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
1,803 Commits 21 Branches 73 Tags 29 MiB
cd52be86e6
Commit Graph

1803 Commits

This Branch
This Branch
All Branches
Author SHA1 Message Date
Christopher Desiniotis
64fb26b086 Add option to nvidia-ctk to enable CDI in docker 
Signed-off-by: Christopher Desiniotis <cdesiniotis@nvidia.com>
Signed-off-by: Evan Lezar <elezar@nvidia.com>
 2023-11-23 10:15:58 +01:00
Evan Lezar
16a4de1a2b Merge branch 'CNT-4645/add-nvswitch-devices' into 'main' 
Add support for NVIDIA_NVSWITCH envvar to inject nvidia-nvswitch device nodes

See merge request nvidia/container-toolkit/container-toolkit!502
 2023-11-22 21:00:54 +00:00
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
886c6b973e Merge branch 'library-search-path-cdi-generate' into 'main' 
Locate libnvidia-egl-gbm.so.*

See merge request nvidia/container-toolkit/container-toolkit!504
 2023-11-22 20:58:55 +00:00
Evan Lezar
1ab3ef0af4 Locate libnvidia-egl-gbm.so.* 
Searching for a pattern allows platforms where no `.so` symlink
exists to function as expected.

Signed-off-by: Evan Lezar <elezar@nvidia.com>
 2023-11-22 21:57:36 +01:00
Evan Lezar
dd9b13cb58 Merge branch 'bump-changelog' into 'main' 
Add missing changelog

See merge request nvidia/container-toolkit/container-toolkit!503
 2023-11-22 20:46:09 +00:00
Evan Lezar
8a7a6e8a70 Add missing changelog 
Signed-off-by: Evan Lezar <elezar@nvidia.com>
 2023-11-22 20:51:12 +01:00
Evan Lezar
1909b1fe60 Merge branch 'library-search-path-cdi-generate' into 'main' 
Allow search paths when locating libcuda.so

See merge request nvidia/container-toolkit/container-toolkit!462
 2023-11-22 19:49:15 +00:00
Evan Lezar
881e440d22 Merge branch 'include-nvoptix' into 'main' 
Update list of graphics mounts

See merge request nvidia/container-toolkit/container-toolkit!501
 2023-11-22 19:47:52 +00:00
Evan Lezar
7d79b311d8 Include vulkan/icd.d/nvidia_layers.json 
This change includes vulkan/icd.d/nvidia_layers.json in the list of
possible graphics mounts.
 2023-11-22 13:54:12 +01:00
Evan Lezar
b46bc10c44 Include nvidia/nvoptix.bin in graphics mounts 
Signed-off-by: Evan Lezar <elezar@nvidia.com>
 2023-11-22 13:53:59 +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
f20ab793a2 Add support for specifying search paths 
Signed-off-by: Evan Lezar <elezar@nvidia.com>
 2023-11-22 13:27:47 +01:00
Evan Lezar
e5391760e6 Remove duplicate not found error 
Signed-off-by: Evan Lezar <elezar@nvidia.com>
 2023-11-21 18:09:42 +01:00
Evan Lezar
5505886655 Use options for NewLibraryLocator 
Signed-off-by: Evan Lezar <elezar@nvidia.com>
 2023-11-21 18:08:53 +01:00
Evan Lezar
64f554ef41 Add builder for file locator 
Signed-off-by: Evan Lezar <elezar@nvidia.com>
 2023-11-21 18:07:47 +01:00
Evan Lezar
fc8c5f82dc Merge branch 'fix-ldconfig-resolution' into 'main' 
Resolve LDConfig path

See merge request nvidia/container-toolkit/container-toolkit!490
 2023-11-21 16:45:21 +00:00
Evan Lezar
d792e64f38 Resolve ldconfig path in update-ldcache hook 
Signed-off-by: Evan Lezar <elezar@nvidia.com>
 2023-11-21 15:31:12 +01:00
Evan Lezar
232df647c1 Resolve LDConfig path passed to nvidia-container-cli 
Instead of relying solely on a static config, we resolve the path
to ldconfig. The path is checked for existence and a .real suffix is preferred.

Signed-off-by: Evan Lezar <elezar@nvidia.com>
 2023-11-21 15:31:12 +01:00
Evan Lezar
adc516fd59 Merge branch 'ctk-hook-chmod-improve-eperm-handling' into 'main' 
nvidia-ctk hook chmod: Improve permission error handling

See merge request nvidia/container-toolkit/container-toolkit!496
 2023-11-21 11:05:03 +00:00
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
Christopher Desiniotis
2768023ff5 Merge branch 'gen-cdi-spec-at-runtime' into 'main' 
Automatically generate CDI spec for the runtime.nvidia.com/gpu=all device

See merge request nvidia/container-toolkit/container-toolkit!492
 2023-11-20 22:15:05 +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
e315d7d74b Merge branch 'allow-separate-dev-root' into 'main' 
Add devRoot option to CDI api

See merge request nvidia/container-toolkit/container-toolkit!497
 2023-11-20 21:10:12 +00:00
Evan Lezar
b4c6832828 Add additional debug 
Signed-off-by: Evan Lezar <elezar@nvidia.com>
 2023-11-20 21:29:35 +01:00
Evan Lezar
3a96a00362 Simplify meta device discovery 
Signed-off-by: Evan Lezar <elezar@nvidia.com>
 2023-11-20 21:29:35 +01:00
Evan Lezar
00a712d018 Add --dev-root option to CDI spec generation 
Signed-off-by: Evan Lezar <elezar@nvidia.com>
 2023-11-20 21:29:35 +01:00
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
Ievgen Popovych
9085cb7dd5 nvidia-ctk hook chmod: Move file mode parsing into flag validation function 
Signed-off-by: Ievgen Popovych <jmennius@gmail.com>
 2023-11-20 14:49:29 +02:00
Evan Lezar
f6e3593a72 Merge branch 'update-go-nvlib' into 'main' 
Update to github.com/NVIDIA/go-nvlib@9fd385bace0d2b8949cf60d9fcaab6169bde87ef

