When I upgraded gcc to 3.2 I had glibc-2.2.5-r5 installed and win4lin worked OK. Now after attempting `emerge -u world` I have glibc-2.2.5-r7 *and* 2.2.5-r2. Starting Win4Lin provokes: $ win /opt/win4lin/bin/print_uid: relocation error: /opt/win4lin/bin/print_uid: undefined symbol: atexit /opt/win4lin/bin/print_uid: relocation error: /opt/win4lin/bin/print_uid: undefined symbol: atexit winconfig: relocation error: winconfig: undefined symbol: atexit /opt/win4lin/bin/print_uid: relocation error: /opt/win4lin/bin/print_uid: undefined symbol: atexit /opt/win4lin/charconv: relocation error: /opt/win4lin/charconv: undefined symbol: atexit /opt/win4lin/x_msg: relocation error: /opt/win4lin/x_msg: undefined symbol: atexit I tried to remerge glibc... # emerge glibc Calculating dependencies ...done! >>> emerge sys-libs/glibc-2.2.5-r7 to / !!! emerge aborting on /usr/portage/sys-libs/glibc/glibc-2.2.5-r7.ebuild .
This is the smae error I'm getting with realplayer, perhaps they are related?
Yep - realplay does that for me too
I get the same problem with vmware... 2.2.5-r6 works perfectly
I had the same problem.. I downgraded to -r6 and the problem is fixed.. glibc-2.2.5-sandbox.diff seems to be the problematic patch... Would it be possible to pull it back?
Was it a build from scratch, or did you update a 2.95.3 install ?
My system was updated from 3.1.1 and previously 3.1 (originally from a 1.3a iso) I recompiled glibc after updating to 3.2 (I update by changing the /etc/make.profile symlink and doing a emerge -u system)...
I had a 1.2 system that I upgraded, following Spider's guidelines.
same problem here, "undefined symbol: atexit" in the avmailgate virus-scanner, since 4 or 5 days. i have updated glibc in the last days, i think from 2.2.5-r5 to 2.2.5-r7, i'm using gcc-2.95.3-r7.
Hmm, ok. For the guys with the fast boxes out there, try: # emerge glibc gcc && emerge glibc Seems that we are not rid of the atexit bug yet :( Quake3 and UT works fine here ... any of you guys can test them ?
I can confirm that q3 and UT work correctly, but realplay and vmware break.. with -r7. I'm trying to do an "emerge glibc gcc && emerge glibc" right now.. I should have results soon...
tried the "emerge glibc gcc && emerge glibc" trick and I still get the same problem going back to -r6. The glibc-2.2.5-sandbox.diff patch is definitely broken.. it should be killed
# ld --version and paste please.
GNU ld version 2.13.90.0.4 20020814
$ ld --version GNU ld version 2.11.92.0.12 20011121 Copyright 2001 Free Software Foundation, Inc. This program is free software; you may redistribute it under the terms of the GNU General Public License. This program has absolutely no warranty. But that is now that I have abandoned 1.4 and restored 1.2 from backup.
GNU ld version 2.13.90.0.4 20020814 I get the same atexit problem as everybody else. 2.95.3, atexit, etc...
This version is masked again.
Is it fixed by masking? or what? I see no resolution here.
> This version is masked again. Meaning, -r7 is masked, meaning it will revert to -r6 with a emerge -u world, or new system.