Re: [RE]Promise FastTrak66 - dmraid doesn't recognize striped array on pair of Maxto
Woody Suwalski <[email protected]>
| Newsgroups | gmane.linux.ataraid |
|---|---|
| Message-ID | <[email protected]> |
Bob Stewart wrote: > --- James Olson <[email protected]> wrote: > Hi Bob, > I ran into a similar problem with the same controller on certain older drives. > The problem has to do with the way the fasttrak bios sees the drive size and > geometry as opposed to the way linux sees it as reported by the old drive's > firmware. Normally the drive metadata sector is located on the last cylinder > of the drive, about 63 sectors before the end. Anyway you can probably get it > to work if you look through the data on your drive to find out what sector the > metadata is on, and then modify the dmraid source (pdc.c?) to add it to the > list of sector offsets from the end of the disk to try. > -James > > Dmraiders, > > This was actually pretty easy to do. May I suggest the following change to > pdc.h? Seems like there should be some rhyme or reason for the offset, though. > Or is it just dependent on the model and revision of the controller? This is > a "hacked" Utra66 controller with BIOS V1.30. > > OLD: > #define PDC_CONFIGOFFSETS 63,255,256,16,399,735 > NEW: > #define PDC_CONFIGOFFSETS 63,255,256,16,399,735,675 > > Bob > > Actually we have found yet another PDC location=991: #define PDC_CONFIGOFFSETS 63,255,256,16,399,735,675,991 I think I have sent Heinz a patch for this, however it somehow got lost? Anyway, thanks for the 675... Woody, Xandros