If we do not specify libdir explicitly, Meson chooses something like
lib/x86_64-linux-gnu, which causes problems when trying to find libraries
and pkg-config files.
Spack packages installed using spack buildcache were not running
post-install hooks, which create module files and manage licenses
(if necessary).
This was already occurring for Spack packages installed with
spack install --use-cache
Spack can now be configured to assign permissions to the files installed by a package.
In the `packages.yaml` file under `permissions`, the attributes `read`, `write`, and `group` control the package permissions. These attributes can be set per-package, or for all packages under `all`. If permissions are set under `all` and for a specific package, the package-specific settings take precedence. The `read` and `write` attributes take one of `user`, `group`, and `world`.
packages:
all:
permissions:
write: group
group: spack
my_app:
permissions:
read: group
group: my_team
* Better default CLI arguments for CDash reporting
--log-format=cdash is now implied if you specify the --cdash-upload-url
option to spack install.
We also now default to writing CTest XML files to cdash_report/ when using
the CDash reporter if no --log-file argument was specified.
* Improved documentation on how to use the CDash reporter
* Push default flag handlers into module scope
* Preserve backwards compatibility of builtin flag handler names
Ensure Spack continues to work for packages using the `Package.env_flags` idiom and equivalent.
* update docs and tests to match
* Update packages to match new syntax
Fix two bugs with module file parsing:
* Detection of the CRAY_LD_LIBRARY_PATH variable was broken by #9100.
This fixes it and adds a test for it.
* For module names like "foo-bar/1.0", the associated PACKAGE_DIR
environment variable name would be "FOO_BAR_DIR", but Spack was not
parsing the components and not converting "-" to "_"
Fixes#9166
This is intended to reduce errors related to lock timeouts by making
the following changes:
* Improves error reporting when acquiring a lock fails (addressing
#9166) - there is no longer an attempt to release the lock if an
acquire fails
* By default locks taken on individual packages no longer have a
timeout. This allows multiple spack instances to install overlapping
dependency DAGs. For debugging purposes, a timeout can be added by
setting 'package_lock_timeout' in config.yaml
* Reduces the polling frequency when trying to acquire a lock, to
reduce impact in the case where NFS is overtaxed. A simple
adaptive strategy is implemented, which starts with a polling
interval of .1 seconds and quickly increases to .5 seconds
(originally it would poll up to 10^5 times per second).
A test is added to check the polling interval generation logic.
* The timeout for Spack's whole-database lock (e.g. for managing
information about installed packages) is increased from 60s to
120s
* Users can configure the whole-database lock timeout using the
'db_lock_timout' setting in config.yaml
Generally, Spack locks (those created using spack.llnl.util.lock.Lock)
now have no timeout by default
This does not address implementations of NFS that do not support file
locking, or detect cases where services that may be required
(nfslock/statd) aren't running.
Users may want to be able to more-aggressively release locks when
they know they are the only one using their Spack instance, and they
encounter lock errors after a crash (e.g. a remote terminal disconnect
mentioned in #8915).
When a Spack Executable was configured to capture stderr and the
process failed, the error messages of the process were discarded.
This made it difficult to understand why the process failed. The
exception is now updated to include the stderr of the process when
the Executable captures stderr.
Adds 'code' to the list of suffixes that are excluded from version
parsing of URLs, such that if a URL contains the string
'cistem-1.0.0-beta-source-code', a version X will substitute in to
produce a URL with cistem-X-source-code ('source' was already excluded).
The 'cistem' package version is updated to make use of this (and fix
a fetching bug with the cistem package). A unit test is added to check
this parsing case.
Improve Spack's parsing of module show to eliminate some false
positives (e.g. accepting MODULEPATH when it is in fact looking for
PATH). This makes the following changes:
* Updates the pattern searching for several paths to avoid the case
where they are prefixes of unwanted paths
* Adds a warning message when an extracted path doesn't exist (which
may help catch future module parsing bugs faster)
* Adds a test with the content mentioned in #9083
Spack originally handled environment modifications in the following
order:
1. clear environment variables
(unless Spack was invoked with --dirty)
2. apply spack-specific environment variable updates,
including variables set by Spack core like CC/PKG_CONFIG_PATH
and those set by installed dependencies (e.g. in
setup_dependent_environment)
3. load all external/compiler modules
1 and 2 were done together. This splits 1 into its own function and
imposes the following order for environment modifications:
1. clear environment variables
2. load all external/compiler modules
3. apply spack-specific environment variable updates
As a result, prepend-path actions taken by Spack (or installed Spack
dependencies) take precedence over prepend-path actions from compiler
and external modules. Additionally, when Spack (or a package
dependency) sets/unsets an environment variable, that will override
the actions of external/compiler modules.
* Add 'extra_env' argument to Executable.__call__: this will be added
to the environment but does not affect whether the current
environment is reused. If 'env' is not set, then the current
environment is copied and the variables from 'extra_env' are added
to it.
* MakeExecutable can take a 'jobs_env' parameter that specifies the
name of an environment variable used to set the level of parallelism.
This is added to 'extra_env' (so does not affect whether the current
environment is reused).
* CMake-based Spack packages set 'jobs_env' when executing the 'test'
target for make and ninja (which does not use -j)
Consolidate prefix calculation logic for intel packages into the
IntelPackage class.
Add documentation on installing Intel packages with Spack an
(alternatively) adding them as external packages in Spack.
The functions returning the default scope to be modified or listed
have been moved from spack.cmd to spack.config.
Lmod now writes the guessed core compiler in the default modify scope
instead of the 'site' scope.
closes#8916
Currently Spack ends with an error if asked to write lmod modules files
and the 'core_compilers' entry is not found in `modules.yaml`. After
this PR an attempt will be made to guess that entry and the site
configuration file will be updated accordingly.
This is similar to what Spack already does to guess compilers on first
run.
- Support for Python 3.3 isn't really needed, as nothing uses it as the
default system Python, and nearly everyone will have a newer Python 3
version installed.
#8223 replaced regex-based makefile target parsing with an invocation of
"make -q". #8818 discovered that "make -q" can result in an error for some
packages.
