1 |
On Fri, 08 Mar 2013 15:44:16 +0100 |
2 |
Thomas Sachau <tommy@g.o> wrote: |
3 |
|
4 |
> Alexis Ballier schrieb: |
5 |
> > On Thu, 07 Mar 2013 19:59:35 +0100 |
6 |
> > Thomas Sachau <tommy@g.o> wrote: |
7 |
> >> |
8 |
> >> I dont have a list of binaries, i either noticed myself some |
9 |
> >> abi-specific behaviour or got user reports for abi-specific |
10 |
> >> behaviour. As an example i remember, dev-libs/libIDL has a config |
11 |
> >> binary not matching the usual *-config scheme (libIDL-config-2), |
12 |
> >> so instead of adding a random list of patterns, i simply added |
13 |
> >> that package to the list of packages with wrapped binaries. The |
14 |
> >> same applies to mysql, which has a mysql_config binary. |
15 |
> > |
16 |
> > Ok, so those make perfect sense for being wrapped and should be done |
17 |
> > per-binary not per-package: We do not really want to wrap all mysql |
18 |
> > binaries just for mysql_config. |
19 |
> > Ebuilds should declare the binaries they want to be wrapped, so we |
20 |
> > can grep the tree for getting a list of what to fight against if we |
21 |
> > want a cleaner multilib system :) |
22 |
> |
23 |
> This is possible, but will make the solution more complex or limited |
24 |
> (either the solution does not allow users to opt-in for wrapping all |
25 |
> binaries of a package or you need an additional option for users to |
26 |
> opt-in beside the variable, which includes the needed binaries). |
27 |
|
28 |
Why do you want _users_ to do that? Ebuilds should do it. We are |
29 |
talking about multi_lib_ not multi_arch_ after all :) |
30 |
If you want to wrap all binaries, it may be allowable to have a regexp |
31 |
there so that '*/bin/*' would wrap them all, but I don't think it is a |
32 |
good idea to do that. |
33 |
|
34 |
> >> I am not sure about the target of your qmake question, so as a |
35 |
> >> general answer: |
36 |
> >> |
37 |
> >> qmake is something like configure for qmake based build systems. If |
38 |
> >> you want to see the difference between qmake (32bit) and qmake |
39 |
> >> (64bit), run a 64bit qmake on a qmake based package and do the same |
40 |
> >> with a 32bit qmake and check the difference between the 2 runs. |
41 |
> > |
42 |
> > Well, I'm asking this because I don't have access to a 32bit qmake |
43 |
> > here so a diff of the files generated by the two will be useful :P |
44 |
> > And I believe it makes sense to study in details this case to |
45 |
> > understand whether we want to wrap it or not. As Davide said, it is |
46 |
> > likely that overriding the correct variable may make qmake output |
47 |
> > not abi-specific. The difference between g++-64 and g++-32 qmake |
48 |
> > mkspec is only a -m32 vs -m64 cflag which I think is overridden |
49 |
> > elsewhere in our ebuilds so this should not matter much. |
50 |
> |
51 |
> If you want an example diff, can you point me to a small qmake based |
52 |
> package? |
53 |
|
54 |
media-video/smplayer or media-sound/qjackctl but this one also uses |
55 |
autoconf before |
56 |
|
57 |
|
58 |
Alexis. |