Commit graph

85 commits

Author SHA1 Message Date
John Ericson f18ddabee7 Merge remote-tracking branch 'upstream/master' into lib-float 2018-05-10 18:13:00 -04:00
John Ericson 1fe81a4bcd lib: Clean up float/fpu options
ARM ABIs now have a float field. This is used as a fallback to lessen
our use of `platform.gcc.float`. I didn't know what the MIPs convention
is so I kept using `platform.gcc.float` in that case.
2018-05-10 18:02:00 -04:00
John Ericson 58b2e875c2 lib/systems: Prohibit "gnu" ABI (*-gnu) with 32-bit ARM
It is ambiguous, and therefore banned within GCC.
2018-05-10 15:05:23 -04:00
John Ericson feb648ce59 Merge commit '70963b382f3f820ba6d3bc3b3aaf50a2957ec1ff' into lib-platform-simplify 2018-05-10 01:40:38 -04:00
John Ericson a02be2bd85 treewide: Get rid of *Platform.arch
Use `parsed.cpu.name` or `platform.gcc.arch` instead.
2018-05-10 01:37:31 -04:00
John Ericson f063a860d6 xbursttools: Cleanup slightly 2018-05-09 23:40:13 -04:00
John Ericson e42a7a5c0b lib/systems: Add uClibc just like MUSL 2018-05-09 23:39:23 -04:00
John Ericson 3fa0ba9177 lib/systems: Parse more arm cpu types 2018-05-09 18:57:39 -04:00
John Ericson db4f96b3cb lib/systems: Fix eval for iphone32* examples
Whoops messed up 9a845de873 slightly.
2018-05-01 13:04:57 -04:00
John Ericson 9a845de873 lib/systems: Update iOS examples
The commented-out configs are @shlevy's old known-good ones. I changed
them as needed to play nice with lib.systems.parse but did not test so
leaving them as comments for now.
2018-05-01 01:16:27 -04:00
John Ericson 57723e947a Merge remote-tracking branch 'upstream/master' into aarch32 2018-04-30 23:06:59 -04:00
John Ericson b9acfb4ecf treewide: isArm -> isAarch32
Following legacy packing conventions, `isArm` was defined just for
32-bit ARM instruction set. This is confusing to non packagers though,
because Aarch64 is an ARM instruction set.

The official ARM overview for ARMv8[1] is surprisingly not confusing,
given the overall state of affairs for ARM naming conventions, and
offers us a solution. It divides the nomenclature into three levels:

```
ISA:             ARMv8   {-A, -R, -M}
                 /    \
Mode:     Aarch32     Aarch64
             |         /   \
Encoding:   A64      A32   T32
```

At the top is the overall v8 instruction set archicture. Second are the
two modes, defined by bitwidth but differing in other semantics too, and
buttom are the encodings, (hopefully?) isomorphic if they encode the
same mode.

The 32 bit encodings are mostly backwards compatible with previous
non-Thumb and Thumb encodings, and if so we can pun the mode names to
instead mean "sets of compatable or isomorphic encodings", and then
voilà we have nice names for 32-bit and 64-bit arm instruction sets
which do not use the word ARM so as to not confused either laymen or
experienced ARM packages.

[1]: https://developer.arm.com/products/architecture/a-profile

(cherry picked from commit ba52ae5048)
2018-04-25 15:50:41 -04:00
John Ericson ba52ae5048 treewide: isArm -> isAarch32
Following legacy packing conventions, `isArm` was defined just for
32-bit ARM instruction set. This is confusing to non packagers though,
because Aarch64 is an ARM instruction set.

The official ARM overview for ARMv8[1] is surprisingly not confusing,
given the overall state of affairs for ARM naming conventions, and
offers us a solution. It divides the nomenclature into three levels:

```
ISA:             ARMv8   {-A, -R, -M}
                 /    \
Mode:     Aarch32     Aarch64
             |         /   \
Encoding:   A64      A32   T32
```

At the top is the overall v8 instruction set archicture. Second are the
two modes, defined by bitwidth but differing in other semantics too, and
buttom are the encodings, (hopefully?) isomorphic if they encode the
same mode.

The 32 bit encodings are mostly backwards compatible with previous
non-Thumb and Thumb encodings, and if so we can pun the mode names to
instead mean "sets of compatable or isomorphic encodings", and then
voilà we have nice names for 32-bit and 64-bit arm instruction sets
which do not use the word ARM so as to not confused either laymen or
experienced ARM packages.

