1 |
On 6/25/2011 8:04 AM, Dale wrote: |
2 |
|
3 |
>> We restructured the dependency chain for fortran support, |
4 |
>> which includes |
5 |
>> a compile test now. The failure can be seen above. |
6 |
>> |
7 |
>> The Problem was in short, USE=fortran was enabled by |
8 |
>> default for linux |
9 |
>> arches, but people tend to disable it. Depending on |
10 |
>> gcc[fortran] doesn't |
11 |
>> work completely as gcc:4.4[fortran] and gcc:4.5[-fortran] |
12 |
>> with gcc-4.5 |
13 |
>> select can be installed, which would full fill the |
14 |
>> dependency but |
15 |
>> nevertheless doesn't give a working compiler. |
16 |
>> |
17 |
>> So now packages depend on virtual/fortran and use an |
18 |
>> eclass to check for |
19 |
>> a working compiler. So if you see this message, this means |
20 |
>> you somehow |
21 |
>> worked around gcc[fortran]. |
22 |
|
23 |
> That make sense? |
24 |
|
25 |
Yes. He's saying they didn't change the USE flag, they |
26 |
changed the fortran dependency test to actually do a |
27 |
run-time check for fortran because the USE flag alone wasn't |
28 |
sufficient. |
29 |
|
30 |
Which means you most likely had a non-working cantor and no |
31 |
fortran compiler before and just didn't notice :) |
32 |
|
33 |
--Mike |