On 31/03/07, Wessel, Jason <[EMAIL PROTECTED]> wrote:
uint64_t now = qemu_get_clock(vm_clock); return (uint32_t)((now*3)/125); The optimizer should fix this up with no floating point ops.
Adding to the bikeshed discussion (www.bikeshed.com) this is still not correct as vm_clock is to be used with ticks_per_sec. return muldiv64(qemu_get_clock(vm_clock), 24000000, ticks_per_sec); Regards, Andrzej
