ACPI battery issues on Dell C840 - OK in -test11-mm1, broken in 2.6.0-mm1.

[Valdis . Kletnieks]

(acpi-devel readers please cc: me, am only on lkml list)

Summary: Battery support was working ok for me in -test11-mm1 and earlier,
but is broken in 2.6.0-mm1 and later.  I suspect 2.6.0-mm1's inclusion of 
acpi-20031203.patch, but I haven't found where in the 154K of patch it goes
astray yet.

Symptoms:

At boot, the batteries are found:

Jan  5 14:53:12 turing-police kernel: ACPI: Battery Slot [BAT0] (battery present)
Jan  5 14:53:12 turing-police kernel: ACPI: Battery Slot [BAT1] (battery present)

and files created:

% ls -lR /proc/acpi/battery
/proc/acpi/battery:
total 0
dr-xr-xr-x  2 root root 0 Jan  5 18:33 BAT0
dr-xr-xr-x  2 root root 0 Jan  5 18:33 BAT1

/proc/acpi/battery/BAT0:
total 0
-rw-r--r--  1 root root 0 Jan  5 18:33 alarm
-r--r--r--  1 root root 0 Jan  5 18:33 info
-r--r--r--  1 root root 0 Jan  5 18:33 state

/proc/acpi/battery/BAT1:
total 0
-rw-r--r--  1 root root 0 Jan  5 18:33 alarm
-r--r--r--  1 root root 0 Jan  5 18:33 info
-r--r--r--  1 root root 0 Jan  5 18:33 state

but the contents are broken:

%  cat /proc/acpi/battery/BAT0/info
present:                 yes
design capacity:         0 mWh
last full capacity:      0 mWh
battery technology:      non-rechargeable
design voltage:          0 mV
design capacity warning: 0 mWh
design capacity low:     0 mWh
capacity granularity 1:  0 mWh
capacity granularity 2:  0 mWh
model number:            
serial number:           
battery type:            
OEM info:                
% cat /proc/acpi/battery/BAT0/state
present:                 yes
capacity state:          ok
charging state:          unknown
present rate:            0 mA
remaining capacity:      0 mAh
present voltage:         0 mV

under -test11-mm1, we had in 'info':
present:                 yes
design capacity:         66000 mWh
last full capacity:      62400 mWh
battery technology:      rechargeable
design voltage:          14800 mV
design capacity warning: 3000 mWh
design capacity low:     1000 mWh
capacity granularity 1:  200 mWh
capacity granularity 2:  200 mWh
model number:            0004M778
serial number:           1084
battery type:            LION
OEM info:                SANYO

and 'state':
present:                 yes
capacity state:          ok
charging state:          unknown
present rate:            unknown
remaining capacity:      66000 mWh
present voltage:         16682 mV

Even while booted on 2.6.1-rc1-mm2, dmidecode 2.2 can find the info:

Handle 0x1600
        DMI type 22, 26 bytes.
        Portable Battery
                Location: Left Module Bay 
                Manufacturer: SANYO           
                Name: 0004M778        
                Design Capacity: 66000 mWh
                Design Voltage: 14800 mV
                SBDS Version: 1.0
                Maximum Error: 4%
                SBDS Serial Number: 043C
                SBDS Manufacture Date: 2002-04-12
                SBDS Chemistry: LION            
                OEM-specific Information: 0x00000001
Handle 0x1601
        DMI type 22, 26 bytes.
        Portable Battery
                Location: Right Module Bay
                Manufacturer: SANYO           
                Name: 0001J433        
                Design Capacity: 66000 mWh
                Design Voltage: 14800 mV
                SBDS Version: 1.0
                Maximum Error: 4%
                SBDS Serial Number: 014F
                SBDS Manufacture Date: 2002-03-26
                SBDS Chemistry: LION            
                OEM-specific Information: 0x00000001

So my acpi can't be TOO insane, right?

I'd be happy to debug further - the kernel(s) in question already have
CONFIG_ACPI_DEBUG built in, but I'm at a loss as to what to echo
onto the /proc/acpi/debug_layer and debug_level files to actually toggle
the printk's to something useful.

Any comments/hints/ideas/patches? ;)

Attachment: pgp00000.pgp
Description: PGP signature