Le dimanche, 27 janvier 2008 03.42:46 Michael Biebl, vous avez écrit : > Didier Raboud wrote: > > (...) > > This is still a bug in hal though. :) > > Depends. A lot of information (on my system) is missing in the new /sys > interface (that's also why I prefer proc atm). Also, what you describe > is simply hal reporting the values provided by the kernel. So it might > actually be a kernel issue after all. > > > Cheers, > Michael
Hi,
I don't think so, and I prove it. Attached to this mail, you'll find a graph
showing the differences between values in /sys/ and reported by hal and the
script I used to generate the data. (battery discharging, AC adapter plugged
in, then battery charging and then # /etc/init.d/hal restart).
So hal is visibly faulty and not the kernel (at least on this specific issue).
---- hal-0.5.10/hald/linux/acpi.c ----
// (...)
/* collect batteries */
acpi_synthesize ("/proc/acpi/battery", ACPI_TYPE_BATTERY, TRUE);
/* collect processors */
acpi_synthesize ("/proc/acpi/processor", ACPI_TYPE_PROCESSOR, TRUE);
/* collect fans */
acpi_synthesize ("/proc/acpi/fan", ACPI_TYPE_FAN, TRUE);
/* collect AC adapters */
acpi_synthesize ("/proc/acpi/ac_adapter", ACPI_TYPE_AC_ADAPTER, TRUE);
// (...)
--------
Regards,
Didier
<<attachment: faulty_hal_2-6-24.png>>
test_hal_sys.sh
Description: Script used to test hal and /sys correspondancy
signature.asc
Description: This is a digitally signed message part.
