International Business Machines Corporation

Srinivasa Rao Muppala

Douglas J Griffith

Sreenivas Makineedi

Anil Kalavakolanu

Srinivasa M Raghavan

Yanhua Yang

Mathew Accapadi

Evelyn Tingmay Yeung

Anand Teerth Desai

Evan Allen Zoss

Rama Mothey Tenjarla

Shanna Beck

Vidya Makineedi

Emmanuelle Samir Hanna Matta

Embodiments are disclosed for managing heap metadata corruption. The techniques include detecting a metadata corruption error in a first heap disposed in a first region of memory. The techniques also include generating a second heap in a free memory region that is disposed beyond a break value address of a memory allocation system. The techniques further include updating a first entry for the first heap in a heap directory. Additionally, the techniques include generating a second entry for the second heap in the heap directory. The techniques also include processing a call to the memory allocation system for the first heap based on the first entry and the second entry.

Managing heap metadata corruption

Threads of a multithreaded application may be scheduled to different cores and executed in various orders and at various frequencies. Controlling how the threads are scheduled and clock rates of processor cores enables testing multiple possible execution scenarios, which may force previously unknown timing window problems to occur. These timing window problems may then be detected.

Method and technique to find timing window problems

Embodiments include method, systems and computer program products for performing Spectre mitigation on a workload. The method includes starting, by at least one processor of a plurality of processors, a process. The at least one processor determines that the process is a kernel process. The at least one processor determines that an interrupt has occurred in response to the determination that the process is a kernel process. The at least one processor processes the interrupt in response to determining that an interrupt has occurred. The at least one processor suppresses a malware mitigation to be applied to the kernel process in response to interrupt being processed.

Interrupt context switching using dedicated processors

Examples of techniques for deferred disclaim of memory pages are described herein. An aspect includes, based on freeing of a last allocation on a first memory page, placing, by a processor, the first memory page on a deferred disclaim list. Another aspect includes, based on freeing of the last allocation on the first memory page, setting, by the processor, a first hidden flag in a first page table entry corresponding to the first memory page.

Deferred disclaim of memory pages

One embodiment provides a method for allocating assignments based on workload and physiological stress. The method comprises maintaining a queue of logged-in service representatives, and, in response to receiving a service request for a new assignment, scanning the queue to determine a subset of potential service representatives to assign the new assignment to. A corresponding workload level of each service representative of the subset is the smallest among all service representatives of the queue. For each service representative of the subset, a corresponding anxiety level is determined based on physiological sensor data captured by a wearable tracking device attached to the service representative. The new assignment is assigned to a service representative of the subset, the service representative having a corresponding anxiety level that is the smallest among all service representatives of the subset.

Call center anxiety feedback processor (CAFP) for biomarker based case assignment

A server and rack system includes at least one server rack; multiple server slots within the rack; multiple near field communication (NFC) tags; a server located within one of the server slots; an NFC reader within and powered by the server; and an entity that needs to receive locations of servers within racks of the system. Each of the multiple NFC tags is positioned on the rack so as to individually correspond, on a one-to-one basis, with each of the multiple server slots, and each of the NFC tags is programmed with an identifier of the specific rack and slot with which it corresponds, and the NFC reader is positioned such that, when power is supplied to the NFC reader by the server, the NFC reader will read the identifier that corresponds to the slot the server is located within and automatically communicate that identifier to the entity.

Growing community of inventors

Round Rock, TX, United States of America

Srinivasa Rao Muppala