Edit: short of it- no one should use yum to upgrade from Core 2 to Core 3 yet, you'll never see a console again. http://www.redhat.com/archives/fedora-test-list/2004-October/msg02697.html

I started with a fresh Fedora Core 2 install and upgraded it to Core 3 (replaced yum and the base package, then 'yum upgrade'). It appears to have worked, as I can boot, and then ssh in, but lish is having issues. I cannot get into lish, as in, I ssh to host35, and enter my password, and then nothing happens (using 2.4.28-linode37-1um), or sometimes I can get in- but the moment I try to connect to the console, or even logview, I get something like this:

ixion login: Kernel panic - not syncing: Kernel mode fault at addr 0x8, ip 0x400e5a18
 
EIP: 0073:[<400e5a18>] CPU: 0 Not tainted ESP: 007b:bffffc6c EFLAGS: 00000246
    Not tainted
EAX: 00000000 EBX: 080bf598 ECX: 00000004 EDX: 000007d0
ESI: 000007d0 EDI: 4014dff4 EBP: bffffc78 DS: 007b ES: 007b
Call Trace: 
 [<80034549>] vprintk+0xe9/0x13c
 [<800421ee>] notifier_call_chain+0x1e/0x38
 [<80033a9e>] panic+0x56/0xbc
 [<8001d907>] segv+0x83/0x1c8
 [<8001d9d6>] segv+0x152/0x1c8
 [<8030f1d7>] sigemptyset+0x17/0x30
 [<8001bd91>] change_signals+0x41/0x6c
 [<8030f116>] sigprocmask+0x1a/0x44
 [<8001dd3c>] segv_handler+0x160/0x208
 [<8001dcab>] segv_handler+0xcf/0x208
 [<80022d2c>] sig_handler_common_skas+0x7c/0x98
 [<80022d45>] sig_handler_common_skas+0x95/0x98
 [<8001bed4>] set_signals+0x6c/0x10c
 [<8001de2a>] sig_handler+0x32/0x34
 [<8030eef8>] __restore+0x0/0x8
 [<80023963>] chan_window_size+0x7/0x44
 [<8030f1d7>] sigemptyset+0x17/0x30
 [<80019dff>] um_virt_to_phys+0x43/0x7c
 [<8001bed4>] set_signals+0x6c/0x10c
 [<80022dae>] maybe_map+0x22/0x78
 [<8030edfd>] __libc_longjmp+0x3d/0x50
 [<8030eded>] __libc_longjmp+0x2d/0x50
 [<80022e5c>] do_op+0x58/0x60
 [<80022eda>] do_buffer_op+0x76/0x120
 [<80023080>] copy_chunk_to_user+0x0/0x2c
 [<80023080>] copy_chunk_to_user+0x0/0x2c
 [<8001ebf7>] setjmp_wrapper+0x4b/0x50

 [<8030f1d7>] sigemptyset+0x17/0x30
 [<8001bed4>] set_signals+0x6c/0x10c
 [<8004f500>] free_hot_cold_page+0xe0/0x11c
 [<8004f4c5>] free_hot_cold_page+0xa5/0x11c
 [<80076d24>] poll_freewait+0x3c/0x44
 [<80078039>] sys_poll+0x209/0x26c
 [<80076d2c>] __pollwait+0x0/0xa8
 [<800228a6>] execute_syscall_skas+0xaa/0xec
 [<80025149>] winch_interrupt+0x49/0xa8
 [<80017047>] handle_IRQ_event+0x27/0x6c
 [<800171e0>] do_IRQ+0x68/0xc4
 [<80022d84>] user_signal+0x3c/0x44
 [<80021efb>] userspace+0x18f/0x1e4
 [<8030eef8>] __restore+0x0/0x8
 [<8030f151>] kill+0x11/0x20

 
[screen is terminating]
[drake@host35 lish]#

Yes, /lib/tls is moved to /lib/tls-geborken. Any ideas?

Actually, it's not okay: with an ssh login, attempting to do almost anything results in the oom-killer going psychotic, killing everything, including init!

Out of Memory: Killed process 1 (init).
oom-killer: gfp_mask=0x1d2
DMA per-cpu:
cpu 0 hot: low 8, high 24, batch 4
cpu 0 cold: low 0, high 8, batch 4
Normal per-cpu: empty
HighMem per-cpu: empty

and then finally dying, as expected:

Kernel panic - not syncing: Out of memory and no killable processes...

 
EIP: 0073:[<0804cf2c>] CPU: 0 Not tainted ESP: 007b:bfffebbc EFLAGS: 00010293
    Not tainted
EAX: 00000010 EBX: bfffebe0 ECX: 0809b638 EDX: 0809b834
ESI: 0000059b EDI: 0000000c EBP: bfffec68 DS: 007b ES: 007b
Call Trace: 
 [<80034549>] vprintk+0xe9/0x13c
 [<800421ee>] notifier_call_chain+0x1e/0x38
 [<80033a9e>] panic+0x56/0xbc
 [<8004ea63>] select_bad_process+0x5b/0x80

and so on, but the panic is in select_bad_process. I can see how, after killing init, you have absolutely nothing left. For reference, I had about 200MB of swap free at the time.