Also, the "make -q" strategy relied on interpreting the error code, which only
worked for GNU Make and not BSD Make (which was deemed acceptable at
the time). As an added bonus, this implementation ignores the exit code and
instead parses STDERR for any indications that the target does not exist; this
works for both GNU Make and BSD Make.
#8223 also updated ninja target detection to use "ninja -t targets". This does
not change that behavior but makes it more-explicit with "ninja -t targets all"
This also adds tests for detection of "make" and "ninja" targets.
Fixes#9001#8289 added support for install_tree and copy_tree to merge into an existing
directory structure. However, it did not properly handle relative symlinks and
also removed support for the 'ignore' keyword. Additionally, some of the tests
were overly-strict when checking the permissions on the copied files.
This updates the install_tree/copy_tree methods and their tests:
* copy_tree/install_tree now preserve relative link targets (if the symlink in the
source directory structure is relative, the symlink created in the destination
will be relative)
* Added support for 'ignore' argument back to copy_tree/install_tree (removed
in #8289). It is no longer the object output by shutil.ignore_patterns: you pass a
function that accepts a path relative to the source and returns whether that
path should be copied.
* The openfoam packages (currently the only ones making use of the 'ignore'
argument) are updated for the new API
* When a symlink target is absolute, copy_tree and install_tree now rewrite the
source prefix to be the destination prefix
* copy_tree tests no longer check permissions: copy_tree doesn't enforce
anything about permissions so its tests don't check for that
* install_tree tests no longer check for exact permission matching since it can add
file permissions
- `imp` is deprecated and seems to have started having some weird
issues on certain Linux versions.
- In particular, the file argument to `load_source` is ignored on
arch linux with Python 3.7.
- `imp` is the only way to do imports in 2.6, so we'll keep it around for
now and use it if importlib won't work.
- `importlib` is the new import system, and it allows us to get
lower-level access to the import implementation.
- This consolidates all import logic into `spack.util.imp`, and make it
use `importlib` if it's avialable.
Replace use of `shutil.copytree` with `copy_tree` and `install_tree` functions in `llnl.util.filesystem`.
- `copy_tree` copies without setting permissions. It should be used to copy files around in the build directory.
- `install_tree` copies files and sets permissions. It should be used to copy files into the installation directory.
- `install` and `copy` are analogous single-file functions.
- add more extensive tests for these functions
- update packages to use these functions.
- dependency patching test didn't attempt to apply patches; just to see
whether they were on the spec.
- it applies the patch now and verifies that that patch was applied.
* Branch with the meson build-system
* Fix build_environment for dual loads and add create code
* Add documentation
* Fixed option list
* Update build_system_guess for meson
* Fixed documentation errors
* Added meson to build and configure and updated documentation
* fix typos
- cc cleanup caused a parsing regression in flag handling
- We added proper quoting to array expansions, but flag variables were
never actually converted to arrays. Old code relied on this.
This commit:
- Adds reads to convert flags to arrays.
- Makes the cc test check for improper space handling to prevent future
regressions.
- flags were prepended in reverse order to args, but this makes it hard
to see what order they'll be in on the final command line.
- add them in the order they'll appear to make cc easier to maintain.
- simplify code for assembling the command line
- fix separator used in SPACK_SYSTEM_DIRS test
- This corrects most of the issues found by shellcheck
- This also uses ':' as the delimiter for SPACK_SYSTEM_DIRS, for
consistency with other variables.
- filtering using sed causes most builds to slow down quite a bit, as the
compiler wrapper has to run sed many times, and *it* runs many times
- do the system directory parsing directly in bash
- Add tests to ensure that RPATHs are not added in cc mode, which can
cause some builds to fail.
- Change cc.py to use pytest style
- Instead of writing out all the flags, break the flags down into
variables so that it's easy to read what each test is supposed to
check. This should make cc.py more maintainable in the future.
- Adding -L and -Wl,-rpath to compile-only command lines ("cc mode" or
"-c") causes clang (if not also other compilers) to emit warnings that
confuse configure systems.
- Clang will print warnings about unused command-line arguments.
- This fix ensures that -L and -Wl,-rpath are not added if the compile
line is just building an object file with -c
- This also cleans up the cc script in several places.
Spack currently prepends include paths, library paths, and rpaths to the
compile line. This causes problems when a header or library in the package
has the same name as one exported by one of its dependencies. The
*dependency's* header will be preferred over the package's, which is not
what most builds expect. This also breaks some of our production codes.
This restores the original cc behavior (from *very* early Spack) of parsing
compiler arguments out by type (`-L`, `-I`, `-Wl,-rpath`) and reconstituting
the full command at the end.
`<includes> <other_args> <library dirs> <rpaths>`
This differs from the original behavior in one significant way, though: it
*appends* the library arguments so that dependency libraries do not shadow
those in the build.
This is safe because semantics aren't affected by *interleaving* `-I`, `-L`,
and `-Wl,-rpath` arguments with others, only with each other (so the order of
two `-L` args affects the search path, but we search for all libraries on the
command line using the same search path).
We preserve the following:
1. Any system directory in the paths will be listed last.
2. The root package's include/library/RPATH flags come before flags of the
same type for any dependency.
3. Order will be preserved within flags passed by the build (except system
paths, which are moved to be last)
4. Flags for dependencies will appear between the root flags and the system
flags, and the flags for any dependency will come before those for *its*
dependencies (this is for completeness -- we already guarantee this in
`build_environment.py`)
* Fix performance issue when compiling.
Spack was doing active wait when compiling, spoiling one core.
My fix consists in not setting any timeout for select, instead of
the previous 0 second.
* Fix comments about select.select timeout
- This was a nasty workaround due to the way our compiler wrappers used
to work. We don't want to have to modify our elfutils installation to
install libdwarf.
- Since cd9691de5, we no longer need this because the package will always
come before dependencies in our include order.
Spack currently prepends include paths, library paths, and rpaths to the compile line. This causes problems when a header or library in the package has the same name as one exported by one of its dependencies. The *dependency's* header will be preferred over the package's, which is not what most builds expect. This also breaks some of our production codes.
