cc-wrapper has essentially two separate codepaths: the `gccForLibs`
codepath, used only by non-gcc (i.e. clang) compilers, and the
"other" codepath.
This PR allows non-clang compilers to opt-in to the `gccForLibs`
codepath (off by default). To allow this, a new parameter
`ccForLibs` is exposed, since it would be extremely confusing for
gcc to be able to use `gccForLibs` but not do so by default.
Fixes "No such file or directory" error when running prefetch-npm-deps
in its default mode (print hashes, for update scripts etc),
in an environment that doesn't have a `nix` binary (for `nix hash`).
otherwise, eval fails when the experimental no-url-literals feature is activated
unquoted urls are discouraged after https://github.com/NixOS/rfcs/pull/45
not all linkers have a ld binary in bin
also note the '${ld:-}' which allows users to set the ld path with a env
var
> '${foo:-val}' $foo, or val if unset (or null)
Used in cases where you need to get the vendor of a target. Such as when
you need to perform dependency resolution outside of Cargo (eg in
Kolloch's crate2nix).
Currently cargo-setup-hook instructs the builder upon cargoSha256 or
cargoHash being out-of-date compared to the Cargo.lock file.
The instructions can be simplified a bit, because nowadays it is fine to
keep a hash empty, instead of filling it with
`0000000000000000000000000000000000000000000000000000`.
Nix nowadays outputs SRI hashes, which should usually be placed in
`cargoHash` instead of `cargoSha256`, but the instructions are still
only referring to `cargoSha256`.
Lastly, the output of Nix doesn't include `got: sha256: ` anymore, as it
now outputs `got: sha256-XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX=`.
It would be nice to make it clear that the trailing `=` is important as
well, so the full example SRI hash is mentioned.
The current `//` override to `agda.passthru.tests` is non-recursive so
it destroys everything else under `passthru`, and furthermore does not
go through `mkDerivation` so that we end up with different values for
`agda.tests` and `agda.passthru.tests`.
Fix it by moving the `allPackages` test to the definition of
`withPackages`.
Upstream changes:
singularity 3.8.7 (the legacy) -> apptainer 1.1.3 (the renamed) / singularity 3.10.4 (Sylabs's fork)
Build process:
* Share between different sources
* Fix the sed regexp to make defaultPath patch work
* allowGoReference is now true
* Provied input parameter removeCompat (default to false)
that removes the compatible "*singularity*" symbolic links
and related autocompletion files when projectName != "singularity"
* Change localstatedir to /var/lib
* Format with nixpkgs-fmt
* Fix the defaultPath patching
and use it instead of the `<executable> path` config directive
deprecated in Apptainer
* Provide dependencies for new functionalities such as
squashfuse (unprivileged squashfs mount)
* Provide an attribute `defaultPathInputs` to override
prefix of container runtime default PATH
NixOS module programs.singularity:
* Allow users to specify packages
* Place related directories to /var/lib
* Format with nixpkgs-fmt
singularity-tools:
* Allow users to specify packages
* Place related directories to /var/lib when building images in VM
See docs.
Follow-up work:
- Existing packages should be converted
- `defaultPkgConfigPackages` should assert on `meta.pkgConfigModules`
and let `tests.pkg-config` alone test the build results.
CC @sternenseemann
Co-authored-by: Robert Hensing <roberth@users.noreply.github.com>
This reverts commit 8c80bd08b7
("build-support/cc-wrapper: pass in non-existent --sysroot= to untangle
from libc").
This change was good in spirit: we caught a few genuine problems with
`scons` based packages (`godot`, `fluxus`) and unexpected `-idirafter`
includes in various boot loadres (`ipxe`, wimboot`):
https://github.com/NixOS/nixpkgs/pull/210004#issuecomment-1407162693
Unfortunately `--sysroot=` also has a negative impact on libary search
order for DT_NEEDED libraries and RUNPATHs of linked libraries. This
unexpectedly broke `dmd`, `d-seams`, `llvmPackages_rocm.compiler-rt`).
An interesting case of unexpected breakage is `usbmuxd2` where the bug
exposed incomplete library move on `libstdc++fs` in `gcc`.
The library breakage is very non-intuitive (on top of already unusual
layout of `cc-wrapper` driver). Let's revert this change for now.
Once it lands we can undo `--sysroot=/` workarounds merged for
`staging-next`.
