Ex Parte Khan et al

Patent Trial and Appeal Board

Mar 26, 2013

11591010 (P.T.A.B. Mar. 26, 2013)

UNITED STATES PATENT AND TRADEMARK OFFICE
____________

BEFORE THE PATENT TRIAL AND APPEAL BOARD
____________

Ex parte GAZALA KHAN, SALEEM MOHIDEEN,
and MANISH AHLUWALIA
____________

Appeal 2010-007160
Application 11/591,010
Technology Center 2100
____________

Before SCOTT R. BOALICK, JOHN A. JEFFERY, and
DENISE M. POTHIER, Administrative Patent Judges.

JEFFERY, Administrative Patent Judge.

DECISION ON APPEAL
Appellants appeal under 35 U.S.C. § 134(a) from the Examiner’s
rejection of claims 1, 5-11, 14-16, 19, and 20. We have jurisdiction under
35 U.S.C. § 6(b). We affirm.

STATEMENT OF THE CASE
During execution of a binary program, Appellants’ invention flushes
pages in a file cache corresponding to a binary file before continuing
program execution. Flushing smaller pages in this fashion enables later use
Appeal 2010-007160
Application 11/591,010

2
of larger page sizes during execution. See generally Abstract; Spec 6:8–
7:25; Figs. 3A-3B. Claim 1 is illustrative:
1. A method of executing a binary program by a computer system, the
method comprising:
beginning execution of a first instance of the binary program;
reading program header table data for the binary program from a
binary file on disk storage;
storing the program header table data in cache memory for use by
subsequent instances of the binary program; and
applying a flush procedure relating to flushing pages in a file cache
that correspond to the binary file prior to continuing with execution of said
first instance, the flush procedure making a determination whether or not to
flush said pages if a preferred text page size for the binary file is greater than
a first threshold size.

THE REJECTION
The Examiner rejected claims 1, 5-11, 14-16, 19, and 20 under 35
U.S.C. § 103(a) as unpatentable over Case (US 7,278,008 B1; Oct. 2, 2007;
filed Jan. 30, 2004) and Loechel (US 5,895,488; Apr. 20, 1999). Ans. 3-7.1

CONTENTIONS
Regarding claim 1, the Examiner finds that Case (1) begins executing
a first instance of a binary program that accesses a cluster table; (2) reads
“program header table data” (i.e., the cluster table) from a binary file on disk
storage; and (3) stores this data in cache memory via an on-chip cache. Ans.
3. The Examiner acknowledges that Case does not flush cache entries, but

1 Throughout this opinion, we refer to (1) the Appeal Brief filed October 21,
2009 (“App. Br.”); (2) the Examiner’s Answer mailed January 19, 2010
(“Ans.”); and (3) the Reply Brief filed March 17, 2010 (“Reply Br.”).
Appeal 2010-007160
Application 11/591,010

3
cites Loechel as teaching this feature in concluding that the claim would
have been obvious. Ans. 3-4, 7-9.
Appellants argue that since Case’s cluster table is not a program
header table, the cited prior art does not read a program header table from a
binary file on disk storage as claimed. App. Br. 4-6; Reply Br. 1-2.
Appellants add that Loechel’s cache flushing based on a percentage of
“dirty” cache lines does not teach flushing based on determining whether or
not a preferred text page size for a binary file is greater than a first threshold
size as claimed. App. Br. 6-7; Reply Br. 2-3. Appellants also contend that
the Examiner’s motivation to combine the cited references is misplaced in
view of the claimed invention’s fragmentation reduction capabilities. App.
Br. 5-6.

ISSUES
I. Under § 103, has the Examiner erred in rejecting claim 1 by finding
that Case and Loechel collectively would have taught or suggested (1)
reading program header table data from a binary file on disk storage, and (2)
applying a flush procedure that determines whether or not to flush pages in a
file cache if a preferred text page size for the binary file is greater than a first
threshold size?
II. Is the Examiner’s combining the teachings of these references
supported by articulated reasoning with some rational underpinning to
justify the Examiner’s obviousness conclusion?

