar <elezar@nvidia.com>
 2022-05-05 13:43:13 +02:00
Evan Lezar
c224832a6d Add log-level config option for nvidia-container-runtime 
Signed-off-by: Evan Lezar <elezar@nvidia.com>
 2022-04-08 13:56:17 +02:00
Evan Lezar
5211960fc3 Merge branch 'detect-gpus-flag' into 'master' 
Detect use of --gpus flag in experimental mode

See merge request nvidia/container-toolkit/container-toolkit!125
 2022-04-08 11:18:11 +00:00
Evan Lezar
cfca18a5f8 Merge branch 'refactor-csv-mount-spec-discovery' into 'master' 
Refactor CSV discovery to make char device discovery clearer

See merge request nvidia/container-toolkit/container-toolkit!129
 2022-04-08 10:54:06 +00:00
Evan Lezar
dab6f4b768 Specify --force flag when invoking nvidia-container-runtime-hook 
Signed-off-by: Evan Lezar <elezar@nvidia.com>
 2022-04-08 12:03:22 +02:00
Evan Lezar
2563c1b87c Export GetDefaultRuntimeConfig 
Signed-off-by: Evan Lezar <elezar@nvidia.com>
 2022-04-08 12:03:22 +02:00
Evan Lezar
62f608a3fe Make order of discoverers deterministic 
Signed-off-by: Evan Lezar <elezar@nvidia.com>
 2022-04-08 11:59:26 +02:00
Evan Lezar
2c1e356370 Refactor CSV discovery to make char device discovery clearer 
Signed-off-by: Evan Lezar <elezar@nvidia.com>
 2022-04-08 11:47:47 +02:00