about summary refs log tree commit diff
path: root/pkgs/development/cuda-modules
Commit message (Collapse)AuthorAge
* cudaPackages.autoAddOpenGLRunpath: keep an alias for compatibilitySomeone Serge2024-03-15
|
* cudaPackages: generalize and refactor setup hookYann Hamdaoui2024-03-15
| | | | | | | | | | | | | | | | | | | | | This PR refactor CUDA setup hooks, and in particular autoAddOpenGLRunpath and autoAddCudaCompatRunpathHook, that were using a lot of code in common (in fact, I introduced the latter by copy pasting most of the bash script of the former). This is not satisfying for maintenance, as a recent patch showed, because we need to duplicate changes to both hooks. This commit abstract the common part in a single shell script that applies a generic patch action to every elf file in the output. For autoAddOpenGLRunpath the action is just addOpenGLRunpath (now addDriverRunpath), and is few line function for autoAddCudaCompatRunpathHook. Doing so, we also takes the occasion to use the newer addDriverRunpath instead of the previous addOpenGLRunpath, and rename the CUDA hook to reflect that as well. Co-Authored-By: Connor Baker <connor.baker@tweag.io>
* cudaPackages.nccl: 2.19.3-1 -> 2.20.3-1R. Ryantm2024-02-16
|
* cudaPackages.nccl-tests: 2.13.8 -> 2.13.9R. Ryantm2024-02-07
|
* Merge pull request #285527 from lheckemann/cudatoolkit-libglvndSomeone2024-02-04
|\ | | | | cudaPackages.cudatoolkit: add libglvnd to inputs
| * cudaPackages.cudatoolkit: add libglvnd to inputsLinus Heckemann2024-02-01
| | | | | | | | | | At least some of the older versions (we noticed the problem with 11.4) need libEGL.so. libglvnd seems like the obvious place to get it.
* | cudaPackage: don't rely on non-portable builtinssternenseemann2024-01-31
| | | | | | | | | | | | | | builtins.groupBy was added in Nix 2.5 (https://nixos.org/manual/nix/stable/release-notes/rl-2.5), but Nixpkgs has to support Nix >= 2.3. Problem introduced in 9a33f8ce5ba7329600b296996d944ec9ddaa7d38.
* | Merge pull request #282185 from SomeoneSerge/fix/cu-multiplexSomeone2024-01-22
|\ \ | | | | | | | | | | | | cudaPackages: fix version solving (again) Unbreaks tensorflowWithCuda and cudaPackagesGoogle.cudnn_8_6
| * | cudaPackagesGoogle.cudnn_8_6: ensure present on all platformsSomeone Serge2024-01-20
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | ...including jetson Previously: ❯ nix eval -f . --argstr system aarch64-linux --arg config '{ cudaCapabilities = [ "6.2" ]; cudaEnableForwardCompat = false; cudaSupport = true; allowUnfree = true; }' -L cudaPackagesGoogle.cudnn_8_6.outPath attribute ... in selection path .... not found Now: ❯ nix eval -f . --argstr system aarch64-linux --arg config '{ cudaCapabilities = [ "6.2" ]; cudaEnableForwardCompat = false; cudaSupport = true; allowUnfree = true; }' -L cudaPackagesGoogle.cudnn_8_6.outPath error: Package ‘cudnn-8.6.0.163’ in ... is not available on the requested hostPlatform:
| * | cudaPackagesGoogle.cudnn_8_6: fix evalSomeone Serge2024-01-20
| | | | | | | | | | | | | | | | | | | | | | | | | | | Now: ❯ nix eval -f . --arg config '{ allowUnfree = true; }' cudaPackagesGoogle.cudnn_8_6.outPath "/nix/store/eeeeeeeeeeeeeeeeeeeeeeeeeeeeeeee-cudnn-8.6.0.163" Previously: meta.broken = true
* | | cudaPackages.backendStdenv: fix eval error blocking the channelSomeone Serge2024-01-21
| | | | | | | | | | | | ...post https://github.com/NixOS/nixpkgs/pull/282220
* | | cudaPackages.backendStdenv: drop the commented out codeSomeone Serge2024-01-20
| | |
* | | cudaPackages.cuda_nvcc: back-end cc already exposes the c++ stdlibSomeone Serge2024-01-20
