From owner-cvs-sys  Sat Jan 11 10:23:23 1997
Return-Path: <owner-cvs-sys>
Received: (from root@localhost)
          by freefall.freebsd.org (8.8.4/8.8.4) id KAA27452
          for cvs-sys-outgoing; Sat, 11 Jan 1997 10:23:23 -0800 (PST)
Received: (from nate@localhost)
          by freefall.freebsd.org (8.8.4/8.8.4) id KAA27444;
          Sat, 11 Jan 1997 10:23:22 -0800 (PST)
Date: Sat, 11 Jan 1997 10:23:22 -0800 (PST)
From: Nate Williams <nate>
Message-Id: <199701111823.KAA27444@freefall.freebsd.org>
To: CVS-committers, cvs-all, cvs-sys
Subject: cvs commit:  src/sys/pccard driver.h pccard.c pcic.c slot.h
Sender: owner-cvs-sys@FreeBSD.ORG
X-Loop: FreeBSD.org
Precedence: bulk

nate        97/01/11 10:23:22

  Modified:    sys/pccard  driver.h pccard.c pcic.c slot.h
  Log:
  Update the PCIC controller's imask with individual slot IRQ's.
  
  Assuming that the intr_mask[] was updated by changing the maskptrs (the
  existing update_intr_masks() function will not work) this code was
  written so the PCIC controller insertion/removal events will not
  interrupt the card IRQ handler events.
  
  Some possible scenarios:
  + Card is removed during IRQ handler:
  
    - PCIC card handler is allowed to interrupt
    - card removal event is called, removing the driver and data structures
    * card interrupt handler continues w/out driver, data structures, and hardware
  
  OR (the code just committed)
  
    * card IRQ handler has no hardware to read/write to, but has code and
      data to run on (XXX- Assume it completes and doesn't spin forever)
    - PCIC card handler unloads the card driver
  
  The current situation at least leaves the card interrupt handlers the
  drivers and data structures to work with although the hardware can't be
  guaranteed.
  
  Reviewed by:	bde
  
  Revision  Changes    Path
  1.2       +2 -2      src/sys/pccard/driver.h
  1.30      +13 -3     src/sys/pccard/pccard.c
  1.26      +1 -0      src/sys/pccard/pcic.c
  1.9       +2 -1      src/sys/pccard/slot.h