Re: port-i386/41624: atheros-card works in 4.0 but not in 5.0

[David Holland <dholland-bugs%netbsd.org@localhost>]

The following reply was made to PR port-i386/41624; it has been noted by GNATS.

From: David Holland <dholland-bugs%netbsd.org@localhost>
To: gnats-bugs%NetBSD.org@localhost
Cc: 
Subject: Re: port-i386/41624: atheros-card works in 4.0 but not in 5.0
Date: Tue, 8 Sep 2015 04:01:12 +0000

 On Mon, Jul 06, 2009 at 11:25:01AM +0000, Nino on NetBSD 4.0 wrote:
  >  Now ath0 DOES work. It was quite interesting, I attribute it to a
  > quirky card (it blue-screened Windows, it said something about
  > "memory parity error", which was definitely incorrect, as I tried
  > several RAM-chips). So not a problem of NetBSD.
  >  
  >  If you want to know details, I am sending you below a) the dmesg
  > while it was NOT working, and b) the dmesg when it WAS working;
  > possibly it can be of help for other cases in the future. What I
  > did to solve it was: reboot in Windows after taking out the
  > wireless-chip, then put in the chip when Windows is running, then
  > reboot Windows with the chip inside, then turn on wireless (using
  > the button on the laptop), then shutdown windows and boot into
  > NetBSD.
 
 Looking at the diffs between those dumps, I wonder if the problem
 is/was that your machine has bad RAM, or some similar problem :(
 
 for example: 
 
   ohci0 at pci0 dev 3 function 0: PCI configuration registers:
     Common header:
 -     0x00: 0x70011039 0x82800017 0x0c03100f 0x00804000
 +     0x00: 0x70011039 0x02800017 0x0c03100f 0x00804000
 
 Not related to the atheros card at all; but it seems like that should
 not be different from one boot to the next. Although I don't know what
 it actually means so maybe I'm jumping to conclusions.
 
 In any event I suspect the interesting part (within the atheros
 section) is this:
 
 -       Parity error detected: on
 +       Parity error detected: off
 
 and that this state caused it to not attach.
 
 However, it was all long enough ago that you've probably junked the
 hardware in the meantime... I hope you found a way to make it work
 reasonably reliably, as opposed to just "sometimes".
 
 I think at this point the PR should be closed...
 
 -- 
 David A. Holland
 dholland%netbsd.org@localhost