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Unbreaks evaluation after the refactoring in 86c28ee650c4393ec1b6f8815a0e1ddc010cd6cb (https://github.com/NixOS/nixpkgs/pull/282220). Explicitly extending nvcc's LIBRARIES may have been redundant (building something mildly cursed like jaxlib may be necessary to verify) Previously: ``` ❯ NIXPKGS_ALLOW_UNFREE=1 nix-instantiate '.' -A tests.cuda ... error: necessary to fix CI ❯ nix build -f '.' --arg config '{ allowUnfree = true; }' -L python3Packages.torchWithCuda ... error: necessary to fix CI ``` Now: ``` ❯ NIXPKGS_ALLOW_UNFREE=1 nix-instantiate '.' -A tests.cuda /nix/store/eeeeeeeeeeeeeeeeeeeeeeeeeeeeeeee-cuda-samples-12.2.drv /nix/store/eeeeeeeeeeeeeeeeeeeeeeeeeeeeeeee-saxpy-unstable-2023-07-11.drv ... ```
* | | cc-wrapper: relocate proprietary-compiler-specific materialAdam Joseph2024-01-20
|/ / | | | | | | | | | | | | | | | | | | | | | | | | | | | | PR #275947, which was self-merged without approvals, inserted functionality specific to a propriteary closed-source compiler (CUDA) into cc-wrapper. This commit relocates this CUDA-specific functionality into the appropritate place: `cuda-modules`. It is unclear to me exactly what this function is supposed to be doing; much of it (like the `.kind` attributes) do not appear to be used *anywhere* in nixpkgs. Making sure we don't insert unexplained deadcode like this is one of the important functions of the review process.
* | Merge pull request #281806 from ↵Someone2024-01-19
|\ \ | |/ |/| | | | | ConnorBaker/fix/cudaPackages-multiplex-builder-isSupported-check cudaPackages: fix multiplex builder isSupported check
| * cudaPackages: fix multiplex builder isSupported checkConnor Baker2024-01-18
| | | | | | | | Co-Authored-By: Yann Hamdaoui <yann.hamdaoui@tweag.io>
* | Merge pull request #281461 from hacker1024/patch-8Someone2024-01-18
|\ \ | |/ |/| cudaPackages.autoAddCudaCompatRunpathHook: Correctly inherit platforms from cuda_compat
| * cudaPackages.autoAddCudaCompatRunpathHook: Correctly inherit platforms from ↵hacker10242024-01-17
| | | | | | | | cuda_compat
* | Merge pull request #269639 from SomeoneSerge/update/cudaPackagesSomeone2024-01-17
|\ \ | | | | | | cudaPackages: bump the default 11 -> 12
| * | cudaPackages_11_4.nsight_systems: clean up the deprecation commentsSomeone Serge2024-01-17
| | |
| * | cudaPackages_12.nsight_systems: unbreakSomeone Serge2024-01-17
| | |
| * | cudaPackages.cuda_cudart: a separate output for stubSomeone Serge2024-01-17
| |/
* / cudaPackages: __structuredAttrs works with autoPatchelf since #272752Connor Baker2024-01-16
|/
* Merge pull request #280386 from SomeoneSerge/fix/tests.cudaSomeone2024-01-15
|\ | | | | tests.cuda: inherit the ready cudaPackages_XX.cuda-samples
| * cudaPackages_11_4.saxpy: clean(er) eval errorSomeone Serge2024-01-12
| |
| * cudaPackages.cuda{,-library}-samples: move to cuda-modules/Someone Serge2024-01-12
| | | | | | | | cudaPackages are part of the package set and should not depend on pkgs/test
* | cudaPackages.backendStdenv: switch to stdenvAdapters.useLibsFromSomeone Serge2024-01-12
| |
* | cudaPackages.cuda_nvcc: use gcc from pkgsHostTargetSomeone Serge2024-01-12
| | | | | | | | i.e. with the same offsets as nvcc itself
* | cc-wrapper: cxxStdlib: expose solib and package separatelySomeone Serge2024-01-12
| |
* | cudaPackages.backendStdenv: use gccForLibsRahul Butani2024-01-12
|/ | | | | | instead of the unreliable libcxx hack (cherry picked from commit dc6a8f9f7912363577e11520bafa040c0db14359)
* Merge pull request #278828 from jonringer/fix-cuda-10.0Someone2024-01-11
|\ | | | | cudaPackage_10_0.cudatoolkit: fix build
| * cudaPackage_10_0.cudatoolkit: fix buildJonathan Ringer2024-01-04
| | | | | | | | | | | | | | | | | | Acknowledge to `autoPatchelfHook` that we don't have ancient versions of ffmeg for the jre shipped. Also, targets/ directory doesn't exist in cuda 10.0, so the preFixup phase collected 0 files and when passed to patchelf would fail the build.
