Setting the SDK root by default allows `overrideSDK` to correctly set
the SDK version when using a different SDK. It also allows the correct
SDK version to be set when using an older deployment target. Not setting
the correct SDK version can result in unexpected behavior at runtime.
Examples:
* Automatic dark mode switching requires linking against an SDK version
of 10.14 or newer. With the current behavior, the only way to do this
is by using a 10.14+ deployment target even when the application
supports older platforms when build with a newer SDK.
* MetalD3D checks that the system version is at least 14.0. The API it
uses returns a compatibility version when the the SDK is older than
11.0, which causes it to display an error and terminate the
application even when even when its requirements are all met.
This is effectively a rewrite of `overrideSDK`. It was required because
`wrapGAppsHook` propagates `depsTargetTarget` with the expectation that
it will effectively be `buildInputs` when the hook is itself used as a
`nativeBuildInput`. This propagates Gtk, which itself propagates the
default Dariwn SDK, making it effectively impossible to override the SDK
when a package depends on Gtk and uses `wrapGAppsHook`.
This rewrite implements the following improvements:
* Cross-compilation should be supported correctly (untested);
* Supports public and private frameworks;
* Supports SDK `libs`;
* Remaps instead of replacing extra (native) build inputs in the stdenv;
* Updates any Darwin framework references in `nix-support`; and
* It updates `xcodebuild` regardless of which input its in.
The implementation avoids recursion for performance reasons. Instead, it
enumerates transitive dependencies and walks the list from the leaf
packages backwards to the parent packages.
- update the hashes and tools needed to extract the bootstrap-tools archive
- unify the x64 and aarch64 unpack process
- unpin libxml for python-minimal
In the code examples for using `allowUnfreePredicate` and
`allowNonSourcePredicate`, we use `lib.getName` (which removes
the version) and not the local `getName` (which keeps the version).
`finalAttrs` is never the first argument. This should have been
called `prevAttrs` all along.
It confused me for a bit, because the callback _must not_ be strict
in `finalAttrs` (the first of _two_ parameters), or it will
recurse infinitely while trying to figure out what the attrNames are.
- unpin LLVM11. fix discrepancy with freshBootstrapTools and the tools
built on hydra. pinning the stdenv for the hydra build doesn't pin the
tools as the included packages are able to change.
- remove unused LLVM tools & libs which reduces the uncompressed and
compressed file sizes by more than 1/2. compressed tarball is now 40M
and uncompressed is around 200M
- add @loader_path/. to dylibs that reference other libs in the archive.
this is needed for libraries with re-exports.
- validate shared objects with @rpath references contain the reference
in lib
- add a test to verify that the @loader_path/ works for libc++ as it
re-exports libc++abi
The Darwin bootstrap currently requires curl, but it is not strictly
required. The bootstrap requires it for two things:
* Fetchers; and
* As a transitive dependency of llvm-manpages (via Sphinx).
For the fetchers, the bootstrap curl can be used. For hatch-vcs, the
dependency, its tests can be disabled. Doing this allows curl to be
dropped from the Darwin stdenv bootstrap.
These two commits make for a cleaner commit history and git blame than
https://github.com/NixOS/nixpkgs/pull/295105, where this refactor
was developed.
See its commit messages for details and design choices, esp. up to
and including 37f76fd4c3b23b324ab7ea06ef3b747338d2080f.
- merge libcxxabi into libcxx for LLVM 12, 13, 14, 15, 16, 17, and git.
- remove the link time workaround `-lc++ -lc++abi` from 58 packages as it is no longer required.
- fixes https://github.com/NixOS/nixpkgs/issues/166205
- provides alternative fixes for. https://github.com/NixOS/nixpkgs/issues/269548https://github.com/NixOS/nix/issues/9640
- pkgsCross.x86_64-freebsd builds work again
This change can be represented in 3 stages
1. merge libcxxabi into libcxx -- files: pkgs/development/compilers/llvm/[12, git]/{libcxx, libcxxabi}
2. update stdenv to account for merge -- files: stdenv.{adapters, cc.wrapper, darwin}
3. remove all references to libcxxabi outside of llvm (about 58 packages modified)
### merging libcxxabi into libcxx
- take the union of the libcxxabi and libcxx cmake flags
- eliminate the libcxx-headers-only package - it was only needed to break libcxx <-> libcxxabi circular dependency
- libcxx.cxxabi is removed. external cxxabi (freebsd) will symlink headers / libs into libcxx.
