Description of problem:

I am facing this not so nice message in dmesg

Mar  5 15:00:54 dhcp-lab-198 kernel: PCI-DMA: Out of SW-IOMMU space for 16384
bytes at device 0000:00:1f.2
Mar  5 15:00:54 dhcp-lab-198 kernel: ata1.00: exception Emask 0x0 SAct 0x800000
SErr 0x0 action 0x0
Mar  5 15:00:54 dhcp-lab-198 kernel: ata1.00: cmd
61/00:b8:19:d7:50/04:00:11:00:00/40 tag 23 ncq 524288 out
Mar  5 15:00:54 dhcp-lab-198 kernel:          res
40/00:b4:e9:c0:50/00:00:11:00:00/40 Emask 0x40 (internal error)
Mar  5 15:00:54 dhcp-lab-198 kernel: ata1.00: status: { DRDY }
Mar  5 15:00:54 dhcp-lab-198 kernel: ata1.00: status: { DRDY }
Mar  5 15:00:54 dhcp-lab-198 kernel: ata1.00: configured for UDMA/133
Mar  5 15:00:54 dhcp-lab-198 kernel: ata1: EH complete
Mar  5 15:00:54 dhcp-lab-198 kernel: SCSI device sda: 488281250 512-byte hdwr
sectors (250000 MB)
Mar  5 15:00:54 dhcp-lab-198 kernel: sda: Write Protect is off
Mar  5 15:00:54 dhcp-lab-198 kernel: SCSI device sda: drive cache: write back

while running `yum upgrade` and might be others operations.

I am not sure whether is it problem or not but might looks kinda scary to someone.

Version-Release number of selected component (if applicable):

Linux dhcp-lab-198.englab.brq.redhat.com 2.6.18-83.el5xen #1 SMP Thu Feb 21
12:44:14 EST 2008 i686 i686 i386 GNU/Linux

How reproducible:
when yum is running

Steps to Reproduce:
1. yum upgrade
2. see dmesg
  
Actual results:

weird msg in dmesg, might be some a bug in recent kernel, because it appeared
first in -83 kernel version.

Expected results:
not sure... no msg when it's normal operation, bug fixed when it's kernel side issue

Additional info:

the HDD is

SATA_WDC_WD2500JS_75_WD_WCANKH571533

(^ HAL info)

machine is 

DELL Precision 490