Bug 161799 – Problems with login, PAM and UDEV !!

• Product: Gentoo Linux
• Status: VERIFIED INVALID

[Description Kevin 2007-01-12 17:53:34 UTC]

I've just finished installing with kde and am experiencing a bunch of problems that are blocking me off completely, from using gentoo. 
 
 I can login as root but, not as a non-root user. When I try to login as a non-root user in kde, it simply puts me back to the login screen. When I login in a text console, here'w aht I get: 
 
 unable to cd to `/home/kevin' 
 
 Here'are the relevant lines from the log: 
-------------------------------------------------------------------
 Jan 10 13:46:01 [login] pam_unix(login:session): session opened for user kevin by LOGIN(uid=0) 
 Jan 10 13:46:01 [login] unable to cd to `/home/kevin' for user `kevin'_ 
 Jan 10 13:46:01 [login] pam_unix(login:session): session closed for user kevin 
-------------------------------------------------------------------

I've checkd the permission of /home/kevin and found them to be 0755. I've double checked /etc/passwd and /etc/group to make sure that root user's uid and  gid both are set to 0 and that of user kevin are set to 1000.

I guess this has to do with pam and I'm no expert in pam. I've attached my /etc/pam.d/login: 
 
 Can somebody help me with this shitty pam issue ?? Plz ?? Is there a way to debug this pam issue ? 
 
 Secondly, I get these errors, actually hundreds of such lines scrolling past during boot/reboot/shutdown: 
 
--------------------------------------------------------------
   Jan 10 13:49:25 [udevd-event] run_program: exec of program '/sbin/udev_run_hotplugd' failed 
 Jan 10 13:49:25 [udevd-event] run_program: exec of program '/sbin/udev_run_devd' failed 
 Jan 10 13:49:25 [udevd-event] run_program: exec of program '/sbin/udev_run_hotplugd' failed 
 Jan 10 13:49:25 [udevd-event] run_program: exec of program '/sbin/udev_run_devd' failed 
 Jan 10 13:49:25 [udevd-event] run_program: exec of program '/sbin/udev_run_hotplugd' failed 
 Jan 10 13:49:25 [udevd-event] run_program: exec of program '/sbin/udev_run_devd' failed 
---------------------------------------------------------------

Here's the snippet from the syslog after installing HAL:
----------------------------SNIP------------------------------------
Jan 10 18:43:27 [hald] 18:43:27.689 [I] blockdev.c:584: block_add: sysfs_path=/sys/block/hda/hda1 dev=/dev/hda1 is_part=1, parent=0x082d2878_
Jan 10 18:43:27 [hald] 18:43:27.690 [I] hald_dbus.c:3173: Client to local_server was disconnected_
Jan 10 18:43:27 [hald] 18:43:27.691 [I] hald_dbus.c:3183: unregistered_
Jan 10 18:43:27 [hald] 18:43:27.695 [I] hald_dbus.c:3195: 25054: Got a connection_
Jan 10 18:43:27 [hald] 18:43:27.696 [I] hald_dbus.c:3196: dbus_connection_get_is_connected = 1_
Jan 10 18:43:27 [hald] 18:43:27.697 [D] hald_dbus.c:1192: udi=/org/freedesktop/Hal/devices/temp/117, key=volume.block_size_
------------------SNIP--------------
Jan 10 18:43:27 [hald] 18:43:27.844 [I] blockdev.c:494: get_luks_uuid: device_file=/dev/rd/1_
Jan 10 18:43:27 [hald] 18:43:27.844 [I] blockdev.c:620: Ignoring hotplug event - no parent_
Jan 10 18:43:27 [hald] 18:43:27.844 [W] blockdev.c:984: Not adding device object_
Jan 10 18:43:27 [hald] 18:43:27.844 [I] hotplug.c:181: /sys/block/ram0 is a block device (subsystem)_
Jan 10 18:43:27 [hald] 18:43:27.845 [I] blockdev.c:584: block_add: sysfs_path=/sys/block/ram0 dev=/dev/rd/0 is_part=1, parent=0x00000000_
Jan 10 18:43:27 [hald] 18:43:27.845 [I] blockdev.c:494: get_luks_uuid: device_file=/dev/rd/0_
Jan 10 18:43:27 [hald] 18:43:27.845 [I] blockdev.c:620: Ignoring hotplug event - no parent_
Jan 10 18:43:27 [hald] 18:43:27.845 [W] blockdev.c:984: Not adding device object_
Jan 10 18:43:27 [hald] 18:43:27.845 [I] acpi.c:1104: acpi_add: acpi_path=/proc/acpi/battery/BAT1 acpi_type=0, parent=0x00000000_
Jan 10 18:43:27 [hald] 18:43:27.864 [I] acpi.c:1076: Add callouts completed udi=/org/freedesktop/Hal/devices/acpi_BAT1_
Jan 10 18:43:27 [hald] 18:43:27.864 [I] hald.c:82: Added device to GDL; udi=/org/freedesktop/Hal/devices/acpi_BAT1_
Jan 10 18:43:27 [hald] 18:43:27.864 [I] acpi.c:1104: acpi_add: acpi_path=/proc/acpi/processor/CPU0 acpi_type=1, parent=0x00000000_
--------------------------SNIP-------------------------
Jan 10 18:43:27 [hald] 18:43:27.956 [I] acpi.c:1076: Add callouts completed udi=/org/freedesktop/Hal/devices/acpi_PWRF_
Jan 10 18:43:27 [hald] 18:43:27.956 [I] hald.c:82: Added device to GDL; udi=/org/freedesktop/Hal/devices/acpi_PWRF_
Jan 10 18:43:27 [hald] 18:43:27.956 [I] acpi.c:1104: acpi_add: acpi_path=/proc/acpi/button/sleep/SLPB acpi_type=10, parent=0x00000000_
Jan 10 18:43:27 [hald] 18:43:27.974 [I] acpi.c:1076: Add callouts completed udi=/org/freedesktop/Hal/devices/acpi_SLPB_
Jan 10 18:43:27 [hald] 18:43:27.974 [I] hald.c:82: Added device to GDL; udi=/org/freedesktop/Hal/devices/acpi_SLPB_
Jan 10 18:43:27 [hald] 18:43:27.974 [I] hald.c:616: Device probing completed_
Jan 10 18:43:27 [hald] 18:43:27.974 [I] hald_dbus.c:3251: entering_
Jan 10 18:43:27 [hald] 18:43:27.975 [E] hald_dbus.c:3258: dbus_bus_get(): Did not receive a reply. Possible causes include: the remote application did not send a reply, the message bus security policy blocked the reply, the reply timeout expired, or the network connection was broken._
---------------------------------------------------------------------

