From owner-cvs-all  Sun Mar 18 23:10:43 2001
Delivered-To: cvs-all@freebsd.org
Received: from freefall.freebsd.org (freefall.freebsd.org [216.136.204.21])
	by hub.freebsd.org (Postfix) with ESMTP
	id 9F12C37B71A; Sun, 18 Mar 2001 23:10:38 -0800 (PST)
	(envelope-from imp@FreeBSD.org)
Received: (from imp@localhost)
	by freefall.freebsd.org (8.11.1/8.11.1) id f2J7AcE76726;
	Sun, 18 Mar 2001 23:10:38 -0800 (PST)
	(envelope-from imp)
Message-Id: <200103190710.f2J7AcE76726@freefall.freebsd.org>
From: Warner Losh <imp@FreeBSD.org>
Date: Sun, 18 Mar 2001 23:10:38 -0800 (PST)
To: cvs-committers@FreeBSD.org, cvs-all@FreeBSD.org
Subject: cvs commit: src/sys/pccard i82365.h pcic.c slot.h
X-FreeBSD-CVS-Branch: HEAD
Sender: owner-cvs-all@FreeBSD.ORG
Precedence: bulk
X-Loop: FreeBSD.ORG

imp         2001/03/18 23:10:38 PST

  Modified files:
    sys/pccard           i82365.h pcic.c slot.h 
  Log:
  Lots of minor cleanup, plus a couple of interesting things.
  o Attempt to disable the slot when we detect that there are problems with
    it in our ISR.  This should make polling mode work better for more cards,
    but more work may be needed.  This "disabling" sets the card interrupt
    register to 0.  This worked for me for lots of tests in polling mode.
  o Now that I've found datasheets, fix a boatload of magic numbers in the
    source to make it easier to understand.
  o Use a table of names rather than a big case statement.
  o Cull a few of the "unused" controller types that we map to other times
    that were a vestiage of PAO code that we never merged in the same way.
  o Enforce legal IRQs.  You are no longer allowed to try to use IRQs that
    will fail on all known ISA/PCI <-> PCMCIA bridges.  The bridges do not
    have pins for these illegal interrupts, and all of them are listed as
    reserved and/or illegeal in the datasheets depending on which one you
    look at.
  o Add comments about how IBM-AT based computers and NEC PC-98 based computers
    map these interrupts and which ones are valid.
  o Always clear the bit that steers the management interrupt either to the
    value listed in the PCIC_STAT_INT register.  I've seen this bit get set
    on suspend/resume and after windows boot, and it does't hurt to clear it.
    NOTE: this might mean we can share this interrupt in the future.
  
  Revision  Changes    Path
  1.11      +61 -18    src/sys/pccard/i82365.h
  1.104     +78 -76    src/sys/pccard/pcic.c
  1.27      +1 -3      src/sys/pccard/slot.h


To Unsubscribe: send mail to majordomo@FreeBSD.org
with "unsubscribe cvs-all" in the body of the message