Jeff Dike writes:
> frank.private@... said:
> > The date skipped back to the date of the host system. Has anybody an
> > idear solving this problem?
> It looks like the easiest way to fix this is to make stime and time act more
> like settimeofday and gettimeofday.
Well I've done an implementation, just for fun, of a working `time'
and `stime' sys_call. They are _almost_ good... `time' I think is
pretty much ok, the problem is with `stime'. I couldn't get the
necessary to define put_user(), into arch/um/kernel/time.c, so I
"hacked" a memcpy() :-(. Besides that, the patch is pretty much basic.
It follows the settimeofday and gettimeofday implementation like you
suggested, and works fine here at home and at the Univ.
Flames and corrections are always appreciated!
best regards to all,