JP2018101448A - Semiconductor device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a semiconductor device capable of obtaining output data corresponding to input data by referencing table data with a simple configuration.SOLUTION: A DTC 3 comprises a first transfer source address register, a second transfer source address register, a transfer destination address register, and a transfer count register. The DTC reads first data from the transfer source address SAR1 specified in the first transfer source address register, operates the first data to generate second data, and transfers data to the transfer destination address DAR specified in the transfer destination address register with the transfer source address SAR2 indicated in the second transfer source address register as the transfer source address. The DTC then generates a new transfer source address SAR2 or a new transfer destination address DAR by adding the second data to or deleting it from the transfer source address SAR2 or the transfer destination address DAR. The DTC repeats the data transfer only the number of times CR specified in the transfer count register.SELECTED DRAWING: Figure 17

Description

  The present invention relates to a semiconductor device, for example, a microcomputer.

 In general, a single chip microcomputer is mainly composed of a central processing unit (CPU), a ROM (Read Only Memory) for holding a program, and a RAM (Random Access Memory) for holding a data. Memory and a functional block such as an input / output circuit for inputting / outputting data or signals are formed on one semiconductor substrate. Such a single chip microcomputer is used for various device control.

  In device control of a single chip microcomputer, there is a request for data transfer in response to an event such as an interrupt. Although the CPU can realize arbitrary processing by a combination of instructions, when interrupt processing is performed, exception processing, stack save / return processing, and execution of a return instruction are required to switch the processing flow. In this case, the operation time of the CPU such as instruction read when performing data transfer tends to be long.

  In order to solve the above-mentioned problems of data transfer, a technology has been proposed in which a data transfer device is provided in a single-chip microcomputer and data is transferred in response to requests from a large number of peripheral processing devices (input / output circuits) with a small amount of hardware (Patent Document 1). In this technique, data transfer information such as a transfer source address indicating the position of a memory in which data to be transferred is stored is stored in a storage device (RAM). In addition, a vector table is provided for storing addresses indicating where in the storage device (RAM) all information necessary for data transfer is stored. A means for referring to the contents of the vector table in response to the activation request and a means for obtaining all information necessary for the data transfer from the contents of the vector table are provided. With this technology, data transfer can be realized with a small amount of hardware, but the contents of the data transfer are not considered.

  On the other hand, in order to expand the application range of data transfer in the data transfer device, a technique for performing different data transfer depending on the data transfer mode has been proposed (Patent Document 2). In this technique, a repeat transfer mode and a block transfer mode have been proposed as data transfer modes. This makes it possible to control the transfer destination and transfer source addresses, select the number of transfers, and the like. For example, if the present technology is applied to a system such as a printer, it is possible to control a stepping motor and print data of the printer. It is also suitable for storing received data in a memory. Also, with this technology, data transfer information is stored as dedicated hardware inside the data transfer device, and the configuration of transfer information in the short address mode / full address mode can be selected in order to effectively use this hardware. It has also been shown to do. In this example, the stepping motor corresponds to the rotation angle and the movement amount, does not require feedback, and only needs to be able to transfer a predetermined number of data in a predetermined order. In this technique, one of the transfer source address and the transfer destination address is a RAM or the like, but it is not considered to use the RAM while updating the stored contents of the RAM.

  A technique has also been proposed in which information necessary for data transfer is stored in a storage device, and data transfer based on at least one piece of information can be specified (chain transfer or chain operation) by one operation of the data transfer device. (Patent Document 3). According to the present technology, an arbitrary number of transfers can be performed by an arbitrary activation factor, so that the present technology can be applied to various applications. As a result, the degree of freedom in system configuration can be improved and usability can be improved.

  There has also been proposed a technique in which an arithmetic unit capable of performing comparison and simple calculation between data set in advance in the data transfer device and data to be transferred has been proposed (Patent Document 4). In this technique, data transfer is executed by a data transfer device that is dedicated hardware, so that data transfer at a speed higher than that of the CPU can be realized. As a result, since the frequency of CPU interrupt processing can be reduced, efficient processing can be performed.

Japanese Patent Laid-Open No. 1-125644 JP-A-5-307516 JP-A-7-129537 JP 2000-194647 A

  However, the inventors have found that the above-described technique has the following problems. In recent years, functions and controlled objects built in microcomputers are increasing, and higher accuracy is required for microcomputer control. For example, it is required to output data corresponding to input data. However, if dedicated hardware is provided for each type of input data, the physical scale of the microcomputer increases. Further, the structure of the microcomputer becomes complicated, and the processing time for obtaining output data also becomes long.

