| Patent application number | Description | Published |
|---|
| 20130036410 | Real-Time Temperature Sensitive Machine Level Code Compilation And Execution - Methods, apparatuses, and computer program products for real-time temperature sensitive machine level code compilation and execution are provided. Embodiments include compiling and executing, by a just-in-time (JIT) compiler, machine level code; during execution of the machine level code, determining, by the JIT compiler, an execution temperature for the machine level code, including reading temperature measurements from one or more processor temperature sensors; based on the determined execution temperature, identifying, by the JIT compiler, a portion of the machine level code that, when executed, caused temperature measurements of one or more processor temperature sensors to exceed a predetermined threshold temperature; recompiling, by the JIT compiler, the machine level code including modifying the identified portion to generate a new execution temperature that is lower than the previously determined execution temperature; and executing, by the JIT compiler, the recompiled machine level code. | 02-07-2013 |
| 20130046952 | Administering Thermal Distribution Among Memory Modules With Call Stack Frame Size Management - Administering thermal distribution among memory modules in a computing system that includes temperature sensors, where each temperature sensor measures temperature of a memory module and thermal distribution is effected by: determining, in real-time by a user-level application in dependence upon the temperature measurements of the temperature sensors, whether a memory module is overheated; if a memory module is overheated and if a current call stack frame is stored on the overheated memory module, increasing, by the user-level application, a size of the current call stack frame to fill remaining available memory space on the overheated memory module, ensuring a subsequent call stack frame is stored on a different memory module. | 02-21-2013 |
| 20130047142 | Thermal-Aware Source Code Compilation - Thermal-aware source code compilation including: receiving, by a compiler, an identification of a target computing system, the identification of the target computing system specifying temperature sensors that measure temperature of a memory module; compiling the source code into an executable application including inserting in the executable application computer program instructions for thermal-aware execution, the computer program instructions, when executed on the target computing system, carry out the steps of: retrieving temperature measurements of one or more of the target computing system's temperature sensors; determining, in real-time in dependence upon the temperature measurements, whether a memory module is overheated; if a memory module is overheated, entering a thermal-aware execution state including, for each memory allocation in the executable application, allocating memory on a different memory module than the overheated memory module; and upon the temperature sensors indicating the memory module is no longer overheated, exiting the thermal-aware execution state. | 02-21-2013 |
| 20130061007 | GENERATING CODE THAT CALLS FUNCTIONS BASED ON TYPES OF MEMORY - In an embodiment, in response to reading a declaration of a function that specifies a name of the function and a type of memory on which the function operates, the name of the function, a pointer to the function, and the type are saved to a template. In response to reading a call statement that specifies the name of the function and an identifier of an object, first code is generated. The first code, when executed, reads the pointer to the function from a virtual function table pointed to by the object, finds an entry in the virtual function table that represents the function, and reads the pointer from the entry in the virtual function table. The call statement, when executed, requests a call of the function. Second code is generated that, when executed, calls the function using the pointer read from the virtual function table. | 03-07-2013 |
| 20130067136 | Administering Thermal Distribution Among Memory Modules Of A Computing System - A computing system includes a number of memory modules and temperature sensors. Each temperature sensor measures a temperature of a memory module. In such a computing system a garbage collector during garbage collection, determines whether a temperature measurement of a temperature sensor indicates that a memory module is overheated and, if a temperature measurement of a temperature sensor indicates a memory module is overheated, the garbage collector reallocates one or more active memory regions on the overheated memory module to a non-overheated memory module. Reallocating the active memory regions includes copying contents of the active memory regions from the overheated memory module to the non-overheated memory module. | 03-14-2013 |
| 20130067155 | Memory Type-Specific Access Control Of A Field Of A Record - A computing system includes computer memory of a number of different memory types. An application program compiled for execution on the computing system controls access to a field of a record in the computer memory of the computing system by defining a record that includes one or more fields, the one or more fields including a restricted field having a specification of restricted accessibility when the restricted field is allocated in a particular memory type; allocating an instance of the record in memory of the particular memory type; and denying each attempted access of the restricted field while the record is allocated in the particular memory type. | 03-14-2013 |
