US20020056035A1 - Data processor instruction system for grouping instruction with or without a common prefix and data processing system that uses two or more instruction grouping methods - Google Patents
Data processor instruction system for grouping instruction with or without a common prefix and data processing system that uses two or more instruction grouping methods Download PDFInfo
- Publication number
- US20020056035A1 US20020056035A1 US09/170,690 US17069098A US2002056035A1 US 20020056035 A1 US20020056035 A1 US 20020056035A1 US 17069098 A US17069098 A US 17069098A US 2002056035 A1 US2002056035 A1 US 2002056035A1
- Authority
- US
- United States
- Prior art keywords
- instruction
- field
- prefix
- data processor
- length
- 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.)
- Granted
Links
- 238000012545 processing Methods 0.000 title claims description 13
- 238000000034 method Methods 0.000 title description 10
- 230000004048 modification Effects 0.000 claims abstract description 17
- 238000012986 modification Methods 0.000 claims abstract description 17
- 230000002093 peripheral effect Effects 0.000 claims description 6
- 230000001143 conditioned effect Effects 0.000 claims description 4
- 230000000977 initiatory effect Effects 0.000 claims 1
- 238000010586 diagram Methods 0.000 description 15
- 230000003068 static effect Effects 0.000 description 6
- 241000761456 Nops Species 0.000 description 4
- 230000007246 mechanism Effects 0.000 description 4
- 108010020615 nociceptin receptor Proteins 0.000 description 4
- 230000006870 function Effects 0.000 description 3
- 238000004891 communication Methods 0.000 description 2
- 206010000210 abortion Diseases 0.000 description 1
- 239000000872 buffer Substances 0.000 description 1
- 230000001413 cellular effect Effects 0.000 description 1
- 230000008094 contradictory effect Effects 0.000 description 1
- 235000019800 disodium phosphate Nutrition 0.000 description 1
- 239000000796 flavoring agent Substances 0.000 description 1
- 235000019634 flavors Nutrition 0.000 description 1
- 230000014509 gene expression Effects 0.000 description 1
- 230000010511 looping mechanism Effects 0.000 description 1
- 230000000873 masking effect Effects 0.000 description 1
- 230000009467 reduction Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F9/00—Arrangements for program control, e.g. control units
- G06F9/06—Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
- G06F9/30—Arrangements for executing machine instructions, e.g. instruction decode
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F9/00—Arrangements for program control, e.g. control units
- G06F9/06—Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
- G06F9/30—Arrangements for executing machine instructions, e.g. instruction decode
- G06F9/38—Concurrent instruction execution, e.g. pipeline, look ahead
- G06F9/3836—Instruction issuing, e.g. dynamic instruction scheduling or out of order instruction execution
- G06F9/3842—Speculative instruction execution
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F9/00—Arrangements for program control, e.g. control units
- G06F9/06—Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
- G06F9/30—Arrangements for executing machine instructions, e.g. instruction decode
- G06F9/30003—Arrangements for executing specific machine instructions
- G06F9/30007—Arrangements for executing specific machine instructions to perform operations on data operands
- G06F9/30018—Bit or string instructions
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F9/00—Arrangements for program control, e.g. control units
- G06F9/06—Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
- G06F9/30—Arrangements for executing machine instructions, e.g. instruction decode
- G06F9/30181—Instruction operation extension or modification
- G06F9/30185—Instruction operation extension or modification according to one or more bits in the instruction, e.g. prefix, sub-opcode
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F9/00—Arrangements for program control, e.g. control units
- G06F9/06—Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
- G06F9/30—Arrangements for executing machine instructions, e.g. instruction decode
- G06F9/38—Concurrent instruction execution, e.g. pipeline, look ahead
- G06F9/3836—Instruction issuing, e.g. dynamic instruction scheduling or out of order instruction execution
- G06F9/3853—Instruction issuing, e.g. dynamic instruction scheduling or out of order instruction execution of compound instructions
Definitions
- the present invention relates generally to data processors, and more particularly to methods and apparatus for grouping data processor instructions and an instruction system for use therewith.
- a dynamic instruction scheduling system special hardware within the data processor maintains a sliding window of visibility into the sequential instruction stream. Each instruction dispatch cycle, the scheduling hardware selects as many of the visible instructions as can be instruction serial constraints. Additional hardware maintains a record of each instruction while in flight and, depending upon system conditions, either aborts or retires the instruction appropriately.
- An example of a dynamically scheduled data processor is the Motorola MPC604 microprocessor.
- the program preparation software tool In a static instruction scheduling system, the program preparation software tool, after it has generated and, perhaps, optimized the serial instruction stream, reexamines that stream and, based upon information describing the hardware configuration and operating characteristics of the target data processor, groups together those instructions that can safely be executed in parallel. Due to the difficulty of predicting the actions of certain program constructs, such as indirect or computed memory references, it is not possible to guarantee optimal scheduling in advance. To accommodate such non-predictable constructs, some hardware interlocks will usually be provided.
- An example of a statically scheduled data processor was the Multiflow Trace 7/14/28.
- the compiler for the Trace machine was commonly referred to as the Bulldog compiler, the name given it by its original authors while at Yale University.
- the Bulldog compiler was constrained to a VLIW having either 7, 14 or 28 fixed function instruction slots, depending upon the machine model. Any instruction slot for which the compiler could not find useful work was simply filled with a no-operation (NOP) instruction (i.e., all zeroes). Rather than store these useless NOPs in memory, the compiler squashed out the NOPs and preceded the set of useful instruction words comprising each VLIW with a bit map which indicated the location of the squashed NOPs (or, viewed conversely, the useful instruction words).
- NOP no-operation
- the Trace cache/memory controller used the information in the bit map word to regenerate the NOPs so that the cache was filled with fully populated VLIWs.
- the bit map was discarded once the corresponding VLIWs were regenerated during prefetch, and no part of the instruction dispatch or execution hardware was even aware of their existence.
- This mechanism even though it increased by one word the logical length of every VLIW in memory, generally tended to reduce the physical length of the stored VLIWs due to the inability of the compiler to fill all of the instruction slots in every VLIW with useful instruction words. On the other hand, for well designed code, this mechanism could significantly increase the actual code size in memory.
- a more recent example of a statically scheduled data processor is the Texas Instruments TMS32C62xx microprocessor family (C62).
- every instruction word includes a dedicated “P” bit which, if set by the program preparation software tool, indicates to the dispatch hardware that the instruction word can be dispatched in parallel with the following instruction word.
- a simultaneously dispatchable “execution packet” is comprised of an instruction word having a clear P bit and up to a maximum number of preceding instruction words, each having a set P bit.
- U.S. Pat. No. 5,560,028 discloses a variation on this mechanism in which the sense of the parallel dispatch control bit is toggled between each set of parallel-dispatchable instruction words.
- DSP digital signal processors
- FIG. 1 is a block diagram of an embodiment of a data processing system.
- FIG. 2 is a diagram that illustrates registers within the core of the system of FIG. 1.
- FIG. 3 is diagram that illustrates a particular embodiment of a portion of the core of the system of FIG. 1 with a single multiply and accumulate unit.
- FIG. 4 is diagram that illustrates another embodiment of a portion of the core of the system of FIG. 1 with two multiply and accumulate units.
- FIG. 5 is diagram that illustrates another embodiment of a portion of the core of the system of FIG. 1 with four multiply and accumulate units.
- FIG. 6 is a diagram that illustrates a particular embodiment of the multiply and accumulate unit of FIGS. 3 - 5 .
- FIG. 7 is a diagram that illustrates a particular embodiment of the bit field and logic unit of the system of FIG. 1.
- FIG. 8 is a diagram that illustrates a particular embodiment of the address generation unit within the core of the system of FIG. 1.
- FIG. 9 is a diagram that illustrates a particular embodiment of a method of pipelining instructions to be executed by the system of FIG. 1.
- FIG. 10 is a diagram that illustrates a set of fetched computing instructions and an execution set of grouped computing instructions to be executed by the system of FIG. 1.
- FIG. 11 is a diagram that illustrates a prefix dispatch and instruction dispatch operation for the core of the system of FIG. 1.
- FIGS. 12 - 17 illustrate various embodiments of an instruction system that may be used with the system of FIG. 1.
- FIG. 18 is a diagram that illustrates an instruction system that uses static grouping.
- FIG. 19 is a diagram that illustrates different values for a type subfield within the instruction system of FIG. 18.
- one aspect of the present invention relates to an instruction system that includes an instruction root having an operation selection field for selecting an operation to be performed by a data processor and an instruction prefix.
- the prefix has a field that may be any one or more of the following: a conditional execution field for selecting a condition under which a data processor will perform the selected operation, an operand length modification field for modifying the selected operation so as to be performed on an operand having a different length, an instruction group field for selecting a length of an instruction group, and a prefix length selection field for selecting a length of the instruction prefix.
- an instruction system for use in a data processor that has a plurality of registers.
- the instruction system includes an instruction root having a register selection field of width m, for selecting one of 2 m registers in the data processor, and an instruction prefix having a register selection extension field of width n, for extending the width of said register selection field to select one of 2 m+n registers in said data processor.
- an instruction system for use in a data processor is provided.
- the data processor is adapted to execute first and second types of operations sequentially and a plurality of the first types of operations in parallel.
- the instruction system has an operation selection field for selecting one of the first and second types of operations to be executed by the data processor.
- the operation selection field has a type subfield of width n, the values of which are assigned as follows: a first value if said first type of operation is to be executed sequentially; a second value if said first type of operation is to be executed in parallel; and a third value selected from the remaining (2 n -2) values if said second type of operation is to be executed.
- the type subfield may optionally be assigned the second value if a selected instruction of the second type of operations is to be executed sequentially.
- a system for instructing a data processor adapted to execute first and second types of operations sequentially and a plurality of said first type of operations in parallel includes an instruction having an operation selection field for selecting one of the first and second types of operations to be executed by the data processor.
- the operation selection field has a type subfield.
- the type subfield has a first value for instructions of said first type to be executed sequentially, a second value for instructions of said first type to be executed in parallel and for a selected instruction of said second type to be executed sequentially, and a third value otherwise.
- the processing system 10 includes a processor core 12 , a system interface unit (SIU) 14 , a direct memory access unit 16 , a peripheral 18 , such as a serial communication port or timer, internal memory modules 20 , 22 , and an external memory interface module 19 .
- the processing system 10 may also be referred to as a data processor.
- the processor core 12 includes an address register file 26 , a program sequencer 24 , data register files 28 , 29 , address arithmetic logic units 30 (also referred to as address generation units (AGU)), multiply and accumulate (MAC) units ( 32 ) (also referred to generally as data arithmetic logic units (DALU)), a bit field and logic unit 34 , and an optional instruction expansion accelerator 36 .
- the address ALUs 30 are coupled to the address register file 26 via internal bus 60 .
- the multiply and accumulate units 32 are coupled to the data register files 28 , 29 via internal bus 62
- bit field unit 34 is coupled to the data register files 28 , 29 via internal bus 64 .
- the program sequencer 24 is coupled via the instruction bus 44 to the address ALUs 30 , the DALUs 32 , the bit field unit 34 , and the instruction expansion accelerator 36 .
- the system 10 further includes a program bus 38 , a first data bus 40 , a second data bus 42 , a peripheral bus 88 , a direct memory access (DMA) bus 84 , and an external memory interface bus 102 .
- the program bus 38 is coupled to the program sequencer 24 via bus 46 , to SIU 14 via bus 66 , and to internal memory 20 , 22 via buses 72 and 82 respectively.
- the data buses 40 , 42 are coupled to address register file 26 via buses 48 , 50 , to data register files 28 , 29 via buses 52 , 54 , and to instruction expansion accelerator 36 via buses 56 , 58 .
