METHOD AND APPARATUS FOR EXTRACTING AND DEPOSITING A STRING OF BITS FROM A SOURCE INTO A DESTINATION
FIELD OF INVENTION:
The present invention relates to the field of computer systems. More particularly, the present invention relates to instructions used to move data within a string from one location to another.
BACKGROUND OF THE INVENTION:
The IEEE standard, "IEEE 1394 Standard For A High Performance Serial Bus," Draft ratified in 1995, is an international standard for implementing an inexpensive high-speed serial bus architecture which supports both asynchronous and isochronous format data transfers. Isochronous data transfers are real-time transfers which deliver data on time without guaranteeing the integrity of the data. Each packet of data transferred isochronously is transferred in its own time period. The IEEE 1394-1995 standard bus architecture provides up to sixty- four (64) channels for isochronous data transfer between applications. A six bit channel number is broadcast with the data to ensure reception by the appropriate application. This allows multiple applications to simultaneously transmit isochronous data across the bus structure. Asynchronous transfers are traditional data transfer operations which guarantee the integrity of the data during delivery using an acknowledgement protocol.
The IEEE 1394-1995 standard provides a high-speed serial bus for interconnecting digital devices thereby providing a universal I/O connection. The IEEE 1394-1995 standard defines a digital interface for the applications thereby eliminating the need for an application to convert digital data to analog data before it is transmitted across the bus. Correspondingly, a receiving application will receive digital data from the bus, not analog data, and will therefore not be required to convert analog data to digital data. Devices can be added and removed from an IEEE 1394-1995 bus while the bus is active. If a device is so added or removed the bus will then automatically reconfigure itself for transmitting data between the then existing nodes. A node is considered a logical entity with a unique identification number on the bus structure. Each node provides an identification ROM, a standardized set of control registers and its own address space.
The IEEE 1394-1995 standard provides for up to sixty- four different isochronous channels to be used within an IEEE 1394-1995 network of devices. However, in current implementations, certain 1394 devices are being built with the capability to only transmit and receive isochronous data over a subset of less than sixty- four channels. When receiving data on an isochronous channel, that data must be processed by the receiving device. This processing includes any or all of displaying, manipulating, forwarding and storing. Often, data received on different isochronous channels is processed differently, depending on the
type of device from which the data is received, the type of data that is received and the desired use of the data. If data received on an isochronous channel is not received and processed efficiently, errors in the display or use of the data can result.
There are a wide variety of computer systems capable of processing digital data. A basic structure of a computer system is shown in Figure 1A. The heart of the computer system 1 is a central processing unit ("CPU") 2. Within a computer system 1 the CPU 2 is coupled to firmware 4, data storage devices 5, ports 3, and random access memory ("RAM") 6 by a bus structure 7. Data storage devices 5 include hard drives, floppy drives, and CD- ROMs. Input/output ("I/O") devices such as a display monitor 8 and an IEEE 1394-1995 device 10, are coupled to the bus structure 7 through ports 3. A keyboard 9 is also coupled to the CPU 2 through one of the ports 3. Ports 3, both serial or parallel, are used to connect the computer system 1 to modems, printers, and other devices, including other computer systems. Figure IB illustrates a computer system 1 coupled to a display monitor 8 and networked to an IEEE 1394-1995 device 10, such as a video camera, through an IEEE 1394-1995 serial cable 11.
In a computer system 1, firmware 4 is used to seek out and load an operating system from one of the data storage devices 5 (usually the hard drive) when the computer system 1 is first turned on. Programs and applications used by the computer system 1 are generally stored on the hard drive and moved at least in part to the RAM 6 during use. Common CPUs 2 included within a computer system 1 include reduced instruction set computation ("RISC") processors or complex instruction set computation ("CISC") processors. Examples of RISC processors are the PowerPC™ processor manufactured by International Business Machines Corporation and the G3 processor manufactured by Motorola Corporation for Apple Computer Corporation personal computers. Examples of CISC processors are the model 80x86 processor and the Pentium™ processor, which are both available from Intel Corporation of Santa Clara, California.
A CPU 2 stores data in internal memory locations, registers, and memory. Registers are used during program execution to temporarily store intermediate results. The advantage of storing data in a register instead of a memory location is that the data within the register can be accessed much faster. Data that is not used during register operation is stored in memory. Memory associated with a processor ("associated memory") is typically located within the CPU 2 itself as LI cache, nearby the CPU 2 as L2 cache, or in an area separate from the CPU 2.