| 20130103885 | ADMINISTERING THERMAL DISTRIBUTION AMONG MEMORY MODULES OF A COMPUTING SYSTEM - A computing system includes a number of memory modules and temperature sensors. Each temperature sensor measures a temperature of a memory module. In such a computing system a garbage collector during garbage collection, determines whether a temperature measurement of a temperature sensor indicates that a memory module is overheated and, if a temperature measurement of a temperature sensor indicates a memory module is overheated, the garbage collector reallocates one or more active memory regions on the overheated memory module to a non-overheated memory module. Reallocating the active memory regions includes copying contents of the active memory regions from the overheated memory module to the non-overheated memory module. | 04-25-2013 |
| 20130103919 | ADMINISTERING THERMAL DISTRIBUTION AMONG MEMORY MODULES WITH CALL STACK FRAME SIZE MANAGEMENT - Administering thermal distribution among memory modules in a computing system that includes temperature sensors, where each temperature sensor measures temperature of a memory module and thermal distribution is effected by: determining, in real-time by a user-level application in dependence upon the temperature measurements of the temperature sensors, whether a memory module is overheated; if a memory module is overheated and if a current call stack frame is stored on the overheated memory module, increasing, by the user-level application, a size of the current call stack frame to fill remaining available memory space on the overheated memory module, ensuring a subsequent call stack frame is stored on a different memory module. | 04-25-2013 |
| 20130104110 | THERMAL-AWARE SOURCE CODE COMPILATION - Thermal-aware source code compilation including: receiving, by a compiler, an identification of a target computing system, the identification of the target computing system specifying temperature sensors that measure temperature of a memory module; compiling the source code into an executable application including inserting in the executable application computer program instructions for thermal-aware execution, the computer program instructions, when executed on the target computing system, carry out the steps of: retrieving temperature measurements of one or more of the target computing system's temperature sensors; determining, in real-time in dependence upon the temperature measurements, whether a memory module is overheated; if a memory module is overheated, entering a thermal-aware execution state including, for each memory allocation in the executable application, allocating memory on a different memory module than the overheated memory module; and upon the temperature sensors indicating the memory module is no longer overheated, exiting the thermal-aware execution state. | 04-25-2013 |
| 20130104111 | REAL-TIME TEMPERATURE SENSITIVE MACHINE LEVEL CODE COMPILATION AND EXECUTION - Methods, apparatuses, and computer program products for real-time temperature sensitive machine level code compilation and execution are provided. Embodiments include compiling and executing, by a just-in-time (JIT) compiler, machine level code; during execution of the machine level code, determining, by the JIT compiler, an execution temperature for the machine level code, including reading temperature measurements from one or more processor temperature sensors; based on the determined execution temperature, identifying, by the JIT compiler, a portion of the machine level code that, when executed, caused temperature measurements of one or more processor temperature sensors to exceed a predetermined threshold temperature; recompiling, by the JIT compiler, the machine level code including modifying the identified portion to generate a new execution temperature that is lower than the previously determined execution temperature; and executing, by the JIT compiler, the recompiled machine level code. | 04-25-2013 |
| 20130110443 | GRANTING AUTHORITY IN RESPONSE TO DEFECT DETECTION | 05-02-2013 |
| 20130179731 | RECOVERING FROM A THREAD HANG - A method, computer-readable storage medium, and computer system are provided. In an embodiment, a detection is made that at least one thread, in a plurality of threads executing in a computer, has hung. A lock order of a plurality of locks used by the plurality of threads is determined. A determination is made that a first thread in the plurality of threads violates the lock order. After the determination that the first thread violates the lock order, the computer is restarted and the first thread that violates the lock order is scheduled on a reduced-speed processor. | 07-11-2013 |
| 20130275948 | Compiling Source Code For Debugging With Expanded Snapshots - Debugging source code includes: tracking, by a debugger during a debug session, duration of user examination of source code locations; providing, by the debugger to an optimizing compiler, a source code examination profile specifying source code locations examined by the user during the debug session; and receiving, by the debugger from the optimizing compiler: compiled source code for debugging, the compiled source code comprising, at each of one or more source code locations specified in the source code examination profile: a snapshot before the source code of the source code location, followed by an expanded snapshot, the expanded snapshot including computer program instructions to enable, during a debug session, examination of variable values changing during execution of the source code at the source code location; and a recording of snapshot locations and expanded snapshot locations. | 10-17-2013 |