After https://github.com/NixOS/nixpkgs/pull/210004 `usbmuxd2` started
failing to build as:
usbmuxd2-unstable> .../ld: cannot find -lstdc++fs: No such file or directory
usbmuxd2-unstable> clang-11: error: linker command failed with exit code 1 (use -v to see invocation)
This started happening because #210004 exposed a long-standing bug in
`gcc` derivation: `cc.lib` is missing `libstdc++fs` library:
$ find $(nix-build --no-link -A stdenv.cc.cc.lib) | fgrep libstdc | unnix
/<<NIX>>/gcc-11.3.0-lib/lib/libstdc++fs.la
/<<NIX>>/gcc-11.3.0-lib/lib/libstdc++.la
/<<NIX>>/gcc-11.3.0-lib/lib/libstdc++.so.6.0.29
/<<NIX>>/gcc-11.3.0-lib/lib/libstdc++.so
/<<NIX>>/gcc-11.3.0-lib/lib/libstdc++.so.6
It was not moved from `cc.out` output:
$ find $(nix-build --no-link -A stdenv.cc.cc) | fgrep libstdc | unnix
/<<NIX>>/gcc-11.3.0/lib/libstdc++.a
/<<NIX>>/gcc-11.3.0/lib/libstdc++fs.a
This change adds `cc` library lookup path back to `staging-next` until
`gcc` is fixed.`
FreeBSD doesn't use LLVM's cxxabi implementation, for backwards
compatibility reasons. Software expects the libcxxrt API when
building on FreeBSD. This fixes the build of
pkgsCross.x86_64-freebsd.boost.
Trivial conflict in release notes, except that the xml/docbook parts
are horrible for (semi-)automatic conflict resolution.
Fortunately that's generated anyway.
With all libcs I'm aware of, libdl is now either empty (Glibc, musl,
uclibc, illumos), a symlink to libc or equivalent (Apple), or does not
exist (FreeBSD, NetBSD). So explicitly linking libdl now does nothing
for the former platforms, and breaks the build for the latter
platforms.
With this patch I've removed -ldl from all overridden linker flags for
all free packages in Nixpkgs. Everything still seems to build.
checkInputs used to be added to nativeBuildInputs. Now we have
nativeCheckInputs to do that instead. Doing this treewide change allows
to keep hashes identical to before the introduction of
nativeCheckInputs.
fetchPypi doesn't use python under the hood and doesn't need to be tied
to a specific version of python. Moving it to top level makes it more
consistent with other fetchers and makes code generation easier.
the output made me think that the test failed but did not fail
testBuildFailure-helloDoesNotFail> Checking /nix/store/x6403x1llpk00i59cmr96iy92mx1f7hb-hello-2.12.1/testBuildFailure.log
testBuildFailure-helloDoesNotFail> testBuildFailure: The builder did not fail, but a failure was expected!
A call to getAllOutputNames was added as part of the structuredAttrs
effort, but this script does not source `setup.sh`, so I repeat the
definition here.
Should be a bit quicker than `source setup.sh`.
Closes#16182
This improves the error message
Error: _assignFirst found no valid variant!
which occurred when the set of outputs was not sufficient to set
the various outputDev, outputBin, etc variables. Specifically, this
would mean that "out" is not among the outputs, which is valid for
a derivation.
This changes the message to something like
error: _assignFirst: could not find a non-empty variable to assign to outputDev. The following variables were all unset or empty: dev out.
If you did not define an "out" output, make sure to define all the specific required outputs: define an output for one of the unset variables.
While this isn't a full explanation of what stdenv can and can not do,
I think it's vast improvement over the 0 bits of information that it
used to provide. This at least gives a clue as to what's going on, and
even suggests a fix, although probably multiple such fixes are required
in an instance where someone starts with a no-out derivation from scratch
(and decide to persist).
The auto-patchelf python script assembles a list of
library (so=shared object) file names and their paths.
This helps speed up the discovery of
library files later when patching elf files.
As further optimization, if a symlink points to a library file,
the script uses the resolved path and file name.
However, this produces a broken list entry if the
symlink's target name doesn't match the symlink's name.
A symptom of the bug, affecting the `tsm-client` package,
is fixed in https://github.com/NixOS/nixpkgs/pull/172372 .
The commit at hand stops resolving symlinks if
the target name differs from the symlink's name.
The commit has been authored by
layus (Guillaume Maudoux <layus.on@gmail.com>)
in pull request comment
https://github.com/NixOS/nixpkgs/pull/172372#issuecomment-1194687183
Fixes#204051. I have tried this on the reproducer stated in the ticket.
