SORT ACCELERATION PROCESSORS, METHODS, SYSTEMS, AND INSTRUCTIONS
A processor of an aspect includes packed data registers, and a decode unit to decode an instruction. The instruction may indicate a first source packed data to include at least four data elements, to indicate a second source packed data to include at least four data elements, and to indicate a destination storage location. An execution unit is coupled with the packed data registers and the decode unit. The execution unit, in response to the instruction, is to store a result packed data in the destination storage location. The result packed data may include at least four indexes that may identify corresponding data element positions in the first and second source packed data. The indexes may be stored in positions in the result packed data that are to represent a sorted order of corresponding data elements in the first and second source packed data.
Latest Intel Patents:
- APPARATUS, SYSTEM AND METHOD OF COMMUNICATING A PHYSICAL LAYER PROTOCOL DATA UNIT (PPDU) INCLUDING A TRAINING FIELD
- USES OF CODED DATA AT MULTI-ACCESS EDGE COMPUTING SERVER
- SELECTIVE PACKING OF PATCHES FOR IMMERSIVE VIDEO
- MULTI-LINK DEVICE RESETUP AND TRANSITION WITH STATION DEVICE ADDRESS AUTHENTICATION
- METHOD AND APPARATUS FOR SHARED VIRTUAL MEMORY TO MANAGE DATA COHERENCY IN A HETEROGENEOUS PROCESSING SYSTEM
Embodiments described herein generally relate to processors. In particular, embodiments described herein generally relate to sorting data in processors.
Background InformationData sorting operations are used extensively in computers, servers, datacenters, sort networks, and the like. For example, sort operations are commonly used in spreadsheets, databases, Structured Query Language (SQL) databases or servers, datacenters, high performance computing (HPC), Apache Hadoop software framework, image processing (e.g., for median filter, gesture recognition, etc.), and neural networks, to name just a few examples. Sort operations, especially when sorting large data sets, can be computationally intensive and may tend to affect overall performance.
The invention may best be understood by referring to the following description and accompanying drawings that are used to illustrate embodiments. In the drawings:
Disclosed herein are sort acceleration instructions, processors to execute the instructions, methods performed by the processors when processing or executing the instructions, and systems incorporating one or more processors to process or execute the instructions. In the following description, numerous specific details are set forth (e.g., specific instruction operations, packed data formats, processor configurations, microarchitectural details, sequences of operations, etc.). However, embodiments may be practiced without these specific details. In other instances, well-known circuits, structures and techniques have not been shown in detail to avoid obscuring the understanding of the description.
During operation, the processor 100 may receive the embodiment of the sort acceleration instruction 102. For example, the instruction may be received from an instruction fetch unit, an instruction queue, or the like. The sort acceleration instruction may represent a macroinstruction, assembly language instruction, machine code instruction, or instruction or control signal of an instruction set of the processor. In some embodiments, the sort acceleration instruction may explicitly specify (e.g., through one or more fields or a set of bits), or otherwise indicate (e.g., implicitly indicate, etc.), a source packed data 110, and may specify or otherwise indicate a destination (e.g., a destination storage location) where a result packed data 114 is to be stored. In some embodiments, the instruction may optionally specify or otherwise indicate a second source packed data 112, and in some embodiments may optionally specify or otherwise indicate a second destination (e.g., a destination storage location) where a second result packed data 116 is to be stored.
Referring again to
In some embodiments, instead of the sort acceleration instruction being provided directly to the decode unit, an instruction emulator, translator, morpher, interpreter, or other instruction conversion module may optionally be used. Various types of instruction conversion modules are known in the arts and may be implemented in software, hardware, firmware, or a combination thereof. In some embodiments, the instruction conversion module may be located outside the processor, such as, for example, on a separate die and/or in a memory (e.g., as a static, dynamic, or runtime emulation module). By way of example, the instruction conversion module may receive the sort acceleration instruction which may be of a first instruction set and may emulate, translate, morph, interpret, or otherwise convert the sort acceleration instruction into one or more corresponding or derived intermediate instructions or control signals which may be of a second different instruction set. The one or more intermediate instructions or control signals of the second instruction set may be provided to a decode unit, which may decode them into one or more lower-level instructions or control signals executable by native hardware of the processor (e.g., one or more execution units).
The processor 100 also includes a set of packed data registers 108. Each of the packed data registers may represent an on-die storage location that is operable to store packed data, vector data, or SIMD data. The packed data registers may represent architecturally-visible registers (e.g., an architectural register file). The architecturally-visible or architectural registers are visible to software and/or a programmer and/or are the registers indicated by instructions of an instruction set of the processor to identify operands. These architectural registers are contrasted to other non-architectural or non-architecturally visible registers in a given microarchitecture (e.g., temporary registers, reorder buffers, retirement registers, etc.). The packed data registers may be implemented in different ways in different microarchitectures using well-known techniques and are not limited to any particular type of circuit. Examples of suitable types of registers include, but are not limited to, dedicated physical registers, dynamically allocated physical registers using register renaming, and combinations thereof.
As shown, in some embodiments, the first source packed data 110, the optional second source packed data 112, the first result packed data 114, and the optional second result packed data 116 may each optionally be stored in the packed data registers. Alternatively, memory locations, or other storage locations, may be used for one or more of these operands. Moreover, although they are shown as being separate in the illustration, in some embodiments, a packed data register used for a source packed data may be reused as a destination storage location (e.g., a result packed data may be written over the source packed data).
Referring again to
The execution unit and/or the processor may include specific or particular logic (e.g., transistors, integrated circuitry, or other hardware potentially combined with firmware (e.g., instructions stored in non-volatile memory) and/or software) that is operable to perform the sort acceleration instruction and/or store the result in response to and/or as a result of the instruction. By way of example, the execution unit may include an arithmetic logic unit, a logic unit, a data rearrangement unit, or the like. In some embodiments, the execution unit may include one or more input structures (e.g., port(s), interconnect(s), an interface) to receive source operands, sort circuitry or logic 107 coupled with the input structure(s) to receive and process the source packed data and generate the result packed data, and one or more output structures (e.g., port(s), interconnect(s), an interface) coupled with the sort circuitry or logic 107 to output the result packed data. By way of example, the sort circuitry or logic may include a compare and swap chain, a value based selection or routing hierarchy, or other sort circuitry or logic.
The method includes receiving the single source sort indexes instruction, at block 226. The instruction may be received at a processor or a portion thereof (e.g., an instruction fetch unit, a decode unit, a bus interface unit). In various aspects, the instruction may be received from an off-die source (e.g., from memory, interconnect, etc.), or from an on-die source (e.g., from an instruction cache, instruction queue). The instruction may specify or otherwise indicate a source packed data including at least four data elements that are not in a sorted order (e.g., not sorted in increasing or decreasing magnitude across the packed data). The instruction may also specify or otherwise indicate a destination storage location.
A result packed data is stored in the destination storage location in response to and/or as a result of the instruction, at block 227. Representatively, an execution unit or processor may perform the instruction and store the result. The result packed data may include at least four indexes. In some embodiments, the indexes may identify corresponding data elements in the source packed data. In some embodiments, the indexes may be stored in positions in the result packed data that represent the sorted order of the corresponding data elements of the source packed data. That is, the indexes may be stored in a sorted order where they are sorted not based on the values of the indexes themselves, but rather based on the values of their corresponding data elements in the source packed data. In some embodiments, the result packed data may include indexes corresponding to all data elements in the source packed data.
In the particular illustrated example, the source packed data has eight data elements. The eight data elements have, from a least significant or lowest order bit position on the right to a most significant or highest order bit position on the left, the values −4, 1, 8, 12, 43, 55, 1, and 12. The values are not in a sorted order (e.g., are not sorted in increasing or decreasing order across the packed data).
A result packed data 314 may be generated (e.g., by an execution unit 106) and stored in a destination storage location in response to the single source sort indexes instruction. The destination storage location may be specified or otherwise indicated by the instruction. In various embodiments, the destination storage location may be a packed data register, a memory location, or other storage location. The destination storage location may either be the same storage location used for the source packed data, or a different storage location.
In some embodiments, the result packed data may include at least four indexes. In the particular illustrated example, the result packed data has eight indexes. Each index may point to, identify, or be associated with, a single corresponding data element in the source packed data. Each of the data elements may have a position within the source packed data that is indicated by the corresponding index. The index may represent the relative position or offset of the data element within the operand. For example, according to one possible convention, which is used in the illustration, index values of 0 through 7 may represent the eight data elements in the first through the eighth positions moving across the packed data from least to most significant bit positions (right to left as viewed). To further illustrate, the data element with value −4 has index of 0, the data element with value 8 has index of 2, the data element with value 43 has index of 4, and so on. Alternatively, various other indexing conventions may optionally be used (e.g., starting with 1 instead of 0, indexing backwards from 7 to 0, an arbitrary mapping convention, etc.). The indexes may have a number of bits sufficient to represent the indexed positions (e.g., 3-bits in the illustrated example).
In some embodiments, the indexes may be stored in positions in the result packed data that are to represent a sorted order of the corresponding data elements in the source packed data. In the illustrated example, the sorted order is an increasing order that increases from least to most significant bit positions, although a decreasing order may optionally be used instead. For the values in the illustrated example, the increasing sorted order would be −4, 1, 1, 8, 12, 12, 43, and 55. As a result, the index value of 0 corresponding to the data element with the value of −4 is stored in the first or lowest order position in the result packed data, the index value of 1 corresponding to the rightmost data element with the value of 1 is stored in the second or next lowest order position, and so on. In the illustration, arrows are used to show the correspondence between the indexes and the data elements.
Notice that the result packed data stores the sorted indexes not the sorted source data elements. In some embodiments, the sorted indexes may optionally be used subsequently by a different instruction to sort the source data elements. For example, in some embodiments, a permute instruction, shuffle instruction, or the like, may use the indexes to sort the source data elements. For example, the permute or shuffle instruction may indicate the result packed data 314 with the indexes as a first source operand and the source packed data 310 a second source operand, and may store a result packed data having the data elements sorted based on the control of the indexes. One particular example of a suitable instruction is the VPERMD instruction, which is described in Intel® Architecture Instruction Set Extensions Programming Reference, 319433-017, published December 2013, available from Intel Corporation, of Santa Clara, Calif.
However, in some implementations, it is advantageous to store the indexes instead of the sorted data elements. For example, the sorted indexes may be used to sort other data in addition to, or instead of, the source data elements of the source packed data 310. For example, this may be case in various different data structures in which the source data elements are used as sort keys and are each associated with multiple data elements. To further illustrate, consider a simple example of a table (e.g., in a spreadsheet) arranged in columns and rows. One column lists inventor names and another column lists inventor identification numbers. Both the names and identification numbers may need to be sorted together and consistently with one another. As one example, a sort may be based on name as the sort key and sorted indexes may be stored as a result. The sorted indexes may then be used to sort both the names and the identification numbers. Accordingly, in some implementations, storing the sorted indexes may offer advantages, such as flexibility and an ability to sort other associated data, as compared to storing the sorted data elements.
