摘要:
A data processing method for storing groups of related information in a storage subsystem of a data processing system in which the storage subsystem includes one or more storage devices having a plurality of block addressable storage locations (blocks or sectors) each of which stores a predetermined fixed number of bytes of said information. The method includes the step of establishing allocatable increments of storage, called physical partitions, which comprise a predetermined number of contiguous addressable blocks, and initially allocating, in response to a request to the operating system, a preselected number of partitions for each group of related information, where the partitions in each group are not necessarily physically contiguous and where the number that is selected is the minimum number of partitions required to store the group of related information. The method further includes the step of automatically allocating one or more additional partitions, located physically on any device, to a previously allocated group of related partitions, dynamically in response to another request of the operating system. A group of partitions that have been allocated to store a group of related information, such as individual files that are related in a hierarchical file system, is called a "Logical Volume". Logical Volumes may be further combined for administration and data management reasons into a "Volume Group".
摘要:
A virtual memory address translation mechanism is provided for converting virtual memory addresses provided by a CPU into real memory addresses within page frames in a large hierachial memories wherein the real memory space is substantially smaller than the scope of the virtual memory. The conversion or translation mechanism includes a combined table in the memory which includes a first list covering the respective virtual address of each memory address (Inverted Page or IPT) and a second list connecting each of a plurality of hashed addresses with a predetermined initial virtual address of a linked group of virtual addresses, each of which when hashed produces the connected hashed address (Hashed Addressed Table, HAT). The system also has means for hashing a selected virtual address to produce a hashed address. Also included in apparatus for sequentially searching through the linked group of virtual addresses in the combined table until a selected virtual address is located as well as apparatus responsive to the location of a particular selected virtual address for accessing from the first list, the real memory address of the located virtual address.
摘要:
A cache for use with input/output devices attached to an input/output bus. Requests for access to system memory by an input/output device pass through the cache. Access authority is checked to determine whether an input/output device is authorised to access that particular page. If it is not, access is denied. Each input/output device has access to a portion of the cache, so that activity by one device will not interfere with activity by another.
摘要:
A cache for use with input/output devices attached to an input/output bus. Requests for access to system memory by an input/output device pass through the cache. Access authority is checked to determine whether an input/output device is authorised to access that particular page. If it is not, access is denied. Each input/output device has access to a portion of the cache, so that activity by one device will not interfere with activity by another.
摘要:
Multiple external page table (XPT) structure geometries provide efficient organisation and management of paged, segmented virtual memory systems. First and second small XPT geometries are provided for file mapping of files of a predetermined size for code segments, and file mapping of files exceeding the predetermined size for process private segments, respectively. A large XPT geometry is for working storage segments containing shared library routines. A uniform addressing scheme accesses page state information for second small and large XPT geometries having root blocks. Selected XPT structure direct blocks are pre-allocated for large XPT structures for address ranges to be referenced to reduce page faults.
摘要:
A memory address translation and related control system for performing the dual functions of converting virtual memory addresses generated by the CPU into real memory addresses in a highly efficient and versatile manner and for controlling certain memory functions such as journalling. The address translation function comprises two steps, the first comprising converting the virtual address into a second virtual address or an effective address and finally the step of converting the effective address into a real memory address. The first step utilizes a set of special registers addressable by a small field of the CPU generated virtual address which converts the virtual address into an expanded form. The second or effective address is then used as the subject of the second or address translation step. To greatly enhance the translation of frequently used virtual addresses, a special set of translation tables referred to herein as a Translation Look-Aside, Buffer (TLB) contain current effective to real address translations for use where frequently referenced addresses are requested. The TLBs are addressed using a subset of the effective address whereupon the contents of the addressed TLB is examined for a match with the effective address. If the addresses match a successful address translation is possible and the real address stored in the address field of the TLB is available for system use. If the desired effective address is not present in the TLB, the page frame tables stored in main memory are accessed and searched for the desired effective address and if found the associated real address is accessed. Further a special data field is provided in both the TLBs and the page frame tables in main memory wherein a bit is provided for each line in the referenced page at a given effective to real address translation which bits may be used to indicate when a line of data has been accessed or altered.
摘要:
Multiple external page table (XPT) structure geometries provide efficient organisation and management of paged, segmented virtual memory systems. First and second small XPT geometries are provided for file mapping of files of a predetermined size for code segments, and file mapping of files exceeding the predetermined size for process private segments, respectively. A large XPT geometry is for working storage segments containing shared library routines. A uniform addressing scheme accesses page state information for second small and large XPT geometries having root blocks. Selected XPT structure direct blocks are pre-allocated for large XPT structures for address ranges to be referenced to reduce page faults.