LinuxQuestions.org

LinuxQuestions.org (/questions/)
-   Linux - Hardware (http://www.linuxquestions.org/questions/linux-hardware-18/)
-   -   fedora 14 report 2.6GB RAM on AMD Zacate E-350 4GB (http://www.linuxquestions.org/questions/linux-hardware-18/fedora-14-report-2-6gb-ram-on-amd-zacate-e-350-4gb-872876/)

trembb22 04-04-2011 09:36 AM

fedora 14 report 2.6GB RAM on AMD Zacate E-350 4GB
 
Hi, I finnaly put my hand on a mini-itx board with the AMD E-350 2 cores Zacate CPU with ASROCK E350M1 E350 A50M R with 4 GB ram. I installed fedora 14 64 bits first, than 32 bits lastly and in both case, linux fedora 14 reported only 2.6GB of my 4GB ram.

On the bios, I read 4096MB with 384MB shared.

I booted a cd with memtest86 V4.20, and it reported 3709MB of ram, that seems ok.

under linux fedora 14: free -m, show: 2637MB,
under monitor, we can read, 2.6GiB

thanks for you help, I think fedora 15 should look at this before release.


Here some more information:

[max@max ~]$ dmesg | grep BIOS
[ 0.000000] BIOS-provided physical RAM map:
[ 0.000000] BIOS-e820: 0000000000000000 - 000000000009ec00 (usable)
[ 0.000000] BIOS-e820: 000000000009ec00 - 00000000000a0000 (reserved)
[ 0.000000] BIOS-e820: 00000000000e0000 - 0000000000100000 (reserved)
[ 0.000000] BIOS-e820: 0000000000100000 - 00000000a7c96000 (usable)
[ 0.000000] BIOS-e820: 00000000a7c96000 - 00000000a7ce5000 (ACPI NVS)
[ 0.000000] BIOS-e820: 00000000a7ce5000 - 00000000a7ce8000 (reserved)
[ 0.000000] BIOS-e820: 00000000a7ce8000 - 00000000a7ceb000 (ACPI NVS)
[ 0.000000] BIOS-e820: 00000000a7ceb000 - 00000000a7cec000 (reserved)
[ 0.000000] BIOS-e820: 00000000a7cec000 - 00000000a7ced000 (ACPI NVS)
[ 0.000000] BIOS-e820: 00000000a7ced000 - 00000000a7d10000 (reserved)
[ 0.000000] BIOS-e820: 00000000a7d10000 - 00000000a7d11000 (usable)
[ 0.000000] BIOS-e820: 00000000a7d11000 - 00000000a7d19000 (ACPI NVS)
[ 0.000000] BIOS-e820: 00000000a7d19000 - 00000000a7d41000 (reserved)
[ 0.000000] BIOS-e820: 00000000a7d41000 - 00000000a7d84000 (ACPI NVS)
[ 0.000000] BIOS-e820: 00000000a7d84000 - 00000000a7f00000 (usable)
[ 0.000000] BIOS-e820: 00000000e0000000 - 00000000f0000000 (reserved)
[ 0.000000] BIOS-e820: 00000000fec00000 - 00000000fec01000 (reserved)
[ 0.000000] BIOS-e820: 00000000fec10000 - 00000000fec11000 (reserved)
[ 0.000000] BIOS-e820: 00000000fed00000 - 00000000fed01000 (reserved)
[ 0.000000] BIOS-e820: 00000000fed61000 - 00000000fed71000 (reserved)
[ 0.000000] BIOS-e820: 00000000fed80000 - 00000000fed90000 (reserved)
[ 0.000000] BIOS-e820: 00000000fef00000 - 0000000100000000 (reserved)
[ 0.000000] BIOS-e820: 0000000100001000 - 0000000140000000 (usable)
[ 0.000000] AMI BIOS detected: BIOS may corrupt low RAM, working around it.
[ 0.000000] #3 [00000fce10 - 0000100000] BIOS reserved ==> [00000fce10 - 0000100000]
[ 0.000000] #5 [000009ec00 - 00000fcba0] BIOS reserved ==> [000009ec00 - 00000fcba0]
[ 0.000000] #6 [00000fcd7c - 00000fce00] BIOS reserved ==> [00000fcd7c - 00000fce00]
[ 1.645256] apm: BIOS not found.
[ 7.799790] hda_codec: ALC892: BIOS auto-probing.
[max@max ~]$

Here some more information:
[root@max max]# dmidecode -q
System Information
Manufacturer: To Be Filled By O.E.M.
Product Name: To Be Filled By O.E.M.
Version: To Be Filled By O.E.M.
Serial Number: To Be Filled By O.E.M.
UUID: 03000200-0400-0500-0006-000700080009
Wake-up Type: Power Switch
SKU Number: To Be Filled By O.E.M.
Family: To Be Filled By O.E.M.

