.. Copyright Spack Project Developers. See COPYRIGHT file for details. SPDX-License-Identifier: (Apache-2.0 OR MIT) .. meta:: :description lang=en: Learn how to use Spack environments to manage reproducible software stacks, making it easy to share and recreate specific sets of packages and their dependencies. .. _environments: Environments (spack.yaml, spack.lock) ===================================== An environment is used to group a set of specs intended for some purpose to be built, rebuilt, and deployed in a coherent fashion. Environments define aspects of the installation of the software, such as: #. *which* specs to install; #. *how* those specs are configured; and #. *where* the concretized software will be installed. Aggregating this information into an environment for processing has advantages over the *à la carte* approach of building and loading individual Spack modules. With environments, you concretize, install, or load (activate) all of the specs with a single command. Concretization fully configures the specs and dependencies of the environment in preparation for installing the software. This is a more robust solution than ad-hoc installation scripts. And you can share an environment or even re-use it on a different computer. Environment definitions, especially *how* specs are configured, allow the software to remain stable and repeatable even when Spack packages are upgraded. Changes are only picked up when the environment is explicitly re-concretized. Defining *where* specs are installed supports a filesystem view of the environment. Yet Spack maintains a single installation of the software that can be re-used across multiple environments. Activating an environment determines *when* all of the associated (and installed) specs are loaded so limits the software loaded to those specs actually needed by the environment. Spack can even generate a script to load all modules related to an environment. Other packaging systems also provide environments that are similar in some ways to Spack environments; for example, `Conda environments `_ or `Python Virtual Environments `_. Spack environments provide some distinctive features though: #. A spec installed "in" an environment is no different from the same spec installed anywhere else in Spack. #. Spack environments may contain more than one spec of the same package. Spack uses a "manifest and lock" model similar to `Bundler gemfiles `_ and other package managers. The environment's user input file (or manifest), is named ``spack.yaml``. The lock file, which contains the fully configured and concretized specs, is named ``spack.lock``. .. _environments-using: Using Environments ------------------ Here we follow a typical use case of creating, concretizing, installing and loading an environment. .. _cmd-spack-env-create: Creating a managed Environment ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ An environment is created by: .. code-block:: console $ spack env create myenv The directory ``$SPACK_ROOT/var/spack/environments/myenv`` is created to manage the environment. .. note:: By default, all managed environments are stored in the ``$SPACK_ROOT/var/spack/environments`` folder. This location can be changed by setting the ``environments_root`` variable in ``config.yaml``. Spack creates the file ``spack.yaml``, hidden directory ``.spack-env``, and ``spack.lock`` file under ``$SPACK_ROOT/var/spack/environments/myenv``. User interaction occurs through the ``spack.yaml`` file and the Spack commands that affect it. Metadata and, by default, the view are stored in the ``.spack-env`` directory. When the environment is concretized, Spack creates the ``spack.lock`` file with the fully configured specs and dependencies for the environment. The ``.spack-env`` subdirectory also contains: * ``repo/``: A subdirectory acting as the repo consisting of the Spack packages used in the environment. It allows the environment to build the same, in theory, even on different versions of Spack with different packages! * ``logs/``: A subdirectory containing the build logs for the packages in this environment. Spack Environments can also be created from another environment. Environments can be created from the manifest file (the user input), the lockfile, or the entire environment at once. Create an environment from a manifest using: .. code-block:: console $ spack env create myenv spack.yaml The resulting environment is guaranteed to have the same root specs as the original but may concretize differently in the presence of different explicit or default configuration settings (e.g., a different version of Spack or for a different user account). Environments created from a manifest will copy any included configs from relative paths inside the environment. Relative paths from outside the environment will cause errors, and absolute paths will be kept absolute. For example, if ``spack.yaml`` includes: .. code-block:: yaml spack: include: [./config.yaml] then the created environment will have its own copy of the file ``config.yaml`` copied from the location in the original environment. Create an environment from a ``spack.lock`` file using: .. code-block:: console $ spack env create myenv spack.lock The resulting environment, when on the same or a compatible machine, is guaranteed to initially have the same concrete specs as the original. Create an environment from an entire environment using either the environment name or path: .. code-block:: console $ spack env create myenv /path/to/env $ spack env create myenv2 myenv The resulting environment will include the concrete specs from the original if the original is concretized (as when created from a lockfile) and all of the config options and abstract specs specified in the original (as when created from a manifest file). It will also include any other files included in the environment directory, such as repos or source code, as they could be referenced in the environment by relative path. .. note:: Environment creation also accepts a full path to the file. If the path is not under the ``$SPACK_ROOT/var/spack/environments`` directory then the source is referred to as an :ref:`independent environment `. The name of an environment can be a nested path to help organize environments via subdirectories. .. code-block:: console $ spack env create projectA/configA/myenv This will create a managed environment under ``$environments_root/projectA/configA/myenv``. Changing ``environment_root`` can therefore also be used to make a whole group of nested environments available. .. _cmd-spack-env-activate: .. _cmd-spack-env-deactivate: Activating an Environment ^^^^^^^^^^^^^^^^^^^^^^^^^ To activate an environment, use the following command: .. code-block:: console $ spack env activate myenv By default, the ``spack env activate`` will load the view associated with the environment into the user environment. The ``-v, --with-view`` argument ensures this behavior, and the ``-V, --without-view`` argument activates the environment without changing the user environment variables. The ``-p`` option to