Ex Parte Delapedraja

Patent Trial and Appeal Board

Sep 18, 2012

11021253 (P.T.A.B. Sep. 18, 2012)

UNITED STATES PATENT AND TRADEMARK OFFICE
____________________

BEFORE THE PATENT TRIAL AND APPEAL BOARD
____________________

Ex parte JAROSLAV DELAPEDRAJA
____________________

Appeal 2010-004115
Application 11/021,253
Technology Center 2100
____________________

Before DENISE M. POTHIER, BRUCE R. WINSOR,
and JENNIFER S. BISK, Administrative Patent Judges.

BISK, Administrative Patent Judge.

DECISION ON APPEAL
SUMMARY
This is an appeal under 35 U.S.C. § 134(a) from the Examiner’s
rejection of claims 1-34, which are all the claims pending in this application.
We have jurisdiction under 35 U.S.C. § 6(b).
We affirm.

The Invention
Appellant’s invention relates to providing efficient access to a data
structure that has an ordered arrangement of elements and is shared by
Appeal 2010-004115
Application 11/021,253

2

multiple tasks in a computing environment. Abstract. The claims describe
associating a spinlock with the data structure. To read the data structure,
each task uses a traversal iterator to point to the element in the data structure
being read. The referenced element also points back to the applicable
traversal iterator. While the reader task is using the pointer of its associated
traversal iterator it holds the spinlock. To delete an element in the data
structure, a writer task holds the spinlock and modifies the pointer of each
traversal iterator that is pointed to by that element. See App. Br. 42-50,
Claims App’x.
Claim 1, reproduced below with emphasis added, is illustrative of the
claimed subject matter.
1. A method for providing efficient access to a data structure
having an ordered arrangement of elements by multiple tasks in a
computing environment, comprising:
associating a spinlock with the data structure;
defining a plurality of traversal iterators, each having a read
pointer pointing to an element within the data structure, wherein each
traversal iterator is associated with a reader task to enable the plurality
of traversal iterators to share access to the data structure concurrently;
traversing the data structure using the reader task, wherein the
reader task holds the spinlock while using the read pointer of the
reader tasks’s associated traversal iterator; and
deleting a selected element from the data structure using a writer
task, wherein the writer task holds the spinlock while modifying the
read pointer of each traversal iterators associated with the reader task
whose read pointers point to the selected element being deleted.

Appeal 2010-004115
Application 11/021,253

3

The Rejections
1. Claims 1-12, 27, 31, and 33 stand rejected under § 103(a) as obvious over
Schimmel (US 6,601,120 B1; July 29, 2003), Kirkman (US 6,581,063
B1; June 17, 2003), and Nitta (US 5,287,521; Feb. 15, 1994). Ans. 3-14.
2. Claims 13-26, 28-30, 32, and 34 stand rejected under § 103(a) as obvious
over Schimmel, Kirkman, Nitta, and Cavanaugh (US 6,016,489; Jan. 18,
2000). Ans. 15-28.

THE OBVIOUSNESS REJECTION OF CLAIMS 1-12, 27, 31, AND 33
The Examiner rejected claims 1-12, 27, 31, and 33 under § 103(a) as
obvious over the combination of Schimmel, Kirkman, and Nitta finding that:
(1) Schimmel teaches “associating a spinlock with a data structure” and
reader and writer tasks, which hold that spinlock when accessing the data
structure (the “associating limitation”); (2) Kirkman teaches a plurality of
traversal iterators; and (3) Nitta teaches enabling the plurality of pointers
from read and write processes to share access to the data structure
concurrently. Ans. 3-4.
Appellant argues that none of the cited references teach or suggest
“deleting a selected element from the data structure using a writer task,
wherein the writer task holds the spinlock while modifying the read pointer
of each traversal iterators” (the “deleting limitation”). App. Br. 27-34.
Appellant also argues that the Examiner improperly combined
Schimmel and Kirkman. Ans. 30-32.

Appeal 2010-004115
Application 11/021,253

4

Issues
1. Under § 103, has the Examiner erred in rejecting claims 1-12, 27,
31, and 33 by finding that Schimmel, Kirkman, and Nitta collectively would
have taught or suggested the deleting limitation?
2. Is the Examiner’s reason to combine the teachings of these
references supported by articulated reasoning with some rational
underpinning to justify the Examiner’s obviousness conclusion?