The location in which data is stored in the registers and memory is identified by an address. A read operation is used to access data found at a specific address. A write operation is used to store data at a specific address. Writing a value to a specific address will erase the value previously found at that address.
Computer systems are controlled by instructions. Instructions are statements specifying an operation to be performed and what data operands are to be processed by the computer system. A queue of pre-selected and sequenced instructions make up each
computer program. Each instruction includes an operation code ("opcode") and operands. The opcode is the part of the instruction that identifies the operation to be performed. Typical operations are ADD, SUBTRACT, and MOVE.
Operands describe the data to be processed as the operation specified by the opcode is carried out. The instruction's operands may be an address location or actual data. Placing actual data within the instruction typically results in faster execution of the instruction. Limitation in the instruction's size, however, usually dictates that most operands are address locations for data stored in memory or registers.
A collection of instructions to be used by a particular computer system 1 are referred to as an instruction set. In RISC architectures, the instructions are of uniform length. In x86
CISC processors, the length of instructions varies widely. The minimum instruction consists of a single opcode byte and is 8 bits long. A long instruction that includes a prefix byte is as long as 104 bits. Longer instructions containing more than a single prefix byte are also possible. One common instruction completed by the CPU 2 is a shift instruction. Shifting is the process of moving data that is stored in a storage device relative to the boundaries of the device, as opposed to moving data in or out of the device. The storage device is often a register designed specifically for shifting ("shift register"). The direction of the shift is either left or right. Vacated bit positions (on the leftmost for shift right operations and on the rightmost for shift left operations) are filled with logical ZEROs. Shift operations are often used in field alignments, packing and unpacking of data items into storage units, and highspeed multiplication and division. Simple shift registers shift data only one space per shift. More advanced shift registers shift data any arbitrary number of spaces per individual shift.
An operation very similar to shifting is rotation. Rotation differs from shifting in that, in a left shift operation, a bit rotated out from the left is placed back into the vacated rightmost bit position. Similarly, in a right shift operation, a bit rotated out from the right is placed back into the vacated leftmost bit position. Otherwise shift and rotate operations are identical.
Another common instruction completed by the CPU 2 is a mask instruction. Masking is used to extract desired information from a storage unit while suppressing the undesired information. In the below example, only the 8 least significant bits of the 16 bit string are extracted from the original register bit string:
01010111 01011100 register bit string
00000000 11111111 mask bit string 00000000 01011100 bit string result
As shown, a bitwise logical AND operation is performed with the register bit string and the mask bit string. Where the value of the mask bit is logical ONE, the corresponding register bit is retained in the bit string result. Where the value of the mask bit is logical ZERO, the register bit is suppressed. The mask bit string is generated during the execution of the instruction from data included within the instruction.
A masking operation is used in combination with bit string read operations, shift registers, bitwise logic operations, and bit string write operations to deposit a string of bits into a specific memory or register location. An extract function is a form of a mask operation. For a source bit string S, a destination bit string D, and a mask bit string Mask, an extract function performs a bitwise logical AND operation with the source bit string S and the mask bit string Mask, then places the bit string result into the destination bit string D. In boolean algebra, the equation reads:
D = S AND Mask
A more complex mask operation is the deposit function. In a deposit function, the bits of the destination string D are preserved in the areas masked in the source string S. In boolean algebra, the equation reads:
D = (S AND Mask) OR (D AND -Mask)
Mask bit strings usually follow predictable patterns. First, the logical ONEs of the mask are typically grouped together. Second, the mask is typically right justified or left justified. Below are examples of 16 bit mask strings.
00000011 11111111 example one
11111111 11000000 example two
Due to their predictable patterns, mask bit strings can be defined in fewer bits than their full length. Defining the mask in fewer bits allows instructions sets to save space within the masking instruction. The cost of saved space, however, is that an additional decoding step is required to generate the mask.
SUMMARY OF THE INVENTION:
The method of and apparatus for extracting a string of bits from a binary bit string and depositing a string of bits onto a binary bit string of the present invention is an improved implementation of deposit and extract instructions wherein the instruction contains an opcode, a source address, a destination address, a shift number, and a K-bit mask string. The opcode describes the operations to be performed upon a J-bit source string and an N-bit destination string. The source address points to the register in the CPU or the location of the J-bit source string. The destination address points to the register in the CPU or the location of the N-bit destination string. The shift number indicates the number of bits the J-bit source string will be shifted to generate a shifted bit string. The combination of the shifted bit string with the N-bit destination string is conducted under the control of the K-bit mask string. The method of and apparatus for extracting a string of bits from a binary bit string and depositing a string of bits onto a binary bit string of the present invention is particularly useful for high speed digital data processing, such as that required by IEEE 1394-1995 compliant devices.