[1]: https://developer.arm.com/products/architecture/a-profile
2018-04-25 15:28:55 -04:00
John Ericson cf0dd1dbd2
Merge pull request #39172 from obsidiansystems/ios-cross
ios-sdk-pkgs: Init from iOS SDK from XCode
2018-04-19 17:00:05 -04:00
Ken Micklas ef3db7d14c ios-sdk-pkgs: Init from iOS SDK from XCode 2018-04-19 16:09:30 -04:00
John Ericson 53686e8995
Merge pull request #38485 from obsidiansystems/nixos-nixpkgs-options
nixpkgs module: Clean up platform options
2018-04-19 14:59:58 -04:00
John Wiegley 4cfdcedb71 Fix a typo: types.openSignifiantByte -> types.openSignificantByte 2018-04-13 18:41:53 -07:00
John Ericson c6f7d43678 nixpkgs module: Clean up platform options
- `localSystem` is added, it strictly supercedes system

 - `crossSystem`'s description mentions `localSystem` (and vice versa).

 - No more weird special casing I don't even understand

TEMP
2018-04-06 12:41:44 -04:00
Shea Levy 26e8d58cb5
libseccomp: Disable only on RISC-V if Linux.
The isSeccomputable flag treated Linux without seccomp as just a
normal variant, when it really should be treated as a special case
incurring complexity debt to support.
2018-03-27 08:16:24 -04:00
Shea Levy cdf9a78a3e
kexectools: Disable only on RISC-V if Linux.
The isKexecable flag treated Linux without kexec as just a normal
variant, when it really should be treated as a special case incurring
complexity debt to support.
2018-03-27 08:15:07 -04:00
Shea Levy 20f51922c0
riscv-pk: Set platforms properly 2018-03-24 08:44:26 -04:00
John Ericson d9a1800239 lib: Add lib.platforms.windows 2018-03-20 12:47:45 -04:00
John Ericson 3c8ae01a45 lib: Make platforms.all actually match all platforms
Otherwise obscure cross-compilations are hampered. `all` breaks all but
the initial derivation (which we can't even write yet) in an open world
setting however, so we really shouldn't have it.
2018-03-20 12:46:19 -04:00
John Ericson 175d4ab1db lib: Make platform predicates greppable
Should have commited on here and on merged master to begin with, but I
didn't, so instead I cherry-pick.

(cherry picked from commit 88c04a8b6b)
2018-03-20 12:35:20 -04:00
John Ericson 88c04a8b6b lib: Make platform predicates greppable 2018-03-19 22:20:48 -04:00
John Ericson bbfa2f9701 lib: Split Darwin into macOS and iOS
I noticed LLVM accepts `ios` as its own OS in platform triples; a
recent change as far as I know. I see it also accepts `macos*` for macOS
(formerly OS X). If it's now customary to distinguish iOS like so
(rather than guessing from the aarch, lets add both so our OSes are
still disjoint, and make Darwin a family instead.

But changing the config everywhere would probably be a mass rebuild, and
I'm not sure how well other software supports OSes besides "darwin", so
I'm keeping that the default name for macOS for now.
2018-03-19 12:02:49 -04:00
John Ericson c26252af3e lib, stdenv: Check meta.platforms against host platform and be open world
First, we need check against the host platform, not the build platform.
That's simple enough.

Second, we move away from exahustive finite case analysis (i.e.
exhaustively listing all platforms the package builds on). That only
work in a closed-world setting, where we know all platforms we might
build one. But with cross compilation, we may be building for arbitrary
platforms, So we need fancier filters. This is the closed world to open
world change.

The solution is instead of having a list of systems (strings in the form
"foo-bar"), we have a list of of systems or "patterns", i.e. attributes
that partially match the output of the parsers in `lib.systems.parse`.
The "check meta" logic treats the systems strings as an exact whitelist
just as before, but treats the patterns as a fuzzy whitelist,
intersecting the actual `hostPlatform` with the pattern and then
checking for equality. (This is done using `matchAttrs`).

The default convenience lists for `meta.platforms` are now changed to be
lists of patterns (usually a single pattern) in
`lib/systems/for-meta.nix` for maximum flexibility under this new
system.