- darwin will re-export the libcxxabi symbols into libcxx so linking `-lc++` is sufficient.
- linux/freebsd `libc++.so` is a linker script `LINK(libc++.so.1, -lc++abi)` making `-lc++` sufficient.
- libcxx/default.nix [12, 17] are identical except for patches and `LIBCXX_ADDITIONAL_LIBRARIES` (only used in 16+)
- git/libcxx/defaul.nix does not link with -nostdlib when useLLVM is true so flag is removed. this is not much different than before as libcxxabi used -nostdlib where libcxx did not, so libc was linked in anyway.
### stdenv changes
- darwin bootstrap, remove references to libcxxabi and cxxabi
- cc-wrapper: remove c++ link workaround when libcxx.cxxabi doesn't exist (still exists for LLVM pre 12)
- adapter: update overrideLibcxx to account for a pkgs.stdenv that only has libcxx
### 58 package updates
- remove `NIX_LDFLAGS = "-l${stdenv.cc.libcxx.cxxabi.libName}` as no longer needed
- swift, nodejs_v8 remove libcxxabi references in the clang override
https://github.com/NixOS/nixpkgs/pull/292043
this equates to -ftrivial-auto-var-init=pattern
clang has removed support for -ftrivial-auto-var-init=zero and
are unlikely to re-add it, so use -ftrivial-auto-var-init=pattern
on both compilers if only to make behaviour more consistent
between the two.
add to pkgsExtraHardening's defaultHardeningFlags.
This is a small simplification of the control flow surrounding these cases. It should make it more obvious when each case happens, and also explicitly defines the current behaviour of --replace.
Example with `clangUseLLVM` which is the default when using `useLLVM`
```nix
config.replaceCrossStdenv = { buildPackages, baseStdenv }:
if baseStdenv.targetPlatform.useLLVM or false
then (buildPackages.stdenvAdapters.overrideCC baseStdenv buildPackages.llvmPackages_16.clangUseLLVM)
else baseStdenv;
```
The conditional necessary, otherwise the other sets(such as `pkgsCross.aarch64-multiplatform.llvmPackages`)
without `useLLVM` will use the stdenv without the necessary conditions to avoid infinite
recursion because of [targetLlvmLibraries](644b234e1c/pkgs/development/compilers/llvm/16/default.nix (L208))
usage.
[`replaceStdenv` is not used when cross-compiling](d77bda728d/pkgs/stdenv/cross/default.nix (L12-L13))
`replaceStdenv` uses an additional stage to replace the stdenv to avoid
infinite recursion and other issues but that should not be necessary for cross.
The aarch64 musl bootstrap tools are woefully outdated. Just getting
them to the point of being able to build new versions of themselves
required a number of hacks[1] that can be reverted once we have new
bootstrap tools, and before that it stdenv didn't even build for the
preceding three years.
[1]: https://github.com/NixOS/nixpkgs/pull/169764
So, following the script established by previous bootstrap tools
updates:
Files came from this Hydra build:
https://hydra.nixos.org/build/246470544
…which used nixpkgs revision dd5621df6d
to instantiate:
/nix/store/g480ass2vjmakaq03z7k2j95xnxh206a-stdenv-bootstrap-tools.drv
…and then built:
/nix/store/95lm0y33dayag4542s8bi83s31bw68dr-stdenv-bootstrap-tools
I downloaded these files from Hydra and prefetched them into the Nix
store with the following commands:
STOREPATH=95lm0y33dayag4542s8bi83s31bw68dr-stdenv-bootstrap-tools
OPTIONS="--option binary-caches https://cache.nixos.org --option trusted-public-keys cache.nixos.org-1:6NCHdD59X431o0gWypbMrAURkbJ16ZPMQFGspcDShjY="
nix --extra-experimental-features nix-command store prefetch-file \
file://$(nix --extra-experimental-features nix-command store add-file --name bootstrap-tools.tar.xz $(nix-store ${OPTIONS} -r /nix/store/${STOREPATH})/on-server/bootstrap-tools.tar.xz)
nix --extra-experimental-features nix-command store prefetch-file --executable \
file://$(nix --extra-experimental-features nix-command store add-path --name busybox $(nix-store ${OPTIONS} -r /nix/store/${STOREPATH})/on-server/busybox)
These commands produced the following output:
warning: you did not specify '--add-root'; the result might be removed by the garbage collector
Downloaded 'file:///nix/store/fm8ys5bb737j82xgyiciyzpcwmzyi9b2-bootstrap-tools.tar.xz' to '/nix/store/jml0gh0q2rnc9sgr87irz0jfbl0fq545-fm8ys5bb737j82xgyiciyzpcwmzyi9b2-bootstrap-tools.tar.xz' (hash 'sha256-ZY9IMOmx1VOn6uoFDpdJbTnPX59TEkrVCzWNtjQ8/QE=').
