I have been running RedHat 6.0 for a number of months. Last week I applied a number of security and recommended patches to my system. I obtained those patches from the RedHat web site. Among those patches was the full suite of Xfree86 3.3.5 rpm's. (I also applied Gnome, CDE, netkit, termcap, rpm, telnet, traceroute patches. I can provide a full detailed list) Starting on the weekend, my system starting logging SCSI errors of the form: Oct 5 04:02:47 willow kernel: SCSI host 0 abort (pid 142691) timed out - resetting Oct 5 04:02:47 willow kernel: SCSI bus is being reset for host 0 channel 0. One symptom for me seems to be that StarOffice has twice now locked up completely. The process is defunct, and not killable. I rebooted once, and ended up having to manually fsck to clean up the disk. The reason I suspect Xfree86 is that I when looking through deja.com and found several posting by Randall J. Parr (one in comp.os.linux.x under the title "bug/conflicts in XFree86 3.3.5 on RH6?" dated Sep 23/1999) He detailed quite well an almost identical situation, only he had done a better job of tracking it down to XFree86. His solution was to downgrade to version 3.3.3. I emailed him and he replied that he has not received a single other response to his several usenet postings. I thought I should try submitting an official bug report, in case this really is something new to you guys. I checked what I could on the redhat support pages and couldn't find anything similar. For completenes, I have a Intel PII-350, ASUS-P2B motherbd, Adaptec 2940UW scsi card, Yamaha crw4416 cd-rw, and seagate ST34520W SCSI hard disk. the other system hardware should be irrelevant, but to cover everything, the video card is an 8MB AGP ati xpert98. Uname -a reports kernel version 2.2.5-22 #1. thank you for your attention.
this is a symptom of faulty hardware, not of bad software. There is NO WAY that XFree86 can cause bad interactions with the SCSI bus. Maybe your hard drive is flakey.