[Xen-bugs] [Bug 1080] New: Stock Xen 3.1, nForce 570 SLI chipset

http://bugzilla.xensource.com/bugzilla/show_bug.cgi?id=1080

           Summary: Stock Xen 3.1, nForce 570 SLI chipset - only 3GB out of
                    4GB RAM seen
           Product: Xen
           Version: unspecified
          Platform: x86-64
        OS/Version: All
            Status: NEW
          Severity: normal
          Priority: P2
         Component: Hypervisor
        AssignedTo: xen-bugs@xxxxxxxxxxxxxxxxxxx
        ReportedBy: eivanec@xxxxxxxxxxxxxxx


Hello all,

Affected system:
K9N SLI Platinum (nForce 570 SLI flashed to latest release)
Athlon X2 6000
4*1GB RAM

On this setup "vanilla" Debian Etch boots and sees the full amount of RAM, but
Xen 3.1 only sees 3071MB.

xm dmesg:

 __  __            _____  _   ___  
 \ \/ /___ _ __   |___ / / | / _ \ 
  \  // _ \ '_ \    |_ \ | || | | |
  /  \  __/ | | |  ___) || || |_| |
 /_/\_\___|_| |_| |____(_)_(_)___/ 

 http://www.cl.cam.ac.uk/netos/xen
 University of Cambridge Computer Laboratory

 Xen version 3.1.0 (shand@localdomain) (gcc version 3.4.4 20050314 (prerelease)
(Debian 3.4.3-13)) Fri May 18 15:44:17 BST 2007
 Latest ChangeSet: Fri May 18 15:44:01 2007 +0100 15041:2d268f9a6a8b

(XEN) Command line: /xen-3.1.0.gz 
(XEN)  0000000000000000 - 000000000009f000 (usable)
(XEN)  0000000000100000 - 00000000c0000000 (usable)
(XEN) System RAM: 3071MB (3145340kB)
(XEN) ACPI: Unable to locate RSDP
(XEN) Xen heap: 13MB (14192kB)
(XEN) Domain heap initialised: DMA width 32 bits
(XEN) Processor #0 15:3 APIC version 16
(XEN) Processor #1 15:3 APIC version 16
(XEN) Enabling APIC mode:  Flat.  Using 1 I/O APICs
(XEN) Using scheduler: SMP Credit Scheduler (credit)
(XEN) Detected 3000.019 MHz processor.
(XEN) HVM: SVM enabled
(XEN) CPU0: AMD Athlon(tm) 64 X2 Dual Core Processor 6000+ stepping 03
(XEN) Mapping cpu 0 to node 255
(XEN) Booting processor 1/1 eip 90000
(XEN) Mapping cpu 1 to node 255
(XEN) AMD: Disabling C1 Clock Ramping Node #0
(XEN) CPU1: AMD Athlon(tm) 64 X2 Dual Core Processor 6000+ stepping 03
(XEN) Total of 2 processors activated.
(XEN) ExtINT not setup in hardware but reported by MP table
(XEN) ENABLING IO-APIC IRQs
(XEN)  -> Using new ACK method
(XEN) Platform timer is 1.193MHz PIT
(XEN) Brought up 2 CPUs
(XEN) *** LOADING DOMAIN 0 ***
(XEN)  Xen  kernel: 64-bit, lsb, compat32
(XEN)  Dom0 kernel: 64-bit, lsb, paddr 0xffffffff80200000 -> 0xffffffff805883f0
(XEN) PHYSICAL MEMORY ARRANGEMENT:
(XEN)  Dom0 alloc.:   00000000ba000000->00000000bc000000 (731120 pages to be
allocated)
(XEN) VIRTUAL MEMORY ARRANGEMENT:
(XEN)  Loaded kernel: ffffffff80200000->ffffffff805883f0
(XEN)  Init. ramdisk: ffffffff80589000->ffffffff81526800
(XEN)  Phys-Mach map: ffffffff81527000->ffffffff81acaf80
(XEN)  Start info:    ffffffff81acb000->ffffffff81acb49c
(XEN)  Page tables:   ffffffff81acc000->ffffffff81add000
(XEN)  Boot stack:    ffffffff81add000->ffffffff81ade000
(XEN)  TOTAL:         ffffffff80000000->ffffffff81c00000
(XEN)  ENTRY ADDRESS: ffffffff80200000
(XEN) Dom0 has maximum 2 VCPUs
(XEN) Initrd len 0xf9d800, start at 0xffffffff80589000
(XEN) Scrubbing Free RAM: .done.
(XEN) Xen trace buffers: disabled
(XEN) Std. Loglevel: Errors and warnings
(XEN) Guest Loglevel: Nothing (Rate-limited: Errors and warnings)
(XEN) Xen is relinquishing VGA console.
(XEN) *** Serial input -> DOM0 (type 'CTRL-a' three times to switch input to
Xen).

After searching the lists I found a couple of related threads:

http://lists.xensource.com/archives/html/xen-users/2007-04/msg00862.html
http://lists.xensource.com/archives/html/xen-devel/2007-05/msg00010.html *

... which seem to suggest this derives from a known GRUB issue - for some
reason it's not paying attention to the e820 bios-provided memory map, and
consequently falling back to a second method which only maps memory up to the
3GB PCI memory hole. Hole-related BIOS options didn't do any good.

I consequently applied these two patches to GRUB:
http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=419994;msg=5
http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=419994;msg=15

But didn't see any benefit come from it.

Please note that in the xen-devel discussion I marked with a (*) developer Keir
Fraser mentions a possible patch/workaround, and in the following thread:

http://www.nabble.com/X86_64-and-4GB-RAM-t3662141.html

The user mentions he did get the patch, even though it doesn't seem it was
posted anywhere.

Given that it would seem that Xen enforces a possibly problematic method of
memory discovery (other kernels circumvent the problem) and at least one
developer considered this issue worth a workaround, I'm posting this here in
case it proves productive. Of course I'm not very much acquainted with the
inner workings of the system, so please apologize any errors in the information
and correct me when I'm wrong.

Thanks in advance!


-- 
Configure bugmail: 
http://bugzilla.xensource.com/bugzilla/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are the assignee for the bug, or are watching the assignee.

_______________________________________________
Xen-bugs mailing list
Xen-bugs@xxxxxxxxxxxxxxxxxxx
http://lists.xensource.com/xen-bugs
WARNING - OLD ARCHIVES

xen-bugs

[Xen-bugs] [Bug 1080] New: Stock Xen 3.1, nForce 570 SLI chipset - only