Can you reproduce the kernel mode fault with 2.6.9-linode9?

-Chris

With a fresh, clean Fedora Core 2, using 2.6.9-linode9, booted from the web app, connecting via lish and typing the first letter of 'root' gave this:

li8-182 login: rKernel panic - not syncing: Kernel mode fault at addr 0x8bffffd6, ip 0xffffe410
 
EIP: 0073:[<ffffe410>] CPU: 0 Not tainted ESP: 007b:bffff990 EFLAGS: 00000246
    Not tainted
EAX: 00000000 EBX: 0000000b ECX: bffffac0 EDX: 00000000
ESI: 00000000 EDI: bffff9e8 EBP: bffffcd8 DS: 007b ES: 007b
Call Trace: 
 [<80034549>] vprintk+0xe9/0x13c
 [<800421ee>] notifier_call_chain+0x1e/0x38
 [<80033a9e>] panic+0x56/0xbc
...

The rest was the same as before.

Edit: And now I can't reproduce it. Blah.

Post-Edit: This post is all LIES! See below.

From the job logs on your host, it appears your next boot after deploying the new distro was into 2.4, so I don't believe 2.6.9 exhibits this behavior.

I'll send the fault info and trace to the UML developers. The 2.6.9-linode9 kernel has updated console code, so perhaps this is already fixed and can be back-ported to 2.4-um.

-Chris

You're right, it was, I forgot that the Distro Wizard makes a configuration with 2.4 as the default kernel.

Sorry about that, and thanks for the help.

Well, I guess I'm not seeing an console again:

Installing knfsd (copyright (C) 1996 okir@monad.swb.de).
SGI XFS with no debug enabled
SGI XFS Quota Management subsystem
pty: 256 Unix98 ptys configured
SLIP: version 0.8.4-NET3.019-NEWTTY (dynamic channels, max=256).
CSLIP: code copyright 1989 Regents of the University of California.
RAMDISK driver initialized: 16 RAM disks of 4096K size 1024 blocksize
loop: loaded (max 8 devices)
PPP generic driver version 2.4.2
PPP Deflate Compression module registered
PPP BSD Compression module registered
Universal TUN/TAP device driver 1.5 (C)1999-2002 Maxim Krasnyansky
md: linear personality registered as nr 1
md: raid0 personality registered as nr 2
md: raid1 personality registered as nr 3
md: raid5 personality registered as nr 4
raid5: measuring checksumming speed
   8regs     :  3311.600 MB/sec
   32regs    :  1158.800 MB/sec
raid5: using function: 8regs (3311.600 MB/sec)
md: multipath personality registered as nr 7
md: md driver 0.90.0 MAX_MD_DEVS=256, MD_SB_DISKS=27
md: Autodetecting RAID arrays.
md: autorun ...
md: ... autorun DONE.
LVM version 1.0.8(17/11/2003)
Initializing Cryptographic API
Initializing software serial port version 1
mconsole (version 2) initialized on /linodes/cryptwizard/.uml/cryptwizard/mconsole
Partition check:
 hda: unknown partition table
 hdb: unknown partition table
 hdc: unknown partition table
Initializing stdio console driver
Netdevice 0 (fe:fd:43:12:b0:c7) : TUN/TAP backend -
IPv6 v0.8 for NET4.0
IPv6 over IPv4 tunneling driver
NET4: Linux TCP/IP 1.0 for NET4.0
IP Protocols: ICMP, UDP, TCP, IGMP
IP: routing cache hash table of 512 buckets, 4Kbytes
TCP: Hash tables configured (established 4096 bind 8192)
IPv4 over IPv4 tunneling driver
GRE over IPv4 tunneling driver
Linux IP multicast router 0.06 plus PIM-SM
ip_conntrack version 2.1 (463 buckets, 3704 max) - 288 bytes per conntrack
ip_tables: (C) 2000-2002 Netfilter core team
ipt_recent v0.3.1: Stephen Frost <sfrost@snowman.net>.  http://snowman.net/projects/ipt_recent/
arp_tables: (C) 2002 David S. Miller
NET4: Unix domain sockets 1.0/SMP for Linux NET4.0.
ip6_tables: (C) 2000-2002 Netfilter core team
NET4: Ethernet Bridge 008 for NET4.0
802.1Q VLAN Support v1.8 Ben Greear <greearb@candelatech.com>
All bugs added by David S. Miller <davem@redhat.com>
kjournald starting.  Commit interval 5 seconds
EXT3-fs: mounted filesystem with ordered data mode.
VFS: Mounted root (ext3 filesystem) readonly.
Warning: unable to open an initial console.