In some embodiments, the sorting may be done in stable order (e.g., ascending order). By stable it is meant that if two elements have equal search keys, then the relative order between them does not change. For example, if you have two rows with the same inventor name, but different identification numbers, and you sort the table according to the name, then the order of the identification numbers may not change. This is contrasted to an unstable order, in which the relative order is not guaranteed, and elements with equal keys may be shuffled.
The single source sort indexes and data elements instruction may specify or otherwise indicate a source packed data 410 having at least four data elements. The source packed data and data elements may be as described elsewhere herein (e.g., as in
A first result packed data 414 may be generated (e.g., by an execution unit 106) and stored in a first destination storage location in response to the instruction. The first result packed data 414 may have at least four indexes. The first result packed data may be as described elsewhere herein (e.g., as in
In this embodiment, a second result packed data 416 may also be generated and stored in a second destination storage location in response to the instruction. The second destination storage location may be specified or otherwise indicated by the instruction, and may be a packed data register, a memory location, or other storage location. The second result packed data may include the corresponding data elements, which correspond to the indexes stored in the first result packed data 414, stored in positions of the second result packed data 416 that reflect the sorted order. For example, in increasing sorted order the second result packed data may store the values, from right to left in the illustration, −4, 1, 1, 8, 12, 12, 43, and 55. In another embodiment, a decreasing order may be used instead for the indexes and data elements.
Sorted indexes are stored as a result in
The method includes receiving the two source sort indexes instruction, at block 536. In various aspects, the instruction may be received at a processor or a portion thereof (e.g., an instruction fetch unit, a decode unit, a bus interface unit). In various aspects, the instruction may be received from an off-die source (e.g., from memory, interconnect, etc.), or from an on-die source (e.g., from an instruction cache, instruction queue). The two source sort indexes instruction may specify or otherwise indicate a first source packed data including a first set of at least four data elements, and may specify or otherwise indicate a second source packed data including a second set of at least four data elements. The instruction may also specify or otherwise indicate a destination storage location.
Referring again to
In the particular illustrated example, the first source packed data and the second source packed data each have eight data elements. The first source packed data has, from a least significant position on the right to a most significant position on the left, the values of −4, 1, 1, 8, 12, 12, 43, 55. The second source packed data has, from a least significant position on the right to a most significant position on the left, the values of −14, −12, 0, 10, 16, 18, 24, and 60. In this embodiment, the values in the first source packed data, and the values in the second source packed data, are each sorted in increasing order from least to most significant bit positions. In some embodiments, the data elements within each of the first and second source packed data may be assumed (e.g., implicitly assumed for the instruction) or required to be in sorted order within the respective packed data operand (e.g., for the instruction to operate correctly). In other embodiments, the data elements within each of the first and second source packed data may not be assumed or required to be in sorted order (e.g., the instruction may operate correctly when the elements are in an unsorted order).
A result packed data 614 may be generated (e.g., by an execution unit 106) and stored in a destination storage location in response to the instruction. In various embodiments, the destination storage location may be a packed data register, a memory location, or other storage location. The destination storage location may either be the same storage location used for one of the first and second source packed data, or a different storage location.
In some embodiments, the result packed data may include at least four indexes. In some embodiments, the result packed data may include a same number of indexes as a number of data elements in one of the source packed data, although this is not required. In the particular illustrated example, the result packed data has eight indexes. Each index may point to, identify, or be associated with, corresponding data element positions in the first and second source packed data. In some embodiments, the indexes may actually identify single corresponding data elements in one of the first and second source packed data. Each of the data elements may have an indexed position within the first and second source packed data. The index may represent the relative position or offset of the data element within the first and second packed data. For example, according to one possible convention, which is used in the illustrated example, index values of 0 through 7 may index the eight data elements when moving across the first source packed data 610 from least to most significant bit positions (right to left as viewed), and index values of 8 through 15 may index the eight data elements when moving across the second source packed data 612 from least to most significant bit positions (right to left as viewed). In the illustrated example, arrows are used to show the correspondence between indexes and data elements. As shown, the data element with the value −14 has an index of 8, the data element with the value of −12 has an index of 9, and so on. Alternatively, various other indexing conventions desired for the particular implementation may optionally be used (e.g., starting with an index of 1 instead of 0, indexing backwards, an arbitrarily mapped indexing convention, etc.).
The indexes may be stored in positions in the result packed data that are to represent a sorted order of the corresponding data elements in the first and second source packed data. In some embodiments, the indexes may be stored for only an ordered subset of all of the data elements in the first and second source packed data. In some embodiments, the ordered subset may be an ordered smallest subset (e.g., a smallest half) of all of the data elements. In the illustrated example, the eight smallest values are −14, −12, −4, 0, 1, 1, 8, and 10. In the illustrated embodiment, the eight indexes are stored in positions in the result packed data that represent a sorted increasing order (that increases from least to most significant bit positions) of the corresponding eight smallest data elements in the first and second source packed data. As shown, the index of 8 corresponding to the data element with value −14 is stored in the lowest order position in the result packed data, the index of 9 corresponding to the data element with value −12 is stored in the next lowest order position, the index of 0 corresponding to the data element with value −4 is stored in the third lowest order position, and so on. In other embodiments, a decreasing or reverse order may optionally be used. The sort operation merges sorted indexes for both the first and second source packed data.
A result packed data 714 may be generated (e.g., by an execution unit 106) and stored in a destination storage location in response to the instruction. In some embodiments, the result packed data may include at least four indexes. In some embodiments, the indexes may identify single corresponding data elements in one of the first and second source packed data. In some embodiments, the indexes may be stored in positions in the result packed data that are to represent a sorted order of the corresponding data elements in the first and second source packed data.
In the embodiment of
A result packed data 814 may be generated (e.g., by an execution unit 106) and stored in a destination storage location in response to the instruction. In some embodiments, the result packed data may include at least four indexes. In some embodiments, the indexes may identify single corresponding data elements in one of the first and second source packed data. In some embodiments, the indexes may be stored in positions in the result packed data that are to represent a sorted order of the corresponding data elements in the first and second source packed data. In the illustrated embodiment, the result packed data has indexes corresponding to a smallest subset, in this case a smallest half, of all data elements in the first and second source packed data. In another embodiment, the result packed data may have indexes corresponding to a largest subset, for example a largest half, of all data elements in the first and second source packed data.
A first result packed data 914 may be generated (e.g., by an execution unit 106) and stored in a first destination storage location in response to the instruction. The first result packed data 914 may have indexes. The first result packed data and indexes may have similar or common characteristics and details as previously described for
In this embodiment, a second result packed data 916 may also be generated and stored in a second destination storage location in response to the instruction. The second destination storage location may be specified or otherwise indicated by the instruction, and may be a packed data register, a memory location, or other storage location. The second destination storage location may either be the same storage location used for one of the first or second source packed data, or may be a different storage location. The second result packed data may include the corresponding data elements, which correspond to the indexes stored in the first result packed data, stored in positions of the second result packed data that reflect the sorted order. In the illustrated embodiment, the second result packed data has a sorted smallest subset, in this case a sorted smallest half, of all data elements in the first and second source packed data. Specifically, the second result packed data stores, from right to left, the data elements −14, −12, −4, 0, 1, 1, 8, and 10. In another embodiment, the second result packed data may instead have a sorted largest subset, for example a sorted largest half, of all data elements in the first and second source packed data.
Sorted indexes are stored as a result in
In the illustrated embodiment, the values in each of the first source packed data and the second source packed data are sorted in order, although this is not required. In some embodiments, it may be assumed (e.g., implicitly assumed for the instruction) or required that the source data elements are in sorted order within the respective packed data. In the illustrated example, increasing order with increasing bit significance in the operands is used, although in another embodiment a decreasing order may optionally be used. In other embodiments, the data elements within each of the first source packed data and the second source packed data may not be assumed or required to be in sorted order within the respective packed data (e.g., the instruction may operate correctly when the elements are in an unsorted order).
A result packed data 1014 may be generated (e.g., by an execution unit 106) and stored in a destination storage location in response to the instruction. In some embodiments, the result packed data may include at least four indexes. In some embodiments, each index may point to, identify, or be associated with, a pair of corresponding data element positions in the first and second source packed data. The indexes themselves may not actually identify a single corresponding data element, but rather may identify a pair of corresponding data element positions in the first and second packed data, without indicating whether the actual corresponding data element is located in the first or second source packed data. For example, according to one possible convention, which is used in the illustrated example, index values of 0 through 7 may represent the eight data element positions when moving from least to most significant bit positions (right to left as viewed) across the first source packed data, and the same index values of 0 through 7 may represent the eight data element positions when moving from least to most significant bit positions across the second source packed data. For example, the data element with the value 43 and the data element with the value 24 may both have a same index of 6. As previously mentioned, various other conventions for the indexes may optionally be used instead (e.g., starting with an index of one instead of zero, a backwards or decrementing convention, an arbitrarily mapped convention, etc.) In such embodiments, the index alone may be insufficient to identify a single corresponding data element.
In some embodiments, the indexes may be stored in positions in the result packed data that are to represent a sorted order of the corresponding data elements. In the illustrated embodiment, the indexes are stored in positions in the result packed data that are to represent a sorted increasing order of the corresponding data elements. In other embodiments, a decreasing order may optionally be used. Moreover, in the illustrated embodiment, the indexes are stored for only an ordered smallest subset, for example a smallest half, of all of the data elements. Alternatively, another subset, such as a largest subset (e.g., a largest half), or an intermediate subset, may be used. Alternatively, indexes for all data elements in the source packed data may optionally be stored.
Referring again to
In some embodiments, the result mask may include at least four mask elements. In some embodiments, the result mask may include a same number of mask elements as a number of indexes in the result packed data. Each mask element may correspond to a different corresponding one of the indexes in the result packed data. Each mask element may indicate whether a single corresponding data element at the data element position indicated by the corresponding index is located in the first source packed data or instead is located in the second source packed data. That is, each mask element may identify or select one of the first and second source packed data, and thereby identify the corresponding single data element at the indexed position in the identified or selected source packed data.
In some embodiments, each mask element may be a single bit, although two or more bits may alternatively optionally be used (e.g., a most or least significant bit of a multi-bit data element). According to one possible convention, which is used in the illustrated example embodiment, a single bit being set to a binary value of one (i.e., 1) indicates that the data element is located in the first source packed data 1010, whereas the bit being cleared to a binary value of zero (i.e., 0) indicates that the data element is located in the second source packed data 1012. Alternatively, the opposite convention may optionally be used. To further illustrate, the data element with value −14 is identified by the corresponding index of 0 in the rightmost position of the result packed data and the mask element value of 0 in the rightmost position in the result mask (to indicate the second source packed data). Similarly, the data element with the value −4 is identified by the corresponding index of 0 in the third position from the right of the result packed data and the mask element value of 1 in the third position from the right of the result mask (to indicate the first source packed data).
In some embodiments, a convention may be adopted that if two inter-operand data elements in different source packed data have equal values, that elements from one of the source packed data may be interpreted as if it had a smaller value, although this is not required. In some embodiments, a convention may be adopted that if two intra-operand data elements in the same source packed data have equal values, the least significant data element may be interpreted as if it had a smaller value, although this is not required.