- the data buses 40 , 42 are coupled to memory 20 , 22 via buses 74 - 80 .
- the DMA bus 84 is coupled to SIU 14 via bus 90 , to DMA 16 via bus 92 , to peripheral unit 18 via bus 94 , and to memory units 20 , 22 via buses 95 and 97 respectively.
- the peripheral bus 88 is coupled to the SIU 14 via bus 96 , to DMA 16 via bus 98 , and to peripheral unit 18 via bus 100 .
- the external memory bus 102 is coupled to external memory interface module 19 and is coupled to external memory (not shown) in communication with the system 10 .
- the program bus 38 is 128 bits wide, and the other buses 40 , 42 , 84 , and 88 are 32 bits wide.
- the address register file 26 includes registers R 0 -R 7 , stack pointer (SP), N 0 -N 3 , M 0 -M 2 , MCTL, SA 0 -SA 3 , LC 0 -LC 3 .
- the program sequencer 24 includes the program counter, status register, and operating mode and status registers.
- the data register file 28 includes registers D 0 -D 7 and the data register file 29 includes registers D 8 -D 15 .
- only a single register file may be used to save cost, such as with the one or two MAC configurations. In other high performance applications, more than two register files may also be used.
- FIG. 3 a particular embodiment having one multiply accumulator 32 within the core 12 is illustrated.
- the single MAC 32 is coupled to the register files 28 , 29 via bus 62 .
- the bit field unit 34 is coupled to the data register files 28 , 29 via bus 64 .
- the data register files 28 , 29 are coupled to the data buses 40 , 42 via an intermediate shifter/limiter module 104 and via buses 52 , 54 , 106 .
- the shifter/limiter module 104 is used to convert data in a 40 bit format over buses 106 into a 16 bit format over buses 52 , 54 .
- FIG. 4 a particular embodiment having two multiply accumulators 32 within the core 12 is illustrated. This embodiment is similar to the embodiment of FIG. 3, except that there are two MAC 32 units instead of one MAC 32 .
- An additional bus 62 is needed to couple the data register files 28 , 29 to the second MAC unit 32 .
- the buses 52 , 54 are 32 bits instead of 16 bits to carry the extra data from the two MACs.
- FIG. 5 illustrates an embodiment having four MAC units 32 . It is to be understood that the present invention is not to be limited by the above exemplary configurations and is not limited to the particular number of MAC units 32 or the particular arrangements thereof.
- the MAC unit 32 includes a multiplier 110 and an adder 112 .
- the multiplier 110 receives data input from the data register files 28 , 29 , and multiplies the data elements 111 , 114 to produce a multiplied output 116 that is input to the adder 112 .
- the adder sums a second data input 120 and the multiplier result 116 to produce an accumulated result 122 that is output back to the data register file 28 , 29 .
- the bit field logic unit 34 includes a shifter 124 , a mask bit generator 126 , a logic unit 128 , and a multiplexer 129 .
- the bit field and logic unit receives input via bus 64 and produces an output that is the result of either a shift, a masking operations, e.g. removing certain bits from the input data, and/or a logical operation, such as a logical and, or, exclusive or, or invert operation.
- the result of the selected operation is output as a forty bit result over the two-way bus 64 back to a selected register within the register file 28 , 29 .
- the AGU includes registers M 0 -M 2 , N 0 -N 3 , R 0 -R 7 , SP from address register file 26 , and the modified control register (MCTL).
- the AGU also includes address arithmetic logic units 30 .
- the AGU is coupled to the first and second address buses 140 , 142 and to the program address bus 138 via registers 26 .
- the ALUs 30 receive data from the various registers, perform selected arithmetic operations on the input data, and provide an arithmetic result to selected output registers. For example, an address stored in register R 0 may be output over bus 140 . Thereafter, the register R 0 may be added to an offset from a selected register N 0 and the resulting addition output stored in register R 0 .
- FIG. 9 illustrates a pipeline execution method that is used with the system 10 of FIG. 1.
- the pipeline method includes the execution steps of program pre-fetch 200 , program fetch 202 , dispatch and decode 204 , address generation 206 , and execute 208 .
- FIG. 10 illustrates an instruction fetch set and grouped instruction execution sets.
- the instruction fetch set 210 comprises a sequence of instructions to be executed.
- the instruction sequence includes eight instructions, numbered from N to N+7.
- this instruction sequence may be implemented as one or more instruction buffers.
- the instructions comprising fetch set 210 have been statically grouped at program preparation time into, for example, execution sets 216 , 218 , and 220 .
- a first boundary 212 separates the last two instructions #N and #N+1 of first execution set 216 from the four instructions #N+2-#N+5 comprising second execution set 218
- a second boundary 214 separates the second execution set 218 from the first two instructions #N+6 and #N+7 of third execution set 220 .
- Each of the instructions in the same execution set are dispatched by the system concurrently. All of the instructions in a particular execution set may be dispatched or issued at substantially the same time, and in certain cases where each of the instructions in an execution set have the same cycle count, each of the instructions may be retired at the end of the last execution cycle.
- the embodiment illustrates a prefix dispatch unit 222 , the instructions 210 , an instruction dispatch module 224 , various data arithmetic logic units (DALUs) 32 , and address generation units (AGUs) 30 .
- the prefix dispatch unit 222 , instructions 210 and instruction dispatch module 224 may form the program sequencer 24 .
- the maximum number of instructions that may be grouped in an execution set would be six.
- the first two instructions N, N+1 are prefix instructions.
- the prefix instruction is passed to the prefix dispatch unit 222 .
- the prefix instruction may include one or as illustrated more than one prefix word.
- the prefix dispatch unit 222 converts the prefix instructions to control signals, such as the grouping extension, conditional execution, and loop branch control signals, that are fed to the instruction dispatch unit 224 .
- the instruction dispatch unit 224 then groups the instructions into execution sets and performs certain operations as directed by the control signals from the prefix dispatch unit 222 .
- the grouped instructions are then simultaneously dispatched via a routing mechanism to the appropriate execution units 30 , 32 , 34 for parallel decoding and execution. Simultaneous dispatch means that execution of each of the grouped instructions is initiated during a common clock cycle. In the illustrated embodiment of the system 10 , execution of each of the grouped instructions is initiated during a common clock cycle, but one or more of the grouped instructions may complete execution during a different clock cycle.
- the instruction system 300 includes an instruction prefix word 302 and an instruction root word 316 .
- the instruction prefix 302 includes a conditional execution field 304 and other fields 312 .
- the instruction root 316 includes an operation selection field 306 , an optional operand field 308 , and an optional field 310 for other operations, such as grouping multiple instructions without a prefix.
- the conditional execution field 304 contains information on whether the instruction root 316 is to be conditionally executed.
- conditional execution field 304 is used to determine whether or not the entire group of instruction roots 316 are executed.
- conditional execution field 304 can control selective execution or non-execution of one or more instruction roots or subsets thereof that follow the prefix.
- each of a plurality of conditional execution fields corresponds to each of a plurality of instruction roots to individually control execution of each instruction root in an instruction group.
- the prefix 302 and instruction root(s) 316 form a single, integral instruction 314 .
- the instruction 314 is to be unconditionally executed and the prefix 302 is not needed for any of its other uses, to be described hereinafter, it may be deleted from instruction 314 .
- the instruction system 320 includes a prefix 322 , a first instruction root 326 , and up to 2 N instruction roots 330 .
- the prefix 322 includes a group field 324 that is N bits wide.
- Each of the instruction roots 326 - 330 include an operation selection field 331 , 332 and other optional fields.
- the group field 324 indicates how many instruction roots 326 - 330 immediately following the prefix 322 are to be executed in parallel.
- the group field 324 in prefix 322 explicitly groups the instruction roots into execution sets.
- instruction roots having a variety of operation selection field values can be grouped with other instruction roots, while others values may not be grouped due to hardware or timing constraints.
- the instruction words 346 , 348 are combined with the prefix 342 to form an instruction group 358 , whereas the instruction root 350 is not grouped and forms a single word instruction group 360 .
- the instruction system includes a prefix 380 , and an instruction root 384 .
- the instruction system operates on a variable width data operand 388 , such as a 32 bit or 64 bit data item transferred to or from memory.
- the prefix 380 includes an operand length modification field 382
- the instruction root 384 includes an operation selection field 386 .
- the prefix 380 , instruction root 384 and operand 388 together form an instruction 390 .
- the operand length modification field 382 indicates whether the operand 388 has a first predetermined length, or a modified second length.
- the operation selection field 386 indicates that the instruction root 384 is a move operation for a 16-bit word
- the operand length modification field 382 is clear
- the operand has a length of 16-bits.
- the operand length modification field 382 is set, then the operation becomes a move operation for a 32-bit word and the operand has a length of 32-bits.
- the instruction system 400 includes a prefix 402 and an instruction root 408 .
- the prefix 402 includes one or more register bank selection fields, such as a first register bank selection field 404 and a second register bank selection field 406 .
- the first and second register bank selection fields 404 , 406 are N-bits in length.
- the instruction word 408 includes M-bit register selection fields 410 , 412 .
- the first register bank selection field 404 may be used to indicate a particular bank of registers that are to be used for a first execution unit, such as a first DALU unit 32 .
- the second register bank selection field 406 may be used to indicate a particular bank of registers that are to be used for a second execution unit, such as a second DALU unit 32 .
- N is 1 and M is 3.
- the upper bank registers, such as registers R 8 -R 15 of register file 29 may be selectively accessed using the register bank selection fields 404 - 406 of prefix word 402 instead of using extra space within the instruction word 408 , thereby saving instruction word memory space in certain circumstances.
- the instruction system 420 includes a first prefix word 422 , a second prefix word 424 , and an Nth prefix word 426 .
- Each prefix word 422 - 426 includes a respective prefix length field 428 - 432 .
- the prefix length fields 428 - 432 form a chain, each indicating whether the prefix word chain is completed or whether there is an additional prefix word. In this manner, many prefix words may be linked together in a variable length chain.
- the first prefix word 422 does not include the length field since the first and second prefix words have other fields that indirectly indicate whether the prefix is one or two words in length.
- a particular embodiment of an instruction system using an encoded prefix instruction that combines many of the previously illustrated prefix fields is disclosed below.
- this embodiment there is a one word prefix encoding format and a two-word prefix encoding format.
- the one word prefix has the following binary layout:
- aaaa number of instruction words to be grouped including the prefix word, minus 1
- the prefix has the following format:
- the prefix is at least 3 words long (at least one additional word)
- EEE Data register expansion for DALU UNIT 0; each of the E bits serves as a fourth register field encoding bit, effectively turning FFF ⁇ FFFF, etc.
- FFF,J2,J1 are operand fields.
- BBB the same as EEE, for DALU UNIT 2
- HH Data register expansion for AGU instructions.
- registers in register file 29 are used instead of register file 28 when the H bit is set to 1.
- TT Scaling up MOVE instructions for wider transfers; each bit activates one AGU unit. For each unit, the bit applies if the unit also decodes one of the instructions:
- the instruction 450 includes one or more instruction words, such as a first instruction word 452 and a second instruction word 454 .
- the first instruction word 452 contains an operation selection field 456 , also referred to as an opcode, that may be used to select a plurality of different operations that may be performed on a system such as system 10 .
- the operation selection field 456 includes a type subfield 458 that has a width of n bits. In a particular embodiment, n is equal to two. While a two instruction word instruction has been illustrated as an example, the type subfield within the operation selection field and the use of the type subfield as described herein is applicable to single word and to multiword instructions of many types.
- the type subfield is generally used to provide grouping and multiword instruction information for an execution set with multiple instructions and/or multiple instruction words.
- the instructions that may be executed are divided at a high level into type I instructions and type II instructions.
