On 08/09/2011 11:28 AM, limp wrote:

Hi all,

I am having a deeper look on Linux hpet code and I was wondering how exactly is_hpet_capable() works.

...

From the name of it, I assume that this function checks if the system supports HPET or not but I can't really tell how it does that.

The function is the following:

static inline int is_hpet_capable(void) { return !boot_hpet_disable && hpet_address; }

Could anyone help me understand how this works?

Regards.

_______________________________________________ Kernelnewbies mailing list Kernelnewbies@kernelnewbies.org http://lists.kernelnewbies.org/mailman/listinfo/kernelnewbies boot_hpet_disable is a boot time param, which as the name suggests disables the hpet timer(nohpet is the actual parameter given to the kernel by the bootloader I think)...

hpet_address is the address where the hpet timers are mapped, which is read by the kernel from the BIOS ACPI table for hpet I think... If you look at the code that does the actual initialization for hpet, it's a bit more complicated than that, but generally, this function checks whether HPET is disabled at boot time, and if a suitable entry / address for HPET timers is found in the BIOS ACPI tables (if it's not found, the hpet_address will be NULL). So if hpet is enabled, and an address is found, this function returns 1/true. Else 0/false. -- Stratos Psomadakis <s.psomadakis@gmail.com>