Index | index by Group | index by Distribution | index by Vendor | index by creation date | index by Name | Mirrors | Help | Search |
Name: libnvidia-container1 | Distribution: SUSE Linux Enterprise 15 |
Version: 1.11.0 | Vendor: SUSE LLC <https://www.suse.com/> |
Release: 150200.5.6.1 | Build date: Thu Oct 20 11:54:31 2022 |
Group: Unspecified | Build host: ibs-power9-13 |
Size: 2952433 | Source RPM: libnvidia-container-1.11.0-150200.5.6.1.src.rpm |
Packager: https://www.suse.com/ | |
Url: https://github.com/NVIDIA/libnvidia-container | |
Summary: NVIDIA container runtime library |
The nvidia-container library provides an interface to configure GNU/Linux containers leveraging NVIDIA hardware. The implementation relies on several kernel subsystems and is designed to be agnostic of the container runtime. This package requires the NVIDIA driver (>= 340.29) to be installed separately.
BSD-3-Clause AND Apache-2.0 AND GPL-3.0-or-later AND LGPL-3.0-or-later AND MIT AND GPL-2.0-only
* Mon Oct 10 2022 [email protected] - Exclude build for arch i586 - Remove libnvidia-container-fix-revision.patch - Update to version 1.11.0 (jsc#SLE-18750): * Bump version to 1.11.0 * Update release date * Properly specify fedora35 artifact dir * Fix format string * Switch to centos:stream8 container for fedora * Add missing changelog entry * Add fedora35 CI steps * Add support for fedora35 * Bump version to 1.11.0-rc.3 * Also change directory when not running pivot root * Bump version to 1.11.0-rc.2 * Bump version to 1.11.0-rc.1 * Preload libgcc_s.so.1 on arm64 systems * Switch to latest docker and docker dind in CI * Bump version to v1.10.0 * Bump version to 1.10.0-rc.4 * Bump github.com/opencontainers/runc from 1.0.3 to 1.1.2 in /src/nvcgo * Fix bug introduced when adding libcudadebugger.so to list of libraries * Bump version to 1.10.0-rc.3 * Add libcudadebugger.so to list of compute libraries * Fix bug when running host ldconfig using glibc compiled with a non-standard prefix * Correct whitespace in changelog * Bump libtirpc to 1.3.2 * Bump libtirpc to 1.3.2 * Change master references to main * Bump version to v1.10.0-rc.2 * Add option --setopt=best=0 to centos and amazonlinux builds * Remove unneeded rpm-sign package from centos and al docker files * Set returned error context to output error to stdout * Exit with error if dxcore initializes with no adapters found * Revert "Log error and skip dxcore initialization if no adapters are found" * Ignore pending MIG mode when checking whether MIG is enabled * Fix typos in comments and changelog * Switch from centos:8 to centos:stream8 images to build centos8 packages * [CI] Add CI job for each distribuition-arch combination * Skip filtering devices based on capability on WSL2 * Log error and skip dxcore initialization if no adapters are found * Bump version to 1.10.0-rc.1 * Bump version to 1.9.0 * Update Changelog * Add additional check for Tegra in /sys/.../family file in CLI * Update changelog * Change default jetpack-specific CLI option for loading base CSV files * Ensure that firmware paths are mounted at /lib in the container * Factor out mount code to accept explicit source and destination * Update Changelog * Update switch statement to include new brands from latest nvml.h * Update NVML to "CUDA 11.6.1 nvML_DEV 11.6.55" * Fix logic to process --require flags on Jetson platforms * Bump version to 1.9.0-rc.1 * Fix centos8 build * Update (placeholder) changelog for 1.8.2-1 * Add symlink for /dev/fd --> /proc/dev/fd when doing ldconfig in chroot * Update changelog * Fix bug where cgroup mount prefix not stripped from cgroup root * Bump version to 1.8.1 * Add NVC_CONTAINER caps around cgroup version