- Type I instructions may be executed in either a sequential or a parallel manner, i.e. the Type I instructions may be grouped with other instructions for execution in an execution set.
- Type II instructions may be executed sequentially, but have limited ability to be executed in parallel or may not be groupable at all.
- Some type II instructions 510 can not be grouped with any other instruction.
- An example of such an ungroupable instruction is the doze instruction that places the data processor 10 in a low power standby mode.
- the type subfield is set to a first value, such as 01 for segment 502 in the illustrated embodiment.
- the type subfield is set to a second value, such as 00, segment 504 in diagram.
- At least one type II instruction is assigned a third value, such as either 10, segment 510 , or optionally a fourth value, such as 11, segment 512 .
- Using the type subfield with two bits as shown for instruction grouping leads to a reduction in the usable opcode space for the selection of instructions. For example, in a prior system, if a single bit was dedicated for grouping information, the opcode space would disadvantageously be reduced by fifty percent (%50). With a system of two bits, the opcode space would be reduced twenty five percent (25%).
- certain multiword instructions re-use a portion of the opcode space for the second value, i.e. 00, of the type subfield.
- Segment 506 for multiword instructions that are to be executed in parallel with at least one other instruction and segment 508 for multiword instructions that are to be executed sequentially use a portion of the opcode quadrant that shares the type I parallel one word instructions.
- this re-use technique is to use combinations of instructions that would not otherwise be allowed, such as combining a type I groupable instruction with a type II non-groupable instruction.
- multiword sequences may be encoded. In this manner, valuable opcode space is conserved leading to an improved instruction system.
- Type I sequential 502 increment register D 0 (inc d 0 )
- Type I parallel 504 increment register d 0 and move word at address of register r 0 to register d 1 (inc d 0 , move.w (r 0 ), d 1 ) (two single word paired instructions)
- Multiword 506 move address in register after adding offset to register d 3 (inc d 0 , move.b (R 3 +1000), d 3 ) (one single word and one multiword instruction)
- Multiword Sequential 508 add 1000 to d 0 and store in d 1 (add #1000, d 0 , d 1 )
- Type II 510 or 512 take a break in execution (doze)
- Type II instructions in segments 510 , 512 may be the same type, in a particular embodiment, the instructions could be divided such that the instructions with no grouping permitted are placed in segment 510 and the instructions with limited grouping permitted are placed in segment 512 .
- Instruction words having a type subfield of 00 is followed by additional instruction words to be grouped.
- Instruction words having a type subfield other than 00 are the last word in the instruction.
- a 10 in the last instruction word where there is at least two instruction words in the execution set indicates there is at least one multiword instruction to be executed.
- a bit after the type subfield in the second word in an instruction group can be used to distinguish two word instructions with pairing versus three word instructions.
- the distinguishing bit is the bit immediately following the type subfield in the second instruction word of a group of instruction words. Using additional bits in the same or other instruction words may distinguish other instructions such as four word instructions versus three word instructions, and paired one or two word instructions.
- type grouping permitted 00 1 grouped with the next instruction 01 1 end of exec set -- may be grouped with previous type 1 instructions (00) 11 2 end of exec set -- may be grouped with previous type 1 instructions (00) (no fully groupable version) 10 4 only instruction in the set, cannot be grouped
- word L-1 word L 00 10 This is an illegal situation for two one word instructions since 10 cannot be grouped with 00. Therefore this situation is kept for 2 & 3 word instructions word L-2 word L-1 word L don't care 000 10 (L-1,L) is a two word instr. (type 3) -- may be grouped with the previous instruction not 00 001 10 two words, ungroupable (type 4). (word L-2 belongs to the previous exec set) 00 001 10 (L-2, L-1, L) is a three word group instr. (type 3) -- may be grouped with the previous instruction
- an instruction system includes a prefix word having a conditional execution field for selecting a condition under which the data processor will perform an operation to be selected and an instruction word having an operation selection field for selecting the operation to be performed by the data processor.
- the conditional execution field may select a condition under which the data processor will not perform the selected operation.
- the prefix word and the instruction word comprise an instruction or an execution step, the performance of which by the data processor is conditioned upon the conditional execution field of said prefix word.
- the instruction may include more than one instruction word.
- the prefix is optional in the instruction.
- a system for instructing a data processor adapted to perform selected operations where the system includes a prefix word having a conditional execution field for selecting a condition under which the data processor will perform an operation to be selected; and an instruction word having an operation selection field for selecting the operation to be performed by the data processor.
- the conditional execution field may select a condition under which said data processor will not perform the selected operation.
- the prefix word and the instruction word comprise an instruction, the performance of which by the data processor is conditioned upon the conditional execution field of the prefix word.
- the instruction may include more than one instruction word.
- the prefix may be optional in the instruction.
- the present disclosure relates to a system for instructing a data processor adapted to perform selected operations, where the system includes a prefix word having a conditional execution field for selecting a condition under which the data processor will perform an operation to be selected, and an instruction word having an operation selection field for selecting said operation to be performed by the data processor.
- the conditional execution field may select a condition under which the data processor will not perform said selected operation.
- the prefix word and the instruction word may comprise an instruction, the performance of which by the data processor is conditioned upon the conditional execution field of the prefix word.
- the instruction may include more than one instruction word.
- Another embodiment relates to an instruction system in a data processor adapted to perform operations on operands having first and second lengths.
- the instruction system includes a prefix word having an operand length modification field for modifying an operation to be selected so as to be performed on an operand having said second length, and an instruction word having an operation selection field for selecting the operation to be performed on an operand having said first length.
- the operands have a plurality of lengths
- the operation selection field selects an operation to be performed on an operand having a predetermined one of said plurality of lengths
- the operand length modification field modifies the operation so as to be performed on an operand having a different one of said plurality of lengths.
- an instruction system includes a prefix word having an instruction group field for selecting a length of an instruction group, and a plurality of instruction words, each instruction word having an operation selection field for selecting an operation to be performed by said data processor.
- the data processor simultaneously initiates the performance of all operations selected by the instruction words comprising said group.
- the prefix word and the instruction word may form an instruction.
- the prefix word precedes each instruction word in said instruction.
- the instruction group field is of length n and may group 2n of said plurality of instruction words. There are also certain instruction words that cannot be grouped with any of the plurality of instruction words.
- a system for instructing a data processor includes an instruction root and an instruction prefix.
- the instruction root has an operation selection field for selecting an operation to be performed by said data processor.
- the instruction prefix has a field selected from at least one of the following: a conditional execution field for selecting a condition under which the data processor will perform said selected operation, an operand length modification field for modifying the selected operation so as to be performed on an operand having a different length, an instruction group field for selecting a word length of an instruction group that includes the instruction root, and a prefix length selection field for selecting a word length of said instruction prefix.
- the instruction may include more than one instruction root.
- the prefix typically precedes at least one of the instruction roots.
- a system for instructing a data processor having a plurality of registers includes an instruction root having a register selection field of width m, for selecting one of 2 m registers in said data processor; and an instruction prefix having a register selection extension field of width n, for extending the width of said register selection field to select one of 2 m+n registers in said data processor.
Abstract
Description
- The present invention relates generally to data processors, and more particularly to methods and apparatus for grouping data processor instructions and an instruction system for use therewith.
- Many different types of data processors are available. Some data processors have multiple execution units that may be used concurrently. Scheduling of instructions for such data processors can be either dynamic or static. Both types of systems operate on a sequential instruction stream which has been prepared for execution using conventional program preparation software tools, including optimizing assemblers and compilers. In general, dynamic systems require significantly more hardware in the data processor, while static systems require more sophisticated program preparation software techniques. The common goal, however, is to identify and exploit instruction level parallelism inherent in the instruction stream while maintaining the appearance of sequentiality of execution.
- In a dynamic instruction scheduling system, special hardware within the data processor maintains a sliding window of visibility into the sequential instruction stream. Each instruction dispatch cycle, the scheduling hardware selects as many of the visible instructions as can be instruction serial constraints. Additional hardware maintains a record of each instruction while in flight and, depending upon system conditions, either aborts or retires the instruction appropriately. An example of a dynamically scheduled data processor is the Motorola MPC604 microprocessor.
- In a static instruction scheduling system, the program preparation software tool, after it has generated and, perhaps, optimized the serial instruction stream, reexamines that stream and, based upon information describing the hardware configuration and operating characteristics of the target data processor, groups together those instructions that can safely be executed in parallel. Due to the difficulty of predicting the actions of certain program constructs, such as indirect or computed memory references, it is not possible to guarantee optimal scheduling in advance. To accommodate such non-predictable constructs, some hardware interlocks will usually be provided. An example of a statically scheduled data processor was the Multiflow Trace 7/14/28. The compiler for the Trace machine was commonly referred to as the Bulldog compiler, the name given it by its original authors while at Yale University.
- In grouping instructions for the Trace, which was a Very Long Instruction Word (VLIW) machine, the Bulldog compiler was constrained to a VLIW having either 7, 14 or 28 fixed function instruction slots, depending upon the machine model. Any instruction slot for which the compiler could not find useful work was simply filled with a no-operation (NOP) instruction (i.e., all zeroes). Rather than store these useless NOPs in memory, the compiler squashed out the NOPs and preceded the set of useful instruction words comprising each VLIW with a bit map which indicated the location of the squashed NOPs (or, viewed conversely, the useful instruction words). At prefetch time, the Trace cache/memory controller used the information in the bit map word to regenerate the NOPs so that the cache was filled with fully populated VLIWs. The bit map was discarded once the corresponding VLIWs were regenerated during prefetch, and no part of the instruction dispatch or execution hardware was even aware of their existence. This mechanism, even though it increased by one word the logical length of every VLIW in memory, generally tended to reduce the physical length of the stored VLIWs due to the inability of the compiler to fill all of the instruction slots in every VLIW with useful instruction words. On the other hand, for well designed code, this mechanism could significantly increase the actual code size in memory.
- A more recent example of a statically scheduled data processor is the Texas Instruments TMS32C62xx microprocessor family (C62). In the C62, every instruction word includes a dedicated “P” bit which, if set by the program preparation software tool, indicates to the dispatch hardware that the instruction word can be dispatched in parallel with the following instruction word. Thus, a simultaneously dispatchable “execution packet” is comprised of an instruction word having a clear P bit and up to a maximum number of preceding instruction words, each having a set P bit. U.S. Pat. No. 5,560,028 discloses a variation on this mechanism in which the sense of the parallel dispatch control bit is toggled between each set of parallel-dispatchable instruction words. In the above statically scheduled systems, by dedicating a bit of each instruction to the grouping function, a significant portion of the instruction is not useable for other functions, such as encoding data processing operations.
- A premium is placed on execution speed for processing data and instructions using such a multiple execution data processor. However, as the complexity of the data processor architecture is increased using multiple execution units, the computer instruction code size tends to increase. In many applications, increased code size is undesirable due to increased cost and space needed for additional memory. One such application is a class of data processors known as digital signal processors (DSP). DSPs are used in many applications, such as cellular phones, where a premium is placed on small size and low power. It would be desirable for a multiple execution unit data processor to provide faster instruction processing without significantly expanding instruction code size.
- Accordingly, there is a need for improved methods and apparatus for grouping computing instructions and for an improved instruction system.
- While the invention is pointed out with particularity in the appended claims, other features of the invention are disclosed by the following detailed description taken in conjunction with the accompanying drawings, in which:
- FIG. 1 is a block diagram of an embodiment of a data processing system.
- FIG. 2 is a diagram that illustrates registers within the core of the system of FIG. 1.
- FIG. 3 is diagram that illustrates a particular embodiment of a portion of the core of the system of FIG. 1 with a single multiply and accumulate unit.
