 espie
|
5afe4fffea
my bad, line at end of summary
|
7 years ago |
|
espie
|
b437d12dc7
white line at end of summary isn't very visible.
|
7 years ago |
|
espie
|
401c11c155
dpb cleans the permanent logs, always using the same pattern,
|
8 years ago |
|
espie
|
44f3329d90
there is no "equivalence list" if we didn't get any info...
|
8 years ago |
|
espie
|
cc2f4890f5
keep a global cache of distfile checksum while scanning, allows us to avoid
|
8 years ago |
|
espie
|
3388fc4cfd
change drastically the way LISTING starts up: let dependencies be a permanent
|
8 years ago |
|
espie
|
84584cc7af
add SKIPDIRS semantics: automatically add visited dirs to a passed skip
|
8 years ago |
|
espie
|
af4f89c46f
tweak the heuristics to run the engine:
|
8 years ago |
|
espie
|
9efe11300f
add a "summary" log that shows for each non built path, why it was non built.
|
9 years ago |
|
espie
|
2ee3c4a3b9
tweak things so that errors show up everywhere
|
9 years ago |
|
espie
|
33d8c81512
stub_out broken entries instead of deleting their info.
|
9 years ago |
|
espie
|
40499f8bbe
do detain distfiles as well as paths (need to track them over rescan)
|
9 years ago |
|
espie
|
5e55c60de7
handle distfile failure better.
|
9 years ago |
|
espie
|
828e1671f8
fix long-standing race condition where removing a lock could break things.
|
9 years ago |
|
espie
|
b150b6d2ec
rename logger interface to desambiguate vs user.
|
9 years ago |
|
espie
|
480dc42a5d
wrap die so that it drops into the debugger if we run under perl -d.
|
10 years ago |
|
espie
|
95850b1a81
finally tracked down a long-standing bug (hard to reproduce): when
|
10 years ago |
|
espie
|
8b3963e860
refactor the engine code, so that new_path can put some stuff directly on
|
10 years ago |
 afresh1
|
edaddea4f0
Work around perlbug 77706, only hit infrequently on vax.
|
10 years ago |
|
espie
|
cfb2f4c5fd
log when we actually run engine, since this is hard to match with perf log.
|
10 years ago |
|
espie
|
7205c851fa
note to myself, this makes deduplicating paths in the queue complicated.
|
10 years ago |
|
espie
|
9e7908fe2e
if we're rescanning, keep the locks around.
|
11 years ago |
|
espie
|
fb87501764
lock paths during error rescan.
|
11 years ago |
|
espie
|
860742c336
isolate dirty details of fullpkgname remanence in its own function,
|
11 years ago |
|
espie
|
43babaae79
when we rescan paths, we temporarily delete info. Those paths are
|
11 years ago |
|
espie
|
6152760dbd
allow the engine to rebuild_info
|
11 years ago |
|
espie
|
c7a350aabb
add the fullpkgname as a comment at the end of the I: lines.
|
11 years ago |
|
espie
|
8877767529
- use separated Heuristics::Size object
|
11 years ago |
|
espie
|
14c5104e65
have start return useful information, like whether it started something
|
11 years ago |
|
espie
|
a0a1e6b657
okay, fixed version where NoBuild subengine gets only what it actually
|
11 years ago |