| 20130283243 | Compiling Source Code For Debugging With Variable Value Restoration Based On Debugging User Activity - Compiling source code includes receiving, by an optimizing compiler from a debugger, a variable value modification profile that specifies locations in the source code at which variable values modified during a debug session; compiling the source code, including: inserting snapshots at one or more of the locations in the source code at which variable values were modified, each snapshot including a breakpoint; and only for each snapshot at a location in the source code at which variable values were modified: inserting, between the breakpoint and remaining source code at the location of the snapshot, a module of computer program instructions that when executed retrieves a current value of a variable and stores the current value in a register; and recording the location of each inserted snapshot; and providing, to the debugger by the optimizing compiler, the recorded locations of each inserted snapshot along with the compiled source code. | 10-24-2013 |
| 20140074312 | DECREASING THE INTERNAL TEMPERATURE OF A COMPUTER IN RESPONSE TO CORROSION - In an embodiment, a current internal corrosion level at a current time is read from an internal corrosion sensor that is internal to a computer. An internal corrosion difference is calculated between the current internal corrosion level and a previous internal corrosion level. If the internal corrosion difference is more than a first threshold amount, a first action is performed that decreases an internal temperature of the computer. | 03-13-2014 |
| 20140075418 | REPLACING CONTIGUOUS BREAKPOINTS WITH CONTROL WORDS - In an embodiment, a first execution path in a program is determined that has been executed a largest number of times, as compared to at least one other path in the program. Swapped instructions are replaced in the program with breakpoints, wherein the swapped instructions are not on the first execution path. The breakpoints that are not on the first execution path and that are contiguous to each other are replaced in the program with a control word. | 03-13-2014 |
| 20140089608 | POWER SAVINGS VIA DYNAMIC PAGE TYPE SELECTION - An operating system monitors a performance metric of a direct memory access (DMA) engine on an I/O adapter to update a translation table used during DMA operations. The translation table is used during a DMA operation to map a virtual address provided by the I/O adapter to a physical address of a data page in the memory modules. If the DMA engine is being underutilized, the operating system updates the translation table such that a virtual address maps to physical address corresponding to a memory location in a more energy efficient memory module. However, if the DMA engine is over-utilized, the operating system may update the translation table such that the data used in the DMA engine is stored in memory modules that provide quicker access times—e.g., the operating system may map virtual addresses to physical addresses in DRAM rather than phase change memory. | 03-27-2014 |
| 20140089631 | POWER SAVINGS VIA DYNAMIC PAGE TYPE SELECTION - An operating system monitors a performance metric of a direct memory access (DMA) engine on an I/O adapter to update a translation table used during DMA operations. The translation table is used during a DMA operation to map a virtual address provided by the I/O adapter to a physical address of a data page in the memory modules. If the DMA engine is being underutilized, the operating system updates the translation table such that a virtual address maps to physical address corresponding to a memory location in a more energy efficient memory module. However, if the DMA engine is over-utilized, the operating system may update the translation table such that the data used in the DMA engine is stored in memory modules that provide quicker access times—e.g., the operating system may map virtual addresses to physical addresses in DRAM rather than phase change memory. | 03-27-2014 |
| 20140165041 | CRASH NOTIFICATION BETWEEN DEBUGGERS - Crash notification between debuggers, including: initiating, by a first debugger, a first debug session of a first application; detecting, by the first debugger, an error condition in the first application; determining, by the first debugger, whether any variables utilized by the first application are related to variables utilized by a second application, wherein the second application is being debugged in a second debug session by a second debugger; and communicating, by the first debugger to a second debugger, information associated with the error condition in the first application. | 06-12-2014 |
| 20140165042 | CRASH NOTIFICATION BETWEEN DEBUGGERS - Crash notification between debuggers, including: initiating, by a first debugger, a first debug session of a first application; detecting, by the first debugger, an error condition in the first application; determining, by the first debugger, whether any variables utilized by the first application are related to variables utilized by a second application, wherein the second application is being debugged in a second debug session by a second debugger; and communicating, by the first debugger to a second debugger, information associated with the error condition in the first application. | 06-12-2014 |
| 20140173140 | DETECTING FIRMWARE VERSION FOR AN INPUT/OUTPUT ADAPTER - An indication of a version of a firmware stored in an input/output adapter may be provided by a method that includes detecting whether a first pin is connected to an external circuit, detecting whether a second pin is unconnected to an external circuit, and causing the indication to be provided if the first pin is connected and the second pin is unconnected. The indication may be provided on the first pin. The first pin may include a power supply pin and the indication may be an average rate of power supplied to the input/output adapter. | 06-19-2014 |