and find_cgroup_path calls * Bump version to 1.8.0 * Remove amazonlinux1 build targets * Bump version to 1.8.0-rc.3 * Specify docker platform args for build and run commands * Update centos8 mirrors * Update changelog * Include libnvidia-pkcs11.so in compute libraries * Support writing entire device filter eBPF program (if none present) * Remove custom code for getting existing eBPF program for device filters * Update cgroupv2 code to use cilium/ebpf to get existing eBPF program * Update celium/ebpf dependency to 0.8.0 in nvcgo * Include firmware paths in list command * Rename (n)dirs members to (n)firmwares * Fix permissions when mounting GSP firmware * Fix bug to support cgroupv2 on linux kernels < 5.5 * Bump version to 1.8.0-rc.2 * Fix bug in cgroupv2 logic when in mixed v1 / v2 environment * Set WITH_NVCGO=yes in mk/docker.mk * Bump github.com/opencontainers/runc from 1.0.2 to 1.0.3 in /src/nvcgo * Turn on WITH_NVCGO feature gate by default * Update changelog for 1.8.0-rc.1 * Allow build-all jobs to be triggered earlier * Ensure that artifact directory exists for docker build * Suppress lintian warnings regarding exported symbols * Install libnvidia-container-go from deps * Fix DESTDIR for deps make target * Do a blanket s/ctx/rpc/ over rpc.c to make it clear what type we have * Introduce rpc_prog struct to wrap fields for defining an RPC program * Cleanup the nvc_shutdown() path when there is an error in RPC services * Add notion of "progname" to RPC service for better logging and errors * Add CAP_NET_ADMIN to NVC_MOUNT permissions * Explicitly set CAP_EFFECTIVE to NVC_MOUNT when setting device rules * Add new cgroup.c file to support both cgroupv1 and cgroupv2 * Move cgroup.c to cgroup_legacy.c in preparation for nvcgo implementation * Add new RPC server for running commands from the nvcgo library * Add nvcgo library that can be called from C for running Go code * Add vendor folder in preperation for new nvcgo package * Bump post 1.7.0 release * Move nvml_path into driver context instead of explicit RPC argument * Move logic to chroot and drop privileges from RPC init into driver init * Move src/driver_rpc.x -> src/nvc_rpc.x to accomodate more RPC programs * Move from an nvc_context specific RPC driver service to a global one * Rename rpc->rpc_* to just rpc->* * Split parts of driver_init/shutdown() into rpc_init/shutdown() * Split parts of driver.{c,h} into rpc.{c,h} * Rename 'driver' service definitions to 'rpc' service definitions * Move initialization of driver->nvml_dl to RPC handler of driver_init() * Make RPC prognum, versnum, and dispatch function part of driver context * Bump version to 1.7.0 * Add -fplan9-extensions to Makefile for future use * Remove the unreferenced declaration of device_mig_device in driver.h * Make device_handles struct in driver.c static * Add replacement for versions in debian symbol file * Bump post 1.7.0-rc.1 release * Move SONAME_LIBNVML and SONAME_LIBCUDA to nvc_internal.h * Remove erroneous cuda_dl field in 'struct driver' * Remove erroneous "unused parameter" warning * Fix typos in changelog * Update changelogs * Add dependency on libnvidia-container0 on ubuntu18.04-arm64 * Update flags in CLI to be library version specific (as necessary) * Correctly bump NVC_MINOR to 7 * Add make logic to check that version is correctly specified * Separate cli-version from lib-version in version output of CLI * Backport some CLI configurations from the jetson branch to the mainline * Dynamically load v0 of libnvidia-container when on tegra without NVML * Fix typo in rpm changelog * Only print device paths in the CLI list command if path != NULL * Move to calling all NVC functions through function