the ``spack env activate`` command modifies the user's prompt to begin with the environment name in brackets. .. code-block:: console $ spack env activate -p myenv [myenv] $ ... The ``activate`` command can also be used to create a new environment, if it is not already defined, by adding the ``--create`` flag. Managed and independent environments can both be created using the same flags that `spack env create` accepts. If an environment already exists then Spack will simply activate it and ignore the create-specific flags. .. code-block:: console $ spack env activate --create -p myenv # ... # [creates if myenv does not exist yet] # ... [myenv] $ ... To deactivate an environment, use the command: .. code-block:: console $ spack env deactivate or the shortcut alias .. code-block:: console $ despacktivate If the environment was activated with its view, deactivating the environment will remove the view from the user environment. .. _independent_environments: Independent Environments ^^^^^^^^^^^^^^^^^^^^^^^^ Independent environments can be located in any directory outside of Spack. .. note:: When uninstalling packages, Spack asks the user to confirm the removal of packages that are still used in a managed environment. This is not the case for independent environments. To create an independent environment, use one of the following commands: .. code-block:: console $ spack env create --dir my_env $ spack env create ./my_env As a shorthand, you can also create an independent environment upon activation if it does not already exist: .. code-block:: console $ spack env activate --create ./my_env For convenience, Spack can also place an independent environment in a temporary directory for you: .. code-block:: console $ spack env activate --temp Environment-Aware Commands ^^^^^^^^^^^^^^^^^^^^^^^^^^ Spack commands are environment-aware. For example, the ``find`` command shows only the specs in the active environment if an environment has been activated. Otherwise it shows all specs in the Spack instance. The same rule applies to the ``install`` and ``uninstall`` commands. .. code-block:: spec $ spack find ==> 0 installed packages $ spack install zlib@1.2.11 [+] q6cqrdt zlib@1.2.11 ~/spack/opt/spack/linux-rhel7-broadwell/gcc-8.1.0/zlib-1.2.11-q6cqrdto4iktfg6qyqcc5u4vmfmwb7iv (12s) $ spack env activate myenv $ spack find ==> In environment myenv ==> No root specs ==> 0 installed packages $ spack install zlib@1.2.8 [+] yfc7epf zlib@1.2.8 ~/spack/opt/spack/linux-rhel7-broadwell/gcc-8.1.0/zlib-1.2.8-yfc7epf57nsfn2gn4notccaiyxha6z7x (12s) ==> Updating view at ~/spack/var/spack/environments/myenv/.spack-env/view $ spack find ==> In environment myenv ==> Root specs zlib@1.2.8 ==> 1 installed package -- linux-rhel7-broadwell / gcc@8.1.0 ---------------------------- zlib@1.2.8 $ despacktivate $ spack find ==> 2 installed packages -- linux-rhel7-broadwell / gcc@8.1.0 ---------------------------- zlib@1.2.8 zlib@1.2.11 Note that when we installed the abstract spec ``zlib@1.2.8``, it was presented as a root of the environment. All explicitly installed packages will be listed as roots of the environment. All of the Spack commands that act on the list of installed specs are environment-aware in this way, including ``install``, ``uninstall``, ``find``, ``extensions``, etc. In the :ref:`environment-configuration` section we will discuss environment-aware commands further. .. _cmd-spack-add: Adding Abstract Specs ^^^^^^^^^^^^^^^^^^^^^ An abstract spec is the user-specified spec before Spack applies defaults or dependency information. You can add abstract specs to an environment using the ``spack add`` command. This adds the abstract spec as a root of the environment in the ``spack.yaml`` file. The most important component of an environment is a list of abstract specs. Adding abstract specs does not immediately install anything, nor does it affect the ``spack.lock`` file. To update the lockfile, the environment must be :ref:`re-concretized `, and to update any installations, the environment must be :ref:`(re)installed `. The ``spack add`` command is environment-aware. It adds the spec to the currently active environment. An error is generated if there isn't an active environment. .. code-block:: spec $ spack env activate myenv $ spack add mpileaks or .. code-block:: spec $ spack -e myenv add python .. note:: All environment-aware commands can also be called using the ``spack -e`` flag to specify the environment. .. _cmd-spack-concretize: Concretizing ^^^^^^^^^^^^ Once user specs have been added to an environment, they can be concretized. There are three different modes of operation to concretize an environment, explained in detail in :ref:`environments_concretization_config`. Regardless of which mode of operation is chosen, the following command will ensure all of the root specs are concretized according to the constraints that are prescribed in the configuration: .. code-block:: console [myenv]$ spack concretize In the case of specs that are not concretized together, the command above will concretize only the specs that were added and not yet concretized. Forcing a re-concretization of all of the specs can be done by adding the ``-f`` option: .. code-block:: console [myenv]$ spack concretize -f Without the option, Spack guarantees that already concretized specs are unchanged in the environment. The ``concretize`` command does not install any packages. For packages that have already been installed outside of the environment, the process of adding the spec and concretizing is identical to installing the spec assuming it concretizes to the exact spec that was installed outside of the environment. The ``spack find`` command can show concretized specs separately from installed specs using the ``-c`` (``--concretized``) flag. .. code-block:: console [myenv]$ spack add zlib [myenv]$ spack concretize [myenv]$ spack find -c ==> In environment myenv ==> Root specs zlib ==> Concretized roots -- linux-rhel7-x86_64 / gcc@4.9.3 ------------------------------- zlib@1.2.11 ==> 0 installed packages .. _installing-environment: Installing an Environment ^^^^^^^^^^^^^^^^^^^^^^^^^ In addition to adding individual specs to an environment, one can install the entire environment at once using the command .. code-block:: console [myenv]$ spack install If the environment has been concretized, Spack will install the concretized specs. Otherwise, ``spack install`` will concretize the environment before installing the concretized specs. .. note:: Every ``spack install`` process builds one package at a time with multiple build jobs, controlled by the ``-j`` flag and the ``config:build_jobs`` option (see :ref:`build-jobs`). To speed up environment builds further, independent packages can be installed in parallel by launching more Spack instances. For example, the following will build at most four packages in parallel using three background jobs: .. code-block:: console [myenv]$ spack install & spack install & spack install & spack install Another option is to generate a ``Makefile`` and run ``make -j`` to control the