CPU and Memory information

DavidDiggs · 03-19-2008, 08:48 PM

I've put together a system with two AMD FX-72's and four 2gig Sticks of memory.

The question is does the FX-72 show up as a single CPU or two? I only get one when I do "cat /proc/cpuinfo" (see below). Is this correct?

Also I have four sitcks of 2 gig memory total 8 gigs but "cat /proc/meminfo" (see below) shows "MemTotal: 906792 kB"? I dont get this.

The system is running Debian 4.0 r3

Code:

linux_host:~# uname -a
Linux linux_host 2.6.18-6-486 #1 Sun Feb 10 22:06:33 UTC 2008 i686 GNU/Linux

Code:

linux_host:~# cat /proc/cpuinfo
processor       : 0
vendor_id       : AuthenticAMD
cpu family      : 15
model           : 193
model name      : AMD Athlon(tm) 64 FX-72 Processor
stepping        : 3
cpu MHz         : 2800.141
cache size      : 1024 KB
fdiv_bug        : no
hlt_bug         : no
f00f_bug        : no
coma_bug        : no
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 ht syscall nx mmxext fxsr_opt lm 3dnowext 3dnow pni cx16 lahf_lm cmp_legacy svm cr8legacy ts fid vid ttp tm stc
bogomips        : 5604.42

Code:

linux_host:~# cat /proc/meminfo
MemTotal:       906792 kB
MemFree:        615460 kB
Buffers:         90832 kB
Cached:          85428 kB
SwapCached:          0 kB
Active:         127644 kB
Inactive:       118424 kB
HighTotal:           0 kB
HighFree:            0 kB
LowTotal:       906792 kB
LowFree:        615460 kB
SwapTotal:     5301368 kB
SwapFree:      5301368 kB
Dirty:              12 kB
Writeback:           0 kB
AnonPages:       69828 kB
Mapped:          28884 kB
Slab:            27540 kB
PageTables:       1316 kB
NFS_Unstable:        0 kB
Bounce:              0 kB
CommitLimit:   5754764 kB
Committed_AS:   162936 kB
VmallocTotal:   122576 kB
VmallocUsed:      7664 kB
VmallocChunk:   114580 kB

johnsfine · 03-19-2008, 09:04 PM

Why did you select a 32-bit kernel for a system with 8Gb of ram?

A 32-bit kernel can be compiled to support that much ram. Maybe an expert can tell from your uname output whether your kernel should. Sorry I can't tell. Similarly I expect an expert can tell whether that kernel includes SMP support (multi processor).

But even if your 32-bit kernel supports 8Gb, using a 32-bit kernel with 8Gb is a strange choice.

BTW, I used the 64-bit Mepis distribution (based on Debian) for my system with 8Gb ram. The kernel is 2.6.22-1-mepis64-smp #1 SMP PREEMPT Mon Feb 18 23:03:02 EST 2008 x86_64 GNU/Linux
The cpuinfo is

Code:

processor       : 0
vendor_id       : AuthenticAMD
cpu family      : 15
model           : 67
model name      : AMD Athlon(tm) 64 X2 Dual Core Processor 6400+
stepping        : 3
cpu MHz         : 3200.000
cache size      : 1024 KB
physical id     : 0
siblings        : 2
core id         : 0
cpu cores       : 2
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 ht syscall nx mmxext fxsr_opt rdtscp lm 3dnowext 3dnow pni cx16 lahf_lm cmp_legacy svm extapic cr8_legacy
bogomips        : 6407.51
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 tm stc

processor       : 1
vendor_id       : AuthenticAMD
cpu family      : 15
model           : 67
model name      : AMD Athlon(tm) 64 X2 Dual Core Processor 6400+
stepping        : 3
cpu MHz         : 3200.000
cache size      : 1024 KB
physical id     : 0
siblings        : 2
core id         : 1
cpu cores       : 2
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 ht syscall nx mmxext fxsr_opt rdtscp lm 3dnowext 3dnow pni cx16 lahf_lm cmp_legacy svm extapic cr8_legacy
bogomips        : 6407.51
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 tm stc

Notice the two cores of my one CPU are reported as two processors with the same physical id.

