GDM2 no longer works with -current

[Timur I. Bakeyev]

On Thu, May 08, 2003 at 11:16:11AM -0400, Joe Marcus Clarke wrote:
> 
> I tried it this time with no malloc.conf, and a clean gdm.conf, and it
> still works.  Can you get a stacktrace with symbols from the dying gdm?

Here is the output of gdb:

> bt

#0  0x28891833 in poll () from /usr/lib/libc.so.5
#1  0x28a4b651 in _thread_kern_sched_state_unlock () from /usr/lib/libc_r.so.5
#2  0x28a4b050 in _thread_kern_scheduler () from /usr/lib/libc_r.so.5
#3  0xd0d0d0d0 in ?? ()
#4  0x08050e86 in gdm_start_first_unborn_local (delay=134701160) at gdm.c:781
#5  0x080521c6 in main (argc=-1077937188, argv=0xbfbffc2c) at gdm.c:1550
#6  0x0804f475 in _start ()

> fr 4

> l

776                             /* only the first local display has
777                              * timed login going on */
778                             if (gdm_first_login)
779                                     d->timed_login_ok = TRUE;
780
781                             if ( ! gdm_display_manage (d)) {
782                                     gdm_display_unmanage (d);
783                                     /* only the first local display gets
784                                      * autologged in */
785                                     gdm_first_login = FALSE;

> fr 5

#5  0x080521c6 in main (argc=-1077937188, argv=0xbfbffc2c) at gdm.c:1550
1550        gdm_start_first_unborn_local (0 /* delay */);
(gdb) l
1545            gdm_xdmcp_init();
1546
1547        create_connections ();
1548
1549        /* Start local X servers */
1550        gdm_start_first_unborn_local (0 /* delay */);
1551
1552        /* Accept remote connections */
1553        if (GdmXdmcp) {
1554            gdm_debug ("Accepting XDMCP connections...");

Process stays in memory and eats 90% of CPU... Also, for fist few seconds after
X server is shuted down, gdm-binary sits in Giant1 state...

Previous version(2.1.4.3) of gdm still works fine :()

Another common thing between Glenn's system and mine, as I noticed is GForce
video chipset...

Hm... Maybe I'd try new snapshot of CURRENT and NVidia drivers...

Regards,
	Timur.