An instruction includes an opcode, a source address, a destination address, a shift number, and a mask bit string. The opcode describes the operations to be performed upon a particular source bit string and destination bit string. The operations include an extract left instruction, an extract right instruction, a deposit left instruction, and a deposit right
instruction. The source address points to the register in the CPU or the location of the source bit string. The destination address points to the register in the CPU or the location of the destination bit string. The shift number indicates the number of bits the source bit string is to be shifted to generate a shifted bit string. The direction of shift is dictated by the shift value or the opcode. The combination of the shifted bit string with the destination bit string is conducted under the control of a mask bit string. The more specific implementations of the present invention are the extract and deposit instructions.
The deposit instruction also begins with an instruction comprising an opcode, a source address, a destination address, a shift value, and a K-bit mask bit string. The CPU first reads a J-bit source string located at the source address and an N-bit destination string located at the destination address. The CPU shifts the J-bit source string as determined by the shift number and the opcode to obtain a shifted bit string. The CPU then combines the shifted bit string and the N-bit destination string under control of the K-bit mask string to obtain an N-bit final string, such that: (i) individual bits of the shifted bit string are included in the N-bit final string where the corresponding individual bits of the K-bit mask string have a value equal to logical ONE; and individual bits of the N-bit destination string are included in the N-bit final string where the coπesponding individual bits of the K-bit mask string have a value equal to logical ZERO. In a final step, the CPU writes the N-bit final string to the destination address. There are three additional implementations of the deposit instruction. In the first additional implementation, the numeric values of J, K, and N are equal. In the second additional implementation, the combination step is performed by the following steps: (i) performing a logical AND operation of the shifted bit string and the K-bit mask string to obtain a first bit string; (ii) performing a logical AND operation of the N-bit destination string and the logical complement of the K-bit mask string to obtain a second bit string; and (iii) performing a logical OR operation of the first bit string and the second bit string to obtain the
N-bit final string. In the third additional implementation, the processing steps are performed by an embedded stream processor and the registers within the embedded stream processor contain the source address and the destination address.
Like the deposit instruction, the extract instruction begins with an instruction comprising an opcode, a source address, a destination address, a shift number, and a K-bit mask string. The CPU or equivalent means first reads a J-bit source string located at the source address. The CPU shifts the J-bit source string as determined by the shift number and the opcode to obtain a shifted bit string. The CPU then combines the shifted bit string and the K-bit mask string to obtain an N-bit final string, such that: (i) individual bits of the shifted bit string are included in the N-bit final string where the corresponding individual bits of the K- bit mask string have a value of logical ONE; and (ii) remaining individual bits of the N-bit final string have a value of logical ZERO. In a final step, the CPU writes the N-bit final string to the destination address.
There are three additional implementations of the extract instruction. In the first additional implementation, the numeric values of J, K, and N are equal. In the second additional implementation, the combination step is accomplished by performing a bitwise logical AND operation with the shifted bit string and the K-bit mask string. In the third additional implementation, the processing steps are performed by an embedded stream processor and the registers within the embedded stream processor contain the source address and the destination address.
BRIEF DESCRIPTION OF THE DRAWINGS: Figure 1A illustrates a block diagram showing the basic components of an exemplary computer system.
Figure IB shows a computer system networked to an IEEE 1394-1995 device through an IEEE 1394-1995 serial cable.
Figures 2A through 2E show different combinations of source bit strings and destination bit strings.
Figure 3 shows a format of the deposit and extract instructions according to the preferred embodiment of the present invention.
Figure 4 shows a block diagram of a circuit for completing an extract left instruction according to the preferred embodiment of the present invention. Figure 5 shows a block diagram of a circuit for completing an extract right instruction according to the preferred embodiment of the present invention.
Figure 6 shows a block diagram of a circuit for completing a deposit left instruction according to the preferred embodiment of the present invention.
Figure 7 shows a block diagram of a circuit for completing a deposit right instruction. Figure 8 illustrates a block diagram showing the basic components of an exemplary computer system with an IEEE 1394-1995 interface circuit and attached IEEE 1394-1995 devices.
DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENT: The method of and apparatus for extracting a string of bits from a binary bit string and depositing a string of bits onto a binary bit string of the present invention preferably include the mask bit string within the deposit and extract instructions, thereby eliminating a mask generation step during the execution of the instruction.
Figures 2A through 2E illustrate possible combinations of a source bit string, from which a string of bits is extracted, and a destination bit string, to which a string of bits is deposited. The section L within the source bit string represents the bits selected for extraction. The section M within the destination bit string represents the bits to be replaced by the section L bits extracted from the source bit string. The bits designated by "s" in the source bit string represent bits that are removed by masking during an extract instruction. The bits designated by "d" in the destination bit string represent bits that are either zeroed out
if the operands are subjected to an extract instruction or that will remain if the operands are subjected to a deposit instruction. The designation "X" represents the number of bits to the right of the section L in the source bit string. The designation "Y" represents the number of bits to the right of the section M in the destination bit string. In the source and destination bit strings illustrated in Figure 2 A, the value of X in the source bit string 60 and the value of Y in the destination bit string 72 are both greater than zero. In this example, the value of Y is also greater than the value of X. In the source and destination bit strings illustrated in Figure 2B, the value of X in the source bit string 62 and the value of Y in the destination bit string 74 are both equal to zero. In the source and destination bit strings illustrated in Figure 2C, the value of X in the source bit string 64 is equal to zero and the value of Y in the destination bit string 76 is greater than zero. In the source and destination bit strings illustrated in Figure 2D, the value of X in the source bit string 66 is greater than zero and the value of Y in the destination bit string 78 is equal to zero. In the source and destination bit strings illustrated in Figure 2E, the value of X in the source bit string 68 and the value of Y in the destination bit string 80 are both greater than zero. In this example, the value of X is also greater than the value of Y.
The direction of the shifting step in an extract or deposit instruction is determined by the relative positions of the section L in the source bit string and the section M in the destination bit string. As an example, in Figures 2A and 2C, the section M is located to the left of the section L. To move the bits within the section L to the location within the section
M, a shift left step must be performed. Similarly, in Figures 2D and 2E, the section M is located to the right of the section L and a shift right step must be performed to move the bits within the section L to the location within the section M. Because section M and section L are both right justified in the example of Figure 2B, no shifting is required. Within the preferred embodiment of the present invention, the direction of the shifting step is contained in the opcode. Extract left and deposit right instructions shift the bits within the section L to the left. Extract right and deposit right instructions shift the bits within the section L to the right.
The magnitude of each shift operation is determined by the difference between the X and Y values. In the preferred embodiment, the shift value is the absolute value of the difference between the X and Y values. In an alternative embodiment, the shift value is a signed number and the sign of the shift value determines the shift direction.
The instruction format of the preferred embodiment is illustrated in Figure 3. The instruction 20 is preferably 64 bits wide. The prefeπed format of the instruction 20 is set forth below in Table I.
TABLE I: Instruction Format
The opcode 21 is located within the six most significant bits of the instruction 20 and informs the processor which operation is to be performed. The source address 22 is located within the next eight most significant bits and specifies the location for the source bit string. The destination address 23 is located within the next eight most significant bits and specifies the location of the destination bit string. The shift value 24 is located within the next six most significant bits and represents the number of bits the source bit string must be shifted to properly place the section L of the source bit string into the section M of the destination bit string. The shift value 24 is not an address. The mask bit string 25 is located within the thirty- four least significant bits of the instruction 20 and is similar to the source bit string and the destination bit string in that it contains 34 bits and the two most significant bits are flag bits. The flag bits indicate first whether the value in the source address 22 is a data packet header and second whether there is another packet that will be sent that is related to the present packet. Unlike the source bit string and the destination bit string, the mask bit string 25 is located within the instruction 20. Like the shift value 24, the mask bit string 25 is generated prior to the execution of the instruction 20. In alternative embodiments of the instruction 20, the length of the source bit string, the destination bit string, and the mask bit string 25 can be of different lengths.
In the preferred embodiment, the direction of the shift operation is dictated by the opcode 21. The extract left instruction, for example, has a different opcode 21 than the extract right instruction. In an alternative embodiment, the shift direction is included in the shift value 24. In this alternative embodiment, the difference between a left shift versus a right shift is encoded in the shift value 24 by use of a flag bit or use of signed numbers.
