Bugzilla – Bug 119377
Second-stage install doesn't create /dev/hdc
Last modified: 2006-03-08 08:19:45 UTC
I started installing Suse 10 RC4 from CDs on my Thinkpad T41p. I got through the first CD, then the installer rebooted me into the second-stage install. However, when the second-stage install prompted me to insert CD 2, I inserted it, but the installer kept asking me for it. It turns out that the /dev/hdc device node had not been created. I went into Ctrl-Alt-F2, did "mknod /dev/hdc b 22 0", and was able to proceed with the install normally after that.
can you please append YaST2 logs and hwinfo? (see http://www.opensuse.org/Bug_Reporting_FAQ#YaST)
Kay, didn't we had the same report (fixed WONTFIX) with a Dell laptop a few weeks ago?!?
Yes, sounds similar to #116234.
Maybe we can investigate this time what's going wrong. Is /dev/hdc there once you reboot the fully installed machine?
I had a similar Problem with my Sony Vaio PCG-GR214MP. I have been testing opensuse since the beta1 release, and didn't use the CD-drive until RC1. With RC1 I recognized the missing /dev/hdc too and created it manually. guest@fm:~> dmesg | grep ^hdc hdc: UJDA710, ATAPI CD/DVD-ROM drive hdc: ATAPI 24X DVD-ROM CD-R/RW drive, 2048kB Cache, UDMA(33) hdc was in dmesg even when the device wasn't there under /dev/ HTH Felix
Now, this is getting interesting ;) Taking AJ into CC (just precautionary).
Created attachment 51165 [details] yast-logs.tar.bz2
Created attachment 51167 [details] hwinfo.txt
(In reply to comment #4) > Maybe we can investigate this time what's going wrong. > > Is /dev/hdc there once you reboot the fully installed machine? Yes, it is there once I reboot after the installation is complete.
I couldn't reproduce it here. I would probably close this bugzilla and document it somewhere; might be a timing issue. However, as we can't reproduce it here and there is a workaround available we might as well close it. aj, ok?
*** Bug 116234 has been marked as a duplicate of this bug. ***
The duplicate mentions a Dell box. This bug's comment #5 mentions that the bug also happens on a Sony Vaio. And the bug itself is for a Thinkpad. So, we have three fairly common machines on which the bug has happened.
Federico, we tried to reproduce it on a Thinkpad here locally but could not. Do you have a reproduceable way for us to reproduce it? We'll try any hardware ;-)
#130943 may be a similar thing. But still no idea so far...
In the case of bug 130943 it is OK in most cases, only sometimes after machine reboot /dev/hdc is missing. Can there be any race (e. g. loading ide-cd during udevstart)?
*** Bug 130943 has been marked as a duplicate of this bug. ***
It's likely the event replay from initramfs, which loses events. We have no idea how to locate the problem so far, as we can't reproduce it. 10.1 will no longer do initramfs event replay and should not have this problem.
I can confirm this bug, happens on every boot. If that helps, this is from hal: [..] 15:09:08.367 [I] blockdev.c:547: block_add: sysfs_path=/sys/block/hdc dev= is_part=0, parent=0x08155738 15:09:08.368 [I] blockdev.c:768: parent_bus is ide scandir: No such file or directory 15:09:08.368 [I] blockdev.c:398: Probing storage device 8217: 15:09:08.371: probe-storage.c:168: Doing probe-storage for (bus ide) (drive_type cdrom) (udi=/org/freedesktop/Hal/devices/temp/77) (--only-check-for-fs==0) 8217: 15:09:08.371: probe-storage.c:182: Doing open ("", O_RDONLY | O_NONBLOCK) 8217: 15:09:08.371: probe-storage.c:185: Cannot open : No such file or directory 15:09:08.372 [I] hald_dbus.c:3007: 8183: Got a connection 15:09:08.372 [I] hald_dbus.c:3008: dbus_connection_get_is_connected = 1 15:09:08.373 [I] hald_dbus.c:2969: local_server_message_handler: destination=(null) obj_path=/org/freedesktop/DBus/Local interface=org.freedesktop.DBus.Local method=Disconnected 15:09:08.373 [I] hald_dbus.c:2985: Client to local_server was disconnected 15:09:08.373 [I] hald_dbus.c:2995: unregistered [..] I have installed the latest hal-0.5.4-6.4 (25.11.05)
After not having this problem for a long time it occured to me again. # dmesg | grep hdc ide1: BM-DMA at 0x1868-0x186f, BIOS settings: hdc:DMA, hdd:pio hdc: UJDA710, ATAPI CD/DVD-ROM drive hdc: ATAPI 24X DVD-ROM CD-R/RW drive, 2048kB Cache, UDMA(33) # LANG=CC ls -l /dev/hdc /bin/ls: /dev/hdc: No such file or directory # udevstart # LANG=CC ls -l /dev/hdc brw-r----- 1 root disk 22, 0 Dec 3 15:33 /dev/hdc So the problem is solved by running udevstart
I have no idea how to fix this. It's hopefully fixed in 10.1 with the completely new coldplug logic. Closing the bug.
I think that I have hit a variant of this bug. I have windows partitions on /dev/hda (Maxtor 6L200P0), the second - data, should be mounted at startup. But during startup I get: udev[1322]: udev_make_node: mknod(/dev/hda2, 060640, 3, 2) failed with error 'No such file or directory' And the partition will not be mounted. This happens at every boot. Bringing up a root shell later I can: # mount /dev/hda2 Without any complaints. This does always work. Sounds like a race to me.