1 |
Mauro Maroni <mmaroni@××××××.ar> posted |
2 |
200702142151.54947.mmaroni@××××××.ar, excerpted below, on Wed, 14 Feb |
3 |
2007 21:51:54 -0300: |
4 |
|
5 |
> While doing a general upgrade yesterday, I noticed that mldonkey fails |
6 |
> almost at the end (see below) and emerge ends. I tried again today but got |
7 |
> the same result. I did not find any bug related to this. Did anyone know |
8 |
> where is the problem? |
9 |
> |
10 |
|
11 |
>>>> Merging net-p2p/mldonkey-2.8.2-r1 to / |
12 |
> --- /usr/ |
13 |
> --- /usr/bin/ |
14 |
>>>> /usr/bin/mlnet |
15 |
>>>> /usr/bin/mld_hash |
16 |
>>>> /usr/bin/get_range |
17 |
>>>> /usr/bin/copysources |
18 |
>>>> /usr/bin/make_torrent |
19 |
>>>> /usr/bin/subconv |
20 |
>>>> /usr/bin/mldonkey |
21 |
> --- /etc/ |
22 |
> --- /etc/conf.d/ |
23 |
> Traceback (most recent call last): |
24 |
> File "/usr/bin/emerge", line 5299, in ? |
25 |
> retval = emerge_main() |
26 |
> File "/usr/bin/emerge", line 5294, in emerge_main |
27 |
> myopts, myaction, myfiles, spinner) |
28 |
> File "/usr/bin/emerge", line 4761, in action_build |
29 |
> retval = mergetask.merge(pkglist, favorites, mtimedb) |
30 |
> File "/usr/bin/emerge", line 2999, in merge |
31 |
> prev_mtimes=ldpath_mtimes) |
32 |
> File "/usr/lib/portage/pym/portage.py", line 3699, in doebuild |
33 |
> vartree=vartree, prev_mtimes=prev_mtimes) |
34 |
> File "/usr/lib/portage/pym/portage.py", line 3907, in merge |
35 |
> mydbapi=mydbapi, prev_mtimes=prev_mtimes) |
36 |
> File "/usr/lib/portage/pym/portage.py", line 7768, in merge |
37 |
> cleanup=cleanup, mydbapi=mydbapi, prev_mtimes=prev_mtimes) |
38 |
> File "/usr/lib/portage/pym/portage.py", line 7364, in treewalk |
39 |
> if |
40 |
> self.mergeme(srcroot,destroot,outfile,secondhand,"",cfgfiledict,mymtime): |
41 |
> File "/usr/lib/portage/pym/portage.py", line 7640, in mergeme |
42 |
> join(offset, x), cfgfiledict, thismtime): |
43 |
> File "/usr/lib/portage/pym/portage.py", line 7640, in mergeme |
44 |
> join(offset, x), cfgfiledict, thismtime): |
45 |
> File "/usr/lib/portage/pym/portage.py", line 7661, in mergeme |
46 |
> if self.isprotected(mydest): |
47 |
> File "/usr/lib/portage/pym/portage.py", line 7116, in isprotected |
48 |
> mydata = self._installed_instance.getcontents().get(filename, None) |
49 |
> AttributeError: 'NoneType' object has no attribute 'get' amd64 ~ # |
50 |
> amd64 ~ # |
51 |
|
52 |
I track portage updates and watch what bugs they are fixing. As a result, I |
53 |
see a lot of portage bugs, and this would appear to be one -- at least |
54 |
indirectly. Even if a file is corrupt or something, portage should quit |
55 |
with an intelligent error. That it crashes with a traceback therefore |
56 |
indicates either a direct portage bug, or at minimum (and perhaps more |
57 |
likely), a bug in that a file or something was corrupt that it should have |
58 |
been prepared to handle but wasn't. |
59 |
|
60 |
I'd suggest filing it as a portage bug (of course, make sure your portage is |
61 |
up to date and try it again before you file, if it's not, in case the bug |
62 |
has already been fixed) and letting the experts deal with it. They are |
63 |
likely to tell you something either in your currently merged version or in |
64 |
portage's record of it is corrupt, but will be able to point you to what it |
65 |
is, and in the process likely fix the traceback such that portage quits with |
66 |
a coherent error that at least lets the user know what was wrong so they |
67 |
know where to start looking to fix it. |
68 |
|
69 |
You may be able to fix the issue yourself by unmerging whatever version you |
70 |
have currently merged, or possibly deleting the record of it in /var/db/ |
71 |
portage, then remerging the old version to give portage a correct record so |
72 |
it can then properly unmerge it when you upgrade. However, that will likely |
73 |
destroy the evidence that the portage folks need to trace and fix their bug, |
74 |
and there's a tiny chance it could make things worse. It'd be much better |
75 |
to file the bug and let the portage experts have at it. They should be able |
76 |
to tell you what's happening and how to fix it, tho from the bugs I've seen, |
77 |
they do sometimes have you run an intermediate patch that doesn't fix it, |
78 |
but helps them to better pin down the problem. |
79 |
|
80 |
-- |
81 |
Duncan - List replies preferred. No HTML msgs. |
82 |
"Every nonfree program has a lord, a master -- |
83 |
and if you use the program, he is your master." Richard Stallman |
84 |
|
85 |
-- |
86 |
gentoo-amd64@g.o mailing list |