- FIG. 4 is diagram that illustrates another embodiment of a portion of the core of the system of FIG. 1 with two multiply and accumulate units.
- FIG. 5 is diagram that illustrates another embodiment of a portion of the core of the system of FIG. 1 with four multiply and accumulate units.
- FIG. 6 is a diagram that illustrates a particular embodiment of the multiply and accumulate unit of FIGS.3-5.
- FIG. 7 is a diagram that illustrates a particular embodiment of the bit field and logic unit of the system of FIG. 1.
- FIG. 8 is a diagram that illustrates a particular embodiment of the address generation unit within the core of the system of FIG. 1.
- FIG. 9 is a diagram that illustrates a particular embodiment of a method of pipelining instructions to be executed by the system of FIG. 1.
- FIG. 10 is a diagram that illustrates a set of fetched computing instructions and an execution set of grouped computing instructions to be executed by the system of FIG. 1.
- FIG. 11 is a diagram that illustrates a prefix dispatch and instruction dispatch operation for the core of the system of FIG. 1.
- FIGS.12-17 illustrate various embodiments of an instruction system that may be used with the system of FIG. 1.
- FIG. 18 is a diagram that illustrates an instruction system that uses static grouping.
- FIG. 19 is a diagram that illustrates different values for a type subfield within the instruction system of FIG. 18.
- Generally, one aspect of the present invention relates to an instruction system that includes an instruction root having an operation selection field for selecting an operation to be performed by a data processor and an instruction prefix. The prefix has a field that may be any one or more of the following: a conditional execution field for selecting a condition under which a data processor will perform the selected operation, an operand length modification field for modifying the selected operation so as to be performed on an operand having a different length, an instruction group field for selecting a length of an instruction group, and a prefix length selection field for selecting a length of the instruction prefix.
- In accordance with another aspect of the present invention, an instruction system for use in a data processor that has a plurality of registers is provided. The instruction system includes an instruction root having a register selection field of width m, for selecting one of 2m registers in the data processor, and an instruction prefix having a register selection extension field of width n, for extending the width of said register selection field to select one of 2m+n registers in said data processor.
- In accordance with another aspect of the present invention, an instruction system for use in a data processor is provided. The data processor is adapted to execute first and second types of operations sequentially and a plurality of the first types of operations in parallel. The instruction system has an operation selection field for selecting one of the first and second types of operations to be executed by the data processor. The operation selection field has a type subfield of width n, the values of which are assigned as follows: a first value if said first type of operation is to be executed sequentially; a second value if said first type of operation is to be executed in parallel; and a third value selected from the remaining (2n-2) values if said second type of operation is to be executed. In a particular embodiment, the type subfield may optionally be assigned the second value if a selected instruction of the second type of operations is to be executed sequentially.
- In accordance with another aspect of the present invention, a system for instructing a data processor adapted to execute first and second types of operations sequentially and a plurality of said first type of operations in parallel is provided. The system includes an instruction having an operation selection field for selecting one of the first and second types of operations to be executed by the data processor. The operation selection field has a type subfield. The type subfield has a first value for instructions of said first type to be executed sequentially, a second value for instructions of said first type to be executed in parallel and for a selected instruction of said second type to be executed sequentially, and a third value otherwise.
- It should be noted that the particular terms and expressions employed and the particular structural and operational details disclosed in the detailed description and accompanying drawings are for illustrative purposes only and are not intended to in any way limit the scope of the invention as described in the appended claims.
- Referring to FIG. 1, an embodiment of a
processing system 10 is illustrated. Theprocessing system 10 includes aprocessor core 12, a system interface unit (SIU) 14, a directmemory access unit 16, a peripheral 18, such as a serial communication port or timer,internal memory modules memory interface module 19. Theprocessing system 10 may also be referred to as a data processor. - The
processor core 12 includes anaddress register file 26, aprogram sequencer 24, data register files 28, 29, address arithmetic logic units 30 (also referred to as address generation units (AGU)), multiply and accumulate (MAC) units (32) (also referred to generally as data arithmetic logic units (DALU)), a bit field andlogic unit 34, and an optionalinstruction expansion accelerator 36. Theaddress ALUs 30 are coupled to theaddress register file 26 viainternal bus 60. The multiply and accumulateunits 32 are coupled to the data register files 28, 29 viainternal bus 62, and bitfield unit 34 is coupled to the data register files 28, 29 viainternal bus 64. Theprogram sequencer 24 is coupled via theinstruction bus 44 to theaddress ALUs 30, theDALUs 32, thebit field unit 34, and theinstruction expansion accelerator 36. - The
system 10 further includes aprogram bus 38, afirst data bus 40, asecond data bus 42, aperipheral bus 88, a direct memory access (DMA)bus 84, and an externalmemory interface bus 102. Theprogram bus 38 is coupled to theprogram sequencer 24 viabus 46, toSIU 14 viabus 66, and tointernal memory buses data buses register file 26 viabuses buses instruction expansion accelerator 36 viabuses 56, 58. Thedata buses memory - The
DMA bus 84 is coupled toSIU 14 via bus 90, toDMA 16 via bus 92, toperipheral unit 18 viabus 94, and tomemory units buses peripheral bus 88 is coupled to theSIU 14 viabus 96, toDMA 16 viabus 98, and toperipheral unit 18 viabus 100. Theexternal memory bus 102 is coupled to externalmemory interface module 19 and is coupled to external memory (not shown) in communication with thesystem 10. In the illustrated embodiment, theprogram bus 38 is 128 bits wide, and theother buses - Referring to FIG. 2, a particular embodiment of registers within the
core 12 of thesystem 10 is disclosed. As illustrated, theaddress register file 26 includes registers R0-R7, stack pointer (SP), N0-N3, M0-M2, MCTL, SA0-SA3, LC0-LC3. Theprogram sequencer 24 includes the program counter, status register, and operating mode and status registers. The data registerfile 28 includes registers D0-D7 and the data registerfile 29 includes registers D8-D15. In an alternative embodiment, only a single register file may be used to save cost, such as with the one or two MAC configurations. In other high performance applications, more than two register files may also be used. - Referring to FIG. 3, a particular embodiment having one multiply
accumulator 32 within thecore 12 is illustrated. In this embodiment, thesingle MAC 32 is coupled to the register files 28, 29 viabus 62. Thebit field unit 34 is coupled to the data register files 28, 29 viabus 64. The data register files 28, 29 are coupled to thedata buses limiter module 104 and viabuses limiter module 104 is used to convert data in a 40 bit format overbuses 106 into a 16 bit format overbuses - Referring to FIG. 4, a particular embodiment having two multiply
accumulators 32 within thecore 12 is illustrated. This embodiment is similar to the embodiment of FIG. 3, except that there are twoMAC 32 units instead of oneMAC 32. Anadditional bus 62 is needed to couple the data register files 28, 29 to thesecond MAC unit 32. Also, thebuses MAC units 32. It is to be understood that the present invention is not to be limited by the above exemplary configurations and is not limited to the particular number ofMAC units 32 or the particular arrangements thereof. - Referring to FIG. 6, a particular embodiment of a
MAC unit 32 is illustrated. TheMAC unit 32 includes amultiplier 110 and anadder 112. Themultiplier 110 receives data input from the data register files 28, 29, and multiplies thedata elements output 116 that is input to theadder 112. The adder sums asecond data input 120 and themultiplier result 116 to produce an accumulatedresult 122 that is output back to the data registerfile - Referring to FIG. 7, a particular embodiment of the bit
field logic unit 34 is disclosed. The bitfield logic unit 34 includes ashifter 124, amask bit generator 126, alogic unit 128, and amultiplexer 129. The bit field and logic unit receives input viabus 64 and produces an output that is the result of either a shift, a masking operations, e.g. removing certain bits from the input data, and/or a logical operation, such as a logical and, or, exclusive or, or invert operation. The result of the selected operation is output as a forty bit result over the two-way bus 64 back to a selected register within theregister file - Referring to FIG. 8, a particular embodiment of the address generation unit (AGU) is illustrated. The AGU includes registers M0-M2, N0-N3, R0-R7, SP from
address register file 26, and the modified control register (MCTL). The AGU also includes addressarithmetic logic units 30. The AGU is coupled to the first andsecond address buses program address bus 138 via registers 26. As illustrated in FIG. 8, theALUs 30 receive data from the various registers, perform selected arithmetic operations on the input data, and provide an arithmetic result to selected output registers. For example, an address stored in register R0 may be output overbus 140. Thereafter, the register R0 may be added to an offset from a selected register N0 and the resulting addition output stored in register R0. - FIG. 9 illustrates a pipeline execution method that is used with the
system 10 of FIG. 1. The pipeline method includes the execution steps ofprogram pre-fetch 200, program fetch 202, dispatch and decode 204,address generation 206, and execute 208. - FIG. 10 illustrates an instruction fetch set and grouped instruction execution sets. The instruction fetch set210 comprises a sequence of instructions to be executed. In the illustrated embodiment, the instruction sequence includes eight instructions, numbered from N to N+7. In a hardware implementation, this instruction sequence may be implemented as one or more instruction buffers. The instructions comprising fetch set 210 have been statically grouped at program preparation time into, for example, execution sets 216, 218, and 220. In the illustrated example, a
first boundary 212 separates the last two instructions #N and #N+1 of first execution set 216 from the four instructions #N+2-#N+5 comprising second execution set 218, and asecond boundary 214 separates the second execution set 218 from the first two instructions #N+6 and #N+7 of third execution set 220. Each of the instructions in the same execution set are dispatched by the system concurrently. All of the instructions in a particular execution set may be dispatched or issued at substantially the same time, and in certain cases where each of the instructions in an execution set have the same cycle count, each of the instructions may be retired at the end of the last execution cycle. - Referring to FIG. 11, a particular embodiment of a portion of the
system 10 is illustrated. The embodiment illustrates aprefix dispatch unit 222, theinstructions 210, aninstruction dispatch module 224, various data arithmetic logic units (DALUs) 32, and address generation units (AGUs) 30. Theprefix dispatch unit 222,instructions 210 andinstruction dispatch module 224 may form theprogram sequencer 24. In the illustrated embodiment, since there are six execution units, the maximum number of instructions that may be grouped in an execution set would be six. In the illustrated example, the first two instructions N, N+1 are prefix instructions. The prefix instruction is passed to theprefix dispatch unit 222. The prefix instruction may include one or as illustrated more than one prefix word. Theprefix dispatch unit 222 converts the prefix instructions to control signals, such as the grouping extension, conditional execution, and loop branch control signals, that are fed to theinstruction dispatch unit 224. Theinstruction dispatch unit 224 then groups the instructions into execution sets and performs certain operations as directed by the control signals from theprefix dispatch unit 222. The grouped instructions are then simultaneously dispatched via a routing mechanism to theappropriate execution units system 10, execution of each of the grouped instructions is initiated during a common clock cycle, but one or more of the grouped instructions may complete execution during a different clock cycle. - Referring to FIG. 12, a particular embodiment of an
instruction system 300 that may be used with thesystem 10 is illustrated. Theinstruction system 300 includes aninstruction prefix word 302 and aninstruction root word 316. Theinstruction prefix 302 includes aconditional execution field 304 andother fields 312. Theinstruction root 316 includes anoperation selection field 306, anoptional operand field 308, and anoptional field 310 for other operations, such as grouping multiple instructions without a prefix. Theconditional execution field 304 contains information on whether theinstruction root 316 is to be conditionally executed. For example, if the conditional execution field is a first value, then the instruction root will be unconditionally executed; if a second value, then the instruction root will be conditionally executed, depending upon the value of a predetermined bit within a status or data register. Where there is more than one instruction root that is grouped in an execution set, theconditional execution field 304 is used to determine whether or not the entire group ofinstruction roots 316 are executed. However, in an alternative embodiment, not shown, theconditional execution field 304 can control selective execution or non-execution of one or more instruction roots or subsets thereof that follow the prefix. In another embodiment, each of a plurality of conditional execution fields corresponds to each of a plurality of instruction roots to individually control execution of each instruction root in an instruction group. In a preferred embodiment, theprefix 302 and instruction root(s) 316 form a single,integral instruction 314. In such a system, if theinstruction 314 is to be unconditionally executed and theprefix 302 is not needed for any of its other uses, to be described hereinafter, it may be deleted frominstruction 314. - Referring to FIG. 13, an embodiment of an
