From: Stuart M. <Stu...@st...> - 2000-08-15 23:02:25
|
Folks I've been looking at an odd problem for the last few days, and not getting anywhere fast, so I was just wondering if anybody else has seen something similar. I have an IDE disk attached via a PCI card, which used to work fine. However recently (unfortunatly I don't really know when) I started seeing a problem: hda: lost interrupt followed by some more debugging information which would appear to indicate that the command was a disk write. The problem appears to be unique to writes (mounting the file system read only, everything is fine). However there are some odd features. For example turning off CPU caches appears to also fix the problem, and whether disk is the root file system, or mounted off a root filesystem on a ramdisk also has an effect, but not conclusivly. This is using the 'lastest' kernel from SourceForge (updating this afternoon), but has also been seen in a 2.4.0-test5 system from a few weeks ago. Any suggestions welcome (please!)? Stuart |