Lennart Sorensen wrote: > On Tue, Aug 28, 2007 at 09:54:27PM -0400, Lennart Sorensen wrote: >> Well on a high note, I managed to boot a Miata 433 MHz with 256MB ram, >> after upgrading the firmware to the latest version and switching it to >> SRM. As soon as I get a working HD in it I will get Etch going on it. >> >> Does anyone know if I could move a 600MHz 21164A into the Miata in place >> of the 433MHz 21164?
If the CPUs are soldered to the daughtercards, it'd be quite difficult. If not, it still might be no-go, as the higher speed may need some ancillary chips to be higher speed as well. You could probably exchange daughtercards of different speeds, though... Oh, and the 433MHz is also a 21164A, aka EV56, just slower. > So have I hit the PYXIS DMA crossing page boundary bug here? I found > some patches from years ago for the ide driver to ensure that it never > tries to cross a page boundary when doing DMA to the IDE device, but > looking at current kernel code it certainly doesn't look like that is > something that is part of the standard kernel. Is it normal to have to > patch this in yourself if you want to be able to run IDE with DMA on > this system? Definitely, the 433 is an early MIATA and thus doesn't have all the "Pyxis fixes" that a MIATA-GL (latest) would have, and so could indeedbe susceptible to the DMA bug... :-\ > The reason for going to IDE is that the Qlogic scsi card only wants to > see narrow scsi drives, not wide. Could be the cable, but I don't have > a spare cable to try with yet, and unfortunately the only narrow scsi > drive I have around has lots of bad sectors so I can't really use it. MIATA Qlogic cards were typically 1020-based, which is only narrow, IIRC. If you do have a 1040-based controller, that should do wide, and may depend on the cable and/or termination. Good luck. --Jay++ -- Jay A Estabrook HPTC - XC I & B Hewlett-Packard Company - ZKO1-3/D-B.8 (603) 884-0301 110 Spit Brook Road, Nashua NH 03062 [EMAIL PROTECTED] -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]