Including the extra packed data operand selection bits (e.g., the mask elements) in the result mask, instead of incorporating them into the indexes (as in the approach shown in
In some embodiments, in addition to the instruction that generates the result mask, the instruction set may also include a second instruction that is able to indicate or access the result mask as a source predicate operand or conditional control operand that is used to predicate, conditionally control, or mask whether or not corresponding operations are to be performed and/or corresponding results are to be stored. One specific example of an instruction that may indicate the result mask as a source predicate operand in certain implementations is a VMOVDQA32 instruction, which is described in Intel® Architecture Instruction Set Extensions Programming Reference, 319433-017, published December 2013. The VMOVDQA32 instruction is able to move aligned packed doubleword integer values from a source packed data operand to a result packed data operand using a source writemask for predication. The result mask may be indicated by the VMOVDQA32 instruction as the source writemask. In some implementations and/or for some algorithms the result mask may be used by such predicated instructions to offer certain performance and/or efficiency advantages from an overall algorithmic perspective. In some embodiments, the result mask may be used even if the result packed data 1014 has enough bits that it could instead have integrated the extra bit into the indexes.
A result packed data 1114 may be generated (e.g., by an execution unit 106) and stored in a destination storage location in response to the instruction. In some embodiments, the result packed data may include at least four indexes. In some embodiments, each index may point to, identify, or be associated with, a pair of corresponding data element positions in the first and second source packed data. Likewise, a result mask 1120 may be stored in response to the instruction. The result mask may have mask elements corresponding to the indexes.
In the embodiment of
In the illustrated embodiment, the eight indexes are stored in positions in the result packed data that are to represent a sorted increasing order of the corresponding eight largest data elements in the first and second source packed data. In other embodiments, a decreasing or reverse order may optionally be used.
A result packed data 1214 may be generated (e.g., by an execution unit 106) and stored in a destination storage location in response to the instruction. In some embodiments, the result packed data may include at least four indexes. In some embodiments, each index may point to, identify, or be associated with, a pair of corresponding data element positions in the first and second source packed data. Likewise, a result mask 1220 may be stored in response to the instruction. The result mask may have mask elements corresponding to the indexes.
In the illustrated embodiment, the indexes and mask elements correspond to a smallest subset, in this case a smallest half, of all data elements in the first and second source packed data. In another embodiment, the indexes and mask elements may corresponding to a largest subset, for example a largest half, of all data elements in the first and second source packed data.
In the illustrated embodiment, the eight indexes are stored in positions in the result packed data that are to represent a sorted increasing order of the corresponding eight largest data elements in the first and second source packed data. In other embodiments, a decreasing or reverse order may optionally be used.
A first result packed data 1314 may be generated (e.g., by an execution unit 106) and stored in a destination storage location in response to the instruction. In some embodiments, the result packed data may include at least four indexes. In some embodiments, each index may point to, identify, or be associated with, a pair of corresponding data element positions in the first and second source packed data. Likewise, a result mask 1320 may be stored in response to the instruction. The result mask may have mask elements corresponding to the indexes.
In this embodiment, a second result packed data 1316 may also be generated and stored in a second destination storage location in response to the instruction. The second destination storage location may be specified or otherwise indicated by the instruction, and may be a packed data register, a memory location, or other storage location. The second destination storage location may either be the same storage location used for one of the first or second source packed data, or may be a different storage location. The second result packed data may include the corresponding data elements, which correspond to the indexes stored and the mask elements, stored in positions of the second result packed data that reflect the sorted order.
In the illustrated embodiment, the sorted data elements, indexes, and mask elements correspond to a smallest subset, in this case a smallest half, of all data elements in the first and second source packed data. In another embodiment, the sorted data elements, indexes, and mask elements may correspond to a largest subset, for example a largest half, of all data elements in the first and second source packed data.
In the illustrated embodiment, the sorted data elements, indexes, and mask elements are stored in positions to represent an increasing order. In other embodiments, a decreasing or reverse order may optionally be used.
Sorted indexes are stored as a result in
The operations of
An instruction set may include one or more of the instructions disclosed herein. For example, in some embodiments, an instruction set may optionally include a first instruction (e.g., as shown or described for one of
The following code snippet represents an example embodiment of an algorithm using examples of instructions as disclosed herein to sort 32 integers. The sortassistd instruction sorts doubleword elements in assenting order with an operation similar to that shown in
sort32:
-
- vmovdqu32 (%rdi), %zmm0
- vmovdqu32 64(%rdi), %zmm1
- sortassistd %zmm0, %zmm2
- sortassistd %zmm1, %zmm3
- vpermd %zmm0, %zmm2, %zmm0
- vpermd %zmm1, %zmm3, %zmm1
- sortedmergedassist1 %zmm1, %zmm0, %zmm2, %k1
- sortedmergedassisth %zmm1, %zmm0, %zmm3, %k2
- vpermd %zmm0, %zmm2, %zmm4
- vpermd %zmm1, %zmm2, %zmm5
- vpermd %zmm0, %zmm3, %zmm6
- vpermd %zmm1, %zmm3, %zmm7
- vmovdqu32 %zmm5, %zmm4, {%k1}
- vmovdqu32 %zmm7, %zmm6, {%k2}
- vmovdqu32 %zmm4, {%rdi}
- vmovdqu32 %zmm6, 64{%rdi}
ret
In some embodiments, an instruction format may include an operation code or opcode. The opcode may represent a plurality of bits or one or more fields that are operable to identify the instruction and/or the operation to be performed (e.g., a sort index operation). Depending upon the particular instruction, the instruction format may also optionally include one or more source and/or destination specifiers. By way of example, each of these specifiers may include bits or one or more fields to specify an address of a register, memory location, or other storage location. Alternatively, instead of such an explicit specifier, one or more sources and/or destinations may optionally be implicit to the instruction instead of being explicitly specified. In addition, it may optionally be implicit that a source is to be reused as a destination. In addition, the instruction format may optionally add additional fields, may overlap certain fields, etc. Fields need not include contiguous sequences of bits but rather may be composed of non-contiguous or separated bits. In some embodiments, the instruction format may follow a VEX or EVEX encoding or instruction format, although the scope of the invention is not so limited.
An instruction set includes one or more instruction formats. A given instruction format defines various fields (number of bits, location of bits) to specify, among other things, the operation to be performed (opcode) and the operand(s) on which that operation is to be performed. Some instruction formats are further broken down though the definition of instruction templates (or subformats). For example, the instruction templates of a given instruction format may be defined to have different subsets of the instruction format's fields (the included fields are typically in the same order, but at least some have different bit positions because there are less fields included) and/or defined to have a given field interpreted differently. Thus, each instruction of an ISA is expressed using a given instruction format (and, if defined, in a given one of the instruction templates of that instruction format) and includes fields for specifying the operation and the operands. For example, an exemplary ADD instruction has a specific opcode and an instruction format that includes an opcode field to specify that opcode and operand fields to select operands (source1/destination and source2); and an occurrence of this ADD instruction in an instruction stream will have specific contents in the operand fields that select specific operands. A set of SIMD extensions referred to the Advanced Vector Extensions (AVX) (AVX1 and AVX2) and using the Vector Extensions (VEX) coding scheme, has been, has been released and/or published (e.g., see Intel® 64 and IA-32 Architectures Software Developers Manual, October 2011; and see Intel® Advanced Vector Extensions Programming Reference, June 2011).
Exemplary Instruction FormatsEmbodiments of the instruction(s) described herein may be embodied in different formats. Additionally, exemplary systems, architectures, and pipelines are detailed below. Embodiments of the instruction(s) may be executed on such systems, architectures, and pipelines, but are not limited to those detailed.
Generic Vector Friendly Instruction FormatA vector friendly instruction format is an instruction format that is suited for vector instructions (e.g., there are certain fields specific to vector operations). While embodiments are described in which both vector and scalar operations are supported through the vector friendly instruction format, alternative embodiments use only vector operations the vector friendly instruction format.
While embodiments of the invention will be described in which the vector friendly instruction format supports the following: a 64 byte vector operand length (or size) with 32 bit (4 byte) or 64 bit (8 byte) data element widths (or sizes) (and thus, a 64 byte vector consists of either 16 doubleword-size elements or alternatively, 8 quadword-size elements); a 64 byte vector operand length (or size) with 16 bit (2 byte) or 8 bit (1 byte) data element widths (or sizes); a 32 byte vector operand length (or size) with 32 bit (4 byte), 64 bit (8 byte), 16 bit (2 byte), or 8 bit (1 byte) data element widths (or sizes); and a 16 byte vector operand length (or size) with 32 bit (4 byte), 64 bit (8 byte), 16 bit (2 byte), or 8 bit (1 byte) data element widths (or sizes); alternative embodiments may support more, less and/or different vector operand sizes (e.g., 256 byte vector operands) with more, less, or different data element widths (e.g., 128 bit (16 byte) data element widths).
The class A instruction templates in
The generic vector friendly instruction format 1500 includes the following fields listed below in the order illustrated in
Format field 1540—a specific value (an instruction format identifier value) in this field uniquely identifies the vector friendly instruction format, and thus occurrences of instructions in the vector friendly instruction format in instruction streams. As such, this field is optional in the sense that it is not needed for an instruction set that has only the generic vector friendly instruction format.
Base operation field 1542—its content distinguishes different base operations.
Register index field 1544—its content, directly or through address generation, specifies the locations of the source and destination operands, be they in registers or in memory. These include a sufficient number of bits to select N registers from a P×Q (e.g. 32×512, 16×128, 32×1024, 64×1024) register file. While in one embodiment N may be up to three sources and one destination register, alternative embodiments may support more or less sources and destination registers (e.g., may support up to two sources where one of these sources also acts as the destination, may support up to three sources where one of these sources also acts as the destination, may support up to two sources and one destination).
Modifier field 1546—its content distinguishes occurrences of instructions in the generic vector instruction format that specify memory access from those that do not; that is, between no memory access 1505 instruction templates and memory access 1520 instruction templates. Memory access operations read and/or write to the memory hierarchy (in some cases specifying the source and/or destination addresses using values in registers), while non-memory access operations do not (e.g., the source and destinations are registers). While in one embodiment this field also selects between three different ways to perform memory address calculations, alternative embodiments may support more, less, or different ways to perform memory address calculations.
Augmentation operation field 1550—its content distinguishes which one of a variety of different operations to be performed in addition to the base operation. This field is context specific. In one embodiment of the invention, this field is divided into a class field 1568, an alpha field 1552, and a beta field 1554. The augmentation operation field 1550 allows common groups of operations to be performed in a single instruction rather than 2, 3, or 4 instructions.
Scale field 1560—its content allows for the scaling of the index field's content for memory address generation (e.g., for address generation that uses 2scale*index+base).
Displacement Field 1562A—its content is used as part of memory address generation (e.g., for address generation that uses 2scale*index+base+displacement).
Displacement Factor Field 1562B (note that the juxtaposition of displacement field 1562A directly over displacement factor field 1562B indicates one or the other is used)—its content is used as part of address generation; it specifies a displacement factor that is to be scaled by the size of a memory access (N)—where N is the number of bytes in the memory access (e.g., for address generation that uses 2scale*index+base+scaled displacement). Redundant low-order bits are ignored and hence, the displacement factor field's content is multiplied by the memory operands total size (N) in order to generate the final displacement to be used in calculating an effective address. The value of N is determined by the processor hardware at runtime based on the full opcode field 1574 (described later herein) and the data manipulation field 1554C. The displacement field 1562A and the displacement factor field 1562B are optional in the sense that they are not used for the no memory access 1505 instruction templates and/or different embodiments may implement only one or none of the two.