This restores the original cc behavior (from *very* early Spack) of parsing compiler arguments out by type (`-L`, `-I`, `-Wl,-rpath`) and reconstituting the full command at the end.
`<includes> <other_args> <library dirs> <rpaths>`
This differs from the original behavior in one significant way, though: it *appends* the library arguments so that dependency libraries do not shadow those in the build.
This is safe because semantics aren't affected by *interleaving* `-I`, `-L`, and `-Wl,-rpath` arguments with others, only with each other (so the order fo two `-L` args affects the search path, but we search for all libraries on the command line using the same search path).
We preserve the following:
1. Any system directory in the paths will be listed last.
2. The root package's include/library/RPATH flags come before flags of the same type for any dependency.
3. Order will be preserved within flags passed by the build (except system paths, which are moved to be last)
4. Flags for dependencies will appear between the root flags and the system flags, and the flags for any dependency will come before those for *its* dependencies (this is for completeness -- we already guarantee this in `build_environment.py`)
- previously, output could be confusing when deptypes were only shown for
one dependent when a node had *multiple* dependents
- also fix default coverage of `Spec.tree()`: it previously defaulted to
cover only build and link dependencies, but this is a holdover from
when those were the only types.
- Previously, Spack didn't check the arguments you put in version()
directives.
- So, you could do something like this, where there are arguments for a
URL fetcher AND for a git fetcher:
version('1.0', md5='abc123', git='https://foo.bar', commit='feda2343')
- Now, we check the arguments before constructing a fetcher, to ensure
that each package has *only* arguments for a single type of fetcher.
- Also added `test_package_version_consistency()` to the `package_sanity`
test, so that all builtin packages are required to have valid
`version()` directives.
- packagers can specify two top-level fetch URLs if one is `url`
- e.g., `url` and `git` or `url` and `svn`
- allow only one VCS fetcher so we can differentiate between URL and VCS.
- also clean up fetcher logic and class structure
- Packages can remove the top-level `url` attribute and still work
- These are now legal:
- Packages with *only* version-specific URLs (even with gaps)
- Packages with a top-level git/hg/svn attribute and `version`
directives for that.
- If a package has both a top-level hg/git/svn attribute AND a top-level
url attribute, the url attribute takes precedence.
Some packages do not have a `url` and are instead downloaded via `git`,
`hg`, or `svn`. Some packages like `spectrum-mpi` cannot be downloaded at
all, and are placeholder packages for system installations. Previously,
`__init__()` in `PackageBase` crashed if a package did not have a `url`
attribute defined.
I hacked this section of code out, but I have no idea what the
repercussions of that are.
- This hard-codes the hash lengths rather than computing them on import.
- Also cleans up the code in `spack.util.crypto` to make it easier to
understand.
Fix this output error:
```
$ spack -m module loads mpileaks
==> Error: `spack module loads -m t -m c -m l ...` has been moved. Try this instead:
$ spack module t loads mpileaks
$ spack module c loads mpileaks
$ spack module l loads mpileaks
```
In case a deprecated form of the module command is used, the program
will exit non-zero and print an informative error message suggesting
which command should be used instead.
As requested in the review all the commands meant to manage module
files have been grouped under the `spack module` command.
Unit tests have been refactored to match the new command structure.
fixes#2215fixes#2570fixes#6676fixes#7281closes#3827
This PR reverts the use of `spack module loads` in favor of
`spack module find` when loading module files via Spack. After this PR
`spack load` will accept a single spec at a time, and will be able
to interpret correctly the `--dependencies` option.
fixes#4400
The feature requested in #4400 was already part of the module file
configuration, but it was neither tested nor documented. This
commit takes care of adding a few lines in the documentation and a
regression test.
This just because the fixture has been moved one level above the one
it was originally defined. In this more general context there's more
than one configuration file that could be patched for tests.
'spack module' has been split into multiple commands, each one tied to a
specific module type. This permits the specialization of the new
commands with features that are module type specific (e.g. set the
default module file in lmod when multiple versions of the same package
are installed at the same time).
- repo membership test was broken by the refactor of spack/__init__.py
- refactor singleton so that 'spec in repo' works again for `spack.repo.path`
- fix spec command and add basic tests for `spack spec` and `spack spec --yaml`
- There was a lot of documentation in `PackageBase` dating back to the
very first versions of Spack.
- It was repetitive and out of date, and the docs at spack.readthedocs.io
are better.
- Remove the outdated specifics, and leave the minimal useful set of
developer docs in `package.py`.
- This changes `get_checksums_for_versions` to generate code that uses an
explicit `sha256` argument instead if the bare `md5` hash we used to
generate.
- also use a generic digest parameter for the `version` directive, rather
than a specific `md5` parameter.
- Add command-line scope option to Spack
- Rework structure of main to allow configuration system to raise
errors more naturally
Co-authored-by: Todd Gamblin <tgamblin@llnl.gov>
- Fixes a bug in `llnl.util.lock`
- Locks in the current directory would fail because the parent directory
was the empty string.
- Fix this and return '.' for the parent of locks in the current
directory.
Replace regex-based target detection for Makefiles with a preliminary "make -q"
to check if a target exists. This does not work for NetBSD make; additional work
is required to detect if NetBSD make is present and to use a regex in that case.
The affected makefile target checks are only performed when the "--test" flag is
added to a "spack install" invocation.
Fixes#8036
Before this PR Package.installed was returning True if the spec prefix
existed, without checking the DB. This is wrong for external packages,
whose prefix exists before being registered into the DB. Now the property
checks for both the prefix and a DB entry.
Spack provides a number of classes based on commonly-used build systems
that users can extend when writing packages; the classes provide functionality
to perform the actions relevant to the build system (e.g. running "configure" for
an Autotools-based package). This adds documentation for classes supporting the
following build systems:
* Makefile
* Autotools
* CMake
* QMake
* SCons
* Waf
This includes build systems for managing extensions of the following packages:
* Perl
* Python
* R
* Octave
This also adds documentation on implementing packages that use a custom build
system (e.g. Perl/CMake).