warning: you did not specify '--add-root'; the result might be removed by the garbage collector
Downloaded 'file:///nix/store/rn9fx55mw2s0qg90zdjxa9xn11nlmjpg-busybox' to '/nix/store/9qylz8gqll63pprwkwlyfs9g4zilak2m-rn9fx55mw2s0qg90zdjxa9xn11nlmjpg-busybox' (hash 'sha256-WuOaun7U5enbOy8SuuCo6G1fbGwsO16jhy/oM8K0lAs=').
I used the hashes from the output above to create the fetchurl
invocation which is part of this commit.
I then started the bootstrap with the following command:
nix --extra-experimental-features nix-command build -L -f . --arg localSystem '(import ./lib).systems.examples.aarch64-multiplatform-musl' hello
As @lovesegfault requested, here are the the sha256sums of all the
on-server components for extra verification:
sha256sum /nix/store/${STOREPATH}/on-server/*
…which produced the following output:
658f4830e9b1d553a7eaea050e97496d39cf5f9f53124ad50b358db6343cfd01 /nix/store/95lm0y33dayag4542s8bi83s31bw68dr-stdenv-bootstrap-tools/on-server/bootstrap-tools.tar.xz
20cdfecb084ddb6b6b958f2b78fd2cc1d9641632f81ec7d5a48fae0a963ad0fa /nix/store/95lm0y33dayag4542s8bi83s31bw68dr-stdenv-bootstrap-tools/on-server/busybox
This PR updates the bootstrap tarballs for riscv64-linux with new Hydra-generated ones.
Fixes#275848 (bootstrap assembler too old to build gcc 13).
I'll be following the script used in #151399, #168199, #183487, and #188334.
Files came from [this](https://hydra.nixos.org/build/246376732#tabs-summary) Hydra build, which used nixpkgs revision 160cedc144 to instantiate:
```
/nix/store/cpiajh4l83b08pynwiwkpxj53d78pcxr-stdenv-bootstrap-tools-riscv64-unknown-linux-gnu.drv
```
and then built:
```
/nix/store/8a92pj40awdw585mcb9dvm4nyb03k3q3-stdenv-bootstrap-tools-riscv64-unknown-linux-gnu
```
I downloaded these files from Hydra and prefetched them into the nix store with the following commands:
```
STOREPATH=8a92pj40awdw585mcb9dvm4nyb03k3q3-stdenv-bootstrap-tools-riscv64-unknown-linux-gnu
OPTIONS="--option binary-caches https://cache.nixos.org --option trusted-public-keys cache.nixos.org-1:6NCHdD59X431o0gWypbMrAURkbJ16ZPMQFGspcDShjY="
nix store prefetch-file \
file://$(nix store add-file --name bootstrap-tools.tar.xz $(nix-store ${OPTIONS} -r /nix/store/${STOREPATH})/on-server/bootstrap-tools.tar.xz)
nix store prefetch-file --executable \
file://$(nix store add-path --name busybox $(nix-store ${OPTIONS} -r /nix/store/${STOREPATH})/on-server/busybox)
```
These commands produced the following output:
```
Downloaded 'file:///nix/store/xs74wcxq3qc12plfc70fds8inbndhcfm-bootstrap-tools.tar.xz' to '/nix/store/3fal4gikp92013kac6rdmfbrch2s859b-xs74wcxq3qc12plfc70fds8inbndhcfm-bootstrap-tools.tar.xz' (hash 'sha256-0LxRd7fdafQezNJ+N2tuOfm0KEwgfRSts5fhP0e0r0s=').