A block diagram of a circuit for implementing an extract left instruction is illustrated in Figure 4. In this implementation, an extract left source address 112 from the instruction 20 is loaded into a source register 26. A J-bit source string 27 located at the extract left source address is then loaded into a shift register 28 from the source register 26. An extract left shift value 114 from the instruction 20 and an extract left opcode 1 10 from the instruction 20 are also loaded into the shift register 28. Taking the left shift direction from the extract left opcode 110 and the amount of shift from the extract left shift value 114, the shift register 28 then produces a shifted bit string 29. This shifted bit string 29 is then provided as an input to a logical AND gate 115. An extract left K-bit mask string 1 16 from the instruction 20 is also
provided as an input to the logical AND gate 115. The logical AND gate 115 performs a logical AND operation on the shifted bit string 29 and the K-bit mask string 116 and provides an N-bit final string 30 as an output. The N-bit final string 30 is provided from the logical AND gate 115 to the destination register 31. A destination register 31 writes the N-bit final string 30 to an extract left destination address 118 dictated by the instruction 20.
A block diagram of a circuit for implementing an extract right instruction is illustrated in Figure 5. In this implementation, an extract right source address 122 from the instruction 20 is loaded into the source register 26. A J-bit source string 27 located at the extract right source address 122 is then loaded into a shift register 28 from the source register 26. An extract right shift value 124 from the instruction 20 and an extract right opcode 120 from the instruction 20 are also loaded into the shift register 28. Taking the right shift direction from the extract right opcode and the amount of shift from the extract right shift value 124, the shift register 28 then produces a shifted bit string 29. This shifted bit string 29 is then provided as an input to the logical AND gate 1 15. An extract right K-bit mask string 126 from the instruction 20 is also provided as an input to the logical AND gate 115. The logical
AND gate 115 performs a logical AND operation on the shifted bit string 29 and the K-bit mask string 126 and provides an N-bit final string 30 as an output. The N-bit final string 30 is provided from the logical AND gate 115 to the destination register 31. The destination register 31 writes the N-bit final string 30 to an extract right destination address 128 dictated by the instruction 20.
A block diagram of a circuit for implementing a deposit left instruction is illustrated in Figure 6. In this implementation, a deposit left source address 132 from the instruction 20 is loaded into the source register 26. A J-bit source string 27 located at the deposit left source address 132 is then loaded into a shift register 28 from the source register 26. A deposit left shift value 136 and deposit left opcode 130 from the instruction 20 are also loaded into the shift register 28. Taking the left shift direction from the deposit left opcode 130 and the amount of shift from the deposit left shift value 136, the shift register 28 then produces a shifted bit string 29. This shifted bit string 29 is then provided as an input to the logical AND gate 115. A deposit left K-bit mask string 138 from the instruction 20 is also provided as an input to the logical AND gate 115. The logical AND gate 115 performs a logical AND operation on the shifted bit string 29 and the K-bit mask string 138 and provides a first bit string 32 as an output. The first bit string 32 is provided from the logical AND gate 115 as an input to the logical OR gate 117. A deposit left destination address 134 from the instruction 20 is loaded into a destination register 34 and a destination register 31. The destination register 34 receives the deposit left destination address 134 from the instruction 20 and reads the N-bit destination string 35 located at the deposit left destination address 134. The N-bit destination string 35 is provided as an input to a logical AND gate 119. The deposit left K-bit mask string 138 from the instruction 20 is provided as an input to an inverter circuit 121. The inverter circuit 121 inverts the bits within the deposit left K-bit mask string 138. The bitwise complement 36 of the deposit left K-bit mask string 138 is provided as an output from the
inverter circuit 121 to the logical AND gate 119 as an input. The logical AND gate 119 performs a logical AND operation on the N-bit destination string 35 and the bitwise complement 36 and provides a second bit string 33 as an output. The second bit string 33 is provided from the logical AND gate 119 as an input to the logical OR gate 117. The logical OR gate 117 performs a logical OR operation on the first bit string 32 and the second bit string 33 and provides an N-bit final string 30 as an output. The N-bit final string 30 is provided from the logical OR gate 117 to the destination register 31. The destination register 31 writes the N-bit final string 30 to a deposit left destination address 134 dictated by the instruction 20. A block diagram of a circuit for implementing a deposit right instruction is illustrated in Figure 7. In this implementation, a deposit right source address 152 from the instruction 20 is loaded into the source register 26. A J-bit source string 27 located at the deposit right source address 152 is then loaded into a shift register 28 from the source register 26. A deposit right shift value 156 and deposit right opcode 150 from the instruction 20 are also loaded into the shift register 28. Taking the right shift direction from the deposit right opcode
150 and the amount of shift from the deposit right shift value 156, the shift register 28 then produces a shifted bit string 29. This shifted bit string 29 is then provided as an input to the logical AND gate 115. A deposit right K-bit mask string 158 from the instruction is also provided as an input to the logical AND gate 115. The logical AND gate 115 performs a logical AND operation on the shifted bit string 29 and the K-bit mask string 158 and provides a first bit string 32 as an output. The first bit string 32 is provided from the logical AND gate 115 as an input to a logical OR gate 117.
