CA1153824A - Address generating mechanism for multiple virtual spaces - Google Patents
Address generating mechanism for multiple virtual spacesInfo
- Publication number
- CA1153824A CA1153824A CA000368423A CA368423A CA1153824A CA 1153824 A CA1153824 A CA 1153824A CA 000368423 A CA000368423 A CA 000368423A CA 368423 A CA368423 A CA 368423A CA 1153824 A CA1153824 A CA 1153824A
- Authority
- CA
- Canada
- Prior art keywords
- address
- register
- arcv
- program
- instruction
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Expired
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F12/00—Accessing, addressing or allocating within memory systems or architectures
- G06F12/14—Protection against unauthorised use of memory or access to memory
- G06F12/1458—Protection against unauthorised use of memory or access to memory by checking the subject access rights
- G06F12/1466—Key-lock mechanism
- G06F12/1475—Key-lock mechanism in a virtual system, e.g. with translation means
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F12/00—Accessing, addressing or allocating within memory systems or architectures
- G06F12/02—Addressing or allocation; Relocation
- G06F12/06—Addressing a physical block of locations, e.g. base addressing, module addressing, memory dedication
- G06F12/0615—Address space extension
- G06F12/0623—Address space extension for memory modules
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F12/00—Accessing, addressing or allocating within memory systems or architectures
- G06F12/02—Addressing or allocation; Relocation
- G06F12/08—Addressing or allocation; Relocation in hierarchically structured memory systems, e.g. virtual memory systems
- G06F12/10—Address translation
- G06F12/1027—Address translation using associative or pseudo-associative address translation means, e.g. translation look-aside buffer [TLB]
- G06F12/1036—Address translation using associative or pseudo-associative address translation means, e.g. translation look-aside buffer [TLB] for multiple virtual address spaces, e.g. segmentation
Abstract
ABSTRACT OF THE DISCLOSURE
The detailed embodiment associates access registers (AR's) with the general purpose registers (GPR's) in a data processor. The AR's are each loaded with a unique STD (segment table descriptor). The STD comprises a segment table address in main storage and a segment table length field. There are 15 AR's associated respectively with 15 GPR's in a processor to define a subset of up to 15 data address spaces. The STD in an AR is selected for address translation when the associated GPR is selected as a storage operand base register, such as being the GPR selected by the B-field in an IBM System/370 instruction. The invention allows each AR to specify that it does not use the STD in its associated AR to define its data address space, but instead uses the STD in the program address space AR.
However, the STD content of an AR is not selected for an address translation if the associated GPR is selected for a purpose other than as a storage operand base register, such as if a GPR is selected as an index (X) register or as a data source or sink register (R) for an instruction. A sixteenth AR may be provided to define and control the executing program address space, which may also contain data.
The embodiment obtains authority and other control for access to and use of the content in each address space by also associating an AR Control Vector (ARCV) register with each AR.
The detailed embodiment associates access registers (AR's) with the general purpose registers (GPR's) in a data processor. The AR's are each loaded with a unique STD (segment table descriptor). The STD comprises a segment table address in main storage and a segment table length field. There are 15 AR's associated respectively with 15 GPR's in a processor to define a subset of up to 15 data address spaces. The STD in an AR is selected for address translation when the associated GPR is selected as a storage operand base register, such as being the GPR selected by the B-field in an IBM System/370 instruction. The invention allows each AR to specify that it does not use the STD in its associated AR to define its data address space, but instead uses the STD in the program address space AR.
However, the STD content of an AR is not selected for an address translation if the associated GPR is selected for a purpose other than as a storage operand base register, such as if a GPR is selected as an index (X) register or as a data source or sink register (R) for an instruction. A sixteenth AR may be provided to define and control the executing program address space, which may also contain data.
The embodiment obtains authority and other control for access to and use of the content in each address space by also associating an AR Control Vector (ARCV) register with each AR.
Description
~9-79-012 ADDRESS GENERATING ~ECHANISM FOR MULTIPLE VIRTUAL SPA OE S
FIELD OF THE INVENTION
The invention rela~es to a method and means for enabling a program being executed in a data processing system to have concurrent access to multiple virtual address spaces.
DESCRIPTION OF THE PRIOR ART
Data processing systems using virtual addressing in multiple virtual address spaces are well known, such as the IB~ System/370 systems using MVS system con~rol ~rogramming. Such systems define each address space as having linear addressability from 0 to (224-l) bytes using a segment table, in which each entry contains an address of a page table, and each page table defines a plurality of page entries in the address space. Each page entry may contain the address of any assigned page frame in main storage. Each address space is referred to by a pointer called a segment table origin (STO) which locates the address space segment table in main storage. In current System/
370 architecture, each address space has up to 16 mega-bytes (224) of linear virtual addressability beginning at address zero. In current System~370 MVS Operation, an unlimited number of address spaces are a~ailable;
different address spaces are assigned to different users of the system, and the address spaces obtain isolation between the users of the system, i.e. one user can only address within his own address space and is isolated from the address spaces of the otner users. The exception is when an address space, or a part thereof, is defined in common among all users, which is the subject of U.S.A. Patents 4,096,573 *Registered Trade Mark ~9-79-012 ~5.3~4 entitled "DLAT Synonym Control Means ~or Common Portions Of All Address Spaces" to A. R. ~eller et al, and 4,136,385 entitled "Synonym Control Means For Multiple Virtual Storage Systems to P. M. Gannon et al, both assigned to the same assignee as the subject appli-cation. The result is either total isolation between address spaces with no inter-space access permitted or total accessibility of an address space, or a part thereof, by all users.
The IBM Series/l Data Processing System supports concurrent accessing in plural address spaces in a different manner from the subject invention. The Series/l processor relates the source operand in an instruction to one address space, relates any sink operand in the instruction to a second address space, and relates the fetching of instructions to a third address space. On the other hand, the subject invention does not use the source or sink characteristic of an instruction operand to relate it to an address space (AS). Instead, the subject invention uses an operand designation of a general purpose register (GPR) as a base address register to select an address space associated with the GPR. As a result, the subject invention can easily switch address spaces for any operand without the intervention of any address space switching instruction, e.g. source operands in succeeding instructions may be in different address spaces. The following patents owned by the assignee of the subject application relate to the address space control used in the IBM Series/l architecture:
U.S.A. Patent No. 4,035,779 entitled "Supervisor Address Key Control System" by R. E. Birney et al, discloses a level status register (LSR) that has an (alternate protect mode~ APM bit, and a supervisor :1~ 53~
state bit; they control a limited supervlsor addressa-bility without disturbing a user address key contained in the user key register (U~R). When both bits are set, each instruction fetch must be in the (O key) supervisor key area, while each operand access is in the user key area identified by the current key in the UKR. The supervisor can never access any other key area in the main memory. When only the supervisor bit is on, all instruction and operand accesses are in the supervisor key area only. If both bits are off, only user accesses are allowed.
U.S.A. Patent No. 4,037,207 entitled "System For Controlling Address Keys Under Interrupt Conditions"
by R. E. Birney et al, discloses each processor AAK
(ignoring each I/O AAK) being inputted as the processor last key saved (i.e. LKSA) in a processor last AAK
register. On the occurrence of a machine check or program check interrupt, the last AAK register has its input blocked (so that it can retain the LKSA).
Then the LKSA is outgated to ~he source operand key section (OPlK) of the AKR and the OP2K and ISK are set to the supervisor key (so that the supervisor program can be initiated by the interrupt with LKSA addressa-bility which probably contains the cause of the interrupt.
U.S.A. Patent No. 4,037,214 entitled "Key Register Controlled Accessing System" by R. E. Birney et al, discloses active address key (AAK) select circuits to correlate key register sections respectively containing address space identifiers in a processor AKR, and in I/O subchannels, with machine-identifiable types of storage accesses available in the main memory. Different machine-identifiable access types includer for example, the instruction fetch, source operand fetch, sink operand fetch/store, and I/O subchannel access.
P0~7s-012 1~3829L
U.S.A. Patent No. 4,037,215 entitled "Key Controlled Address Relocation Translation System" by R. ~. Birney et al, discloses an address relocation translation system combined with a plurality of AKR sections, and AAK
select circuits to translate machine-selected address keys and program selected logical addresses in order to manage the space in a physical main memory.
U.S.A. Patent No. 4,038,645 entitled "Non-Trans-latable Storage Protection Control System" by R. E.
Birney et al, discloses the AKR sections combined with the AAK select circuits with a conventional type of storage protect key memory control to obtain address sub-range protectability among different machine-identified storage-access-types being accessed in a main memory. Also, address sub-range(s~ associated with a common key are made shareable by users of the system; and all I/O accesses override a sub-range read-only protect flag.
U.S.A. ~atent No. 4,042,913 entitled "Address Key Register Load/Store Instruction System" by R. E.
Birney et al, discloses instruction operated controls for the loading and storing of address keys (i.e.
address space identifiers) into or from an address key register (AKR) to control addressabilities within a system.
U.S.A. Patent No. 4,050,060 entitled "Equated Operand Address Space Control System" by R. E. Birney et al, discloses equate operand spaces (EOS) control circuits, which when enabled, force all operand accesses (i.e. source and sink) to occur within the sink operand addressability provided by the OP2K section in the AKR, even though the AKR provides a different addressability with its OPlK section for source operand accesses. When ~09-79-012 ~S~382~
the EOS feature is disabled, the source operand addressability provided by OPlK section is used when fetching source operands.
SUMMARY OF THE INV~NTION
The subject invention is concerned with obtaining precise control over the degree of isolation or shareability among subsets of address spaces in a system in relation to an executable program. The shareability control is obtained in a processor by means of a unique architectural method and its means of implementations, which is compatible with current IBM System/370 architecture.
Objects of this invention are to provide processor means:
1. Which can directly address named collections of data in different address spaces with relative byte addressing.
FIELD OF THE INVENTION
The invention rela~es to a method and means for enabling a program being executed in a data processing system to have concurrent access to multiple virtual address spaces.
