| Commit message (Collapse) | Author | Age | Files | Lines |
|
|
|
| |
This happens if depstring evaluates to nothing.
|
|
|
|
| |
Broken since commit b4f98ec5a1eb2e010a455b19db5bef8d61def241.
|
| |
|
|
|
|
|
| |
This restores some desired logic that was left out of commit
5655b4dcfe5e9dae5e9d6352d791c3d04953baf7.
|
|
|
|
|
| |
Since commit 878f166c5a06b0ab6c5a34dd658c878cc7605fe7, empty implies
deep = True, so these checks are redundant.
|
| |
|
|
|
|
|
|
| |
Always prefer parents for which the parent atom is unsatisfied by the
child node. This seems to be the intention of the code, though it may
not have behaved correctly due to continuation of the parent loop.
|
|
|
|
|
| |
There is an existing handler for EnvironmentError, so raise that
instead of FileNotFound.
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
If the user specifies FETCHCOMMAND_*, Portage should honor this when grabbing
Packages files. This allows users to setup support for grabbing Packages files
from other protocols.
BUG=chrome-os-partner:2026
TEST=Try downloading prebuilts from gs:// when FETCHCOMMAND_GS is setup in make.conf
Change-Id: I96b239819351633dd02d608954e81a1c363a4687
Review URL: http://codereview.chromium.org/6458015
|
|
|
|
| |
This will fix bug #354747.
|
|
|
|
|
|
| |
Now the masking display is disabled by the --quiet option if the
--verbose option is not enabled simultaneously. Also, docs are
updated as per bug 58416, comment #11 and #12.
|
| |
|
|
|
|
| |
This will fix bug #354687.
|
| |
|
| |
|
| |
|
|
|
|
|
|
| |
Recursively traversed virtual dependencies, and their direct
dependencies, are considered to have the same depth as direct
dependencies.
|
|
|
|
|
| |
Control over recursion depth hasn't behaved properly since commit
6503980e0e3bcfce9fbaff85c33d87f616e955a9. Now it is fixed and tested.
|
|
|
|
|
|
|
| |
This won't change the existing depgraph behavior, but it will allow
logic to be simplified in some cases such that solely the deep param
needs to be consulted (rather than needing to consult both empty and
deep params).
|
|
|
|
| |
This will fix bug #354387.
|
|
|
|
|
|
| |
Errors returned from --update-use-local-desc and --update-changelogs
are no longer ignored with --tolerant. This will fix part of
bug #354687.
|
|
|
|
| |
This will fix bug #354685.
|
|
|
|
|
|
|
|
| |
Atoms are stored in the graph as (atom, id(atom)) tuples since each
atom is considered to be a unique entity. For example, atoms that
appear identical may behave differently in USE matching, depending on
their unevaluated form. Also, specially generated virtual atoms may
appear identical while having different _orig_atom attributes.
|
| |
|
|
|
|
| |
This will fix bug #354437.
|
|
|
|
|
|
|
|
| |
In _add_pkg_dep_string(), _add_pkg() was called with generated
Dependency instances that created fake/distorted edges in the
graph for new-style virtuals (similar to those that were fixed
in commit 5b16e0d800f34dc90cdd8c505172ad16b97f0af4). Now there
won't be any such distorted edges.
|
|
|
|
|
|
|
|
| |
As mentioned in bug #353933, comment #4, special new-style virtual
atoms generated/distorted by _expand_new_virtuals() can prevent
extract_affecting_use() from working properly. This is fixed by
saving the original atoms so that the depgraph can map them back
into place at the appropriate step in dependency evaluation.
|
|
|
|
| |
This will fix bug #354441.
|
|
|
|
| |
This will fix bug #354137.
|
|
|
|
| |
This will fix bug #354141.
|
|
|
|
| |
AttributeError: 'tuple' object has no attribute 'use'
|
|
|
|
|
|
| |
We tolerate global scope import failures for optional modules, so that
ImportModulesTestCase can succeed (or possibly alert us about
unexpected import failures).
|
| |
|
|
|
|
| |
The repoman.herdbase module requires python with xml support.
|
|
|
|
| |
ENODATA is undefined on FreeBSD. This will fix bug #354057.
|
| |
|
| |
|
| |
|
|
|
|
| |
This will fix bug #354003.
|
| |
|
|
|
|
|
|
| |
We need to evaluate conditionals for dependency matching purposes,
though the unevaluated atom will still be available later for other
purposes.
|
|
|
|
|
|
|
|
| |
This leads to a better unsatisfied USE message than the one reported
in bug 353933, comment #3. TODO: Fix it so that these generated
atoms don't break depgraph._get_dep_chain(), due to not being
identical to the original atoms and therefore breaking
extract_affecting_use().
|
|
|
|
| |
This will fix bug #353836.
|
|
|
|
| |
This will fix bug #353897.
|
|
|
|
|
|
|
| |
It's essential to feed back autounmask config changes, especially
USE changes, in order to make the last run as successful as possible.
This fixes a regression in the master branch, since commit
2186f87c22e0c4bdb86065855c2034bdbb66fa45.
|
| |
|
|
|
|
|
|
| |
This reverts commit 52f6b55d51e2accaa446bc5d6920421959289baa.
Since split() never return empty tokens, this patch does nothing.
|
| |
|
| |
|
| |
|