See merge request nvidia/container-toolkit/container-toolkit!495
 2023-11-20 11:25:25 +00:00
Evan Lezar
f2ef7ee661 Update to github.com/NVIDIA/go-nvlib@9fd385bace0d2b8949cf60d9fcaab6169bde87ef 
Signed-off-by: Evan Lezar <elezar@nvidia.com>
 2023-11-20 12:25:08 +01:00
Evan Lezar
27777f4dab Merge branch 'bump-libnvidia-container' into 'main' 
Update libnvidia-container

See merge request nvidia/container-toolkit/container-toolkit!494
 2023-11-20 11:11:15 +00:00
Evan Lezar
34175f15d3 Update libnvidia-container 
Signed-off-by: Evan Lezar <elezar@nvidia.com>
 2023-11-20 10:57:35 +01:00
Ievgen Popovych
eb35d9b30a nvidia-ctk hook chmod: Ignore permission errors 
In some cases we might get a permission error trying to chmod -
most likely this is due to something beyond our control
like whole `/dev` being mounted.
Do not fail container creation in this case.

Due to loosing control of the program after `exec()`-ing `chmod(1)` program
and therefore not being able to handle errors -
refactor to use `chmod(2)` syscall instead of `exec()` `chmod(1)` program.

Fixes: #143
Signed-off-by: Ievgen Popovych <jmennius@gmail.com>
 2023-11-20 01:29:51 +02:00
Ievgen Popovych
f1d32f2cd3 nvidia-ctk hook chmod: Only chmod if desired permissions are different 
This is to avoid any unnecessary potential errors (e.g. due to permissions).

Fixes: #143
Signed-off-by: Ievgen Popovych <jmennius@gmail.com>
 2023-11-20 01:18:36 +02:00
Evan Lezar
ee713adf33 Merge branch 'fix-update-ldcache' into 'main' 
Allow ldcache update in container to be skipped

See merge request nvidia/container-toolkit/container-toolkit!485
 2023-11-18 11:37:56 +00:00
Christopher Desiniotis
33cb1b68df Merge branch 'deduplicate-symlinks' into 'main' 
Deduplicate symlinks

See merge request nvidia/container-toolkit/container-toolkit!493
 2023-11-17 17:09:18 +00:00
Evan Lezar
6dc9ee3f33 Allow ldcache update in container to be skipped 
This change skips the update of ld.cache in the container if it
doesn't exist. Instead, the -N flag is used to only create the
relevant symlinks.

Signed-off-by: Evan Lezar <elezar@nvidia.com>
 2023-11-17 11:56:19 +01:00
Evan Lezar
e609e41a64 Allow multiple pattern matches for symlinks 
Since we allow pattern inputs for locating symlinks we could have
multiples. The error being checked is resolved by the deduplication.

Signed-off-by: Evan Lezar <elezar@nvidia.com>
 2023-11-17 10:43:52 +01:00
Evan Lezar
80ecd024ee Add tests for library locator 
Signed-off-by: Evan Lezar <elezar@nvidia.com>
 2023-11-17 10:43:52 +01:00
Evan Lezar
e8dbb216a5 Return empty ldcache if cache does not exist 
Signed-off-by: Evan Lezar <elezar@nvidia.com>
 2023-11-17 10:14:03 +01:00
Christopher Desiniotis
f5d8d248b7 Deduplicate symlinks 
Signed-off-by: Christopher Desiniotis <cdesiniotis@nvidia.com>
 2023-11-16 17:57:31 -08:00
Evan Lezar
5d7ee25b37 Merge branch 'migrate-go-nvlib' into 'main' 
Use github.com/NVIDIA/go-nvlib imports

See merge request nvidia/container-toolkit/container-toolkit!491
 2023-11-15 20:39:35 +00:00
Evan Lezar
2ff2d84283 Update vendoring 
Signed-off-by: Evan Lezar <elezar@nvidia.com>
 2023-11-15 21:38:54 +01:00
Evan Lezar
c63fb35ba8 Use github.com/NVIDIA/go-nvlib imports 
Signed-off-by: Evan Lezar <elezar@nvidia.com>
 2023-11-15 21:38:26 +01:00
Evan Lezar
da0755769f Merge branch 'improve-library-lookup' into 'main' 
Make CDI-based library discovery more robust

See merge request nvidia/container-toolkit/container-toolkit!488
 2023-11-06 18:17:25 +00:00
Evan Lezar
04b28d116c Make library lookups more robust 
These changes make library lookups more robust. The core change is that
library lookups now first look a set of predefined locations before checking
the ldcache. This also handles cases where an ldcache is not available more
gracefully.

Signed-off-by: Evan Lezar <elezar@nvidia.com>
 2023-11-06 12:15:28 -06:00
Evan Lezar
65b0b2b5e0 Merge branch 'make-graphics-optional' into 'main' 
Make discovery of graphics libraries optional

See merge request nvidia/container-toolkit/container-toolkit!489
 2023-11-03 21:22:47 +00:00