Fixes #30902
2018-03-15 00:44:34 -04:00
John Ericson f79f80dbf2 treewide: get rid of platforms.allBut
Negative reasoning like `allBut` is a bad idea with an open world of
platforms. Concretely, if we add a new, quite different sort of
platform, existing packages with `allBut` will claim they work on it
even though they probably won't.
2018-03-14 18:44:42 -04:00
John Ericson 2482e2858e prebuilt android tools: Init using SDK
Expose as an option for the cross stdenv.
2018-02-27 14:15:39 -05:00
John Ericson dfc5d7835d
Merge pull request #35247 from telent/mips32
lib, treewide: Add missing MIPS arches, and fix existing usage
2018-02-27 14:01:15 -05:00
Shea Levy 3c57e770cf
Only build libseccomp on supported systems 2018-02-24 23:05:08 -05:00
Shea Levy 0ac6d4aeb6
gnu-efi: Fix aarch64 cross-build 2018-02-24 22:15:48 -05:00
Shea Levy 87270fb306
gnu-efi: Only build on efi-enabled systems 2018-02-24 21:49:31 -05:00
Shea Levy 63fa1fcb04
kexectools: Only build on kexecable architectures. 2018-02-24 10:37:54 -05:00
Daniel Barlow 9c50ae6898 lib, treewide: Add missing MIPS arches, and fix existing usage
Existing "mips64el" should be "mipsel".

This is just the barest minimum so that nixpkgs can recognize them as
systems - although required for building individual derivations onto
MIPS boards, it is not sufficient if you want to actually build nixos on
those targets
2018-02-23 20:43:42 -05:00
Shea Levy 32d3533b29
riscv: Compile console support into the kernel. 2018-02-19 22:43:18 -05:00
Shea Levy 6173f2f945
linux_riscv: Add 4.16-rc1.
Fixes #35148.
2018-02-19 12:14:22 -05:00
Shea Levy e288febee0
Add riscv{32,64} crossSystems. 2018-02-18 00:09:25 -05:00
Tuomas Tynkkynen e4bfe5aac9 treewide: Drop kernelHeadersBaseConfig
Nothing actually needs this.
2018-02-14 01:47:28 +02:00
Will Dietz f090bbb248 Drop "isGlibc", but keep isMusl.
gnu "abi" doesn't mean glibc (mingw, apparently).
2018-02-11 14:55:17 -06:00
Will Dietz 2dfee94fe7 lib/systems: musl, libc predicates
Note this doesn't actually provide musl support yet,
just improves our "system" code to understand
musl-based triples and non-glibc linux configurations.
2018-02-11 14:20:14 -06:00
Tuomas Tynkkynen c9d1bf3e76 platforms.nix: Include RPi 3 serial port in the kernel config 2018-02-06 22:14:09 +02:00
John Ericson afb87a66b3 lib: Avoid double import 2018-01-31 00:13:11 -05:00
John Ericson ad78ba1efa lib: Better use the module type system in platform parsing
I need some module system types here so I can next fix meta-checks for
derivations. I'd like to use a "proper" record type here, but submodule
types seem overkill so holding off with ad-hoc stuff for now. In
practice, all I need for the next step are the `.check` functions so
this is good, especially as the submodule check function is shallow,
saving full inductive type-checking for a later step.
2018-01-30 22:03:06 -05:00
John Ericson 57b01b1bcf lib, openssl: Get rid of openssl.system
We compute it on the fly, careful to avoid any mass rebuilds for now.
2018-01-26 21:22:00 -05:00
John Ericson 16a50f5a07 lib: Remove examples platforms' bigEndian attr
They still have `parsed.cpu.significantByte` which has the same info.
2018-01-26 21:22:00 -05:00
John Ericson 71f814a889 lib, glibc: Get rid of withTLS
glibc removed the underlying flag in 2011 in
83cd14204559abbb52635006832eaf4d2f42514a [1].

This gets us one step closer to fixing #34274: the cross stdenv for
aarch64-unknown-linux-gnu at least evals now.

Thanks to @Dezgeg for doing all the research for this.

[1]: https://sourceware.org/git/?p=glibc.git;a=commit;h=83cd14204559abbb52635006832eaf4d2f42514a
2018-01-26 23:29:06 +02:00
John Ericson d1478c91c3 lib: Allow parsing platform configs with arch of {riscv,wasm}{32,64}
Also add `isRiscv` and `isWasm` predicates.
2018-01-26 12:44:05 -05:00
John Ericson 992bd2f6d3 Merge commit 'ab77a6bb1e7d2ff475210ad392f1a9bd1bb6ba3a' into gcc-simplify-flags 2017-12-05 17:41:15 -05:00
John Ericson 93cd0685c5 Merge commit '71186e73455a4e06e96a31da34b76f84e545ba1f' into gcc-simplify-flags 2017-12-05 17:09:41 -05:00