Analysis
We are not persuaded by Appellant’s arguments. Appellant argues
that Schimmel does not disclose a plurality of traversal iterators or
modifying the read pointer of each traversal iterator. App. Br 28-29. This
argument is not consistent with the Examiner’s rejection which cites to
Kirkman, not Schimmel, as teaching this limitation. Ans. 4, 29-30.
Similarly, Appellant’s argument that Nitta does not teach the deleting
limitation does not relate to the rejection because the Examiner only relied
on Nitta for teaching a plurality of processes sharing access to the data
structure concurrently. App. Br. 32-33; Ans. 4, 34.
Appellant also argues that Kirkman does not make up for the
deficiencies in Schimmel because Kirkman’s iterators do not function in the
recited manner. App. Br. 29. Kirkman describes a linked list with an
associated data structure used by tasks to update the linked list. See
Kirkman, Abstract. The data structure is used to block subsets of the linked
list from being altered by other tasks and includes “inspector objects,” which
traverse the linked list without changing it (similar to a read task) and
“mutator objects,” which alter the list (similar to a write task). See id.
Appeal 2010-004115
Application 11/021,253

5

Mutators have associated blocker objects to block access to selected list
elements while they are being altered. Id. Access is blocked by altering
inspector object pointers to point to the blocker object instead of the linked
list element. Id. The blocker object in turn points to a section of the linked
list that is not blocked. Id. This way, an iterator object by-passes the blocked
portion of the linked list. Id.
The Examiner reasonably finds that the pointers associated of
Kirkman’s inspector object are necessarily modified to point to the next
available element on the list to prevent a “dangling pointer”—something that
would have been well known to a person of ordinary skill in the art. Ans. 30.
Appellant disputes this finding, stating that Kirkman only explicitly
describes changing the blocking object, but does not explicitly describe
modifying the inspector object pointers. App. Br. 29-30. This argument is
not effective to rebut the Examiner’s finding. Appellant does not provide
persuasive argument or evidence that a person of skill in the art would not
know that the pointers of the inspector object are necessarily modified in
order to prevent dangling pointers. Mere arguments unsupported by factual
evidence are entitled to little probative value. Cf., In re Geisler, 116 F.3d
1465, 1470 (Fed. Cir. 1997).
We are also not persuaded that the Examiner improperly combined
Schimmel and Kirkman. Appellant argues that modification of Schimmel is
not necessary because of its stated assumption that tasks that write to the
data structure are rare. App. Br. 31 (citing Schimmel col. 3, ll. 27-29). This
argument does not address the relevant inquiry—whether the claimed
subject matter would have been obvious to those of ordinary skill in the art
in light of the combined teachings of the references. See In re Keller, 642
Appeal 2010-004115
Application 11/021,253

6

F.2d 413, 425 (CCPA 1981). The Examiner’s findings regarding what
Schimmel would teach a person of ordinary skill in the art are reasonable.
Ans. 4-6; 10-14; see In re Mouttet, 686 F.3d 1322, 1331 (Fed. Cir. 2012)
(“A reference may be read for all that it teaches, including uses beyond its
primary purpose.”).
Appellant also argues that a person of ordinary skill would not have
looked to the teachings of Kirkman, because Kirkman’s invention eliminates
the needs for locks and states that locks have the potential to affect
performance. App. Br. 31 (citing Kirkman col. 2, ll. 32-45). The Examiner,
however, finds that a person of ordinary skill would have realized that a spin
lock was efficient and would look to Kirkman for a general solution of
updating shared data structures in a lock environment. Ans. 32-33.
Moreover, the Examiner explains that a person of ordinary skill in the art
would have been motivated to combine the teachings of Schimmel and
Kirkman “to update . . . linked lists in a multiprocessor environment which
is less disruptive of tasks executing in other processors.” Ans. 4 (citing
Kirkman col. 2, ll. 45-49). Appellant provides no persuasive evidence to
rebut these findings.
For these reasons, we conclude that Appellant has not rebutted the
Examiner’s prima facie case of obviousness for representative claim 1 or
claims 2, 3, 6-12, 27, 31, and 33, which were not argued separately. App.
Br. 33-34.
Appellant argues that claims 4 and 5 are patentable for the additional
reason that the cited references fail to teach “releasing the spinlock held by
the reader task immediately after” (a) each element is visited by the reader
task as recited in claim 4 or (b) removing the selected element as recited in
Appeal 2010-004115
Application 11/021,253