Data element width field 1564—its content distinguishes which one of a number of data element widths is to be used (in some embodiments for all instructions; in other embodiments for only some of the instructions). This field is optional in the sense that it is not needed if only one data element width is supported and/or data element widths are supported using some aspect of the opcodes.
Write mask field 1570—its content controls, on a per data element position basis, whether that data element position in the destination vector operand reflects the result of the base operation and augmentation operation. Class A instruction templates support merging-writemasking, while class B instruction templates support both merging- and zeroing-writemasking. When merging, vector masks allow any set of elements in the destination to be protected from updates during the execution of any operation (specified by the base operation and the augmentation operation); in other one embodiment, preserving the old value of each element of the destination where the corresponding mask bit has a 0. In contrast, when zeroing vector masks allow any set of elements in the destination to be zeroed during the execution of any operation (specified by the base operation and the augmentation operation); in one embodiment, an element of the destination is set to 0 when the corresponding mask bit has a 0 value. A subset of this functionality is the ability to control the vector length of the operation being performed (that is, the span of elements being modified, from the first to the last one); however, it is not necessary that the elements that are modified be consecutive. Thus, the write mask field 1570 allows for partial vector operations, including loads, stores, arithmetic, logical, etc. While embodiments of the invention are described in which the write mask field's 1570 content selects one of a number of write mask registers that contains the write mask to be used (and thus the write mask field's 1570 content indirectly identifies that masking to be performed), alternative embodiments instead or additional allow the mask write field's 1570 content to directly specify the masking to be performed.
Immediate field 1572—its content allows for the specification of an immediate. This field is optional in the sense that is it not present in an implementation of the generic vector friendly format that does not support immediate and it is not present in instructions that do not use an immediate.
Class field 1568—its content distinguishes between different classes of instructions. With reference to
In the case of the non-memory access 1505 instruction templates of class A, the alpha field 1552 is interpreted as an RS field 1552A, whose content distinguishes which one of the different augmentation operation types are to be performed (e.g., round 1552A.1 and data transform 1552A.2 are respectively specified for the no memory access, round type operation 1510 and the no memory access, data transform type operation 1515 instruction templates), while the beta field 1554 distinguishes which of the operations of the specified type is to be performed. In the no memory access 1505 instruction templates, the scale field 1560, the displacement field 1562A, and the displacement scale filed 1562B are not present.
No-Memory Access Instruction Templates—Full Round Control Type Operation
In the no memory access full round control type operation 1510 instruction template, the beta field 1554 is interpreted as a round control field 1554A, whose content(s) provide static rounding. While in the described embodiments of the invention the round control field 1554A includes a suppress all floating point exceptions (SAE) field 1556 and a round operation control field 1558, alternative embodiments may support may encode both these concepts into the same field or only have one or the other of these concepts/fields (e.g., may have only the round operation control field 1558).
SAE field 1556—its content distinguishes whether or not to disable the exception event reporting; when the SAE field's 1556 content indicates suppression is enabled, a given instruction does not report any kind of floating-point exception flag and does not raise any floating point exception handler.
Round operation control field 1558—its content distinguishes which one of a group of rounding operations to perform (e.g., Round-up, Round-down, Round-towards-zero and Round-to-nearest). Thus, the round operation control field 1558 allows for the changing of the rounding mode on a per instruction basis. In one embodiment of the invention where a processor includes a control register for specifying rounding modes, the round operation control field's 1550 content overrides that register value.
No Memory Access Instruction Templates—Data Transform Type Operation
In the no memory access data transform type operation 1515 instruction template, the beta field 1554 is interpreted as a data transform field 1554B, whose content distinguishes which one of a number of data transforms is to be performed (e.g., no data transform, swizzle, broadcast).
In the case of a memory access 1520 instruction template of class A, the alpha field 1552 is interpreted as an eviction hint field 1552B, whose content distinguishes which one of the eviction hints is to be used (in
Vector memory instructions perform vector loads from and vector stores to memory, with conversion support. As with regular vector instructions, vector memory instructions transfer data from/to memory in a data element-wise fashion, with the elements that are actually transferred is dictated by the contents of the vector mask that is selected as the write mask.
Memory Access Instruction Templates—Temporal
Temporal data is data likely to be reused soon enough to benefit from caching. This is, however, a hint, and different processors may implement it in different ways, including ignoring the hint entirely.
Memory Access Instruction Templates—Non-Temporal
Non-temporal data is data unlikely to be reused soon enough to benefit from caching in the 1st-level cache and should be given priority for eviction. This is, however, a hint, and different processors may implement it in different ways, including ignoring the hint entirely.
Instruction Templates of Class BIn the case of the instruction templates of class B, the alpha field 1552 is interpreted as a write mask control (Z) field 1552C, whose content distinguishes whether the write masking controlled by the write mask field 1570 should be a merging or a zeroing.
In the case of the non-memory access 1505 instruction templates of class B, part of the beta field 1554 is interpreted as an RL field 1557A, whose content distinguishes which one of the different augmentation operation types are to be performed (e.g., round 1557A.1 and vector length (VSIZE) 1557A.2 are respectively specified for the no memory access, write mask control, partial round control type operation 1512 instruction template and the no memory access, write mask control, VSIZE type operation 1517 instruction template), while the rest of the beta field 1554 distinguishes which of the operations of the specified type is to be performed. In the no memory access 1505 instruction templates, the scale field 1560, the displacement field 1562A, and the displacement scale filed 1562B are not present.
In the no memory access, write mask control, partial round control type operation 1510 instruction template, the rest of the beta field 1554 is interpreted as a round operation field 1559A and exception event reporting is disabled (a given instruction does not report any kind of floating-point exception flag and does not raise any floating point exception handler).
Round operation control field 1559A—just as round operation control field 1558, its content distinguishes which one of a group of rounding operations to perform (e.g., Round-up, Round-down, Round-towards-zero and Round-to-nearest). Thus, the round operation control field 1559A allows for the changing of the rounding mode on a per instruction basis. In one embodiment of the invention where a processor includes a control register for specifying rounding modes, the round operation control field's 1550 content overrides that register value.
In the no memory access, write mask control, VSIZE type operation 1517 instruction template, the rest of the beta field 1554 is interpreted as a vector length field 1559B, whose content distinguishes which one of a number of data vector lengths is to be performed on (e.g., 128, 256, or 512 byte).
In the case of a memory access 1520 instruction template of class B, part of the beta field 1554 is interpreted as a broadcast field 1557B, whose content distinguishes whether or not the broadcast type data manipulation operation is to be performed, while the rest of the beta field 1554 is interpreted the vector length field 1559B. The memory access 1520 instruction templates include the scale field 1560, and optionally the displacement field 1562A or the displacement scale field 1562B.
With regard to the generic vector friendly instruction format 1500, a full opcode field 1574 is shown including the format field 1540, the base operation field 1542, and the data element width field 1564. While one embodiment is shown where the full opcode field 1574 includes all of these fields, the full opcode field 1574 includes less than all of these fields in embodiments that do not support all of them. The full opcode field 1574 provides the operation code (opcode).
The augmentation operation field 1550, the data element width field 1564, and the write mask field 1570 allow these features to be specified on a per instruction basis in the generic vector friendly instruction format.
The combination of write mask field and data element width field create typed instructions in that they allow the mask to be applied based on different data element widths.
The various instruction templates found within class A and class B are beneficial in different situations. In some embodiments of the invention, different processors or different cores within a processor may support only class A, only class B, or both classes. For instance, a high performance general purpose out-of-order core intended for general-purpose computing may support only class B, a core intended primarily for graphics and/or scientific (throughput) computing may support only class A, and a core intended for both may support both (of course, a core that has some mix of templates and instructions from both classes but not all templates and instructions from both classes is within the purview of the invention). Also, a single processor may include multiple cores, all of which support the same class or in which different cores support different class. For instance, in a processor with separate graphics and general purpose cores, one of the graphics cores intended primarily for graphics and/or scientific computing may support only class A, while one or more of the general purpose cores may be high performance general purpose cores with out of order execution and register renaming intended for general-purpose computing that support only class B. Another processor that does not have a separate graphics core, may include one more general purpose in-order or out-of-order cores that support both class A and class B. Of course, features from one class may also be implement in the other class in different embodiments of the invention. Programs written in a high level language would be put (e.g., just in time compiled or statically compiled) into an variety of different executable forms, including: 1) a form having only instructions of the class(es) supported by the target processor for execution; or 2) a form having alternative routines written using different combinations of the instructions of all classes and having control flow code that selects the routines to execute based on the instructions supported by the processor which is currently executing the code.
Exemplary Specific Vector Friendly Instruction FormatIt should be understood that, although embodiments of the invention are described with reference to the specific vector friendly instruction format 1600 in the context of the generic vector friendly instruction format 1500 for illustrative purposes, the invention is not limited to the specific vector friendly instruction format 1600 except where claimed. For example, the generic vector friendly instruction format 1500 contemplates a variety of possible sizes for the various fields, while the specific vector friendly instruction format 1600 is shown as having fields of specific sizes. By way of specific example, while the data element width field 1564 is illustrated as a one bit field in the specific vector friendly instruction format 1600, the invention is not so limited (that is, the generic vector friendly instruction format 1500 contemplates other sizes of the data element width field 1564).
The generic vector friendly instruction format 1500 includes the following fields listed below in the order illustrated in
EVEX Prefix (Bytes 0-3) 1602—is encoded in a four-byte form.
Format Field 1540 (EVEX Byte 0, bits [7:0])—the first byte (EVEX Byte 0) is the format field 1540 and it contains 0x62 (the unique value used for distinguishing the vector friendly instruction format in one embodiment of the invention).
The second-fourth bytes (EVEX Bytes 1-3) include a number of bit fields providing specific capability.
REX field 1605 (EVEX Byte 1, bits [7-5])—consists of a EVEX.R bit field (EVEX Byte 1, bit [7]—R), EVEX.X bit field (EVEX byte 1, bit [6]—X), and 1557BEX byte 1, bit[5]—B). The EVEX.R, EVEX.X, and EVEX.B bit fields provide the same functionality as the corresponding VEX bit fields, and are encoded using is complement form, i.e. ZMM0 is encoded as 1111B, ZMM15 is encoded as 0000B. Other fields of the instructions encode the lower three bits of the register indexes as is known in the art (rrr, xxx, and bbb), so that Rrrr, Xxxx, and Bbbb may be formed by adding EVEX.R, EVEX.X, and EVEX.B.
REX′ field 1510—this is the first part of the REX′ field 1510 and is the EVEX.R′ bit field (EVEX Byte 1, bit [4]—R′) that is used to encode either the upper 16 or lower 16 of the extended 32 register set. In one embodiment of the invention, this bit, along with others as indicated below, is stored in bit inverted format to distinguish (in the well-known x86 32-bit mode) from the BOUND instruction, whose real opcode byte is 62, but does not accept in the MOD R/M field (described below) the value of 11 in the MOD field; alternative embodiments of the invention do not store this and the other indicated bits below in the inverted format. A value of 1 is used to encode the lower 16 registers. In other words, R′Rrrr is formed by combining EVEX.R′, EVEX.R, and the other RRR from other fields.