Downloaded 'file:///nix/store/9ndpna6jrlac4y9fappdjm0sxx0g2bja-busybox' to '/nix/store/kb7wyy30y1gxcmdajljr26kxxac606qa-9ndpna6jrlac4y9fappdjm0sxx0g2bja-busybox' (hash 'sha256-OGO96QUzs2n5pGipn/V87AxzUY9OWKZl417nE8HdZIE=').
```
I used the hashes from the output above to create the `fetchurl` invocation which is part of this commit.
I then started the bootstrap with the following command:
```
nix build -L -f . --arg localSystem '(import ./lib).systems.examples.riscv64' hello
```
As @lovesegfault requested, here are the the `sha256sum`s of all the `on-server` components for extra verification:
```
sha256sum /nix/store/${STOREPATH}/on-server/*
```
which produced the following output:
```
d0bc5177b7dd69f41eccd27e376b6e39f9b4284c207d14adb397e13f47b4af4b /nix/store/8a92pj40awdw585mcb9dvm4nyb03k3q3-stdenv-bootstrap-tools-riscv64-unknown-linux-gnu/on-server/bootstrap-tools.tar.xz
65f9433abb598f63c932d33351b14f686551512b1cece1e64c2d0e76aa0ec52e /nix/store/8a92pj40awdw585mcb9dvm4nyb03k3q3-stdenv-bootstrap-tools-riscv64-unknown-linux-gnu/on-server/busybox
```
stdenv.cc may throw, e.g. in the case of pkgsCross.ghcjs where we must
not force it for the purpose of attribute accessing (`or` doesn't
implicitly tryEval…).
Regression introduced in 1a5bd697ad.
Before the change `pkgsLLVM` attributes were failing to pull in
`compiler-rt` on `x86_64-linux`:
$ nix build --no-link -f. pkgsLLVM.asciidoc-full
error:
error: attribute 'llvmPackages_13' missing
at pkgs/stdenv/adapters.nix:86:32:
85| inherit libcxx;
86| extraPackages = [ cxxabi pkgs.pkgsTargetTarget."llvmPackages_${lib.versions.major llvmLibcxxVersion}".compiler-rt ];
| ^
87| });
It happens because `pkgs.pkgsTargetTarget` are always empty for
cross-packages like `pkgsLLVM.`, `pkgsCross.*.` or
`--arg crossSystem '...'`.
Before the change `pkgsMusl.adobe-reader` was failing the interpreter:
$ nix-instantiate --eval --strict --expr 'with import ./. {}; builtins.tryEval pkgsMusl.adobe-reader'
error:
error: evaluation aborted with the following error message: 'unsupported platform for the pure Linux stdenv'
After the change `pkgsMusl.adobe-reader` returns catchable excation:
$ nix-instantiate --eval --strict --expr 'with import ./. {}; builtins.tryEval pkgsMusl.adobe-reader'
{ success = false; value = false; }
Noticed when was exploring `nixpkgs` for uncatchable evaluation errors.
Ideally those should only happen when there is a code bug in the use
site. In this case it's just a package with incompatible constraints for
`musl`.
Changed uncatchable `abort` to `throw`.
Aka `checkMeta` goes brrr.
Using the module system type checking works OK & generates good error messages.
The performance of using it however is terrible because of the value merging it does being very allocation heavy.
By implementing a very minimal type checker we can drastically improve the performance when nixpkgs is evaluated with `checkMeta = true`.
