Patent Application
20060165094
This invention relates to the art of computer system emulation and, more particularly, to a computer system emulator in which the functions normally performed by the hardware in a central processor unit are emulated by a software program.
Users of mainframe computers running a proprietary operating system may have a very large investment in proprietary application software and, further, may be comfortable with using the application software because it has been developed and improved over a period of years, even decades, to achieve a very high degree of reliability and efficiency.
As manufacturers of very fast and powerful commodity processors continue to improve the capabilities of their products, it has become practical to emulate the proprietary operating systems of powerful older computers such that the manufacturers of the older computers can provide new systems which allow the users to continue to use their highly-regarded proprietary software by emulating the older or “legacy” computer and in particular the central processing units of the legacy system. Accordingly, computer system manufacturers are developing such emulator systems for the users of their legacy systems, and the emulation process used by a given system manufacturer is itself subject to ongoing refinement and increases in efficiency and reliability.
In one specific state-of-the-art example, a 64-bit Itanium Intel processor is used to emulate the Bull DPS 9000 36-bit memory space and the GCOS 8 instruction set of the Bull DPS 9000. Within the memory space of the emulator, the 36-bit word of the DPS 9000 is stored right justified (least significant bits) in the least significant 36 bits of the “host” (Itanium) 64-bit word. The upper 28 bits of the 64-bit word are typically zero for “legacy” code. Sometimes, certain specific bits in the upper 28 bits of the containing word are used as flags or for other temporary purposes, but in normal operation these bits are usually zero and in any case are typically viewed by older programs in the “emulated” view of the world as being non-existent. That is, only the emulation program itself uses these bits.
For some purposes, such as providing new or more direct communication with programs or services running in the 64-bit system, it is advantageous to provide the emulated system with full access to the entire “containing” word, which in this case is the 64-bit containing word, for purposes of both loading or storing the 64-bit word from the viewpoint of the emulated software into visible space within the 36-bit environment. The capability of viewing and manipulating 64 bits can also be used to improve the machine architecture by such expedients as adding new opcodes with more functionality, increasing the address space, or other similar things which can utilize more bits in the instruction or data words.
It is also possible to enhance the basic instruction set of the newly defined system beyond what was provided in the older legacy system. Opcodes can be selected and their functionality can be defined in the emulation program which extends the instruction set beyond, or even far beyond the functionality of the original “legacy” system. This can be done by either enhancing or enriching the functionality of the opcodes already defined on the legacy system, or by utilizing previously unused opcodes to specify new functionality.
The present invention is directed to achieving this end, that is, to enhance the instruction set of the new emulated system to include functionality which encompasses that previously performed by the operating system using only the older legacy instructions. The purpose of the invention is to increase performance of the emulated system, and in particular to increase the performance of programs which make heavy or large use of the operating system to perform the processing required by that program.
It is therefore a broad object of this invention to increase the performance of a computer system by defining new opcodes which enhance the instruction set of an emulated central processor unit with these new opcodes implementing functionality previously performed by a sequence of legacy instructions inside the code of the operating system.
It is another broad object of this invention to improve the performance of the emulated central processor unit and also the performance of the emulated system by writing programs or subroutines which will run in machine code native on the host system to perform the functions previously performed by the operating system in a manner which surpasses the performance that could be achieved by emulating a sequence of the original legacy instructions which performed the original function. Typically one newly defined instruction will replace the functionality of a section of operating system code that would require the processing of many legacy instructions.
It is a more particular object of this invention to provide two types of new opcodes. The first type is to newly utilize previously unused opcodes or other encoding of the original legacy instruction word to specify and provide functionality which was previously provided by a sequence of instructions in the original legacy operating system code. The second type is to utilize unused bits in the host instruction word to define new instructions and new functionality that replaces operating system code. Both of these instruction formats allow for increased performance by allowing for host system machine code to replace a sequence of emulated legacy instructions with a single instruction. Use of previously unused bits in the host system word, also allows for the overall encoding of the instructions to be less compact which in turn allows for faster decode and quicker determination of the precise work to be done by the code of the emulation system.
Briefly, these and other objects of the invention are achieved by providing new opcodes or other means of encoding the specification of new functionality into the instruction set of an emulated processing unit in an emulated computer system. The newly added opcodes specifically implement functionality previously performed by the legacy operating system with performance improvement of the overall system being achieved through more efficient coding of operating system functions in a language or machine code native to the host system, instead of the prior art approach of interpreting a series of legacy instructions. This approach enables more efficient coding with resulting improved performance of the overall computer system.