DESCRIPTION OF THE PRIOR ART
Data processing systems using virtual addressing in multiple virtual address spaces are well known, such as the IB~ System/370 systems using MVS system con~rol ~rogramming. Such systems define each address space as having linear addressability from 0 to (224-l) bytes using a segment table, in which each entry contains an address of a page table, and each page table defines a plurality of page entries in the address space. Each page entry may contain the address of any assigned page frame in main storage. Each address space is referred to by a pointer called a segment table origin (STO) which locates the address space segment table in main storage. In current System/
370 architecture, each address space has up to 16 mega-bytes (224) of linear virtual addressability beginning at address zero. In current System~370 MVS Operation, an unlimited number of address spaces are a~ailable;
different address spaces are assigned to different users of the system, and the address spaces obtain isolation between the users of the system, i.e. one user can only address within his own address space and is isolated from the address spaces of the otner users. The exception is when an address space, or a part thereof, is defined in common among all users, which is the subject of U.S.A. Patents 4,096,573 *Registered Trade Mark ~9-79-012 ~5.3~4 entitled "DLAT Synonym Control Means ~or Common Portions Of All Address Spaces" to A. R. ~eller et al, and 4,136,385 entitled "Synonym Control Means For Multiple Virtual Storage Systems to P. M. Gannon et al, both assigned to the same assignee as the subject appli-cation. The result is either total isolation between address spaces with no inter-space access permitted or total accessibility of an address space, or a part thereof, by all users.
The IBM Series/l Data Processing System supports concurrent accessing in plural address spaces in a different manner from the subject invention. The Series/l processor relates the source operand in an instruction to one address space, relates any sink operand in the instruction to a second address space, and relates the fetching of instructions to a third address space. On the other hand, the subject invention does not use the source or sink characteristic of an instruction operand to relate it to an address space (AS). Instead, the subject invention uses an operand designation of a general purpose register (GPR) as a base address register to select an address space associated with the GPR. As a result, the subject invention can easily switch address spaces for any operand without the intervention of any address space switching instruction, e.g. source operands in succeeding instructions may be in different address spaces. The following patents owned by the assignee of the subject application relate to the address space control used in the IBM Series/l architecture:
U.S.A. Patent No. 4,035,779 entitled "Supervisor Address Key Control System" by R. E. Birney et al, discloses a level status register (LSR) that has an (alternate protect mode~ APM bit, and a supervisor :1~ 53~
state bit; they control a limited supervlsor addressa-bility without disturbing a user address key contained in the user key register (U~R). When both bits are set, each instruction fetch must be in the (O key) supervisor key area, while each operand access is in the user key area identified by the current key in the UKR. The supervisor can never access any other key area in the main memory. When only the supervisor bit is on, all instruction and operand accesses are in the supervisor key area only. If both bits are off, only user accesses are allowed.
U.S.A. Patent No. 4,037,207 entitled "System For Controlling Address Keys Under Interrupt Conditions"
by R. E. Birney et al, discloses each processor AAK
(ignoring each I/O AAK) being inputted as the processor last key saved (i.e. LKSA) in a processor last AAK
register. On the occurrence of a machine check or program check interrupt, the last AAK register has its input blocked (so that it can retain the LKSA).
Then the LKSA is outgated to ~he source operand key section (OPlK) of the AKR and the OP2K and ISK are set to the supervisor key (so that the supervisor program can be initiated by the interrupt with LKSA addressa-bility which probably contains the cause of the interrupt.
U.S.A. Patent No. 4,037,214 entitled "Key Register Controlled Accessing System" by R. E. Birney et al, discloses active address key (AAK) select circuits to correlate key register sections respectively containing address space identifiers in a processor AKR, and in I/O subchannels, with machine-identifiable types of storage accesses available in the main memory. Different machine-identifiable access types includer for example, the instruction fetch, source operand fetch, sink operand fetch/store, and I/O subchannel access.
P0~7s-012 1~3829L
U.S.A. Patent No. 4,037,215 entitled "Key Controlled Address Relocation Translation System" by R. ~. Birney et al, discloses an address relocation translation system combined with a plurality of AKR sections, and AAK
select circuits to translate machine-selected address keys and program selected logical addresses in order to manage the space in a physical main memory.
U.S.A. Patent No. 4,038,645 entitled "Non-Trans-latable Storage Protection Control System" by R. E.
Birney et al, discloses the AKR sections combined with the AAK select circuits with a conventional type of storage protect key memory control to obtain address sub-range protectability among different machine-identified storage-access-types being accessed in a main memory. Also, address sub-range(s~ associated with a common key are made shareable by users of the system; and all I/O accesses override a sub-range read-only protect flag.
U.S.A. ~atent No. 4,042,913 entitled "Address Key Register Load/Store Instruction System" by R. E.
Birney et al, discloses instruction operated controls for the loading and storing of address keys (i.e.
address space identifiers) into or from an address key register (AKR) to control addressabilities within a system.
U.S.A. Patent No. 4,050,060 entitled "Equated Operand Address Space Control System" by R. E. Birney et al, discloses equate operand spaces (EOS) control circuits, which when enabled, force all operand accesses (i.e. source and sink) to occur within the sink operand addressability provided by the OP2K section in the AKR, even though the AKR provides a different addressability with its OPlK section for source operand accesses. When ~09-79-012 ~S~382~
the EOS feature is disabled, the source operand addressability provided by OPlK section is used when fetching source operands.
SUMMARY OF THE INV~NTION
The subject invention is concerned with obtaining precise control over the degree of isolation or shareability among subsets of address spaces in a system in relation to an executable program. The shareability control is obtained in a processor by means of a unique architectural method and its means of implementations, which is compatible with current IBM System/370 architecture.
Objects of this invention are to provide processor means:
1. Which can directly address named collections of data in different address spaces with relative byte addressing.
2. With direct addressability to data sets which may be handled as named collections by both processor storage and I/O storage.
3. With concurrent access to a subset of plural address spaces.
4. Which is compatible with existing data processing system architecture to permit the execution of both programs designed for existing architecture and for the new archi~ecture of the subject invention.
Po9-79-~.2 ~3~
Po9-79-~.2 ~3~
5. ~or slmultaneously having executable access to a subset of address spaces which are easy to use in a direct manner by user application programs as well as by existing access method programs,
6. Which at any one time has concurrent access to an executing program's address space and a subset of data address spaces, e.g. up to 15 data spaces.
7. With a separate access register associated with each base register that defines relative by~e addressing in a concurrently accessible address space.
8. To support a maximum number of currently accessible data spaces dependent upon the number of general purpose registers used by the processor architecture.
9. To support relative byte addressing in each address space having up to 2N linear byte addresses, in which N is the number of bits in a general purpose register.
10. That can at any time: change the subset of concurrently accessible address spaces by reloading the access registers for accessing different address spaces.
ll. In which program address spaces may share specified data address spaces but may not share other data address spaces, and may be allowed or prohibited from sharing access to each otherls program address space~
~0~--79-012 3~
--7~
12. Having authority controls on each access register to permit different progra~s to have different authorities to access the same data address space.
13. Having descriptor fields with each access regis~er defining attributes of the associated address space, or the data elements within the space, which determine the operations of instruc-tions relative to that space.
14. With an addressing mechanism which need not connect a data address space to a program until access is required by the program.
15. ~ith an addressing mechanism which can support data address spaces independent of any 1~ program when the program is made not executable in the processor, e.g. when the program is not dispatched.
16. Which allows an access register to be disabled to permit its general purpose register to provide relative addressing of data in the program address space, but also allows the access register to be enabled so that its general purpose register controls relative addressing in a data address space specified by its access register separately from the program address space.
17. Which allows access registers to be disabled while the processor is executing programs under current architecture to eliminate the need for loading the access registers prior to execution of such programs.
1~5~82~
18. Having authority controls with each program space separately authorizing programs in that program space to use specified privileged instruc-tions.
19. Which can have sequential operand requests flexibly access different address spaces identi-fied in a plurality of access registers. No intervention is needed by any address space switching instruction, and without any required relationship between any address space and the source or sink state of the requesting operand in an instruction.
It is therefore a feature of this invention to control access to data in a subset of address spaces by an executing program without requiring any inter-ruption in the execution of the program when it changes its accessing of data to a different address space in the subset. The manner of access can be varied among the different address spaces; for example, the program may be permitted to read (but not write) data from one address space, write (but not read) data in a second address space, and both read and write data in a third address space.
The invention associates access registers (AR's) with the general purpose registers (GPR's) in a data processor. The AR's may each be loaded with an address space identifier, for example, a unique STD (segment table descriptor). The STD comprises a STO field and a segment table length field. Thus, there may be 16 AR's 3~ associated respectively with 16 GPR's in a processor.
The address s~ace identifier in an AR is selected for address translation when the associated GPR is selected as a storage operand base register, such as being the GPR selected by the B-field in an IBM System/370 instruc-_g_ tion. However, the address space identifier content of an AR is not selected for an address translation if the associated GPR is selected for a purpose other than as a storage operand base register, such as if a GPR is selected as an index (X) register or as a data source or sink register (R) for an instruction.
The invention also obtains authority control for each program's access to each address space by also associating an AR Control Vector (ARCV) register wi~h each AR to control the type of access permitted to the associated address space by an executing program.
A special field in either an AR or its associated ARCV may indicate whether the AR content is enabled and thereby contains a useable address space identifier; or is disabled and the address space identifier must be obtained elsewhere. This special field with each AR
permits the GPR of any disabled AR to specify a base value associated with the address space defined by another AR; for example, a disabled GPR may specify a base address for data in the program address space defined by ARO.
Thus, the invention controls which address space may contain an executing program and which other address spaces may be accessed for data by the executing program.
This may be done by permanently assigning one of the A~'s to contain the STD of ~he executing program (e.g. AR0).
Then the other AR's may be permanently available to con~ain the address space identifiers of other address spaces which have data accessible to the executing program.
Their associated ARCV xegisters then control the type of access permitted into such data address spaces by the executing program. The executing progranl address space may also contain data available to the program. For example in a System/370 machine, AR0 may be assigned for PO9-79-o_, lh53~2~
containing the STD of the executing program of a pro-cessor, while part or all of ARl through ARl5 available for containing the STD's defining address spaces con-taining data available to the program.
