php85: init at 8.5.2

[piotrkwiecinski]

Supersedes #422308

Notes:
opcache is now built-in. I run a test using var_dump(function_exists('opcache_get_status'));

uri extension works. Tested using:

$uri = Uri\Rfc3986\Uri::parse('https://example.com/path?query=value#fragment');  
var_dump($uri->getScheme());

Things done

• Built on platform:
  • x86_64-linux
  • aarch64-linux
  • x86_64-darwin
  • aarch64-darwin
• Tested, as applicable:
  • NixOS tests in nixos/tests.
  • Package tests at passthru.tests.
  • Tests in lib/tests or pkgs/test for functions and "core" functionality.
• Ran nixpkgs-review on this PR. See nixpkgs-review usage.
• Tested basic functionality of all binary files, usually in ./result/bin/.
• Nixpkgs Release Notes
  • Package update: when the change is major or breaking.
• NixOS Release Notes
  • Module addition: when adding a new NixOS module.
  • Module update: when the change is significant.
• Fits CONTRIBUTING.md, pkgs/README.md, maintainers/README.md and other READMEs.

---

Add a 👍 reaction to pull requests you find important.

[piotrkwiecinski]

@Al2Klimov would you test it. There are still some failures as we need to merge other PRs for extension upgrades, but core should work.

[piotrkwiecinski]

nixpkgs-review result

Generated using nixpkgs-review.

Command: nixpkgs-review pr 481104
Commit: 1c614d75004b9eb1ecda6ddeb959c4f544403de5

---

x86_64-linux

❌ 1 package failed to build:

• php85Extensions.imagick

✅ 171 packages built:

• php82Extensions.rrd
• php83Extensions.rrd
• phpExtensions.rrd (php84Extensions.rrd)
• php85
• php85Extensions.amqp
• php85Extensions.apcu
• php85Extensions.apcu.dev
• php85Extensions.ast
• php85Extensions.bcmath
• php85Extensions.bcmath.dev
• php85Extensions.blackfire
• php85Extensions.bz2
• php85Extensions.bz2.dev
• php85Extensions.calendar
• php85Extensions.calendar.dev
• php85Extensions.ctype
• php85Extensions.ctype.dev
• php85Extensions.curl
• php85Extensions.curl.dev
• php85Extensions.dba
• php85Extensions.dba.dev
• php85Extensions.decimal
• php85Extensions.dom
• php85Extensions.dom.dev
• php85Extensions.ds
• php85Extensions.enchant
• php85Extensions.enchant.dev
• php85Extensions.event
• php85Extensions.excimer
• php85Extensions.exif
• php85Extensions.exif.dev
• php85Extensions.ffi
• php85Extensions.ffi.dev
• php85Extensions.fileinfo
• php85Extensions.fileinfo.dev
• php85Extensions.filter
• php85Extensions.filter.dev
• php85Extensions.ftp
• php85Extensions.ftp.dev
• php85Extensions.gd
• php85Extensions.gd.dev
• php85Extensions.gettext
• php85Extensions.gettext.dev
• php85Extensions.gmp
• php85Extensions.gmp.dev
• php85Extensions.gnupg
• php85Extensions.iconv
• php85Extensions.iconv.dev
• php85Extensions.imap
• php85Extensions.inotify
• php85Extensions.intl
• php85Extensions.intl.dev
• php85Extensions.ldap
• php85Extensions.ldap.dev
• php85Extensions.luasandbox
• php85Extensions.mailparse
• php85Extensions.maxminddb
• php85Extensions.mbstring
• php85Extensions.mbstring.dev
• php85Extensions.memcached
• php85Extensions.meminfo
• php85Extensions.memprof
• php85Extensions.mongodb
• php85Extensions.msgpack
• php85Extensions.mysqli
• php85Extensions.mysqli.dev
• php85Extensions.mysqlnd
• php85Extensions.mysqlnd.dev
• php85Extensions.oci8
• php85Extensions.openssl
• php85Extensions.openssl.dev
• php85Extensions.opentelemetry
• php85Extensions.pcntl
• php85Extensions.pcntl.dev
• php85Extensions.pcov
• php85Extensions.pdlib
• php85Extensions.pdo
• php85Extensions.pdo.dev
• php85Extensions.pdo_dblib
• php85Extensions.pdo_dblib.dev
• php85Extensions.pdo_mysql
• php85Extensions.pdo_mysql.dev
• php85Extensions.pdo_oci
• php85Extensions.pdo_odbc
• php85Extensions.pdo_odbc.dev
• php85Extensions.pdo_pgsql
• php85Extensions.pdo_pgsql.dev
• php85Extensions.pdo_sqlite
• php85Extensions.pdo_sqlite.dev
• php85Extensions.pgsql
• php85Extensions.pgsql.dev
• php85Extensions.pinba
• php85Extensions.posix
• php85Extensions.posix.dev
• php85Extensions.protobuf
• php85Extensions.pspell
• php85Extensions.rdkafka
• php85Extensions.readline
• php85Extensions.readline.dev
• php85Extensions.redis
• php85Extensions.relay
• php85Extensions.rrd
• php85Extensions.session
• php85Extensions.session.dev
• php85Extensions.shmop
• php85Extensions.shmop.dev
• php85Extensions.simplexml
• php85Extensions.simplexml.dev
• php85Extensions.smbclient
• php85Extensions.snmp
• php85Extensions.snmp.dev
• php85Extensions.snuffleupagus
• php85Extensions.soap
• php85Extensions.soap.dev
• php85Extensions.sockets
• php85Extensions.sockets.dev
• php85Extensions.sodium
• php85Extensions.sodium.dev
• php85Extensions.spx
• php85Extensions.sqlite3
• php85Extensions.sqlite3.dev
• php85Extensions.sqlsrv
• php85Extensions.ssh2
• php85Extensions.systemd
• php85Extensions.sysvmsg
• php85Extensions.sysvmsg.dev
• php85Extensions.sysvsem
• php85Extensions.sysvsem.dev
• php85Extensions.sysvshm
• php85Extensions.sysvshm.dev
• php85Extensions.tidy
• php85Extensions.tidy.dev
• php85Extensions.tokenizer
• php85Extensions.tokenizer.dev
• php85Extensions.uuid
• php85Extensions.uv
• php85Extensions.vld
• php85Extensions.wikidiff2
• php85Extensions.xdebug
• php85Extensions.xml
• php85Extensions.xml.dev
• php85Extensions.xmlreader
• php85Extensions.xmlreader.dev
• php85Extensions.xmlwriter
• php85Extensions.xmlwriter.dev
• php85Extensions.xsl
• php85Extensions.xsl.dev
• php85Extensions.yaml
• php85Extensions.zend_test
• php85Extensions.zend_test.dev
• php85Extensions.zip
• php85Extensions.zip.dev
• php85Extensions.zlib
• php85Extensions.zlib.dev
• php85Extensions.zstd
• php85Packages.castor
• php85Packages.composer
• php85Packages.composer-local-repo-plugin
• php85Packages.cyclonedx-php-composer
• php85Packages.grumphp
• php85Packages.phan
• php85Packages.phing
• php85Packages.phive
• php85Packages.php-codesniffer
• php85Packages.php-cs-fixer
• php85Packages.php-parallel-lint
• php85Packages.phpinsights
• php85Packages.phpmd
• php85Packages.phpspy
• php85Packages.phpstan
• php85Packages.psalm

