The daily build of the Etch installer 20070227(?), and the Etch Installer RC1, both tickle the problem reported in Debian bug 333831. Despite discussions with upstream, which suggested this issue might have been fixed in 2.6.18, the 2.6.18-4 kernel in the Debian-Installer tickles this problem on the Acer Travelmate 200DX. Simplest workaround is to boot with the kernel option "acpi=off". In the Debian Installer this means typing; linux acpi=off Indicators of this problem include various stack dumps in the kernel output during boot, and; ide-cd: cmd 0x5a timed out hdc: lost interrupt Obviously some hardware will have other device names (hdd etc). I have looked at hacking the ACPI assembly code to fix the issue, but even if this worked the resulting assembler would be copyright of Acer (or their software suppliers) and impossible to validate (i.e. might fry peoples hardware), and upstream seem less than 100% certain that this is a problem with the ACPI code on the laptop.
Attachment:
signature.asc
Description: OpenPGP digital signature