Date: Tue, 18 May 1999 13:34:34 +0300 (IDT)
From: Eli Zaretskii <eliz AT is DOT elta DOT co DOT il>
X-Sender: eliz AT is
To: Sean <sproctor AT enter DOT net>
cc: djgpp AT delorie DOT com
Subject: Re: uclock trouble
In-Reply-To: <373ED67A.7C7E261E@enter.net>
Message-ID: <Pine.SUN.3.91.990518133401.24330F-100000@is>
MIME-Version: 1.0
Content-Type: TEXT/PLAIN; charset=US-ASCII
Reply-To: djgpp AT delorie DOT com
X-Mailing-List: djgpp AT delorie DOT com
X-Unsubscribes-To: listserv AT delorie DOT com
Precedence: bulk


On Sun, 16 May 1999, Sean wrote:

> Each time I ran
> it, it would count for a while, then jump down to the same point, then
> count up and continue up through 0 normally, so the first thing I used
> uclock to time would be off by around 70000 uclocks, I think.

65536, to be exact ;-).

This is a peculiarity of Windows 9X, or so it seems; thanks for paying
attention to it.  Please try the version of `uclock' below and tell me
if it works okay; it seems to solve this problem in my machine.

(If you want to know what do I think about the cause for the problem,
read the comments in the code below.)

--------------------------------------------------------------------
/* Copyright (C) 1996 DJ Delorie, see COPYING.DJ for details */
/* Copyright (C) 1995 DJ Delorie, see COPYING.DJ for details */
#include <time.h>
#include <errno.h>
#include <pc.h>
#include <libc/farptrgs.h>
#include <go32.h>
#include <dpmi.h>
#include <libc/bss.h>

static int uclock_bss = -1;

/* tics = about 18.2 * 65536 (1,192,755)

   actually, it's   0x1800b0 tics/day (FAQ)
                  / 24*60*60 sec/day
                  * 65536 utics/tic
		  = 1,193,180 utics/sec */

uclock_t
uclock(void)
{
  static uclock_t base = 0;
  static unsigned long last_tics = 0;
  unsigned char lsb, msb;
  unsigned long tics, otics;
  uclock_t rv;

  if (uclock_bss != __bss_count)
  {
    int e = errno;

    /* switch the timer to mode 2 (rate generator) */
    /* rather than mode 3 (square wave), which doesn't count linearly. */

    outportb(0x43, 0x34);
    outportb(0x40, 0xff);
    outportb(0x40, 0xff);

    base = 0;
    last_tics = 0;
    uclock_bss = __bss_count;

    /* It seems like Windows 9X virtualization of the timer device
       delays the actual execution of the above command until the
       next timer tick.  Or maybe it only consults the actual device
       once per tick.  In any case, the values returned during the
       first 55 msec after the timer was reprogrammed still look as
       if the timer worked in mode 3.  So we simply wait for one clock
       tick when we run on Windows.  */
    _farsetsel(_dos_ds);
    otics = _farnspeekl(0x46c);
    do {
      errno = 0;
      __dpmi_yield();	/* will set errno to ENOSYS on plain DOS */
    } while (errno == 0 && _farnspeekl(0x46c) == otics);
    errno = e;
  }

  /* Make sure the numbers we get are consistent */
  do {
    otics = _farpeekl(_dos_ds, 0x46c);
    outportb(0x43, 0x00);
    lsb = inportb(0x40);
    msb = inportb(0x40);
    tics = _farpeekl(_dos_ds, 0x46c);
  } while (otics != tics);

  /* calculate absolute time */
  msb ^= 0xff;
  lsb ^= 0xff;
  rv = ((uclock_t)tics << 16) | (msb << 8) | lsb;

  if (base == 0L)
    base = rv;

  if (last_tics > tics) /* midnight happened */
    base -= 0x1800b00000LL;

  last_tics = tics;

  /* return relative time */
  return rv - base;
}