International Business Machines Corporation

Zhi Gan

Xiao Jun Dai

Subhendu Das

Zhang Yue

Rui Han

Xian Liu

Yao Dong Qi

Aravind Srinivasan

Ying Chun Guo

Mo Jiong Qiu

Rahul Kurane

Managing access to a cache memory includes dividing said cache memory into multiple of cache areas, each cache area having multiple entries; and providing at least one separate lock attribute for each cache area such that only a processor thread having possession of the lock attribute corresponding to a particular cache area can update that cache area.

Locks to enable updating data and a data replacement order in cache areas

A cache memory device includes a plurality of cache areas, each the cache area comprising a plurality of entries. The cache memory device is configured to maintain a separate lock attribute for each the cache area and temporarily assign possession of a lock attribute for a particular the cache area to a processor thread attempting to update the particular the cache area, the processor thread being unable to update the particular the cache area without possession of the lock attribute for the particular the cache area.

Managing processor thread access to cache memory using lock attributes

Virtual machines are managed by obtaining software hierarchy information of a current virtual machine to be installed. Then logical memory assigned to the current virtual machine is divided into a private part and a shared part based at least in part upon existing software hierarchy information of at least one virtual machine already installed and the software hierarchy information of the current virtual machine. Then, the shared part of the logical memory is mapped to shared segments of a physical memory, wherein the shared segments are used by at least one installed virtual machine.

Managing memory in multiple virtual machines

A method of accessing a shared data structure in parallel by multiple threads in a parallel application program is disclosed. A lock of the shared data structure is granted to one thread of the multiple threads, an operation of the thread which acquires the lock is performed on the shared data structure, an operation of each thread of the multiple threads which does not acquire the lock is buffered, and the buffered operations are performed on the shared data structure when another thread of the multiple threads subsequently acquires the lock. A corresponding apparatus and program product are also disclosed.

Method and apparatus for accessing a shared data structure in parallel by multiple threads

A computer-implementable method, system and computer-usable medium for extending the portability of code to a limited-class environment are disclosed. In a preferred embodiment, the method includes the steps of: identifying any full-version-only objects that are not in a reduced-version of a language, wherein the reduced-version of the language comprises a reduced portion of objects that are in a full-version of the language; associating a full-version-only object with a set of one or more objects in the reduced-version of the language, wherein the full-version-only object is functionally equivalent to the set of one or more objects in the reduced-version of the language; identifying any full-version-only objects in a software program; and creating a reduced-version of the software program by replacing any identified full-version-only objects with a functionally equivalent set of one or more objects from the reduced-version of the language.

Growing community of inventors

Beijing, China

Zhi Gan