  Other problems and novel features will become apparent from the description of the specification and the accompanying drawings.

  According to one embodiment, a semiconductor device includes a data transfer device that transfers data from a transfer source address region to a transfer destination address region based on data transfer information, and the data transfer device includes a first transfer source address. The second transfer source address information is calculated based on the data read from the information, the read is performed based on the calculation result, and the read data is written based on the transfer destination address information.

  According to an embodiment, output data corresponding to input data can be obtained by referring to table data with a simple configuration in a semiconductor device.

1 is a block diagram schematically illustrating a configuration of a microcomputer according to a first embodiment. It is a figure which shows the basic data flow of a data transfer apparatus (DTC). It is a figure which shows the structure of data transfer information. It is a block diagram showing typically the composition of a data transfer device (DTC). It is a figure which shows the state transition of a data transfer apparatus (DTC). It is a figure which shows the data transfer in a branch mode. It is a figure which shows an example of designation | designated of the data transfer information in branch mode. It is a figure which shows the data transfer in shift mode. It is a figure which shows an example of the designation | designated method of the table used by shift mode. It is a figure which shows the data transfer in offset mode. It is a figure which shows an example of the designation | designated method of the table used by offset mode. It is a figure which shows the data transfer in a proportional mode. It is a figure which shows the data transfer in differential mode. It is a figure which shows the data transfer in integral mode. It is a figure which shows data transfer in case transfer mode is block addition mode. It is a figure which shows the data transfer in the modification of branch mode. It is a state transition diagram of DTC in the modification of a branch mode. It is a figure which shows the data transfer in an address calculation mode. It is a figure which shows an example of the designation | designated method of the table used in address calculation mode. It is a block diagram of an interrupt controller (INT). 1 is a block diagram of a camera system including a microcomputer (MCU).

  Embodiments of the present invention will be described below with reference to the drawings. In the drawings, the same elements are denoted by the same reference numerals, and redundant description is omitted as necessary.

Embodiment 1
A microcomputer 100 that is a semiconductor device according to the first embodiment will be described. FIG. 1A is a block diagram schematically showing the configuration of the microcomputer 100 according to the first embodiment. FIG. 1B shows a basic data flow of a data transfer device (DTC) included in the microcomputer 100. Hereinafter, the configuration of the microcomputer 100 will be described with reference to FIG. 1A. Hereinafter, the microcomputer is abbreviated as MCU (Micro Controller Unit).

  The MCU 100 includes a central processing unit (CPU) 1, an interrupt controller (hereinafter referred to as INT) 2, a data transfer device (data transfer controller (hereinafter referred to as DTC)) 3, a read-only memory ( ROM) 4, random access memory (RAM) 5, bus controller (hereinafter referred to as BSC) 6, timer 7, communication module 8, analog module 9, input / output port (I / O) 10 and internal bus 11 Etc. As the communication module 8, for example, a serial communication interface or the like is provided. For example, an analog / digital (A / D) converter and a digital / analog (D / A) converter are provided as the analog module 9.

  In the MCU 100, the CPU 1 is the main operation. The CPU 1 mainly operates by reading instructions from the ROM 4. The DTC 3 performs data transfer on behalf of the CPU 1 based on the setting of the CPU 1.

  The INT 2 receives an interrupt request from the timer 7, the communication module 8, the analog module 9, etc., or an interrupt request from the I / O 10 based on a plurality of external interrupt request signals from outside the MCU 1, and sends it to the CPU 1 or DTC 3. An interrupt request or a data transfer request is output. In addition, the INT 2 sends an interrupt clear signal, which is a signal for clearing the interrupt when processing by the interrupt request signal is started or ended, to the interrupt signal of each of the timer 7, the communication module 8, the analog module 9, and the I / O 10. Output in response to the interrupt factor flag.

  The BSC 6 receives the bus request signals from the CPU 1 and the DTC 3, arbitrates the occupation of the internal bus 11, and outputs a bus use permission signal. The BSC 6 interfaces with the CPU 1 and the DTC 3 such as a bus request, a bus acknowledge, a bus command, a wait, an address, and data, and implements read / write to the RAM and other functional blocks and modules connected to the internal bus 11.