number of parallel install processes. See :ref:`cmd-spack-env-depfile` for details. As it installs, ``spack install`` creates symbolic links in the ``logs/`` directory in the environment, allowing for easy inspection of build logs related to that environment. The ``spack install`` command also stores a Spack repo containing the ``package.py`` file used at install time for each package in the ``repos/`` directory in the environment. The ``--no-add`` option can be used in a concrete environment to tell Spack to install specs already present in the environment but not to add any new root specs to the environment. For root specs provided to ``spack install`` on the command line, ``--no-add`` is the default, while for dependency specs, it is optional. In other words, if there is an unambiguous match in the active concrete environment for a root spec provided to ``spack install`` on the command line, Spack does not require you to specify the ``--no-add`` option to prevent the spec from being added again. At the same time, a spec that already exists in the environment, but only as a dependency, will be added to the environment as a root spec without the ``--no-add`` option. .. _cmd-spack-develop: Developing Packages in a Spack Environment ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ The ``spack develop`` command allows one to develop Spack packages in an environment. It will configure Spack to install the package from local source. By default, ``spack develop`` will also clone the package to a subdirectory in the environment for the local source. These choices can be overridden with the ``--path`` argument, and the ``--no-clone`` argument. Relative paths provided to the ``--path`` argument will be resolved relative to the environment directory. All of these options are recorded in the environment manifest, although default values may be left implied. .. code-block:: console $ spack develop --path src/foo foo@develop $ cat `spack location -e`/spack.yaml spack: ... develop foo: spec: foo@develop path: src/foo When ``spack develop`` is run in a concretized environment, Spack will modify the concrete specs in the environment to reflect the modified provenance. Any package built from local source will have a ``dev_path`` variant, and the hash of any dependent of those packages will be modified to reflect the change. The value of the ``dev_path`` variant will be the absolute path to the package source directory. If the develop spec conflicts with the concrete specs in the environment, Spack will raise an exception and require the ``spack develop --no-modify-concrete-specs`` option, followed by a ``spack concretize --force`` to apply the ``dev_path`` variant and constraints from the develop spec. When concretizing an environment with develop specs, the version, variants, and other attributes of the spec provided to the ``spack develop`` command will be treated as constraints by the concretizer (in addition to any constraints from the packages ``specs`` list). If the ``develop`` configuration for the package does not include a spec version, Spack will choose the **highest** version of the package. This means that any "infinity" versions (``develop``, ``main``, etc.) will be preferred for specs marked with the ``spack develop`` command, which is different from the standard Spack behavior to prefer the highest **numeric** version. These packages will have an automatic ``dev_path`` variant added by the concretizer, with a value of the absolute path to the local source Spack is building from. Spack will ensure the package and its dependents are rebuilt any time the environment is installed if the package's local source code has been modified. Spack's native implementation is to check if ``mtime`` is newer than the installation. A custom check can be created by overriding the ``detect_dev_src_change`` method in your package class. This is particularly useful for projects using custom Spack repos to drive development and want to optimize performance. When ``spack develop`` is run without any arguments, Spack will clone any develop specs in the environment for which the specified path does not exist. When working deep in the graph it is often desirable to have multiple specs marked as ``develop`` so you don't have to restage and/or do full rebuilds each time you call ``spack install``. The ``--recursive`` flag can be used in these scenarios to ensure that all the dependents of the initial spec you provide are also marked as develop specs. The ``--recursive`` flag requires a pre-concretized environment so the graph can be traversed from the supplied spec all the way to the root specs. For packages with ``git`` attributes, git branches, tags, and commits can also be used as valid concrete versions (see :ref:`version-specifier`). This means that for a package ``foo``, ``spack develop foo@git.main`` will clone the ``main`` branch of the package, and ``spack install`` will install from that git clone if ``foo`` is in the environment. Further development on ``foo`` can be tested by re-installing the environment, and eventually committed and pushed to the upstream git repo. If the package being developed supports out-of-source builds then users can use the ``--build_directory`` flag to control the location and name of the build directory. This is a shortcut to set the ``package_attributes:build_directory`` in the ``packages`` configuration (see :ref:`assigning-package-attributes`). The supplied location will become the build-directory for that package in all future builds. .. admonition:: Potential pitfalls of setting the build directory :class: warning Spack does not check for out-of-source build compatibility with the packages and so the onus of making sure the package supports out-of-source builds is on the user. For example, most ``autotool`` and ``makefile`` packages do not support out-of-source builds while all ``CMake`` packages do. Understanding these nuances is up to the software developers and we strongly encourage developers to only redirect the build directory if they understand their package's build-system. Modifying Specs in an Environment ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ The ``spack change`` command allows the user to change individual specs in a Spack environment. By default, ``spack change`` operates on the abstract specs of an environment. The command a list of spec arguments. For each argument, the root spec with the same name as the provided spec is modified to satisfy the provided spec. For example, in an environment with the root spec ``hdf5+mpi+fortran``, then .. code-block:: console spack change hdf5~mpi+cxx will change the root spec to ``hdf5~mpi+cxx+fortran``. When more complex matching semantics are necessary, the ``--match-spec`` argument replaces the spec name as the selection criterion. When using the ``--match-spec`` argument, the spec name is not required. In the same environment, .. code-block:: console spack change --match-spec "+fortran" +hl will constrain the ``hdf5`` spec to ``+hl``. By default, the ``spack change`` command