Patents by Inventor Ching-Fu Kung

Ching-Fu Kung has filed for patents to protect the following inventions. This listing includes patent applications that are pending as well as patents that have already been granted by the United States Patent and Trademark Office (USPTO).

  • Patent number: 10019288
    Abstract: A hypervisor hosted by a computing system performs a method to allocate a contiguous physical memory space to a device. A given region of physical memory is marked as migratable. From an operating system (OS) kernel, the hypervisor receives a request for memory allocation to the device, the request indicating a first set of available virtualized pages in a virtualized memory. In response to the request, the hypervisor identifies a set of contiguous frames in the given region to be allocated to the device. The set of contiguous frames are mapped to a second set of virtualized pages. The hypervisor disables the mapping for the first set of available virtualized pages and the second set of virtualized pages. Then one or more occupied frames in the set of contiguous frames are migrated out of the given region to allow for allocation of the set of contiguous frames to the device.
    Type: Grant
    Filed: September 12, 2016
    Date of Patent: July 10, 2018
    Assignee: MediaTek, Inc.
    Inventors: Ching-Fu Kung, Sheng-Yu Chiu
  • Publication number: 20180074863
    Abstract: A hypervisor hosted by a computing system performs a method to allocate a contiguous physical memory space to a device. A given region of physical memory is marked as migratable. From an operating system (OS) kernel, the hypervisor receives a request for memory allocation to the device, the request indicating a first set of available virtualized pages in a virtualized memory. In response to the request, the hypervisor identifies a set of contiguous frames in the given region to be allocated to the device. The set of contiguous frames are mapped to a second set of virtualized pages. The hypervisor disables the mapping for the first set of available virtualized pages and the second set of virtualized pages. Then one or more occupied frames in the set of contiguous frames are migrated out of the given region to allow for allocation of the set of contiguous frames to the device.
    Type: Application
    Filed: September 12, 2016
    Publication date: March 15, 2018
    Inventors: Ching-Fu Kung, Sheng-Yu Chiu
  • Publication number: 20170060783
    Abstract: An apparatus for performing secure memory allocation control in an electronic device and an associated method are provided. The electronic device may include a plurality of bus master circuits, each of which has capability of accessing data through a bus of the electronic device, and may further include a plurality of master side memory address filters (MAFs) that are coupled between the bus and the bus master circuits, where the apparatus may include a control circuit that is coupled to the master side MAFs. In addition, the control circuit may be arranged for controlling secure memory allocation of the electronic device through the master side MAFs, to restrict any unauthorized access to any portion of secure data within the electronic device. Additionally, the master side MAFs may be arranged for selectively restricting data accessing activities of the bus master circuits through memory address filtering.
    Type: Application
    Filed: March 9, 2016
    Publication date: March 2, 2017
    Inventors: Sheng-Yu Chiu, Ching-Fu Kung, Chih-Pin Su, Ming-Hsien Hsieh
  • Patent number: 8452863
    Abstract: An operation method for a server system includes: (A) under control of a hardware abstraction layer (HAL), a plurality of node management units sharing a hardware resource; (B) if one of the node management units needs to use the hardware resource, the node management unit sending an instruction or a data to the HAL and accordingly the HAL using the hardware resource in represent of the node management unit; and (C) if an external instruction is received, the HAL identifying which transmission port of the hardware resource receives the external instruction, so to send the external instruction to a corresponding node management unit, and after the external instruction is executed, the corresponding node management unit sending back an information to the HAL so that the HAL sends back the information to an external system administrator.
    Type: Grant
    Filed: December 3, 2010
    Date of Patent: May 28, 2013
    Assignee: Quanta Computer Inc.
    Inventors: Te-Hsien Lai, Yu-Cheng Chen, Ching-Fu Kung
  • Patent number: 8130763
    Abstract: A data item interval identifier lookup method and system is proposed, which is designed for integration to an information processing system for finding which predefined interval the value of an input data item, such as an IP (Internet Protocol) address, belongs. The proposed method and system is characterized by the use of a multi-stage lookup-table data structure having a number of cascaded lookup tables constructed by partitioning the data format of the input data item into a number of segments, each being mapped to one stage of lookup table data structure whose key-value relationships are predefined based on a predefined interval-and-identifier definition table. In operation, the values of the partitioned segments are sequentially used as lookup keys to search through the multi-stage lookup-table data structure until the corresponding interval identifier is found. This feature allows the implementation to have low memory requirement and enhanced system performance.