Spack also provides extendable classes which aggregate functionality for related
sets of packages, e.g. those using CUDA. Documentation is added for
CudaPackage.
- The setup-env.sh script currently makes two calls to spack, but it
should only need to make one.
- Add a fast-path shell setup routine in `main.py` to allow the shell
setup to happen in a single, fast call that doesn't load more than it
needs to.
- This simplifies setup code, as it has to eval what Spack prints
- TODO: consider eventually making the whole setup script the output of a
spack command
* update help of `clean --all` to include `-p`
* remove old orphaned `.pyc` removal
* restrict removal or orphaned pyc files to `lib/spack` and `var/spack`
- Clean up error messages for when a lock can't be created, or when an
exclusive (write) lock can't be taken on a file.
- Add a number of subclasses of LockError to distinguish timeouts from
permission issues.
- Add an explicit check to prevent the user from taking a write lock on a
read-only file.
- We had a check for this for when we try to *upgrade* a lock on an RO
file, but not for an initial write lock attempt.
- Add more tests for different lock permission scenarios.
- write locks previously wrote information about the lock holder (host
and pid), and read locks woudl read this in.
- This is really only for debugging, so only enable it then
- add some tests that target debug info, and improve multiproc lock test
output
When a user specifies a URL for a specific version of a package, Spack originally
would use that URL for all newer versions of the package. This behavior has
proven to be generally more harmful than useful, so this PR removes the feature
such that a version-specific URL override affects only that version.
If the user sets "ccache: true" in spack's config.yaml, Spack will use an available
ccache executable when compiling c/c++ code. This feature is disabled by default
(i.e. "ccache: false") and the documentation is updated with how to enable
ccache support
Functional updates:
- `python` now creates a copy of the `python` binaries when it is added
to a view
- Python extensions (packages which subclass `PythonPackage`) rewrite
their shebang lines to refer to python in the view
- Python packages in the same namespace will not generate conflicts if
both have `...lib/site-packages/namespace-example/__init__.py`
- These `__init__` files will also remain when removing any package in
the namespace until the last package in the namespace is removed
Generally (Updated 2/16):
- Any package can define `add_files_to_view` to customize how it is added
to a view (and at the moment custom definitions are included for
`python` and `PythonPackage`)
- Likewise any package can define `remove_files_from_view` to customize
which files are removed (e.g. you don't always want to remove the
namespace `__init__`)
- Any package can define `view_file_conflicts` to customize what it
considers a merge conflict
- Global activations are handled like views (where the view root is the
spec prefix of the extendee)
- Benefit: filesystem-management aspects of activating extensions are
now placed in views (e.g. now one can hardlink a global activation)
- Benefit: overriding `Package.activate` is more straightforward (see
`Python.activate`)
- Complication: extension packages which have special-purpose logic
*only* when activated outside of the extendee prefix must check for
this in their `add_files_to_view` method (see `PythonPackage`)
- `LinkTree` is refactored to have separate methods for copying a
directory structure and for copying files (since it was found that
generally packages may want to alter how files are copied but still
wanted to copy directories in the same way)
TODOs (updated 2/20):
- [x] additional testing (there is some unit testing added at this point
but more would be useful)
- [x] refactor or reorganize `LinkTree` methods: currently there is a
separate set of methods for replicating just the directory structure
without the files, and a set for replicating everything
- [x] Right now external views (i.e. those not used for global
activations) call `view.add_extension`, but global activations do not
to avoid some extra work that goes into maintaining external views. I'm
not sure if addressing that needs to be done here but I'd like to
clarify it in the comments (UPDATE: for now I have added a TODO and in
my opinion this can be merged now and the refactor handled later)
- [x] Several method descriptions (e.g. for `Package.activate`) are out
of date and reference a distinction between global activations and
views, they need to be updated
- [x] Update aspell package activations
- Spack was assuming that a group with gid == current uid would always exist.
- This was breaking the travis build for macos.
- also fix issue with the DB tarball test finding coverage filesx
- pytest was not reporing the correct version from pytest.__version__.
It reported 'unknown'
- this fixes issues on some systems where system-installed pytest plugins
would try to use the version and convert it to an int
The following improvements are made to cxx standard support
(e.g. compiler.cxxNN_flag functions) in compilers:
* Add cxx98_flag property
* Add support for throwing an exception when a flag is not supported (previously
if a flag was not supported the application was terminated with tty.die)
* The name of the flag associated with e.g. c++14 standard support changes for
different compiler versions (e.g. c++1y vs c++14). This makes a few corrections
on what flag to return for which version.
* Added tests to confirm that versions report expected flags for various c++
standards (or raise an exception for versions that don't provide a given cxx
standard)
Note that if a given cxx standard is the default, the associated flag property will
return ""; cxx98 is assumed to be the default standard so this is the behavior for
the associated property in the base compiler class.
Package changes:
* Improvements to the boost spec to take advantage of the improved standard
flag facility.
* Update the clingo spec to catch the new exception rather than look for an
empty flag to indicate non-support (which is not part of the compiler flag API)
Fixes#7885#7193 added the patches_to_apply function to collect patches which are then
applied in Package.do_patch. However this only collects patches that are
associated with the Package object and does not include Spec-related patches
(which are applied by dependents, added in #5476).
Spec.patches already collects patches from the package as well as those applied
by dependents, so the Package.patches_to_apply function isn't necessary. All
uses of Package.patches_to_apply are replaced with Package.spec.patches.
This also updates Package.content_hash to require the associated spec to be
concrete: Spec.patches is only set after concretization. Before this PR, it was
possible for Package.content_hash to be valid before concretizing the associated
Spec if all patches were associated with the Package (vs. being applied by
dependents). This behavior was unreliable though so the change is unlikely to
be disruptive.
Fixes#8345
Spack environment modifications are applied before modules are loaded; this
includes settings to CC, FC, F77, and CXX, which point to the Spack compiler
wrappers. If the loaded modules set CC, this overrides the Spack compiler
wrappers. This PR adds a context manager to preserve the values of CC etc. that
are set by Spack: any effects on the CC, FC, F77, and CXX variables from modules
are undone and their original values are restored.
