Hey guys, I'm new here. I recently got a Gateworks GW2348-4 running Kamikaze 7.09 ixp4xx squashfs. I've got an AR5005 card in it and everything generally works pretty well. But once every 1 to 3 days it just seems to freeze. Drops all connections and won't allow any new ones, not HTTP, not SSH, not anything. Needs to be power cycled. This happens seemingly randomly - not when the CPU is loaded down or when I have a lot of open connections, just any time.  The network lights on it still flash.

I was reading some other threads and one of them gave me the idea of seeing if I can ping it from the WAN interface when this happens, but I'll have to wait for it to happen. The thing's got a serial port but I have to unplug it and move it near my computer to use that so it's working then.

dmesg:

Linux version 2.6.21.6 (nbd@ds10) (gcc version 4.1.2) #2 Sun Sep 30 20:44:34 CES                                                                                                  T 2007
CPU: XScale-IXP42x Family [690541c1] revision 1 (ARMv5TE), cr=000039ff
Machine: Gateworks Avila Network Platform
Memory policy: ECC disabled, Data cache writeback
On node 0 totalpages: 16384
  DMA zone: 128 pages used for memmap
  DMA zone: 0 pages reserved
  DMA zone: 16256 pages, LIFO batch:3
  Normal zone: 0 pages used for memmap
CPU0: D VIVT undefined 5 cache
CPU0: I cache: 32768 bytes, associativity 32, 32 byte lines, 32 sets
CPU0: D cache: 32768 bytes, associativity 32, 32 byte lines, 32 sets
Built 1 zonelists.  Total pages: 16256
Kernel command line: root=/dev/mtdblock2 rootfstype=squashfs,jffs2 noinitrd cons                                                                                                  ole=ttyS0,115200 init=/etc/preinit
PID hash table entries: 256 (order: 8, 1024 bytes)
Dentry cache hash table entries: 8192 (order: 3, 32768 bytes)
Inode-cache hash table entries: 4096 (order: 2, 16384 bytes)
Memory: 64MB = 64MB total
Memory: 62848KB available (1788K code, 167K data, 76K init)
Calibrating delay loop... 532.48 BogoMIPS (lpj=2662400)
Mount-cache hash table entries: 512
CPU: Testing write buffer coherency: ok
NET: Registered protocol family 16
IXP4xx: Using 16MiB expansion bus window size
PCI: IXP4xx is host
PCI: IXP4xx Using direct access for memory space
PCI: bus0: Fast back to back transfers enabled
dmabounce: registered device 0000:00:02.0 on pci bus
NET: Registered protocol family 2
Time: OSTS clocksource has been installed.
IP route cache hash table entries: 1024 (order: 0, 4096 bytes)
TCP established hash table entries: 2048 (order: 2, 16384 bytes)
TCP bind hash table entries: 2048 (order: 1, 8192 bytes)
TCP: Hash tables configured (established 2048 bind 2048)
TCP reno registered
NetWinder Floating Point Emulator V0.97 (double precision)
squashfs: version 3.0 (2006/03/15) Phillip Lougher
Registering mini_fo version $Id$
JFFS2 version 2.2. (NAND) (C) 2001-2006 Red Hat, Inc.
io scheduler noop registered
io scheduler deadline registered (default)
IXP4xx Watchdog Timer: heartbeat 60 sec
Serial: 8250/16550 driver $Revision: 1.90 $ 2 ports, IRQ sharing disabled
serial8250.0: ttyS0 at MMIO 0xc8000000 (irq = 15) is a XScale
serial8250.0: ttyS1 at MMIO 0xc8001000 (irq = 13) is a XScale
IXP4XX Q Manager 0.2.1 initialized.
IXP4XX NPE driver Version 0.3.0 initialized
ixp4xx_crypto 0.0.1 registered successfully
IXP4XX-Flash.0: Found 1 x16 devices at 0x0 in 16-bit bank
Intel/Sharp Extended Query Table at 0x0031
Using buffer write method
cfi_cmdset_0001: Erase suspend on write enabled
erase region 0: offset=0x0,size=0x20000,blocks=128
Searching for RedBoot partition table in IXP4XX-Flash.0 at offset 0xfe0000
5 RedBoot partitions found on MTD device IXP4XX-Flash.0
Creating 5 MTD partitions on "IXP4XX-Flash.0":
0x00000000-0x00080000 : "RedBoot"
npe: searching for firmware...
npe: found at 0x12104, unknown/NPE-A func: 03, rev: 5.8, size: 1074729484, id: 5                                                                                                  003c508
npe: found at 0x3fd1c, IXP465/NPE-A func: 81, rev: 2.0, size: 12848, id: 1081020                                                                                                  0
npe: found at 0x42f4c, IXP425/NPE-B func: 01, rev: 2.0, size: 12848, id: 0101020                                                                                                  0
Image loaded to NPE-B Func:1, Rel: 2:0, Status: 82400000