A deposit right destination address 154 from the instruction 20 is loaded into the destination register 34 and a destination register 31. The destination register 34 receives the deposit right destination address 154 from the instruction 20 and reads the N-bit destination string 35 located at the deposit right destination address 154. The N-bit destination string 35 is provided as an input to a logical AND gate 119. The deposit right K-bit mask string 158 from the instruction 20 is provided as an input to an inverter circuit 121. The inverter circuit 121 inverts the bits within the deposit right K-bit mask string 158. The bitwise complement 36 of the deposit right K-bit mask string 158 is provided as an output from the inverter circuit
121 to the logical AND gate 119 as an input. The logical AND gate 119 performs a logical AND operation on the N-bit destination string 35 and the bitwise complement 36 and provides a second bit string 33 as an output. The second bit string 33 is provided from the logical AND gate 119 as an input to the logical OR gate 117. The logical OR gate 117 performs a logical OR operation on the first bit string 32 and the second bit string 33 and provides an N-bit final string 30 as an output. The N-bit final string 30 is provided from the logical OR gate 117 to the destination register 31. The destination register 31 writes the N-bit final string 30 to a deposit right destination address 154 dictated by the instruction 20.
In the deposit and extract instructions of the present invention, the length of the J-bit source string 27, the K-bit mask string 25, and the N-bit destination string 35 is often the
same length. The 34 bit length is used in the preferred embodiment described in Table I because many video packets have 32 bit headers and 2 flag bits. The flag bits indicate whether additional packets exist and whether the 32 bits are header information or data.
Deposit and extract instructions can be executed by the CPU 2 of a computer system 1 or by an embedded stream processor. Use of an embedded stream processor within a computer system 1 is illustrated in Figure 8. Figure 8 shows a computer system 1 that includes a CPU 2, firmware 4, data storage 5, RAM 6, a video card 17, an embedded stream processor 15, and an IEEE 1394-1995 interface circuit 13, all of which are intercoupled by a bus structure 7. A display monitor 8 is coupled to the video card 17. A keyboard 9 is coupled to the CPU 2. The embedded stream processor 15 is coupled to the bus structure 7 and the IEEE 1394-1995 interface circuit 13. The IEEE 1394-1995 interface circuit 13 is coupled to a video camera 16 by an IEEE 1394-1995 serial bus cable 92. The video camera 16 is coupled to a second IEEE 1394-1995 device 10 by a second IEEE 1394-1995 serial bus cable 90. The prefeπed embodiment of the embedded stream processor 15 is taught within U.S.
Patent Application Serial Number 08/612,322, filed on March 7, 1996, and entitled "Isochronous Data Pipe for Managing and Manipulating a High-Speed Stream of Isochronous Data Flowing Between an Application and a Bus Structure" which is hereby incorporated by reference. The embedded stream processor 15 is programmable and will execute a series of instructions on a stream of data in order to perform operations and manipulations on the data as required to place the data in the appropriate format. Within the present invention, the embedded stream processor 15 is tasked with converting digital data from one application format to another. Digital data is generally transferred in data packets over the IEEE 1394-1995 serial bus. The data packets contain both header and data fields. The header provides information such as the size and format of the data packet. Using the extract instructions of the present invention, the embedded stream processor 15 selectively captures and modifies appropriate header and data bits. The captured header and data bits are then provided to appropriate locations using the deposit instructions of the present invention. The present invention has been described in terms of specific embodiments incorporating details to facilitate the understanding of principles of construction and operation of the invention. References to specific embodiments and details of the invention are not intended to limit the scope of the appended claims. It will be apparent to those skilled in the art that modifications may be made in the illustrated embodiment without departing from the spirit and scope of the invention. Specifically, it will be apparent to those skilled in the art that while the prefeπed embodiment of the present invention is used with an IEEE
1394-1995 serial bus structure, the present invention could also be implemented on any other appropriate systems or bus structures, including other or later versions of the IEEE 1395 serial bus.