apm problem

[M. Warner Losh]

In message: <20040623135554.V87272 at root.org>
            Nate Lawson <nate at root.org> writes:
: On Fri, 23 Jun 2000, Liam J. Foy wrote:
: > On Wed, 23 Jun 2004 12:40:40 -0700 (PDT)
: > We decided to go with -1. The apm.c patch currently will not apply due to
: > the re-structure of apm. The following patch will:
: >
: > http://liamfoy.kerneled.org/apm.diff
: 
: Ok, good.  That should help Warner.
: 
: > After more digging, apm -l should return 255(stated in man page and acpi spec).
: > The following patch will make it work:
: >
: > --- /usr/src/sys/dev/acpica/acpi_cmbat.c	Tue Jun 22 16:40:35 2004
: > +++ /hd2/acpi_cmbat.c	Tue Jun 22 17:02:18 2004
: > @@ -449,7 +449,7 @@
: >      static int	bat_units = 0;
: >      static struct acpi_cmbat_softc **bat = NULL;
: >
: > -    cap = min = -1;
: > +    cap = min = 255;
: >      batt_stat = ACPI_BATT_STAT_NOT_PRESENT;
: >      error = 0;
: >
: > @@ -545,7 +545,7 @@
: >
: >      /* Battery life */
: >      if (valid_units == 0) {
: > -	cap = -1;
: > +	cap = 255;
: >  	batt_stat = ACPI_BATT_STAT_NOT_PRESENT;
: >      } else {
: >  	cap = total_cap / valid_units;
: > @@ -649,7 +649,7 @@
: >      }
: >
: >      if (!sc->present) {
: > -	battinfo->cap = -1;
: > +	battinfo->cap = 255;
: >  	battinfo->min = -1;
: >  	battinfo->state = ACPI_BATT_STAT_NOT_PRESENT;
: >      } else {
: 
: I disagree with this.  Capacity and time remaining are both quantities
: represented by ints that we calculate.  The man page should be updated if
: it's incorrect.

That's why I've been reluctant to make these changes...

Warner