instruction system 320 is disclosed. Theinstruction system 320 includes aprefix 322, afirst instruction root 326, and up to 2Ninstruction roots 330. Theprefix 322 includes agroup field 324 that is N bits wide. Each of the instruction roots 326-330 include anoperation selection field group field 324 indicates how many instruction roots 326-330 immediately following theprefix 322 are to be executed in parallel. Thus, thegroup field 324 inprefix 322 explicitly groups the instruction roots into execution sets. In the preferred embodiment, instruction roots having a variety of operation selection field values can be grouped with other instruction roots, while others values may not be grouped due to hardware or timing constraints. As illustrated in FIG. 14, theinstruction words prefix 342 to form an instruction group 358, whereas theinstruction root 350 is not grouped and forms a singleword instruction group 360. - Referring to FIG. 15, another embodiment of an instruction system is illustrated. The instruction system includes a
prefix 380, and aninstruction root 384. The instruction system operates on a variablewidth data operand 388, such as a 32 bit or 64 bit data item transferred to or from memory. Theprefix 380 includes an operandlength modification field 382, and theinstruction root 384 includes anoperation selection field 386. Theprefix 380,instruction root 384 andoperand 388 together form aninstruction 390. The operandlength modification field 382 indicates whether theoperand 388 has a first predetermined length, or a modified second length. For example, where theoperation selection field 386 indicates that theinstruction root 384 is a move operation for a 16-bit word, and where the operandlength modification field 382 is clear, the operand has a length of 16-bits. However, if the operandlength modification field 382 is set, then the operation becomes a move operation for a 32-bit word and the operand has a length of 32-bits. By encoding the length modification field into theprefix word 380, the number of operations that may be selected for execution is increased but without having to add any extra length to the operationselect field 386 within theinstruction root 384, thereby reducing code size in certain circumstances. - Referring to FIG. 16, an embodiment of an
instruction system 400 is disclosed. Theinstruction system 400 includes aprefix 402 and aninstruction root 408. Theprefix 402 includes one or more register bank selection fields, such as a first registerbank selection field 404 and a second registerbank selection field 406. The first and second register bank selection fields 404, 406 are N-bits in length. Theinstruction word 408 includes M-bitregister selection fields bank selection field 404 may be used to indicate a particular bank of registers that are to be used for a first execution unit, such as afirst DALU unit 32. The second registerbank selection field 406 may be used to indicate a particular bank of registers that are to be used for a second execution unit, such as asecond DALU unit 32. In a particular embodiment, N is 1 and M is 3. In this embodiment, there are two different register banks, such as the register files 28 and 29 of thesystem 10, and each of the register banks has eight registers. In this manner, the upper bank registers, such as registers R8-R15 ofregister file 29 may be selectively accessed using the register bank selection fields 404-406 ofprefix word 402 instead of using extra space within theinstruction word 408, thereby saving instruction word memory space in certain circumstances. - Referring to FIG. 17, an embodiment of an
instruction system 420 with variable length prefix words is shown. Theinstruction system 420 includes afirst prefix word 422, asecond prefix word 424, and an Nth prefix word 426. Each prefix word 422-426 includes a respective prefix length field 428-432. The prefix length fields 428-432 form a chain, each indicating whether the prefix word chain is completed or whether there is an additional prefix word. In this manner, many prefix words may be linked together in a variable length chain. However, in the preferred embodiment thefirst prefix word 422 does not include the length field since the first and second prefix words have other fields that indirectly indicate whether the prefix is one or two words in length. - A particular embodiment of an instruction system using an encoded prefix instruction that combines many of the previously illustrated prefix fields is disclosed below. In this embodiment, there is a one word prefix encoding format and a two-word prefix encoding format.
- In the one word prefix encoding format, the one word prefix has the following binary layout:
- 1001
aaaa 11|0-pjcc - where:
- aaaa: number of instruction words to be grouped including the prefix word,
minus 1 - if aaaa==0: this is a NOP instruction, which is not dispatched for 4 DALU flavor, valid values for aaaa are 0 to 6 for 4 DALU system
- cc: conditional execution of the entire execution set
- 00 unconditionally executed
- 01 reserved
- conditionally executed, if status register T is set
- 11 conditionally executed, if status register T is clear
- pj: two bits used in a looping mechanism
- “-”: reserved for future use such as additional looping information.
- In the two-word form, the prefix has the following format:
- src0 src1 dest
- encoding: 0011 aaaa TTHH pjcc 101M bBeE bBeE bBeE
- where:
- aaaa, cc, pj: same as for one word prefix
- For example:
- aaaa==3→two word prefix+2 grouped instruction words
- M: reserved for future expansion for longer prefix words.
- 0: the prefix is two-words
- 1: the prefix is at least 3 words long (at least one additional word)
- EEE: Data register expansion for
DALU UNIT 0; each of the E bits serves as a fourth register field encoding bit, effectively turning FFF→FFFF, etc. - --1: high data reg used for the FFF field
- -1-: high data reg used for the J2 field
- 1--: high data reg used for the J1 field
- Note: FFF,J2,J1 are operand fields.
- For example:
- DALU DJ1,DJ2,DF
- In case of two operand DALU unit instructions, only -EE is used.
- In case of one operand DALU unit instructions, only --E is used.
- eee: the same as EEE, for
DALU UNIT 1 - BBB: the same as EEE, for
DALU UNIT 2 - bbb: the same as EEE, for
DALU UNIT 3 - HH: Data register expansion for AGU instructions.
- -1 high data reg used for MOVE operation in
AGU 0 - 1- high data reg used for MOVE operation in
AGU 1 - For example, in the
system 10, the registers inregister file 29 are used instead ofregister file 28 when the H bit is set to 1. - TT: Scaling up MOVE instructions for wider transfers; each bit activates one AGU unit. For each unit, the bit applies if the unit also decodes one of the instructions:
- MOVE.2W, MOVE.2F, MOVE.L
- If the T bit is 0: leave these instructions as originally defined if the T bit is 1:
- MOVE.2W→MOVE.4W
- MOVE.2F→MOVE.4F
- MOVES.2F→MOVES.4F
- MOVE.L ->MOVE.2L
- -1 make instruction in AGU0 be interpreted as a wider move
- 1- make instruction in
AGU 1 be interpreted as a wider move - In principle these combinations are reserved for the following, if defined:
- MOVE.2L
- MOVE.4W
- MOVE.4F
- In accordance with another embodiment, a method of grouping instructions without using a prefix instruction is now disclosed. Referring to FIG. 18, an instruction without a prefix word that may be used to perform static grouping is illustrated. The
instruction 450 includes one or more instruction words, such as afirst instruction word 452 and asecond instruction word 454. Thefirst instruction word 452 contains anoperation selection field 456, also referred to as an opcode, that may be used to select a plurality of different operations that may be performed on a system such assystem 10. Theoperation selection field 456 includes atype subfield 458 that has a width of n bits. In a particular embodiment, n is equal to two. While a two instruction word instruction has been illustrated as an example, the type subfield within the operation selection field and the use of the type subfield as described herein is applicable to single word and to multiword instructions of many types. - Referring to FIG. 19, a diagram is presented that illustrates how different types of instructions that are encoded by the operation selection field within an instruction may be mapped to the value within the type subfield. The type subfield is generally used to provide grouping and multiword instruction information for an execution set with multiple instructions and/or multiple instruction words. The instructions that may be executed are divided at a high level into type I instructions and type II instructions. Type I instructions may be executed in either a sequential or a parallel manner, i.e. the Type I instructions may be grouped with other instructions for execution in an execution set. Type II instructions may be executed sequentially, but have limited ability to be executed in parallel or may not be groupable at all. Some type II
instructions 510 can not be grouped with any other instruction. An example of such an ungroupable instruction is the doze instruction that places thedata processor 10 in a low power standby mode. - As shown in FIG. 19, for type I instructions that are to be executed sequentially, the type subfield is set to a first value, such as 01 for
segment 502 in the illustrated embodiment. For type I instructions that are to be executed in parallel with other one word instructions, the type subfield is set to a second value, such as 00,segment 504 in diagram. At least one type II instruction is assigned a third value, such as either 10,segment 510, or optionally a fourth value, such as 11,segment 512. Using the type subfield with two bits as shown for instruction grouping leads to a reduction in the usable opcode space for the selection of instructions. For example, in a prior system, if a single bit was dedicated for grouping information, the opcode space would disadvantageously be reduced by fifty percent (%50). With a system of two bits, the opcode space would be reduced twenty five percent (25%). - However, in the illustrated embodiment, to conserve useful opcode space, certain multiword instructions re-use a portion of the opcode space for the second value, i.e. 00, of the type subfield.
Segment 506 for multiword instructions that are to be executed in parallel with at least one other instruction andsegment 508 for multiword instructions that are to be executed sequentially use a portion of the opcode quadrant that shares the type I parallel one word instructions. One method of implementation this re-use technique is to use combinations of instructions that would not otherwise be allowed, such as combining a type I groupable instruction with a type II non-groupable instruction. By using otherwise contradictory instruction combinations, multiword sequences may be encoded. In this manner, valuable opcode space is conserved leading to an improved instruction system. - An example of the different types of instructions described generally above is presented below. The actual assembly instructions are in parentheses.
- Type I sequential502: increment register D0 (inc d0)
- Type I parallel504: increment register d0 and move word at address of register r0 to register d1 (inc d0, move.w (r0), d1) (two single word paired instructions)
- Multiword506: move address in register after adding offset to register d3 (inc d0, move.b (R3+1000), d3) (one single word and one multiword instruction)
- Multiword Sequential508: add 1000 to d0 and store in d1 (add #1000, d0, d1)
- Type II510 or 512: take a break in execution (doze)
- While the Type II instructions in
segments segment 510 and the instructions with limited grouping permitted are placed insegment 512. - With instructions having the type subfield as described above, general rules for grouping instructions may be used. An example of such rules are:
- 1. Instruction words having a type subfield of 00 is followed by additional instruction words to be grouped.
- 2. Instruction words having a type subfield other than 00 are the last word in the instruction.
- 3. A 10 in the last instruction word where there is at least two instruction words in the execution set indicates there is at least one multiword instruction to be executed.
- 4. A bit after the type subfield in the second word in an instruction group can be used to distinguish two word instructions with pairing versus three word instructions. In one embodiment, the distinguishing bit is the bit immediately following the type subfield in the second instruction word of a group of instruction words. Using additional bits in the same or other instruction words may distinguish other instructions such as four word instructions versus three word instructions, and paired one or two word instructions.
- In a particular illustrative embodiment, there are 3 ways to convey grouping information for grouping instructions into execution sets:
- using the 2-3 most significant bits in the instruction—no prefix is used
- using a 1-word prefix for an execution set
- using a 2-word prefix for an execution set
- The following grouping algorithm chooses what option to use (when the instructions to be grouped are given) to reduce the instruction code size.
- The algorithm is:
- are registers d8-d15 used in the execution set?
- yes→use 2-word prefix
- no→continue
- is the set conditionally executed (Ifc), or should it convey looping information?
- yes→use 1-word prefix.
