Commit graph

28 commits

Author SHA1 Message Date
Nikolay Amiantov
044a49febf cudatoolkit: remove OpenCL library
libOpenCL is provided by ocl-icd with implementation in NVIDIA driver instead.
2017-02-28 17:22:06 +03:00
Eelco Dolstra
c2d4b14637 cudatoolkit: Multiple output improvements
* Add a "lib" output containing (for the moment) only libOpenCL. This
  reduces the closure size of opensubdiv and blender by about 2 GiB.

* Add a "doc" output (about ~200 MiB).

* Remove the "sdk" output since it was worse than useless: all of the
  "sdk" output was also included in "out", so it actually increased
  disk space usage.

* Run patchelf on ELF binaries only.
2017-02-28 11:21:33 +01:00
Eelco Dolstra
b5c619cbfe cudatoolkit: 8.0.44 -> 8.0.61 2017-02-28 11:21:33 +01:00
Frederik Rietdijk
ccf4d5e3d5 cudatoolkit5: remove this ancient version 2017-02-13 14:42:24 +01:00
Eelco Dolstra
5ecb427368 cudatoolkit: Add 8.0.44 2016-11-01 22:52:26 +01:00
Eelco Dolstra
5ade8fff79 cudatoolkit: Fix building dependent packages
- Set a cmake flag to allow cmake to find CUDA automatically.

- Pass -D_FORCE_INLINES to work around

    /nix/store/8sl4jfs3nq0pkq4gg655s3axrxdx7z29-glibc-2.24-dev/include/string.h: In function 'void* __mempcpy_inline(void*, const void*, size_t)':
    /nix/store/8sl4jfs3nq0pkq4gg655s3axrxdx7z29-glibc-2.24-dev/include/string.h:650:42: error: 'memcpy' was not declared in this scope

  https://github.com/BVLC/caffe/issues/4046

This fixes OpenSubdiv and Blender.
2016-09-17 00:40:41 +02:00
Nikolay Amiantov
87ebab128a replace ${stdenv.cc.cc}/lib occurences 2016-05-01 00:13:23 +03:00
Eelco Dolstra
bd4720a7d6 cudatoolkit: Don't barf on GCC > 4.9
Blender appears to build/run fine with GCC 5.3.
2016-04-08 22:23:14 +02:00
Eelco Dolstra
6d97de951d cudatoolkit: Add version 7.5.18 2016-03-02 21:35:46 +01:00
Eelco Dolstra
6c1e3a82de cudatoolkit: Merge into one file and use callPackages 2016-03-02 21:35:46 +01:00
Vladimír Čunát
88c9f8b574 xlibs: replace occurrences by xorg
This seems to have been confusing people, using both xlibs and xorg, etc.
- Avoided renaming local (and different) xlibs binding in gcc*.
- Fixed cases where both xorg and xlibs were used.
Hopefully everything still works as before.
2015-09-15 12:54:34 +02:00
Bjørn Forsman
22321f2e58 cudatoolkit: don't move $out/include to $out/usr_include
This effectively reverts 86c283824f
("If cuda headers are presented to nix [...]") and all the following
workarounds that was added due to that commit.

As far as I can tell[1] this hack isn't needed anymore. And moving
includes to $out/usr_include causes pain for cudatoolkit users, so
better get rid of it.

In patches that did more than the $out/usr_include workaround, I only
changed the line back to $out/include instead of re-generating the
patches and fully removing the changed line.

[1]: I build tested blender and caffe, and temporarily added
recurseIntoAttrs to rPackages and haskellPackages so that nox-review
could get proper coverage. However, many of the packages do not build
even before this patch. I also built CUDA samples with cudatoolkit7
that ran fine.
2015-09-10 10:35:40 +02:00
Bjørn Forsman
836fa19651 cudatoolkit: fixup path to samples
This is needed for cuda 6.5 or else nsight will not find them. In cuda 7.0 the
samples are already in $out/samples. Older versions do not build due to broken
dependency (python-minimal).
2015-08-18 17:06:42 +02:00
William Casarin
d1590482cf cudatoolkit: generalize derivations, add cuda7
There is not much different between the cudatoolkit
expressions:

  pkgs/development/compilers/cudatoolkit/5.5.nix
  pkgs/development/compilers/cudatoolkit/6.0.nix
  pkgs/development/compilers/cudatoolkit/6.5.nix

This commit removes those and replaces them with

  pkgs/development/compilers/cudatoolkit/generic.nix

and adds cudatoolkit version 7
2015-07-04 17:05:50 -07:00
Aristid Breitkreuz
c6d17dfa24 fix up cudatoolkit 6.5 so that nvprof can run 2015-05-02 12:36:20 +02:00
Eelco Dolstra
6f33aa8465 orig-gcc -> orig-cc 2015-03-04 22:15:46 +01:00
Eelco Dolstra
97fcb33ac2 cudatoolkit: Fix gcc -> cc
Urgh, we shouldn't have renamed gcc to cc...
2015-03-04 22:15:46 +01:00
Eelco Dolstra
ea0d6a7fc0 Add CUDA toolkit 6.5 2015-03-02 00:24:00 +01:00
Eelco Dolstra
2ba9ae0129 Fix bad licenses on CUDA packages
They were marked as "nonfree" rather than "unfree", causing them to be
allowed with allowUnfree = false.
2015-02-15 22:29:47 +01:00
Eric Seidel
f3c6827373 rename all occurrences of stdenv.cc.gcc to stdenv.cc.cc 2015-01-14 20:27:55 -08:00
John Wiegley
28b6fb61e6 Change occurrences of gcc to the more general cc
This is done for the sake of Yosemite, which does not have gcc, and yet
this change is also compatible with Linux.
2014-12-26 11:06:21 -06:00
Daniel Zinn
86c283824f If cuda headers are presented to nix in $out/include they are added to future gcc calls via a -isystem flag. However, cuda does not allow kernel calls from template function if these are located in system-headers. We thus move headers from $out/include to $out/usr_include and add a custom hook to add these headers via -I. 2014-06-23 06:11:34 -07:00
Daniel Zinn
ca0de0f1f7 Added cuda6 compiler as package cudatoolkit6. 2014-06-19 04:24:21 -07:00
Vladimír Čunát
34091bf265 cudatoolkit: fix multiple outputs (close #2466) 2014-06-11 21:32:13 +02:00
David Guibert
42911b1903 cuda: install the SDK using multiple-outputs support 2014-01-29 15:48:55 +01:00
Bjørn Forsman
a74ab8c0a1 cudatoolkit: fix missing dependencies
Some of the java tools failed to run (notably nsight) because of missing
libraries. This fixes it.
2013-08-27 14:11:23 +02:00
Justin Bedo
6c2985e8d7 Update CUDA Toolkit to 5.5 2013-08-27 13:44:56 +02:00
Andres Löh
7aa85ac536 Adding NVIDIA CUDA toolkit and haskell-accelerate-cuda plus deps.
svn path=/nixpkgs/trunk/; revision=34186
2012-05-20 17:38:03 +00:00