```
[nix-develop]$ $(nix-build -I nixpkgs=/home/shana/programming/nixpkgs --no-out-link)/tests/foo
running 1 test
test check_module_name ... ok
test result: ok. 1 passed; 0 failed; 0 ignored; 0 measured; 0 filtered out; finished in 0.00s
```
nixpkgs-release-checks> Nixpkgs is not allowed to use <nixpkgs> to refer to itself.
nixpkgs-release-checks> The offending files: /nix/store/mpwvmd4mr2nx1ymw1b2r44lm34ff2vs7-source/pkgs/build-support/deterministic-uname/default.nix
Gcc does not allow `-march=` on PowerPC:
https://gcc.gnu.org/onlinedocs/gcc-12.2.0/gcc/RS_002f6000-and-PowerPC-Options.html#RS_002f6000-and-PowerPC-Options
Instead, `-mcpu=` should be used to set the minimum instruction set
and `-mtune=` is used to optimize instruction scheduling for a
specific processor. Both flags take the same set of valid values,
which includes `native`.
This commit causes `isGccArchSupported` to return `false` for PowerPC
targets so we never pass an `-march=` flag, since that will always be
rejected by gcc.
I would like to add an extra `gcc` build step during linux bootstrap
(https://github.com/NixOS/nixpkgs/issues/208412). This makes it early
bootstrap compiler linked and targeted against `bootstrapTools` `glibc`
including it's headers.
Without this change `gcc`'s spec files always prefer `bootstrapTools` `glibc`
for header search path (passed in as --with-native-system-header-dir=). We'can't
override it with:
- `-I` option as it gets stacked before gcc-specific headers, we need to keep
glibc headers after gcc as gcc cleans namespace up for C standard by using
#include_next and by undefining system macros.
- `-idirafter` option as it gets appended after existing `glibc`-includes
This `--sysroot=/nix/store/does/not/exist` hack allows us to remove existing
`glibc` headers and add new ones with `-idirafter`.
We use `cc-cflags-before` instead of `libc-cflags` to allow user to define
their own `--sysroot=` (like `firefox` does).
To keep it working prerequisite cross-symlink in gcc.libs is required:
https://github.com/NixOS/nixpkgs/pull/209153
to not unnecessarily print
```
hello> ++ type -f strip
hello> strip is /nix/store/xzqzq99kbbkad23s2bf4bbrjppg6x0vj-bootstrap-stage2-gcc-wrapper-/bin/strip
```
printing what the strip command is is already done in stripDirs
`echo "stripping (with command $cmd and flags $stripFlags) in $paths"`
Inside `nix-shell`, `TMPDIR` (used by `mktemp`) is set to
`/run/user/<uid>` which is usually a tmpfs stored in RAM.
When fetching a large dotnet deps tree to this tmpdir from a
nix-shell (e.g. via `btcpayserver/update.sh`), this can easily exceed
system RAM and `fetch-deps` fails.
mktemp arg `-t` is deprecated and can be omitted.
Switches that gnatmake needs to pass to gcc must be given as
"-cargs <gcc_switches>" after at least the files to compile (see the
gnatmake docs for all the subtleties). This commit makes that happen,
in a way that triggers rebuilds only for things that actually depend on
GNAT, and not the other compilers contained in GCC.
For reasons explained in the commit contents, in order to build the
native gnat package for x86_64-darwin, the native gnatboot package for
x86_64-darwin must have access to both the Clang integrated assembler
and the cctools GNU assembler for that platform. This commit creates a
package with both of those assemblers that x86_64-darwin gnatboot can
then be wrapped with.
Previously srcOnly would not work for e.g.
`srcOnly haskell.compiler.ghcjs`, as the custom `installPhase` would be
skipped if `dontInstall` is set. Consequently, we need to ensure that
the two skippable phases we need will always be executed.
the hook doesn't output the following if $prefix doesn't exist
find: '/nix/store/...': No such file or directory
some other hooks that do this: prune-libtool-files.sh, audit-tmpdir.sh separate-debug-info.sh
This change switches to using GCC 11 by default on aarch64-linux, as well as passing `-lgcc` to the linker, per #201485.
See #201254 and #208412 for wider context on the issue.
This reverts commit b6fc00b8f4.
Rust 1.66.0 contains a fix for libiconv being linked unconditionally on macOS, but this only applies to packages that don't depend on older versions of `libc`.
For now, let's go back to including libiconv in `buildInputs` by default for packages that use `buildRustPackage`. As packages bump their `libc` versions, we can eventually stop including it by default, and manually add it where needed.
The go toolchain calls pkg-config using "--" to separate flags and
packages. For example:
pkg-config --cflags -- zlib
This breaks if addFlags are appended. This commit therefore prepends the
addFlags to the command.
Single quotes must be used when using graves in a quote, else Bash will interpret it as a command to run.
Thanks to Dimitri for catching this (8e651111b7 (commitcomment-92100762)).
glib expression is messy enough as is.
Also rename the `glib-schema-to-var` argument to `schemaIdToVariableMapping` to better match Nixpkgs coding style.
