Appeal 2007-2127
Reexamination Control No. 90/006,621
definition in the '603 patent is inconsistent with the definition in the '604
patent and the meaning in the art, it is not entitled to any weight.
Moreover, Patent Owner is incorrect in stating that the claims do not
require more than one thread to be interrupted. Independent claims 1, 4, 6,
18, 22, 24, 26, 69, and 75 all require preempting an executing thread in
response to each actuation of the interrupt operation, which can only happen
if all threads are interruptible. That is, if a thread is executing when its
timeslice expires, these claims require that it must be preempted.
Independent claims 10, 14, and 17 only recite interrupting execution of a
first thread, but do not preclude interrupting execution of all threads and
Patent Owner considers these claims to read on conventional multithreading.
Independent claims 1, 18, 22, and dependent claim 25 specifically recite that
the threads alternate and perform successive incremental portions of their
respective subtasks, which necessarily requires that both threads are
interrupted before finishing their subtasks; a thread that is not finished with
its subtask would not voluntarily relinquish any part of its timeslice.
Patent Owner argues that the Examiner erred in construing
"preemptive execution of a plurality of threads" in the quotation above from
the '603 patent as requiring the system to "be preemptive for all threads,"
because this reads a limitation into that phrase that is not there (Br. 36). It is
argued that "preemptive" is an adjective modifying "execution" and "[a]s
long as preemption occurs during the execution of the 'application program',
'preemptive execution' occurs, whether or not each and every thread of the
program is preempted" (Br. 36).
The phrase "the concurrent time-sliced preemptive execution of a
plurality of threads of instructions" requires more than "preemptive
60
Page: Previous 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 Next
Last modified: September 9, 2013