7

claim 5. App. Br. 34-36. Appellant’s sole argument is that Nitta does not
disclose releasing the lock immediately after each element is visited because
it requires a “shared mode counter,” as found by the Examiner. App. Br. 34.
Appellant asserts that this “shared mode counter” would not be necessary if
the lock was released by a reader task. App. Br. 34-36. We are not persuaded
by this argument. The Examiner explains that Nitta teaches that the unlock
program can be called by a task at any time. Ans. 35 (citing Nitta col. 6, ll.
15-25). This necessarily includes unlocking immediately after a read. See
Ans. 35. Thus, although not explicitly stated in Nitta, Nitta does teach
unlocking (releasing the spinlock) immediately after an element is visited by
a reader task. “A person of ordinary skill is also a person of ordinary
creativity, not an automaton.” KSR Int’l Co. v. Teleflex Inc., 550 U.S. 398,
421 (2007). And a reference may be read for all that it teaches, including
uses beyond its primary purpose. Mouttet, 686 F.3d at 1331. Appellant’s
argument that a person of ordinary skill would not realize this because it is
not explicitly stated in the reference, made without reference to persuasive
supporting evidence, is not convincing.
Thus, we also sustain the Examiner’s rejections of claims 4 and 5.

THE OBVIOUSNESS REJECTION OF CLAIMS 13-26, 28-30, 32, and 34
The Examiner rejected claims 13-26, 28-30, 32, and 34 under § 103(a)
as obvious over the combination of Schimmel, Kirkman, Nitta, and
Cavanaugh. The Examiner uses the same reasoning as in the previously
discussed rejection, adding that Cavanaugh teaches that the referenced
element points back to the traversal iterator. Ans. 16.
Appeal 2010-004115
Application 11/021,253

8

Appellant argues that Cavanaugh does not overcome the deficiencies
of Schimmel, Kirkman, and Nitta argued in the previous rejection. App. Br.
36-38. Appellant adds that the Examiner improperly combined Cavanaugh
with the other references. App. Br. 38-39.

Issue
Is the Examiner’s reason to combine the teachings of these references
supported by articulated reasoning with some rational underpinning to
justify the Examiner’s obviousness conclusion?

Analysis
We are not persuaded by Appellant’s arguments. First, as discussed
above, we do not agree that Schimmel, Kirkman, and Nitta are deficient.
Second, we are not persuaded by Appellant’s argument that Cavanaugh’s
iterators cannot be combined with the iterators of Schimmel, Kirkman, and
Nitta because of the way they are implemented. App. Br. 38-39. The
relevant inquiry is whether the claimed subject matter would have been
obvious to those of ordinary skill in the art in light of the combined
teachings of the references. See Keller, 642 F.2d at 425. “Combining the
teachings of references does not involve an ability to combine their specific
structures.” In re Nievelt, 482 F.2d 965, 968 (CCPA 1973). Appellant does
not provide persuasive argument or evidence that combining the various
types of iterators from the references would be “uniquely challenging or
difficult for one of ordinary skill in the art” at the time of Appellant’s
invention. Leapfrog Enters., Inc. v. Fisher-Price, Inc., 485 F.3d 1157, 1162
(Fed. Cir. 2007) (citing KSR, 550 U.S. at 418). Finally, Appellant asserts
Appeal 2010-004115
Application 11/021,253

9

that the Examiner did not provide a reason for the combination of Nitta and
Cavanaugh. App. Br. 38. To the contrary, the Examiner states that it would
have been obvious to combine the teachings of Nitta with Schimmel and
Cavanaugh “in order to improve overall processing capability and upgrade
the concurrent processing level.” Ans. 37 (citing Nitta col. 3, ll. 30-35).
Appellant argues that claims 14 and 17 are patentable for the
additional reason that the cited references fail to teach the additional
limitations recited by these dependent claims. App. Br. 34-36. Appellant’s
sole argument is identical to that advanced for claims 4 and 5—Nitta does
not disclose releasing the lock immediately after each element is visited
because it requires a “shared mode counter,” as found by the Examiner. As
discussed above, we are not persuaded by this argument.
For these reasons, we sustain the rejection of claims 13-26, 28-30, 32,
and 34.

DECISION
The Examiner’s decision rejecting claims 1-34 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