Opcode map field 1615 (EVEX byte 1, bits [3:0]—mmmm)—its content encodes an implied leading opcode byte (0F, 0F 38, or 0F 3).
Data element width field 1564 (EVEX byte 2, bit [7]—W)—is represented by the notation EVEX.W. EVEX.W is used to define the granularity (size) of the datatype (either 32-bit data elements or 64-bit data elements).
EVEX.vvvv 1620 (EVEX Byte 2, bits [6:3]—vvvv)—the role of EVEX.vvvv may include the following: 1) EVEX.vvvv encodes the first source register operand, specified in inverted (is complement) form and is valid for instructions with 2 or more source operands; 2) EVEX.vvvv encodes the destination register operand, specified in is complement form for certain vector shifts; or 3) EVEX.vvvv does not encode any operand, the field is reserved and should contain 1111b. Thus, EVEX.vvvv field 1620 encodes the 4 low-order bits of the first source register specifier stored in inverted (is complement) form. Depending on the instruction, an extra different EVEX bit field is used to extend the specifier size to 32 registers.
EVEX.U 1568 Class field (EVEX byte 2, bit [2]—U)—If EVEX.U=0, it indicates class A or EVEX.U0; if EVEX.U=1, it indicates class B or EVEX.U1. Prefix encoding field 1625 (EVEX byte 2, bits [1:0]—pp)—provides additional bits for the base operation field. In addition to providing support for the legacy SSE instructions in the EVEX prefix format, this also has the benefit of compacting the SIMD prefix (rather than requiring a byte to express the SIMD prefix, the EVEX prefix requires only 2 bits). In one embodiment, to support legacy SSE instructions that use a SIMD prefix (66H, F2H, F3H) in both the legacy format and in the EVEX prefix format, these legacy SIMD prefixes are encoded into the SIMD prefix encoding field; and at runtime are expanded into the legacy SIMD prefix prior to being provided to the decoder's PLA (so the PLA can execute both the legacy and EVEX format of these legacy instructions without modification). Although newer instructions could use the EVEX prefix encoding field's content directly as an opcode extension, certain embodiments expand in a similar fashion for consistency but allow for different meanings to be specified by these legacy SIMD prefixes. An alternative embodiment may redesign the PLA to support the 2 bit SIMD prefix encodings, and thus not require the expansion.
Alpha field 1552 (EVEX byte 3, bit [7]—EH; also known as EVEX.EH, EVEX.rs, EVEX.RL, EVEX.write mask control, and EVEX.N; also illustrated with α)—as previously described, this field is context specific.
Beta field 1554 (EVEX byte 3, bits [6:4]—SSS, also known as EVEX.s2-0, EVEX.r2-0, EVEX.rr1, EVEX.LL0, EVEX.LLB; also illustrated with βββ)—as previously described, this field is context specific.
REX′ field 1510—this is the remainder of the REX′ field and is the EVEX.V′ bit field (EVEX Byte 3, bit [3]—V′) that may be used to encode either the upper 16 or lower 16 of the extended 32 register set. This bit is stored in bit inverted format. A value of 1 is used to encode the lower 16 registers. In other words, V′VVVV is formed by combining EVEX.V′, EVEX.vvvv.
Write mask field 1570 (EVEX byte 3, bits [2:0]—kkk)—its content specifies the index of a register in the write mask registers as previously described. In one embodiment of the invention, the specific value EVEX.kkk=000 has a special behavior implying no write mask is used for the particular instruction (this may be implemented in a variety of ways including the use of a write mask hardwired to all ones or hardware that bypasses the masking hardware).
Real Opcode Field 1630 (Byte 4) is also known as the opcode byte. Part of the opcode is specified in this field.
MOD R/M Field 1640 (Byte 5) includes MOD field 1642, Reg field 1644, and R/M field 1646. As previously described, the MOD field's 1642 content distinguishes between memory access and non-memory access operations. The role of Reg field 1644 can be summarized to two situations: encoding either the destination register operand or a source register operand, or be treated as an opcode extension and not used to encode any instruction operand. The role of R/M field 1646 may include the following: encoding the instruction operand that references a memory address, or encoding either the destination register operand or a source register operand.
Scale, Index, Base (SIB) Byte (Byte 6)—As previously described, the scale field's 1550 content is used for memory address generation. SIB.xxx 1654 and SIB.bbb 1656—the contents of these fields have been previously referred to with regard to the register indexes Xxxx and Bbbb.
Displacement field 1562A (Bytes 7-10)—when MOD field 1642 contains 10, bytes 7-10 are the displacement field 1562A, and it works the same as the legacy 32-bit displacement (disp32) and works at byte granularity.
Displacement factor field 1562B (Byte 7)—when MOD field 1642 contains 01, byte 7 is the displacement factor field 1562B. The location of this field is that same as that of the legacy x86 instruction set 8-bit displacement (disp8), which works at byte granularity. Since disp8 is sign extended, it can only address between −128 and 127 bytes offsets; in terms of 64 byte cache lines, disp8 uses 8 bits that can be set to only four really useful values −128, −64, 0, and 64; since a greater range is often needed, disp32 is used; however, disp32 requires 4 bytes. In contrast to disp8 and disp32, the displacement factor field 1562B is a reinterpretation of disp8; when using displacement factor field 1562B, the actual displacement is determined by the content of the displacement factor field multiplied by the size of the memory operand access (N). This type of displacement is referred to as disp8*N. This reduces the average instruction length (a single byte of used for the displacement but with a much greater range). Such compressed displacement is based on the assumption that the effective displacement is multiple of the granularity of the memory access, and hence, the redundant low-order bits of the address offset do not need to be encoded. In other words, the displacement factor field 1562B substitutes the legacy x86 instruction set 8-bit displacement. Thus, the displacement factor field 1562B is encoded the same way as an x86 instruction set 8-bit displacement (so no changes in the ModRM/SIB encoding rules) with the only exception that disp8 is overloaded to disp8*N. In other words, there are no changes in the encoding rules or encoding lengths but only in the interpretation of the displacement value by hardware (which needs to scale the displacement by the size of the memory operand to obtain a byte-wise address offset).
Immediate field 1572 operates as previously described.
Full Opcode FieldWhen U=1, the alpha field 1552 (EVEX byte 3, bit [7]—EH) is interpreted as the write mask control (Z) field 1552C. When U=1 and the MOD field 1642 contains 11 (signifying a no memory access operation), part of the beta field 1554 (EVEX byte 3, bit [4]—S0) is interpreted as the RL field 1557A; when it contains a 1 (round 1557A.1) the rest of the beta field 1554 (EVEX byte 3, bit [6-5]—S2-1) is interpreted as the round operation field 1559A, while when the RL field 1557A contains a 0 (VSIZE 1557.A2) the rest of the beta field 1554 (EVEX byte 3, bit [6-5]—S2-1) is interpreted as the vector length field 1559B (EVEX byte 3, bit [6-5]—L1-0). When U=1 and the MOD field 1642 contains 00, 01, or 10 (signifying a memory access operation), the beta field 1554 (EVEX byte 3, bits [6:4]—SSS) is interpreted as the vector length field 1559B (EVEX byte 3, bit [6-5]—L1-0) and the broadcast field 1557B (EVEX byte 3, bit [4]—B).
Exemplary Register Architecture
In other words, the vector length field 1559B selects between a maximum length and one or more other shorter lengths, where each such shorter length is half the length of the preceding length; and instructions templates without the vector length field 1559B operate on the maximum vector length. Further, in one embodiment, the class B instruction templates of the specific vector friendly instruction format 1600 operate on packed or scalar single/double-precision floating point data and packed or scalar integer data. Scalar operations are operations performed on the lowest order data element position in an zmm/ymm/xmm register; the higher order data element positions are either left the same as they were prior to the instruction or zeroed depending on the embodiment.
Write mask registers 1715—in the embodiment illustrated, there are 8 write mask registers (k0 through k7), each 64 bits in size. In an alternate embodiment, the write mask registers 1715 are 16 bits in size. As previously described, in one embodiment of the invention, the vector mask register k0 cannot be used as a write mask; when the encoding that would normally indicate k0 is used for a write mask, it selects a hardwired write mask of 0xFFFF, effectively disabling write masking for that instruction.
General-purpose registers 1725—in the embodiment illustrated, there are sixteen 64-bit general-purpose registers that are used along with the existing x86 addressing modes to address memory operands. These registers are referenced by the names RAX, RBX, RCX, RDX, RBP, RSI, RDI, RSP, and R8 through R15.
Scalar floating point stack register file (x87 stack) 1745, on which is aliased the MMX packed integer flat register file 1750—in the embodiment illustrated, the x87 stack is an eight-element stack used to perform scalar floating-point operations on 32/64/80-bit floating point data using the x87 instruction set extension; while the MMX registers are used to perform operations on 64-bit packed integer data, as well as to hold operands for some operations performed between the MMX and XMM registers.
Alternative embodiments of the invention may use wider or narrower registers. Additionally, alternative embodiments of the invention may use more, less, or different register files and registers.
Exemplary Core Architectures, Processors, and Computer ArchitecturesProcessor cores may be implemented in different ways, for different purposes, and in different processors. For instance, implementations of such cores may include: 1) a general purpose in-order core intended for general-purpose computing; 2) a high performance general purpose out-of-order core intended for general-purpose computing; 3) a special purpose core intended primarily for graphics and/or scientific (throughput) computing. Implementations of different processors may include: 1) a CPU including one or more general purpose in-order cores intended for general-purpose computing and/or one or more general purpose out-of-order cores intended for general-purpose computing; and 2) a coprocessor including one or more special purpose cores intended primarily for graphics and/or scientific (throughput). Such different processors lead to different computer system architectures, which may include: 1) the coprocessor on a separate chip from the CPU; 2) the coprocessor on a separate die in the same package as a CPU; 3) the coprocessor on the same die as a CPU (in which case, such a coprocessor is sometimes referred to as special purpose logic, such as integrated graphics and/or scientific (throughput) logic, or as special purpose cores); and 4) a system on a chip that may include on the same die the described CPU (sometimes referred to as the application core(s) or application processor(s)), the above described coprocessor, and additional functionality. Exemplary core architectures are described next, followed by descriptions of exemplary processors and computer architectures.
Exemplary Core Architectures In-Order and Out-of-Order Core Block DiagramIn
The front end unit 1830 includes a branch prediction unit 1832 coupled to an instruction cache unit 1834, which is coupled to an instruction translation lookaside buffer (TLB) 1836, which is coupled to an instruction fetch unit 1838, which is coupled to a decode unit 1840. The decode unit 1840 (or decoder) may decode instructions, and generate as an output one or more micro-operations, micro-code entry points, microinstructions, other instructions, or other control signals, which are decoded from, or which otherwise reflect, or are derived from, the original instructions. The decode unit 1840 may be implemented using various different mechanisms. Examples of suitable mechanisms include, but are not limited to, look-up tables, hardware implementations, programmable logic arrays (PLAs), microcode read only memories (ROMs), etc. In one embodiment, the core 1890 includes a microcode ROM or other medium that stores microcode for certain macroinstructions (e.g., in decode unit 1840 or otherwise within the front end unit 1830). The decode unit 1840 is coupled to a rename/allocator unit 1852 in the execution engine unit 1850.
