Install this version:
emerge -a =dev-libs/cudnn-8.9.7.29
If this version is masked, you can unmask it using the autounmask tool or standard emerge options:
autounmask =dev-libs/cudnn-8.9.7.29
Or alternatively:
emerge --autounmask-write -a =dev-libs/cudnn-8.9.7.29
| Version | EAPI | Keywords | Slot |
|---|---|---|---|
| 8.9.7.29 | 8 | ~amd64 ~amd64-linux | 0/8 |
# Copyright 1999-2023 Gentoo Authors
# Distributed under the terms of the GNU General Public License v2
EAPI=8
BASE_V="$(ver_cut 0-3)"
# supports 12.x but URL has a specific version number
CUDA_MA="12"
CUDA_MI="8"
CUDA_V="${CUDA_MA}.${CUDA_MI}"
DESCRIPTION="NVIDIA Accelerated Deep Learning on GPU library"
HOMEPAGE="https://developer.nvidia.com/cudnn"
SRC_URI="https://developer.download.nvidia.com/compute/cudnn/redist/cudnn/linux-x86_64/cudnn-linux-x86_64-${PV}_cuda${CUDA_MA}-archive.tar.xz"
S="${WORKDIR}/cudnn-linux-x86_64-${PV}_cuda${CUDA_MA}-archive"
LICENSE="NVIDIA-cuDNN"
SLOT="0/8"
KEYWORDS="~amd64 ~amd64-linux"
RESTRICT="mirror"
RDEPEND="=dev-util/nvidia-cuda-toolkit-12*"
QA_PREBUILT="/opt/cuda/targets/x86_64-linux/lib/*"
src_install() {
insinto /opt/cuda/targets/x86_64-linux
doins -r include lib
}
| Type | File | Size | Source URLs |
|---|---|---|---|
| DIST | cudnn-linux-x86_64-8.9.7.29_cuda12-archive.tar.xz | 864984964 bytes | https://developer.download.nvidia.com/compute/cudnn/redist/cudnn/linux-x86_64/cudnn-linux-x86_64-8.9.7.29_cuda12-archive.tar.xz |