HAVE_ZLIB has to be defined in mysqlnd.h for compression support to be
turned on, but the configure script doesn't actually define it even
when zlib is available.
NoMachine removes each old release as soon as a new one is available,
resulting in failed downloads. Thankfully, the Internet Archive
provides backups of old downloads, so we can use it as a fallback.
The couchbase extension depends on the igbinary PECL which needs to be
loaded and loaded prior to it. It also seems like the pcs extension
isn't actually needed - it at least builds and loads without it. Since
the pcs extension dependency was the reason couchbase didn't build on
PHP 7.4 it now does, so let's unmark it broken.
The pcs extension fails to load at runtime with PHP 7.3, so let's mark
it broken from 7.3 onwards. It also depends on the tokenizer internal
extension.
Fix an issue brought up in #86463, where the apcu_bc extension isn't
loaded correctly since it produces a .so with a different name than
the extension name. Also, the apcu extension has to be loaded and
loaded prior to loading this extension.
Some PECLs depend on other PECLs and, like internal PHP extension
dependencies, need to be loaded in the correct order. This makes this
possible by adding the argument "peclDeps" to buildPecl, which adds
the extension to buildInputs and is treated the same way as
internalDeps when the extension config is generated.
With recent work done on the PHP packaging, the PHP executable is by
default a wrapper script. Darwin doesn't like scripts in the shebang
of another script, so we have to wrap arcanist, feeding the script as
an argument to PHP instead. See #86881 and #23018.
Instead of using two different php packages in php-packages.nix, one
wrapper and one unwrapped, simply use the wrapper and use its
"unwrapped" attribute when necessary. Also, get rid of the packages
and extensions attributes from the base package, since they're no
longer needed.
Since the introduction of php.unwrapped there's no real need for the
phpXXbase attributes, so let's remove them to lessen potential
confusion and clutter. Also update the docs to make it clear how to
get hold of an unwrapped PHP if needed.
Some extensions depend on other extensions. Previously, these had to
be added manually to the list of included extensions, or we got a
cryptic error message pointing to strings-with-deps.nix, which wasn't
very helpful. This makes sure all required extensions are included in
the set from which textClosureList chooses its snippets.
Rework withExtensions / buildEnv to handle currently enabled
extensions better and make them compatible with override. They now
accept a function with the named arguments enabled and all, where
enabled is a list of currently enabled extensions and all is the set
of all extensions. This gives us several nice properties:
- You always get the right version of the list of currently enabled
extensions
- Invocations chain
- It works well with overridden PHP packages - you always get the
correct versions of extensions
As a contrived example of what's possible, you can add ImageMagick,
then override the version and disable fpm, then disable cgi, and
lastly remove the zip extension like this:
{ pkgs ? (import <nixpkgs>) {} }:
with pkgs;
let
phpWithImagick = php74.withExtensions ({ all, enabled }: enabled ++ [ all.imagick ]);
phpWithImagickWithoutFpm743 = phpWithImagick.override {
version = "7.4.3";
sha256 = "wVF7pJV4+y3MZMc6Ptx21PxQfEp6xjmYFYTMfTtMbRQ=";
fpmSupport = false;
};
phpWithImagickWithoutFpmZip743 = phpWithImagickWithoutFpm743.withExtensions (
{ enabled, all }:
lib.filter (e: e != all.zip) enabled);
phpWithImagickWithoutFpmZipCgi743 = phpWithImagickWithoutFpmZip743.override {
cgiSupport = false;
};
in
phpWithImagickWithoutFpmZipCgi743
Make buildEnv take earlier overridden values into account by
forwarding all arguments (a merge of generic's arguments, all previous
arguments and the current arguments) to the next invocation of
buildEnv.
Make all arguments to a PHP build overridable; i.e, both configuration
flags, such as valgrindSupport, and packages, such as valgrind:
php.override { valgrindSupport = false; valgrind = valgrind-light; }
This applies to packages built by generic and buildEnv/withExtensions;
i.e, it works with both phpXX and phpXXBase packages.