will result in an error and no change to the environment if it will modify more than one abstract spec. Use the ``--all`` option to allow ``spack change`` to modify multiple abstract specs. The ``--concrete`` option allows ``spack change`` to modify the concrete specs of an environment as well as the abstract specs. Multiple concrete specs may be modified, even for a change that modifies only a single abstract spec. The ``--all`` option does not affect how many concrete specs may be modified. .. warning:: Concrete specs are modified without any constraints from the packages. The ``spack change --concrete`` command may create invalid specs that will not build properly if applied without caution. The ``--concrete-only`` option allows for modifying concrete specs without modifying abstract specs. It allows changes to be applied to non-root nodes in the environment, and other changes that do not modify any root specs. Loading ^^^^^^^ Once an environment has been installed, the following creates a load script for it: .. code-block:: console $ spack env loads -r This creates a file called ``loads`` in the environment directory. Sourcing that file in Bash will make the environment available to the user, and can be included in ``.bashrc`` files, etc. The ``loads`` file may also be copied out of the environment, renamed, etc. .. _environment_include_concrete: Including Concrete Environments ------------------------------- Spack can create an environment that includes information from already concretized environments. You can think of the new environment as a combination of existing environments. It uses information from the existing environments' ``spack.lock`` files in the creation of the new environment. When such an environment is concretized it will generate its own ``spack.lock`` file that contains relevant information from the included environments. Creating combined concrete environments ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ To create a combined concrete environment, you must have at least one existing concrete environment. You will use the command ``spack env create`` with the argument ``--include-concrete`` followed by the name or path of the environment you'd like to include. Here is an example of how to create a combined environment from the command line:: $ spack env create myenv $ spack -e myenv add python $ spack -e myenv concretize $ spack env create --include-concrete myenv combined_env You can also include concrete environments directly in the ``spack.yaml`` file. It involves adding the absolute paths to the concrete environments ``spack.lock`` under the new environment's ``include`` heading. Spack-specific configuration variables, such as ``$spack``, and environment variables can be used in the include paths as long as the expression expands to an absolute path. (See :ref:`config-file-variables` for more information.) For example, .. code-block:: yaml spack: include: - /absolute/path/to/environment1/spack.lock - $spack/../path/to/environment2/spack.lock specs: [] concretizer: unify: true will include the specs from ``environment1`` and ``environment2`` where the second environment's path is the absolute path of the directory that is relative to the spack root. .. note:: Once the ``spack.yaml`` file is updated you must concretize the new environment to get the concrete specs from the included environments. This will produce the combined ``spack.lock`` file. Updating a combined environment ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ If you want changes made to one of the included environments reflected in the combined environment, then you will need to re-concretize the included environment **then** the combined environment for the change to be incorporated. For example:: $ spack env create myenv $ spack -e myenv add python $ spack -e myenv concretize $ spack env create --include-concrete myenv combined_env $ spack -e myenv find ==> In environment myenv ==> Root specs python ==> 0 installed packages $ spack -e combined_env find ==> In environment combined_env ==> No root specs ==> Included specs python ==> 0 installed packages Here we see that ``combined_env`` contains the python package from ``myenv`` environment. But if we were to add another spec to ``myenv``, ``combined_env`` will not know about the other spec. .. code-block:: spec $ spack -e myenv add perl $ spack -e myenv concretize $ spack -e myenv find ==> In environment myenv ==> Root specs perl python ==> 0 installed packages $ spack -e combined_env find ==> In environment combined_env ==> No root specs ==> Included specs python ==> 0 installed packages It isn't until you run the ``spack concretize`` command that the combined environment will get the updated information from the re-concretized ``myenv``. .. code-block:: console $ spack -e combined_env concretize $ spack -e combined_env find ==> In environment combined_env ==> No root specs ==> Included specs perl python ==> 0 installed packages .. _environment-configuration: Configuring Environments ------------------------ A variety of Spack behaviors are changed through Spack configuration files, covered in more detail in the :ref:`configuration` section. Spack Environments provide an additional level of configuration scope between the custom scope and the user scope discussed in the configuration documentation. There are two ways to include configuration information in a Spack Environment: #. Inline in the ``spack.yaml`` file #. Included in the ``spack.yaml`` file from another file. Many Spack commands also affect configuration information in files automatically. Those commands take a ``--scope`` argument, and the environment can be specified by ``env:NAME`` (to affect environment ``foo``, set ``--scope env:foo``). These commands will automatically manipulate configuration inline in the ``spack.yaml`` file. Inline configurations ^^^^^^^^^^^^^^^^^^^^^ Inline environment-scope configuration is done using the same yaml format as standard Spack configuration scopes, covered in the :ref:`configuration` section. Each section is contained under a top-level yaml object with its name. For example, a ``spack.yaml`` manifest file containing some package preference configuration (as in a ``packages.yaml`` file) could contain: .. code-block:: yaml spack: # ... packages: all: providers: mpi: [openmpi] # ... This configuration sets the default ``mpi`` provider to be ``openmpi``. Included configurations ^^^^^^^^^^^^^^^^^^^^^^^ Spack environments allow an ``include`` heading in their yaml schema. This heading pulls in external configuration files and applies them to the environment. .. code-block:: yaml spack: include: - environment/relative/path/to/config.yaml - path: https://github.com/path/to/raw/config/compilers.yaml sha256: 26e871804a92cd07bb3d611b31b4156ae93d35b6a6d6e0ef3a67871fcb1d258b - /absolute/path/to/packages.yaml - path: /path/to/$os/$target/environment optional: true - path: /path/to/os-specific/config-dir when: os == "ventura" Included configuration files are required *unless* they are explicitly optional or the