Bug#462229: Problem not present in older kernel.

Rolled back to 2.6.18-4-amd64, and I'm not seeing this problem. Will try to test again with 2.6.18-5-amd64 when I have the opportunity (server is in production).

Bug#462229: I'm currently experiencing the same problem.

Debian server is a VMware server host. Each time a VM experiences moderate disk activity I typically see the following messages: kernel: ata1: EH complete kernel: ata1.00: exception Emask 0x0 SAct 0x0 SErr 0x0 action 0x0 kernel: ata1.00: (BMDMA stat 0x20) kernel: ata1.00: tag 0 cmd 0x25 Emas