Base Board Information
Manufacturer: ASRock
Product Name: E350M1
Version:
Serial Number:
Asset Tag:
Features:
Board is a hosting board
Board is replaceable
Location In Chassis:
Type: Motherboard

Chassis Information
Manufacturer: To Be Filled By O.E.M.
Type: Desktop
Lock: Not Present
Version: To Be Filled By O.E.M.
Serial Number: To Be Filled By O.E.M.
Asset Tag: To Be Filled By O.E.M.
Boot-up State: Safe
Power Supply State: Safe
Thermal State: Safe
Security Status: None
OEM Information: 0x00000000
Height: Unspecified
Number Of Power Cords: 1
Contained Elements: 0

Processor Information
Socket Designation: CPUSocket
Type: Central Processor
Family: <OUT OF SPEC>
Manufacturer: AMD E-350 Process
ID: 10 0F 50 00 FF FB 8B 17
Version: AMD E-350 Processor
Voltage: 1.3 V
External Clock: 100 MHz
Max Speed: 1600 MHz
Current Speed: 1600 MHz
Status: Populated, Enabled
Upgrade: None
Serial Number: To Be Filled By O.E.M.
Asset Tag: To Be Filled By O.E.M.
Part Number: To Be Filled By O.E.M.
Core Count: 2
Core Enabled: 2
Characteristics:
64-bit capable

Cache Information
Socket Designation: L1-Cache
Configuration: Enabled, Not Socketed, Level 1
Operational Mode: Write Back
Location: Internal
Installed Size: 128 kB
Maximum Size: 128 kB
Supported SRAM Types:
Unknown
Installed SRAM Type: Unknown
Speed: Unknown
Error Correction Type: None
System Type: Unified
Associativity: 2-way Set-associative

Cache Information
Socket Designation: L2-Cache
Configuration: Enabled, Not Socketed, Level 2
Operational Mode: Varies With Memory Address
Location: Internal
Installed Size: 1024 kB
Maximum Size: 1024 kB
Supported SRAM Types:
Unknown
Installed SRAM Type: Unknown
Speed: Unknown
Error Correction Type: None
System Type: Data
Associativity: 16-way Set-associative

System Slot Information
Designation: PCIE1
Type: x4 PCI Express
Current Usage: In Use
Length: Long
ID: 17
Characteristics:
3.3 V is provided
Opening is shared
PME signal is supported
Bus Address: 0000:00:04.0

OEM Strings
String 1: To Be Filled By O.E.M.

Physical Memory Array
Location: System Board Or Motherboard
Use: System Memory
Error Correction Type: None
Maximum Capacity: 8 GB
Number Of Devices: 2

Memory Array Mapped Address
Starting Address: 0x00000000000
Ending Address: 0x000FFFFFFFF
Range Size: 4 GB
Partition Width: 0

Memory Device
Total Width: 64 bits
Data Width: 64 bits
Size: 2048 MB
Form Factor: DIMM
Set: None
Locator: A1_DIMM1
Bank Locator: A1_BANK1
Type: DDR3
Type Detail: Synchronous
Speed: 1333 MHz
Manufacturer: Kingston
Serial Number: 6A0E78C4
Asset Tag: A1_AssetTagNum1
Part Number: 9905471-001.A00LF
Rank: 2

Memory Device Mapped Address
Starting Address: 0x00000000000
Ending Address: 0x0007FFFFFFF
Range Size: 2 GB
Partition Row Position: 1

System Boot Information
Status: No errors detected

BIOS Information
Vendor: American Megatrends Inc.
Version: P1.10
Release Date: 02/10/2011
Address: 0xF0000
Runtime Size: 64 kB
ROM Size: 4096 kB
Characteristics:
PCI is supported
BIOS is upgradeable
BIOS shadowing is allowed
Boot from CD is supported
Selectable boot is supported
BIOS ROM is socketed
EDD is supported
5.25"/1.2 MB floppy services are supported (int 13h)
3.5"/720 kB floppy services are supported (int 13h)
3.5"/2.88 MB floppy services are supported (int 13h)
Print screen service is supported (int 5h)
8042 keyboard services are supported (int 9h)
Serial services are supported (int 14h)
Printer services are supported (int 17h)
ACPI is supported
USB legacy is supported
BIOS boot specification is supported
Targeted content distribution is supported
BIOS Revision: 4.6