Furthermore, the a~ove described arrangement is useable for executing current programs designed to run on IBM System/360 or System/370 architecture by disabling the use of the AR's and ARCV's by the processor, since the GPR's may be the same as found in IBM System/370 processors. For example, a field in a control register or the program status word (PSW) can be assigned to indi-cate whether the processor is running in conventional System/360 or System/370 mode or whether the processor is running in the cross-virtual-memory mode of this invention~
A new non-privileged instruction (load AR/ARCV) is also provided by this invention to load the AR's and ARCV registers from fields in main storage and set the special field to the enabled state. The area in main storage containing the AR and ARCV values to be loaded for a particular program may be defined in an address and extent in one of the processor control registers, which may be called a DSI CR. The AR and ARCV values to be loaded for a given program are contained in the data space inventory (DSI) defined by ~he DSI CR. The load AR/ARCV instruction uses the DSI CR address to load into a specified AR, a DSI entry for an address space to be accessed by the program. The loading of the DSI CR
is done by the privileged System/370 Load Control (LCTL) instruction. Therefore, a program using the load AR/ARCV
instruction may not have authority for issuing the privileged LCTL instruction. Thus, the only access to the DSI addressed by the DSI CR is by means of the load AR/ARCV instruction. A non-privileged store AR/~RCV
instruction can only store the current AR values into the DSI currently defined in the DSI CR. The extent field specified in the DSI CR prevents the accessing by either the load AR/ARCV or store AR/ARCV instructions outside the specified DSI. The origin and extent for the DSI may be in two separate CR's, called a DSI CR
pair. Another non-privileged instruction may be pro-vided to set the disabled state in the special field for any AR to disable the AR.
BRIEF DESCRIPTION OF THE DRAWINGS
FIGURE 1 shows the invention comprising an address space identifier selection mechanism in combination with an instruction decoder inserted into an otherwise conventional address translation environment in a data processing system.
FIGURE 2 is a preferred embodiment of the STD
selection mechanism.
FIGURE 3 is an embodiment of circuitry supporting the load AR/ARCV instruction.
FIGURE 4 is an addition to the embodiment in FIGURE 2.
FIGURE 5 is a modification to the embodiment in FIGURE 2 in which a single array contains bo~h the AR
and ARCV registers.
BRIEF DESCRIPTION OF THE PREFERRED EMBODIMENTS
FIGURE 1 shows symbolically a conventional instruc-tion decoder 9, such as may be found in any of a number of commercial processors such as an I~M 3033 processor.
PO~--79--012 ~S~
Decoder 9 can decode any of the IBM System/370 instruc-tions which have an operation code indicating the instruction length and type. The operation code defines which fields of the instruction are R, X or B fields that specify GPR's~ These GPR fields are selected by gates 11, 12, 13 and 14 which have outputs provided on a four bit address bus 9A to a general register array 17 which contains sixteen GPR's. The signal on bus 9A
selects a required GPR in array 17.
Gates 11, 12, 13 and 14 detect the base register addresses on busses llA, 12A, 13A and 14A to a STD selec-tion mechanism 20, which selects an AR/ARC~ pair associ-ated with the GPR simultaneously selected in array 17.
The X, B, D components of each logical address are sent by array 17 and instruction decoder 9 to a conven-tional address adder 18 to generate the logical address in a register 19, which provides inputs to a conven-tional address translation mechanism with a conventional translation lookaside buffer (TLB3 27 of the type imple-mented within the IBM System/370 M3033 processor.
Mechanism 20 outputs a STO from the selected AR and sends the STO to the conventional TLB and address trans-lation mechanism. The TLB uses the STO with the segment index and page index from logical address register 19 to determine if the translation was previously done and is still valid in TLB 27. If that is the case, TLB
addressing means 26 causes the translated real address to be immediately taken from TLB 27 and put in real address register 33, and the translation process is terminated. If the translation is not available in TLB 27, the full address translation process is executed accessing the segment table 22 addressed by the STO
provided by mechanism 20 and the page table 24 accessed PO~-79-012 1~3~9 through tha, segment table using the segment and page indexes from logical address reglster 19.
The basic STD selection mechanism 20 is shown in FIGURE 2. It has an OR circuit 40 which receives the storage operand base register address provided on one of the lines llA, 12A~ 13A o.r 14A from the instruction decoder 9. The output of circuit 40 sets an AR select register 41 which is decoded by a sixteen way decoder 42 to activate one of sixteen output lines, of which address 0 activates the select AR0 line to inverter 51, while each of the other fifteen output lines, select AR 1-15, simultaneously selects a corres-ponding access register in the AR array 46 and a corresponding ARCV register in ARCV array 48.
The content of the selected AR in array 46 is set into an AR out register 47, while the content of the corresponding ARCV register in array 48 is set into an ARCV out register 49.
The content of register 47 is the STD in the selected AR in array 46.
AR0 register 72 contains the STD which specifies the address space containing the program which can access the data space specified by the selected STD
in register 47. The program's STD may be directly loaded into register 72 from main storage, or the program STD may be transferred from a predetermined control resister 71 which may be CRl in the IBM
System/370 architecture environment which requires that the STO in CRl define the address space for the executing program and its accessible data.
1~5;3~4 The ARCV content in register 49 includes at least three bits: enable (E), read (R) and write (W), which determine the access rights of the e~ecuting program to the data space defined by the selected STD in register 47. The E bit must be set to 1 to enable the use of the corresponding AR register. Also a four bit storage key protect field (KEY) may be included in the ARCV to compare with the main storage protect key for an accessed 2K block wherein an equal comparison is required to allow the main storage access.
However, any address space may be accessed in main storage only when its STD is transferred to a main storage STO register 74. Therefore, the program space specified in AR0 register 72, or the data space specified in AR out register 47 may be accessed in main storage only when its STD is transferred into MS STO
register 74. This access-request STD transfer is controlled by gates 54 and 73, which are oppositely enabled by the output of an AND gate 53. Gate 73 is enabled when gate 53 is disabled due to inverter 56 connected to the output of gate 53.
An instruction fetch request requires an access to the program address space. The request is outputted from an OR circuit 43 which receives a branch (BR) instruction access request on line 16C, or a next sequential instruction request on the I fetch line 16D. The instruction re~uest signal from circuit 43 is inverted in lnverter 52 to disable gate 53. When gate 53 is disabled, the AR0 STD content is transferred by gate 73 into MS STO register 74 to control the instruction fetch request address.
However, during any operand access request, gate 53 will be enabled as long as the E bit is on in the corresponding ARCV. The STD in AR out register 47 is Po~ 7~-012 3~Z~
thereby transferred into MS request STO register 74, which contains the STO of the current request. When-ever the selected AR content is disabled, i.e. the E
bit is off in the selected ARCV in register 49, gate 73 will be enabled and the program STD in ARO will be transferred to request STO register 74.
The authority for an operand access request is regulated by the value of the R and W bits in the selected ARCV in register 49. If the R bit is 1, fetch accesses are permitted, but they are not permitted if R is zero. Likewise, if W is 1, store accesses are permitted, but stores are not permitted if W is 0. If both W and R are set to one, then both read and write accesses are permitted in the data space. The AND gates 63 and 64 enforce the use of the W and R bits by outputting a program interrupt signal on line 66 if a storage request violates the bit setting. Thus, gate 63 outputs an interrupt signal if the W bit is off during a write request on line 16B from the instruction unit, and gate 64 outputs an interrupt signal if the R bit is off during a read request on line 16A from the instruction unit.
When ARO is used as the STD for a storage access, ARCVO register 48A has R and W bits which are routed to inverter circuits 62A and 61A. The output of inverter 61A is routed to gate 63A which receives other inputs from line 16B and inverter 56. This checks authorization of the program to do a store in the program space and results in a program check if store accesses are not authorized. Similarly, gate 64A
with inputs from inverters 62A and 56 and from line 16A
checks authorization for an operand fetch access.
P09-79- ~12 3~t A storage protect key (KEY) field in each ARCV
provides additional authority control over accesses to the data in the related address space. An AND
gate 50 receives the KEY and the E bit from ARCV out register 49. If the E bit is enabled (i.e. set to 1), the key is transferred from the ARCV to the conven-tional key compare circuits in main storage, where the ARCV key must compare equal to the key for the 2KE storage block to permit the access, and the conventional PSW key is not used. If the E bit is disabled (i.e. set to 0), the ARCV key is not used, but the conventional PSW key is used instead.
FIGURE 3 represents the circuits needed to support the execution of a load AR/ARCV instruction which is shown with an operation code LE in instruction register 10. The Rl field in bits 24-27 of the LE instruction contains the address of the AR and ARCV register pair to be loaded. The R2 field in bits 28-31 of register 10 contains the address of a GPR which contains the logical address of a double word in a data space inventory (DS ) in main storage which is to be loaded into the specified AR/ARCV pair. This logical address is in the DSI whose real storage origin is specified by the content of a control register (CRX~ 80A. ~he R2 specified address in the LE instruction is a logical address transferred from the GPR array 17 into the GPR
out register 17A offset to the left by three bit positions and to the displacement address register 19, from which it is transferred to adder means 27 where it is combined with the contents of DSI CR 80A
into a real address put into register 33 which is used to access the DSI double word in main storage 82, which is outputted to storage operand register 83.
DSI CR 8OB contains the maximum DSI number that can be used by the executing program space. It is compared PO9-79~012 by comparator 80C with the content of register 17A, and if the register 17A content exceeds the con~ent of DSI CR 80B, a program check is generated. The doubleword in register 83 contains field 0-31 which is loaded into the selected ARCV register in array 48 and field 32-63 which is loaded into the selected AR
in arra~ 46. Bit 0 in the ARCV fleld contains the E
bit, which must be set to 1 to enable the selected AR/ARCV pair. If off, the E bit sets a trigger 84 which outpu's a program interrupt to tell the operating system that the AR being loaded is not enabled.
A store AR/ARCV instruction is also provided. It is also represented by FIGURE 3 except that the operation code is SE (instead of LE) in instruction register 10. Then the content of the AR/ARCV pair specified by the Rl field of the SE instruction is stored into the main storage doubleword specified by the R2 field of the SE instruction relative to a real main storage address in control register 80.
