attached is a patch which:
1. adds an __ATTR_N macro to device.h,
which takes an extra _index arg, and initializes device_attibute's new
.index member/
2. replaces lots of nested-macro-uses of these:
static DEVICE_ATTR(temp##offset##_input, S_IRUGO, \
show_temp##offset##_input, NULL); \
with array initializers. like these:
static struct device_attribute dev_attr_temp_input[] =
{
__ATTR_N(temp1_input, S_IRUGO, show_temp_input, NULL, 0),
__ATTR_N(temp2_input, S_IRUGO, show_temp_input, NULL, 1),
__ATTR_N(temp3_input, S_IRUGO, show_temp_input, NULL, 2),
__ATTR_N(temp4_input, S_IRUGO, show_temp_input, NULL, 3),
__ATTR_N(temp5_input, S_IRUGO, show_temp_input, NULL, 4),
__ATTR_N(temp6_input, S_IRUGO, show_temp_input, NULL, 5),
};
3. de-macros the functions called by the individual items.
the functions use devattr->index to determine which item theyre working on.
this is an interim version, and has ifdefs to support back and forth
debugging:
grep define pc87360.c |grep ATTR_LOOP
#define FAN_ATTR_LOOP 0
#define PWM_ATTR_LOOP 0
#define VOLT_ATTR_LOOP 1
#define TEMP_ATTR_LOOP 0
using loops results in a large savings in obj code.
baseline: -rw-rw-r-- 1 jimc jimc 168326 Jul 27 23:10
drivers/hwmon/pc87360.ko
VOLT_ATTR_LOOP -rw-rw-r-- 1 jimc jimc 135940 Jul 27 23:11
drivers/hwmon/pc87360.ko
+ TEMP_ATTR_LOOP -rw-rw-r-- 1 jimc jimc 127441 Jul 27 23:16
drivers/hwmon/pc87360.ko
+ FAN/PWM_ATTR_LOOP -rw-rw-r-- 1 jimc jimc 121062 Jul 27 23:55
drivers/hwmon/pc87360.ko
I have a couple of questions that Im hoping you can clarify for me;
its late, and I could sleep now, and see your responses when I wake.
line 887 starts the show_and_set_temp(offset) macro,
which looks to be a duplicate of
#define show_and_set_therm(offset) on line 651,
or if theyre not duplicates (I havent looked carefully), its an
unfortunate naming,
creating functions which are disambiguated only by the offset.
OTOH, they could be different cuz the physical units on the chip have
different capabilities.
ATM code has some unused data, but probly not another 10k
Using loops to call device_create_file also simplifies some of the logic
wrt data->innr, tempnr, fannr;
the if-variations are hidden/abstracted by 0..configd_max. Of course,
I could have oversimplified.
Ive done cursory testing: new code does this.
root at soekris:~# sensors
pc87366-isa-6620
Adapter: ISA adapter
avi0: +3.01 V (min = +0.00 V, max = +3.01 V)
VCORE: +2.02 V (min = +0.00 V, max = +3.01 V)
VCC: +5.01 V (min = +0.00 V, max = +6.03 V)
VPWR: +14.01 V (min = +0.00 V, max = +60.56 V)
+12V: +11.91 V (min = +0.00 V, max = +14.46 V)
-12V: -12.28 V (min = -60.61 V, max = -2.76 V)
GND: +0.00 V (min = +0.00 V, max = +3.01 V)
Vsb: +3.33 V (min = +0.00 V, max = +6.03 V)
Vdd: +2.91 V (min = +0.00 V, max = +6.03 V)
Vbat: +3.01 V (min = +0.00 V, max = +3.01 V)
AVdd: +3.31 V (min = +0.00 V, max = +6.03 V)
Temp: +107 C (low = -55 C, high = +127 C)
Critical: +127 C
root at soekris:~#
root at soekris:~# cd /sys/bus/i2c/devices/0-6620/
alarms_temp in2_max in6_max temp1_crit temp4_input
bus in2_min in6_min temp1_input temp4_max
driver in2_status in6_status temp1_max temp4_min
in0_input in3_input in7_input temp1_min temp4_status
in0_max in3_max in7_max temp1_status temp5_crit
in0_min in3_min in7_min temp2_crit temp5_input
in0_status in3_status in7_status temp2_input temp5_max
in10_input in4_input in8_input temp2_max temp5_min
in10_max in4_max in8_max temp2_min temp5_status
in10_min in4_min in8_min temp2_status temp6_crit
in10_status in4_status in8_status temp3_crit temp6_input
in1_input in5_input in9_input temp3_input temp6_max
in1_max in5_max in9_max temp3_max temp6_min
in1_min in5_min in9_min temp3_min temp6_status
in1_status in5_status in9_status temp3_status
in2_input in6_input name temp4_crit
I have no fans on this board, and have never gotten temp readings I trust.
But the voltages look right :-)
thanks
jimc
-------------- next part --------------
An embedded and charset-unspecified text was scrubbed...
Name: diff.linux.20050728.001755
Url: http://lists.lm-sensors.org/pipermail/lm-sensors/attachments/20050728/c827e00f/attachment.pl