Memory Device
Total Width: 64 bits
Data Width: 64 bits
Size: 2048 MB
Form Factor: DIMM
Set: None
Locator: A1_DIMM0
Bank Locator: A1_BANK0
Type: DDR3
Type Detail: Synchronous
Speed: 1333 MHz
Manufacturer: Kingston
Serial Number: 6A0E92C5
Asset Tag: A1_AssetTagNum0
Part Number: 9905471-001.A00LF
Rank: 2

Memory Device Mapped Address
Starting Address: 0x00080000000
Ending Address: 0x000FFFFFFFF
Range Size: 2 GB
Partition Row Position: 1

[root@max max]#

trembb22 04-04-2011 09:39 AM

exact uname for fedora 14
 
and my current uname -r from fedora 14 is:
2.6.35.6-45.fc14.i686

cascade9 04-04-2011 11:38 AM

Seems like it could be a linux/driver problem, somebody here has the same problem-

http://lime-technology.com/forum/ind...?topic=11258.0

UNRAID 4.7 (based on slackware), and seems to have a 3GB limit.

I think that what is the last post (well, the last post currently) might have the reason-

Quote:

supposedly final AMD Fusion/Zacate drivers were not available until 2.6.38
If the drivers arent working right, it could explain the 3GB limit.

You could try getting a 2.6.38 kernel, and see if that fixes the problem.

TobiSGD 04-04-2011 11:57 AM

I also would have a look at the BIOS if there is an option for memory remapping and enable that.

johnsfine 04-04-2011 12:06 PM

Quote:

Originally Posted by TobiSGD (Post 4313574)
I also would have a look at the BIOS if there is an option for memory remapping and enable that.

Don't bother. The "BIOS-provided physical RAM map" in the first post shows the BIOS has remapped the ram that seems to be missing, so that BIOS option is not the problem.

Based on that map, 64 bit Linux or PAE 32 bit Linux should see the missing memory.

I don't know why it isn't working and normally would suggest some investigation into obscure possibilities (though at the moment I haven't yet thought of exactly what to investigate).

However, Cascade9's post has the feel of accuracy. I don't have any specific knowledge to say it is accurate, but the post feels plausible (in contrast to looking for a BIOS option that the first post shows is not the problem).

Other than that, are you sure about the reported 64 bit behavior? (I always suspect operator error for otherwise unlikely malfunctions. But I don't have any specific operator error in mind).

Quote:

than 32 bits
If the 32 bit wasn't PAE then its behavior was as expected.

Make sure you have a PAE kernel for 32 bit. Wild stab in the dark: If there is an obscure driver problem, maybe that driver is only broken for 64 bit and works for 32 bit PAE.

You might have read and believed the common misinformation that PAE is only required for more than 4GB. That is not correct. Your BIOS map shows a chunk of ram which is (just a trivial amount short of) 1GB in size that a 32 bit kernel without PAE cannot use and a 32 bit kernel with PAE should be able to use. That almost 1GB would be in addition to the 2.6GB you have access to now.

jefro 04-04-2011 03:35 PM

Memtest is correct and it is usable.

See this for some other ways to see ram instead of free. http://www.linuxquestions.org/questi...m-size-213382/

And top. http://www.computerhope.com/unix/top.htm

TobiSGD 04-04-2011 06:55 PM

Quote:

Originally Posted by johnsfine (Post 4313585)
Don't bother. The "BIOS-provided physical RAM map" in the first post shows the BIOS has remapped the ram that seems to be missing, so that BIOS option is not the problem.

...

(in contrast to looking for a BIOS option that the first post shows is not the problem).

You are right, I should have read that post more carefully. I apologize for my mistake.

trembb22 04-05-2011 02:22 PM

used PAE kernel without sucess
 
Hi, thanks to all..

now uname -r show
2.6.35.11-83.fc14.i686.PAE

free -m show
total of 2642 MB
the same as without PAE

also, I still see 2.6GiB under system Monitor

same info from cat /proc/meminfo and top

now maybe I should try the: 2.6.38 kernel
but I dont know how to install a new kernel without yum, so let me know how to do it please.

thanks for your help

johnsfine 04-05-2011 02:34 PM

Quote:

Originally Posted by trembb22 (Post 4314906)
now uname -r show
2.6.35.11-83.fc14.i686.PAE

free -m show
total of 2642 MB
the same as without PAE

That is pretty convincing and no longer sounds to me like operator error is a viable theory. So I have no guesses at the moment other than Cascade9's earlier theory.

