Deeply-curried functions are pretty error-prone in untyped languages
like Nix. This is a particularly bad case because
`top-level/splice.nix` *also* declares a makeScopeWithSplicing, but
it takes *two fewer arguments*.
Let's switch to attrset-passing form, to provide some minimal level
of sanity-checking.
This is needed to avoid confusing and repeated boilerplate for
`fooForTarget`. The vast majority of use-cases can still use
`buildPackages or `targetPackages`, which are now defined in terms of
these.
Otherwise extra packages in scope can't be made to work for cross. As
much as I think splicing is an evil trick, I think it's best to do this
and at least have it work consistently for now.
It would seems simpler to expose a `newScopeWithSplicing`, but there's a
few attrs (like `buildPackages` or `buildHaskellPackages`) that
shouldn't be spliced. Users should instead splice, override the splicing
on those packages, and apply `newScope` to that.
Not only does the suffix unnecessarily reduce sharing, but it also breaks
unpacker setup hooks (e.g. that of `unzip`) which identify interesting tarballs
using the file extension.
This also means we can get rid of the splicing hacks for fetchers.
- No more `or {}`, this was misleading at best since those values
wouldn't be used unless the attr they are defined from was present
anyways.
- `tryEval` for get outputs. This ensures that if some derivations fail, they
won't take out the others. This benefited from the `or {}` before, but that
was never good enough. `tryEval` is strictly better.
Each bootstrapping stage ought to just depend on the previous stage, but
poorly-written compilers break this elegence. This provides an easy-enough
way to depend on the next stage: targetPackages. PLEASE DO NOT USE IT
UNLESS YOU MUST!
I'm hoping someday in a pleasant future I can revert this commit :)
- `pkgs` is self-similar, and thus already spliced
- `buildPackages` is an ingredient of splicing and should be kept as is
- The platforms are not packages or package sets and couldn't be spliced
There's probably other things that shouldn't be spliced too. The best long-
term solution is simply to stop splicing altogether.
[N.B., this package also applies to the commits that follow it in the same
PR.]
In most cases, buildPackages = pkgs so things work just as before. For
cross compiling, however, buildPackages is resolved as the previous
bootstrapping stage. This allows us to avoid the mkDerivation hacks cross
compiling currently uses today.
To avoid a massive refactor, callPackage will splice together both package
sets. Again to avoid churn, it uses the old `nativeDrv` vs `crossDrv` to do
so. So now, whether cross compiling or not, packages with get a `nativeDrv`
and `crossDrv`---in the non-cross-compiling case they are simply the same
derivation. This is good because it reduces the divergence between the
cross and non-cross dataflow. See `pkgs/top-level/splice.nix` for a comment
along the lines of the preceding paragraph, and the code that does this
splicing.
Also, `forceNativeDrv` is replaced with `forceNativePackages`. The latter
resolves `pkgs` unless the host platform is different from the build
platform, in which case it resolves to `buildPackages`. Note that the
target platform is not important here---it will not prevent
`forcedNativePackages` from resolving to `pkgs`.
--------
Temporarily, we make preserve some dubious decisions in the name of preserving
hashes:
Most importantly, we don't distinguish between "host" and "target" in the
autoconf sense. This leads to the proliferation of *Cross derivations
currently used. What we ought to is resolve native deps of the cross "build
packages" (build = host != target) package set against the "vanilla
packages" (build = host = target) package set. Instead, "build packages"
uses itself, with (informally) target != build in all cases.
This is wrong because it violates the "sliding window" principle of
bootstrapping stages that shifting the platform triple of one stage to the
left coincides with the next stage's platform triple. Only because we don't
explicitly distinguish between "host" and "target" does it appear that the
"sliding window" principle is preserved--indeed it is over the reductionary
"platform double" of just "build" and "host/target".
Additionally, we build libc, libgcc, etc in the same stage as the compilers
themselves, which is wrong because they are used at runtime, not build
time. Fixing this is somewhat subtle, and the solution and problem will be
better explained in the commit that does fix it.
Commits after this will solve both these issues, at the expense of breaking
cross hashes. Native hashes won't be broken, thankfully.
--------
Did the temporary ugliness pan out? Of the packages that currently build in
`release-cross.nix`, the only ones that have their hash changed are
`*.gcc.crossDrv` and `bootstrapTools.*.coreutilsMinimal`. In both cases I
think it doesn't matter.
1. GCC when doing a `build = host = target = foreign` build (maximally
cross), still defines environment variables like `CPATH`[1] with
packages. This seems assuredly wrong because whether gcc dynamically
links those, or the programs built by gcc dynamically link those---I
have no idea which case is reality---they should be foreign. Therefore,
in all likelihood, I just made the gcc less broken.
2. Coreutils (ab)used the old cross-compiling infrastructure to depend on
a native version of itself. When coreutils was overwritten to be built
with fewer features, the native version it used would also be
overwritten because the binding was tight. Now it uses the much looser
`BuildPackages.coreutils` which is just fine as a richer build dep
doesn't cause any problems and avoids a rebuild.
So, in conclusion I'd say the conservatism payed off. Onward to actually
raking the muck in the next PR!
[1]: https://gcc.gnu.org/onlinedocs/gcc/Environment-Variables.html
