1 |
Davide Pesavento schrieb: |
2 |
> On Thu, Mar 7, 2013 at 10:59 AM, Thomas Sachau <tommy@g.o> wrote: |
3 |
>> Alexis Ballier schrieb: |
4 |
>>> On Mon, 04 Mar 2013 21:17:50 +0100 |
5 |
>>> Thomas Sachau <tommy@g.o> wrote: |
6 |
>>>> dev-db/mysql abiwrapper |
7 |
>>>> dev-lang/perl abiwrapper |
8 |
>>>> dev-lang/python abiwrapper |
9 |
>>>> dev-lang/ruby abiwrapper |
10 |
>>>> dev-libs/gobject-introspection abiwrapper |
11 |
>>>> dev-libs/libIDL abiwrapper |
12 |
>>>> dev-scheme/guile abiwrapper |
13 |
>>>> net-libs/courier-authlib abiwrapper |
14 |
>>>> dev-qt/qtcore abiwrapper |
15 |
>>>> dev-qt/qtgui abiwrapper |
16 |
>>>> media-libs/fontconfig abiwrapper |
17 |
>>>> www-servers/apache abiwrapper |
18 |
>>>> x11-libs/pango abiwrapper |
19 |
>>>> x11-libs/gtk+ abiwrapper |
20 |
>>> |
21 |
>>> |
22 |
>>> Since you are doing this per-package and not per-binary: Do you happen |
23 |
>>> to remember/have a list of the binaries that have abi-specific output? |
24 |
>>> |
25 |
>>> Also, for the qmake example, do you have an example of difference |
26 |
>>> between qmake32 and qmake64? |
27 |
>>> |
28 |
>>> Alexis. |
29 |
>>> |
30 |
>> |
31 |
>> I dont have a list of binaries, i either noticed myself some |
32 |
>> abi-specific behaviour or got user reports for abi-specific behaviour. |
33 |
>> As an example i remember, dev-libs/libIDL has a config binary not |
34 |
>> matching the usual *-config scheme (libIDL-config-2), so instead of |
35 |
>> adding a random list of patterns, i simply added that package to the |
36 |
>> list of packages with wrapped binaries. The same applies to mysql, which |
37 |
>> has a mysql_config binary. |
38 |
>> |
39 |
>> I am not sure about the target of your qmake question, so as a general |
40 |
>> answer: |
41 |
>> |
42 |
>> qmake is something like configure for qmake based build systems. If you |
43 |
>> want to see the difference between qmake (32bit) and qmake (64bit), run |
44 |
>> a 64bit qmake on a qmake based package and do the same with a 32bit |
45 |
>> qmake and check the difference between the 2 runs. |
46 |
>> |
47 |
> |
48 |
> Isn't this solvable by passing the right mkspec to a 64-bit qmake? |
49 |
> |
50 |
> Thanks, |
51 |
> Pesa |
52 |
> |
53 |
> |
54 |
|
55 |
Bug 304971 - dev-qt/qtcore stores machine-specific information in |
56 |
/usr/share/qt4/mkspecs |
57 |
|
58 |
in short: since the install location for the mkspec files is not |
59 |
abi-specific, you only have either 32bit or 64bit ones around, but not both. |
60 |
|
61 |
So if you get upstream to place those files into an abi-specific dir, |
62 |
get qmake to use the mkspec files from the right dir for the currently |
63 |
active ABI and got all older versions of qmake removed from user |
64 |
systems, then we might not need to anymore wrap qmake. |
65 |
|
66 |
-- |
67 |
|
68 |
Thomas Sachau |
69 |
Gentoo Linux Developer |