After updating media-libs/ilmbase to 2.0.1 from 2.0.0 yesterday a lot of stuff needed to be rebuild to link against the new files, so I started doing so, one of them is DevIL which now cannnot find libstdc++ but I said ok, maybe things will work themselves out as they usually do with these problems in gentoo so I just continued on my updating, which after updating ICU to 51.2 I needed to rebuild sqlite, but now sqlite can't find libstdc++.so.6 (which is where it should be and pointing to libstdc++.so.6.0.17) so I thought I would post this on bugs.gentoo since no one else has this problem evidently, or they know how to fix it. I dont know because all the fixes I come to online have something to do with the wrong symlink being in place which it is not wrong, its just as it should be.
Build log, emerge --info and emerge -pqv are attached, thanks for any help in advance.
Created attachment 354420 [details]
Created attachment 354422 [details]
Created attachment 354424 [details]
What I am talking about is these lines:
/usr/lib/gcc/x86_64-pc-linux-gnu/4.7.3/../../../../x86_64-pc-linux-gnu/bin/ld: warning: libstdc++.so.6, needed by /usr/lib/gcc/x86_64-pc-linux-gnu/4.7.3/../../../../lib64/libicui18n.so, not found (try using -rpath or -rpath-link)
/usr/lib/gcc/x86_64-pc-linux-gnu/4.7.3/../../../../lib64/libicui18n.so: undefined reference to `operator delete(void*)@GLIBCXX_3.4'
but libstdc++.so.6 IS there and linked properly for ALL versions of gcc installed, this is ridiculous.
No this is not resolved, rebuilding ICU did nothing to fix the problem, ICU builds just fine, but then sqlite says the icu file cant find libstdc++.so.6, for DevIL it has --nothing-- to do with ICU and fails on a similar message about not finding libstdc++.so.6. Sorry to change the status here but that does not fix the problem. If it would be helpful I'll throw the DevIL compile log up here too so you can see multiple packages were affected right after updating media-libs/ilmbase to 2.0.1.
Created attachment 354466 [details]
build log for DevIL which fails the same way
This program fails to find libstdc++.so.6 as well but has nothing to do with ICU, this time its a media-libs/ilmbase file that cant find the library, and no rebuilding ilmbase will not fix the problem.
(In reply to Robert from comment #7)
> Created attachment 354466 [details]
> build log for DevIL which fails the same way
> This program fails to find libstdc++.so.6 as well but has nothing to do with
> ICU, this time its a media-libs/ilmbase file that cant find the library, and
> no rebuilding ilmbase will not fix the problem.
Rebuild media-libs/openexr, then.
I just rebuild media-libs/openexr and the same error happens with DevIL... Im not sure what openexr has to do with sqlite failing though... Im sure there are more packages affected by this if I where to go through them.. If I can do anything else on my end to help diagnosing this problem please let me know and Ill get it posted asap.
Also wanted to say I rebuild gcc:4.7 and 4.6 last night to rebuild libstdc++, so thats not it either.... just retried sqlite, same error...
Created attachment 354482 [details]
libxml2 shows a very similar error, no libstdc++.so.6
So I started to go through the other packages like I said and libxml2 will fail with a very similar message, but it seems to have something to do with icu this time as well. The error is:
/usr/lib/gcc/x86_64-pc-linux-gnu/4.7.3/../../../../x86_64-pc-linux-gnu/bin/ld: warning: libstdc++.so.6, needed by /usr/lib64/libicuuc.so, not found (try using -rpath or -rpath-link)
/usr/lib64/libicuuc.so: undefined reference to `operator delete(void*)@GLIBCXX_3.4'
libicuuc.so might be icu integration for libxml2, not sure but any way this is definitely a large problem.....
no one has any ideas on this? should I reboot and reinstall? When I downgraded ilmbase openexr and icu thing still fail to build, devil, sqlite and libxml2 still fail.
*** Bug 480008 has been marked as a duplicate of this bug. ***
No clue at all what happens here. Do you still have this problem?
(In reply to Andreas K. Hüttel from comment #14)
> No clue at all what happens here. Do you still have this problem?
See Bug 549994 for a solution.