Appeal 2010-007160
Application 11/591,010

4
ANALYSIS
Claims 1 and 5-10
We begin by construing a key disputed term of representative claim 1,
namely “program header table data” (emphasis added). We emphasize the
term “data” here, for the claim does not recite a table per se, but rather table
data. Although this data is read from a binary file and stored in cache
memory for use in subsequent program instances in claim 1, whether this
data is derived from a table associated with a program header is not
functionally tied to the recited flush procedure which flushes pages from a
file cache based on page text size. In that sense, then, the particular type of
data (i.e., program header table data) is non-functional descriptive material.2
Therefore, even if a file’s program header table is “an array of
structures describing information a computing system requires in order to
prepare or execute a program or create an executable image” as Appellants
contend (App. Br. 5), that data comes from a header table is irrelevant at
least with respect to the disputed steps of claim 1. As such, leaving aside the
fact that a program header table is undefined in the Specification as the
Examiner indicates (Ans. 8), and can include a wide variety of data
including metadata and unspecified “other data” according to Appellants’
non-limiting example (App. Br. 5), Appellants’ arguments regarding Case’s
cluster table not constituting a program header table (App. Br. 5-6; Reply Br.
1-2) are not commensurate with the scope of the claim which recites, among
other things, reading and storing data and not a table.

2 See In re Ngai, 367 F.3d 1336, 1339 (Fed. Cir. 2004); see also Ex parte
Nehls, 88 USPQ2d 1883, 1887-89 (BPAI 2008) (precedential) (discussing
cases pertaining to non-functional descriptive material).
Appeal 2010-007160
Application 11/591,010

5
We are therefore unpersuaded of error in the Examiner’s reliance on
Case’s cluster table in connection with the recited data reading step, for Case
at least suggests reading data from a binary file on disk storage by reading
data from the cluster table for address translation. See Ans. 3 (citing Case,
col. 3, ll. 30-40); Case, col. 11, ll. 26-44. Although Case does not expressly
state that the cluster table’s data is read from a binary file, Case nonetheless
indicates that the table can reside in main system memory or elsewhere in
the system as Appellants acknowledge (Reply Br. 2 (citing Case, col. 11, ll.
30-32)), which reasonably suggests some sort of associated stored binary file
for this data. See Case, Fig. 1 (showing system memory 104 with cluster
table 140, system disk 128, and removable storage device 129).
Nor are we persuaded of error in the Examiner’s reliance on Loechel
for teaching the recited flush procedure. As Appellants acknowledge,
Loechel flushes a cache if a determined percentage of “dirty”3 cache lines
exceeds a predetermined threshold. App. Br. 6; Reply Br. 2 (citing Loechel,
Abstract; col. 6, ll. 52-59; Fig. 5). Despite Appellants’ arguments to the
contrary, we see no error in the Examiner’s equating the percentage of dirty
cache lines to the recited preferred text page size. Ans. 8-9. This finding is
based on the notion that a “page” comprises multiple cache lines (Ans. 9): an
unrebutted finding that has a rational basis given the Specification’s
reference to pages as contiguous blocks of memory address space. See Spec.
1:27-29. To the extent that Appellants argue that this “page” size is not a
text page size, we find this distinction to be without a functional difference

3 “Dirty” data is data not yet written to persistent memory. Loechel, col. 2,
ll. 15-18.
Appeal 2010-007160
Application 11/591,010

6
here, for the Examiner’s finding is based on data size—data that skilled
artisans would recognize could include text.
Combining Loechel’s teaching with Case’s system that caches cluster
data as the Examiner proposes would predictably yield a procedure that
would flush “pages” from the cache if they exceed a threshold page size.
Since skilled artisans would recognize that this cached cluster data is derived
at least partly from an associated binary file for the cluster table as noted
above, the preferred page size is therefore for a binary file as claimed.
Lastly, we find unpersuasive Appellants’ contention that the
Examiner’s motivation to combine the cited references is misplaced in view
of the claimed invention’s fragmentation reduction capabilities. App. Br. 5-
6. Applying Loechel’s cache-flushing technique to Case’s cache which
stores cluster information would predictably flush the cache when needed to
make the cache available for additional data as the Examiner indicates. Ans.
4. Since this enhancement predictably uses prior art elements according to
their established functions (see KSR Int’l Co. v. Teleflex Inc., 550 U.S. 398,
417 (2007)), we find the Examiner’s combining the cited references’
teachings supported by articulated reasoning with some rational
underpinning to justify the Examiner’s obviousness conclusion.
We are therefore not persuaded that the Examiner erred in rejecting
representative claim 1, and claims 5-10 not separately argued with
particularity.

Claims 11, 14-16, 19, and 20
Despite nominally arguing claims 11, 14-16, 19, and 20 separately,
Appellants reiterate similar arguments made in connection with claim 1 (see
Appeal 2010-007160
Application 11/591,010

7
App. Br. 8-15; Reply Br. 4) which we find unpersuasive for the reasons
previously discussed.

CONCLUSION
The Examiner did not err in rejecting claims 1, 5-11, 14-16, 19, and
20 under § 103.

ORDER
The Examiner’s decision rejecting claims 1, 5-11, 14-16, 19, and 20 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

babc