Bug 115891

Only difference is: During installation, an UP kernel is used but for the
installed system a SMP kernel is trying to be booted.

*** Bug 115885 has been marked as a duplicate of this bug. ***

I've looked at several other similar bug reports, and have tried a variety of 
things to work past this problem.   
 
For example, I first tried setting the following as grub boot options: 
 
early_printk=vga vga=1 
 
But that doesn't work, because it hangs before anything can get logged, or at 
least before I can switch to the log console. 
 
So I tried these grub boot options: 
 
pci=noacpi noapic acpi_irq_balance 
 
These also didn't change anything. 
 
So I looked at the options that are set by "Linux (failsafe)", and wanted to 
find the combination of options that would make it work.  I guessed that it 
has something to do with powermanagement, so I tried: 
 
apm=off acpi=off 
 
That worked!  But I wanted to get the smallest number of option to make it 
easier to debug.  So I tried: 
 
apm=off 
 
That didn't work.  The server still hanged.  So I tried: 
 
acpi=off 
 
And that worked!  The acpi switch being off is the lone switch that makes the 
difference. 
 
I'm attaching /var/log/messages and dmesg. 

Ohoh, maybe this is one of the apci modules now loaded through initrd.
I expect the processor module...

I didn't manage to write the code to disable them via boot param.
Increase severity, whether there is still time for a fast hack (just declaring
three global varialbes for __setup and go out of init function of
fan/thermal/processor modules ...).

Could you please try:
boot into working system.
delete thermal,fan,processor modules from:
INITRD_MODULES="sata_promise sata_via via82cxxx processor thermal fan reiserfs ..."
in /etc/sysconfig/kernel
then invoke mkinitrd and try whether you can boot.

early_printk=vga vga=1
-> you should see something, delete other vga=XXX and splash= options in
/boot/grub/menu.lst and you should see something.

Removed thermal, fan and processor from INITRD_MODULES.  Then ran mkinitrd. 
 
Rebooted. 
 
No improvement.  Still doesn't work. 

I got the early_printk to work, thanks to your help, by removing vga=XXX and 
splash= from menu.lst.  Here's the abreviated output: 
 
ACPI: Looking for DSDT in initrd... not found! 
 not found! 
Using local APIC timer interrupts. 
Detected 12.528 MHz APIC timer. 
cpu_up: attempt to bring up CPU 1 failed 
Unable to handle kernel paging request at 0000006f812c5160 RIP: 
<ffffffff8016f4cb>{free_block+123} 
PGD 0 
Oops: 0000 [1] SMP 
 
All the register and callback data goes here.... I can include it if you 
want/need. 
 
<0> Kernel panic - not syncing: Attempted to kill init! 
 
 

Please include the full log.

That's easiest if you use earlyprintk=serial,ttyS0,baud and a null modem
cable to another machine. 

I was hoping you wouldn't ask me to do that.  ;-) 
 
But now that I've found a null modem cable, I've got the output: 
 
time.c: Using 3.579545 MHz PM timer. 
time.c: Detected 1804.115 MHz processor. 
Console: colour VGA+ 80x25 
Dentry cache hash table entries: 262144 (order: 9, 2097152 bytes) 
Inode-cache hash table entries: 131072 (order: 8, 1048576 bytes) 
Memory: 1019548k/1048544k available (2418k kernel code, 0k reserved, 932k 
data, 212k init) 
Calibrating delay using timer specific routine.. 3615.70 BogoMIPS 
(lpj=7231418) 
Security Framework v1.0.0 initialized 
SELinux:  Disabled at boot. 
Mount-cache hash table entries: 256 
CPU: L1 I Cache: 64K (64 bytes/line), D cache 64K (64 bytes/line) 
CPU: L2 Cache: 1024K (64 bytes/line) 
CPU 0(2) -> Node 0 -> Core 0 
mtrr: v2.0 (20020519) 
checking if image is initramfs... it is 
ACPI: Looking for DSDT in initrd... not found! 
 not found! 
Using local APIC timer interrupts. 
Detected 12.528 MHz APIC timer. 
cpu_up: attempt to bring up CPU 1 failed 
Unable to handle kernel paging request at 000000ef81cb1fb0 RIP: 
<ffffffff8016f4cb>{free_block+123} 
PGD 0 
Oops: 0000 [1] SMP 
CPU 0 
Modules linked in: 
Pid: 1, comm: swapper Not tainted 2.6.13-3-smp 
RIP: 0010:[<ffffffff8016f4cb>] <ffffffff8016f4cb>{free_block+123} 
RSP: 0000:ffff81003ffb1e58  EFLAGS: 00010012 
RAX: 0000001e002fca9e RBX: ffff810002532100 RCX: 000f0017e54f0000 
RDX: 0000000000000001 RSI: 0000000000000010 RDI: f000ff54f0000073 
RBP: 0000000000000010 R08: ffff81003ffb0000 R09: 0000000000000001 
R10: 0000000000019a28 R11: 0000000000000000 R12: ffff810002532508 
R13: 0000000000000000 R14: 0000000000000001 R15: ffff810002532528 
FS:  0000000000000000(0000) GS:ffffffff80508800(0000) knlGS:0000000000000000 
CS:  0010 DS: 0018 ES: 0018 CR0: 000000008005003b 
CR2: 000000ef81cb1fb0 CR3: 0000000000101000 CR4: 00000000000006e0 
Process swapper (pid: 1, threadinfo ffff81003ffb0000, task ffff81003ffaf500) 
Stack: ffffffff7fffffff 0000000000000000 ffff810002532100 ffff810002532568 
       0000000000000001 0000000000000000 ffffffff804cb460 ffffffff80170167 
       0000000100000001 ffffffff803d3620 
Call Trace:<ffffffff80170167>{cpuup_callback+455} 
<ffffffff8014a56f>{notifier_call_chain+31} 
       <ffffffff80156a01>{cpu_up+225} <ffffffff8010c15c>{init+268} 
       <ffffffff8010f92e>{child_rip+8} <ffffffff8010c050>{init+0} 
       <ffffffff8010f926>{child_rip+0} 
 
Code: 48 8b 14 c5 c0 ca 4c 80 48 8d 04 cd 00 00 00 00 48 c1 e1 06 
RIP <ffffffff8016f4cb>{free_block+123} RSP <ffff81003ffb1e58> 
CR2: 000000ef81cb1fb0 
 <0>Kernel panic - not syncing: Attempted to kill init! 
 

Change status back to ASSIGNED. 

Can you give me the full log starting with the beginning of the boot?

Sorry.  Adding screen log.  Attached. 

Created attachment 49414 [details]
screen log

Your BIOS is somewhat broken and generates an invalid SRAT table. The fallback
error code had a quirk that lead to the crash later. 

It should work if you boot with numa=noacpi 

Probably didn't make RC2, but you can get a kotd later with that 
change and then drop the command line option again

- patches.arch/srat-fallback: (#115891) Backport some x86-64 srat
fallback fixes