* pybind11: test support
Add a test functionality to pybind11.
* CMake: test also on "make check"
Some projects use non-CTest manual targets for tests.
* extend Prefix class with join() member to support dynamic directories
* add more tests for Prefix.join()
* more tests for Prefix.join()
* add docstring
* add example to docstring of Prefix class
* cleanup Prefix.join() tests
* use Prefix.join() in Packaging Guide
Fixes#8217
Trying to relocate a distribution when the new and old paths are
equal leads to failure, because the test that ensures that no
unrelocated bits are left over always fails. As an example, this
occurs if a user installs a package, generates a binary with it
using 'spack buildcache', uninstalls it, and then attempts to
reinstall into the same spack installation using the generated
binary package.
This updates the relocation check to accept the presence of the
old prefix in binaries if the package is being reinstalled into
its original location.
* allow user to constrain dependencies that are added conditionally
* remove check for not-visited deps from normalize, move it to concretize. The check now runs after the concretization loop completes (so an error is only reported if the user-mentioned spec doesnt appear anywhere in the dag)
* remove separate full_spec_deps variable; rename spec_deps to all_spec_deps to clarify that it merges user-specified dependencies with derived dependencies
* add unit test to confirm new functionality
- `spack config blame` is similar to `spack config get`, but it prints
out the config file and line number that each line of the merged
configuration came from.
- This is a debugging tool for understanding where Spack config settings
come from.
- add tests for config blame
Fixes: #8258#8090 altered import behavior so that import spack no longer
provides access to many other Spack modules. This addresses
a case which depended on the prior behavior and was not
updated as part of #8090. This particular import error only
came up when users were setting compiler flags on specs.
See also: #8194
- there were some leftover spack.* names being used after we removed
globals and moved everything in the top-level namespace to spack.pkgkit
- point those references to their new homes
- remove most `import spack` statements, except for files that need
`spack_version`
- import spack is no longer sufficient to use submodules
(e.g. spack.directives).
- these submodules must be imported directly. Update references
accordingly.
- Spack packages were originally expected to call `from spack import *`
themselves, but it has become difficult to manage imports in the
Spack core.
- the top-level namespace polluted by package symbols, and it's not
possible to avoid circular dependencies and unnecessary module loads in
the core, given all the stuff the packages need.
- This makes the top-level `spack` package essentially empty, save for a
version tuple and a version string, and `from spack import *` is now
essentially a no-op.
- The common routines and directives that packages need are now in
`spack.pkgkit`, and the import system forces packages to automatically
include this so that old packages that call `from spack import *`
will continue to work without modification.
- Since `from spack import *` is no longer required, we could consider
removing ``from spack import *`` from packages in the future and
shifting to ``from spack.pkgkit import *``, but we can wait a while to
do this.
- spack.util.lock behaves the same as llnl.util.lock, but Lock._lock and
Lock._unlock do nothing.
- can be disabled with a control variable.
- configuration options can enable/disable locking:
- `locks` option in spack configuration controls whether Spack will use filesystem locks or not.
- `-l` and `-L` command-line options can force-disable or force-enable locking.
- Spack will check for group- and world-writability before disabling
locks, and it will not allow a group- or world-writable instance to
have locks disabled.
- update documentation
- Spack core has long used llnl.util.filesystem.join_path, but
os.path.join is pretty much the same thing, and is more efficient.
- Use os.path.join in the core Spack code from now on.
- simplify the singleton pattern across the codebase
- reduce lines of code needed for crufty initialization
- reduce functions that need to mess with a global
- Singletons whose semantics changed:
- spack.store.store() -> spack.store
- spack.repo.path() -> spack.repo.path
- spack.config.config() -> spack.config.config
- spack.caches.fetch_cache() -> spack.caches.fetch_cache
- spack.caches.misc_cache() -> spack.caches.misc_cache
- `spack.cmd.all_commands` does a directory listing on
`lib/spack/spack/cmd`, regardless of whether it is needed
- make this lazy so that the directory listing won't happen unless it's
necessary.
- It turns out that jsonschema is one of the more expensive imports.
- move imports of jsonschema into functions to avoid the performance hits
for calls that don't need config.
- spack.store was previously initialized at the spack.store module level,
but this means the store has to be initialized on every spack call.
- this moves the state in spack.store to a singleton so that the store is
only initialized when needed.
- spack.repository module is now spack.repo
- `spack.repo` is now `spack.repo.path()` and loaded lazily
- Added `spack.repo.get()` and `spack.repo.all_package_names()` as
convenience functions to simplify the new lazy interface.
- updated tests and code
- no longer require `spack_version` to be a Version (it isn't used that
way anyway)
- use a simple tuple `spack_version_info` with major, minor, patch
versions
- generate `spack_version` from the tuple
- replace `spack.config.get_configuration()` with `spack.config.config()`
- replace `get_config`/`update_config` with `get`, `set`
- add a path syntax that can be used to refer to specific config options
without firt getting the entire configuration dict
- update usages of `get_config` and `update_config` to use `get` and `set`
- Current configuration code forces the config system to be initialized
at module scope, so configs are parsed on every Spack run, essentially
before anything else.
- We need more control over configuration init order, so move the config
scopes into a class and reduce global state in config.py
Fixes#2781
This PR introduces a new attribute for packages called
`archive_files`, which designates files that should be saved from
a package build (e.g. the config.log generated during autotools
builds).
The attribute contains a list of glob expressions; Any file that
matches will be archived in the `<prefix>/.spack/archived-files`
directory. Errors that occur when archiving files are collected and
reported in a file named `<prefix>/.spack/archived-files/errors.txt`.
`AutotoolsPackage` and `CMakePackage` provide a sensible default
override for this attribute.
fixes#7941
Modified string representation of Specs to add a space before deps
Unit-tests have been modified accordingly
Added a test for regression on #7941
Fixes#7924
Spack's yaml schema validator was initializing all instances of
unspecified properties with the same object, so that updating the
property for one entry was updating it for others (e.g. updating
versions for one package would update it for other packages).