Without the change bootstrapTools build fails as:
https://cache.nixos.org/log/g5wyq9xqshan6m3kl21bjn1z88hx48rh-stdenv-bootstrap-tools.drv
error: install_name_tool: changing install names or rpaths can't be redone for: /nix/store/0hxg356h7lnl2hck23wrdpbib3cckx41-stdenv-bootstrap-tools/bin/tac (for architecture x86_64) because larger updated load commands do not fit (the program must be relinked, and you may need to use -headerpad or -headerpad_max_install_names)
this makes it a lot easier to create a modified stdenv with a
different set of defaultHardeningFlags and as a bonus allows us
to inject the correct defaultHardeningFlags into toolchain wrapper
scripts, reducing repetition.
while most hardening flags are arguably more of a compiler thing,
it works better to put them in bintools-wrapper because cc-wrapper
can easily refer to bintools but not vice-versa.
mkDerivation can still easily refer to either when it is constructed.
this also switches fortran-hook.sh to use the same defaults for
NIX_HARDENING_ENABLE as for C. previously NIX_HARDENING_ENABLE
defaults were apparently used to avoid passing problematic flags
to a fortran compiler, but this falls apart as soon as mkDerivation
sets its own NIX_HARDENING_ENABLE - cc.hardeningUnsupportedFlags
is a more appropriate mechanism for this as it actively filters
out flags from being used by the wrapper, so switch to using that
instead.
this is still an imperfect mechanism because it doesn't handle a
compiler which has both langFortran *and* langC very well - applying
the superset of the two's hardeningUnsupportedFlags to either
compiler's invocation. however this is nothing new - cc-wrapper
already poorly handles a langFortran+langC compiler, applying two
setup hooks that have contradictory options.
If a CMake target has a non-default LINKER_LANGUAGE set, CMake will
manually add the libraries it has detected that language's compiler as
linking implicitly. When it does this, it'll pass -Bstatic and
-Bdynamic options based on the vibes it gets from each such detected
library. This in itself isn't a problem, because the compiler
toolchain, or our wrapper, or something, seems to be smart enough to
ignore -Bdynamic for those libraries. But it does create a problem if
the compiler adds extra libraries to the linker command line after
that final -Bdynamic, because those will be linked dynamically. Since
our compiler is static by default, CMake should reset to -Bstatic
after it's done manually specifying libraries, but CMake didn't
actually know that our compiler is static by default. The fix for
that is to tell it, like so.
Until recently, this problem was difficult to notice, because it would
result binaries that worked, but that were dynamically linked. Since
e08ce498f0 ("cc-wrapper: Account for NIX_LDFLAGS and NIX_CFLAGS_LINK
in linkType"), though, -Wl,-dynamic-linker is no longer mistakenly
passed for executables that are supposed to be static, so they end up
created with a /lib interpreter path, and so don't run at all on
NixOS.
This fixes pkgsStatic.graphite2.
As reported in #241692, since the `llvmPackages` bump the
bootstrap-tools started failing to build due to a mismatch in LLVM
versions used to build certain tools.
By overlaying the imported package set to specify `llvmPackages`, we get
everything built with the expected LLVM version.
Provide a `runPhase` function which wraps the phase running action of
genericBuild. The new function can be used as an interface by `nix
develop`, i.e. `nix develop some#flake --build` may just call `runPhase
build`, which makes its behavior more consistent with `nix build`.
In preparation of fixing https://github.com/NixOS/nix/issues/6202
The 10.12 Libsystem is not located as a sub-attribute of
`darwin.apple_sdk_10_12`. This will be fixed as part of the SDK changes
planned for post-23.11. In the meantime, special case it so the adapter
can be used to change the deployment target.
This was taken from #264091 to use in the interim before that PR lands
(sometime after the release of 23.11). It allows different versions of
clang to link the same libc++, allowing dependencies to be linked when
they are built with a different version of clang than the stdenv.
This patch switches the CoreFoundation on x86_64-darwin from the open
source swift-corelibs-foundation (CF) to the system CoreFoundation.
This change was motivated by failures building packages for the current
staging-next cycle #263535 due to an apparent incompatibility with the
rpath-based approach to choosing CF or CoreFoundation and macOS 14. This
error often manifests as a crash with an Illegal Instruction.
