I have compiled a kernel from the sources in xen-3.0.3_0-src.tgz. I generated an initrd via mkinitrd /boot/initrd-2.6.16.29-xen.img 2.6.16.29-xen (after doing a depmod) I did a make world/make install from the source directory, per instructions in the README. I extracted the files from the initrd file to see what the init script looked like. It has a line inserting the ext3.ko module. The last few lines from the output: (XEN) Xen trace buffers: disabled (XEN) Xen is relinquishing VGA console. (XEN) *** Serial input -> DOM0 (type ''CTRL-a'' three times to switch input to Xen). Kernel panic - not syncing: VFS: Unable to mount root fs on unknown-block(0,0) (XEN) Domain 0 crashed: ''noreboot'' set - not rebooting. The menu.lst (trimmed) serial --unit=0 --speed=19200 --word=8 --parity=no --stop=1 terminal --timeout=10 serial console title Fedora Core (2.6.16-29_FC4xen0) root (hd1,1) kernel /boot/xen.gz dom0_mem=262144 noreboot com1=19200n81 module /boot/vmlinuz-2.6.16-xen ro root=/dev/hdb2 quiet console=tty0 console=ttyS0,19200n8 module /boot/initrd-2.6.16.29-xen.img This seems to conform to everything I''ve ever been able to google on making Domain 0 boot. Here is the section that boots normally. title Fedora Core (2.6.11-1.1369_FC4) root (hd1,1) kernel /boot/vmlinuz-2.6.11-1.1369_FC4 ro root=LABEL=/12 rhgb quiet initrd /boot/initrd-2.6.11-1.1369_FC4.img Note that I had to replace the root=LABEL=/12 with root=/dev/hdb2 because it complained about not finding that label. There are no errors during compilation and no errors in the boot messages prior to the failure. The first few lines of the boot messages: Booting ''Fedora Core (2.6.16-29_FC4xen0)'' root (hd1,1) Filesystem type is ext2fs, partition type 0x83 kernel /boot/xen.gz dom0_mem=262144 noreboot com1=19200n81 [Multiboot-elf, <0x100000:0x815d8:0x48a28>, shtab=0x1ca078, entry=0x100000] module /boot/vmlinuz-2.6.16-xen ro root=/dev/hdb2 quiet console=tty0 console=tt yS0,19200n8 [Multiboot-module @ 0x1cb000, 0x2d703c bytes] module /boot/initrd.img [Multiboot-module @ 0x4a3000, 0x282200 bytes] Any suggestions as what to try next would be appreciated. I''ve googled and read all I could find on FC and xen to no avail. Thanks, Bo _______________________________________________ Xen-users mailing list Xen-users@lists.xensource.com http://lists.xensource.com/xen-users
Tim Post
2006-Dec-17 06:15 UTC
Re: [Xen-users] VFS: Unable to mount root fs on unknown-block(0,0)
On Sat, 2006-12-16 at 15:35 -0500, Bo wrote:> I have compiled a kernel from the sources in xen-3.0.3_0-src.tgz. I > generated an initrd via mkinitrd /boot/initrd-2.6.16.29-xen.img > 2.6.16.29-xen (after doing a depmod) > > I did a make world/make install from the source directory, per > instructions in the README. I extracted the files from the initrd file > to see what the init script looked like. It has a line inserting the > ext3.ko module. > > The last few lines from the output: > > (XEN) Xen trace buffers: disabled > (XEN) Xen is relinquishing VGA console. > (XEN) *** Serial input -> DOM0 (type ''CTRL-a'' three times to switch > input to Xen). > Kernel panic - not syncing: VFS: Unable to mount root fs on > unknown-block(0,0) > (XEN) Domain 0 crashed: ''noreboot'' set - not rebooting. >Bo, is it possible (with your hardware) to boot without an initrd? I''d look at the initrd itself. You notice unknown-block(0,0), when its obvious that you passed the correct root as a kernel paramater, which is correct at 1,1. My *guess*, and this is only a guess is that you could try swapping the order of ro and root= in your boot config. I don''t work much with Fedora, however it looks like "ro" is being passed as the real_root_fs, and a default value of 0,0 is being passed to pivot root because obviously no block device named ''ro'' exists. In other words, the order of the arguments could be an issue. Your initrd is "dumb". I''d try switching them, then try booting without the initrd. The chain loading process of pivoting to the real root FS is broken in between vmlinuz and your initrd. This should not be an issue, but doesn''t mean it isn''t an issue :) Again, just a guess but a simple one to try. Hope this helps - Best, -Tim _______________________________________________ Xen-users mailing list Xen-users@lists.xensource.com http://lists.xensource.com/xen-users
Bo
2006-Dec-17 13:02 UTC
Re: [Xen-users] VFS: Unable to mount root fs on unknown-block(0,0)
Tim Post wrote:> On Sat, 2006-12-16 at 15:35 -0500, Bo wrote: > >> I have compiled a kernel from the sources in xen-3.0.3_0-src.tgz. I >> generated an initrd via mkinitrd /boot/initrd-2.6.16.29-xen.img >> 2.6.16.29-xen (after doing a depmod) >> >> I did a make world/make install from the source directory, per >> instructions in the README. I extracted the files from the initrd file >> to see what the init script looked like. It has a line inserting the >> ext3.ko module. >> >> The last few lines from the output: >> >> (XEN) Xen trace buffers: disabled >> (XEN) Xen is relinquishing VGA console. >> (XEN) *** Serial input -> DOM0 (type ''CTRL-a'' three times to switch >> input to Xen). >> Kernel panic - not syncing: VFS: Unable to mount root fs on >> unknown-block(0,0) >> (XEN) Domain 0 crashed: ''noreboot'' set - not rebooting. >> >> > > Bo, is it possible (with your hardware) to boot without an initrd? I''d > look at the initrd itself. > > You