The first line of meminfo is

Code:

MemTotal:      8109568 kB

I hope those help you know what you should expect from your system once the kernel and/or other problems are fixed.

DavidDiggs · 03-19-2008, 10:09 PM

Oh wow, I think you're right. I can't believe I didn't notice that.

I'm going to try a reinstall with a 64bit version and an SMP kernel. I'll post back with the results.

J.W. · 03-20-2008, 01:40 AM

You might also check out the CONFIG_HIGHMEM parameter

More info here: http://www.netadmintools.com/part379.html

johnsfine · 03-20-2008, 04:22 PM

In several of these recent threads where memory wasn't seen, I was wondering about more easily distinguishing what part of the problem is in the kernel vs. Hardware+BIOS.

I just tested my guess at a good test for that on an 8Gb system with a kernel that can't see 8G.

To look at the beginning of dmesg, I used the command

Code:

less /var/log/dmesg

The part of that I was looking for says

Code:

BIOS-provided physical RAM map:
 BIOS-e820: 0000000000000000 - 000000000009ec00 (usable)
 BIOS-e820: 00000000000f0000 - 0000000000100000 (reserved)
 BIOS-e820: 0000000000100000 - 00000000cfe0ac00 (usable)
 BIOS-e820: 00000000cfe0ac00 - 00000000cfe0cc00 (ACPI NVS)
 BIOS-e820: 00000000cfe0ec00 - 00000000cfe5cc00 (reserved)
 BIOS-e820: 00000000cfe5cc00 - 00000000cfe5ec00 (ACPI data)
 BIOS-e820: 00000000cfe5ec00 - 00000000d0000000 (reserved)
 BIOS-e820: 00000000e0000000 - 00000000f0000000 (reserved)
 BIOS-e820: 00000000fe000000 - 00000000ff000000 (reserved)
 BIOS-e820: 00000000ffb00000 - 0000000100000000 (reserved)
 BIOS-e820: 0000000100000000 - 0000000230000000 (usable)
Warning only 4GB will be used.
Use a PAE enabled kernel.
3200MB HIGHMEM available.
896MB LOWMEM available.
found SMP MP-table at 000fe710

For those who don't know how to read such things

Code:

BIOS-e820: 0000000000100000 - 00000000cfe0ac00 (usable)

The above line says there is just over 3326Mb in the area starting at the 1Mb boundary of address space.

Code:

 BIOS-e820: 0000000100000000 - 0000000230000000 (usable)

The above line says there is exactly 4864Mb in the area starting at the 4Gb boundary of address space. Without a PAE enabled kernel, I can't use that.

So I have 8192Mb of ram of which
1 and a fraction are lost to the BIOS and to address space fragmentation
4864 are lost by lack of PAE
3326 and a fraction are usable.

Code:

3200MB HIGHMEM available.
896MB LOWMEM available.

I don't understand that part. Both of those have to be with the 3326 Mb that is usable. So they aren't disjoint. Is LOWMEM entirely contained within HIGHMEM or do they just overlap?

BTW from /proc/meminfo

Code:

MemTotal:      3367684 kB

divide that by 1024 to see that it is 3288 and a fraction Mb. So part of the 3326 is outside that total (but I think actually used for something, not wasted like the 4864).

syg00 · 03-20-2008, 05:13 PM

Quote:

Originally Posted by johnsfine

Code:

3200MB HIGHMEM available.
896MB LOWMEM available.

I don't understand that part. Both of those have to be with the 3326 Mb that is usable. So they aren't disjoint. Is LOWMEM entirely contained within HIGHMEM or do they just overlap?

Both ...

Back in the day, Linux was based on the assumption that *no-one* would ever have more than a Gig of memory. The 896 Meg is basically that Gig minus some buffers.
HIGHMEM was a "fiddle" added later to accommodate larger memories. Not one of Linus more favourite adaptions judging by comments he has made.
The _64 kernels don't suffer from this.

DavidDiggs · 03-20-2008, 08:36 PM

Looks like it was the HIGHMEM(32bit only) and non SMP kernel that was doing it.

I installed Debian 4.0r3 AMD64 and everything looks good. 4 cores visible and 8gigs ram.

Thanks guys.