Re: [2.624-rc1 regression] lost battery information

[Alexey Starikovskiy <aystarik@gmail.com>]

[-- Attachment #1: Type: text/plain, Size: 2159 bytes --]

Andrey Borzenkov wrote:
> On Friday 26 October 2007, Alexey Starikovskiy wrote:
>> Andrey Borzenkov wrote:
>>> On Friday 26 October 2007, Alexey Starikovskiy wrote:
>>>> Andrey Borzenkov wrote:
>>>>> I have lost battery in 2.6.24-rc1. Without CONFIG_ACPI_PROCFS I have
>>>>> no /proc/acpi/battery and cannot test netlink interface because right
>>>>> now there is no consumer of this.
>>>> for /sysfs interface you need to enable power_supply driver.
>>>> or you could apply this patch and power_supply will be selected by
>>>> battery itself.
>>> I already have power_supply module, battery depends on it and it is
>>> autloaded. But I fail to see where I can get battery info in /sys
>> Intent was to put into /sysfs same information:
>> aystarik@samsung:~/client_conf$ ls /sys/class/power_supply/BAT1/
>> alarm  charge_full  charge_full_design  charge_now  current_now  device 
>> manufacturer  model_name  present  status  subsystem  technology  type 
>> uevent  voltage_min_design  voltage_now
>>
> 
> is it in -rc1 or can you point me to the patch (I'd rather avoid having to 
> pull from different git trees). Thank you.
No, it should be rc1.
> 
> And what about ACPI_PROCFS case? It still needs attention I believe.
As you can see, there is info in /proc too:
>> aystarik@samsung:~/client_conf$ cat /proc/acpi/battery/BAT1/*
>> alarm:                   unsupported
>> present:                 yes
>> design capacity:         4300 mAh
>> last full capacity:      4172 mAh
>> battery technology:      rechargeable
>> design voltage:          14800 mV
>> design capacity warning: 208 mAh
>> design capacity low:     41 mAh
>> capacity granularity 1:  41 mAh
>> capacity granularity 2:  41 mAh
>> model number:            Bat1
>> serial number:           001
>> battery type:            LION
>> OEM info:                Pacifi
>> present:                 yes
>> capacity state:          ok
>> charging state:          charged
>> present rate:            0 mA
>> remaining capacity:      4172 mAh
>> present voltage:         16518 mV

Your cat's "Bad address" means -EFAULT, according to "man errno".
Please apply this patch to see what exactly failed...


[-- Attachment #2: x.patch --]
[-- Type: text/x-diff, Size: 1372 bytes --]

diff --git a/drivers/acpi/battery.c b/drivers/acpi/battery.c
index 02a396d..55e9a8e 100644
--- a/drivers/acpi/battery.c
+++ b/drivers/acpi/battery.c
@@ -262,21 +262,29 @@ static int extract_package(struct acpi_battery *battery,
 {
 	int i, *x;
 	union acpi_object *element;
-	if (package->type != ACPI_TYPE_PACKAGE)
+	if (package->type != ACPI_TYPE_PACKAGE) {
+printk (KERN_ERR PREFIX "package->type = %x\n", package->type);
 		return -EFAULT;
+	}
 	for (i = 0; i < num; ++i) {
-		if (package->package.count <= i)
+		if (package->package.count <= i) {
+printk (KERN_ERR PREFIX "package.count = %d, i = %d\n", package->package.count, i);
 			return -EFAULT;
+		}
 		element = &package->package.elements[i];
 		if (offsets[i].mode) {
 			if (element->type != ACPI_TYPE_STRING &&
-			    element->type != ACPI_TYPE_BUFFER)
+			    element->type != ACPI_TYPE_BUFFER) {
+printk (KERN_ERR PREFIX "element[%d]->type = %x, expected string\n", i, element->type);
 				return -EFAULT;
+			}
 			strncpy((u8 *)battery + offsets[i].offset,
 				element->string.pointer, 32);
 		} else {
-			if (element->type != ACPI_TYPE_INTEGER)
+			if (element->type != ACPI_TYPE_INTEGER) {
+printk (KERN_ERR PREFIX "element[%d]->type = %x, expected integer\n", i, element->type);
 				return -EFAULT;
+			}
 			x = (int *)((u8 *)battery + offsets[i].offset);
 			*x = element->integer.value;
 		}