I am not sure if this is relevant for your case. I assembled a PC based on the MSI-6570 K7N2 Delta (Nforce2) MB with an Athlon XP (Barton) CPU. The first I noticed was that it would now and then freeze during e2fsck on large file systems. I did an extensive memory test without finding any faults, so the situation was unclear.
Then I stumbled across a description of how to lower the CPU temperature by enabling the powersaving mode of the AMD Athlon/Duron processors. I found the Linux program athcool at http://members.jcom.home.ne.jp/jacobi/linux/softwares.html The following statement caught my attention: WARNING: Depending your motherboard and/or hardware components, enabling Athlon powersaving mode sometimes causes that (1) noisy sound playback (2) slowdown harddisk performance (3) system hangup or instable Evidently I was having (3) without having enabled powersaving. BUT: it turned out that the powersaving bit was turned on from the very start, presumably by the bios. After having turned off the powersaving bit by athcool I have NEVER had any single case of instability. I have had a few hangups during the boot-up procedure before powersaving was turned on. I did post a report on this problem on the MSI forum, but I have had absolutely ZERO response. Bjarne Thomsen lspcidrake -v unknown : Nvidia Corporation|nForce2 AGP Controller [BRIDGE_HOST] (vendor:10de device:01e0) unknown : Nvidia Corporation|nForce2 Memory Controller [MEMORY_RAM] (vendor:10de device:01eb subv:1462 subd:5700) unknown : Nvidia Corporation|nForce2 Memory Controller [MEMORY_RAM] (vendor:10de device:01ee subv:1462 subd:5700) unknown : Nvidia Corporation|nForce2 Memory Controller [MEMORY_RAM] (vendor:10de device:01ed subv:1462 subd:5700) unknown : Nvidia Corporation|nForce2 Memory Controller [MEMORY_RAM] (vendor:10de device:01ec subv:1462 subd:5700) unknown : Nvidia Corporation|nForce2 Memory Controller [MEMORY_RAM] (vendor:10de device:01ef subv:1462 subd:5700) unknown : Nvidia Corporation|nForce2 LPC / Legacy / System Management [BRIDGE_ISA] (vendor:10de device:0060 subv:1462 subd:5700) unknown : Nvidia Corporation|nForce2 SMBus 2.0 Controller [SERIAL_SMBUS] (vendor:10de device:0064 subv:1462 subd:5700) usb-ohci : Nvidia Corporation|nForce2 USB 1.0 OHCI Controller [SERIAL_USB] (vendor:10de device:0067 subv:1462 subd:5700) usb-ohci : Nvidia Corporation|nForce2 USB 1.0 OHCI Controller [SERIAL_USB] (vendor:10de device:0067 subv:1462 subd:5700) ehci-hcd : Nvidia Corporation|nForce2 USB 2.0 Enhanced Controller [SERIAL_USB] (vendor:10de device:0068 subv:1462 subd:5700) nvnet : Nvidia Corporation|nForce2 MCP Networking Adapter [NETWORK_ETHERNET] (vendor:10de device:0066 subv:1462 subd:570c) unknown : Nvidia Corporation|nForce2 APU [MULTIMEDIA_AUDIO] (vendor:10de device:006b subv:1462 subd:5700) snd-intel8x0 : Nvidia Corporation|nForce2 Audio Codec Interface [MULTIMEDIA_AUDIO] (vendor:10de device:006a subv:1462 subd:5700) unknown : Nvidia Corporation|nForce2 External PCI Bridge [BRIDGE_PCI] (vendor:10de device:006c) unknown : Nvidia Corporation|nForce2 UDMA 100 IDE Controller [STORAGE_IDE] (vendor:10de device:0065 subv:1462 subd:5700) ohci1394 : Nvidia Corporation|nForce2 Firewire Controller [SERIAL_FIREWIRE] (vendor:10de device:006e subv:1462 subd:570d) unknown : Nvidia Corporation|nForce2 AGP Host to PCI Bridge [BRIDGE_PCI] (vendor:10de device:01e8) aic7xxx : Adaptec|7892B [STORAGE_SCSI] (vendor:9005 device:0081 subv:9005 subd:62a1) 3c59x : 3Com Corporation|3c905C-TX [Fast Etherlink] [NETWORK_ETHERNET] (vendor:10b7 device:9200 subv:10b7 subd:1000) pdc-ultra : Promise Technology|PDC20376 FastTrak 376 Controller [STORAGE_RAID] (vendor:105a device:3376 subv:105a subd:6620) Card:NVIDIA GeForce4 (generic): NVidia|0x322 [DISPLAY_VGA] (vendor:10de device:0322 subv:1462 subd:9171) unknown : Unknown|USB OHCI Root Hub [Hub|Root Hub] (vendor:0000 device:0000) unknown : Unknown|USB OHCI Root Hub [Hub|Root Hub] (vendor:0000 device:0000) Removable:memory_card: JMTek|USBdisk [Mass Storage|SCSI|Bulk (Zip)] (vendor:0c76 device:0005) On Sat, 2003-08-16 at 12:21, Adam Williamson wrote: > On Sat, 2003-08-16 at 06:52, Tom Brinkman wrote: > > On Friday August 15 2003 08:08 pm, Adam Williamson wrote: > > > K, so to reply to myself again, I realised I was once more being > > > silly, and drew the obvious link between the two, and yes, the > > > problem is heavy network usage. Basically, downloading a file > > > from the internet will trigger it; do it any which way (urpmi, > > > galeon, wget) and the system immediately begins to resemble > > > treacle. I hope someone not using nvnet is seeing this, because > > > if they aren't, it might be difficult to sort > > > > you never should'a chose a nforce chipset to begin with > > > > no problems here, kt400a, xp3000+ overclocked to 2288, and better > > performance than nF2. Face it, nVidia is a last resort, not a > > choice. > > Stop trolling; it's entirely irrelevant. This post was utterly and > entirely pointless. (And besides, nforce2 is consistently shown to be > substantially faster than the kt400, and kt400a was not available when I > built the system. There are also no SFF PCs available based around the > VIA chipsets. Is that enough reasons for you yet?) -- Bjarne Thomsen <[EMAIL PROTECTED]> Institute of Physics & Astronomy