Subject: Change to cons.c for device cloning should be backed out?
To: None <port-pmax@NetBSD.ORG>
From: Jonathan Stone <jonathan@DSG.Stanford.EDU>
List: port-pmax
Date: 12/26/1994 19:38:07
Since the change in interface to device open routines has been undone,
the relevant change in src/sys/arch/pmax/pmax/cons.c should also be
undone. Simply reverting to the previous revision is sufficient:

I wonder if it's possible to eliminate much of the pmax-specific
cons.c altogether, and src/sys/dev/cons.c used instead.  It seems that
the pmax console driver never gets changed when corresponding files
elsewhere get changed...


((No, this  _isn't_ an accidentally reversed patch))

*** cons.c.OLD-CURRENT	Wed Nov 23 04:09:29 1994
--- cons.c	Mon Nov 21 14:40:55 1994
***************
*** 1,4 ****
! /*	$NetBSD: cons.c,v 1.5 1994/11/22 18:59:57 dean Exp $	*/
  
  /*
   * Copyright (c) 1988 University of Utah.
--- 1,4 ----
! /*	$NetBSD: cons.c,v 1.4 1994/10/26 21:10:18 cgd Exp $	*/
  
  /*
   * Copyright (c) 1988 University of Utah.
***************
*** 72,87 ****
  	(struct tty *)0,
  };
  
! cnopen(dev, flag, mode, p, fp)
  	dev_t dev;
  	int flag, mode;
  	struct proc *p;
- 	struct file *fp;
  {
  	if (cn_tab.cn_dev == NODEV)
  		return (0);
  	dev = cn_tab.cn_dev;
! 	return ((*cdevsw[major(dev)].d_open)(dev, flag, mode, p, fp));
  }
   
  cnclose(dev, flag, mode, p)
--- 72,86 ----
  	(struct tty *)0,
  };
  
! cnopen(dev, flag, mode, p)
  	dev_t dev;
  	int flag, mode;
  	struct proc *p;
  {
  	if (cn_tab.cn_dev == NODEV)
  		return (0);
  	dev = cn_tab.cn_dev;
! 	return ((*cdevsw[major(dev)].d_open)(dev, flag, mode, p));
  }
   
  cnclose(dev, flag, mode, p)