entry's condition evaluates to ``false``. Optional includes are specified with the ``optional`` clause and conditional with the ``when`` clause. (See :ref:`include-yaml` for more information on optional and conditional entries.) Files are listed using paths to individual files or directories containing them. Path entries may be absolute or relative to the environment or specified as URLs. URLs to individual files must link to the **raw** form of the file's contents (e.g., `GitHub `_ or `GitLab `_) **and** include a valid sha256 for the file. Only the ``file``, ``ftp``, ``http`` and ``https`` protocols (or schemes) are supported. Spack-specific, environment and user path variables can be used. (See :ref:`config-file-variables` for more information.) .. warning:: Recursive includes are not currently processed in a breadth-first manner so the value of a configuration option that is altered by multiple included files may not be what you expect. This will be addressed in a future update. Configuration precedence ^^^^^^^^^^^^^^^^^^^^^^^^ Inline configurations take precedence over included configurations, so you don't have to change shared configuration files to make small changes to an individual environment. Included configurations listed earlier will have higher precedence, as the included configs are applied in reverse order. Manually Editing the Specs List ------------------------------- The list of abstract/root specs in the environment is maintained in the ``spack.yaml`` manifest under the heading ``specs``. .. code-block:: yaml spack: specs: - ncview - netcdf - nco - py-sphinx Appending to this list in the yaml is identical to using the ``spack add`` command from the command line. However, there is more power available from the yaml file. .. _environments_concretization_config: Spec concretization ^^^^^^^^^^^^^^^^^^^ An environment can be concretized in three different modes and the behavior active under any environment is determined by the ``concretizer:unify`` configuration option. The *default* mode is to unify all specs: .. code-block:: yaml spack: specs: - hdf5+mpi - zlib@1.2.8 concretizer: unify: true This means that any package in the environment corresponds to a single concrete spec. In the above example, when ``hdf5`` depends down the line of ``zlib``, it is required to take ``zlib@1.2.8`` instead of a newer version. This mode of concretization is particularly useful when environment views are used: if every package occurs in only one flavor, it is usually possible to merge all install directories into a view. A downside of unified concretization is that it can be overly strict. For example, a concretization error would happen when both ``hdf5+mpi`` and ``hdf5~mpi`` are specified in an environment. The second mode is to *unify when possible*: this makes concretization of root specs more independent. Instead of requiring reuse of dependencies across different root specs, it is only maximized: .. code-block:: yaml spack: specs: - hdf5~mpi - hdf5+mpi - zlib@1.2.8 concretizer: unify: when_possible This means that both ``hdf5`` installations will use ``zlib@1.2.8`` as a dependency even if newer versions of that library are available. The third mode of operation is to concretize root specs entirely independently by disabling unified concretization: .. code-block:: yaml spack: specs: - hdf5~mpi - hdf5+mpi - zlib@1.2.8 concretizer: unify: false In this example ``hdf5`` is concretized separately, and does not consider ``zlib@1.2.8`` as a constraint or preference. Instead, it will take the latest possible version. The last two concretization options are typically useful for system administrators and user support groups providing a large software stack for their HPC center. .. note:: The ``concretizer:unify`` config option was introduced in Spack 0.18 to replace the ``concretization`` property. For reference, ``concretization: together`` is replaced by ``concretizer:unify:true``, and ``concretization: separately`` is replaced by ``concretizer:unify:false``. .. admonition:: Re-concretization of user specs The ``spack concretize`` command without additional arguments will *not* change any previously concretized specs. This may prevent it from finding a solution when using ``unify: true``, and it may prevent it from finding a minimal solution when using ``unify: when_possible``. You can force Spack to ignore the existing concrete environment with ``spack concretize -f``. .. _environment-spec-matrices: Spec Matrices ^^^^^^^^^^^^^ Entries in the ``specs`` list can be individual abstract specs or a spec matrix. A spec matrix is a yaml object containing multiple lists of specs, and evaluates to the cross-product of those specs. Spec matrices also contain an ``excludes`` directive, which eliminates certain combinations from the evaluated result. The following two environment manifests are identical: .. code-block:: yaml spack: specs: - zlib %gcc@7.1.0 - zlib %gcc@4.9.3 - libelf %gcc@7.1.0 - libelf %gcc@4.9.3 - libdwarf %gcc@7.1.0 - cmake .. code-block:: yaml spack: specs: - matrix: - [zlib, libelf, libdwarf] - ["%gcc@7.1.0", "%gcc@4.9.3"] exclude: - libdwarf%gcc@4.9.3 - cmake Spec matrices can be used to install swaths of software across various toolchains. .. _spec-list-references: Spec List References ^^^^^^^^^^^^^^^^^^^^ The last type of possible entry in the specs list is a reference. The Spack Environment manifest yaml schema contains an additional heading ``definitions``. Under definitions is an array of yaml objects. Each object has one or two fields. The one required field is a name, and the optional field is a ``when`` clause. The named field is a spec list. The spec list uses the same syntax as the ``specs`` entry. Each entry in the spec list can be a spec, a spec matrix, or a reference to an earlier named list. References are specified using the ``$`` sigil, and are "splatted" into place (i.e. the elements of the referent are at the same level as the elements listed separately). As an example, the following two manifest files are identical. .. code-block:: yaml spack: definitions: - first: [libelf, libdwarf] - compilers: ["%gcc", "%intel"] - second: - $first - matrix: - [zlib] - [$compilers] specs: - $second - cmake .. code-block:: yaml spack: specs: - libelf - libdwarf - zlib%gcc - zlib%intel - cmake .. note:: Named spec lists in the definitions section may only refer to a named list defined above itself. Order matters. In short files like the example, it may be easier to simply list the included specs. However for more complicated examples involving many packages across many toolchains, separately factored lists make environments substantially more manageable. Additionally, the ``-l`` option to the ``spack add`` command allows one to add to named lists in the definitions section of the manifest file directly from the command line. The ``when`` directive can be used to conditionally add specs to a named list. The ``when`` directive takes a string of Python code referring to a restricted set of variables, and evaluates to a boolean. The specs listed are