1 |
On Mon, 2009-03-16 at 19:45 -0400, Mike Frysinger wrote: |
2 |
> On Monday 16 March 2009 18:49:04 Ned Ludd wrote: |
3 |
> > On Mon, 2009-03-16 at 17:05 -0400, Mike Frysinger wrote: |
4 |
> > > On Monday 16 March 2009 14:35:15 Ned Ludd wrote: |
5 |
> > > > On Mon, 2009-03-16 at 18:34 +0200, Amit Dor-Shifer wrote: |
6 |
> > > > > Hi all. |
7 |
> > > > > |
8 |
> > > > > While working on my overlay, I stumbled on an issue where qfile |
9 |
> > > > > refused to acknowledge an installed file as being part of my package. |
10 |
> > > > > |
11 |
> > > > > Looking into q's implementation (portage-utils-0.1.29), I see: |
12 |
> > > > > |
13 |
> > > > > amit0 portage-utils-0.1.29 # grep -A 2 next_entry |
14 |
> > > > > ./libq/vdb_get_next_dir.c next_entry: |
15 |
> > > > > ret = readdir(dir); |
16 |
> > > > > if (ret == NULL) { |
17 |
> > > > > -- |
18 |
> > > > > goto next_entry; |
19 |
> > > > > if (strchr(ret->d_name, '-') == NULL) |
20 |
> > > > > if ((strcmp(ret->d_name, "virtual")) != 0) |
21 |
> > > > > goto next_entry; |
22 |
> > > > > |
23 |
> > > > > I encountered this since I used a new category, which only contained |
24 |
> > > > > a single word. Adding a hyphen and a 2nd token solved my issue, and |
25 |
> > > > > now qfile knows the file's association. |
26 |
> > > > > |
27 |
> > > > > Is this assumption, that category should be "stringA-stringB" |
28 |
> > > > > documented somewhere? |
29 |
> > > > |
30 |
> > > > We made that assumption for portage-utils as they can be used on a |
31 |
> > > > device which has no $PORTDIR at all. So when there is no categories |
32 |
> > > > file that exists we fell back to the rules that have been working well |
33 |
> > > > for the past %d years. |
34 |
> > > > |
35 |
> > > > We changed that behavior however a while ago. I thought this was in the |
36 |
> > > > tree. But I guess not if you are hitting it. |
37 |
> > > > |
38 |
> > > > http://sources.gentoo.org/viewcvs.py/gentoo-projects/portage-utils/libq |
39 |
> > > >/vdb _get_next_dir.c?r1=1.2&r2=1.3 |
40 |
> > > |
41 |
> > > we should do a new release already |
42 |
> > |
43 |
> > Why yes.. Yes you should :) |
44 |
> |
45 |
> if you dont do it before me, i'll probably try and do it this weekend. |
46 |
|
47 |
|
48 |
|
49 |
I'd prefer it if you could do it this time. (thanks in advance) |
50 |
|
51 |
> btw, i |
52 |
> went through the bug reports and saw qcache crashes ... are those still |
53 |
> relevant ? |
54 |
> -mike |
55 |
|
56 |
Yeah. tcort was the guy who wrote most of that. He's retired now. |
57 |
I never really looked into it much but I think there are some NULL |
58 |
values he did not check for in the metacache. |
59 |
|
60 |
There is also a bug with atom parsing iirc on 32bit platforms. gradm was |
61 |
the test case. Think we need to change from int to long.. Maybe another |
62 |
with -rX parsing. |
63 |
|
64 |
-- |
65 |
Ned Ludd <solar@g.o> |
66 |
Gentoo Linux |