  What's with this ? I'm not yet familiar with the way gentoo handles udev and hotplug. I'm trying to figure out the cause here. 

Soon after I extract stage3 and portage tarballs and compiling the kernel and later booting into it works fine. I mean, I can log in as non-root user, udev runs fine without any problem. But, right I after I run "emerge -eDNa world" things get screwed up and I end up with the above described scenario. 

Just to ensure, I repeated the whole process with another partition and soon after emerge -eDNa world things go berserk !! 

Oh ! The first installation was carried in a chrooted environment while I were working in BLFS. I went up to the stage of installing kde and tons of other apps only to find later that I can't login as non-root user and hence into kde and X wouldn't lauch properly because of issues with udev/hal !! In my frustration I trashed it all !!

It could be possible that something is wrong with the USE flags or CFLAGS I'm using. I've attached /etc/make.conf I used.

Is udev/hal/pam somehow broken in 2006.1 profile ??

I'd appreciate any response in this regard, as I'm eager to reinstall gentoo from scratch onel last time !!

Thanks.

Kevin

[Comment 1 Kevin 2007-01-12 17:54:22 UTC]

Created attachment 106707 [details]
/etc/make.conf

[Comment 2 Jakub Moc (RETIRED) 2007-01-12 17:56:00 UTC]

1/ Kindly review http://bugs.gentoo.org/page.cgi?id=fields.html#bug_severity
2/ Kindly don't file a summary bug about totally unrelated issues.
3/ Kindly note that Gentoo Bugzilla is for bugs, it's not a support forum.

[Comment 3 Kevin 2007-01-12 17:56:13 UTC]

Created attachment 106709 [details]
/etc/pam.d/login

[Comment 4 Kevin 2007-01-12 18:19:03 UTC]

I don't understand ! I have been able to reproduce the problem for the second time. Like I have said, things work fine until I run "emerge -eDNa world" which updates the packages pam, shadow, cracklib and udev/hal etc and that's when things stop working altogether. 

I were thinking it be a bug of some kind. 

Please, tell me what could be wrong  with my setup ? Why do the binaries /sbin/udev_run_devd and /sbin/udev_run_hotplugd get deleted after performing an update ?

I'm lost completely !!

[Comment 5 Kevin 2007-01-12 18:30:07 UTC]

I was of the opinion that this was a bug in pam or udev, not sure ! 

Things work just fine until I do a "emerge -eDNa world" which updates udev, pam, shadow, cracklib and hal etc. That's when things just stop working. 

I'm just trying to figure out what is a way out of this ! I've tried asking for help in the gentoo, hal and pam mailing lists and that didn't seem to help.

Thanks anyway !

[Comment 6 Jakub Moc (RETIRED) 2007-01-12 18:31:11 UTC]

This is NOT a support forum. We have forums.gentoo.org or #gentoo on freenode for questions.

[Comment 7 Jakub Moc (RETIRED) 2007-01-12 18:37:35 UTC]

Once you've made sure that's there's actually a *bug* somewhere, file a new bug, one each for separate issues. Most of your issues look like you've failed to run etc-update or dispatch-conf altogether.

Closing this mess.

[Comment 8 Kevin 2007-01-13 02:12:37 UTC]

All right ! I'm gonna a file the two bugs I think exist in two seperate reports. I felt that the problems were related to ebuild or otherwise, I put them together in on report. As far as me failing to run etc-update or dispatch-conf, you would be glad (?) to know that I din't fail to run dispatch-conf. I've run dispatch-conf on some while I merged some manually. As to closing the bug, plz do so. I've decided to stick with BLFS and never to install gentoo. Thanks very much.