A disable AR instruction is also provided having a DAR operation code with a single Rl operand which specifies the ARCV in which the E bit is set to zero (i.e. to off state) to disable the associated P.R .
FIGURE 4 illustrates an addition to the circuits in FIGURE 2 to obtain additional functions for the STD
selection mechanism by providing an ARCV-O and further controls with ARCV 1-15. The ARCV-O register settings control the execution of selected subclasses of privileged instructions in the program address space defined in AR-O .
The further controls with ARCV 1-15 provide an additional subset of operation decoding within a seneric operation code of the current instruction being decode~
by the instruction decoder 9 in FIGURE 1.
The ARCV-0 privileged instruction subclass controls are independent of the additional ARCV 1-15 controls, and either may be used without the other, or neither may be used as is the case in FIGURE 2.
Operation decoder 16A includes additional out-puts Pl, P2...Pn which respectively signal different types of privileged instruction subclasses, as follows:
1. I/O instructions, e.g. start I/O, etc.
2. Multiprocessor instructions, e.g. SIGP, etc.
3. Special instructions, e.g. RPQ instructions, etc.
4. Maintenance instructions, e.g. diagnose, etc.
5. Load Control instruction.
6. Resource Management instructions, e.g. IPTE, set key, purge TLB, set clock, set interrupt masks, set CPU timer, etc.
An ARCV-0 register 48A contains the previously described authority control bits R and W and also contains a privileged instruction subclass mask bit positions Sl, S2...Sn which correspond to the decoder 25 16A output signal lines P1, P2.. Pn. The 0 or 1 setting of each subclass mask bit respectively enables or disables the execution of its corresponding subclass of privileged instructions accessed in the proper address space defined by the STD in AR0. The subclass enablement/disablement is done by AND gates 91, 92 through 95, each receiving a corresponding mask bit from the ARCV-0 register 48A and a corresponding POg-79-012 ~;3~2~
signal line from operation decoder 16A. The outputs Cl, C2...Cn o~ gates 91 through 95 provide program check signals to the program interrupt circuits in the processor. Thus, if the S bit is set to 1 to disable a subclass, and operation decoder 16A issues a privileged instruction in the disabled subclass, the corresponding AND gate C will output a signal which causes a program interrupt, from which the system control program interrupt handler routine will stop further execution of the program in the current AR0 address space and indicate the reason for the stoppage.
Unprivileged instructions are always permitted to execute. Therefore, if no privileged instructions are permitted to execute in an application program address space in AR0, all of the privileged instruction subclass bits Sl...Sn are set to 1 to disable all privileged instruction execution.
The bits Tl...Tn in ARCV 1-15 in FIGURE 4 are used as an extension of the operation code of the current instruction being interpreted in instruction decoder 9. For example, the operation code for an instruction can generically signal the multiplication of its operands. Then the bits Tl...Tn will indicate the data type in the address space in the corresponding ARl-15 and signal that data type on bus 49B to operation decoder 16A so that the instruction unit can perform the multiplication on the data type indicated as being accessed in the corresponding address space. E~amples of the meanings of bits Tl...Tn areo 1. Fixed point 2. Single Precision floating point 3. Double precision floating point 4. Extended precision floating point Additional controls may be implemented in the ARCV's.
Indicators may be included in the ARCV's to specify 24 or 31 bit addressing mode, whether or not an address space contains segments common to other address spaces, whether the origin of the space is zero or some other value, what ranges of logical addresses are to be allowed access and even whether the logical address is to be translated to real (DAT on) or treated as a real address (DAT off). The circuits and implementation for such additional controls are straight forward and are not shown in the drawings.
It is understood that the main storage real address resulting from a translation, including from a TLB look-up, may access the main storage data in a cache or high speed buffer (instead of in main storage) in any system having the da~a previously transferred from main storage to a cache or high speed buffer.
The address translation means shown in FIGURE 1 using segment and page tables is only one e~ample of an address translation means, and the invention may be used with any other address translation means, such as for example, hashing address translation means as is described in the "IBM System/38 Technical Developments"
manual (Form No. G580-0237) on page 23.
FIGURE 5 shows a modification to the embodiment in FIGURE 2 in which both the AR array 4 6 and the ARCV
array 48 are combined into a single array 46A in which each entry is associated with a respective GPR and contains bcth the associated AR register and ARCV regis-ters; and logic may be used to inhibit the ARJARcv output when a non-address containing GPR is selected.
3~ ~2 9!
Although this invention has been shown and described with respect to plural embodiments thereof, it should be understood by those skilled in the art that various changes and omissions in the form and details of the S invention ma~ be made therein without departing from the spirit and the scope oE the invention.
ll. In which program address spaces may share specified data address spaces but may not share other data address spaces, and may be allowed or prohibited from sharing access to each otherls program address space~
~0~--79-012 3~
--7~
12. Having authority controls on each access register to permit different progra~s to have different authorities to access the same data address space.
13. Having descriptor fields with each access regis~er defining attributes of the associated address space, or the data elements within the space, which determine the operations of instruc-tions relative to that space.
14. With an addressing mechanism which need not connect a data address space to a program until access is required by the program.
15. ~ith an addressing mechanism which can support data address spaces independent of any 1~ program when the program is made not executable in the processor, e.g. when the program is not dispatched.
16. Which allows an access register to be disabled to permit its general purpose register to provide relative addressing of data in the program address space, but also allows the access register to be enabled so that its general purpose register controls relative addressing in a data address space specified by its access register separately from the program address space.
17. Which allows access registers to be disabled while the processor is executing programs under current architecture to eliminate the need for loading the access registers prior to execution of such programs.
1~5~82~
18. Having authority controls with each program space separately authorizing programs in that program space to use specified privileged instruc-tions.
19. Which can have sequential operand requests flexibly access different address spaces identi-fied in a plurality of access registers. No intervention is needed by any address space switching instruction, and without any required relationship between any address space and the source or sink state of the requesting operand in an instruction.
It is therefore a feature of this invention to control access to data in a subset of address spaces by an executing program without requiring any inter-ruption in the execution of the program when it changes its accessing of data to a different address space in the subset. The manner of access can be varied among the different address spaces; for example, the program may be permitted to read (but not write) data from one address space, write (but not read) data in a second address space, and both read and write data in a third address space.
The invention associates access registers (AR's) with the general purpose registers (GPR's) in a data processor. The AR's may each be loaded with an address space identifier, for example, a unique STD (segment table descriptor). The STD comprises a STO field and a segment table length field. Thus, there may be 16 AR's 3~ associated respectively with 16 GPR's in a processor.
The address s~ace identifier in an AR is selected for address translation when the associated GPR is selected as a storage operand base register, such as being the GPR selected by the B-field in an IBM System/370 instruc-_g_ tion. However, the address space identifier content of an AR is not selected for an address translation if the associated GPR is selected for a purpose other than as a storage operand base register, such as if a GPR is selected as an index (X) register or as a data source or sink register (R) for an instruction.
The invention also obtains authority control for each program's access to each address space by also associating an AR Control Vector (ARCV) register wi~h each AR to control the type of access permitted to the associated address space by an executing program.
A special field in either an AR or its associated ARCV may indicate whether the AR content is enabled and thereby contains a useable address space identifier; or is disabled and the address space identifier must be obtained elsewhere. This special field with each AR
permits the GPR of any disabled AR to specify a base value associated with the address space defined by another AR; for example, a disabled GPR may specify a base address for data in the program address space defined by ARO.
Thus, the invention controls which address space may contain an executing program and which other address spaces may be accessed for data by the executing program.
This may be done by permanently assigning one of the A~'s to contain the STD of ~he executing program (e.g. AR0).
Then the other AR's may be permanently available to con~ain the address space identifiers of other address spaces which have data accessible to the executing program.
Their associated ARCV xegisters then control the type of access permitted into such data address spaces by the executing program. The executing progranl address space may also contain data available to the program. For example in a System/370 machine, AR0 may be assigned for PO9-79-o_, lh53~2~
containing the STD of the executing program of a pro-cessor, while part or all of ARl through ARl5 available for containing the STD's defining address spaces con-taining data available to the program.
Furthermore, the a~ove described arrangement is useable for executing current programs designed to run on IBM System/360 or System/370 architecture by disabling the use of the AR's and ARCV's by the processor, since the GPR's may be the same as found in IBM System/370 processors. For example, a field in a control register or the program status word (PSW) can be assigned to indi-cate whether the processor is running in conventional System/360 or System/370 mode or whether the processor is running in the cross-virtual-memory mode of this invention~
A new non-privileged instruction (load AR/ARCV) is also provided by this invention to load the AR's and ARCV registers from fields in main storage and set the special field to the enabled state. The area in main storage containing the AR and ARCV values to be loaded for a particular program may be defined in an address and extent in one of the processor control registers, which may be called a DSI CR. The AR and ARCV values to be loaded for a given program are contained in the data space inventory (DSI) defined by ~he DSI CR. The load AR/ARCV instruction uses the DSI CR address to load into a specified AR, a DSI entry for an address space to be accessed by the program. The loading of the DSI CR
is done by the privileged System/370 Load Control (LCTL) instruction. Therefore, a program using the load AR/ARCV
instruction may not have authority for issuing the privileged LCTL instruction. Thus, the only access to the DSI addressed by the DSI CR is by means of the load AR/ARCV instruction. A non-privileged store AR/~RCV
instruction can only store the current AR values into the DSI currently defined in the DSI CR. The extent field specified in the DSI CR prevents the accessing by either the load AR/ARCV or store AR/ARCV instructions outside the specified DSI. The origin and extent for the DSI may be in two separate CR's, called a DSI CR
pair. Another non-privileged instruction may be pro-vided to set the disabled state in the special field for any AR to disable the AR.
BRIEF DESCRIPTION OF THE DRAWINGS
FIGURE 1 shows the invention comprising an address space identifier selection mechanism in combination with an instruction decoder inserted into an otherwise conventional address translation environment in a data processing system.
FIGURE 2 is a preferred embodiment of the STD
selection mechanism.
FIGURE 3 is an embodiment of circuitry supporting the load AR/ARCV instruction.
