lkml.org 
[lkml]   [1997]   [Jan]   [27]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
/
Date
From
SubjectRe: AHA1542 SCSI reset code still broken???

>i also have a 1542cf with the floppy disabled and a single old 400m
>quantum. i can't even get the kernel to find the drive at bootup. i have
>the controller at buson=11,busoff=4,dmaspeed=5, for whatever that is
>worth. the dos diagnostic driver can find and verify the drive just fine,
>which is annoying to no end :) there are no dma/irq conflicts as far as i
>can tell, it just seems that the kernel is really ticking off the 1542..
>more later..

Probably the best way to fix the problem with kernel hanging
is to throw away the 1542 and get a decent scsi card. The main problem with
the 1542 is that it is pretty easy for errors on the scsi bus to cause
the controller to get wedged, and this requires you to cycle power on the
system to bring it back to life. The problem seems to be aggravated when
you have multiple outstanding commands on the same bus.

It appears that as long as none of the devices on the bus are
misbehaving, the 1542 does OK. The performance is mediocre compared
to a PCI board of some kind, but the main limitation is that the
documented commands to abort outstanding commands simply serve to
wedge the controller.

The error handling in the 1542 driver relies upon being able
to send BUS_DEVICE_RESET and BUS_RESET commands to the 1542. Many people
reported that the error handling situation improved when I added the code
to do this, but the fact of the matter is that it is still pretty easy
for the controller to get wedged.

-Eric


\
 
 \ /
  Last update: 2005-03-22 13:38    [W:0.054 / U:0.172 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site