Re: ULE status, invalid load, buildkernel times.

From: Attilio Rao <attilio_at_FreeBSD.org>
Date: Fri, 27 Jul 2007 12:08:55 +0200
Peter Wemm wrote:
> On Sunday 22 July 2007, Milos Vyletel wrote:
>> On Sun, Jul 22, 2007 at 01:48:46PM +0200, Milos Vyletel wrote:
> [...]
>>> No problem,
>>>
>>> I've extracted it and made a patch. If someone is intrested, it's
>>> on
>>>
>>> http://rulez.sk/~mv/cpu.patch
>> Well, i've just updated my kernel and it paniced right after
>> identifying cpu.
>>
>> CPU: AMD Athlon(tm) 64 X2 Dual Core Processor 4200+ (2205.01-MHz
>> K8-class CPU) Origin = "AuthenticAMD"  Id = 0x20f32  Stepping = 2
>>  
>> Features=0x178bfbff<FPU,VME,DE,PSE,TSC,MSR,PAE,MCE,CX8,APIC,SEP,MTRR,
>> PGE,MCA,CMOV,PAT,PSE36,CLFLUSH,MMX,FXSR,SSE,SSE2,HTT>
>> Features2=0x1<SSE3>
>>   AMD Features=0xe2500800<SYSCALL,NX,MMX+,FFXSR,LM,3DNow!+,3DNow!>
>>   AMD Features2=0x3<LAHF,CMP>
>>   Cores per package: 2
>> usable memory = 3211776000 (3062 MB)
>> avail memory  = 3105628160 (2961 MB)
>> kernel trap 12 with interrupts disabled
>>
>>
>> Fatal trap 12: page fault while in kernel mode
>> cpuid = 0; apic id = 00
>> fault virtual address   = 0x310
>> fault code              = supervisor read data, page not present
>> instruction pointer     = 0x8:0xffffffff8033953c
>> stack pointer           = 0x10:0xffffffff80855c70
>> frame pointer           = 0x10:0xffffffff80855c80
>> code segment            = base 0x0, limit 0xfffff, type 0x1b
>>                         = DPL 0, pres 1, long 1, def32 0, gran 1
>> processor eflags        = resume, IOPL = 0
>> current process         = 0 ()
>>
>> this is output from from dmesg.
>>
>> Thanks for suggestions.
>> Milos
> 
> Unfortunately this isn't much help.  What would be useful would be to 
> get a backtrace.  If you're not running GENERIC, a copy of your kernel 
> config would be useful.  Any other patches?

The patch is not going to work as the slot for SI_ORDER_SECOND was 
alredy held by kern/subr_smp.c::mp_start() function.

Could you please try this comprehensive patch?
It has a fix for the mp_start / lapic_init confusion and some tricks 
with CNXT-ID bit which should exactly identify HTT:
http://people.freebsd.org/~attilio/machdep_lapic.diff

I have still to stress test it, so please use caution. I will update 
soon if I will have more information (a 'work for me' / 'don't work for 
me' would be very appreciated though).

Thanks,
Attilio
Received on Fri Jul 27 2007 - 08:09:52 UTC

This archive was generated by hypermail 2.4.0 : Wed May 19 2021 - 11:39:15 UTC