The subject matter of the invention is particularly pointed out and distinctly claimed in the concluding portion of the specification. The invention, however, both as to organization and method of operation, may best be understood by reference to the following description taken in conjunction with the subjoined claims and the accompanying drawing of which:
The target operating system reference space 15 also contains suitable information about the interconnection and interoperation among the various target system elements and components and a complete directory of the target system operating system commands which includes information on the steps the host system must take to “execute” each target system command in a program originally prepared to run on a physical machine using the target system operating system. It can loosely be considered that, to the extent that the target system 1 can be said to “exist” at all, it is in the target operating system reference space 15 of the host system memory 12. Thus, an emulator program running on the host system 2 can replicate all the operations of an application program written in the target system operating system as if the application program were running on a physical target system.
Referring now to
Any number of pieces of the operating system can be rewritten and compiled and run as native code on the host central processing unit. The original operating system code is modified such that the large sequences of instructions previously used to implement a function are replaced by new single instructions, possibly with arguments, and the code for performing these functions is now in essence part of the emulation system rather than the operating system, although of course conceptually it can be viewed as remaining part of the operating system.
The processing in
In a current state-of-the-art example chosen to illustrate the invention, a 64-bit Itanium Intel processor is used to emulate the Bull DPS9000 36-bit memory space and the instruction set of the DPS9000 with its proprietary GCOS 8 operating system. Within the memory space of the emulator, the 36-bit word of the DPS9000 is stored right justified in the least significant 36 bits of the “host” (Itanium) 64-bit word during the emulation process. The upper 28 bits of the 64-bit word are typically zero; however, sometimes, certain specific bits in the “upper” 28 bits of the “containing” word are used as flags or for other temporary purposes. Within the instruction set of the legacy DPS9000 the numbers 3 and 4 (among others) were unused in the legacy instruction set, and a fault was generated if they were encountered. For purposes of improving the performance of the emulated computer system, opcodes 3 and 4 were chosen as a mechanism to implement a call to host native emulation code with the address field of the instruction for opcode 3 indicating which of many operating system functions was being invoked. When opcode number 3 is encountered by the emulation system, one of many new functions as specified by the address field is invoked and processing of that single instruction as part the host native emulation code replaces the processing of many instructions in the original operating system code.
The newly defined instructions for specifying operating system functions are called “hybrid” instructions and these instructions differ from typical instructions from the legacy instruction set. Hybrid instructions would typically be more complex than typical instructions from the legacy instruction set. Hybrid instructions have full access to the resources of the host machine for performing their services, which the instructions of the legacy instruction set typically do not utilize. Hybrid instructions may access files or utilize communication functions of the host operating system. Hybrid functions may use multiple threads to achieve higher degrees of parallelism than can be achieved by attacking a problem with a sequence of legacy instructions. Hybrid functions can utilize recursion and resources for calls and returns that are native to the host system hardware. Hybrid instructions can use internal caching of data and internal variable to avoid recalculation of certain data when that possibility is recognized.
Hybrid functions may be called from only one place in the operating system. That is, since they are replacing a specific section of code, they may be called or invoked only from that specific place in the new source code for the legacy operating system. Thus, the hybrid instructions do not need to be designed and tailored for general use and since they may be invoked from only one place they can be optimized for most efficient operation in precisely the environment from which they are invoked.
Within the functions of the operating system it is possible that the code implementing an operating system function may itself require calls back to the operating system itself. That is, the function may require calls to other operating system functions which may or may not also contain or be implemented with other “hybrid” instructions. This requires that the mechanism of the invention may for some functions require a method or procedure for the saving and subsequent restoring of the program state after the return from the operating system calls. The mechanism must allow for the stacking of program states as the code could be called from multiple places and lower level functions could also themselves make calls to the same function. If the program is written in a higher level language such as C, or C++ then the internal state of the C program must itself be saved as part of the program state so that if the emulated processor is dispatched to another task on behalf of another user, the variables for the processing of the first program must be restored when the first program is resumed.
It may also be required that the implementation of the program for performing the operating system function be given access to the program visible registers, or other internal variables and structures of the emulation program. This can be accomplished in many ways well known in the state of the art such as passing pointers to variable structures as part of the call to the function implementing the program, or by identifying certain structures and variables as being contained in global or common memory space.
It is noted that the invention which replaces sections of operating system code with an alternative implementation in machine code of the host system is not limited to exact replacement of the function, nor does the function have to be precisely or even approximately equivalent to the original function. The application of the invention does not preclude continued improvement in algorithms or in approach to the overall operating system function as legacy code is migrated into instructions which invoke machine code of the host system.
While the principles of the invention have now been made clear in an illustrative embodiment, there will be immediately obvious to those skilled in the art many modifications of structure, arrangements, proportions, the elements, materials, and components, used in the practice of the invention which are particularly adapted for specific environments and operating requirements without departing from those principles.