* | Merge pull request #274319 from ConnorBaker/feat/cudaPackages-all-packages-evalSomeone2024-01-11
|\ \ | | | | | | tree-wide: cudaPackages attributes should not cause default eval to fail
| * | cudaPackages: default badPlatformsConditions to emptyConnor Baker2024-01-10
| | |
| * | cudaPackages: set badPlatforms when cudaSupport is falseConnor Baker2024-01-10
| | |
| * | cudaPackages: unsupported platform should not set broken to trueConnor Baker2024-01-10
| | |
| * | tree-wide: cudaPackages should not break default evalConnor Baker2024-01-10
| | | | | | | | | | | | cudaPackages: guard expressions against null values
* | | Merge pull request #280076 from SomeoneSerge/fix/apptainer-nv-cherrySomeone2024-01-11
|\ \ \ | | | | | | | | apptainer: unbreak --nv (userns+nvliblist.conf variant)
| * | | cudaPackages.saxpy: 1/3 the runtime closureSomeone Serge2024-01-10
| | | | | | | | | | | | | | | | (cherry picked from commit 3e003a17bafb8ad2c9026843c241412e2d61a541)
* | | | Merge pull request #280073 from SomeoneSerge/fix/cuda-pkg-configSomeone2024-01-10
|\ \ \ \ | |_|/ / |/| | | cudaPackages: move .pc to hook-defined location
| * | | cudaPackages: move .pc to hook-defined locationSomeone Serge2024-01-09
| | | |
* | | | Merge pull request #277213 from danielfullmer/cuda-runpath-libsConnor Baker2024-01-09
|\ \ \ \ | |_|/ / |/| | | cudaPackages: apply runpath setup hooks to non-executable ELF files
| * | | cudaPackages: apply runpath setup hooks to non-executable ELF filesDaniel Fullmer2023-12-27
| | |/ | |/|
* / | cudaPackages.autoAddCudaCompatRunpathHook: Fix eval w/o cuda_compatSilvan Mosberger2024-01-07
|/ /
* | Merge pull request #276800 from SomeoneSerge/fix/cuda-no-throwSomeone2023-12-26
|\ \ | | | | | | cudaPackages: eliminate exceptions
| * | cudaPackages: manifest-builder: fake url/sha256 instead of exceptionsSomeone Serge2023-12-26
| | |
| * | cudaPackages: eliminate exceptionsSomeone Serge2023-12-26
| | | | | | | | | | | | | | | as they break common nixpkgs tools like lib.meta, cf. https://github.com/NixOS/nixpkgs/issues/276795
* | | cudaPackages.cudart: stubs: add the libcuda.so.1 sonameSomeone Serge2023-12-25
|/ /
* | Merge pull request #275921 from SomeoneSerge/fix/cuda-pkg-configSomeone2023-12-25
|\| | | | | cudaPackages: replace the FHS paths in pkg-config files
| * cudaPackages: allow FHS references by defaultSomeone Serge2023-12-25
| | | | | | | | ...harden gradually instead
generated by cgit-pink 1.4.1 (git 2.36.1) at 2024-06-28 10:17:47 +0000