Bug 86437
Summary: | Glibc update breaks wine | ||
---|---|---|---|
Product: | [Retired] Red Hat Linux | Reporter: | drew.middlesworth |
Component: | glibc | Assignee: | Jakub Jelinek <jakub> |
Status: | CLOSED ERRATA | QA Contact: | Brian Brock <bbrock> |
Severity: | high | Docs Contact: | |
Priority: | medium | ||
Version: | 8.0 | CC: | alessandro.volpi5, bugzilla, chris.ricker, dcb, djuran, fweimer, hcamp, mitr, sdp |
Target Milestone: | --- | ||
Target Release: | --- | ||
Hardware: | i686 | ||
OS: | Linux | ||
Whiteboard: | |||
Fixed In Version: | Doc Type: | Bug Fix | |
Doc Text: | Story Points: | --- | |
Clone Of: | Environment: | ||
Last Closed: | 2003-04-09 19:21:27 UTC | Type: | --- |
Regression: | --- | Mount Type: | --- |
Documentation: | --- | CRM: | |
Verified Versions: | Category: | --- | |
oVirt Team: | --- | RHEL 7.3 requirements from Atomic Host: | |
Cloudforms Team: | --- | Target Upstream Version: | |
Embargoed: |
Description
drew.middlesworth
2003-03-21 22:53:13 UTC
Seeing this as well. Has broken quite a bit of which relies on wine locally. Currently downloading src rpm to see if that'll do the trick. Can you please try ftp://people.redhat.com/jakub/glibc/errata/8.0/ ? I've observed two problems with wine since the glibc upgrade. The first had to do with lstat()ing the wineserver socket. That one is fixed by your newer packages. The second, which shows up with Half-Life for me, is: XIO: fatal IO error 0 (Success) on X server ":0.0" after 6175 requests (6147 known processed) with 0 events remaining. This pops up anywhere from just after the main menu starts drawing to right after clicking a menu option. This is not fixed by your newer packages. Can you please try ftp://people.redhat.com/jakub/glibc/errata/8.0/*4.80.3* ? 4.80.1 was still using only /lib libs, not /lib/i686 and thus non-FLOATING_STACKS libpthread. I have the exact same problem as the other posters. I tried the latest glibc that you offered at comment #4 and I still have the problem, wine won't work: XIO: fatal IO error 0 (Success) on X server ":0.0" after 169 requests (169 known processed) with 0 events remaining. *** Bug 87569 has been marked as a duplicate of this bug. *** glibc 2.3.2-4.80.3 works for me, fixes the problem I have also seen this. Updating to glibc 80.3 didn't fix it. I've noticed that it seems to be timing dependant. I'm running on a dual 300 Mhz SMP system. If I load one CPU down the problem seems to go away, at least it doesn't happen as often. Updating from RH's RPM wine to the latest Wine (3/18/2003) also didn't fix it. Finally, I tried installing fonts into Wine by copying a bunch of Windows fonts into C:/WINDOWS/Fonts, and the problem vanished. So here's what seems to work: glibc 80.3 wine 3/18/2003 RH 8.0 SMP w/ all updates as of 4/1/2003, except for glibc as above Required fonts in the Windows/Fonts folder I suspect that the fonts trick was all that was required. On a dual Athlon, I get the XIO: fatal IO error 0 (Success) on X server ":0.0" after 12161 requests (11917 known processed) with 0 events remaining. reliably, after a few random seconds. Only booting non-SMP kernel solves the problem. glibc-2.3.2-4.80.3 wine-20020605-2 (also tried a few other versions) Got the same problem on a single processor I686 500MHz RH 8.0 updated to 2004-04-04 Wine 20020605 glibc-2.3.2-4.80.4 update plus four VGA fonts in /Windows/Fonts solved/bypassed it. There is another thing. wine (with exception of latest CVS wine) assumes it can override __errno_location symbol and __h_errno_location symbol. But this is not possible in glibc 2.3.2 any longer, since glibc calls that function directly, bypassing .plt. I've built glibc-2.3.2-4.80.6 in the same location which reverts the changes which made __errno_location and __h_errno_location unoverridable. Can you please check it out? This (glibc-2.3.2-4.80.6) fixes all the problems I've seen with: cxoffice-1.2 (tested only with wordview/xlview but...) missing /lib/libnss*.so.1 files (we have some semi-statically linked commercial apps which need them) random crashed with apps using pthreads (I guess this was the /lib/i686/ issue mentioned above). I know that QA can take a long time, but already this seems much better than plain glibc-2.3.2-4.80, do you have any idea how long before we can expect to see an official update RPM for RH8? [ Do some fo these issues also affect the RH9 version of glibc, I suspect that the ones related to kernel fun don't but some of the others probably do. I've not yet dared test RH9 (still waiting for feedback from those who paid to beta-test it :-) ] An errata has been issued which should help the problem described in this bug report. This report is therefore being closed with a resolution of ERRATA. For more information on the solution and/or where to find the updated files, please follow the link below. You may reopen this bug report if the solution does not work for you. http://rhn.redhat.com/errata/RHBA-2003-136.html I'm still getting: XIO: fatal IO error 0 (Success) on X server ":0.0" after 3765 requests (3752 known processed) with 0 events remaining. on RHL 9 with the errata glibc installed. Any word on what might be causing this? Specs: Red Hat Linux 9 1.0 GHZ CPU 256 Megs ram glibc-common-2.3.2-27.9 glibc-2.3.2-27.9 glibc-devel-2.3.2-27.9 glibc-kernheaders-2.4-8.10 RedHat 8 shuld asupposedly be supported untill Dec. 31st 2003. Has the support period been changed? This bug is reported against RedHat 8. But the errata glibc is still not published for RedHat8, only for RedHat 9! This is not true. See https://rhn.redhat.com/errata/RHSA-2003-089.html Concerning wine on RHL9, unlike RHL8 you need to run very recent CVS wine and configure it for NPTL. FWIW, I've tried both latest wine and crossover office 2.0 on RHL 9 + glibc errata, and they both work.... |