- no→continue
- does the execution set have just one instruction?
- yes→no prefix needed
- no→continue
- are the instructions in the set groupable according to their type?
- yes→no prefix needed—encode grouping using the 2 most significant bits only
- no→use 1-w prefix
- A further description in tabular form of a static grouping embodiment is disclosed below.
- One-word instruction type encoding:
- 2 most significant bits of the
type subfield 458: type grouping permitted 00 1 grouped with the next instruction 01 1 end of exec set -- may be grouped with previous type 1 instructions (00)11 2 end of exec set -- may be grouped with previous type 1 instructions (00) (no fullygroupable version) 10 4 only instruction in the set, cannot be grouped - Two and three word identification:
- (L—last instruction in the execution set)
word L-1 word L 00 10 This is an illegal situation for two one word instructions since 10 cannot be grouped with 00. Therefore this situation is kept for 2 & 3 word instructions word L-2 word L-1 word L don't care 000 10 (L-1,L) is a two word instr. (type 3) -- may be grouped with the previous instruction not 00 001 10 two words, ungroupable (type 4). (word L-2 belongs to the previous exec set) 00 001 10 (L-2, L-1, L) is a three word group instr. (type 3) -- may be grouped with the previous instruction - Certain additional aspects of the present disclosure are summarized below.
- In a data processor adapted to perform selected operations, an instruction system includes a prefix word having a conditional execution field for selecting a condition under which the data processor will perform an operation to be selected and an instruction word having an operation selection field for selecting the operation to be performed by the data processor. The conditional execution field may select a condition under which the data processor will not perform the selected operation. In one embodiment, the prefix word and the instruction word comprise an instruction or an execution step, the performance of which by the data processor is conditioned upon the conditional execution field of said prefix word. The instruction may include more than one instruction word. In another embodiment, the prefix is optional in the instruction.
- A system for instructing a data processor adapted to perform selected operations, where the system includes a prefix word having a conditional execution field for selecting a condition under which the data processor will perform an operation to be selected; and an instruction word having an operation selection field for selecting the operation to be performed by the data processor. The conditional execution field may select a condition under which said data processor will not perform the selected operation. In one embodiment, the prefix word and the instruction word comprise an instruction, the performance of which by the data processor is conditioned upon the conditional execution field of the prefix word. The instruction may include more than one instruction word. Also, the prefix may be optional in the instruction.
- In another embodiment, the present disclosure relates to a system for instructing a data processor adapted to perform selected operations, where the system includes a prefix word having a conditional execution field for selecting a condition under which the data processor will perform an operation to be selected, and an instruction word having an operation selection field for selecting said operation to be performed by the data processor. The conditional execution field may select a condition under which the data processor will not perform said selected operation.
- The prefix word and the instruction word may comprise an instruction, the performance of which by the data processor is conditioned upon the conditional execution field of the prefix word. The instruction may include more than one instruction word.
- Another embodiment relates to an instruction system in a data processor adapted to perform operations on operands having first and second lengths. The instruction system includes a prefix word having an operand length modification field for modifying an operation to be selected so as to be performed on an operand having said second length, and an instruction word having an operation selection field for selecting the operation to be performed on an operand having said first length. In one embodiment, the operands have a plurality of lengths, the operation selection field selects an operation to be performed on an operand having a predetermined one of said plurality of lengths, and the operand length modification field modifies the operation so as to be performed on an operand having a different one of said plurality of lengths.
- In a data processor, an instruction system includes a prefix word having an instruction group field for selecting a length of an instruction group, and a plurality of instruction words, each instruction word having an operation selection field for selecting an operation to be performed by said data processor. The data processor simultaneously initiates the performance of all operations selected by the instruction words comprising said group. The prefix word and the instruction word may form an instruction. In one embodiment, the prefix word precedes each instruction word in said instruction. In another embodiment, the instruction group field is of length n and may group 2n of said plurality of instruction words. There are also certain instruction words that cannot be grouped with any of the plurality of instruction words.
- A system for instructing a data processor is provided. The system includes an instruction root and an instruction prefix. The instruction root has an operation selection field for selecting an operation to be performed by said data processor. The instruction prefix has a field selected from at least one of the following: a conditional execution field for selecting a condition under which the data processor will perform said selected operation, an operand length modification field for modifying the selected operation so as to be performed on an operand having a different length, an instruction group field for selecting a word length of an instruction group that includes the instruction root, and a prefix length selection field for selecting a word length of said instruction prefix. The instruction may include more than one instruction root. The prefix typically precedes at least one of the instruction roots.
- A system for instructing a data processor having a plurality of registers. The system includes an instruction root having a register selection field of width m, for selecting one of 2m registers in said data processor; and an instruction prefix having a register selection extension field of width n, for extending the width of said register selection field to select one of 2m+n registers in said data processor.
- Thus, there has been described herein an embodiment including at least one preferred embodiment of an improved method and apparatus for grouping data processor instructions and embodiments of instruction systems. It will be apparent to those skilled in the art that the disclosed subject matter may be modified in numerous ways and may assume many embodiments other than the preferred form specifically set out and described above. For example, the specific elements and configuration of the
data processor 10 is described by way of example only. Many other alternative systems may be used. For example, theDMA 16,memory
Claims (24)
Priority Applications (10)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
US09/170,690 US6418527B1 (en) | 1998-10-13 | 1998-10-13 | Data processor instruction system for grouping instructions with or without a common prefix and data processing system that uses two or more instruction grouping methods |
SG9904821A SG95605A1 (en) | 1998-10-13 | 1999-09-24 | Data processor system and instruction system using grouping |
AT99118986T ATE266226T1 (en) | 1998-10-13 | 1999-09-27 | DATA PROCESSING SYSTEM WITH CONDITIONAL EXECUTION OF ADVANCED COMPOUND COMMANDS |
ES99118986T ES2221282T3 (en) | 1998-10-13 | 1999-09-27 | DATA PROCESSING SYSTEM WITH CONDITIONAL EXECUTION OF EXTENDED COMPOSITE INSTRUCTIONS. |
EP99118986A EP0994413B1 (en) | 1998-10-13 | 1999-09-27 | Data processing system with conditional execution of extended compound instructions |
DE69916962T DE69916962T2 (en) | 1998-10-13 | 1999-09-27 | Data processing system with conditional execution of extended compound instructions |
TW088117156A TW497073B (en) | 1998-10-13 | 1999-10-05 | Data processor system and instruction system using grouping |
JP11288062A JP2000122864A (en) | 1998-10-13 | 1999-10-08 | Data processing system and instruction system using aggregating |
CN99121094A CN1129843C (en) | 1998-10-13 | 1999-10-12 | Use composite data processor systemand instruction system |
KR1019990044023A KR100690225B1 (en) | 1998-10-13 | 1999-10-12 | Data processor system and instruction system using grouping |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
US09/170,690 US6418527B1 (en) | 1998-10-13 | 1998-10-13 | Data processor instruction system for grouping instructions with or without a common prefix and data processing system that uses two or more instruction grouping methods |
Publications (2)
Publication Number | Publication Date |
---|---|
US20020056035A1 true US20020056035A1 (en) | 2002-05-09 |
US6418527B1 US6418527B1 (en) | 2002-07-09 |
Family
ID=22620876
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
US09/170,690 Expired - Lifetime US6418527B1 (en) | 1998-10-13 | 1998-10-13 | Data processor instruction system for grouping instructions with or without a common prefix and data processing system that uses two or more instruction grouping methods |
Country Status (10)
Country | Link |
---|---|
US (1) | US6418527B1 (en) |
EP (1) | EP0994413B1 (en) |
JP (1) | JP2000122864A (en) |
KR (1) | KR100690225B1 (en) |
CN (1) | CN1129843C (en) |
AT (1) | ATE266226T1 (en) |
DE (1) | DE69916962T2 (en) |
ES (1) | ES2221282T3 (en) |
SG (1) | SG95605A1 (en) |
TW (1) | TW497073B (en) |
Cited By (12)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20040039897A1 (en) * | 2002-06-24 | 2004-02-26 | Seiko Epson Corporation | Data processing device and electronic equipment |
US20050223379A1 (en) * | 2000-05-19 | 2005-10-06 | Rehg James M | On-line scheduling of constrained dynamic applications for parallel targets |
US20060218379A1 (en) * | 2005-03-23 | 2006-09-28 | Lucian Codrescu | Method and system for encoding variable length packets with variable instruction sizes |
GB2486737A (en) * | 2010-12-24 | 2012-06-27 | Cambridge Silicon Radio Ltd | Prefix instruction, which modifies an operand from a table lookup to an immediate value |
GB2486740A (en) * | 2010-12-24 | 2012-06-27 | Cambridge Silicon Radio Ltd | Long instructions for a data processing system encoded as a series of shorter instructions |
US20140019990A1 (en) * | 2011-03-30 | 2014-01-16 | Freescale Semiconductor, Inc. | Integrated circuit device and method for enabling cross-context access |
US20160139925A1 (en) * | 2014-11-17 | 2016-05-19 | International Business Machines Corporation | Techniques for identifying instructions for decode-time instruction optimization grouping in view of cache boundaries |