pointers * Bump for post 1.6.0 development * Bump version to 1.6.0 * Bump nvidia-modprobe version to 495.44 * Bump version post 1.6.0-rc.2 release * Fix bug for missing files in /proc/driver/nvidia * Bump version to 1.6.0-rc.2 * Add aarch64 build for Amazon Linux 2 * Introduce a get_cgroup_version() call to the cgroupp API * Update cgroup API to be device specific * Refactor cgroup logic into its own set of files * Update major, minor, and patch version * Bump for post v1.5.1 development * Bump version to 1.5.1 * Use defined root when resolving firmware path * Free allocated dirs * Update changelogs for merged changes * Allow to override build date with SOURCE_DATE_EPOCH * getpgrp is required to run nvcr.io/nvidia/tensorrt:20.12-py3 * Bump for post v1.5.0 development * Add NVC_TAG to version header file * Bump version to 1.5.0 * Only copy .deb and .rpm files to dist folder * Allow REVISION to be specified as a make variable * Correct debian version tag * Bump version to 1.5.0~rc.2 * Read TAG from version string * Revert "Mount nvlink and nvswitch devices into container" * Revert "Add --fabric-device option to select nvlink and nvswitch devices" * Revert "Check for pointer null before strcmp" * Check for pointer null before strcmp * lseek64/_llseek is required for ldconfig * Bump version to 1.5.0~rc.1 * Mount NVIDIA firmware directory if present * Add --fabric-device option to select nvlink and nvswitch devices * Mount nvlink and nvswitch devices into container * Add dist and .vscode to gitignore * Use pmake and bmake from mageia 7.1 in amazonlinux build * WSL - Add code to map the binaries from the driver store if they are redistributed with the display driver * WSL - Add code to map the binaries from the driver store if they are redistributed with the display driver * Explicitly check whether branch is a tag before releasing to Artifactory * Add release step to Jenkinsfile * Wed Jul 14 2021 [email protected] - Update libnvidia-container to 1.4.0 - Update nvidia-modprobe to 450.57 * Wed Jun 03 2020 [email protected] - Update to version 1.1.1 - Fix libnvidia-container-fix-makefile.patch for SLE15 * Tue May 26 2020 [email protected] - Cleanup the spec file, fully unbundle elftoolchain part. - Add no-manual-debuginfo.patch, we automatically generate it. * Tue May 19 2020 [email protected] - Fix libnvidia-container-tools runtime dependencies * Wed Apr 22 2020 [email protected] - Remove vendored libtirpc and add libnvidia-container-fix-makefile.patch as substitution * Wed Apr 08 2020 [email protected] - Update to version 1.0.6 - Fix build for tumbleweed * Tue Feb 11 2020 [email protected] - Update to version 1.0.5: * Fix release number in rpms * Fix changelog * Bump version to 1.0.5 * Add pkg-config package to fix build * Bump version to 1.0.4 * Update DSL to output the first element instead of the last in case of failure * opensuse-leap: ensure /mnt is writeable by the unprivileged user * mk/nvidia-modprobe: fix parallel build issue * add support for openSUSE Leap 15 * Add initial support for Optix * Add libnvidia-container-fix-revision.patch
/usr/lib64/libnvidia-container-go.so /usr/lib64/libnvidia-container-go.so.1.11.0 /usr/lib64/libnvidia-container.so.1 /usr/lib64/libnvidia-container.so.1.11.0 /usr/share/doc/packages/libnvidia-container1 /usr/share/doc/packages/libnvidia-container1/README.md /usr/share/licenses/libnvidia-container-1.11.0 /usr/share/licenses/libnvidia-container-1.11.0/COPYING /usr/share/licenses/libnvidia-container-1.11.0/COPYING.LESSER /usr/share/licenses/libnvidia-container-1.11.0/LICENSE /usr/share/licenses/libnvidia-container-1.11.0/NOTICE
Generated by rpm2html 1.8.1
Fabrice Bellet, Tue Jul 9 19:51:39 2024