kernel errors in log/messages

[Tim Wunder]

Using FC3 (kernel 2.6.9-1.681_FC3) with XFS filesystem on an old K6-500 system 
with a fairly new Seagate 80 GB HDD. 
I'm trying to copy a couple websites over from one server to another using 
tar, and when untar'ing the file, I keep getting errors like this:

tar: pn/html/parameters/Sports/modules/NS-Banners/lang/eng: Cannot mkdir: 
Input/output error

/var/log/messages has the following:
Dec 18 10:19:04 jeeves kernel: hda: dma_intr: status=0x51 { DriveReady 
SeekComplete Error }
Dec 18 10:19:04 jeeves kernel: hda: dma_intr: error=0x40 
{ UncorrectableError }, LBAsect=96175142, high=5, low=12289062
, sector=96175142
Dec 18 10:19:04 jeeves kernel: ide: failed opcode was: unknown
Dec 18 10:19:04 jeeves kernel: end_request: I/O error, dev hda, sector 
96175142
Dec 18 10:19:04 jeeves kernel: I/O error in filesystem ("hda8") meta-data dev 
hda8 block 0x49b7c10       ("xfs_trans_re
ad_buf") error 5 buf count 4096
Dec 18 10:19:08 jeeves kernel: hda: dma_intr: status=0x51 { DriveReady 
SeekComplete Error }
Dec 18 10:19:08 jeeves kernel: hda: dma_intr: error=0x40 
{ UncorrectableError }, LBAsect=70408902, high=4, low=3300038,
 sector=70408902
Dec 18 10:19:08 jeeves kernel: ide: failed opcode was: unknown
Dec 18 10:19:08 jeeves kernel: end_request: I/O error, dev hda, sector 
70408902
Dec 18 10:19:08 jeeves kernel: I/O error in filesystem ("hda8") meta-data dev 
hda8 block 0x31252b2       ("xfs_trans_re
ad_buf") error 5 buf count 512
Dec 18 10:19:12 jeeves kernel: hda: dma_intr: status=0x51 { DriveReady 
SeekComplete Error }
Dec 18 10:19:12 jeeves kernel: hda: dma_intr: error=0x40 
{ UncorrectableError }, LBAsect=96175142, high=5, low=12289062
, sector=96175142
Dec 18 10:19:12 jeeves kernel: ide: failed opcode was: unknown
Dec 18 10:19:12 jeeves kernel: end_request: I/O error, dev hda, sector 
96175142
Dec 18 10:19:12 jeeves kernel: I/O error in filesystem ("hda8") meta-data dev 
hda8 block 0x49b7c10       ("xfs_trans_re
ad_buf") error 5 buf count 4096
Dec 18 10:19:16 jeeves kernel: hda: dma_intr: status=0x51 { DriveReady 
SeekComplete Error }

FWIW:
# hdparm /dev/hda

/dev/hda:
 multcount    = 16 (on)
 IO_support   =  1 (32-bit)
 unmaskirq    =  1 (on)
 using_dma    =  1 (on)
 keepsettings =  0 (off)
 readonly     =  0 (off)
 readahead    = 256 (on)
 geometry     = 16383/255/63, sectors = 80026361856, start = 0

# hdparm -i /dev/hda

/dev/hda:

 Model=ST380013A, FwRev=3.06, SerialNo=5JV8TL2L
 Config={ HardSect NotMFM HdSw>15uSec Fixed DTR>10Mbs RotSpdTol>.5% }
 RawCHS=16383/16/63, TrkSize=0, SectSize=0, ECCbytes=4
 BuffType=unknown, BuffSize=8192kB, MaxMultSect=16, MultSect=16
 CurCHS=4047/16/255, CurSects=16511760, LBA=yes, LBAsects=156301488
 IORDY=on/off, tPIO={min:240,w/IORDY:120}, tDMA={min:120,rec:120}
 PIO modes:  pio0 pio1 pio2 pio3 pio4
 DMA modes:  mdma0 mdma1 mdma2
 UDMA modes: udma0 udma1 udma2 udma3 *udma4 udma5
 AdvancedPM=no WriteCache=enabled
 Drive conforms to: ATA/ATAPI-6 T13 1410D revision 2:

 * signifies the current active mode

Any idea what's going on?
Thanks, 
Tim

-- 
Fedora Core 2, Kernel 2.6.9-1.6_FC2,  KDE 3.3.2, Xorg 6.7.0
 10:20:00 up 9 days, 16:43,  4 users,  load average: 0.40, 0.42, 0.37
It's what you learn after you know it all that counts