notice unknown-block(0,0), when its obvious that you passed the > correct root as a kernel paramater, which is correct at 1,1. > > My *guess*, and this is only a guess is that you could try swapping the > order of ro and root= in your boot config. I don''t work much with > Fedora, however it looks like "ro" is being passed as the real_root_fs, > and a default value of 0,0 is being passed to pivot root because > obviously no block device named ''ro'' exists. > > In other words, the order of the arguments could be an issue. Your > initrd is "dumb". > > I''d try switching them, then try booting without the initrd. The chain > loading process of pivoting to the real root FS is broken in between > vmlinuz and your initrd. > > This should not be an issue, but doesn''t mean it isn''t an issue :) > > Again, just a guess but a simple one to try. > > Hope this helps - > Best, > -Tim > > > > >Using this: root (hd1,1) kernel /boot/xen.gz dom0_mem=262144 noreboot com1=19200n81 module /boot/vmlinuz-2.6.16-xen root=/dev/hdb2 ro quiet console=tty0 > module /boot/initrd-2.6.16.29-xen.img I get : Red Hat nash version 4.2.15 starting mount: error 6 mounting ext3 ERROR opening /dev/console!!!!: 2 error dup2''ing fd of 0 to 0 error dup2''ing fd of 0 to 1 error dup2''ing fd of 0 to 2 switchroot: mount failed: 22 Kernel panic - not syncing: Attempted to kill init! (XEN) Domain 0 crashed: ''noreboot'' set - not rebooting. Removing the initrd line results in the exact same error as before, i. e. Kernel panic - not syncing: VFS: Unable to mount root fs on unknown-block(0,0) Here''s an entry from the init script from the initrd echo Creating /dev mount -o mode=0755 -t tmpfs /dev /dev mknod /dev/console c 5 1 So I know there''s a console being created The first few lines of that file are: mount -t proc /proc /proc setquiet echo Mounted /proc filesystem echo Mounting sysfs mount -t sysfs /sys /sys You will notice, the message says Nash is starting, but the first "echo" isn''t displayed. Does that mean it can''t mount proc? I''m not sure I understand all I should know about proc but I would think it would be something built into the kernel and not a module. In any case, AFAIK it isn''t a ext3 file system. I think I''m confused. /dev/hdb2 is the root. It''s the only partition used by FC4. The system also has a Debian partition and an old Suse partition, but they won''t be involved here. So I went into the init script and put an echo as the first line. No output. I''m assuming that the problem is occurring before Nash actually starts processing input, but what? _______________________________________________ Xen-users mailing list Xen-users@lists.xensource.com http://lists.xensource.com/xen-users
Bo
2006-Dec-17 16:22 UTC
Re: [Xen-users] VFS: Unable to mount root fs on unknown-block(0,0)
Tim Post wrote:> On Sat, 2006-12-16 at 15:35 -0500, Bo wrote: > > Bo, is it possible (with your hardware) to boot without an initrd? I''d > look at the initrd itself. > > You notice unknown-block(0,0), when its obvious that you passed the > correct root as a kernel paramater, which is correct at 1,1. > > My *guess*, and this is only a guess is that you could try swapping the > order of ro and root= in your boot config. I don''t work much with > Fedora, however it looks like "ro" is being passed as the real_root_fs, > and a default value of 0,0 is being passed to pivot root because > obviously no block device named ''ro'' exists. > > In other words, the order of the arguments could be an issue. Your > initrd is "dumb". > > I''d try switching them, then try booting without the initrd. The chain > loading process of pivoting to the real root FS is broken in between > vmlinuz and your initrd. > > This should not be an issue, but doesn''t mean it isn''t an issue :) > > Again, just a guess but a simple one to try. > > Hope this helps - > Best, > -Tim > >I''ve gotten a bit further. I took out a setquiet in the init script, I put ide specs on the kernel line and I think it is finding the partition. kernel now looks like: module /boot/vmlinuz-2.6.16-xen root=/dev/hdb2 ide0=1x1f0,0x3f6,14 ide1=1x170,0x376,15 console=tty0 console=ttyS0,19200n8 Here''s the latest output: Loading ide-disk.ko module Loading jbd.ko module Loading ext3.ko module Creating root device Mounting root filesystem mount: error 6 mounting ext3 Switching to new root ERROR opening /dev/console!!!!: 2 error dup2''ing fd of 0 to 0 error dup2''ing fd of 0 to 1 error dup2''ing fd of 0 to 2 unmounting old /proc unmounting old /sys switchroot: mount failed: 22 Kernel panic - not syncing: Attempted to kill init! Google seems to find a lot of hits on the mount: error 6, but almost all of them are from people with roots on scsi or sata drives. This MB has a sata controller in addition to the ata controller, but I don''t have any sata devices connected. I have 3 drives, hda, hdb and hdd. With hdc being the CDROM drive. In comparing the output from this boot and the output from dmesg (FC4) I see the ide-disk module must have been compiled into the kernel on FC4, since there is output prefaced with VP_IDE: and probing the drives. Does anyone have any clues as what to try next? By the way this is a CPU0: AMD Athlon(TM) XP 2200+ system. Thanks, Jim.> > >_______________________________________________ Xen-users mailing list Xen-users@lists.xensource.com http://lists.xensource.com/xen-users