For example, building aws-sdk-cpp for building Nix will fail this way.
https://hydra.nixos.org/build/239459417/nixlog/1
Application Specific Information:
CF objects must have a non-zero isa
Error Formulating Crash Report:
PC register does not match crashing frame (0x0 vs 0x7FF8094DD640)
Thread 0 Crashed:: Dispatch queue: com.apple.main-thread
0 CoreFoundation 0x7ff8094dd640 CF_IS_OBJC.cold.1 + 14
1 CoreFoundation 0x7ff8094501d0 CF_IS_OBJC + 60
2 CoreFoundation 0x7ff8093155e8 CFRelease + 40
3 ??? 0x10c7a2c61 s_aws_secure_transport_ctx_destroy + 65
4 ??? 0x10c87ba32 aws_ref_count_release + 34
5 ??? 0x10c7b7adb aws_tls_connection_options_clean_up + 27
6 ??? 0x10c596db4 Aws::Crt::Io::TlsConnectionOptions::~TlsConnectionOptions() + 20
7 ??? 0x10c2d249c Aws::CleanupCrt() + 92
8 ??? 0x10c2d1ff0 Aws::ShutdownAPI(Aws::SDKOptions const&) + 64
9 ??? 0x102d9bc6f main + 335
10 dyld 0x202f333a6 start + 1942
According to a [post][1] on the Apple developer forums, hardening was
added to CoreFoundation, and this particular message occurs when you
attempt to release an object it does not recognize as a valid CF object.
(Thank you to @lilyinstarlight for finding this post).
When I switched aws-sdk-cpp to link against CoreFoundation instead of
CF, the error went away. Somehow both libraries were being used. To
prevent dependent packages from linking the wrong CoreFoundation, it
would need to be added as a propagated build input.
Note that there are other issues related to mixing CF and CoreFoundation
frameworks. #264503 fixes an issue with abseil-cpp where it propagates
CF, causing issues when using a different SDK version. Mixing versions
can also cause crashes with Python when a shared object is loaded that
is linked to the “wrong” CoreFoundation.
`NIX_COREFOUNDATION_RPATH` is supposed to make sure the right
CoreFoundation is being used, but it does not appear to be enough on
macOS 14 (presumably due to the hardening). While it is possible to
propagate CoreFoundation manually, the cleaner solution is to make it
the default. CF remains available as `darwin.swift-corelibs-foundation`.
[1]: https://developer.apple.com/forums/thread/739355
- These new-cli commands can be used with `-f`, in which case they're
evaluated with pure evaluation disabled.
- Nix 2.4+ is not part of the condition; "flakes" is fully descriptive
and more relatable.
- Don't suggest that it only enables this variable.
- Just don't say too much.
This is a replacement for using `darwin.apple_sdk_<ver>.callPackage`.
Instead of injecting the required packages, it provides a stdenv adapter
that modifies the derivation’s build inputs to use the requested SDK
versions. This modification extends to any build inputs propagated to it
as well. The `callPackage` approach is not deprecated yet, but it is
expected that it will be eventually.
Note that this is an MVP. It should work with most packages, but it only
handles build inputs and also only handles frameworks. Once more SDKs
are added (after #229210 is merged) and the SDK structure is normalized,
it can be extended to handle any package in the SDK namespace.
Cross-compilation may or may not work. Any cross-related issues can be
addressed after #256590 is merged.
While there is no fetcher or builder (in nixpkgs) that takes an `md5` parameter,
for some inscrutable reason the nix interpreter accepts the following:
```nix
fetchurl {
url = "https://www.perdu.com";
hash = "md5-rrdBU2a35b2PM2ZO+n/zGw==";
}
```
Note that neither MD5 nor SHA1 are allowed by the syntax of SRI hashes.
curl needs to link against several frameworks, but building the
frameworks (directly or indirectly) depends on curl via fetchurl and
fetchFromGitHub. Break the infinite recursion by building the SDKs’
dependencies in the last stage of the stdenv bootstrap using the prior
stage’s fetchurl and fetchFromGitHub.
Make both ready for cross with prefixes
Currently
`pkgsCross.aarch64-multiplatform.mold.passthru.tests.{wrapped,adapter}`
fail with
```
Testing running the 'hello' binary which should be linked with 'mold'
Hello, world!