This updates the schema validator to instantiate defaults with
separate object instances and adds a test to confirm this behavior
(and also confirms #7924 without this change).
* Use reported version of IBM XL Fortran compiler for compiler versions >= 16.0.
Starting with the April 2018 release, the IBM XL C and Fortran compilers report the same version, 16.0. Consequently, there is no need to downgrade the Fortran compiler version to match that of the C compiler.
* Use GitLab's API endpoint for fetching a git snapshot.
* More GitLab packages use the API.
* find_list_url for GitLab's API URLs.
* Flake8
* Url for 'hacckernels'.
* Check GitLab API regexps before the non-API ones.
Activating a package that is already activated now sends a `tty.msg`
and returns.
```
-bash-4.2$ ~/spack/bin/spack activate aspell6-en
==> Package aspell6-en/lc4v24f is already activated.
```
* Better error message for spack providers
fixes#1355
`spack providers` now outputs a sensible error message if non-virtual
specs are provided as arguments:
```
$ spack providers mpi zlib petsc
==> Error: non-virtual specs cannot be part of the query [zlib, petsc]
```
Formatting of the output changed slightly.
* Calling 'spack providers' without arguments print the virtual pkg list
Also, the error message in case of a wrong parameter has been improved
to show the list of valid packages.
* Avoid printing headers if stdout is not a tty
* The provider list is formatted with colify if not in a tty
* Added a test to check the list of providers returned from the command
Popping the when spec from kwargs in the extends directive breaks
class inheritance. Inheriting classes do not find their when spec.
We now get the when spec from kwargs instead, leaving it to be found
by any downstream package classes.
fixes#7705
Package.provides now checks constraints to ensure that a spec provides
a given virtual package. Note that 'strict=True' is not passed to
satisfies as this function is also used during concretization.
Fixes#7548
This updates the "spack view" command to use the same parsing logic
as "spack install" on the user-provided specs. For example you can
provide a DAG hash to refer to an exact installed spec instead of
specifying name, compiler, etc.
This fixes a check that decides when to skip buildcache relocation.
Originally the check was flawed in two ways: it would skip if the
source prefix matched the destination prefix, which no longer
matters since the source prefix is replaced with a placeholder
(so it always needs to be updated); it also would skip relocation
if the rpaths were not relative, when in fact it should be the
opposite (binaries without relative rpaths *should* be relocated,
and those without don't need it).
- FastPackageChecker was being called at startup every time Spack runs,
which takes a long time on networked filesystems. Startup was taking
5-7 seconds due to this call.
- The checker was intended to avaoid importing all packages (which is
really expensive) when all it needs is to stat them. So it's only
"fast" for parts of the code that *need* it.
- This commit makes repositories instantiate the checker lazily, so it's
only constructed when needed.
- This was needed when we transitioned to all lowercase packages because
git didn't handle case changes well on case-insensitive filesystems.
- Now it just adds extra stat calls to startup, and we check for
all-lowercase package names in tests, so we'll remove it.
- people using really old versions of Spack can re-clone.
* Create unload_module method
Extract code from load_module into unload_module.
* Unload modules to create a clean env on Cray
removes cray-libsci, cray-mpich and darshan to prevent any silent
linking with those packages.
* Add format to separate target and os for path
spec format can now handle separations of target and os for setting
up the path.
* Added ${PLATFORM} et al to spec.format()
${PLATFORM}, ${OS}, ${TARGET}
* Update tests
Updated tests and got rid of unnecessary code.
* Also update documentation to reflect this new ability.
* Add default path scheme to config.yaml
Added default path scheme to config.yaml. Users can overwrite this
section if they want.
* Speedup the default 'libs' property search - important for external
packages.
* As advised by @alalazo, use tuples instead of lists inside
_libs_default_handler.
* Added installation date and time to the database
Information on the date and time of installation of a spec is recorded
into the database. The information is retained on reindexing.
* Expose the possibility to query for installation date
The DB can now be queried for specs that have been installed in a given
time window. This query possibility is exposed to command line via two
new options of the `find` command.
* Extended docstring for Database._add
* Use timestamps since the epoch instead of formatted date in the DB
* Allow 'pretty date' formats from command line
* Substituted kwargs with explicit arguments
* Simplified regex for pretty date strings. Added unit tests.
This updates architecture concretization to
* Search for the nearest parent in the DAG for architecture information
rather than defaulting to the root of the DAG
* Propagate architecture settings transitively, such that if for
example the target is set at the root of the dag it will set the
same target on indirect dependencies (assuming no intermediate
dependency specifies a separate target). Previously this occurred
in general but under some conditions did not, for example if an
intermediate dependency specified some subset of architecture
properties.
* Create mirror for system with different compilers
Spack concretizes the spec provided by the user in
"spack mirror create" to ensure downloading the right
dependencies. Under normal circumstances concretization
requires that the chosen compiler exists on the system,
but this is not required when creating download mirrors
for other systems, so this requirement is removed in that
case.
* Add test for disabling compiler existence check
* Update compiler existence checking logic
* improve test for disabling compiler existence check
* make py-setuptools a run-time-only dep for py-basemap and patch python package to only apply setuptools flag for build deps
* py-qtconsole does not require setuptools
* This allows Spack to work with MD5 hashes on machines with openssl in FIPS mode.
* We are still using MD5 for validation in many places, and a later PR will replace all uses of MD5 with SHA256.
* This is a quick fix until that happens.
- transitive dependencies were not being handled correctly
- restructure code to do recursion and mark visited packages properly
- add `-V` option to *not* expand virtuals in spack dependencies
This re-adds the option to create and install unsigned tarballs, now
with the -u option (--unsigned) rather than the -y option.
This also changes the "keys" command, replacing the -y/--yes-to-all
option with the -t/--trust option (which has the same effect but is
more-clearly named).
Fixes#7130
shutil.move expects a source path like "/x/y/" to be a directory and
fails if "/x/y" is a symlink. This invokes realpath on the source
path to avoid the issue.