Quote:

now maybe I should try the: 2.6.38 kernel
but I dont know how to install a new kernel without yum, so let me know how to do it please.
On the specifics of kernel version in fc14, I can't help you at all.

If no one else jumps in to answer that, you might want to start a new thread for it. How to diagnose a wrong report of ram size is a specialized topic attracting a specific set of experts such as myself and maybe not including the right experts for the very different question:

How to install 2.6.38 kernel in fc14

I don't know the significance of your qualifier "without yum" (I know that little about fc14). If that qualifier conveys some information an fc14 expert wouldn't already know, put it in your post or even in the subject line.

I use yum in Centos, and to install a new kernel I think my choices are install it easily with yum if the right version is available or get the whole kernel source code and rebuild it (at the limits of my skill set, so I'm far short of being able to tell someone else how even for Centos).

trembb22 04-05-2011 04:48 PM

Hi, thanks anyway... I used yum list kernel* to see all available kernel a can directly install under fedora 14 with yum and that's the way I used to install de PAE version but the new kernel 2.6.38 wasn't listed with yum so installing that kernel will require another methode other than yum, that's what I was meaning by..."without yum". So I am open to try installing a new kernel with little help from anyone or wait fedora 15 but, I think I have an issue that might not be fixed for fedora 15 if we let things going, the AMD E350 APU is so new that something wrong is probably happening with linux. What do you think?

trembb22 04-07-2011 10:11 AM

Tested Fedora 15 Alpha with Amd E350 Zacate APU
 
Hi, I booted fedora 15 Alpha Desktop to teste the new kernel.

uname -r gave my:
2.6.38-0.rc5.git1.1.fc15.i686

so I found a simple way to test kernel 2.6.38 and found the following error into var log messages.

pci_root PNP0A08.00: address space collision: host bridge windows <mem 0x000C8000-0x000dffff> conflic with video ROM <mem 0x000c000-0x000e1ff>

And System Monitor shown again 2.6Gig of total ram.

so, fedora 15 is not going to fix that problem with kernel 2.6.38.

Is there any expert that participate in federa developpement that could look at this? I will try to report that problem direcly to the fedora team.

johnsfine 04-07-2011 11:00 AM

Quote:

Originally Posted by trembb22 (Post 4317081)
Hi, I booted fedora 15 Alpha Desktop to teste the new kernel.

Good test. Sorry it was a bad result.

Quote:

pci_root PNP0A08.00: address space collision: host bridge windows <mem 0x000C8000-0x000dffff> conflic with video ROM <mem 0x000c000-0x000e1ff>
Trivial issue, not related to your significant problem.

I expect some message that is related to the real problem is in the kernel's boot log. Since you looked in the right place and found an unimportant message, I expect the important message does not have an obvious meaning.

Quote:

Is there any expert that participate in federa developpement that could look at this? I will try to report that problem direcly to the fedora team.
I think it is very unlikely that the issue is specific to Fedora (or even to the Red Hat family). So I think you should want attention from a kernel developer rather than from a Fedora developer (other than the possibility that there may be some overlap between the two groups).

I don't know how hard it is to get attention from a kernel developer (if they look at individual problems from ordinary users they would never have time to do anything else).

Try searching the log again. Something in there might narrow it down to a searchable issue that someone else has already reported and you can look up the status.

trembb22 04-08-2011 10:51 AM

thanks a lot johnsfine to let me know that I didn't yet put the finger on the problem, I remember having seen couples messages under var log messages that show the wrong amount of RAM during boot, so the moment when it happen in the log might help a kernel developper to know where it happen in the kernel.

I found this:

The official place for reporting bugs in the Linux kernel
http://www.tux.org/lkml/

I will prepare some more information and post the bug there..

thanks a lot

cascade9 04-10-2011 08:13 AM

Phoronix tested a E350/4GB RAM system with ubuntu 10.10 and dont seem to have reported any missing RAM-

http://www.phoronix.com/scan.php?pag...ion_e350&num=1

though they dont actually stae it clearly, I'd believe from the article that the "AMD Fusion/Zacate drivers were not available until 2.6.38" issue I posted above is more to do with video/networking drivers.

Maybe posting a question at the phoronix forums might clear things up for you.


All times are GMT -5. The time now is 07:20 AM.