The following changes were also made to facilitate this:
- generic and generic' are merged into one function
- generic now takes all required arguments for a complete build and
is meant to be called by callPackage
- The main function called from all-packages.nix no longer takes all
required arguments for a complete build - all arguments passed to it
are however forwarded to the individual builds, thus default
arguments can still be overridden from all-packages.nix
This implements the override pattern for builds done with buildEnv, so
that we can, for example, write
php.override { fpmSupport = false; }
and get a PHP package with the default extensions enabled, but PHP
compiled without fpm support.
mkDerivation uses the dev output in buildInputs if it exits, hence the
php-with-extensions package was never built or put into the path of
packages dependent on it during build. With this fix, the php packages
built with buildEnv or withExtensions don't have any dev outputs;
packages which need the dev output can refer to the phpXXbase packages
instead.
This provides a means to build a PHP package based on a list of
extensions from another.
For example, to generate a package with all default extensions
enabled, except opcache, but with ImageMagick:
php.withExtensions (e:
(lib.filter (e: e != php.extensions.opcache) php.enabledExtensions)
++ [ e.imagick ])
Fix extensions that were broken by the extension refactoring and
remove pthreads and pinba, which had asserts which broke evaluation,
were in need of refactoring and of dubious value.
Make mkExtension put headers in the dev output and use them, instead of
a different part of the current source tree, when referring to another
extension by using internalDeps.
This means external extensions can be built against the internal ones.
This means php packages can now refer to other php packages by looking
them up in the php.packages attribute and gets rid of the internal
recursive set previously defined in php-packages.nix. This also means
that in applications where previously both the php package and the
corresponding version of the phpPackages package set had to be
specified, the php package will now suffice.
This also adds the phpWithExtensions parameter to the
php-packages.nix, which can be used by extensions that need a fully
featured PHP executable.
A slight rewrite of buildEnv which:
1. Makes buildEnv recursively add itself to its output, so that it can
be accessed from any php derivation.
2. Orders the extension text strings according to their internalDeps
attribute - dependencies have to be put before dependants in the
php.ini or they will fail to load due to missing symbols.
The tests for many of the extensions run just fine, for some a small
portion fail. This runs the tests by default and disables the tests
extensions with any failing tests.
This moves yet more extensions from the base build to
phpPackages.ext. Some of the extensions are a bit quirky and need
patching for this to work, most notably mysqlnd and opcache.
Two new parameters are introduced for mkExtension - internalDeps and
postPhpize. internalDeps is used to specify which other internal
extensions the current extension depends on, in order to provide them
at build time. postPhpize is for when patches and quirks need to be
applied after running phpize.
Patch notes:
- For opcache, older versions of PHP have a bug where header files are
included in the wrong order.
- For mysqlnd, the config.h is never included, so we include it in the
main header file, mysqlnd.h. Also, the configure script doesn't add
the necessary library link flags, so we add them to the variable
configure should have added them to.
Running haproxy with "DynamicUser = true" doesn't really work, since
it prohibits specifying a TLS certificate bundle with limited
permissions. This revives the haproxy user and group, but makes them
dynamically allocated by NixOS, rather than statically allocated. It
also adds options to specify which user and group haproxy runs as.
The way ruby loads gems and keeps track of their paths seems to not
always work very well when the gems are accessed through
symlinks. Ruby will then complain that the same files are loaded
multiple times; it relies on the file's full path to determine whether
the file is loaded or not.
This adds an option to simply copy all gem files into the environment
instead, which gets rid of this issue, but may instead result in major
file duplication.
Includes the following fixes for Darwin:
- Build an application bundle
- Don't install the .desktop file and icons
- Get rid of unused cmake variables
- Don't list avahi as a dependency, since it's not used
This gets rid of a few patches that aren't needed any more, adds one
to make the tests work again and updates the MacOS patch. It also
introduces two builds - one with the Qt application and one without.
The patch to get the tests working will be submitted upstream and
hopefully not be needed for future releases.
GitLab Shell now has the go.mod and go.sum files in the root of the
repo; the go subdirectory has been removed and all the code in it has
been moved up to the root.
For some reason this untagged commit is the one referred to in the
main repository; this might be a mistake, but we'll have to package it
for now to follow upstream.
Kim Lindberger