From Bugzilla Helper: User-Agent: Mozilla/5.0 (Windows; U; Windows NT 5.0; en-US; rv:1.4) Gecko/20030624 Description of problem: Problem originally manifested scp'ing a >3GB file on LAN, lately occurs during booting, every time. Changed BIOS to latest Promise BIOS, no change. Changed controller with FastTrak 100 TX2, no change. Changed controller with FastTrak 33, works fine. Am running one software RAID 1 device over two drives, and using two other partitions not in RAID between the two, all formatted ext3. Hang usually occurs right after red "Red Hat. . . I for interactive startup" prompt. Sometimes a couple of lines afterward. With the FastTrak 33, at that point get "Spurious 8259A Interrupt 7" error, but boot continues OK. IBM/Hitachi drive advanced diagnostics with Ultra 133 TX2 passes with flying colors, so I don't think it's a hardware problem. Does not seem temperature related either. Version-Release number of selected component (if applicable): Maxtor BIOS 2.20.0050.10, Promise BIOS 2.20.0.15 How reproducible: Always Steps to Reproduce: 1. Try to boot (I am using boot floppy with 2.4.20-18.4) Actual Results: Locks up solid at point mentioned above. Power switch won't do anything, have to hardware reset or disconnect power. Expected Results: Boots Additional info: Tried nmi_watchdog options, no oops reported at hang. Apparently nothing gets logged to syslog, even though file system for /var (and basically everything else) is mounted on a couple of SCSI drives with software RAID. I've subbed out power supply, processor, motherboard, controller, cables, everything except for the drives (which test good)
Red Hat Linux 7.2 is no longer supported by Red Hat. Please try upgrading to a newer OS release that is currently supported, such as Fedora Core 1 or Red Hat Enterprise Linux 3, and if the problem still occurs, feel free to file a new bug report. Thanks. Closing as CURRENTRELEASE
Perhaps reopen this bug? I just ran into a situation where I had 2 linux software RAID5 drives go flakey within a day of eachother. To try to recover data I am doing a dd from a flakey drive to a new drive. The flakey drive has multiple bad sectors and causes many syslog errors. Here's the interesting part: if the flakey drive is plugged into my Promise PDC20268 Ultra100 TX2 IDE controller, it freezes the entire system when dd hits certain sectors on the drive. If the flakey drive is plugged into the motherboard (Intel E7205) controller then error messages are generated but the system does not freeze (the correct behaviour). I reproduced the problem a number of times (probably 20+) trying to get the data off before I put the drive onto the mobo instead. I'm using the latest FC3 with kernel-2.6.9-1.667 Symptom: complete system freeze Expected: errors logged, but no freeze Log right before complete freeze is below. I tried turning off the DMA in the config files and rebooting and after that I wouldn't get dma errors and the bus wouldn't try to reset -- the system would just freeze or output a drive error then freeze. Nov 24 00:00:10 pog kernel: hdg: dma_intr: status=0x51 { DriveReady SeekComplete Error } Nov 24 00:00:10 pog kernel: hdg: dma_intr: error=0x01 { AddrMarkNotFound }, LBAsect=212833983, high=12, low=11507391, sector=212833896 Nov 24 00:00:10 pog kernel: ide: failed opcode was: unknown Nov 24 00:00:12 pog kernel: hdg: dma_intr: status=0x51 { DriveReady SeekComplete Error } Nov 24 00:00:12 pog kernel: hdg: dma_intr: error=0x01 { AddrMarkNotFound }, LBAsect=212833983, high=12, low=11507391, sector=212833896 Nov 24 00:00:12 pog kernel: ide: failed opcode was: unknown Nov 24 00:00:12 pog kernel: hdg: DMA disabled Nov 24 00:00:12 pog kernel: PDC202XX: Secondary channel reset. Nov 24 00:00:12 pog kernel: ide3: reset: success