US20170024216A1 (en) * | 2014-03-12 | 2017-01-26 | Samsung Electronics Co., Ltd. | Method and device for processing vliw instruction, and method and device for generating instruction for processing vliw instruction |
US20170192788A1 (en) * | 2016-01-05 | 2017-07-06 | Intel Corporation | Binary translation support using processor instruction prefixes |
US9733940B2 (en) | 2014-11-17 | 2017-08-15 | International Business Machines Corporation | Techniques for instruction group formation for decode-time instruction optimization based on feedback |
CN108139911A (en) * | 2015-10-22 | 2018-06-08 | 德州仪器公司 | In the same execution packet of vliw processor specification is performed using having ready conditions for the instruction for expansion slot of having ready conditions |
US20190236442A1 (en) * | 2016-08-05 | 2019-08-01 | Cambricon Technologies Corporation Limited | Operation unit, method and device capable of supporting operation data of different bit widths |
Families Citing this family (29)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
EP1039375A1 (en) * | 1999-03-19 | 2000-09-27 | Motorola, Inc. | Method and apparatus for implementing zero overhead loops |
US6606700B1 (en) * | 2000-02-26 | 2003-08-12 | Qualcomm, Incorporated | DSP with dual-mac processor and dual-mac coprocessor |
US7020763B2 (en) * | 2000-03-08 | 2006-03-28 | Sun Microsystems, Inc. | Computer processing architecture having a scalable number of processing paths and pipelines |
US6725360B1 (en) * | 2000-03-31 | 2004-04-20 | Intel Corporation | Selectively processing different size data in multiplier and ALU paths in parallel |
GB2362733B (en) * | 2000-05-25 | 2002-02-27 | Siroyan Ltd | Processors having compressed instructions. |
US6415376B1 (en) * | 2000-06-16 | 2002-07-02 | Conexant Sytems, Inc. | Apparatus and method for issue grouping of instructions in a VLIW processor |
US6877084B1 (en) * | 2000-08-09 | 2005-04-05 | Advanced Micro Devices, Inc. | Central processing unit (CPU) accessing an extended register set in an extended register mode |
AU2002348842A1 (en) * | 2001-11-26 | 2003-06-10 | Koninklijke Philips Electronics N.V. | Wlim architecture with power down instruction |
US7697946B2 (en) * | 2002-06-04 | 2010-04-13 | Forster Ian J | Reflective communication using radio-frequency devices |
US6944749B2 (en) * | 2002-07-29 | 2005-09-13 | Faraday Technology Corp. | Method for quickly determining length of an execution package |
US6865662B2 (en) * | 2002-08-08 | 2005-03-08 | Faraday Technology Corp. | Controlling VLIW instruction operations supply to functional units using switches based on condition head field |
US7861062B2 (en) * | 2003-06-25 | 2010-12-28 | Koninklijke Philips Electronics N.V. | Data processing device with instruction controlled clock speed |
US7340588B2 (en) * | 2003-11-24 | 2008-03-04 | International Business Machines Corporation | Extending the number of instruction bits in processors with fixed length instructions, in a manner compatible with existing code |
US7873815B2 (en) | 2004-03-04 | 2011-01-18 | Qualcomm Incorporated | Digital signal processors with configurable dual-MAC and dual-ALU |
US20060149926A1 (en) * | 2004-12-23 | 2006-07-06 | Yuval Sapir | Control words for instruction packets of processors and methods thereof |
US20060150171A1 (en) * | 2004-12-28 | 2006-07-06 | Ceva D.S.P. Ltd. | Control words for instruction packets of processors and methods thereof |
US20060149922A1 (en) * | 2004-12-28 | 2006-07-06 | Ceva D.S.P. Ltd. | Multiple computational clusters in processors and methods thereof |
US7350040B2 (en) * | 2005-03-03 | 2008-03-25 | Microsoft Corporation | Method and system for securing metadata to detect unauthorized access |
US7793078B2 (en) * | 2005-04-01 | 2010-09-07 | Arm Limited | Multiple instruction set data processing system with conditional branch instructions of a first instruction set and a second instruction set sharing a same instruction encoding |
JP5217431B2 (en) * | 2007-12-28 | 2013-06-19 | 富士通株式会社 | Arithmetic processing device and control method of arithmetic processing device |
US8407680B2 (en) * | 2008-12-16 | 2013-03-26 | International Business Machines Corporation | Operand data structure for block computation |
US8327345B2 (en) * | 2008-12-16 | 2012-12-04 | International Business Machines Corporation | Computation table for block computation |
US8285971B2 (en) * | 2008-12-16 | 2012-10-09 | International Business Machines Corporation | Block driven computation with an address generation accelerator |
US8458439B2 (en) * | 2008-12-16 | 2013-06-04 | International Business Machines Corporation | Block driven computation using a caching policy specified in an operand data structure |
US8281106B2 (en) * | 2008-12-16 | 2012-10-02 | International Business Machines Corporation | Specifying an addressing relationship in an operand data structure |
US8898433B2 (en) * | 2012-04-26 | 2014-11-25 | Avago Technologies General Ip (Singapore) Pte. Ltd. | Efficient extraction of execution sets from fetch sets |
US10761849B2 (en) * | 2016-09-22 | 2020-09-01 | Intel Corporation | Processors, methods, systems, and instruction conversion modules for instructions with compact instruction encodings due to use of context of a prior instruction |
CN111813446A (en) * | 2019-04-12 | 2020-10-23 | 杭州中天微系统有限公司 | Processing method and processing device for data loading and storing instructions |
CN116917859A (en) * | 2022-01-26 | 2023-10-20 | 谷歌有限责任公司 | Parallel decoded instruction set computer architecture with variable length instructions |
Family Cites Families (14)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JPH0827716B2 (en) * | 1985-10-25 | 1996-03-21 | 株式会社日立製作所 | Data processing device and data processing method |
EP0221577B1 (en) * | 1985-11-08 | 1996-01-17 | Nec Corporation | Microprogram control unit |
US5214763A (en) * | 1990-05-10 | 1993-05-25 | International Business Machines Corporation | Digital computer system capable of processing two or more instructions in parallel and having a coche and instruction compounding mechanism |
DE69123629T2 (en) * | 1990-05-04 | 1997-06-12 | Ibm | Machine architecture for scalar compound instruction set |
JPH04156613A (en) * | 1990-10-20 | 1992-05-29 | Fujitsu Ltd | Instruction buffer device |
DE69429061T2 (en) * | 1993-10-29 | 2002-07-18 | Advanced Micro Devices Inc | Superskalarmikroprozessoren |
US5689672A (en) * | 1993-10-29 | 1997-11-18 | Advanced Micro Devices, Inc. | Pre-decoded instruction cache and method therefor particularly suitable for variable byte-length instructions |
DE69427265T2 (en) * | 1993-10-29 | 2002-05-02 | Advanced Micro Devices Inc | Superskalarbefehlsdekoder |
EP0950946B1 (en) | 1993-11-05 | 2001-08-16 | Intergraph Corporation | Software scheduled superscaler computer architecture |
DE69433124T2 (en) * | 1993-11-05 | 2004-05-27 | Intergraph Corp., Huntsville | Command memory with associative crossbar switch |
US5630083A (en) * | 1994-03-01 | 1997-05-13 | Intel Corporation | Decoder for decoding multiple instructions in parallel |
US5822778A (en) * | 1995-06-07 | 1998-10-13 | Advanced Micro Devices, Inc. | Microprocessor and method of using a segment override prefix instruction field to expand the register file |
JPH09265397A (en) * | 1996-03-29 | 1997-10-07 | Hitachi Ltd | Processor for vliw instruction |
US6275927B2 (en) * | 1998-09-21 | 2001-08-14 | Advanced Micro Devices. | Compressing variable-length instruction prefix bytes |
-
1998
- 1998-10-13 US US09/170,690 patent/US6418527B1/en not_active Expired - Lifetime
-
1999
- 1999-09-24 SG SG9904821A patent/SG95605A1/en unknown
- 1999-09-27 ES ES99118986T patent/ES2221282T3/en not_active Expired - Lifetime
- 1999-09-27 EP EP99118986A patent/EP0994413B1/en not_active Expired - Lifetime
- 1999-09-27 DE DE69916962T patent/DE69916962T2/en not_active Expired - Fee Related
- 1999-09-27 AT AT99118986T patent/ATE266226T1/en not_active IP Right Cessation
- 1999-10-05 TW TW088117156A patent/TW497073B/en not_active IP Right Cessation
- 1999-10-08 JP JP11288062A patent/JP2000122864A/en active Pending
- 1999-10-12 KR KR1019990044023A patent/KR100690225B1/en not_active IP Right Cessation
- 1999-10-12 CN CN99121094A patent/CN1129843C/en not_active Expired - Fee Related
Cited By (30)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20050223379A1 (en) * | 2000-05-19 | 2005-10-06 | Rehg James M | On-line scheduling of constrained dynamic applications for parallel targets |
US7010788B1 (en) * | 2000-05-19 | 2006-03-07 | Hewlett-Packard Development Company, L.P. | System for computing the optimal static schedule using the stored task execution costs with recent schedule execution costs |
US7073175B2 (en) * | 2000-05-19 | 2006-07-04 | Hewlett-Packard Development Company, Inc. | On-line scheduling of constrained dynamic applications for parallel targets |
US20040039897A1 (en) * | 2002-06-24 | 2004-02-26 | Seiko Epson Corporation | Data processing device and electronic equipment |
US7340589B2 (en) * | 2002-06-24 | 2008-03-04 | Seiko Epson Corporation | Shift prefix instruction decoder for modifying register information necessary for decoding the target instruction |
US20060218379A1 (en) * | 2005-03-23 | 2006-09-28 | Lucian Codrescu | Method and system for encoding variable length packets with variable instruction sizes |
US7526633B2 (en) * | 2005-03-23 | 2009-04-28 | Qualcomm Incorporated | Method and system for encoding variable length packets with variable instruction sizes |
GB2565242A (en) * | 2010-12-24 | 2019-02-06 | Qualcomm Technologies Int Ltd | Encapsulated instruction set |
GB2486737B (en) * | 2010-12-24 | 2018-09-19 | Qualcomm Technologies Int Ltd | Instruction execution |
US9032187B2 (en) | 2010-12-24 | 2015-05-12 | Cambridge Silicon Radio Limited | Instruction execution |
US9152421B2 (en) | 2010-12-24 | 2015-10-06 | Qualcomm Incorporated | Assembling long instruction encapsulated in short instructions according to method identified by format bits |
GB2565242B (en) * | 2010-12-24 | 2019-04-03 | Qualcomm Technologies Int Ltd | Encapsulated instruction set |
GB2486740A (en) * | 2010-12-24 | 2012-06-27 | Cambridge Silicon Radio Ltd | Long instructions for a data processing system encoded as a series of shorter instructions |
GB2486740B (en) * | 2010-12-24 | 2019-02-13 | Qualcomm Technologies Int Ltd | Encapsulated instruction set |
GB2486737A (en) * | 2010-12-24 | 2012-06-27 | Cambridge Silicon Radio Ltd | Prefix instruction, which modifies an operand from a table lookup to an immediate value |
US20140019990A1 (en) * | 2011-03-30 | 2014-01-16 | Freescale Semiconductor, Inc. | Integrated circuit device and method for enabling cross-context access |
US20170024216A1 (en) * | 2014-03-12 | 2017-01-26 | Samsung Electronics Co., Ltd. | Method and device for processing vliw instruction, and method and device for generating instruction for processing vliw instruction |
US10599439B2 (en) * | 2014-03-12 | 2020-03-24 | Samsung Electronics Co., Ltd. | Method and device for allocating a VLIW instruction based on slot information stored in a database by a calculation allocation instruction |
US20160139925A1 (en) * | 2014-11-17 | 2016-05-19 | International Business Machines Corporation | Techniques for identifying instructions for decode-time instruction optimization grouping in view of cache boundaries |
