| 1 |
Ed W wrote: |
| 2 |
> Hi |
| 3 |
> |
| 4 |
> Thanks for the tips. |
| 5 |
> |
| 6 |
>> uclibc and hardened are current at odd with each other. Current uclibc |
| 7 |
>> demands a newer gcc than what is provided in tree. |
| 8 |
> |
| 9 |
> Could this be the problem also? |
| 10 |
> http://lists.uclibc.org/pipermail/uclibc/2008-November/041404.html |
| 11 |
|
| 12 |
Update (and for the benefit of anyone else): |
| 13 |
|
| 14 |
- It appears that gcc 3.4.6 doesn't properly compile uclibc 2.9.30 and |
| 15 |
in particular the -fomit-frame-pointer causes problems for anything |
| 16 |
linked to -lpthread |
| 17 |
|
| 18 |
There is a patch to uclibc here |
| 19 |
http://lists.uclibc.org/pipermail/uclibc/2008-November/041416.html |
| 20 |
|
| 21 |
I will create a bug report and file it (if any maintainer is watching |
| 22 |
here then perhaps you could just slap it in directly anyway?). Tested |
| 23 |
here on x86, compiled with gcc3.4.6 and all seems fine. |
| 24 |
|
| 25 |
The upgrade had minor problems breaking openssl, but curiously |
| 26 |
revdep-rebuild isn't effective in finding broken packages (I guess |
| 27 |
nothing rev-ed for it to spot it was broken). Rebuilding world right |
| 28 |
now which is flagging up loads more broken packages... |
| 29 |
|
| 30 |
I guess the road ahead is going to continue to be rocky with |
| 31 |
uclibc+gcc3.4... |
| 32 |
|
| 33 |
Ed W |
Archives