Re: Problems with more than 8 AMD Opteron Cores per System

Re: Problems with more than 8 AMD Opteron Cores per System

[Daniel J Blueman]

I find the ondemand governor works as expected with 2.6.12 on my
Athlon64 Winchester [1]; as soon as I bzip2 a file, processor freq is
pinned at 1.8GHz and drops to 1GHz when idle.

--- [1]

$ cat /proc/cpuinfo
processor       : 0
vendor_id       : AuthenticAMD
cpu family      : 15
model           : 31
model name      : AMD Athlon(tm) 64 Processor 3000+
stepping        : 0
cpu MHz         : 1004.646
cache size      : 512 KB
fpu             : yes
fpu_exception   : yes
cpuid level     : 1
wp              : yes
flags           : fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge
mca cmov pat pse36 clflush mmx fxsr sse sse2 syscall nx mmxext
fxsr_opt lm 3dnowext 3dnow
bogomips        : 1988.83
TLB size        : 1024 4K pages
clflush size    : 64
cache_alignment : 64
address sizes   : 40 bits physical, 48 bits virtual
power management: ts fid vid ttp

Oliver Weihe wrote:
> Hello,
> 
> I've two Iwill 8way Opteron equipped with 8 Opteron 875 CPUs each.
> (In the past we build some systems with singlecore CPUs and they went
> very well)
> 
> The Problem now is that the machines crash during boot when maxcpus is
> greater than 8.
> 2.6.12-rc4 works well with maxcpus=8, with 9 or more it freezes after
> "Testing NMI Watchdog... OK"
> 
> 2.6.12-rc5 and above have more problems even with maxcpus=4 or less very
> early during booting.
> 
> 2.6.13-X crashes later during boot (from 2 to 16 cores it's the same
> behavior)
> 
> The last I can so on the console (kernel 2.6.13-rc2-git4, maxcpus=2..16)
> is:
> 
> NET: Registered protocol family 1
> Using IPI Shortcut mode
> int3: 0000 [1] SMP
> CPU4
> Modules linked in:
> Freeing unused kernel memory: 212k freed
> Pid: 0, comm: swapper Not tainted 2.6.13-rc2-git4-default
> RIP: 0010:[<ffffffff8050fc00>]
> 
> After that the machines are totally freezed.
> 
> With maxcpus=1 all (tested) versions >=2.6.12-rc3 are able to boot.
> 
> Any hints/ideas/what ever?
> 
> Regards,
>    Oliver Weihe
> 
> P.S. if you answer CC me ;)
___
Daniel J Blueman

Re: Problems with more than 8 AMD Opteron Cores per System

[Daniel J Blueman]

Ensure your BIOS supports the stepping of the Opterons you have. Eg, if
you have the Iwill H8501, BIOS v1.8 [1] covers all the current
steppings. You may even encounter problems if you mix different
steppings.

--- [1]

http://www.iwill.net/product_2s.asp?p_id=90&tp=BIOS

Oliver Weihe wrote:
> Hello,
>
> I've two Iwill 8way Opteron equipped with 8 Opteron 875 CPUs each.
> (In the past we build some systems with singlecore CPUs and they went
> very well)
>
> The Problem now is that the machines crash during boot when maxcpus is
> greater than 8.
> 2.6.12-rc4 works well with maxcpus=8, with 9 or more it freezes after
> "Testing NMI Watchdog... OK"
>
> 2.6.12-rc5 and above have more problems even with maxcpus=4 or less very
> early during booting.
>
> 2.6.13-X crashes later during boot (from 2 to 16 cores it's the same
> behavior)
>
> The last I can so on the console (kernel 2.6.13-rc2-git4, maxcpus=2..16)
> is:
>
> NET: Registered protocol family 1
> Using IPI Shortcut mode
> int3: 0000 [1] SMP
> CPU4
> Modules linked in:
> Freeing unused kernel memory: 212k freed
> Pid: 0, comm: swapper Not tainted 2.6.13-rc2-git4-default
> RIP: 0010:[<ffffffff8050fc00>]
>
> After that the machines are totally freezed.
>
> With maxcpus=1 all (tested) versions >=2.6.12-rc3 are able to boot.
>
> Any hints/ideas/what ever?
>
> Regards,
>    Oliver Weihe
>
> P.S. if you answer CC me ;)
___
Daniel J Blueman

Problems with more than 8 AMD Opteron Cores per System

[Oliver Weihe]

Hello,

I've two Iwill 8way Opteron equipped with 8 Opteron 875 CPUs each.
(In the past we build some systems with singlecore CPUs and they went
very well)

The Problem now is that the machines crash during boot when maxcpus is
greater than 8.
2.6.12-rc4 works well with maxcpus=8, with 9 or more it freezes after
"Testing NMI Watchdog... OK"

2.6.12-rc5 and above have more problems even with maxcpus=4 or less very
early during booting.

2.6.13-X crashes later during boot (from 2 to 16 cores it's the same
behavior)

The last I can so on the console (kernel 2.6.13-rc2-git4, maxcpus=2..16)
is:

NET: Registered protocol family 1
Using IPI Shortcut mode
int3: 0000 [1] SMP
CPU4
Modules linked in:
Freeing unused kernel memory: 212k freed
Pid: 0, comm: swapper Not tainted 2.6.13-rc2-git4-default
RIP: 0010:[<ffffffff8050fc00>]

After that the machines are totally freezed.

With maxcpus=1 all (tested) versions >=2.6.12-rc3 are able to boot.

Any hints/ideas/what ever?

Regards,
   Oliver Weihe

P.S. if you answer CC me ;)