From owner-freebsd-commit  Mon Dec 25 17:21:56 1995
Return-Path: owner-commit
Received: (from root@localhost)
          by freefall.freebsd.org (8.7.3/8.7.3) id RAA10471
          for freebsd-commit-outgoing; Mon, 25 Dec 1995 17:21:56 -0800 (PST)
Received: (from root@localhost)
          by freefall.freebsd.org (8.7.3/8.7.3) id RAA10457
          for cvs-all-outgoing; Mon, 25 Dec 1995 17:21:48 -0800 (PST)
Received: (from root@localhost)
          by freefall.freebsd.org (8.7.3/8.7.3) id RAA10447
          for cvs-sys-outgoing; Mon, 25 Dec 1995 17:21:42 -0800 (PST)
Received: (from bde@localhost)
          by freefall.freebsd.org (8.7.3/8.7.3) id RAA10440
          Mon, 25 Dec 1995 17:21:40 -0800 (PST)
Date: Mon, 25 Dec 1995 17:21:40 -0800 (PST)
From: Bruce Evans <bde>
Message-Id: <199512260121.RAA10440@freefall.freebsd.org>
To: CVS-committers, cvs-sys
Subject: cvs commit: src/sys/kern subr_prof.c
Sender: owner-commit@FreeBSD.ORG
Precedence: bulk

bde         95/12/25 17:21:40

  Modified:    sys/kern  subr_prof.c
  Log:
  Unstaticized addupc_task().  It is supposed to be called from trap().
  See the comments for addupc_intr() and the NetBSD implementation.
  We use dummy versions of fuswintr() and susiwintr(), so addupc_intr()
  always pushes the work to trap() (this is inefficient), and trap()
  calls the special i386 function addupc() instead of addupc_task().
  addupc() is more efficient than addupc_intr(), so some of the lost
  efficiency is recovered.  However, addupc() may be broken on plain
  i386's since it doesn't check for write permission like copyout().
  
  Revision  Changes    Path
  1.15      +2 -2      src/sys/kern/subr_prof.c