| 20140201352 | SELECTIVE LOGGING OF NETWORK REQUESTS BASED ON SUBSETS OF THE PROGRAM THAT WERE EXECUTED - In an embodiment, in response to detecting a request by a program to access a network, if the request is the first time that the program requests to access the network, a subset of instructions is replaced in the program with supervisor call instructions. The supervisor call instructions cause respective interrupts of execution of the program. In response to each of the respective interrupts of execution of the program, the supervisor call instructions that caused the respective interrupts are replaced with the respective swapped instructions, and if a number of the respective interrupts of execution exceed a trap threshold, all remaining of the respective swapped instructions are stored to the program. | 07-17-2014 |
| 20140201718 | ANALYZING CONCURRENT DEBUGGING SESSIONS - Methods, apparatuses, and computer program products for analyzing concurrent debugging sessions are provided. Embodiments include a first debugger initiating a conditional breakpoint in a first debug session of a first application. The conditional breakpoint stops execution of the first application based on a condition of a value of a variable of a second application. Embodiments also include the first debugger requesting and receiving the value of the variable from a second debugger. Embodiments also include the first debugger evaluating the condition of the conditional breakpoint based on the received value of the variable. | 07-17-2014 |
| 20140201719 | ANALYZING CONCURRENT DEBUGGING SESSIONS - Methods, apparatuses, and computer program products for analyzing concurrent debugging sessions are provided. Embodiments include a first debugger initiating a conditional breakpoint in a first debug session of a first application. The conditional breakpoint stops execution of the first application based on a condition of a value of a variable of a second application. Embodiments also include the first debugger requesting and receiving the value of the variable from a second debugger. Embodiments also include the first debugger evaluating the condition of the conditional breakpoint based on the received value of the variable. | 07-17-2014 |
| 20150058593 | MERGING DIRECT MEMORY ACCESS WINDOWS - A computing device may merge two translation tables used when performing a DMA operation into a single, combined translation table. To merge the translation tables, the computing device may update a register in the IOMMU to include a pointer to the combined translation table. In addition, the IOMMU may clear one of the registers from having a pointer to one of the merged translation table. Doing so means the entries in this translation table are now no longer assigned. The IOMMU may update the register with the pointer to the combined translation table to include the unassigned entries in the combined translation table. In this manner, the entries from the two translation tables are merged into the single, combined table. The combined translation table may be owned or assigned to a service provider that originally owned one of the merged translation tables or to a completely different service provider. | 02-26-2015 |
| 20150058594 | SPLITTING DIRECT MEMORY ACCESS WINDOWS - A computing device may split a translation table used when performing a DMA operation into two different translation tables. To split the translation table, the computing device may update the registers in the IOMMU to include pointers to the two different translation tables. For example, the IOMMU may update one register to point to the same starting address as the original translation table but assign a shorter length (i.e., fewer entries) to that table. The extra entries may then be used to form the other translation table by adding a new pointer to one of the IOMMU registers. The two translation tables may be owned by the same service provider or two different service providers. Alternatively, the computing device may assign the two tables to the same service provider which in turn assigns the tables to respective client devices executed by the service provider. | 02-26-2015 |
| 20150058596 | MERGING DIRECT MEMORY ACCESS WINDOWS - A computing device may merge two translation tables used when performing a DMA operation into a single, combined translation table. To merge the translation tables, the computing device may update a register in the IOMMU to include a pointer to the combined translation table. In addition, the IOMMU may clear one of the registers from having a pointer to one of the merged translation table. Doing so means the entries in this translation table are now no longer assigned. The IOMMU may update the register with the pointer to the combined translation table to include the unassigned entries in the combined translation table. In this manner, the entries from the two translation tables are merged into the single, combined table. The combined translation table may be owned or assigned to a service provider that originally owned one of the merged translation tables or to a completely different service provider. | 02-26-2015 |
| 20150058597 | SPLITTING DIRECT MEMORY ACCESS WINDOWS - A computing device may split a translation table used when performing a DMA operation into two different translation tables. To split the translation table, the computing device may update the registers in the IOMMU to include pointers to the two different translation tables. For example, the IOMMU may update one register to point to the same starting address as the original translation table but assign a shorter length (i.e., fewer entries) to that table. The extra entries may then be used to form the other translation table by adding a new pointer to one of the IOMMU registers. The two translation tables may be owned by the same service provider or two different service providers. Alternatively, the computing device may assign the two tables to the same service provider which in turn assigns the tables to respective client devices executed by the service provider. | 02-26-2015 |