Checking for mold in the '.comment' section
No mention of 'mold' detected in the '.comment' section
The command was:
aarch64-unknown-linux-gnu-readelf -p .comment ...bin/hello
The output was:
String dump of section '.comment':
[ 0] GCC: (GNU) 12.3.0
```
This will allow buliding bootstrap tools for platforms with
non-default libcs, like *-unknown-linux-musl.
This gets rid of limitedSupportSystems/systemsWithAnySupport. There
was no need to use systemsWithAnySupport for supportDarwin, because it
was always equivalent to supportedSystems for that purpose, and the
only other way it was used was for determining which platforms to
build the bootstrap tools for, so we might as well use a more explicit
parameter for that, and then we can change how it works without
affecting the rest of the Hydra jobs.
Not affecting the rest of the Hydra jobs is important, because if we
changed all jobs to use config triples, we'd end up renaming every
Hydra job. That might still be worth thinking about at some point,
but it's unnecessary at this point (and would be a lot of work).
I've checked by running
nix-eval-jobs --force-recurse pkgs/top-level/release.nix
that the actual bootstrap tools derivations are unaffected by this
change, and that the only other jobs that change are ones that depend
on the hash of all of Nixpkgs. Of the other jobset entrypoints that
end up importing pkgs/top-level/release.nix, none used the
limitedSupportedSystems parameter, so they should all be unaffected as
well.
When specifying the `builder` attribute in `stdenv.mkDerivation`, this
will be effectively transformed into
builtins.derivation {
builder = stdenv.shell;
args = [ "-e" builder ];
}
This also means that `default-builder.sh` is never sourced and as a
result it's not guaranteed that `$NIX_ATTRS_SH_FILE` is set to a correct
location[1].
Also, we need to source `.attrs.sh` to source `$stdenv`. So, the
following is done now:
* If `$NIX_ATTRS_SH_FILE` points to a correct location, then use it.
Directly using `.attrs.sh` is problematic for `nix-shell(1)` usage
(see previous commit for more context), so prefer the environment
variable if possible.
* Otherwise, if `.attrs.sh` exists, then use it. See [1] for when this
can happen.
* If neither applies, it can be assumed that `__structuredAttrs` is
turned off and thus nothing needs to be done.
[1] It's possible that it doesn't exist at all - in case of Nix 2.3 or
it can point to a wrong location on older Nix versions with a bug in
`__structuredAttrs`.
Relying on `.attrs.sh` to exist in `$NIX_BUILD_TOP` is problematic
because that's not compatible with how `nix-shell(1)` behaves. It places
`.attrs.{json,sh}` into a temporary directory and makes them accessible via
`$NIX_ATTRS_{SH,JSON}_FILE` in the environment[1]. The sole reason that
`nix-shell(1)` still works with structured-attrs enabled derivations
is that the contents of `.attrs.sh` are sourced into the
shell before sourcing `$stdenv/setup` (if `$stdenv` exists) by `nix-shell`.
However, the assumption that two files called `.attrs.sh` and
`.attrs.json` exist in `$NIX_BUILD_TOP` is wrong in an interactive shell
session and thus an inconsistency between shell debug session and actual
builds which can lead to unexpected problems.
To be precise, we currently have the following problem: an expression
like
with import ./. {};
runCommand "foo" { __structuredAttrs = true; foo.bar = [ 1 2 3 ]; }
''
echo "''${__structuredAttrs@Q}"
touch $out
''
prints `1` in its build-log. However when building interactively in a
`nix-shell`, it doesn't.
Because of that, I'm considering to propose a full deprecation of
`$NIX_BUILD_TOP/.attrs.{json,sh}`. A first step is to only mention the
environment variables, but not the actual paths anymore in Nix's
manual[2]. The second step - this patch - is to fix nixpkgs' stdenv
accordingly.
Please note that we cannot check for `-e "$NIX_ATTRS_JSON_FILE"` because
certain outdated Nix minors (that are still in the range of supported
Nix versions in `nixpkgs`) have a bug where `NIX_ATTRS_JSON_FILE` points
to the wrong file while building[3].
Also, for compatibility with Nix 2.3 which doesn't provide these
environment variables at all we still need to check for the existence of
.attrs.json/.attrs.sh here. As soon as we bump nixpkgs' minver to 2.4,
this can be dropped.
