Very Computer

> Hi!
> Don't ask me why, but i want to disable and stop the system time.
> Every tick from the quartz the irq 0 is invoked.
> The irq 0 increments the jiffies and the bottom half is marked.
> The bottom half now sets xtime.
> The scheduler don't needs the xtime, the system will run
> without problems, if i stop counting the xtime.

> Now i changed do_gettimeofday, do_settimeofday and update_wall_time in
> order not to manipulate xtime.
> After recompiling the kernel the system time is still running and
> date() shows the right time.
> Does somebody know which kernel function also sets xtime or is
> accountable doing the system time ?

>> Hi!
>> Don't ask me why, but i want to disable and stop the system time.
>> Every tick from the quartz the irq 0 is invoked.
>> The irq 0 increments the jiffies and the bottom half is marked.
>> The bottom half now sets xtime.
>> The scheduler don't needs the xtime, the system will run
>> without problems, if i stop counting the xtime.

>> Now i changed do_gettimeofday, do_settimeofday and update_wall_time in
>> order not to manipulate xtime.
>> After recompiling the kernel the system time is still running and
>> date() shows the right time.
>> Does somebody know which kernel function also sets xtime or is
>> accountable doing the system time ?

> Hi Sundial Services !

>> Perhaps it would be easier, and perhaps :-) it would even be possible, to
>> arrange for the clock to /appear/ to be stopped, as viewed by a
>> particular
>> application.  Without actually stopping the system-clock.

>> For example, the "date()" function could perhaps be overridden so that
>> your application, when compiled, does not call the "true" version of this
>> function ... does not actually interrogate the system clock, but works as
>> if it did.

> I want to set the system clock by my own program.
> If the ethernet card receives a packet, the xtime should increment
> by one.
> It is necessary to disable the system clock.