From: Michel <mi...@da...> - 2002-07-24 12:33:47
|
On Wed, 2002-07-24 at 14:27, Rene Brothuhn wrote: > On 2002.07.24 12:33 Ken Tyler wrote: > >=20 > > I came across a pait of 4.3 Gig UW SCSI drives at a give away price the > > other day. I thought I might grab them and have ago. > >=20 > > What's the status of the driver ? > >=20 > > I know it doesn't work and that Fred H (possibly someone else too ) hav= e > > tried modding existing drivers. > >=20 > > Which driver almost worked (can't remember the details - some sort of > > cache coherency problem ?). >=20 >=20 > Hi, Ken! >=20 > I'm working on the driver. But I have to give up my work for several=20 > month, due to lack of time. But for some weeks, I got a little time back=20 > and contiune my work on the driver. >=20 > So, for now it seems that all the cache coherency problems are solved. An= d=20 > the communication between the lowlevel and midlevel SCSI driver thru DMA=20 > seems working correctly now (which wasn't since i stopped my work). The=20 > driver detects the devices connected to the SCSI bus and prints out the=20 > names, revision numbers and so on. Great to hear you're making progress! Can you please commit what you have, so others can look at it? > After that, when the driver tries to reselect a device (for reading out=20 > the capacity...), I got an illigal instruction exception from the SCRIPTS= .=20 > But there is no illegal instruction. The documentation of the 53c770 says= ,=20 > that illigal instruction exception also occur on some illegal phases on=20 > the SCSI bus. I don't know much about SCSI bus arbitration, but I'm=20 > learning... I don't really know anything about this, but if the CPU hits an illegal instruction, there could still be coherency issues? --=20 Earthling Michel D=E4nzer (MrCooper)/ Debian GNU/Linux (powerpc) developer XFree86 and DRI project member / CS student, Free Software enthusiast |