Finally, dropped the check for ATTRS_SH_FILE because that was never
relevant. In nix#4770 the ATTRS_SH_FILE variable was introduced[4] and
in a review iteration prefixed with NIX_[5]. In other words, these
variables were never part of a release and you'd only have this problem
if you'd use a Nix from a git revision of my branch from back then. In
other words, that's dead code.
[1] https://github.com/nixos/nix/pull/4770#issuecomment-834718851
[2] https://github.com/NixOS/nix/pull/9032
[3] https://github.com/NixOS/nix/issues/6736
[4] 3944a120ec
[5] 27ce722638
Without this, you get error messages during the install phase along the
lines of: "file RPATH_CHANGE could not write new RPATH:".
This is unsurprising because the static binaries do not have any dynamic
linker and thus, no runpath to rewrite either.
Tell cmake it doesn't need to do RPATH manipulation by passing
cmakeFlags.
While we're here, I also renamed `finalAttrs` to `args` and fixed the
indentation; this improves consistency with the surrounding code and
eliminates a point of confusion: because it was named `finalAttrs` I
presumed I should be able to influence it with an overrideAttrs setting
dontAddStaticConfigureFlags, but this turns out not to be possible;
adding prevAttrs as well doesn't work because of a limitation of
overrideAttrs whereby it gives an infinite recursion if the set of
attribute keys being returned depends on finalAttrs.
Signed-off-by: Peter Waller <p@pwaller.net>
Fixes `pkgsCross.musl64.llvmPackages_16.clang.cc` on `x86_64-linux`,
which used to fail with `/bin/sh: clang-tblgen: not found`.
Same hack is used in other projects:
https://github.com/search?q=%2FCMAKE_CROSSCOMPILING_EMULATOR.%2B%5C%2Fusr%5C%2Fbin%5C%2Fenv%2F+NOT+is%3Afork&type=code
Comment from 30435a9d0f/build/cmake/HostLinuxToolchain.cmake (L64)
> Required to run host Linux executables during the build itself.
> An example would be https://gitub.com/KhronosGroup/Vulkan-Loader and
> its "asm_offset" program.
>
> NOTE: Alternatives have been tried unsuccessfully, i.e.:
>
> With $(set CMAKE_CROSSCOMPILING_EMULATOR), the build fails because
> the CMake ninja/Make script tries to find the executable in the current
> path, as in:
>
> [3/16] Generating gen_defines.asm
> FAILED: loader/gen_defines.asm
> cd /tmp/cc/build-Vulkan-Loader/loader && asm_offset GAS
> /bin/sh: asm_offset: command not found
> ninja: build stopped: subcommand failed.
>
> With $(set CMAKE_CROSSCOMPILING_EMULATOR ""), the build fails because
> the shell cannot find the "" program as in:
>
> [3/16] Generating gen_defines.asm
> FAILED: loader/gen_defines.asm
> cd /tmp/cc/build-Vulkan-Loader/loader && "" /tmp/cc/build-Vulkan-Loader/loader/asm_offset GAS
> /bin/sh: : command not found
> ninja: build stopped: subcommand failed.
>
> It seems that the root of the problem comes from how the CMake function
> cmCustomCommandGenerator::GetArgc0Location() computes the target
> executable's location. At this point it's unclear whether this is a CMake
> bug or a feature.
Risicle discovered this hack.
Co-authored-by: Robert Scott <code@humanleg.org.uk>
In the default `fixupPhase` the output of `substituteAllStream` is
streamed to setup-hook.
`stdenv.cc.bintools.overrideAttrs { NIX_DEBUG = 6; }`
With `NIX_DEBUG` contains:
```
@expandResponseParams@ -> /nix/store/yl01rd58vp4m8bbhkihpk132cprfmx6f-expand-response-params/bin/expand-response-params
...
```
To work around intermitent build failures with clang 16, the stdenv
attempted to pass arguments on the command-line on newer versions of
macOS. Unfortunately, the larger `ARG_MAX` is still not large enough to
build qtwebengine. This commit reverts the `NIX_CC_NO_RESPONSE_FILE`
logic in the stdenv. The changes to cc-wrapper in #245282 are needed for
clang 16 to prevent the above-mentioned build failures.