    Type: Grant
    Filed: June 19, 2008
    Date of Patent: March 6, 2012
    Assignee: National Taiwan University
    Inventors: Ching-Fu Kung, Sheng-De Wang
  • Publication number: 20120023210
    Abstract: An operation method for a server system includes: (A) under control of a hardware abstraction layer (HAL), a plurality of node management units sharing a hardware resource; (B) if one of the node management units needs to use the hardware resource, the node management unit sending an instruction or a data to the HAL and accordingly the HAL using the hardware resource in represent of the node management unit; and (C) if an external instruction is received, the HAL identifying which transmission port of the hardware resource receives the external instruction, so to send the external instruction to a corresponding node management unit, and after the external instruction is executed, the corresponding node management unit sending back an information to the HAL so that the HAL sends back the information to an external system administrator.
    Type: Application
    Filed: December 3, 2010
    Publication date: January 26, 2012
    Applicant: Quanta Computer Inc.
    Inventors: Te-Hsien LAI, Yu-Cheng Chen, Ching-Fu Kung
  • Patent number: 7792120
    Abstract: A computer network packet classification method and system based on a nonoverlapping rule group encoding scheme is proposed, which is designed for integration to a network system for classification of packets within the network system. The proposed method and system is characterized by the use of a nonoverlapping rule group encoding scheme which organizes a database of rules into nonoverlapping rule groups and creates a number of consecutive projected intervals over the dimension of each classification-related field of the packet header, whereby a projected-interval to encoded-bit-vector lookup table and an encoded-bit-vector to rule-group lookup table can be established. During the operation of packet classification, these two lookup tables are used to find the corresponding rule for each incoming packet. This scheme allows the encoded bit vectors to have a reduced bit length, and therefore allows the packet classification to be implemented with low memory requirement and enhanced performance.
    Type: Grant
    Filed: June 27, 2008
    Date of Patent: September 7, 2010
    Assignee: National Taiwan University
    Inventors: Ching-Fu Kung, Sheng-De Wang
  • Publication number: 20090196291
    Abstract: A computer network packet classification method and system based on a nonoverlapping rule group encoding scheme is proposed, which is designed for integration to a network system for classification of packets within the network system. The proposed method and system is characterized by the use of a nonoverlapping rule group encoding scheme which organizes a database of rules into nonoverlapping rule groups and creates a number of consecutive projected intervals over the dimension of each classification-related field of the packet header, whereby a projected-interval to encoded-bit-vector lookup table and an encoded-bit-vector to rule-group lookup table can be established. During the operation of packet classification, these two lookup tables are used to find the corresponding rule for each incoming packet. This scheme allows the encoded bit vectors to have a reduced bit length, and therefore allows the packet classification to be implemented with low memory requirement and enhanced performance.
    Type: Application
    Filed: June 27, 2008
    Publication date: August 6, 2009
    Applicant: NATIONAL TAIWAN UNIVERSITY
    Inventors: Ching-Fu Kung, Sheng-De Wang
  • Publication number: 20090190597
    Abstract: A data item interval identifier lookup method and system is proposed, which is designed for integration to an information processing system for finding which predefined interval the value of an input data item, such as an IP (Internet Protocol) address, belongs. The proposed method and system is characterized by the use of a multi-stage lookup-table data structure having a number of cascaded lookup tables constructed by partitioning the data format of the input data item into a number of segments, each being mapped to one stage of lookup table data structure whose key-value relationships are predefined based on a predefined interval-and-identifier definition table. In operation, the values of the partitioned segments are sequentially used as lookup keys to search through the multi-stage lookup-table data structure until the corresponding interval identifier is found. This feature allows the implementation to have low memory requirement and enhanced system performance.
    Type: Application
    Filed: June 19, 2008
    Publication date: July 30, 2009
    Applicant: NATIONAL TAIWAN UNIVERSITY
    Inventors: Ching-Fu Kung, Sheng-De Wang
  • Patent number: RE47289
    Abstract: An operation method for a server system includes: (A) under control of a hardware abstraction layer (HAL), a plurality of node management units sharing a hardware resource; (B) if If one of the node management units needs to use the hardware resource, the method includes the node management unit sending an instruction or a data to the HAL and accordingly the HAL using the hardware resource in represent of the node management unit; and (C) if. If an external instruction is received, the method includes the HAL identifying which transmission port of the hardware resource receives the external instruction, so to send the external instruction to a corresponding node management unit, and. Then, after the external instruction is executed, the method includes the corresponding node management unit sending back an information to the HAL so that the HAL sends back the information to an external system administrator.
    Type: Grant
    Filed: May 4, 2015
    Date of Patent: March 12, 2019
    Assignee: QUANTA COMPUTER INC.
    Inventors: Te-Hsien Lai, Yu-Cheng Chen, Ching-Fu Kung