The execution engine unit 1850 includes the rename/allocator unit 1852 coupled to a retirement unit 1854 and a set of one or more scheduler unit(s) 1856. The scheduler unit(s) 1856 represents any number of different schedulers, including reservations stations, central instruction window, etc. The scheduler unit(s) 1856 is coupled to the physical register file(s) unit(s) 1858. Each of the physical register file(s) units 1858 represents one or more physical register files, different ones of which store one or more different data types, such as scalar integer, scalar floating point, packed integer, packed floating point, vector integer, vector floating point, status (e.g., an instruction pointer that is the address of the next instruction to be executed), etc. In one embodiment, the physical register file(s) unit 1858 comprises a vector registers unit, a write mask registers unit, and a scalar registers unit. These register units may provide architectural vector registers, vector mask registers, and general purpose registers. The physical register file(s) unit(s) 1858 is overlapped by the retirement unit 1854 to illustrate various ways in which register renaming and out-of-order execution may be implemented (e.g., using a reorder buffer(s) and a retirement register file(s); using a future file(s), a history buffer(s), and a retirement register file(s); using a register maps and a pool of registers; etc.). The retirement unit 1854 and the physical register file(s) unit(s) 1858 are coupled to the execution cluster(s) 1860. The execution cluster(s) 1860 includes a set of one or more execution units 1862 and a set of one or more memory access units 1864. The execution units 1862 may perform various operations (e.g., shifts, addition, subtraction, multiplication) and on various types of data (e.g., scalar floating point, packed integer, packed floating point, vector integer, vector floating point). While some embodiments may include a number of execution units dedicated to specific functions or sets of functions, other embodiments may include only one execution unit or multiple execution units that all perform all functions. The scheduler unit(s) 1856, physical register file(s) unit(s) 1858, and execution cluster(s) 1860 are shown as being possibly plural because certain embodiments create separate pipelines for certain types of data/operations (e.g., a scalar integer pipeline, a scalar floating point/packed integer/packed floating point/vector integer/vector floating point pipeline, and/or a memory access pipeline that each have their own scheduler unit, physical register file(s) unit, and/or execution cluster—and in the case of a separate memory access pipeline, certain embodiments are implemented in which only the execution cluster of this pipeline has the memory access unit(s) 1864). It should also be understood that where separate pipelines are used, one or more of these pipelines may be out-of-order issue/execution and the rest in-order.
The set of memory access units 1864 is coupled to the memory unit 1870, which includes a data TLB unit 1872 coupled to a data cache unit 1874 coupled to a level 2 (L2) cache unit 1876. In one exemplary embodiment, the memory access units 1864 may include a load unit, a store address unit, and a store data unit, each of which is coupled to the data TLB unit 1872 in the memory unit 1870. The instruction cache unit 1834 is further coupled to a level 2 (L2) cache unit 1876 in the memory unit 1870. The L2 cache unit 1876 is coupled to one or more other levels of cache and eventually to a main memory.
By way of example, the exemplary register renaming, out-of-order issue/execution core architecture may implement the pipeline 1800 as follows: 1) the instruction fetch 1838 performs the fetch and length decoding stages 1802 and 1804; 2) the decode unit 1840 performs the decode stage 1806; 3) the rename/allocator unit 1852 performs the allocation stage 1808 and renaming stage 1810; 4) the scheduler unit(s) 1856 performs the schedule stage 1812; 5) the physical register file(s) unit(s) 1858 and the memory unit 1870 perform the register read/memory read stage 1814; the execution cluster 1860 perform the execute stage 1816; 6) the memory unit 1870 and the physical register file(s) unit(s) 1858 perform the write back/memory write stage 1818; 7) various units may be involved in the exception handling stage 1822; and 8) the retirement unit 1854 and the physical register file(s) unit(s) 1858 perform the commit stage 1824.
The core 1890 may support one or more instructions sets (e.g., the x86 instruction set (with some extensions that have been added with newer versions); the MIPS instruction set of MIPS Technologies of Sunnyvale, Calif.; the ARM instruction set (with optional additional extensions such as NEON) of ARM Holdings of Sunnyvale, Calif.), including the instruction(s) described herein. In one embodiment, the core 1890 includes logic to support a packed data instruction set extension (e.g., AVX1, AVX2), thereby allowing the operations used by many multimedia applications to be performed using packed data.
It should be understood that the core may support multithreading (executing two or more parallel sets of operations or threads), and may do so in a variety of ways including time sliced multithreading, simultaneous multithreading (where a single physical core provides a logical core for each of the threads that physical core is simultaneously multithreading), or a combination thereof (e.g., time sliced fetching and decoding and simultaneous multithreading thereafter such as in the Intel® Hyperthreading technology).
While register renaming is described in the context of out-of-order execution, it should be understood that register renaming may be used in an in-order architecture. While the illustrated embodiment of the processor also includes separate instruction and data cache units 1834/1874 and a shared L2 cache unit 1876, alternative embodiments may have a single internal cache for both instructions and data, such as, for example, a Level 1 (L) internal cache, or multiple levels of internal cache. In some embodiments, the system may include a combination of an internal cache and an external cache that is external to the core and/or the processor. Alternatively, all of the cache may be external to the core and/or the processor.
Specific Exemplary in-Order Core Architecture
The local subset of the L2 cache 1904 is part of a global L2 cache that is divided into separate local subsets, one per processor core. Each processor core has a direct access path to its own local subset of the L2 cache 1904. Data read by a processor core is stored in its L2 cache subset 1904 and can be accessed quickly, in parallel with other processor cores accessing their own local L2 cache subsets. Data written by a processor core is stored in its own L2 cache subset 1904 and is flushed from other subsets, if necessary. The ring network ensures coherency for shared data. The ring network is bi-directional to allow agents such as processor cores, L2 caches and other logic blocks to communicate with each other within the chip. Each ring data-path is 1012-bits wide per direction.
Processor with Integrated Memory Controller and Graphics
Thus, different implementations of the processor 2000 may include: 1) a CPU with the special purpose logic 2008 being integrated graphics and/or scientific (throughput) logic (which may include one or more cores), and the cores 2002A-N being one or more general purpose cores (e.g., general purpose in-order cores, general purpose out-of-order cores, a combination of the two); 2) a coprocessor with the cores 2002A-N being a large number of special purpose cores intended primarily for graphics and/or scientific (throughput); and 3) a coprocessor with the cores 2002A-N being a large number of general purpose in-order cores. Thus, the processor 2000 may be a general-purpose processor, coprocessor or special-purpose processor, such as, for example, a network or communication processor, compression engine, graphics processor, GPGPU (general purpose graphics processing unit), a high-throughput many integrated core (MIC) coprocessor (including 30 or more cores), embedded processor, or the like. The processor may be implemented on one or more chips. The processor 2000 may be a part of and/or may be implemented on one or more substrates using any of a number of process technologies, such as, for example, BiCMOS, CMOS, or NMOS.
The memory hierarchy includes one or more levels of cache within the cores, a set or one or more shared cache units 2006, and external memory (not shown) coupled to the set of integrated memory controller units 2014. The set of shared cache units 2006 may include one or more mid-level caches, such as level 2 (L2), level 3 (L3), level 4 (L4), or other levels of cache, a last level cache (LLC), and/or combinations thereof. While in one embodiment a ring based interconnect unit 2012 interconnects the integrated graphics logic 2008, the set of shared cache units 2006, and the system agent unit 2010/integrated memory controller unit(s) 2014, alternative embodiments may use any number of well-known techniques for interconnecting such units. In one embodiment, coherency is maintained between one or more cache units 2006 and cores 2002-A-N.
In some embodiments, one or more of the cores 2002A-N are capable of multi-threading. The system agent 2010 includes those components coordinating and operating cores 2002A-N. The system agent unit 2010 may include for example a power control unit (PCU) and a display unit. The PCU may be or include logic and components needed for regulating the power state of the cores 2002A-N and the integrated graphics logic 2008. The display unit is for driving one or more externally connected displays.
The cores 2002A-N may be homogenous or heterogeneous in terms of architecture instruction set; that is, two or more of the cores 2002A-N may be capable of execution the same instruction set, while others may be capable of executing only a subset of that instruction set or a different instruction set.
Exemplary Computer ArchitecturesReferring now to
The optional nature of additional processors 2115 is denoted in
The memory 2140 may be, for example, dynamic random access memory (DRAM), phase change memory (PCM), or a combination of the two. For at least one embodiment, the controller hub 2120 communicates with the processor(s) 2110, 2115 via a multi-drop bus, such as a frontside bus (FSB), point-to-point interface such as QuickPath Interconnect (QPI), or similar connection 2195.
In one embodiment, the coprocessor 2145 is a special-purpose processor, such as, for example, a high-throughput MIC processor, a network or communication processor, compression engine, graphics processor, GPGPU, embedded processor, or the like. In one embodiment, controller hub 2120 may include an integrated graphics accelerator.
There can be a variety of differences between the physical resources 2110, 2115 in terms of a spectrum of metrics of merit including architectural, microarchitectural, thermal, power consumption characteristics, and the like.
In one embodiment, the processor 2110 executes instructions that control data processing operations of a general type. Embedded within the instructions may be coprocessor instructions. The processor 2110 recognizes these coprocessor instructions as being of a type that should be executed by the attached coprocessor 2145. Accordingly, the processor 2110 issues these coprocessor instructions (or control signals representing coprocessor instructions) on a coprocessor bus or other interconnect, to coprocessor 2145. Coprocessor(s) 2145 accept and execute the received coprocessor instructions.
Referring now to
Processors 2270 and 2280 are shown including integrated memory controller (IMC) units 2272 and 2282, respectively. Processor 2270 also includes as part of its bus controller units point-to-point (P-P) interfaces 2276 and 2278; similarly, second processor 2280 includes P-P interfaces 2286 and 2288. Processors 2270, 2280 may exchange information via a point-to-point (P-P) interface 2250 using P-P interface circuits 2278, 2288. As shown in
Processors 2270, 2280 may each exchange information with a chipset 2290 via individual P-P interfaces 2252, 2254 using point to point interface circuits 2276, 2294, 2286, 2298. Chipset 2290 may optionally exchange information with the coprocessor 2238 via a high-performance interface 2239. In one embodiment, the coprocessor 2238 is a special-purpose processor, such as, for example, a high-throughput MIC processor, a network or communication processor, compression engine, graphics processor, GPGPU, embedded processor, or the like.
A shared cache (not shown) may be included in either processor or outside of both processors, yet connected with the processors via P-P interconnect, such that either or both processors' local cache information may be stored in the shared cache if a processor is placed into a low power mode.
Chipset 2290 may be coupled to a first bus 2216 via an interface 2296. In one embodiment, first bus 2216 may be a Peripheral Component Interconnect (PCI) bus, or a bus such as a PCI Express bus or another third generation I/O interconnect bus, although the scope of the present invention is not so limited.