  As described above, the DTC 3 performs data transfer based on the data transfer information. FIG. 2 shows the configuration of the data transfer information. The data transfer information includes a mode register (also referred to as MR), a first source address register (also referred to as SAR1), a second source address register (also referred to as SAR2), a destination address register (also referred to as DAR), and a data register (also referred to as DR). And a transfer count register (also referred to as CR).

  MR designates the mode of data transfer performed by DTC3. The DTC 3 uses necessary data from DR, SAR1, SAR2, DAR, and CR depending on the data transfer mode specified by the MR.

  The CR is divided into a block transfer count register (also referred to as BTCR) and a transfer count register (also referred to as TCR). The TCR is divided into 8-bit TCRH and 8-bit TCRL.

  These registers have one set of circuits in the DTC 3 and are not particularly limited, but do not exist in the address space of the CPU 1. The data transfer information to be stored in these registers is arranged in a predetermined data transfer information arrangement area such as the RAM 5 in the required number of sets in the address space of the CPU 1.

  Next, the MR bit configuration will be described. The MR bits 31 to 16 are used to set the data transfer mode and to specify and input data for the table transfer data transfer function. Bits 15 to 0 of MR perform designation of table reference of the data transfer function for table reference and designation of data transfer of the normal data transfer function. At this time, the table is obtained by arranging a plurality of data or parameters in advance in a predetermined order on the address space.

  Bit 31 of MR is a TLU bit and selects whether or not to perform table reference in data transfer. When the TLU bit is cleared to “0”, the table reference is not executed. The data transfer information at this time varies depending on the value of the DRE bit described later. When the TLU bit is “1” and the DRE bit is cleared to “0”, the data transfer information is 132 bits (32 bits × 4) of MR, SAR2, DAR, and CR. When the TLU bit is “1” and the DRE bit is set to “1”, the data transfer information is 160 bits (32 bits × 5) of MR, DR, SAR2, DAR, and CR.

  When the TLU bit is set to “1”, a table reference is executed. At this time, the table reference mode is designated based on the SFM bit, OFM bit, PRM bit, DRM bit, INM bit, BRM bit, AEM bit, and NOP bit, which will be described later. The table reference mode will be described later.

  The MR bits 30 to 23 are the SFM bit, OFM bit, PRM bit, DRM bit, INM bit, BRM bit, AEM bit and NOP bit, respectively. These bits are information for designating a table reference mode when executing table reference.

  When the SFM bit is “1”, the table reference mode is the shift mode. When the OFM bit is “1”, the table reference mode is the offset mode. When the PRM bit is “1”, the table reference mode is the proportional mode. When the DRM bit is “1”, the table reference mode is the differential mode. When the INM bit is “1”, the table reference mode is the integration mode. When the BRM bit is “1”, the table reference mode is a branch mode. When the AEM bit is “1”, the table reference mode is the address calculation mode. When the NOP bit is “1”, the table reference mode is the no operation mode.

  In the shift mode, the data transfer information changes depending on the value of the CRE bit described later. When the CRE bit is set to “1”, the data transfer information is 160 bits (32 bits × 5) of MR, SAR1, SAR2, DAR, and CR. When the CRE bit is cleared to “0”, the data transfer information is 132 bits (32 bits × 4) of MR, SAR1, SAR2, and DAR.

  In the offset mode, the proportional mode, and the differential mode, the data transfer information changes depending on the value of the CRE bit described later. When the CRE bit is set to “1”, the data transfer information is 182 bits (32 bits × 6) of MR, DR, DAR1, SAR2, DAR, and CR. When the CRE bit is cleared to “0”, the data transfer information is 160 bits (32 bits × 5) of MR, DR, SAR1, SAR2, and DAR.

  In the integration mode and the address calculation mode, the data transfer information is 160 bits (32 bits × 5) of MR, SAR1, SAR2, DAR, and CR.

  In the branch mode, the data transfer information is 96 bits (32 bits × 3) of MR, SAR1, and SAR2.

  In the no operation mode, the data transfer information is 32 bits (32 bits × 1) of MR only.

  Details of the data transfer in each table reference mode will be described later.

  Bits 22 to 20 are SF [2: 0] bits and specify shift of input data by +3 to -3 bits. When +, perform left arithmetic shift, and when-, perform right arithmetic shift.

  Bit 19 and bit 18 are S1M [1: 0] bits, respectively, which specify whether SAR1 is incremented, decremented, or fixed after data transfer.

