#164 BSOD with 20091105 - VMALLOC_END overlaps host shared memory

v0.8.x (devel)
closed-fixed
Henry N.
7
2010-03-06
2009-11-06
Anonymous
No

with snapshot 0.8.0-20091105, the host got BSOD. no problem with 0.8.0-20090905.
downloaded from http://www.colinux.org/snapshots/
os: windows xp sp3 with all hotfix installed

i have retried many time, double check file permission, no luck

Discussion

  • Henry N.

    Henry N. - 2009-11-06

    Do clean upgrade steps by step:
    * Download archive "daemons-0.8.0-20091105.zip" form:
    http://www.henrynestler.com/colinux/testing/devel-0.8.0/20091105-Snapshot/packages/
    * Unpack all the files into new directory, for example to c:\coLinux-2.6.25
    * Copy your colinux config file or your BAT file from old installation into the new directory. Edit the new config- or bat- file and add the full path for your image files (cobd, scsi or hda). - If you not have used full path names before.
    * Remove or comment out the "initrd=initrd.gz" from config, and save the config.
    * Stop all running colinux
    * Open command prompt on old installation directory
    * Remove the driver
    colinux-daemon --remove-driver
    * Check, that the driver is truely removed:
    colinux-daemon --status-driver
    * Reboot your windows now!
    * Open command prompt on the new directory and perform the manual install:
    colinux-daemon --install-driver
    * Reboot your Windows now!
    (This step is need, if the start would crash at next step.)
    * Run your new colinux versioon from new directory.

     
  • Xuefer

    Xuefer - 2009-11-07

    reproduced only with big ram configured for colinux

    ram=960 -> no problem
    ram=992 (960+32) -> guest crash
    ram=1024 -> host crash

    ===================
    colinux-daemon output:
    Cooperative Linux Daemon, 0.8.0
    Daemon compiled on Thu Nov 5 23:08:04 2009

    using 'vmlinux' as kernel image
    configuring 992 MB of virtual RAM
    selected cobd0 for sda2, mapping to '\Device\Harddisk0\Partition2'
    selected cobd1 for sda3, mapping to '\Device\Harddisk0\Partition3'
    selected cobd2 for sda5, mapping to '\Device\Harddisk0\Partition4'
    configured NDIS bridge at 'coLinux' device as eth1
    MAC address: 00:ff:4f:0d:e7:c0
    mapping cofs0 to \DosDevices\c:\app\system\colinux\ mapping cofs1 to \DosDevices\c:\ mapping cofs2 to \DosDevices\d:\ mapping cofs3 to \DosDevices\i:\ mapping cofs4 to \DosDevices\g:\ kernel boot parameters: 'video0=4 root=/dev/sda3 bootlevel=c.boot softlevel=c.default'
    creating monitor
    PID: 2300
    colinux: launching console
    executing: colinux-console-fltk -a 2300
    launching daemon for conet1
    executing: colinux-ndis-net-daemon -i 2300 -u 1 -n "coLinux" -mac 00:ff:4f:0d:e7:c0 -p 1
    colinux: booting
    Ndis bridge probe on "1394 连接"
    Ndis bridge probe on "本地连接"
    Ndis bridge probe on "WAN"
    Ndis bridge probe on "Bluetooth"
    Ndis bridge probe on "VirtualBox Host-Only Network"
    Ndis bridge probe on "VirtualBox Host-Only Network"
    Ndis bridge probe on "VirtualBox Host-Only Network"
    Ndis bridge probe on "VirtualBox Host-Only Network"
    Ndis bridge probe on "coLinux"
    conet-ndis-daemon: Bridge on: coLinux
    Linux version 2.6.25.20-co-0.8.0 (hn@hn-dt) (gcc version 4.2.1 (SUSE Linux)) #1 PREEMPT Thu Nov 5 23:06:10 CET 2009
    992MB LOWMEM available.
    Entering add_active_range(0, 0, 253952) 0 entries of 256 used
    Zone PFN ranges:
    Normal 0 -> 253952
    Movable zone start PFN for each node
    early_node_map[1] active PFN ranges
    0: 0 -> 253952
    On node 0 totalpages: 253952
    Normal zone: 1984 pages used for memmap
    Normal zone: 0 pages reserved
    Normal zone: 251968 pages, LIFO batch:31
    Movable zone: 0 pages used for memmap
    Built 1 zonelists in Zone order, mobility grouping on. Total pages: 251968
    Kernel command line: video0=4 root=/dev/sda3 bootlevel=c.boot softlevel=c.default
    Enabling fast FPU save and restore... done.
    Enabling unmasked SIMD FPU exception support... done.
    Initializing CPU#0
    Setting proxy interrupt vectors
    PID hash table entries: 4096 (order: 12, 16384 bytes)
    Console: colour CoCON 80x25
    console [tty0] enabled
    Dentry cache hash table entries: 131072 (order: 7, 524288 bytes)
    Inode-cache hash table entries: 65536 (order: 6, 262144 bytes)
    Memory: 1002624k/1015808k available (2029k kernel code, 0k reserved, 618k data, 156k init, 0k highmem)
    .init : 0xcCPA: page pool initialized 1 of 1 pages preallocated
    Calibrating delay loop... 14522.77 BogoMIPS (lpj=72613888)
    Security Framework initialized
    SELinux: Disabled at boot.
    Capability LSM initialized
    Mount-cache hash table entries: 512
    CPU: L1 I Cache: 64K (64 bytes/line), D cache 64K (64 bytes/line)
    CPU: L2 Cache: 512K (64 bytes/line)
    Compat vDSO mapped to ffffe000.
    CPU: AMD Athlon(tm) 64 X2 Dual Core Processor 3800+ stepping 02
    Checking 'hlt' instruction... OK.
    Freeing SMP alternatives: 0k freed
    net_namespace: 536 bytes
    NET: Registered protocol family 16
    SCSI subsystem initialized
    PCI: Probing PCI hardware
    PCI: Probing PCI hardware (bus 00)
    NET: Registered protocol family 2
    IP route cache hash table entries: 32768 (order: 5, 131072 bytes)
    TCP established hash table entries: 131072 (order: 8, 1048576 bytes)
    TCP bind hash table entries: 65536 (order: 6, 262144 bytes)
    TCP: Hash tables configured (established 131072 bind 65536)
    TCP reno registered
    audit: initializing netlink socket (disabled)
    colinux: Linux VM terminated
    colinux: Kernel panic: Attempted to kill init!
    shutting down
    Pipe broken, exiting
    daemon cleanup
    rc=0b14b800 exit=0