As shown in
Referring now to
Referring now to
Embodiments of the mechanisms disclosed herein may be implemented in hardware, software, firmware, or a combination of such implementation approaches. Embodiments of the invention may be implemented as computer programs or program code executing on programmable systems comprising at least one processor, a storage system (including volatile and non-volatile memory and/or storage elements), at least one input device, and at least one output device.
Program code, such as code 2230 illustrated in
The program code may be implemented in a high level procedural or object oriented programming language to communicate with a processing system. The program code may also be implemented in assembly or machine language, if desired. In fact, the mechanisms described herein are not limited in scope to any particular programming language. In any case, the language may be a compiled or interpreted language.
One or more aspects of at least one embodiment may be implemented by representative instructions stored on a machine-readable medium which represents various logic within the processor, which when read by a machine causes the machine to fabricate logic to perform the techniques described herein. Such representations, known as “IP cores” may be stored on a tangible, machine readable medium and supplied to various customers or manufacturing facilities to load into the fabrication machines that actually make the logic or processor.
Such machine-readable storage media may include, without limitation, non-transitory, tangible arrangements of articles manufactured or formed by a machine or device, including storage media such as hard disks, any other type of disk including floppy disks, optical disks, compact disk read-only memories (CD-ROMs), compact disk rewritable's (CD-RWs), and magneto-optical disks, semiconductor devices such as read-only memories (ROMs), random access memories (RAMs) such as dynamic random access memories (DRAMs), static random access memories (SRAMs), erasable programmable read-only memories (EPROMs), flash memories, electrically erasable programmable read-only memories (EEPROMs), phase change memory (PCM), magnetic or optical cards, or any other type of media suitable for storing electronic instructions.
Accordingly, embodiments of the invention also include non-transitory, tangible machine-readable media containing instructions or containing design data, such as Hardware Description Language (HDL), which defines structures, circuits, apparatuses, processors and/or system features described herein. Such embodiments may also be referred to as program products.
Emulation (Including Binary Translation, Code Morphing, Etc.)In some cases, an instruction converter may be used to convert an instruction from a source instruction set to a target instruction set. For example, the instruction converter may translate (e.g., using static binary translation, dynamic binary translation including dynamic compilation), morph, emulate, or otherwise convert an instruction to one or more other instructions to be processed by the core. The instruction converter may be implemented in software, hardware, firmware, or a combination thereof. The instruction converter may be on processor, off processor, or part on and part off processor.
Components, features, and details described for any of
In the description and claims, the terms “coupled” and/or “connected,” along with their derivatives, may have be used. These terms are not intended as synonyms for each other. Rather, in embodiments, “connected” may be used to indicate that two or more elements are in direct physical and/or electrical contact with each other. “Coupled” may mean that two or more elements are in direct physical and/or electrical contact with each other. However, “coupled” may also mean that two or more elements are not in direct contact with each other, but yet still co-operate or interact with each other. For example, an execution unit may be coupled with a register and/or a decode unit through one or more intervening components. In the figures, arrows are used to show connections and couplings.
The term “and/or” may have been used. As used herein, the term “and/or” means one or the other or both (e.g., A and/or B means A or B or both A and B).
In the description above, specific details have been set forth in order to provide a thorough understanding of the embodiments. However, other embodiments may be practiced without some of these specific details. The scope of the invention is not to be determined by the specific examples provided above, but only by the claims below. In other instances, well-known circuits, structures, devices, and operations have been shown in block diagram form and/or without detail in order to avoid obscuring the understanding of the description. Where considered appropriate, reference numerals, or terminal portions of reference numerals, have been repeated among the figures to indicate corresponding or analogous elements, which may optionally have similar or the same characteristics, unless specified or clearly apparent otherwise.
Certain operations may be performed by hardware components, or may be embodied in machine-executable or circuit-executable instructions, that may be used to cause and/or result in a machine, circuit, or hardware component (e.g., a processor, potion of a processor, circuit, etc.) programmed with the instructions performing the operations. The operations may also optionally be performed by a combination of hardware and software. A processor, machine, circuit, or hardware may include specific or particular circuitry or other logic (e.g., hardware potentially combined with firmware and/or software) is operable to execute and/or process the instruction and store a result in response to the instruction.
Some embodiments include an article of manufacture (e.g., a computer program product) that includes a machine-readable medium. The medium may include a mechanism that provides, for example stores, information in a form that is readable by the machine. The machine-readable medium may provide, or have stored thereon, an instruction or sequence of instructions, that if and/or when executed by a machine are operable to cause the machine to perform and/or result in the machine performing one or operations, methods, or techniques disclosed herein. The machine-readable medium may store or otherwise provide one or more of the embodiments of the instructions disclosed herein.
In some embodiments, the machine-readable medium may include a tangible and/or non-transitory machine-readable storage medium. For example, the tangible and/or non-transitory machine-readable storage medium may include a floppy diskette, an optical storage medium, an optical disk, an optical data storage device, a CD-ROM, a magnetic disk, a magneto-optical disk, a read only memory (ROM), a programmable ROM (PROM), an erasable-and-programmable ROM (EPROM), an electrically-erasable-and-programmable ROM (EEPROM), a random access memory (RAM), a static-RAM (SRAM), a dynamic-RAM (DRAM), a Flash memory, a phase-change memory, a phase-change data storage material, a non-volatile memory, a non-volatile data storage device, a non-transitory memory, a non-transitory data storage device, or the like. The non-transitory machine-readable storage medium does not consist of a transitory propagated signal.
Examples of suitable machines include, but are not limited to, a general-purpose processor, a special-purpose processor, an instruction processing apparatus, a digital logic circuit, an integrated circuit, or the like. Still other examples of suitable machines include a computing device or other electronic device that includes a processor, instruction processing apparatus, digital logic circuit, or integrated circuit. Examples of such computing devices and electronic devices include, but are not limited to, desktop computers, laptop computers, notebook computers, tablet computers, netbooks, smartphones, cellular phones, servers, network devices (e.g., routers and switches.), Mobile Internet devices (MIDs), media players, smart televisions, nettops, set-top boxes, and video game controllers.
Reference throughout this specification to “one embodiment,” “an embodiment,” “one or more embodiments,” “some embodiments,” for example, indicates that a particular feature may be included in the practice of the invention but is not necessarily required to be. Similarly, in the description various features are sometimes grouped together in a single embodiment, Figure, or description thereof for the purpose of streamlining the disclosure and aiding in the understanding of various inventive aspects. This method of disclosure, however, is not to be interpreted as reflecting an intention that the invention requires more features than are expressly recited in each claim. Rather, as the following claims reflect, inventive aspects lie in less than all features of a single disclosed embodiment. Thus, the claims following the Detailed Description are hereby expressly incorporated into this Detailed Description, with each claim standing on its own as a separate embodiment of the invention.
EXAMPLE EMBODIMENTSThe following examples pertain to further embodiments. Specifics in the examples may be used anywhere in one or more embodiments.
Example 1 is a processor including a plurality of packed data registers, and a decode unit to decode an instruction. The instruction is to indicate a first source packed data that is to include a first set of at least four data elements, is to indicate a second source packed data that is to include a second set of at least four data elements, and is to indicate a destination storage location. An execution unit is coupled with the packed data registers and the decode unit. The execution unit, in response to the instruction, is to store a result packed data in the destination storage location. The result packed data is to include at least four indexes. The indexes are to identify corresponding data element positions in the first and second source packed data. The indexes are to be stored in positions in the result packed data that are to represent a sorted order of corresponding data elements in the first and second source packed data.
Example 2 includes the processor of Example 1, in which the execution unit is to store the result packed data in which each of the indexes is to identify a corresponding data element position in each of the first and second source packed data. Also, in which the execution unit, in response to the instruction, is to store a result mask that is to have at least four mask elements, in which each mask element is to correspond to a different one of the indexes. Also, in which each mask element is to indicate whether the data element position for the corresponding index is in the first source packed data or the second source packed data.
Example 3 includes the processor of Example 2, further including a mask register to store the result mask. The instruction is included in an instruction set that includes a second instruction that is able to indicate the result mask as a predicate operand to predicate a packed data operation.
Example 4 includes the processor of Example 1, in which the execution unit is to store the result packed data in which each of the indexes is to identify the corresponding single data element in one of the first and second source packed data.
Example 5 includes the processor of any of Examples 1 to 4, in which the execution unit, in response to the instruction, is to store a second result packed data in a second destination storage location that is to be indicated by the instruction. The second result packed data is to include the data elements from the first and second source packed data that correspond to the indexes stored in positions of the second result packed data that reflect the sorted order.
Example 6 includes the processor of any of Examples 1 to 4, in which the decode unit is to decode the instruction that is to indicate the first source packed data having the at least four data elements that are assumed to be in sorted order for the instruction, and that is to indicate the second source packed data having the at least four data elements that are assumed to be in sorted order for the instruction.
Example 7 includes the processor of any of Examples 1 to 4, in which the decode unit is to decode the instruction that is to indicate the first source packed data having the at least four data elements that are not assumed to be in sorted order for the instruction, and that is to indicate the second source packed data having the at least four data elements that are not assumed to be in sorted order for the instruction.
Example 8 includes the processor of any of Examples 1 to 4, in which the execution unit is to store the result packed data in which the indexes are to be stored in the positions that are to represent the sorted order of the corresponding data elements that are to include a smallest half of all of the data elements of the first and second source packed data.
Example 9 includes the processor of any of Examples 1 to 4, in which the execution unit is to store the result packed data in which the indexes are to be stored in the positions that are to represent the sorted order of the corresponding data elements that are to include a largest half of all of the data elements of the first and second source packed data.
Example 10 includes the processor of any of Examples 1 to 4, in which the decode unit is to decode the instruction that is to indicate the first source packed data that is to include at least eight data elements which are each to have one of 32-bits and 64-bits.
Example 11 is a method in a processor that includes receiving an instruction. The instruction indicating a first source packed data including a first set of at least four data elements, indicating a second source packed data including a second set of at least four data elements, and indicating a destination storage location. Storing result packed data in the destination storage location in response to the instruction. The result packed data including at least four indexes. The indexes identifying corresponding data element positions in the first and second source packed data. The indexes stored in positions in the result packed data that represent a sorted order of corresponding data elements in the first and second source packed data.
Example 12 includes the method of Example 11, in which receiving includes receiving the instruction indicating the first source packed data having the at least four data elements in sorted order.
Example 13 includes the method of any one of Examples 11 to 12, in which storing the result packed data includes storing result packed data in which each of the indexes identifies the corresponding single data element in one of the first and second source packed data.
Example 14 is processor including a plurality of packed data registers, and a decode unit to decode an instruction. The instruction to indicate a source packed data that is to include at least four data elements that are not in a sorted order, and to indicate a destination storage location. An execution unit is coupled with the packed data registers and the decode unit. The execution unit, in response to the instruction, is to store a result packed data in the destination storage location. The result packed data to include at least four indexes. The indexes to identify corresponding data elements in the source packed data. The indexes to be stored in positions in the result packed data that are to represent the sorted order of the corresponding data elements in the source packed data.