appended to the named list if the ``when`` string evaluates to ``True``. In the following snippet, the named list ``compilers`` is ``["%gcc", "%clang", "%intel"]`` on ``x86_64`` systems and ``["%gcc", "%clang"]`` on all other systems. .. code-block:: yaml spack: definitions: - compilers: ["%gcc", "%clang"] - when: arch.satisfies("target=x86_64:") compilers: ["%intel"] .. note:: Any definitions with the same named list with true ``when`` clauses (or absent ``when`` clauses) will be appended together The valid variables for a ``when`` clause are: #. ``platform``. The platform string of the default Spack architecture on the system. #. ``os``. The OS string of the default Spack architecture on the system. #. ``target``. The target string of the default Spack architecture on the system. #. ``architecture`` or ``arch``. A Spack spec satisfying the default Spack architecture on the system. This supports querying via the ``satisfies`` method, as shown above. #. ``arch_str``. The architecture string of the default Spack architecture on the system. #. ``re``. The standard regex module in Python. #. ``env``. The user environment (usually ``os.environ`` in Python). #. ``hostname``. The hostname of the system (if ``hostname`` is an executable in the user's PATH). SpecLists as Constraints ^^^^^^^^^^^^^^^^^^^^^^^^ Dependencies and compilers in Spack can be both packages in an environment and constraints on other packages. References to SpecLists allow a shorthand to treat packages in a list as either a compiler or a dependency using the ``$%`` or ``$^`` syntax respectively. For example, the following environment has three root packages: ``gcc@8.1.0``, ``mvapich2@2.3.1 %gcc@8.1.0``, and ``hdf5+mpi %gcc@8.1.0 ^mvapich2@2.3.1``. .. code-block:: yaml spack: definitions: - compilers: [gcc@8.1.0] - mpis: [mvapich2@2.3.1] - packages: [hdf5+mpi] specs: - $compilers - matrix: - [$mpis] - [$%compilers] - matrix: - [$packages] - [$^mpis] - [$%compilers] This allows for a much-needed reduction in redundancy between packages and constraints. .. _environment-spec-groups: Spec Groups ^^^^^^^^^^^ .. versionadded:: 1.2 Environments can be organized with named spec groups, enabling you to apply localized configuration overrides and establish concretization dependencies. This is extremely useful in a couple of common scenarios, as detailed below. .. _environment-spec-groups-bootstrapping-compiler: Building and using a compiler in a single environment """"""""""""""""""""""""""""""""""""""""""""""""""""" A common use case is to build a recent compiler on top of an existing system and then compile a stack of software with it. For instance, assume we are interested in building ``hdf5`` and ``libtree`` with ``gcc@15.2``. The following manifest file would do exactly that: .. code-block:: yaml spack: specs: - group: compiler specs: - gcc@15.2 - group: apps needs: [compiler] specs: - hdf5 %gcc@15.2 - libtree %gcc@15.2 The ``group:`` attribute allows to name a group of specs, which are then listed under the ``specs:`` attribute in the same object. The simplest example is the ``compiler`` group composed of just the ``gcc@15.2`` spec. To express dependencies among groups of specs the ``needs:`` attribute is used, which is a list of names corresponding to the groups we depend on. The way this works is that group dependencies are always concretized *before* the current group, and their specs are *always* available for reuse when the current group is concretized. .. _environment-spec-groups-configuring-groups: Configuring a group of specs """""""""""""""""""""""""""" Another common scenario is the deployment of different configurations (e.g. CUDA enabled vs. ROCm enabled) of the same set of software. As an example, assume we want to install ``gromacs`` and ``quantum-espresso`` for both ``target=x86_64_v3`` and ``target=x86_64_v4``. That can be done with the following manifest file: .. code-block:: yaml spack: - group: apps-x86_64_v3 specs: - gromacs - quantum-espresso override: packages: all: prefer: - target=x86_64_v3 - group: apps-x86_64_v4 specs: - gromacs - quantum-espresso override: packages: all: prefer: - target=x86_64_v4 The ``override:`` attribute allows us to override the configuration for a single group of specs. The overridden part is always added as the *topmost* scope when the current group is concretized. This ensures the override always takes precedence over other sources of configuration. .. _environment-spec-groups-explicit: Controlling garbage collection with ``explicit: false`` """"""""""""""""""""""""""""""""""""""""""""""""""""""" By default every spec group is treated as a set of *explicit* roots. This means its specs are preserved by ``spack gc`` even when nothing else depends on them. Setting ``explicit: false`` on a group marks its specs as *implicit*, making them eligible for garbage collection once no other installed spec depends on them: .. code-block:: yaml spack: specs: - group: compiler explicit: false specs: - gcc@15.2 - group: apps needs: [compiler] specs: - hdf5 %gcc@15.2 - libtree %gcc@15.2 After the apps are installed, ``spack gc`` will remove the compiler once no installed spec has a link or run dependency on it. .. note:: Flipping ``explicit: false`` on a group that has already been installed does **not** retroactively update the database record for the already-installed specs. The flag takes effect only for specs installed, or re-installed, after the change. To immediately mark an existing spec as implicit, use ``spack mark -i ``. Modifying Environment Variables ------------------------------- Spack Environments can modify the active shell's environment variables when activated. The environment can be configured to set, unset, prepend, or append using ``env_vars`` configuration in ``spack.yaml``: .. code-block:: yaml spack: env_vars: set: ENVAR_TO_SET_IN_ENV_LOAD: "FOO" unset: - ENVAR_TO_UNSET_IN_ENV_LOAD prepend_path: PATH_LIST: "path/to/prepend" append_path: PATH_LIST: "path/to/append" remove_path: PATH_LIST: "path/to/remove" Environment Views ----------------- Spack Environments can have an associated filesystem view, which is a directory with a more traditional structure ``/bin``, ``/lib``, ``/include`` in which all files of the installed packages are linked. By default a view is created for each environment, thanks to the ``view: true`` option in the ``spack.yaml`` manifest file: .. code-block:: yaml spack: specs: [perl, python] view: true The view is created in a hidden directory ``.spack-env/view`` relative to the environment. If you've used ``spack env activate``, you may have already interacted with this view. Spack prepends its ``/bin`` dir to ``PATH`` when the environment is activated, so that you can directly run executables from all installed packages in the environment. Views are highly customizable: you can control where they are put, modify their structure, include and exclude specs, change how files are linked, and you can even generate multiple views for a single environment. .. _configuring_environment_views: Minimal view configuration ^^^^^^^^^^^^^^^^^^^^^^^^^^ The minimal configuration .. code-block:: yaml spack: # ... view: true lets Spack generate a single view with default settings under the ``.spack-env/view`` directory of the environment. Another short way to configure a view is to specify just where to put it: .. code-block:: yaml spack: # ... view: /path/to/view Views can also be disabled by setting ``view: false``. .. _cmd-spack-env-view: Advanced view configuration ^^^^^^^^^^^^^^^^^^^^^^^^^^^ One or more **view descriptors** can be defined under ``view``, keyed by a name. The example from the previous section with ``view: /path/to/view`` is equivalent to defining a view descriptor named ``default`` with a ``root`` attribute: .. code-block:: yaml spack: # ... view: default: # name of the view root: /path/to/view # view descriptor attribute The ``default`` view descriptor name is special: when you ``spack env activate`` your environment, this view will be used to update (among other things) your ``PATH`` variable. View descriptors must contain the root of the view, and optionally projections, ``select`` and ``exclude`` lists and link information via ``link`` and ``link_type``. As a more advanced example, in the following manifest file snippet we define a view named ``mpis``, rooted at ``/path/to/view`` in which all projections use the package name, version, and compiler name to determine the path for a given package. This view selects all packages that depend on MPI, and excludes those built with the GCC compiler at version 8.5. The root specs with their (transitive) link and run type dependencies will be put in the view due to the ``link: all`` option, and the files in the view will be symlinks to the Spack install directories. .. code-block:: yaml spack: # ... view: mpis: root: /path/to/view select: [^mpi] exclude: ["%gcc@8.5"] projections: all: "{name}/{version}-{compiler.name}" link: all link_type: symlink link_dirs: true The default for the ``select`` and ``exclude`` values is to select everything and exclude nothing. The default projection is the default view projection (``{}``). The ``link`` attribute allows the following values: #. ``link: all`` include root specs with their transitive run and link type dependencies (default); #. ``link: run`` include root specs with their transitive run type dependencies; #. ``link: roots`` include root specs without their dependencies. The ``link_type`` defaults to ``symlink`` but can also take the value of ``hardlink`` or ``copy``. .. versionadded:: 1.2 The ``link_dirs`` option controls whether directories are symlinked. This is the default behavior in Spack v1.2 and later. This is an optimization that significantly reduces the time to create views, and reduces the inode usage of the view. It only applies when ``link_type`` is set to ``symlink``. If you want to link only non-directory files, set ``link_dirs: false``. .. tip:: The option ``link: run`` can be used to create small environment views for Python packages. Python will be able to import packages *inside* of the view even when the environment is not activated, and linked libraries will be located *outside* of the view thanks to rpaths. From the command line, the ``spack env create`` command takes an argument ``--with-view [PATH]`` that sets the path for a single, default view. If no path is specified, the default path is used (``view: true``). The argument ``--without-view`` can be used to create an environment without any view configured. The ``spack env view`` command can be used to manage views of an environment. The subcommand ``spack env view enable`` will add a view named ``default`` to an environment. It takes an optional argument to specify the path for the new default view. The subcommand ``spack env view disable`` will remove the view named ``default`` from an environment if one exists. The subcommand ``spack env view regenerate`` will regenerate the views for the environment. This will apply any updates in the environment configuration that have not yet been applied. .. _view_projections: View Projections """""""""""""""" The default projection into a view is to link every package into the root of the view. The projections attribute is a mapping of partial specs to spec format strings, defined by the :meth:`~spack.spec.Spec.format` function, as shown in the example below: .. code-block:: yaml projections: zlib: "{name}-{version}" ^mpi: "{name}-{version}/{^mpi.name}-{^mpi.version}-{compiler.name}-{compiler.version}" all: "{name}-{version}/{compiler.name}-{compiler.version}" Projections also permit environment and Spack configuration variable expansions as shown below: .. code-block:: yaml projections: all: "{name}-{version}/{compiler.name}-{compiler.version}/$date/$SYSTEM_ENV_VARIABLE" where ``$date`` is the Spack configuration variable that will expand with the ``YYYY-MM-DD`` format and ``$SYSTEM_ENV_VARIABLE`` is an environment variable defined in the shell. The entries in the projections configuration file must all be either specs or the keyword ``all``. For each spec, the projection used will be the first non-``all`` entry that the spec satisfies, or ``all`` if there is an entry for ``all`` and no other entry is satisfied by the spec. Where the keyword ``all`` appears in the file does not matter. Given the example above, the spec ``zlib@1.2.8`` will be linked into ``/my/view/zlib-1.2.8/``, the spec ``hdf5@1.8.10+mpi %gcc@4.9.3 ^mvapich2@2.2`` will be linked into ``/my/view/hdf5-1.8.10/mvapich2-2.2-gcc-4.9.3``, and the spec ``hdf5@1.8.10~mpi %gcc@4.9.3`` will be linked into ``/my/view/hdf5-1.8.10/gcc-4.9.3``. If the keyword ``all`` does not appear in the projections configuration file, any spec that does not satisfy any entry in the file will be linked into the root of the view as in a single-prefix view. Any entries that appear below the keyword ``all`` in the projections configuration file will not be used, as all specs will use the projection under ``all`` before reaching those entries. Group of Specs """""""""""""" Views can also be applied to a selected list of :ref:`spec groups `. This can be done by specifying the ``group:`` attribute in the view configuration. For instance, with the following manifest: .. code-block:: yaml spack: concretizer: unify: true packages: all: require: - target=x86_64_v4 specs: - group: compiler specs: - gcc@15.2 - group: apps needs: [compiler] specs: - hdf5~mpi %gcc@15.2 - libtree %gcc@15.2 view: apps: root: ./views/apps group: apps The view will only contain entries from the ``apps`` group, and will not include specs from the ``compiler`` group. Activating environment views ^^^^^^^^^^^^^^^^^^^^^^^^^^^^ The ``spack env activate `` command has two effects: 1. It activates the environment so that further Spack commands such as ``spack install`` will run in the context of the environment. 