[piotrkwiecinski]

There is a separate PR with php 8.5 support for imagick for #474586 which targets staging at the moment.

Otherwise I reckon this one is ready to review @Ma27.

[Ma27]

Thanks!

Even though I'm still not particularly happy about this de-facto vendoring of lexbor, just including the headers where necessary is IMHO the way better solution.

No final approval yet because CI is red.
@NixOS/nixpkgs-ci is there a way to reasonably add versioned attributes into by-name now given that this is what the CI wants from us?

[MattSturgeon]

@NixOS/nixpkgs-ci is there a way to reasonably add versioned attributes into by-name now given that this is what the CI wants from us?

Without looking at the specifics of this PR, yes this is possible.

One approach is to create a by-name package directory for each version. Each version-package then overrides the base package (since it can't reference paths outside its scope).

Alternatively, you can expose versioned-packages as passthru attrs. Each version-package can reference finalAttrs.finalPackage as a base. These won't get picked up directly by hydra or eval-ci, but they will still be indirectly built and cached if they're a dependency of something else.

[Ma27]

@MattSturgeon

Alternatively, you can expose versioned-packages as passthru attrs. Each version-package can reference finalAttrs.finalPackage as a base. These won't get picked up directly by hydra or eval-ci, but they will still be indirectly built and cached if they're a dependency of something else.

Well, this is a new versioned attribute and not used anywhere yet. Also, I disagree with this being acceptable. By this rationale we could kill most Hydra jobs because the packages happen to be built in e.g. VM tests. This isn't helpful at all though when having dedicated jobsets to track regressions after core updates of e.g. glibc.

I honestly think this should be used in exceptional cases at best.

One approach is to create a by-name package directory for each version. Each version-package then overrides the base package (since it can't reference paths outside its scope).

To me, this still feels like a hack to force-push versioned attributes into something that doesn't really fit, by-name.
Also, isn't overrideAttrs discouraged anyways: #421201

[drupol]

Perhaps we could acknowledge that we should improve this and have another dedicated PR for it? So we don't block this long and awaited PR ?

[Ma27]

The reason I'm bringing this up is that the CI is red and IIRC "no PR failures" means we get kicked out of the queue when trying to merge this.

EDIT: to be clear, the approaches outlined by @MattSturgeon are the workarounds I've seen for a while now on trying to push ~everything into by-name. I'd feel very uneasy to do this for versioned attributes like we have here for the reasons I explained above.

[piotrkwiecinski]

@Ma27 @drupol I've created a second pr with some changes to top level definition. vet is green, but I'm not sure if it's a good direction b312e2b.

[jtojnar]

Alternately, we could merge without waiting for requirements and deal with this later.

[philiptaron]

Yeah, it's OK to let the ratchet go back a notch.

[drupol]

Note for later: remove the comment from the script, move it outside.

Add a release note.

[nixpkgs-ci]

Successfully created backport PR for release-25.11:

• [Backport release-25.11] php85: init at 8.5.2 #484734