Fixes#7356
In some cases OperatingSystem (e.g. LinuxDistro) was getting
instantiated with a version that contains dashes. This breaks because
the concretizer later converts this value to a string and re-parses
it, and the '-' character is used to separate architecture components.
This adds a guard in the initializer to convert '-' to '_'.
Fixes#7237Fixes#6404Fixes#6418Fixes#6369
Identify when binary relocation fails to remove all instances of the
source prefix (and report an error to the user unless they specify
-a to allow the old root to appear). This check occurs at two stages:
during "bincache create" all instances of the root are replaced with
a special placeholder string ("@@@@..."), and a failure occurs if the
root is detected at this point; when the binary package is extracted
there is a second check. This addresses #7237 and #6418.
This is intended to be compatible with previously-created binary
packages.
This also adds:
* Better error messages for "spack install --use-cache" (#6404)
* Faster relocation on Mac OS (using a single call to
install_name_tool for all files rather than a call for each file)
* Clean up when "buildcache create" fails (addresses #6369)
* Explicit error message when the spack instance extracting the binary
package uses a different install layout than the spack instance that
created the binary package (since this is currently not supported)
* Remove the option to create unsigned binary packages with -y
This updates Cray.setup_platform_environment to use cray-specific
pkgconfig paths so that all providers of 'pkgconfig' have access
to them (previously the 'pkg-config' provider had this but the
'pkgconf' provider did not).
* [SPACK/spec.py] When a query through ForwardQueryToPackage returns
'None', treat that as query failure and raise RuntimeError with
suitable message. This overrides the current behavior to raise an
AttributeError which is now triggered only when no suitable query
property is found and there is no default handler.
* [spack/spec.py] Fix style.
* [SPACK/spec.py] In case of query failure, i.e. property returning
'None', raise AttributeError instead of RuntimeError in order to
pass the unit test. Also, small update in the logic distinguishing
query failure and lack of relevant property/attribute handling.
This updates the fix_darwin_install_name function to use the Spack
Executable object to run install_name_tool, which ensures that
process output is formatted as a 'str' for python2 and python3.
Originally fix_darwin_install_name was invoking subprocess.Popen
directly.
Fixes#5189
When working with non-normalized paths containing ".." on some
file systems, Spack was found to encounter a permission error when
writing to the path. This normalizes a path written by the
intel-parallel-studio package and also normalizes all paths
written by the license install hook (for all packages) to avoid
this issue for intel-parallel-studio.
Following the discussion with Todd and Adam, find has been modified to
accept glob expressions. This should not affect performance as every
glob implementation I inspected has 3 cases (no wildcard, wildcard but
no directories involved, wildcard and directories involved) and uses
fnmatch underneath.
Mixins have been changed to do by default a non-recursive search (but
a recursive search can still be triggered using the recursive keyword).
Following a comment from Todd, the search path for the files listed in
`filter_compiler_wrappers` can now be narrowed. Anyhow, the function
implementation still makes use of `find`, the rationale being that we
have already seen packages that install artifacts in e.g. architecture
dependent folders. The possibility to have a relative search path might
be a good compromise between the previous approach and the one suggested
in the review.
Also: 'ignore_absent' and 'backup' keyword arguments can be optionally
forwarded to `filter_file`.
Following comments from Todd:
- the call to tty.debug has been moved deeper, to log the filtering of each file
- the shadowing on the name "kwargs" is avoided
Implemented a declarative syntax for the additional behavior that can
get attached to classes. Implemented a function to filter compiler
wrappers that uses the mechanism above.
- command reference now includes usage for all Spack commands as output
by `spack help`. Each command usage links to any related section in
the docs.
- added `spack commands` command which can list command names,
subcommands, and generate RST docs for commands.
- added `llnl.util.argparsewriter`, which analyzes an argparse parser and
calls hooks for description, usage, options, and subcommands
- Shorten Spack command usage for short options. Short options are now
shown as [-abc] instead of as [-a] [-b] [-c]
- fix bug that mixed long and short options for top-level `spack help`
- Add proper help for `spack buildcache` subcommands
- Reorganize the help categories of Spack commands so that buildcache is
in packaging and diy and setup are now in build.
- previously commands with this argument showed a long list of choices
that were platform specific.
- use a better metavar: {defaults,system,site,user}[/PLATFORM]
Fixes#7159
When activating extensions in external views, the --ignore-conflicts
option was being ignored. In this particular issue the conflict was
for the duplicate __init__ file for multiple python packages in the
same namespace, but in general any conflict for extensions would
cause an error whether or not --ignore-conflicts was set.
This also renames the 'force' option of do_activate to
'with_dependencies' and updates views to call do_activate with this
set to False (since it traverses the dependency dag anyway). This
isn't strictly required, it just avoids redundant calls.
This reorganizes most sections and rewords a significant portion of
the content (including all introductions) but keeps all the examples.
* Remove section 'What happens at subscript time' from tutorial:
it is too detailed for a tutorial
* Move the 'Extra query parameters' and 'Attach attributes to other
packages' sections into a separate grouping 'Other packaging topics'
* move the 'Set variables at build time yourself' section after
'Set environment variables in dependents' section since the latter
is more motivating
* start the 'set environment variables at build-time for yourself'
section with qt as an example
* renamed section 'specs build interface' to 'retrieving library
information' and updated section introduction
* renamed section 'a motivating example' to 'accessing library
dependencies'; split out the material which deals with implementing
.libs for netlib-lapack into a separate section called 'providing
libraries to dependents'. consolidated in material from the section
'single package providing multiple virtual specs' since
netlib-lapack is an example of this (this removes the material
about intel-parallel studio)
* Allow dashes in command names and fix command name handling
- Command should allow dashes in their names like the reest of spack,
e.g. `spack log-parse`
- It might be too late for `spack build-cache` (since it is already
called `spack buildcache`), but we should try a bit to avoid
inconsistencies in naming conventions
- The code was inconsistent about where commands should be called by
their python module name (e.g. `log_parse`) and where the actual
command name should be used (e.g. `log-parse`).