FIGURE 4 is an addition to the embodiment in FIGURE 2.
FIGURE 5 is a modification to the embodiment in FIGURE 2 in which a single array contains bo~h the AR
and ARCV registers.
BRIEF DESCRIPTION OF THE PREFERRED EMBODIMENTS
FIGURE 1 shows symbolically a conventional instruc-tion decoder 9, such as may be found in any of a number of commercial processors such as an I~M 3033 processor.
PO~--79--012 ~S~
Decoder 9 can decode any of the IBM System/370 instruc-tions which have an operation code indicating the instruction length and type. The operation code defines which fields of the instruction are R, X or B fields that specify GPR's~ These GPR fields are selected by gates 11, 12, 13 and 14 which have outputs provided on a four bit address bus 9A to a general register array 17 which contains sixteen GPR's. The signal on bus 9A
selects a required GPR in array 17.
Gates 11, 12, 13 and 14 detect the base register addresses on busses llA, 12A, 13A and 14A to a STD selec-tion mechanism 20, which selects an AR/ARC~ pair associ-ated with the GPR simultaneously selected in array 17.
The X, B, D components of each logical address are sent by array 17 and instruction decoder 9 to a conven-tional address adder 18 to generate the logical address in a register 19, which provides inputs to a conven-tional address translation mechanism with a conventional translation lookaside buffer (TLB3 27 of the type imple-mented within the IBM System/370 M3033 processor.
Mechanism 20 outputs a STO from the selected AR and sends the STO to the conventional TLB and address trans-lation mechanism. The TLB uses the STO with the segment index and page index from logical address register 19 to determine if the translation was previously done and is still valid in TLB 27. If that is the case, TLB
addressing means 26 causes the translated real address to be immediately taken from TLB 27 and put in real address register 33, and the translation process is terminated. If the translation is not available in TLB 27, the full address translation process is executed accessing the segment table 22 addressed by the STO
provided by mechanism 20 and the page table 24 accessed PO~-79-012 1~3~9 through tha, segment table using the segment and page indexes from logical address reglster 19.
The basic STD selection mechanism 20 is shown in FIGURE 2. It has an OR circuit 40 which receives the storage operand base register address provided on one of the lines llA, 12A~ 13A o.r 14A from the instruction decoder 9. The output of circuit 40 sets an AR select register 41 which is decoded by a sixteen way decoder 42 to activate one of sixteen output lines, of which address 0 activates the select AR0 line to inverter 51, while each of the other fifteen output lines, select AR 1-15, simultaneously selects a corres-ponding access register in the AR array 46 and a corresponding ARCV register in ARCV array 48.
The content of the selected AR in array 46 is set into an AR out register 47, while the content of the corresponding ARCV register in array 48 is set into an ARCV out register 49.
The content of register 47 is the STD in the selected AR in array 46.
AR0 register 72 contains the STD which specifies the address space containing the program which can access the data space specified by the selected STD
in register 47. The program's STD may be directly loaded into register 72 from main storage, or the program STD may be transferred from a predetermined control resister 71 which may be CRl in the IBM
System/370 architecture environment which requires that the STO in CRl define the address space for the executing program and its accessible data.
1~5;3~4 The ARCV content in register 49 includes at least three bits: enable (E), read (R) and write (W), which determine the access rights of the e~ecuting program to the data space defined by the selected STD in register 47. The E bit must be set to 1 to enable the use of the corresponding AR register. Also a four bit storage key protect field (KEY) may be included in the ARCV to compare with the main storage protect key for an accessed 2K block wherein an equal comparison is required to allow the main storage access.
However, any address space may be accessed in main storage only when its STD is transferred to a main storage STO register 74. Therefore, the program space specified in AR0 register 72, or the data space specified in AR out register 47 may be accessed in main storage only when its STD is transferred into MS STO
register 74. This access-request STD transfer is controlled by gates 54 and 73, which are oppositely enabled by the output of an AND gate 53. Gate 73 is enabled when gate 53 is disabled due to inverter 56 connected to the output of gate 53.
An instruction fetch request requires an access to the program address space. The request is outputted from an OR circuit 43 which receives a branch (BR) instruction access request on line 16C, or a next sequential instruction request on the I fetch line 16D. The instruction re~uest signal from circuit 43 is inverted in lnverter 52 to disable gate 53. When gate 53 is disabled, the AR0 STD content is transferred by gate 73 into MS STO register 74 to control the instruction fetch request address.
However, during any operand access request, gate 53 will be enabled as long as the E bit is on in the corresponding ARCV. The STD in AR out register 47 is Po~ 7~-012 3~Z~
thereby transferred into MS request STO register 74, which contains the STO of the current request. When-ever the selected AR content is disabled, i.e. the E
bit is off in the selected ARCV in register 49, gate 73 will be enabled and the program STD in ARO will be transferred to request STO register 74.
The authority for an operand access request is regulated by the value of the R and W bits in the selected ARCV in register 49. If the R bit is 1, fetch accesses are permitted, but they are not permitted if R is zero. Likewise, if W is 1, store accesses are permitted, but stores are not permitted if W is 0. If both W and R are set to one, then both read and write accesses are permitted in the data space. The AND gates 63 and 64 enforce the use of the W and R bits by outputting a program interrupt signal on line 66 if a storage request violates the bit setting. Thus, gate 63 outputs an interrupt signal if the W bit is off during a write request on line 16B from the instruction unit, and gate 64 outputs an interrupt signal if the R bit is off during a read request on line 16A from the instruction unit.
When ARO is used as the STD for a storage access, ARCVO register 48A has R and W bits which are routed to inverter circuits 62A and 61A. The output of inverter 61A is routed to gate 63A which receives other inputs from line 16B and inverter 56. This checks authorization of the program to do a store in the program space and results in a program check if store accesses are not authorized. Similarly, gate 64A
with inputs from inverters 62A and 56 and from line 16A
checks authorization for an operand fetch access.
P09-79- ~12 3~t A storage protect key (KEY) field in each ARCV
provides additional authority control over accesses to the data in the related address space. An AND
gate 50 receives the KEY and the E bit from ARCV out register 49. If the E bit is enabled (i.e. set to 1), the key is transferred from the ARCV to the conven-tional key compare circuits in main storage, where the ARCV key must compare equal to the key for the 2KE storage block to permit the access, and the conventional PSW key is not used. If the E bit is disabled (i.e. set to 0), the ARCV key is not used, but the conventional PSW key is used instead.
FIGURE 3 represents the circuits needed to support the execution of a load AR/ARCV instruction which is shown with an operation code LE in instruction register 10. The Rl field in bits 24-27 of the LE instruction contains the address of the AR and ARCV register pair to be loaded. The R2 field in bits 28-31 of register 10 contains the address of a GPR which contains the logical address of a double word in a data space inventory (DS ) in main storage which is to be loaded into the specified AR/ARCV pair. This logical address is in the DSI whose real storage origin is specified by the content of a control register (CRX~ 80A. ~he R2 specified address in the LE instruction is a logical address transferred from the GPR array 17 into the GPR
out register 17A offset to the left by three bit positions and to the displacement address register 19, from which it is transferred to adder means 27 where it is combined with the contents of DSI CR 80A
into a real address put into register 33 which is used to access the DSI double word in main storage 82, which is outputted to storage operand register 83.
DSI CR 8OB contains the maximum DSI number that can be used by the executing program space. It is compared PO9-79~012 by comparator 80C with the content of register 17A, and if the register 17A content exceeds the con~ent of DSI CR 80B, a program check is generated. The doubleword in register 83 contains field 0-31 which is loaded into the selected ARCV register in array 48 and field 32-63 which is loaded into the selected AR
in arra~ 46. Bit 0 in the ARCV fleld contains the E
bit, which must be set to 1 to enable the selected AR/ARCV pair. If off, the E bit sets a trigger 84 which outpu's a program interrupt to tell the operating system that the AR being loaded is not enabled.
A store AR/ARCV instruction is also provided. It is also represented by FIGURE 3 except that the operation code is SE (instead of LE) in instruction register 10. Then the content of the AR/ARCV pair specified by the Rl field of the SE instruction is stored into the main storage doubleword specified by the R2 field of the SE instruction relative to a real main storage address in control register 80.
A disable AR instruction is also provided having a DAR operation code with a single Rl operand which specifies the ARCV in which the E bit is set to zero (i.e. to off state) to disable the associated P.R .
FIGURE 4 illustrates an addition to the circuits in FIGURE 2 to obtain additional functions for the STD
selection mechanism by providing an ARCV-O and further controls with ARCV 1-15. The ARCV-O register settings control the execution of selected subclasses of privileged instructions in the program address space defined in AR-O .
The further controls with ARCV 1-15 provide an additional subset of operation decoding within a seneric operation code of the current instruction being decode~
by the instruction decoder 9 in FIGURE 1.
The ARCV-0 privileged instruction subclass controls are independent of the additional ARCV 1-15 controls, and either may be used without the other, or neither may be used as is the case in FIGURE 2.
Operation decoder 16A includes additional out-puts Pl, P2...Pn which respectively signal different types of privileged instruction subclasses, as follows:
1. I/O instructions, e.g. start I/O, etc.
2. Multiprocessor instructions, e.g. SIGP, etc.
3. Special instructions, e.g. RPQ instructions, etc.
4. Maintenance instructions, e.g. diagnose, etc.
5. Load Control instruction.
6. Resource Management instructions, e.g. IPTE, set key, purge TLB, set clock, set interrupt masks, set CPU timer, etc.
An ARCV-0 register 48A contains the previously described authority control bits R and W and also contains a privileged instruction subclass mask bit positions Sl, S2...Sn which correspond to the decoder 25 16A output signal lines P1, P2.. Pn. The 0 or 1 setting of each subclass mask bit respectively enables or disables the execution of its corresponding subclass of privileged instructions accessed in the proper address space defined by the STD in AR0. The subclass enablement/disablement is done by AND gates 91, 92 through 95, each receiving a corresponding mask bit from the ARCV-0 register 48A and a corresponding POg-79-012 ~;3~2~
signal line from operation decoder 16A. The outputs Cl, C2...Cn o~ gates 91 through 95 provide program check signals to the program interrupt circuits in the processor. Thus, if the S bit is set to 1 to disable a subclass, and operation decoder 16A issues a privileged instruction in the disabled subclass, the corresponding AND gate C will output a signal which causes a program interrupt, from which the system control program interrupt handler routine will stop further execution of the program in the current AR0 address space and indicate the reason for the stoppage.
