Update on Dells with 5.1.0pre7

[castelli at cscs.ch]

>Dear List Humans,
>
>I've spent the last couple of weeks after I got back from vacation
>testing and installing our Dell poweredge boxes with the new pre7
>revision, and have the following observations:
>
>a) If I try to boot up a Dell system that has never been booted before
>with a boot floppy/kernel containing the pre7 aic7xxx driver, it hangs
>right after the timeout that occurs after downloading the sequencer
>and during the attached disk probe.  The hang is the usual loop of
>annoying Parity errors etc.  This is highly reproducible.
>
>b) If I boot one of the Dell systems and let Windows NT (preinstalled
>on the hard drive) get through its initial pre-installation routine,
>during which it (presumably) installs its own aic7xxx drivers and the
>like, and THEN boot from the SAME boot floppy, it boots "perfectly"
>except for the single timeout during the attached device probe
>previously reported on the list.  Thereafter, it appears to work
>reasonably well -- I have now installed a dozen or so of the dual P2's
>with 7890U2W and 7860 U onboard adapters without (yet, knock on wood)
>any problems.
>
It work's also with WIN 98 on a asus PS2/B (aic7890 on board), with pentium
400MHz and a 9.1Gb scsi-3 HD, 32xCD-ROM (EIDE). As linux operating system I
work with S.u.S.E. 5.2 with a aic7xxx pre7 driver.

>c) I'm a bit scared to power cycle these systems.  It seems to me that
>there is likely some initialization code or steps missing from the
>existing sequencer download and initialization process -- NT puts the
>adapter in a state from which the linux driver works, but it doesn't
>work from a powerup default state.  This is not in real disagreement
>with my earlier observations of pre-[3-5] drivers that worked for
>while and then "went away" when I unplugged the systems and moved
>them.  Of course, after I install linux I can no longer boot up NT to
>get them "sane" again.
>
>Now, I don't have any idea what is going on or how to contribute
>further to the debugging effort.  I have one system in reserve that
>has never booted up NT from the disk, so I can test any possible fixes
>that come up, if they come up fairly soon (I do need the system in
>production before too long).  I will also test with one of my
>existing systems whether or not a fully installed system still goes
>away when power cycled.
>
>Overall, the pre7 driver is pretty good -- the installed systems boot
>locally and appear to run stably.  The powerup/initialization issue,
>however, looms large and still unresolved.  If I were the paranoid
>type, I'd mutter darkly about the Dell/Adaptec/Microsoft conspiracy
>and some deliberate tweak of the Dell Bios that forces one to use NT,
>but if this were true I'd have to kill somebody, so I'll continue to
>think happy, flowery thoughts while awaiting the addition of some
>magic initialization fix.
>
>BTW, I haven't tried a freebsd boot disk against this problem, but I
>do believe that I have an old one lying around that did indeed boot up
>and find the 7890 AFTER an early NT boot.  I may try to boot my
>un-NT-booted box with this to see if the problem is common to both, as
>it may well be.
>
>   rgb
>
>Robert G. Brown	                       http://www.phy.duke.edu/~rgb/
>Duke University Dept. of Physics, Box 90305
>Durham, N.C. 27708-0305
>Phone: 1-919-660-2567  Fax: 919-660-2525     email:rgb at phy.duke.edu
>
>
>
>
>To Unsubscribe: send mail to majordomo at FreeBSD.org
>with "unsubscribe freebsd-aic7xxx" in the body of the message


--------------------------------------------
Christian Castelli
Laboratory of Cellular Pathology
Regional Institute of Pathology
24, in selva road
ch-6601 Locarno (switzerland)

tel ++41 (0)91 756 2681
fax ++41 (0)91 756 2646
--------------------------------------------





To Unsubscribe: send mail to majordomo at FreeBSD.org
with "unsubscribe freebsd-aic7xxx" in the body of the message