Andrew Halliwell
And verily, didst Peter Dalgaard announce to the hordes:
Argh. I thought this was fixed but apparently not. The same issue existed with 9.0 but was solved with the Patch CD for that release. No such luck with the 9.1 DVD it seems.
As I understand the issue, there are two drivers that claim to deal with the board (sil3112 based): siimage and sata_sil. Of these, the former is broken - the symptom is a "lost interrupt" message and disk access slower than PPP over a 300baud modem... Looks like the cause of the bug is well known, but nevertheless not fixed in current kernels. The sata_sil module which builds on libata is claimed to work perfectly. The problem is that the siimage driver loads unconditionally and claims the device so that the working driver never gets a chance.
Is it possible that we could have a Patch CD for 9.1 too? Or is there a trick that I have overlooked?
I've seen lost interrupt messages before that've been cured by simply dropping to a console and...
hdparm -d1 -u1 /dev/hdX (where X is the ide drive in question) Worth a shot, at least.
I can try, but I don't think you realize just how long it takes to get to the point where there is a console to drop to... I might be able to build a custom 2.6 kernel on the 9.0 system. Anyone have the recipe for using that for installation? -- O__ ---- Peter Dalgaard Blegdamsvej 3 c/ /'_ --- Dept. of Biostatistics 2200 Cph. N (*) \(*) -- University of Copenhagen Denmark Ph: (+45) 35327918 ~~~~~~~~~~ - (p.dalgaard@biostat.ku.dk) FAX: (+45) 35327907