Unprivileged instructions are always permitted to execute. Therefore, if no privileged instructions are permitted to execute in an application program address space in AR0, all of the privileged instruction subclass bits Sl...Sn are set to 1 to disable all privileged instruction execution.
The bits Tl...Tn in ARCV 1-15 in FIGURE 4 are used as an extension of the operation code of the current instruction being interpreted in instruction decoder 9. For example, the operation code for an instruction can generically signal the multiplication of its operands. Then the bits Tl...Tn will indicate the data type in the address space in the corresponding ARl-15 and signal that data type on bus 49B to operation decoder 16A so that the instruction unit can perform the multiplication on the data type indicated as being accessed in the corresponding address space. E~amples of the meanings of bits Tl...Tn areo 1. Fixed point 2. Single Precision floating point 3. Double precision floating point 4. Extended precision floating point Additional controls may be implemented in the ARCV's.
Indicators may be included in the ARCV's to specify 24 or 31 bit addressing mode, whether or not an address space contains segments common to other address spaces, whether the origin of the space is zero or some other value, what ranges of logical addresses are to be allowed access and even whether the logical address is to be translated to real (DAT on) or treated as a real address (DAT off). The circuits and implementation for such additional controls are straight forward and are not shown in the drawings.
It is understood that the main storage real address resulting from a translation, including from a TLB look-up, may access the main storage data in a cache or high speed buffer (instead of in main storage) in any system having the da~a previously transferred from main storage to a cache or high speed buffer.
The address translation means shown in FIGURE 1 using segment and page tables is only one e~ample of an address translation means, and the invention may be used with any other address translation means, such as for example, hashing address translation means as is described in the "IBM System/38 Technical Developments"
manual (Form No. G580-0237) on page 23.
FIGURE 5 shows a modification to the embodiment in FIGURE 2 in which both the AR array 4 6 and the ARCV
array 48 are combined into a single array 46A in which each entry is associated with a respective GPR and contains bcth the associated AR register and ARCV regis-ters; and logic may be used to inhibit the ARJARcv output when a non-address containing GPR is selected.
3~ ~2 9!
Although this invention has been shown and described with respect to plural embodiments thereof, it should be understood by those skilled in the art that various changes and omissions in the form and details of the S invention ma~ be made therein without departing from the spirit and the scope oE the invention.
Claims (22)
1. Address control means in a data processing system having a predetermined number of general purpose registers (GPR's) in which some instructions of the system specify a GPR as a storage operand base register which content takes part in a calculation of the instructions logical storage operand address, comprising:
a plurality of access registers (AR's) respectively associated with the GPR's, means for accessing the content of the AR
associated with a GPR being accessed, the AR
content having an address space identifier (ASI), address translation means receiving a requested storage operand address and the ASI
from the accessing means for translating the storage operand address to a real address in a main storage of the data processing system.
a plurality of access registers (AR's) respectively associated with the GPR's, means for accessing the content of the AR
associated with a GPR being accessed, the AR
content having an address space identifier (ASI), address translation means receiving a requested storage operand address and the ASI
from the accessing means for translating the storage operand address to a real address in a main storage of the data processing system.
2. Address control means as defined in Claim 1, further comprising:
instruction unit means for detecting the GPR
field(s) in each instruction being decoded, including AR detecting means for detecting each GPR field addressing a GPR containing an operand address in an address space, and an AR address signal being outputted by the AR detecting means to the accessing means for selecting an AR.
instruction unit means for detecting the GPR
field(s) in each instruction being decoded, including AR detecting means for detecting each GPR field addressing a GPR containing an operand address in an address space, and an AR address signal being outputted by the AR detecting means to the accessing means for selecting an AR.
3. Address control means as defined in Claim 2, in which the data processing system has a control register for containing a program ASI
for defining the address space of a program being executed, further comprising:
instruction fetch controls providing a logical address for each instruction fetch request, the address translation means receiving the program ASI content of the control register for translating the instruction fetch address to a real address in main storage for the program being executed from which instruction fetches occur, whereby the program is in a program address space and can access data in any address space specified by the content of an AR.
for defining the address space of a program being executed, further comprising:
instruction fetch controls providing a logical address for each instruction fetch request, the address translation means receiving the program ASI content of the control register for translating the instruction fetch address to a real address in main storage for the program being executed from which instruction fetches occur, whereby the program is in a program address space and can access data in any address space specified by the content of an AR.
4. Address space control means as defined in Claim 1, further comprising:
instruction fetch control means providing a logical address for each instruction fetch request, operand access control means providing a logical address for each operand access request, program space access control means for accessing the ASI content of an AR associated with an executing program when an instruction fetch request is provided, data space access control means for accessing the ASI content of an AR associated with a GPR
selected by an operand access request, the translation controls receiving the ASI
provided by either the program space access control means or the data space access control means, whereby the AR accessed by the data space access control means may specify a data address space, while the AR accessed by the program space access control means specifies the program address space.
instruction fetch control means providing a logical address for each instruction fetch request, operand access control means providing a logical address for each operand access request, program space access control means for accessing the ASI content of an AR associated with an executing program when an instruction fetch request is provided, data space access control means for accessing the ASI content of an AR associated with a GPR
selected by an operand access request, the translation controls receiving the ASI
provided by either the program space access control means or the data space access control means, whereby the AR accessed by the data space access control means may specify a data address space, while the AR accessed by the program space access control means specifies the program address space.
5. Address control means as defined in Claim 1, further comprising:
access register control vector (ARCV) registers being respectively associated with the plurality of AR's, each ARCV register containing control bits for regulating the operation of the data processing system when accessing a data address space specified in the associated AR.
access register control vector (ARCV) registers being respectively associated with the plurality of AR's, each ARCV register containing control bits for regulating the operation of the data processing system when accessing a data address space specified in the associated AR.
6. Address control means as defined in Claim 1, further comprising:
access register control vector (ARCV) registers being respectively associated with the plurality of AR's, the content of each ARCV
register controlling the mode of access to an address space identified in the associated AR.
access register control vector (ARCV) registers being respectively associated with the plurality of AR's, the content of each ARCV
register controlling the mode of access to an address space identified in the associated AR.
7. Address control means as defined in Claim 6, further comprising:
each ARCV register containing an enable/
disable bit position set to on/off state, the state of the enable/disable bit position being sensed by the accessing means, transfer means in the accessing means for transferring to the address translation means the content of the AR selected by a GPR designated as a storage operand register field of an instruction if the enable/disable bit position is set to an enable state, the transferring means not transferring said AR contents if the enable/
disable bit position is set to a disabled state, whereby when the enable/disable bit is set to the disabled state, the address translation means interprets the operand address in the same way as in a system in which the AR's and ARCV's do not exist.
each ARCV register containing an enable/
disable bit position set to on/off state, the state of the enable/disable bit position being sensed by the accessing means, transfer means in the accessing means for transferring to the address translation means the content of the AR selected by a GPR designated as a storage operand register field of an instruction if the enable/disable bit position is set to an enable state, the transferring means not transferring said AR contents if the enable/
disable bit position is set to a disabled state, whereby when the enable/disable bit is set to the disabled state, the address translation means interprets the operand address in the same way as in a system in which the AR's and ARCV's do not exist.
8. Address control means as defined in Claim 6, further comprising:
the ARCV register associated with each operand AR containing at least a write authority protection field, the protection field being accessed by the accessing means, means for signalling storage write request signals from an instruction execution unit, program interrupt means being actuated by the write request signal when the write authority protection field is off, whereby the operand address translation is interrupted if the request does not have the required access authority.
the ARCV register associated with each operand AR containing at least a write authority protection field, the protection field being accessed by the accessing means, means for signalling storage write request signals from an instruction execution unit, program interrupt means being actuated by the write request signal when the write authority protection field is off, whereby the operand address translation is interrupted if the request does not have the required access authority.
9. Address control means as defined in Claim 6, further comprising:
the ARCV register associated with each operand AR containing at least a read authority protection field, the protection field being accessed by the accessing means, means for signalling storage read request signals from an instruction execution unit, program interrupt means being actuated by the read request signal when the read authority protection field is off, whereby the operand address translation is interrupted if the request does not have the required access authority.
the ARCV register associated with each operand AR containing at least a read authority protection field, the protection field being accessed by the accessing means, means for signalling storage read request signals from an instruction execution unit, program interrupt means being actuated by the read request signal when the read authority protection field is off, whereby the operand address translation is interrupted if the request does not have the required access authority.
10. Address control means as defined in Claims 8 or 9, further comprising:
a program ARCV register associated with a program AR, the program ARCV register containing an access authority protection field, the program interrupt means being actuated by an operand access request signal to the program address space if the request does not have the authority indicated by the access authority protection field in the program ARCV register.
a program ARCV register associated with a program AR, the program ARCV register containing an access authority protection field, the program interrupt means being actuated by an operand access request signal to the program address space if the request does not have the authority indicated by the access authority protection field in the program ARCV register.
11. Address control means as defined in Claim 1, further comprising:
a control register for containing address information for locating ASI data in main storage which are to be loaded into the AR's, instruction unit means for interpreting a load AR instruction specifying an AR to be loaded and a main storage address of the ASI data to be loaded into the AR, main storage means for accessing the ASI data at the storage address into a storage out register, and means to load the ASI data into the designated AR to identify an address space to the data processing system.
a control register for containing address information for locating ASI data in main storage which are to be loaded into the AR's, instruction unit means for interpreting a load AR instruction specifying an AR to be loaded and a main storage address of the ASI data to be loaded into the AR, main storage means for accessing the ASI data at the storage address into a storage out register, and means to load the ASI data into the designated AR to identify an address space to the data processing system.
