Ex Parte Cheng et al

Board of Patent Appeals and Interferences

Jan 31, 2012

10672180 (B.P.A.I. Jan. 31, 2012)

UNITED STATES PATENT AND TRADEMARK OFFICE
____________

BEFORE THE BOARD OF PATENT APPEALS
AND INTERFERENCES
____________

Ex parte JEFFREY G. CHENG,
HING PONG CHAN, and YINAN JIANG
_____________

Appeal 2009-013735
Application 10/672,180
Technology Center 2100
____________

Before MARC S. HOFF, THOMAS S. HAHN, and
ELENI MANTIS MERCADER, Administrative Patent Judges.

MANTIS MERCADER, Administrative Patent Judge.

DECISION ON APPEAL

Appeal 2009-013735
Application 10/672,180

2
STATEMENT OF THE CASE
Appellants appeal under 35 U.S.C. § 134(a) from the final rejection of
claims 1, 2, 5-7, 10-13, 24-26, 28-30 and, 32-35. We have jurisdiction under
35 U.S.C. § 6(b).
We affirm-in-part.

INVENTION
Appellants’ Figure 1 is depicted below:

Appellants’ Figure 1 shows a hang detector module 106, located in
processor 100, used to detect a hang in a co-processor 102.

Appellants’ claimed invention is directed to a hang detector module
106 which detects a hang in a co-processor 102. The hang detector module
106 examines a flag that indicates whether the co-processor 102 is currently
executing instructions. The same hang detector module 106 then examines
corresponding registers associated with the co-processor 102 to determine
Appeal 2009-013735
Application 10/672,180

3
whether activity in those registers is occurring that reflects the actual
processing of such instructions. If a discrepancy between the activity in
such registers is detected (i.e., a flag indicates the current state as busy but
there is no activity in the registers), the hang detector module 106 then
returns a hang notification signal 110 to the selective processor reset module
108 indicating that a hang has been detected in the co-processor 102. See
Spec.¶¶ [0035], [0049].
Claims 1 and 11, reproduced below, are representative of the subject
matter on appeal.
1. A circuit for monitoring and resetting a co-processor comprising:
a hang detector module operative to detect a hang in the co-processor
by detecting a discrepancy between a current state of the co-processor and a
current activity of the coprocessor; and
a selective processor reset module operative to selectively reset the
co-processor without resetting a processor, in response to detecting a hang in
the co-processor.

11. A circuit for monitoring and resetting a co-processor comprising:
a hang detector module operative to detect a hang in the co-processor;
a halt communications module operative to halt executable instruction
communications with the co-processor, in response to detecting a hang in the
co-processor;
a selective processor reset module operative to selectively reset the
co-processor without resetting a processor, in response to detecting a hang in
the co-processor;
a reset check module operative to detect if the co-processor has been
successfully reset, in response to the resetting of the co-processor; and
a restart communications module operative to restart executable
instruction communications with the co-processor, in response to detecting
that the co-processor has been successfully reset.
Appeal 2009-013735
Application 10/672,180

4
THE REJECTION
The Examiner relies upon the following as evidence of
unpatentability:
Hill US 2002/0093505 A1 Jul. 18, 2002
Forsman US 6,742,139 B1 May 25, 2004
(filed Oct. 19, 2000)

The following rejection is before us for review:
1. The Examiner rejected claims 1-2, 5-7, 10-11, 24-26, 28-30, and
32-35 under 35 U.S.C. § 102(e) as being anticipated by Forsman.
2. The Examiner rejected claims 12 and 13 under 35 U.S.C. § 103(a)
as being unpatentable over Forsman in view of Hill.

ISSUES
The pivotal issues are whether the Examiner erred in finding that
Forsman teaches:
“a hang detector module operative to detect a hang in the co-
processor by detecting a discrepancy between a current state of the co-
processor and a current activity of the co-processor” as recited in claim 1;
and
“a halt communications module operative to halt executable
instruction communications” as recited in claim 11.

Appeal 2009-013735
Application 10/672,180

5
ANALYSIS
1. Analysis with respect to claims 1-2, 5-7, 10-11, 24-26, 28-30,
and 32-35
Appellants argue, with respect to claim 1, that the Examiner erred in
finding that Forsman teaches detecting a discrepancy between a current
status and a current activity of a co-processor (See App. Br. 16-18; Reply Br.
2).
We agree with Appellants’ argument. We agree with Appellants that,
contrary to the Examiners assertions (Ans. 13), Forsman’s status/control
register is not used to detect any hang condition (col. 4, ll. 36-65) but is used
after the detection of hang to determine if conditions exist that pre-empt
from re-setting (col. 4, ll. 25-65). Forsman teaches that a hang is detected
by host 202 by failure to detect a heartbeat signal (col. 4, ll. 25-35).
Additional steps are taken during the initiation of the communications
recovery (i.e., after the hang has already been detected) by the status/control
register 208 to determine if some conditions exist such as handling a critical
event that need to be finished prior to completing the resetting of the
processor 204 based on the detected hang. Thus, Forsman does not teach a
hang detector module operative to detect a hang in the co-processor by
detecting a discrepancy between a current state of the co-processor and a
current activity of the coprocessor.
Appeal 2009-013735
Application 10/672,180

6
Accordingly, we reverse the Examiner’s rejection of claim 1 and the
rejections of independent claims 6, 24, 26, 30, and 34 and their dependent
claims 2-5, 7-10, 25, 27-29, 31-33, and 35 for the same reasons.1
With respect to claim 11, Appellants argue that Forsman does not
teach a halt communication module because communication exists by way
of an acknowledgement message (App. Br. 23-24). We are not persuaded by
Appellants’ argument. Forsman’s acknowledgment statement is that the
service processor has received the warning and it is ready to reset (col. 5, ll.
16-19). If anything the acknowledgment indicates that the halt
communication has been received and the executable instructions have
ceased since the service processor is ready to be reset. The claim does not
recite that all communications have ceased, but rather that the executable
instruction communications have halted.
Thus, we will affirm the Examiner’s rejection of claim 11.
2. Analysis with respect to claims 12 and 13
Appellants present no additional arguments of patentability with respect
to claims 12 and 13.2 Accordingly, we will affirm the Examiner’s rejections
of these claims, as Appellants have not shown error on behalf of the
Examiner.

1 While Appellants group claims 5 and 10 with claim 11 (App. Br. 21), we
consider this an inadvertent error as these claims depend from claims 1 and
6, respectively--not claim 11.
2 Appellants assert that claims 12 and 13 stand or fall with claim 1 (App. Br.
14). However, claims 12 and 13 depend on claim 11.
Appeal 2009-013735
Application 10/672,180

7
CONCLUSIONS
The Examiner erred in finding that Forsman teaches:
“a hang detector module operative to detect a hang in the co-
processor by detecting a discrepancy between a current state of the co-
processor and a current activity of the co-processor” as recited in claim 1.
The Examiner did not err in finding that Forsman teaches:
“a halt communications module operative to halt executable
instruction communications” as recited in claim 11.

ORDER
The decision of the Examiner to reject claims 1, 2, 5-7, 10, 24-26, 28-
30 and 32-35 is reversed. The decision of the Examiner to reject claims 11-
12 and 13 is affirmed.
No time period for taking any subsequent action in connection with
this appeal may be extended under 37 C.F.R. § 1.136(a)(1)(iv).

AFFIRMED-IN-PART

rwk