This is a second attempt at unifying the generic and manual-config
kernel builds (see #412 for the last time).
The set of working kernel packages is a superset of those that work on
master, and as the only objection last time was the size of the $dev
closure and now both $out and $dev combined are 20M smaller than $out on
master (see message for ac2035287f), this
should be unobjectionable.
Signed-off-by: Shea Levy <shea@shealevy.com>
In most cases, this just meant changing kernelDev (now removed from
linuxPackagesFor) to kernel.dev. Some packages needed more work (though
whether that was because of my changes or because they were already
broken, I'm not sure). Specifics:
* psmouse-alps builds on 3.4 but not 3.10, as noted in the comments that
were already there
* blcr builds on 3.4 but not 3.10, as noted in comments that were
already there
* open-iscsi, ati-drivers, wis-go7007, and openafsClient don't build on
3.4 or 3.10 on this branch or on master, so they're marked broken
* A version-specific kernelHeaders package was added
The following packages were removed:
* atheros/madwifi is superceded by official ath*k modules
* aufs is no longer used by any of our kernels
* broadcom-sta v6 (which was already packaged) replaces broadcom-sta
* exmap has not been updated since 2011 and doesn't build
* iscis-target has not been updated since 2010 and doesn't build
* iwlwifi is part of mainline now and doesn't build
* nivida-x11-legacy-96 hasn't been updated since 2008 and doesn't build
Everything not specifically mentioned above builds successfully on 3.10.
I haven't yet tested on 3.4, but will before opening a pull request.
Signed-off-by: Shea Levy <shea@shealevy.com>
Also, Ant no longer has a build-time dependency on a particular JDK.
It finds the JDK via $JAVA_HOME or $PATH (by looking up javac). This
way, we don't need to have separate packages like apacheAntOpenJDK and
apacheAntOracleJDK. It also seems reasonable: after all, installing
GNU Make doesn't give you a C compiler either. It does mean that
instead of
buildInputs = [ ant ];
you now need to write something like
buildInputs = [ ant jdk ];
This unifies the "openjdk" and "openjre" packages. The JDK is placed
in the "out" output, the JRE in "jre".
Also, everything is now stored in $prefix/lib/openjdk, so the JDK/JRE
no longer pollute user environments with files like
"ASSEMBLY_EXCEPTION" at top-level.
This has three major benefits:
1. We no longer have two kernel build processes to maintain
2. The build process is (IMO) cleaner and cleaves more closely to
upstream. In partuclar, we use make install to install the kernel and
development source/build trees, eliminating the guesswork about which
files to copy.
3. The derivation has multiple outputs: the kernel and modules are in
the default `out' output, while the build and source trees are in a
`dev' output. This makes it possible for the full source and build tree
to be kept (which is expected by out-of-tree modules) without bloating
the closure of the system derivation.
In addition, if a solution for how to handle queries in the presence of
imports from derivations ever makes it into nix, a framework for
querying the full configuration of the kernel in nix expressions is
already in place.
Signed-off-by: Shea Levy <shea@shealevy.com>
- Unify the "single" and "float" variants, which are the same thing.
- Enable threads and openmp wrapper by default (they are very small).
- Don't use sse on i686, as I'm quite sure we have no warrant for that.
Cf. http://en.wikipedia.org/wiki/Pentium_Pro
- Disable static libs (big, no indication we need them).
- I tested most builds using fftw (they build OK).
With this, I was able to successfully compile a defconfig kernel for the
sheevaplug, though I didn't actually try to run it (not having a
sheevaplug myself).
For native compiles, the most significant difference is that the
platform's kernel target is built directly rather than hoping the
default make target will pull it in.
Also some stylistic improvements along the way.
Signed-off-by: Shea Levy <shea@shealevy.com>
In addition to replacing use of a feature with IMO confusing semantics,
this will also let us cleanly separate out all-packages.nix into
category-specific files without changing the structure of the pkgs
attrset.
Signed-off-by: Shea Levy <shea@shealevy.com>
This patch also bumps up the HPN version of openssh so that it compiles
on top of 6.4. Along with the bump, a package was added for the high
performance networking version.
The gcmrekey patch was removed as this vulnerability is fixed in
version 6.4 onward. http://www.openssh.org/txt/gcmrekey.adv
It's to separate from other changes coming from master.
Conflicts:
pkgs/development/libraries/glibc/2.18/common.nix (taking both changes)
pkgs/development/libraries/ncurses/5_4.nix (deleted)
This commit, and the "pinnings" that follow should not cause major issues.
I personally checked (build-only) each and every package that depends
on ffmpeg to find the maximum supported version.
I performed this on stdenv-updates+glibc2.18+x-updates but I don't expect
that to matter.
Tested like this:
nix-build -A docbook2odf
./result/bin/docbook2odf --verbose --debug result/share/doc/docbook2odf/examples/book.docbook
swriter book.odt # verify that the document looks ok
"Dzen is a general purpose messaging, notification and menuing program for X11. It was designed to be fast, tiny and scriptable in any language." -- https://github.com/robm/dzen
This can fail randomly with an error like:
building toplev.o
../../gcc-4.6.3/gcc/reload.c: In function 'find_reloads':
../../gcc-4.6.3/gcc/reload.c:3778:11: error: corrupted value profile: value profile counter (93791 out of 95865) inconsistent with basic-block count (95759)
http://hydra.nixos.org/build/7185029
On the plus side, this makes building GCC much faster because it
re-enables parallel building.
When FreeCAD is started it prints that it's missing matplotlib and
pycollada:
$ FreeCAD
[...]
matplotlib not found, Plot module will be disabled
pycollada not found, collada support will be disabled.
The closure size of freecad is currently at 1.9 GiB. Adding these two
deps increase it to 2.0 GiB.
Reflow the expression function argument list.