Example 15 includes the processor of Example 14, in which the execution unit, in response to the instruction, is to store a second result packed data in a second destination storage location that is to be indicated by the instruction, the second result packed data to include the corresponding data elements stored in positions of the second result packed data that reflect the sorted order.
Example 16 includes the processor of any of Examples 14 to 15, in which the result packed data is to include indexes corresponding to all data elements in the source packed data.
Example 17 includes the processor of any of Examples 14 to 15, in which the decode unit is to decode the instruction that is to indicate the source packed data that is to include at least eight data elements that are each to have one of 32-bits and 64-bits.
Example 18 is a method in a processor including receiving an instruction, the instruction indicating a source packed data including at least four data elements that are not in a sorted order, and indicating a destination storage location. Storing result packed data in the destination storage location in response to the instruction. The result packed data including at least four indexes. The indexes identifying corresponding data elements in the source packed data. The indexes stored in positions in the result packed data that represent the sorted order of the corresponding data elements in the source packed data.
Example 19 includes the method of Example 18, further including storing a second result packed data in a second destination storage location indicated by the instruction, the second result packed data including the corresponding data elements stored in positions that reflect the sorted order.
Example 20 includes the method of any one of Examples 18 to 19, in which receiving includes receiving the instruction indicating the source packed data having at least eight data elements that each have one of 32-bits and 64-bits, and in which storing includes storing the result packed data including indexes corresponding to all data elements in the source packed data.
Example 21 is a system to process instructions including an interconnect, and a processor coupled with the interconnect. The processor to receive an instruction, the instruction to indicate a first source packed data that is to include a first set of at least four data elements, to indicate a second source packed data that is to include a second set of at least four data elements, and to indicate a destination register. The processor, in response to the instruction, to store a result packed data in the destination register. The result packed data to include at least four indexes. The indexes to identify corresponding data element positions in the first and second source packed data. The indexes to be stored in positions in the result packed data that are to represent a sorted order of corresponding data elements in the first and second source packed data. A dynamic random access memory (DRAM) is coupled with the interconnect. The DRAM optionally stores an algorithm to use the indexes of the result packed data to sort data.
Example 22 includes the system of claim 21, in which the processor is to store the result packed data in which each of the indexes is to identify the corresponding single data element in one of the first and second source packed data.
Example 23 is an article of manufacture including a non-transitory machine-readable storage medium, the non-transitory machine-readable storage medium storing an instruction. The instruction to indicate a first source packed data that is to include a first set of at least four data elements, to indicate a second source packed data that is to include a second set of at least four data elements, and that is to indicate a destination storage location. The instruction if executed by a machine is to cause the machine to perform operations including storing result packed data in the destination storage location in response to the instruction. The result packed data including at least four indexes. The indexes identifying corresponding data element positions in the first and second source packed data. The indexes stored in positions in the result packed data that represent a sorted order of corresponding data elements in the first and second source packed data.
Example 24 includes the article of manufacture of claim 23, in which the instruction indicates the first source packed data having the at least four data elements in sorted order.
Example 25 is a processor or other apparatus including means for performing the method of any of claims 11-13.
Example 26 is a processor or other apparatus including modules, units, logic, circuitry, means, or any combination thereof, to perform the method of any of claims 11-13.
Example 27 is an article of manufacture including a machine-readable medium, which is optionally a non-transitory machine-readable storage medium, that provides an instruction that if and/or when executed by a processor, computer system, or other machine is operable to cause the machine to perform the method of any of claims 11-13.
Example 28 is a computer system or other electronic device including an interconnect; a processor coupled with the interconnect; and at least one component coupled with the interconnect that is selected from a dynamic random access memory (DRAM), a graphics chip, a wireless communications chip, a phase change memory, and a video camera, the computer system or other electronic device operable to perform the method of any of claims 11-13.
Example 29 is a processor or other apparatus including means for performing the method of any of claims 18-20.
Example 30 is a processor or other apparatus including modules, units, logic, circuitry, means, or any combination thereof, to perform the method of any of claims 18-20.
Example 31 is an article of manufacture including a machine-readable medium, which is optionally a non-transitory machine-readable storage medium, that provides an instruction that if and/or when executed by a processor, computer system, or other machine is operable to cause the machine to perform the method of any of claims 18-20.
Example 32 is a computer system or other electronic device including an interconnect; a processor coupled with the interconnect; and at least one component coupled with the interconnect that is selected from a dynamic random access memory (DRAM), a graphics chip, a wireless communications chip, a phase change memory, and a video camera, the computer system or other electronic device operable to perform the method of any of claims 18-20.
Example 33 is a processor or other apparatus that is operable to perform one or more operations or any method substantially as described herein.
Example 34 is a processor or other apparatus including means for performing one or more operations or any method substantially as described herein.
Example 35 is a processor or other apparatus to perform any of the instructions substantially as described herein.
Example 36 is a processor or other apparatus including means for performing any of the instructions substantially as described herein.
Example 37 includes a method that includes converting a first instruction, which may be any of the instructions substantially as disclosed herein, and which is of a first instruction set, into one or more instructions of a second instruction set. The method also includes decoding and executing the one or more instructions of the second instruction set on a processor. The executing includes storing a result in a destination. The result may include any of the results substantially as disclosed herein for the first instruction.
Example 38 includes a processor or other apparatus including a decode unit that is operable to decode instructions of a first instruction set. The decode unit is to receive one or more instructions that emulate a first instruction, which may be any of the instructions substantially as disclosed herein, and which is to be of a second instruction set. The processor or other apparatus also includes one or more execution units coupled with the decode unit to execute the one or more instructions of the first instruction set. The one or more execution units in response to the one or more instructions of the first instruction set are operable to store a result in a destination. The result may include any of the results substantially as disclosed herein for the first instruction.
Example 39 includes a computer system or other electronic device that includes a processor having a decode unit that is operable to decode instructions of a first instruction set, and having one or more execution units. The computer system also includes a storage device coupled to the processor. The storage device is to store a first instruction, which may be any of the instructions substantially as disclosed herein, and which is to be of a second instruction set. The storage device is also to store instructions to convert the first instruction into one or more instructions of the first instruction set. The one or more instructions of the first instruction set, when executed by the processor, are operable to cause the processor to store a result in a destination. The result may include any of the results substantially as disclosed herein for the first instruction.
Claims
1. A processor comprising:
- a plurality of packed data registers;
- a decode unit to decode an instruction, the instruction to indicate a first source packed data that is to include a first set of at least four data elements, to indicate a second source packed data that is to include a second set of at least four data elements, and to indicate a destination storage location;
- an execution unit coupled with the packed data registers and the decode unit, the execution unit, in response to the instruction, to store a result packed data in the destination storage location,
- the result packed data to include at least four indexes,
- the indexes to identify corresponding data element positions in the first and second source packed data, and
- the indexes to be stored in positions in the result packed data that are to represent a sorted order of corresponding data elements in the first and second source packed data.
2. The processor of claim 1, wherein the execution unit is to store the result packed data in which each of the indexes is to identify a corresponding data element position in each of the first and second source packed data, and wherein the execution unit, in response to the instruction, is to store a result mask that is to have at least four mask elements, in which each mask element is to correspond to a different one of the indexes, and in which each mask element is to indicate whether the data element position for the corresponding index is in the first source packed data or the second source packed data.
3. The processor of claim 2, further comprising a mask register to store the result mask, and wherein the instruction is included in an instruction set that includes a second instruction that is able to indicate the result mask as a predicate operand to predicate a packed data operation.
4. The processor of claim 1, wherein the execution unit is to store the result packed data in which each of the indexes is to identify the corresponding single data element in one of the first and second source packed data.
5. The processor of claim 1, wherein the execution unit, in response to the instruction, is to store a second result packed data in a second destination storage location that is to be indicated by the instruction, the second result packed data to include the data elements from the first and second source packed data that correspond to the indexes stored in positions of the second result packed data that reflect the sorted order.
6. The processor of claim 1, wherein the decode unit is to decode the instruction that is to indicate the first source packed data having the at least four data elements that are assumed to be in sorted order for the instruction, and that is to indicate the second source packed data having the at least four data elements that are assumed to be in sorted order for the instruction.
7. The processor of claim 1, wherein the decode unit is to decode the instruction that is to indicate the first source packed data having the at least four data elements that are not assumed to be in sorted order for the instruction, and that is to indicate the second source packed data having the at least four data elements that are not assumed to be in sorted order for the instruction.
8. The processor of claim 1, wherein the execution unit is to store the result packed data in which the indexes are to be stored in the positions that are to represent the sorted order of the corresponding data elements that are to include a smallest half of all of the data elements of the first and second source packed data.
9. The processor of claim 1, wherein the execution unit is to store the result packed data in which the indexes are to be stored in the positions that are to represent the sorted order of the corresponding data elements that are to include a largest half of all of the data elements of the first and second source packed data.
10. The processor of claim 1, wherein the decode unit is to decode the instruction that is to indicate the first source packed data that is to include at least eight data elements which are each to have one of 32-bits and 64-bits.
11. A method in a processor comprising:
- receiving an instruction, the instruction indicating a first source packed data including a first set of at least four data elements, indicating a second source packed data including a second set of at least four data elements, and indicating a destination storage location; and
- storing result packed data in the destination storage location in response to the instruction, the result packed data including,
- at least four indexes,
- the indexes identifying corresponding data element positions in the first and second source packed data, and
- the indexes stored in positions in the result packed data that represent a sorted order of corresponding data elements in the first and second source packed data.
12. The method of claim 11, wherein receiving comprises receiving the instruction indicating the first source packed data having the at least four data elements in sorted order.
13. The method of claim 11, wherein storing the result packed data comprises storing result packed data in which each of the indexes identifies the corresponding single data element in one of the first and second source packed data.
14. A processor comprising:
- a plurality of packed data registers;
- a decode unit to decode an instruction, the instruction to indicate a source packed data that is to include at least four data elements that are not in a sorted order, and to indicate a destination storage location;
- an execution unit coupled with the packed data registers and the decode unit, the execution unit, in response to the instruction, to store a result packed data in the destination storage location,
- the result packed data to include at least four indexes,
- the indexes to identify corresponding data elements in the source packed data, and
- the indexes to be stored in positions in the result packed data that are to represent the sorted order of the corresponding data elements in the source packed data.
15. The processor of claim 14, wherein the execution unit, in response to the instruction, is to store a second result packed data in a second destination storage location that is to be indicated by the instruction, the second result packed data to include the corresponding data elements stored in positions of the second result packed data that reflect the sorted order.
16. The processor of claim 14, wherein the result packed data is to include indexes corresponding to all data elements in the source packed data.
17. The processor of claim 14, wherein the decode unit is to decode the instruction that is to indicate the source packed data that is to include at least eight data elements that are each to have one of 32-bits and 64-bits.
18. The processor of claim 14, wherein the packed data registers are able to store 512-bit packed data.
Type: Application
Filed: Sep 18, 2017
Publication Date: Jan 4, 2018
Applicant: Intel Corporation (Santa Clara, CA)
Inventors: Shay Gueron (Haifa), Vlad Krasnov (Nesher)
Application Number: 15/707,633