    ===================

    kernel output: (no idea how to paste kernel output as text)
    see http://paste.ubuntu.com.cn/i46371

    cannot attach a screenshot file here as i didn't open this bug

     
  • Xuefer

    Xuefer - 2009-11-07

    btw, no crash if i configure as:

    mem=1024
    init=/bin/sh

    looks like there's some memory overrun colinux compiled with new kernel while 1024M ram is in use

     
  • Henry N.

    Henry N. - 2009-11-07

    Thanks. Think, the new kernel uses more "reserved" memory.
    I have successfully tested up to mem=984 with and without PAE in Windows boot.ini.
    Is 984M also working in your environment? Than I would change the limit to that value.

     
  • Henry N.

    Henry N. - 2009-11-08

    Real memory limit was 984 MB all times.
    With value of 1000 MB "vmalloc" and "lowmem" overlaps the "colinux" shared host memory and would crash, if Linux would use all of this memory later.

    Here is a virtual memory layout with adjusted pointers for vmalloc:
    fixmap : 0xffffc000 - 0xfffff000 ( 12 kB)
    colinux : 0xfe800000 - 0xffc00000 ( 20 MB)
    vmalloc : 0xfe000000 - 0xfe7fe000 ( 7 MB)
    lowmem : 0xc0000000 - 0xfd800000 ( 984 MB)

    Fixed in SVN revision r1285.
    New snapshot available.

     
  • Henry N.

    Henry N. - 2009-11-08
    • assigned_to: nobody --> henryn
    • status: open --> open-fixed
     
  • Nobody/Anonymous

    no crash with ram=any value. looks like it's fixed. thanks for your great work

     
  • Henry N.

    Henry N. - 2009-11-10

    Now, also fixed for the stable release 0.7.6-rc1 and committed to SVN release r1291.
    Download available from Snapshot page.

     
  • Henry N.

    Henry N. - 2009-11-10
    • priority: 5 --> 7
    • status: open-fixed --> closed-fixed
     
  • Henry N.

    Henry N. - 2010-03-06
    • summary: host BSOD on with 20091105 --> BSOD with 20091105 - VMALLOC_END overlaps host shared memory
     
  • Nobody/Anonymous

    Sourceforge.. Reposted it :)

     

Get latest updates about Open Source Projects, Conferences and News.

Sign up for the SourceForge newsletter:





No, thanks