1 |
On 22-04-2010 20:55:58 -0500, Johan Hattne wrote: |
2 |
> As far as I can see, setting the rpath is one way to solve the problem, |
3 |
> and I've been toying around with patchelf to accomplish that (chrpath |
4 |
> won't work, because it doesn't allow rpaths to grow). To my mind, the |
5 |
|
6 |
I'd like to add it to the tree when I've got time to make an ebuild for |
7 |
it. |
8 |
|
9 |
> process is conceptually similar to the "shebang prefixing" that happens |
10 |
> for shell scripts, so I'm wondering whether it would make sense to |
11 |
> implement it in a similar fashion, e.g. |
12 |
> |
13 |
> for each ELF file { |
14 |
> prepend ${EPREFIX} to each rpath element |
15 |
> prepend ${EPREFIX}/lib:${EPREFIX}/usr/lib to rpath |
16 |
> } |
17 |
> |
18 |
> somewhere in portage. At least this allowed me to use icc to compile a |
19 |
> 32 bit helloworld using an x86-prefix on both x86 and amd64. |
20 |
|
21 |
Do you have an idea how to automagically select the correct ELF files? |
22 |
Most already contain the required paths of course. I don't quickly see |
23 |
how we can pick the right ones out there to "fix", but perhaps that's |
24 |
possible. Otherwise, maybe a function in prefix.eclass that does the |
25 |
job somehow? |
26 |
|
27 |
|
28 |
-- |
29 |
Fabian Groffen |
30 |
Gentoo on a different level |