12. Address control means as defined in Claim 11, further comprising:
access register control vector (ARCV) registers being respectively associated with the plurality of AR's, each ARCV register containing control bits for regulating the operation of the data pro-cessing system when accessing a data address space specified in the associated AR, the loading means also loading a subset of the ASI data into an ARCV register associated with the RR being loaded.
access register control vector (ARCV) registers being respectively associated with the plurality of AR's, each ARCV register containing control bits for regulating the operation of the data pro-cessing system when accessing a data address space specified in the associated AR, the loading means also loading a subset of the ASI data into an ARCV register associated with the RR being loaded.
13. Address control means as defined in Claim 11, further comprising:
means to test an enable field in the storage out register to determine if the ASI data is valid, and means to inhibit loading the ASI data into the specified AR if the ASI data is not valid.
means to test an enable field in the storage out register to determine if the ASI data is valid, and means to inhibit loading the ASI data into the specified AR if the ASI data is not valid.
14. Address control means as defined in Claim 1, further comprising:
a control register for containing address information for locating an area in main storage into which the content of an AR can be stored, instruction unit means for interpreting a store AR instruction specifying an AR having contents to be stored at an address in main storage determined from the address information in the control register, and main storage means for storing the AR
contents into the main storage address.
a control register for containing address information for locating an area in main storage into which the content of an AR can be stored, instruction unit means for interpreting a store AR instruction specifying an AR having contents to be stored at an address in main storage determined from the address information in the control register, and main storage means for storing the AR
contents into the main storage address.
15. Address control means as defined in Claim 14, further comprising:
an ARCV register associated with the AR
specified in the store AR instruction, the content of the ARCV register being stored in main storage along with the AR content.
an ARCV register associated with the AR
specified in the store AR instruction, the content of the ARCV register being stored in main storage along with the AR content.
16. Address control means as defined in Claim 5, further comprising:
instruction unit means for interpreting a disable/enable AR instruction, an enable/disable bit associated with a specified AR, means responsive to the instruction unit means for setting the enable/disable bit for the particular AR to an on or off state as indicated in the instruction.
instruction unit means for interpreting a disable/enable AR instruction, an enable/disable bit associated with a specified AR, means responsive to the instruction unit means for setting the enable/disable bit for the particular AR to an on or off state as indicated in the instruction.
17. Address control means as defined in Claim 5, further comprising:
one of the AR and its corresponding ARCV
register (called an AR/ARCV pair) being pre-designated to control a program address space for an executable program, and the remaining AR/ARCV pairs being available to control data address spaces.
one of the AR and its corresponding ARCV
register (called an AR/ARCV pair) being pre-designated to control a program address space for an executable program, and the remaining AR/ARCV pairs being available to control data address spaces.
18. Address control means as defined in Claim 17, further comprising:
an enable/disable field in each ARCV
register, any of the remaining AR/ARCV pairs having its enabled/disabled field set to a disabled state having its corresponding GPR associated with the program address space defined in the one AR/ARCV pair.
an enable/disable field in each ARCV
register, any of the remaining AR/ARCV pairs having its enabled/disabled field set to a disabled state having its corresponding GPR associated with the program address space defined in the one AR/ARCV pair.
19. Address control means as defined in Claim 18, further comprising:
the ARCV register in the one AR/ARCV pair having a plurality of privileged instruction subclass mask bits respectively associated with different subclasses of privileged instructions, program interrupt means actuated by operational decoding of an instruction in a subclass having its mask bit disabled in the one AR/ARCV pair.
the ARCV register in the one AR/ARCV pair having a plurality of privileged instruction subclass mask bits respectively associated with different subclasses of privileged instructions, program interrupt means actuated by operational decoding of an instruction in a subclass having its mask bit disabled in the one AR/ARCV pair.
20. Address control means as defined in Claim 17, further comprising:
each ARCV register having authority control bits to control read or write accesses in the corresponding address space, program interrupt means actuated by any disabled authority bit when an access request attempts to violate the disabled authority.
each ARCV register having authority control bits to control read or write accesses in the corresponding address space, program interrupt means actuated by any disabled authority bit when an access request attempts to violate the disabled authority.
21. Address control means as defined in Claim 1, further comprising:
the address translation means using each ASI as a segment table descriptor (STD) to locate a particular segment table, whereby each requested logical address is accessed into the address space defined by a STD.
the address translation means using each ASI as a segment table descriptor (STD) to locate a particular segment table, whereby each requested logical address is accessed into the address space defined by a STD.
22. Address control means as defined in Claim 17, further comprising:
an array in which each addressed location contains the registers for an AR/ARCV pair.
an array in which each addressed location contains the registers for an AR/ARCV pair.
Applications Claiming Priority (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| US131,570 | 1980-03-19 | ||
| US06/131,570 US4355355A (en) | 1980-03-19 | 1980-03-19 | Address generating mechanism for multiple virtual spaces |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CA1153824A true CA1153824A (en) | 1983-09-13 |
Family
ID=22450031
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CA000368423A Expired CA1153824A (en) | 1980-03-19 | 1981-01-13 | Address generating mechanism for multiple virtual spaces |
Country Status (6)
| Country | Link |
|---|---|
| US (1) | US4355355A (en) |
| EP (1) | EP0036085B1 (en) |
| JP (1) | JPS6022377B2 (en) |
| CA (1) | CA1153824A (en) |
| DE (1) | DE3176834D1 (en) |
| ES (1) | ES500467A0 (en) |
Families Citing this family (63)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JPS5687282A (en) * | 1979-12-14 | 1981-07-15 | Nec Corp | Data processor |
| US4500952A (en) * | 1980-05-23 | 1985-02-19 | International Business Machines Corporation | Mechanism for control of address translation by a program using a plurality of translation tables |
| US4521846A (en) * | 1981-02-20 | 1985-06-04 | International Business Machines Corporation | Mechanism for accessing multiple virtual address spaces |
| JPS57143782A (en) * | 1981-03-03 | 1982-09-06 | Toshiba Corp | Information processor |
| US4513368A (en) * | 1981-05-22 | 1985-04-23 | Data General Corporation | Digital data processing system having object-based logical memory addressing and self-structuring modular memory |
| US4473878A (en) * | 1981-11-23 | 1984-09-25 | Motorola, Inc. | Memory management unit |
| US4472790A (en) * | 1982-02-05 | 1984-09-18 | International Business Machines Corporation | Storage fetch protect override controls |
| US4581702A (en) * | 1983-01-10 | 1986-04-08 | International Business Machines Corporation | Critical system protection |
| EP0170525B1 (en) * | 1984-07-31 | 1997-10-01 | Texas Instruments Incorporated | Cache hierarchy design for use in a memory management unit |
| US4985829A (en) * | 1984-07-31 | 1991-01-15 | Texas Instruments Incorporated | Cache hierarchy design for use in a memory management unit |
| US4677546A (en) * | 1984-08-17 | 1987-06-30 | Signetics | Guarded regions for controlling memory access |
| JPS61206057A (en) * | 1985-03-11 | 1986-09-12 | Hitachi Ltd | Address converting device |
| US4763244A (en) * | 1986-01-15 | 1988-08-09 | Motorola, Inc. | Paged memory management unit capable of selectively supporting multiple address spaces |
| US4682283A (en) * | 1986-02-06 | 1987-07-21 | Rockwell International Corporation | Address range comparison system using multiplexer for detection of range identifier bits stored in dedicated RAM's |
| JPS62237547A (en) * | 1986-04-09 | 1987-10-17 | Hitachi Ltd | Address conversion system |
| JPS6376034A (en) * | 1986-09-19 | 1988-04-06 | Hitachi Ltd | Multiple address space control system |
| EP0282213A3 (en) * | 1987-03-09 | 1991-04-24 | AT&T Corp. | Concurrent context memory management unit |
| JPS63231550A (en) * | 1987-03-19 | 1988-09-27 | Hitachi Ltd | Multiple virtual space control system |
| JPS63278145A (en) * | 1987-05-11 | 1988-11-15 | Nec Corp | Parameter correspondence check system |
| US5008811A (en) * | 1988-02-10 | 1991-04-16 | International Business Machines Corp. | Control mechanism for zero-origin data spaces |
| CA1308202C (en) * | 1988-02-10 | 1992-09-29 | Richard I. Baum | Access register translation means for address generating mechanism for multiple virtual spaces |
| US5023773A (en) * | 1988-02-10 | 1991-06-11 | International Business Machines Corporation | Authorization for selective program access to data in multiple address spaces |
| US5220669A (en) * | 1988-02-10 | 1993-06-15 | International Business Machines Corporation | Linkage mechanism for program isolation |
| US4979098A (en) * | 1988-02-10 | 1990-12-18 | International Business Machines Corporation | Multiple address space token designation, protection controls, designation translation and lookaside |
| US4945480A (en) * | 1988-02-10 | 1990-07-31 | International Business Machines Corporation | Data domain switching on program address space switching and return |
| US5134696A (en) * | 1988-07-28 | 1992-07-28 | International Business Machines Corp. | Virtual lookaside facility |
| JPH0290349A (en) * | 1988-09-28 | 1990-03-29 | Hitachi Ltd | Address space controller for virtual memory system |
| US5226132A (en) * | 1988-09-30 | 1993-07-06 | Hitachi, Ltd. | Multiple virtual addressing using/comparing translation pairs of addresses comprising a space address and an origin address (sto) while using space registers as storage devices for a data processing system |
