1 |
On Thu, 2009-12-17 at 05:21 -0500, Mike Frysinger wrote: |
2 |
> On Thursday 17 December 2009 04:38:36 Markus Duft wrote: |
3 |
> > I recently did some ia64-hpux hacking in prefix, and i stumbled over |
4 |
> > alittle problem: |
5 |
> > |
6 |
> > the native linker collects -L options on the command line, and python |
7 |
> > uses -L. while building. this results in . beeing in rpath. |
8 |
> > |
9 |
> > now when $PWD == $S, there is libpython-*.so, and executing the already |
10 |
> > installed python finds (through rpath ".") this local libpython. this |
11 |
> > (although binary exactly the same) refuses to load installed extensions |
12 |
> > (Fatal Python Error: ....). |
13 |
> > |
14 |
> > The problem was caused by the filter-bash-environment.py call in |
15 |
> > ebuild.sh, _only_ during the "test" phase seemingly. |
16 |
> > |
17 |
> > the attached patch fixes the problem for me - what do you think? I guess |
18 |
> > there are better ways to tackle the problem, but i needed to get this to |
19 |
> > work ;) |
20 |
> |
21 |
> this ignores the real problem. there should never be an installed binary |
22 |
> whose rpath includes $PWD. this is a terrible security issues waiting to bite |
23 |
> you in the ass, or cause random failures depending on what the user's cwd is |
24 |
> at any point in time. |
25 |
> |
26 |
> the second change (adding the || exit) is a good idea in theory, but the |
27 |
> proposed change is incorrect. use the assert helper so that the exit values |
28 |
> of all commands in the pipeline are checked. |
29 |
> -mike |
30 |
|
31 |
Yep to everything Mike said.. |