2. It activates the view so that environment variables such as ``PATH`` are updated to include the view. Without further arguments, the ``default`` view of the environment is activated. If a view with a different name has to be activated, ``spack env activate --with-view `` can be used instead. You can also activate the environment without modifying further environment variables using ``--without-view``. The environment variables affected by the ``spack env activate`` command and the paths that are used to update them are determined by the :ref:`prefix inspections ` defined in your modules configuration; the defaults are summarized in the following table. =================== ========= Variable Paths =================== ========= PATH bin MANPATH man, share/man ACLOCAL_PATH share/aclocal PKG_CONFIG_PATH lib/pkgconfig, lib64/pkgconfig, share/pkgconfig CMAKE_PREFIX_PATH . =================== ========= Each of these paths are appended to the view root, and added to the relevant variable if the path exists. For this reason, it is not recommended to use non-default projections with the default view of an environment. The ``spack env deactivate`` command will remove the active view of the Spack environment from the user's environment variables. .. _cmd-spack-env-depfile: Generating Depfiles from Environments ------------------------------------------ Spack can generate ``Makefile``\s to make it easier to build multiple packages in an environment in parallel. .. note:: Since Spack v1.1, there is a new experimental installer that supports package-level parallelism out of the box with POSIX jobserver support. You can enable it with ``spack config add config:installer:new``. This new installer may provide a simpler alternative to the ``spack env depfile`` workflow described in this section for users primarily interested in speeding up environment installations. Generated ``Makefile``\s expose targets that can be included in existing ``Makefile``\s, to allow other targets to depend on the environment installation. A typical workflow is as follows: .. code-block:: spec $ spack env create -d . $ spack -e . add perl $ spack -e . concretize $ spack -e . env depfile -o Makefile $ make -j64 This generates a ``Makefile`` from a concretized environment in the current working directory, and ``make -j64`` installs the environment, exploiting parallelism across packages as much as possible. Spack respects the Make jobserver and forwards it to the build environment of packages, meaning that a single ``-j`` flag is enough to control the load, even when packages are built in parallel. By default the following phony convenience targets are available: - ``make all``: installs the environment (default target); - ``make clean``: cleans files used by make, but does not uninstall packages. .. tip:: GNU Make version 4.3 and above have great support for output synchronization through the ``-O`` and ``--output-sync`` flags, which ensure that output is printed orderly per package install. To get synchronized output with colors, use ``make -j SPACK_COLOR=always --output-sync=recurse``. Specifying dependencies on generated ``make`` targets ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ An interesting question is how to include generated ``Makefile``\s in your own ``Makefile``\s. This comes up when you want to install an environment that provides executables required in a command for a make target of your own. The example below shows how to accomplish this: the ``env`` target specifies the generated ``spack/env`` target as a prerequisite, meaning that the environment gets installed and is available for use in the ``env`` target. .. code-block:: Makefile SPACK ?= spack .PHONY: all clean env all: env spack.lock: spack.yaml $(SPACK) -e . concretize -f env.mk: spack.lock $(SPACK) -e . env depfile -o $@ --make-prefix spack env: spack/env $(info environment installed!) clean: rm -rf spack.lock env.mk spack/ ifeq (,$(filter clean,$(MAKECMDGOALS))) include env.mk endif This works as follows: when ``make`` is invoked, it first "remakes" the missing include ``env.mk`` as there is a target for it. This triggers concretization of the environment and makes Spack output ``env.mk``. At that point the generated target ``spack/env`` becomes available through ``include env.mk``. As it is typically undesirable to remake ``env.mk`` as part of ``make clean``, the include is conditional. .. note:: When including generated ``Makefile``\s, it is important to use the ``--make-prefix`` flag and use the non-phony target ``/env`` as prerequisite, instead of the phony target ``/all``. Building a subset of the environment ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ The generated ``Makefile``\s contain install targets for each spec, identified by ``--``. This allows you to install only a subset of the packages in the environment. When packages are unique in the environment, it's enough to know the name and let tab-completion fill out the version and hash. The following phony targets are available: ``install/`` to install the spec with its dependencies, and ``install-deps/`` to *only* install its dependencies. This can be useful when certain flags should only apply to dependencies. Below we show a use case where a spec is installed with verbose output (``spack install --verbose``) while its dependencies are installed silently: .. code-block:: console $ spack env depfile -o Makefile # Install dependencies in parallel, only show a log on error. $ make -j16 install-deps/python-3.11.0- SPACK_INSTALL_FLAGS=--show-log-on-error # Install the root spec with verbose output. $ make -j16 install/python-3.11.0- SPACK_INSTALL_FLAGS=--verbose Adding post-install hooks ^^^^^^^^^^^^^^^^^^^^^^^^^ Another advanced use-case of generated ``Makefile``\s is running a post-install command for each package. These "hooks" could be anything from printing a post-install message, running tests, or pushing just-built binaries to a build cache. This can be accomplished through the generated ``[/]SPACK_PACKAGE_IDS`` variable. Assuming we have an active and concrete environment, we generate the associated ``Makefile`` with a prefix ``example``: .. code-block:: console $ spack env depfile -o env.mk --make-prefix example And we now include it in a different ``Makefile``, in which we create a target ``example/push/%`` with ``%`` referring to a package identifier. This target depends on the particular package installation. In this target we automatically have the target-specific ``HASH`` and ``SPEC`` variables at our disposal. They are respectively the spec hash (excluding leading ``/``), and a human-readable spec. Finally, we have an entry point target ``push`` that will update the build cache index once every package is pushed. Note how this target uses the generated ``example/SPACK_PACKAGE_IDS`` variable to define its prerequisites. .. code-block:: Makefile SPACK ?= spack BUILDCACHE_DIR = $(CURDIR)/tarballs .PHONY: all all: push include env.mk example/push/%: example/install/% @mkdir -p $(dir $@) $(info About to push $(SPEC) to a buildcache) $(SPACK) -e . buildcache push --only=package $(BUILDCACHE_DIR) /$(HASH) @touch $@ push: $(addprefix example/push/,$(example/SPACK_PACKAGE_IDS)) $(info Updating the buildcache index) $(SPACK) -e . buildcache update-index $(BUILDCACHE_DIR) $(info Done!) @touch $@