| US5159677A (en) * | 1988-11-21 | 1992-10-27 | International Business Machines Corp. | Method and system for storing data in and retrieving data from a non-main storage virtual data space |
| JPH02202652A (en) * | 1989-02-01 | 1990-08-10 | Hitachi Ltd | Multiple virtical storage control system |
| US5175828A (en) * | 1989-02-13 | 1992-12-29 | Hewlett-Packard Company | Method and apparatus for dynamically linking subprogram to main program using tabled procedure name comparison |
| JPH0650480B2 (en) * | 1989-05-02 | 1994-06-29 | 株式会社日立製作所 | Multiple virtual memory system and address controller |
| DE4019961C2 (en) * | 1989-06-23 | 1994-11-24 | Hitachi Ltd | Controller for accessing an address translation memory in a processor system |
| JP2768503B2 (en) * | 1989-07-25 | 1998-06-25 | 富士通株式会社 | Virtual memory address space access control method |
| JP2825550B2 (en) * | 1989-09-21 | 1998-11-18 | 株式会社日立製作所 | Multiple virtual space address control method and computer system |
| JPH0679296B2 (en) * | 1989-09-22 | 1994-10-05 | 株式会社日立製作所 | Multiple virtual address space access method and data processing device |
| US5237668A (en) * | 1989-10-20 | 1993-08-17 | International Business Machines Corporation | Process using virtual addressing in a non-privileged instruction to control the copying of a page of data in or between multiple media |
| US5317706A (en) * | 1989-11-15 | 1994-05-31 | Ncr Corporation | Memory expansion method and apparatus in a virtual memory system |
| US5469556A (en) * | 1989-12-12 | 1995-11-21 | Harris Corporation | Resource access security system for controlling access to resources of a data processing system |
| US5210841A (en) * | 1990-01-30 | 1993-05-11 | Advanced Micro Devices, Inc. | External memory accessing system |
| EP0508577A1 (en) * | 1991-03-13 | 1992-10-14 | International Business Machines Corporation | Address translation mechanism |
| US5423013A (en) * | 1991-09-04 | 1995-06-06 | International Business Machines Corporation | System for addressing a very large memory with real or virtual addresses using address mode registers |
| US5390312A (en) * | 1991-09-24 | 1995-02-14 | International Business Machines Corporation | Access look-aside facility |
| US5371890A (en) * | 1991-10-30 | 1994-12-06 | International Business Machines Corporation | Problem state cross-memory communication using communication memory domains |
| EP0543032A1 (en) * | 1991-11-16 | 1993-05-26 | International Business Machines Corporation | Expanded memory addressing scheme |
| US5381537A (en) * | 1991-12-06 | 1995-01-10 | International Business Machines Corporation | Large logical addressing method and means |
| CA2078913A1 (en) * | 1991-12-12 | 1993-06-13 | John J. Reilly | Interprocessor communication system and method for multiprocessor circuitry |
| US5379392A (en) * | 1991-12-17 | 1995-01-03 | Unisys Corporation | Method of and apparatus for rapidly loading addressing registers |
| US5325496A (en) * | 1991-12-24 | 1994-06-28 | Intel Corporation | Selectable pointer validation in a computer system |
| US5361356A (en) * | 1992-03-06 | 1994-11-01 | International Business Machines Corporation | Storage isolation with subspace-group facility |
| JP2788836B2 (en) * | 1992-05-15 | 1998-08-20 | インターナショナル・ビジネス・マシーンズ・コーポレイション | Digital computer system |
| US5237616A (en) * | 1992-09-21 | 1993-08-17 | International Business Machines Corporation | Secure computer system having privileged and unprivileged memories |
| US5548746A (en) * | 1993-11-12 | 1996-08-20 | International Business Machines Corporation | Non-contiguous mapping of I/O addresses to use page protection of a process |
| US5802574A (en) * | 1993-12-28 | 1998-09-01 | Intel Corporation | Method and apparatus for quickly modifying cache state |
| US5583806A (en) * | 1994-10-17 | 1996-12-10 | Advanced Micro Devices, Inc. | Optimized binary adder for concurrently generating effective and intermediate addresses |
| US5734817A (en) * | 1995-03-01 | 1998-03-31 | Unisys Corporation | Method for making a data base available to a user program during data base recovery |
| US5802359A (en) * | 1995-03-31 | 1998-09-01 | International Business Machines Corporation | Mapping processor state into a millicode addressable processor state register array |
| US5761740A (en) * | 1995-11-30 | 1998-06-02 | Unisys Corporation | Method of and apparatus for rapidly loading addressing registers |
| US5732404A (en) * | 1996-03-29 | 1998-03-24 | Unisys Corporation | Flexible expansion of virtual memory addressing |
| US6356991B1 (en) * | 1997-12-31 | 2002-03-12 | Unisys Corporation | Programmable address translation system |
| JP3476376B2 (en) * | 1998-12-16 | 2003-12-10 | 富士通株式会社 | Virtual storage address space access control method and apparatus therefor |
| US6289435B1 (en) * | 1999-05-17 | 2001-09-11 | Creative Technology Ltd. | Re-use of special purposed registers as general purpose registers |
| US6691306B1 (en) * | 2000-12-22 | 2004-02-10 | Lsi Logic Corporation | Use of limited program space of general purpose processor for unlimited sequence of translated instructions |
Family Cites Families (18)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US3412382A (en) * | 1965-11-26 | 1968-11-19 | Massachusetts Inst Technology | Shared-access data processing system |
| JPS5040738B1 (en) * | 1970-06-11 | 1975-12-26 | ||
| JPS532296B2 (en) * | 1973-03-19 | 1978-01-26 | ||
| FR122199A (en) * | 1973-12-17 | |||
| US3949378A (en) * | 1974-12-09 | 1976-04-06 | The United States Of America As Represented By The Secretary Of The Navy | Computer memory addressing employing base and index registers |
| US4042913A (en) * | 1976-04-30 | 1977-08-16 | International Business Machines Corporation | Address key register load/store instruction system |
| US4038645A (en) * | 1976-04-30 | 1977-07-26 | International Business Machines Corporation | Non-translatable storage protection control system |
| US4037214A (en) * | 1976-04-30 | 1977-07-19 | International Business Machines Corporation | Key register controlled accessing system |
| US4035779A (en) * | 1976-04-30 | 1977-07-12 | International Business Machines Corporation | Supervisor address key control system |
| US4050060A (en) * | 1976-04-30 | 1977-09-20 | International Business Machines Corporation | Equate operand address space control system |
| US4037207A (en) * | 1976-04-30 | 1977-07-19 | International Business Machines Corporation | System for controlling address keys under interrupt conditions |
| US4037215A (en) * | 1976-04-30 | 1977-07-19 | International Business Machines Corporation | Key controlled address relocation translation system |
| JPS52149444A (en) * | 1976-06-08 | 1977-12-12 | Fujitsu Ltd | Multiplex virtual space processing data processing system |
| US4084227A (en) * | 1976-09-24 | 1978-04-11 | Sperry Rand Corporation | Virtual address translator |
| US4128875A (en) * | 1976-12-16 | 1978-12-05 | Sperry Rand Corporation | Optional virtual memory system |
| US4096573A (en) * | 1977-04-25 | 1978-06-20 | International Business Machines Corporation | DLAT Synonym control means for common portions of all address spaces |
| US4136385A (en) * | 1977-03-24 | 1979-01-23 | International Business Machines Corporation | Synonym control means for multiple virtual storage systems |
| FR2385147A1 (en) * | 1977-03-24 | 1978-10-20 | Ibm | Control method for dynamic translation of buffer address - uses resolution register to divide each address space into common and private portions by use of status fields |
-
1980
- 1980-03-19 US US06/131,570 patent/US4355355A/en not_active Expired - Lifetime
-
1981
- 1981-01-13 CA CA000368423A patent/CA1153824A/en not_active Expired
- 1981-02-16 DE DE8181101074T patent/DE3176834D1/en not_active Expired
- 1981-02-16 EP EP81101074A patent/EP0036085B1/en not_active Expired
- 1981-03-10 JP JP56033288A patent/JPS6022377B2/en not_active Expired
- 1981-03-18 ES ES500467A patent/ES500467A0/en active Granted
Also Published As
| Publication number | Publication date |
|---|---|
| ES8202968A1 (en) | 1982-03-01 |
| DE3176834D1 (en) | 1988-09-08 |
| EP0036085B1 (en) | 1988-08-03 |
| JPS6022377B2 (en) | 1985-06-01 |
| ES500467A0 (en) | 1982-03-01 |
| JPS56140576A (en) | 1981-11-02 |
| EP0036085A3 (en) | 1983-09-14 |
| US4355355A (en) | 1982-10-19 |
| EP0036085A2 (en) | 1981-09-23 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CA1153824A (en) | Address generating mechanism for multiple virtual spaces | |
| US5075842A (en) | Disabling tag bit recognition and allowing privileged operations to occur in an object-oriented memory protection mechanism | |
| US5075845A (en) | Type management and control in an object oriented memory protection mechanism | |
| US4945480A (en) | Data domain switching on program address space switching and return | |
| US5577231A (en) | Storage access authorization controls in a computer system using dynamic translation of large addresses | |
| US5075848A (en) | Object lifetime control in an object-oriented memory protection mechanism | |
| US5230069A (en) | Apparatus and method for providing private and shared access to host address and data spaces by guest programs in a virtual machine computer system | |
| EP1701269B1 (en) | Invalidating storage, clearing buffer entries | |
| US5574936A (en) | Access control mechanism controlling access to and logical purging of access register translation lookaside buffer (ALB) in a computer system | |
| EP0548315B1 (en) | Translation buffer for virtual machines with address space match | |
| US5280614A (en) | Apparatus and method for controlling access to data using domains | |
| US4979098A (en) | Multiple address space token designation, protection controls, designation translation and lookaside | |
| US4967342A (en) | Data processing system having plurality of processors and channels controlled by plurality of system control programs through interrupt routing | |
| US5339417A (en) | Computer system with two levels of guests | |
| US4472790A (en) | Storage fetch protect override controls | |
| EP0264216B1 (en) | Implied domain addressing | |
| JPS6248258B2 (en) | ||
| US5787309A (en) | Apparatus for protecting storage blocks from being accessed by unwanted I/O programs using I/O program keys and I/O storage keys having M number of bits | |
| US5615327A (en) | Apparatus and method for controlling the initiating of a synchronization protocol for synchronizing a plurality of processors in a multi-processor system | |
| US5280592A (en) | Domain interlock | |
| KR100218617B1 (en) | Method and system for efficient memory management in a data processing system utilizing a dual mode translation lookaside buffer | |
| US5802397A (en) | System for storage protection from unintended I/O access using I/O protection key by providing no control by I/O key entries over access by CP entity | |
| US5305458A (en) | Multiple virtual storage system and address control apparatus having a designation table holding device and translation buffer | |
| EP0171475B1 (en) | Data processing system with logical processor facility | |
| EP0264215A2 (en) | Fast entry to emulation |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| MKEX | Expiry |