npe: found at 0x4617c, IXP425/NPE-C func: 01, rev: 2.0, size: 12848, id: 0201020                                                                                                  0
Image loaded to NPE-C Func:1, Rel: 2:0, Status: 80800000
0x00080000-0x00180000 : "kernel"
0x00180000-0x00fe0000 : "rootfs"
0x00280000-0x00fe0000 : "rootfs_data"
0x00fe0000-0x00fff000 : "FIS directory"
0x00fff000-0x01000000 : "RedBoot config"
i2c /dev entries driver
nf_conntrack version 0.5.0 (512 buckets, 4096 max)
ip_tables: (C) 2000-2006 Netfilter Core Team
TCP westwood registered
NET: Registered protocol family 1
NET: Registered protocol family 17
802.1Q VLAN Support v1.8 Ben Greear <greearb@candelatech.com>
All bugs added by David S. Miller <davem@redhat.com>
XScale DSP coprocessor detected.
ixp4xx_mac driver 0.3.1: eth0 on NPE-B with PHY[0] initialized
ixp4xx_mac driver 0.3.1: eth1 on NPE-C with PHY[1] initialized
drivers/rtc/hctosys.c: unable to open rtc device (rtc0)
VFS: Mounted root (squashfs filesystem) readonly.
Freeing init memory: 76K
Warning: unable to open an initial console.
mini_fo: using base directory: /
mini_fo: using storage directory: /jffs
device eth0 entered promiscuous mode
eth0: link up, 100Mbps, full-duplex, lpa 0xCDE1
Link of eth0 is full-duplex
br-lan: port 1(eth0) entering learning state
br-lan: topology change detected, propagating
br-lan: port 1(eth0) entering forwarding state
eth1: link up, 100Mbps, full-duplex, lpa 0x41E1
Link of eth1 is full-duplex
PPP generic driver version 2.4.2
IPP2P v0.8.1_rc1 loading
IMQ starting with 2 devices...
IMQ driver loaded successfully.
        Hooking IMQ before NAT on PREROUTING.
        Hooking IMQ after NAT on POSTROUTING.
wlan: 0.8.4.2 (svn r2568)
ath_hal: module license 'Proprietary' taints kernel.
ath_hal: 0.9.30.13 (AR5210, AR5211, AR5212, AR5416, RF5111, RF5112, RF2413, RF54                                                                                                  13, RF2133, REGOPS_FUNC)
ath_rate_minstrel: Minstrel automatic rate control algorithm 1.2 (svn r2568)
ath_rate_minstrel: look around rate set to 10%
ath_rate_minstrel: EWMA rolloff level set to 75%
ath_rate_minstrel: max segment size in the mrr set to 6000 us
wlan: mac acl policy registered
ath_pci: 0.9.4.5 (svn r2568)
PCI: enabling device 0000:00:02.0 (0340 -> 0342)
ath_pci: switching rfkill capability off
ath_pci: switching per-packet transmit power control off
wifi0: 11b rates: 1Mbps 2Mbps 5.5Mbps 11Mbps
wifi0: 11g rates: 1Mbps 2Mbps 5.5Mbps 11Mbps 6Mbps 9Mbps 12Mbps 18Mbps 24Mbps 36                                                                                                  Mbps 48Mbps 54Mbps
wifi0: turboG rates: 6Mbps 12Mbps 18Mbps 24Mbps 36Mbps 48Mbps 54Mbps
wifi0: H/W encryption support: WEP AES AES_CCM TKIP
wifi0: mac 7.9 phy 4.5 radio 5.6
wifi0: Use hw queue 1 for WME_AC_BE traffic
wifi0: Use hw queue 0 for WME_AC_BK traffic
wifi0: Use hw queue 2 for WME_AC_VI traffic
wifi0: Use hw queue 3 for WME_AC_VO traffic
wifi0: Use hw queue 8 for CAB traffic
wifi0: Use hw queue 9 for beacons
wifi0: Atheros 5212: mem=0x48000000, irq=27
device ath0 entered promiscuous mode
br-lan: port 2(ath0) entering learning state
br-lan: topology change detected, propagating
br-lan: port 2(ath0) entering forwarding state

Hmm, looks like it could be a memory leak? I can sit here and watch the RAM usage just keep going up. Not sure what happens when it reaches 100% but I can only assume that's when it freezes....

Seems to be related to QoS. Not happening with that off....

In my opinion, this - ath_pci: 0.9.4.5 (svn r2568) - is the troublemaker, use more recent version of madwifi and i bet it will be ok. There were pretty nasty bugs in madwifi for this platform, they are hopefully gone in trunk tho. I have put kamikaze 7.07 on my Compex WP188 - which is about the same as this board - and in usually died - panicked - after minutes, not hours or days with 4 minipci cards on. You could somehow prove this theory by using echo "5" > /proc/sys/kernel/panic_on_oops . Then board should, if my theory is correct, reboot itself instead of awaiting powercycle

It's most definitely QoS that was eating my memory. I'll try updating to trunk though.

I tried to compile OpenWRT. I mostly succeeded except for the ethernet ports. I downloaded IPL_ixp400NpeLibrary-2_4.zip and put it in the dl folder and that kinda worked. But it wouldn't bring up the interfaces, just said NPE-B and such were off.

I'll try just compiling madwifi separately and installing that on top of 7.9 for now.