View previous topic :: View next topic |
Author |
Message |
CaptainBlood Advocate
Joined: 24 Jan 2010 Posts: 3595
|
Posted: Fri Mar 11, 2011 10:33 pm Post subject: mpfr problem after update [Solved] |
|
|
Hi dear all,
I've just updated mpfr to stable 3.0.0. Thet I did a Code: | revdep-rebuild --library '/usr/lib64/libmpfr.so.1' | which didn't rebuild anything, then I did Code: | rm '/usr/lib64/libmpfr.so.1' | ..Afterwards I had to recompile a kernel, which suprisingly failed. So I decided to before trying to rebuild the previous package.but it also faled.
Now I can't compile my kernel nor pmfr which is now uninstalled from my system.I have a strange feeling my system in now in a terrible situation. Is there anything I should do to recover this situation?
Thanks for your attention.
Last edited by CaptainBlood on Fri Mar 11, 2011 10:57 pm; edited 1 time in total |
|
Back to top |
|
|
Caligatio n00b
Joined: 13 Feb 2010 Posts: 48
|
Posted: Fri Mar 11, 2011 10:40 pm Post subject: |
|
|
I ran into the exact same problem and found that symlinking the new /usr/lib64/libmpfr.so as /usr/lib64/libmpfr.so.1 allowed GCC to compile (I realize it's ugly by it worked) |
|
Back to top |
|
|
CaptainBlood Advocate
Joined: 24 Jan 2010 Posts: 3595
|
Posted: Fri Mar 11, 2011 10:54 pm Post subject: |
|
|
Thanks for your quick anwser.
I must confess I didn't realized there is a Post-It to recover from such a situation.
Seems like the ebuild is not stable yet. I haven't decided whether once recovered I will upgrade or not.
Thanks for your attention, interest and support |
|
Back to top |
|
|
CaptainBlood Advocate
Joined: 24 Jan 2010 Posts: 3595
|
Posted: Tue Mar 22, 2011 7:04 am Post subject: |
|
|
Hi again;
The mpfr package has been remasked after its release to stable.
It has been switched to stable again.
The issue remains It is very likely the issue is caused by the gcc-4.5.2 which seems to link to /usr/lib64/libmpfr.so.1.
So there seems to be two solutions there: Either not applying the manual deletion recommended at the end of the ebuild report, or creating a link such as below: Code: | ln -s /usr/lib64/libmpfr.so.4 /usr/lib64/libmpfr.so.1 |
Thanks for your attention, interest ans support. |
|
Back to top |
|
|
phajdan.jr Retired Dev
Joined: 23 Mar 2006 Posts: 1777 Location: Poland
|
Posted: Tue Mar 22, 2011 10:39 am Post subject: Re: mpfr problem after update [Solved] |
|
|
CaptainBlood wrote: | I've just updated mpfr to stable 3.0.0. Thet I did a Code: | revdep-rebuild --library '/usr/lib64/libmpfr.so.1' | which didn't rebuild anything |
That's strange, I tested the upgrade path and a correct revdep-rebuild line was written. Could you post your emerge --info?
CaptainBlood wrote: | then I did Code: | rm '/usr/lib64/libmpfr.so.1' |
|
It's always a good idea to have a backup for a while.
CaptainBlood wrote: | Afterwards I had to recompile a kernel, which suprisingly failed. |
mpfr itself doesn't require kernel recomplication, so I guess it was something unrelated.
CaptainBlood wrote: | So I decided to before trying to rebuild the previous package.but it also faled. |
Always do a backup with quickpkg before removing critical packages (when in doubt, do make a backup).
And by the way, the right fix is to re-emerge gcc! _________________ http://phajdan-jr.blogspot.com/ |
|
Back to top |
|
|
Aquous l33t
Joined: 08 Jan 2011 Posts: 700
|
Posted: Tue Mar 22, 2011 4:33 pm Post subject: |
|
|
That's strange. My mpfr was updated today (I'm running stable amd64 btw), and it said that I should run revdep-rebuild and remove /usr/lib64/libmpfr.so.1. revdep-rebuild correctly flagged gcc to be rebuilt and all is well for me. |
|
Back to top |
|
|
dnadesign Apprentice
Joined: 23 Dec 2006 Posts: 172 Location: Poland
|
Posted: Wed Mar 23, 2011 9:50 pm Post subject: |
|
|
Yeah, it updated alright. And broke GCC... I didn't remove the libmpfr.so.1 yet, just ran "revdep-rebuild -library 'libmpfr.so.1'" and it FAILED: "configure: error: cannot run C compiled programs."... Boom, headshot. xD
The "revdep..." command is the one that is advised by mpfr package messages and it does a GCC re-emerge. The re-emerge fails. Interesting...
EDIT: Sorry for trollin' It seems I had this issue hidden for some time (I don't use binary apps, so I haven't got any means of detecting the problem). xD _________________ MacBook Pro 5.1
Intel C2D 2.53 GHz, 4GB RAM DDR3, GF 9400M and 9600M GT 512MB |
|
Back to top |
|
|
|