| Age | Commit message (Collapse) | Author | Files | Lines |
|---|
|
Allow flags from different sources (compilers, `require:`, command-line
specs, and `depends_on`) to be merged together, and enforce a consistent
order among them.
The order is based on the sources, e.g. flags on specs from the command
line always come last. Some flag order consistency issues are fixed:
1. Flags from `compilers.yaml` and the command line were always intra- and
inter-source order consistent.
2. Flags from dependents and packages.yaml (introduced via `require:`)
were not: for `-a -b` from one source and `-c` from another, the final
result might rearrange `-a -b`, and would also be inconsistent in terms
of whether `-c` came before or after.
(1) is/was handled by going back to the original source, i.e., flags are
retrieved directly from the command line spec rather than the solver.
(2) is addressed by:
* Keeping track of grouped flags in the solver
* Keeping track of flag sources in the solver on a per-flag basis
The latter info is used in this PR to enforce DAG ordering on flags
applied from multiple dependents to the same package, e.g., for this
graph:
```
a
/|\
b | c
\|/
d
```
If `a`, `b`, and `c` impose flags on `d`, the combined flags on `d` will
contain the flags of `a`, `b`, and `c` -- in that order.
Conflicting flags are allowed (e.g. -O2 and -O3). `Spec.satisifes()` has
been updated such that X satisfies Y as long as X has *at least* all of
the flags that Y has. This is also true in the solver constraints.
`.satisfies` does not account for how order can change behavior (so
`-O2 -O3` can satisfy `-O3 -O2`); it is expected that this can be
addressed later (e.g. by prohibiting flag conflicts).
`Spec.constrain` and `.intersects` have been updated to be consistent
with this new definition of `.satisfies`.
|
|
Spack can now bootstrap two new dependencies on Windows: GnuPG, and file.
These dependencies are modeled as a separate package, and they install a cross-compiled binary.
Details on how they binaries are built are in https://github.com/spack/windows-bootstrap-resources
|
|
|
|
|
|
* Allow deprecating more than one property in config
This internal change allows the customization of errors
and warnings to be printed when deprecating a property.
Signed-off-by: Massimiliano Culpo <massimiliano.culpo@gmail.com>
* fix
Signed-off-by: Massimiliano Culpo <massimiliano.culpo@gmail.com>
* Use a list comprehension for "issues"
Signed-off-by: Massimiliano Culpo <massimiliano.culpo@gmail.com>
---------
Signed-off-by: Massimiliano Culpo <massimiliano.culpo@gmail.com>
|
|
* Windows: Update pytest with echo and remove others
* Fix style
* Remove unused pytest import in undevelop
* Update test_test_part_pass to be Window compatible
* Style
|
|
fixes #40791
Currently stacks behave differently if used in unify:false
environments, which leads to inconsistencies during concretization.
For instance, we might have two abstract user specs that do not
intersect with each other map to the same concrete spec in the
environment. This is clearly wrong.
This PR removes the best effort expansion, so that user specs
are always applied strictly.
|
|
fixes two tests that did not clear the in-memory bits of a database
before calling reindex.
|
|
Ensures that Database instances do not reference a global
`spack.store.STORE.layout`. Simplify Database.{add,reindex} signature.
|
|
This test was possibly meant for the Cray platform, and
currently is a no-op.
Signed-off-by: Massimiliano Culpo <massimiliano.culpo@gmail.com>
|
|
|
|
Signed-off-by: Massimiliano Culpo <massimiliano.culpo@gmail.com>
|
|
|
|
|
|
|
|
|
|
|
|
Update build system conflict between CUDA 12.6 and Clang 18
|
|
Signed-off-by: Georgia Stuart <gstuart@umass.edu>
Co-authored-by: Jordan Galby <67924449+Jordan474@users.noreply.github.com>
|
|
|
|
This PR simplifies the code doing external spec detection by removing
the `DetectedPackage` class. Now, functions accepting or returning lists
of `DetectedPackage`, will accept or return list of specs.
Performance doesn't seem to change if we use `Spec.__reduce__` instead
of `DetectionPackage.__reduce__`.
|
|
so linking to shared libraries works well at runtime on darwin for all packages with cmake_minimum_required < 3.
|
|
|
|
|
|
This function is used only in tests.
|
|
Fix regression introduced in spack 0.22.1 where Spack would ask about checksums
on spack-develop packages.
|
|
(#45522)
|
|
|
|
inspect.isclass -> isinstance(..., type)
inspect.getmro -> cls.__mro__
|
|
Change the signature of Spec.from_detection to set the
external prefix, and the external modules, if they are
present.
Delete "spack.package_prefs.spec_externals" since it
is unused.
|
|
|
|
|
|
This context manager was used to prevent detecting a platform
as "cray". Since now Cray machines are detected as linux, we can
remove the context manager.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Right now if you run `spack ci generate` you get a warning about CDash credentials even
if there's no CDash configuration specified. We should only warn if there was actually a
CDash config.
Signed-off-by: Todd Gamblin <tgamblin@llnl.gov>
|
|
|
|
|
|
|
|
Source mirrors store entries by digest and add a human readable alias of the
form 'name-version'. If no digest is available, the alias is used as the primary
storage location.
Spack erroneously fetches by alias when the digest path does not exist. This is
problematic if `version(..., sha256=...)` changes in package.py, and the mirror
is populated with the old shasum. That would result in an error when a digest
is available, but in case of git versions with a modified commit sha, the wrong
sources would be fetched without error. With this PR, only the digest path is
used, not the alias, in case a digest is available. This is also a small performance
optimization, as the number of request is halved for mirrors that don't contain
the sources.
Further, for git sources the tag was used as a digest, but this is a moving
target. Only commit sha is used now.
Also whenever the alias already existed, Spack used to keep it in place when
updating the mirror cache, which means that aliases would always point to
outdated mirror entries whenever digests are modified. With this PR the alias
is moved in place.
Lastly, fix a recent regression where `Stage.disable_mirrors` disabled mirrors
but not the local download cache, which was the intention.
|
|
|
|
|
|
* Add missing MultiMethodMeta metaclass in builders
and remove the Python 2 fallback option in favor of hard errors to catch
similar issues going forward.
The fallback option can cause about 10K stat calls due to use of
`realpath` in the inspect module, depending on how deep Spack itself is
nested in the file system, which is ... undesirable.
* code shuffling to avoid circular import
* more reshuffling
* move reserved variant names into variants module
|
|
* Add a new audit to find missing package.py files
* Remove directory without package.py
|
|
|
