ACPI/PMU procfs HAL test program. Version 002.

[Richard Hughes]

Okay, version 002 of the hal acpi/pmu test program is now here:

http://hughsie.no-ip.com/write/hal/procfs-hal-002/

It bears little resemblance to my initial version:
 * it's lots simpler! No really, lots.
 * used -pedantic to please older GCC's
 * seporated acpi and generic stuff so PMU is happy.
 * renamed most of the functions to be more descriptive
 * added typical usage to each function to help people debug
 * changed to "procfs_generic_device" style to minimise duplication
 * added "procfs_refresh_acpi_ac_adapter" and friends - 
   this lets us refresh the acpi entries really easy from /proc/*
   pmu isn't so simple, for some things it's okay, but others will
   need an addon to read /dev/pmu.

Currently to search all the entries:
time ./procfs-hal > /dev/null
real    0m0.004s
user    0m0.000s
sys     0m0.004s

If you run it, it will use /proc/acpi/* (or /proc/pmu) to display 
(on console) something like the following:

[7]set_string   linux.powersystem,      acpi
[7]set_string   linux.acpi.version,     20041210
* new_device *
[11]capability  system.ac_adapter
[11]set_int     system.ac_adapter.number,       0
[11]set_string  info.bus,       unknown
[11]set_string  info.udi,       /org/freedesktop/Hal/devices/ac_adapter_0
[11]capability  linux.acpi
[11]set_string  linux.acpi.procfs_path, /proc/acpi/ac_adapter/ADP1
[11]set_string  info.product,   AC Adaptor
[11]set_bool    system.acadaptor.present,       TRUE
* callout_device *
* new_device *
[12]capability  system.fan
[12]set_int     system.fan.number,      0
[12]set_string  info.bus,       unknown
[12]set_string  info.udi,       /org/freedesktop/Hal/devices/fan_0
[12]capability  linux.acpi
[12]set_string  linux.acpi.procfs_path, /proc/acpi/fan/FAN
[12]set_string  info.product,   Fan
[12]set_bool    system.fan.enabled,     FALSE
* callout_device *
* new_device *
[13]capability  system.processor
[13]set_int     system.processor.number,        0
[13]set_string  info.bus,       unknown
[13]set_string  info.udi,       /org/freedesktop/Hal/devices/processor_0
[13]capability  linux.acpi
[13]set_string  linux.acpi.procfs_path, /proc/acpi/processor/CPU0
[13]set_string  info.product,   Processor
[13]set_bool    system.processor.can_throttle,  FALSE
* callout_device *
[7]set_string   linux.acpi.toshiba.version,     0.18

Could you please comment on the code, compile it, run it, before I send any 
patches for HEAD. It's a lot simpler to change and debug a 10k .c file than 
the whole of haldaemon.

Comments about 001 from Sjoerd:
> * You shouldn't need to define maximum lenght stuff when glib

Couldn't find a way round this. Still, string checking is at least safe
from overflow.

> * getprocfsvalue opens and searches the given file each time to get one
> value, better to parse it into a hash once and pull value's from that..

Didn't do this as for the sheer amount of complexity it adds. procfs
stays in memory, so all accesses are really quick.

Richard

-- 

http://www.hughsie.com/PUBLIC-KEY


_______________________________________________
hal mailing list
hal at lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/hal