- This made it hard to make a command with a dash in the name, and it
made `SpackCommand` fail to recognize commands with dashes.
- The code now uses the user-facing name with dashes for function
parameters, then converts that the module name when needed.
* Improve performance of log parsing
- A number of regular expressions from ctest_log_parser have really poor
performance, most due to untethered expressions with * or + (i.e., they
don't start with ^, so the repetition has to be checked for every
position in the string with Python's backtracking regex implementation)
- I can't verify that CTest's regexes work with an added ^, so I don't
really want to touch them. I tried adding this and found that it
caused some tests to break.
- Instead of using only "efficient" regular expressions, Added a
prefilter() class that allows the parser to quickly check a
precondition before evaluating any of the expensive regexes.
- Preconditions do things like check whether the string contains "error"
or "warning" (linear time things) before evaluating regexes that would
require them. It's sad that Python doesn't use Thompson string
matching (see https://swtch.com/~rsc/regexp/regexp1.html)
- Even with Python's slow implementation, this makes the parser ~200x
faster on the input we tried it on.
* Add `spack log-parse` command and improve the display of parsed logs
- Add better coloring and line wrapping to the log parse output. This
makes nasty build output look better with the line numbers.
- `spack log-parse` allows the log parsing logic used at the end of
builds to be executed on arbitrary files, which is handy even outside
of spack.
- Also provides a profile option -- we can profile arbitrary files and
show which regular expressions in the magic CTest parser take the most
time.
* Parallelize log parsing
- Log parsing now uses multiple threads for long logs
- Lines from logs are divided into chnks and farmed out to <ncpus>
- Add -j option to `spack log-parse`
* Marking database tests as slow
* Marking url command tests as slow
* Marking every test that uses database as slow
* Marking tests that import files as slow
* Marking gpg tests as slow
* Marking all versions and one list tests as slow
* Added more markers to unit tests + cli option to skip slow tests
Following a discussion with Axel, the generic 'slowtest' marker has been
split into 'db', 'network' and 'maybeslow'. A brief description of the
meaning of each marker has been added to pytest.ini.
A command line option to run only fast tests has been added to
'spack test'
* Don't use classes to group tests together
Reverted grouping tests under a class, as required in the review
* Minor style changes
This attempts to address one of the complaints at #5996 (comment):
> Repo currently caches package instances by Spec, and those Package instances have a Spec.
> This is unnecessary and causes confusion. I think I thought that we'd need to cache instances
> after loading package classes, but really just caching the classes is fine.
With this update, Repo's package cache is removed and Specs cache the package reference themselves. One consequence is that Specs which compare as equal will store separate instances of a Package class (not doing this creates issues for #4595 (comment)).
There were several references to Spec.package that could be replaced with Spec.package_class without any additional modifications. There are still a couple remaining references to Spec.package in Spec that would require adding functionality before replacing (e.g. calling Package.provides and Package.installed).
Note this makes it difficult to mock fetchers for tests which invokes code that reconstructs specs. test_packaging was one example of this where the updates caused a failure (in that case the error was avoided by not making an unnecessary call).
Details:
* Replace instances of spec.package with spec.package_class where a class method is being called
* Remove instances of Repo.get where Spec.package_class can be used in its place
* remove Repo.get caching instances of Package class for specs
* remove redundant check (which is also incorrect now that each spec stores its own copy of its package)
* avoid creating mirror with specs because it copies specs and those copies dont refer to the mocked fetcher (and it is also not required for the test)
* remove checks that are no longer necessary since repo doesn't cache specs
* Cleaned up JUnit report generation on install
The generation of a JUnit report was previously part of the install
command. This commit factors the logic into its own module, and uses
a template for the generation of the report.
It also improves report generation, that now can deal with multiple
specs installed at once. Finally, extending the list of supported
formats is much easier than before, as it entails just writing a
new template.
* Polished report generation + added tests for failures and errors
The generation of a JUnit report has been polished, so that the
stacktrace is correctly displayed with Jenkins JUnit plugin. Standard
error is still not used.
Added unit tests to cover for installation failures and installation
errors.
Avoid adding an "outside" (local home's) python user site directory
during python package installs.
Implements #6611
Fixes packages with auto-find side effects such as `py-setuptools`
that cause `py-matplotlib` to fail to build #6558
The flag_handlers method was being set as a bound method, but when
reset in the package.py file it was being set as an unbound method
(all python2 issues). This gets the underlying function information,
which is the same in either case.
The bug was uncovered for parmetis in #6858. This is a partial fix.
Included are changes to the parmetis package.py file to make use of
flag_handlers.
The feature added in #4611 is currently broken. This commit fixes the
behavior of the command and adds unit tests to ensure the basic semantic
is maintained.
It also changes slightly the behavior of Spec.concretized to avoid
copying caches before the concretization (as this may result in a
wrong hash computation for the DAG).
- Generating the HTML from for >2300 packages from RST in Sphinx seems to
take forever.
- Add an option to `spack list` to generate straight HTML instead.
- This reduces the doc build time to about a minute (from 5 minutes on a mac laptop).
* Vendor ordereddict for python2.6 only
This commit substitutes the custom module 'ordereddict_backport' with
the more known 'ordereddict' and vendors it only for python 2.6. Other
supported versions of python will use 'collections.OrderedDict'.
* Use absolute imports also for python 2.6
See PEP-328 for more information on the subject
* Added provenance of vendored ordereddict
See #6794
This fixes cases where test-only dependencies were omitted from
consideration when modifying the environment at build time. This
includes an update to the python package definition to add
testing-related python extensions to its specialized environment
setup.
This updates the conflict-checking logic to require that the conflict
spec matches exactly and that all fields mentioned in the conflict
spec are present in the concretized spec in order to report a
conflict. This will automatically skip all conflict checks for
dependencies of externals (since externals strip dependencies). This
will not affect non-external packages since all fields and
dependencies are fully specified for such packages.
* Keep track of source and versions for external libraries
* Note source of more obscure libraries
* We aren't upgrading jsonschema after all
* Add note on modifications made to pytest
