Description of problem: With 2.6.90-{16,17}, gdm seems to hang. Once a cursor appears, it never proceeds. I privately mailed to Jakub about this issue, however anyway I file a bug report so that this bug can be tracked. Version-Release number of selected component (if applicable): glibc-2.6.90-15 (i686) gdm-2.20.0-13.fc8 scim-1.4.7-5.fc8 I don't know if this is Japanese environ specific, however I also write scim version. How reproducible: 100% Steps to Reproduce: 1. init 3 2. install glibc 2.6.90-{16,17} 3. init 5 Actual results: gdm hangs Additional info: Reverting to -15 is okay.
(In reply to comment #0) > Version-Release number of selected component (if applicable): > glibc-2.6.90-15 (i686) > gdm-2.20.0-13.fc8 > scim-1.4.7-5.fc8 This is of course glibc-2.6.90-{16,17} (i686)
I can't reproduce it here, no scim though. What exactly means gdm hang, does it allow you to pick users, enter password and then you don't get a desktop, or does it get stuck earlier?
(In reply to comment #2) > What exactly means gdm hang, does it allow you to pick users, enter password > and then you don't get a desktop No, after all rc sysinit daemons launch, black X screen appears, a mouse pointer appears, then it never proceeds. So no password dialog appears. For scim: I uses ja_JP and [tasaka1@localhost ~]$ rpm -qa | grep scim | sort scim-1.4.7-5.fc8 scim-anthy-1.2.4-2.fc8 scim-bridge-0.4.13-4.fc8 scim-bridge-gtk-0.4.13-4.fc8 scim-debuginfo-1.4.7-5.fc8 scim-devel-1.4.7-5.fc8 scim-gtk-1.4.7-5.fc8 scim-libs-1.4.7-5.fc8 scim-python-0.1.3-1.fc8 scim-rawcode-1.4.7-5.fc8 scim-tomoe-0.6.0-2.fc8 If this can help.
Created attachment 217571 [details] gdb log of gdmgreeter, scim-launcher attached gdb log of gdmgreeter, scim-laucher, and attached strace result log with gzipped.
Well, when I remove scim-python-0.1.3-1.fc8.i386, gdm does not hang and I can log in to GNOME normally.
Created attachment 218281 [details] gdb log again Again gdb log. this time glibc-debuginfo(-common), gdm-debuginfo, scim-debuginfo python-debuginfo and scim-python-debuginfo (and others) are installed.
See Also bug 321761
Based the gdb backtrace and bug 321761 I don't think this is related to scim and should be reproducible without it.
Now marking this bug as a dup of bug 321761 (now this issue is discussed on that bug) *** This bug has been marked as a duplicate of 321761 ***