US10061705B2 (en) | 2014-11-17 | 2018-08-28 | International Business Machines Corporation | Identifying instructions for decode-time instruction optimization grouping in view of cache boundaries |
US9740491B2 (en) | 2014-11-17 | 2017-08-22 | International Business Machines Corporation | Instruction group formation techniques for decode-time instruction optimization based on feedback |
US9733940B2 (en) | 2014-11-17 | 2017-08-15 | International Business Machines Corporation | Techniques for instruction group formation for decode-time instruction optimization based on feedback |
US9940242B2 (en) * | 2014-11-17 | 2018-04-10 | International Business Machines Corporation | Techniques for identifying instructions for decode-time instruction optimization grouping in view of cache boundaries |
US10402199B2 (en) * | 2015-10-22 | 2019-09-03 | Texas Instruments Incorporated | Conditional execution specification of instructions using conditional extension slots in the same execute packet in a VLIW processor |
CN108139911A (en) * | 2015-10-22 | 2018-06-08 | 德州仪器公司 | In the same execution packet of vliw processor specification is performed using having ready conditions for the instruction for expansion slot of having ready conditions |
US11397583B2 (en) | 2015-10-22 | 2022-07-26 | Texas Instruments Incorporated | Conditional execution specification of instructions using conditional extension slots in the same execute packet in a VLIW processor |
US11960892B2 (en) | 2015-10-22 | 2024-04-16 | Texas Instruments Incorporated | Conditional execution specification of instructions using conditional extension slots in the same execute packet in a VLIW processor |
US20170192788A1 (en) * | 2016-01-05 | 2017-07-06 | Intel Corporation | Binary translation support using processor instruction prefixes |
US20190236442A1 (en) * | 2016-08-05 | 2019-08-01 | Cambricon Technologies Corporation Limited | Operation unit, method and device capable of supporting operation data of different bit widths |
US10489704B2 (en) * | 2016-08-05 | 2019-11-26 | Cambricon Technologies Corporation Limited | Operation unit, method and device capable of supporting operation data of different bit widths |
Also Published As
Publication number | Publication date |
---|---|
SG95605A1 (en) | 2003-04-23 |
TW497073B (en) | 2002-08-01 |
DE69916962D1 (en) | 2004-06-09 |
EP0994413B1 (en) | 2004-05-06 |
EP0994413A3 (en) | 2002-01-23 |
DE69916962T2 (en) | 2005-04-07 |
ATE266226T1 (en) | 2004-05-15 |
EP0994413A2 (en) | 2000-04-19 |
KR100690225B1 (en) | 2007-03-12 |
JP2000122864A (en) | 2000-04-28 |
KR20000029005A (en) | 2000-05-25 |
CN1250906A (en) | 2000-04-19 |
CN1129843C (en) | 2003-12-03 |
US6418527B1 (en) | 2002-07-09 |
ES2221282T3 (en) | 2004-12-16 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
US6418527B1 (en) | Data processor instruction system for grouping instructions with or without a common prefix and data processing system that uses two or more instruction grouping methods | |
US5163139A (en) | Instruction preprocessor for conditionally combining short memory instructions into virtual long instructions | |
US7386844B2 (en) | Compiler apparatus and method of optimizing a source program by reducing a hamming distance between two instructions | |
EP2145249B1 (en) | Compact instruction set encoding | |
JP2000222206A (en) | Data processor | |
KR20030067892A (en) | Apparatus and method for dispatching very long instruction word with variable length | |
US20040205326A1 (en) | Early predicate evaluation to reduce power in very long instruction word processors employing predicate execution | |
EP1340141A1 (en) | A high performance vliw processor | |
EP1468367A4 (en) | Multithreaded processor with efficient processing for convergence device applications | |
EP1866804A2 (en) | Speculative load of look up table entries based upon coarse index calculation in parallel with index calculation | |
EP0992894A1 (en) | Apparatus and method for loop execution | |
JPH03218523A (en) | Data processor | |
US8671266B2 (en) | Staging register file for use with multi-stage execution units | |
CN108139911B (en) | Conditional execution specification of instructions using conditional expansion slots in the same execution packet of a VLIW processor | |
EP3295299A1 (en) | Decoding information about a group of instructions including a size of the group of instructions | |
US6857063B2 (en) | Data processor and method of operation | |
JP5112627B2 (en) | Processor architecture | |
CN113703845B (en) | RISC-V based reconfigurable embedded processor micro-architecture and working method thereof | |
EP3314404B1 (en) | Processing header to interpret information regarding a group of instructions | |
US6681319B1 (en) | Dual access instruction and compound memory access instruction with compatible address fields | |
JPH11282674A (en) | Processor | |
EP0992887B1 (en) | Memory access using byte qualifiers | |
US8898433B2 (en) | Efficient extraction of execution sets from fetch sets | |
WO2005036384A2 (en) | Instruction encoding for vliw processors | |
Glossner et al. | The Sandblaster SBX 2.0 Architecture |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
AS | Assignment |
Owner name: MOTOROLA, INC., ILLINOIS Free format text: ASSIGNMENT OF ASSIGNORS INTEREST;ASSIGNORS:ROZENSHEIN, ZVIKA;TOKAR, JACOB;DAYAN, URI;AND OTHERS;REEL/FRAME:010000/0204;SIGNING DATES FROM 19990419 TO 19990525 |
|
STCF | Information on status: patent grant |
Free format text: PATENTED CASE |
|
CC | Certificate of correction | ||
AS | Assignment |
Owner name: FREESCALE SEMICONDUCTOR, INC., TEXAS Free format text: ASSIGNMENT OF ASSIGNORS INTEREST;ASSIGNOR:MOTOROLA, INC.;REEL/FRAME:015698/0657 Effective date: 20040404 Owner name: FREESCALE SEMICONDUCTOR, INC.,TEXAS Free format text: ASSIGNMENT OF ASSIGNORS INTEREST;ASSIGNOR:MOTOROLA, INC.;REEL/FRAME:015698/0657 Effective date: 20040404 |
|
FPAY | Fee payment |
Year of fee payment: 4 |
|
AS | Assignment |
Owner name: CITIBANK, N.A. AS COLLATERAL AGENT, NEW YORK Free format text: SECURITY AGREEMENT;ASSIGNORS:FREESCALE SEMICONDUCTOR, INC.;FREESCALE ACQUISITION CORPORATION;FREESCALE ACQUISITION HOLDINGS CORP.;AND OTHERS;REEL/FRAME:018855/0129 Effective date: 20061201 Owner name: CITIBANK, N.A. AS COLLATERAL AGENT,NEW YORK Free format text: SECURITY AGREEMENT;ASSIGNORS:FREESCALE SEMICONDUCTOR, INC.;FREESCALE ACQUISITION CORPORATION;FREESCALE ACQUISITION HOLDINGS CORP.;AND OTHERS;REEL/FRAME:018855/0129 Effective date: 20061201 |
|
FPAY | Fee payment |
Year of fee payment: 8 |
|
AS | Assignment |
Owner name: CITIBANK, N.A., AS COLLATERAL AGENT,NEW YORK Free format text: SECURITY AGREEMENT;ASSIGNOR:FREESCALE SEMICONDUCTOR, INC.;REEL/FRAME:024397/0001 Effective date: 20100413 Owner name: CITIBANK, N.A., AS COLLATERAL AGENT, NEW YORK Free format text: SECURITY AGREEMENT;ASSIGNOR:FREESCALE SEMICONDUCTOR, INC.;REEL/FRAME:024397/0001 Effective date: 20100413 |
|
AS | Assignment |
Owner name: CITIBANK, N.A., AS NOTES COLLATERAL AGENT, NEW YOR Free format text: SECURITY AGREEMENT;ASSIGNOR:FREESCALE SEMICONDUCTOR, INC.;REEL/FRAME:030633/0424 Effective date: 20130521 |
|
AS | Assignment |
Owner name: CITIBANK, N.A., AS NOTES COLLATERAL AGENT, NEW YOR Free format text: SECURITY AGREEMENT;ASSIGNOR:FREESCALE SEMICONDUCTOR, INC.;REEL/FRAME:031591/0266 Effective date: 20131101 |
|
FPAY | Fee payment |
Year of fee payment: 12 |
|
AS | Assignment |
Owner name: FREESCALE SEMICONDUCTOR, INC., TEXAS Free format text: PATENT RELEASE;ASSIGNOR:CITIBANK, N.A., AS COLLATERAL AGENT;REEL/FRAME:037356/0553 Effective date: 20151207 Owner name: FREESCALE SEMICONDUCTOR, INC., TEXAS Free format text: PATENT RELEASE;ASSIGNOR:CITIBANK, N.A., AS COLLATERAL AGENT;REEL/FRAME:037356/0143 Effective date: 20151207 Owner name: FREESCALE SEMICONDUCTOR, INC., TEXAS Free format text: PATENT RELEASE;ASSIGNOR:CITIBANK, N.A., AS COLLATERAL AGENT;REEL/FRAME:037354/0225 Effective date: 20151207 |
|
AS | Assignment |
Owner name: MORGAN STANLEY SENIOR FUNDING, INC., MARYLAND Free format text: ASSIGNMENT AND ASSUMPTION OF SECURITY INTEREST IN PATENTS;ASSIGNOR:CITIBANK, N.A.;REEL/FRAME:037486/0517 Effective date: 20151207 |
|
AS | Assignment |
Owner name: MORGAN STANLEY SENIOR FUNDING, INC., MARYLAND Free format text: ASSIGNMENT AND ASSUMPTION OF SECURITY INTEREST IN PATENTS;ASSIGNOR:CITIBANK, N.A.;REEL/FRAME:037518/0292 Effective date: 20151207 |
|
AS | Assignment |
Owner name: NORTH STAR INNOVATIONS INC., CALIFORNIA Free format text: ASSIGNMENT OF ASSIGNORS INTEREST;ASSIGNOR:FREESCALE SEMICONDUCTOR, INC.;REEL/FRAME:037694/0264 Effective date: 20151002 |
|
AS | Assignment |
Owner name: NXP, B.V., F/K/A FREESCALE SEMICONDUCTOR, INC., NETHERLANDS Free format text: RELEASE BY SECURED PARTY;ASSIGNOR:MORGAN STANLEY SENIOR FUNDING, INC.;REEL/FRAME:040925/0001 Effective date: 20160912 Owner name: NXP, B.V., F/K/A FREESCALE SEMICONDUCTOR, INC., NE Free format text: RELEASE BY SECURED PARTY;ASSIGNOR:MORGAN STANLEY SENIOR FUNDING, INC.;REEL/FRAME:040925/0001 Effective date: 20160912 |
|
AS | Assignment |
Owner name: NXP B.V., NETHERLANDS Free format text: RELEASE BY SECURED PARTY;ASSIGNOR:MORGAN STANLEY SENIOR FUNDING, INC.;REEL/FRAME:040928/0001 Effective date: 20160622 |
|
AS | Assignment |
Owner name: MORGAN STANLEY SENIOR FUNDING, INC., MARYLAND Free format text: CORRECTIVE ASSIGNMENT TO CORRECT THE REMOVE PATENTS 8108266 AND 8062324 AND REPLACE THEM WITH 6108266 AND 8060324 PREVIOUSLY RECORDED ON REEL 037518 FRAME 0292. ASSIGNOR(S) HEREBY CONFIRMS THE ASSIGNMENT AND ASSUMPTION OF SECURITY INTEREST IN PATENTS;ASSIGNOR:CITIBANK, N.A.;REEL/FRAME:041703/0536 Effective date: 20151207 |
|
AS | Assignment |
Owner name: SHENZHEN XINGUODU TECHNOLOGY CO., LTD., CHINA Free format text: CORRECTIVE ASSIGNMENT TO CORRECT THE TO CORRECT THE APPLICATION NO. FROM 13,883,290 TO 13,833,290 PREVIOUSLY RECORDED ON REEL 041703 FRAME 0536. ASSIGNOR(S) HEREBY CONFIRMS THE THE ASSIGNMENT AND ASSUMPTION OF SECURITYINTEREST IN PATENTS.;ASSIGNOR:MORGAN STANLEY SENIOR FUNDING, INC.;REEL/FRAME:048734/0001 Effective date: 20190217 |
|
AS | Assignment |
Owner name: MORGAN STANLEY SENIOR FUNDING, INC., MARYLAND Free format text: CORRECTIVE ASSIGNMENT TO CORRECT THE REMOVE APPLICATION11759915 AND REPLACE IT WITH APPLICATION 11759935 PREVIOUSLY RECORDED ON REEL 037486 FRAME 0517. ASSIGNOR(S) HEREBY CONFIRMS THE ASSIGNMENT AND ASSUMPTION OF SECURITYINTEREST IN PATENTS;ASSIGNOR:CITIBANK, N.A.;REEL/FRAME:053547/0421 Effective date: 20151207 |
|
AS | Assignment |
Owner name: NXP B.V., NETHERLANDS Free format text: CORRECTIVE ASSIGNMENT TO CORRECT THE REMOVEAPPLICATION 11759915 AND REPLACE IT WITH APPLICATION11759935 PREVIOUSLY RECORDED ON REEL 040928 FRAME 0001. ASSIGNOR(S) HEREBY CONFIRMS THE RELEASE OF SECURITYINTEREST;ASSIGNOR:MORGAN STANLEY SENIOR FUNDING, INC.;REEL/FRAME:052915/0001 Effective date: 20160622 |
|
AS | Assignment |
Owner name: NXP, B.V. F/K/A FREESCALE SEMICONDUCTOR, INC., NETHERLANDS Free format text: CORRECTIVE ASSIGNMENT TO CORRECT THE REMOVEAPPLICATION 11759915 AND REPLACE IT WITH APPLICATION11759935 PREVIOUSLY RECORDED ON REEL 040925 FRAME 0001. ASSIGNOR(S) HEREBY CONFIRMS THE RELEASE OF SECURITYINTEREST;ASSIGNOR:MORGAN STANLEY SENIOR FUNDING, INC.;REEL/FRAME:052917/0001 Effective date: 20160912 |