| 1 |
This week didn't see much movement in the repository due to life |
| 2 |
happening (moving and lack of internet at new apt). On the plus side, |
| 3 |
I believe the majority of the work is complete, and I urge everyone to |
| 4 |
give it a test to help me discover problems. :) |
| 5 |
|
| 6 |
I'm happy with the overall results so far. One thing that irks me is |
| 7 |
the dependence on directory and library masks found in |
| 8 |
/etc/revdep-rebuild/*. I was really hoping to find a more elegant |
| 9 |
solution than parsing masks from config files, but all my attempts at |
| 10 |
a different solution were not successful. Unfortunately, as long as |
| 11 |
masks are used, there is the potential of false positives for |
| 12 |
libraries that are not properly masked, which as far as I know is the |
| 13 |
only major issue. |
| 14 |
|
| 15 |
My work over the last week was largely polishing details in |
| 16 |
docstrings/comments and contemplating final modifications (or future |
| 17 |
endeavors). During this last week of gsoc, I'm going to (hopefully) |
| 18 |
improve the performance of LinkageMap and MissingLibraryConsumerSet |
| 19 |
before the deadline. It was suggested on the gentoo-portage-dev |
| 20 |
mailing list [1] that efficiency in path comparisons in LinkageMap |
| 21 |
could be improved by comparing inodes rather than os.path.realpath in |
| 22 |
python's os.path module, which is what I started to implement |
| 23 |
recently; however, this appears to require changes outside of the |
| 24 |
class, which will take time to finalize e.g. during the unmerge |
| 25 |
process, LinkageMap.rebuild is called after files have been removed |
| 26 |
but before /var/db/pkg has been correspondingly updated, which is |
| 27 |
problematic since. I may not get all the kinks worked out of changing |
| 28 |
LinkageMap before the gsoc deadline, so I'll have something to work on |
| 29 |
afterwards. :) |
| 30 |
|
| 31 |
The attached patches add the library dependency rebuilder as a set, |
| 32 |
@missing-rebuild, to portage-2.2_rc6. Similar to the --library flag |
| 33 |
in revdep-rebuild, the user can additionally emerge the set of |
| 34 |
packages containing consumers of libraries matched by a (python) |
| 35 |
regular expression; however, until a better solution is found, the |
| 36 |
regexp must be passed through the LIBRARY environment variable to |
| 37 |
enable that feature. |
| 38 |
|
| 39 |
Attached are patches for |
| 40 |
/usr/lib/portage/pym/portage/dbapi/vartree.py, |
| 41 |
/usr/lib/portage/pym/portage/sets/libs.py and |
| 42 |
/usr/share/portage/config/sets.conf, which can also be found in the |
| 43 |
repository.[2] After applying the three patches, one can test the set |
| 44 |
with emerge -p @missing-rebuild. If anyone is brave enough to do so, |
| 45 |
let me know how it goes! |
| 46 |
|
| 47 |
Lucian |
| 48 |
|
| 49 |
|
| 50 |
[1] http://archives.gentoo.org/gentoo-portage-dev/msg_abcbac026ed8670f6dc61bb28be6151e.xml |
| 51 |
[2] http://repo.or.cz/w/revdep-rebuild-reimplementation.git?a=tree;h=refs/heads/rc1;hb=rc1 |
Archives