  Bits 17 and 16 are ISz [1: 0] bits, and select whether to read input data in byte size, word size, or long word size.

  Bits 15 and 14 of MR are SM [1: 0] bits, and specify whether SAR2 is incremented, decremented, or fixed after data transfer.

  Bits 13 and 12 of MR are DM [1: 0] bits, and designate whether DAR is incremented, decremented, or fixed after data transfer.

  MR bits 11 and 10 are TMD [1: 0] bits and designate a data transfer mode. When the TMD [1: 0] bits are “00”, the transfer mode is the normal mode. When the TMD [1: 0] bits are “01”, the transfer mode is a repeat mode. When the TMD [1: 0] bits are “10”, the transfer mode is the block transfer mode. When the TMD [1: 0] bits are “11”, the transfer mode is the block addition mode. Details of each transfer mode will be described later.

  Bit 9 of MR is a DIR bit, and designates which of the transfer source and transfer destination is a repeat area and which is a block area.

  Bits 7 and 6 of MR are Sz [1: 0] bits, respectively, and specify whether one data transfer is performed in byte size, word size, or long word size.

  Bit 3 is the DRE bit. When the DRE bit is cleared to “0”, the read data is written without using DR. When the DRE bit is set to “1”, DR is used, and data obtained by subtracting the contents of DR from the read data is written.

  Bit 2 is the CRE bit. When the CRE bit is cleared to “0”, the DTC operation is performed without limitation in accordance with the activation request signal. At this time, if the transfer mode is the normal mode, the CR is not used. When the CRE bit is set to “1”, the DTC 3 operates according to the value initially set in the CR.

  Bit 1 of MR is the NXTE1 bit, and designates whether to end data transfer or perform the next data transfer for one activation factor. When the NXTE1 bit is cleared to “0”, after the data transfer information is read and transferred, the data transfer information is written and the operation of the DTC 3 is terminated. When the NXTE1 bit is set to “1”, the data transfer information is written after the data transfer information read and data transfer. Further, data transfer information is read from successive addresses. Then, after performing the data transfer specified by the data transfer information, the data transfer information is written. An operation in which a series of processes of data transfer information write, data transfer, and data transfer information write is continuously performed a plurality of times is referred to as a chain operation. Bit 0 of MR is the NXTE0 bit, and specifies whether to perform a chain operation when CR becomes “0”.

  A configuration of the DTC 3 will be described. FIG. 3 is a block diagram schematically showing the configuration of DTC3. The DTC 3 includes a data transfer control block (also referred to as DTCCNT) 31, a bus interface (also referred to as BIF) 32, a vector generation block (also referred to as VG) 33, a vector address register (also referred to as VAR) 34, MR, SAR1, SAR2, and DAR. , CR, DR, an arithmetic unit (also referred to as ALU) 35 and an internal bus 36.

  The DTCCNT 31 controls the DTC 3 based on the contents of the activation request signal DCTREQ from the INT 2 and the MR.

  The BIF 32 provides an interface between the internal bus 36 of the DTC 3 and the internal bus 11 of the MCU 1. This interface includes a bus request, a bus acknowledge, a bus command, a wait, an address, data, and the like.

  The VG 33 generates a vector address according to the vector number DTCVEC given from the INT2. For example, the vector number DTCVEC is quadrupled and a predetermined offset is added.

  The VAR 34 stores the start address of the data transfer information read from the vector address.

  MR, SAR1, SAR2, DAR, and CR store data transfer information sequentially read from the head address of the data transfer information.

  The ALU 35 has functions such as logical operation, shift, and arithmetic operation, and executes these operations in a predetermined order. Although not shown, the internal bus 36 has a plurality of buses, and the contents of SAR1, SAR2, DAR, CR, and DR can be given to the ALU 35 through the plurality of buses. The ALU 35 can perform calculations using SAR1, SAR2, DAR, CR, and DR.

Normal mode When the TMD [1: 0] bits are “00”, the transfer mode is the normal mode. In the normal mode, one data transfer is performed from the transfer source address indicated by SAR to the transfer destination address indicated by DAR by one activation. After the data transfer is finished, the SAR is incremented, decremented or fixed according to the SM [1: 0] bits. The DAR is incremented, decremented or fixed according to the DM [1: 0] bits. Thereafter, the TCR is decremented.

  In the normal mode, the data transfer and register operation described above are repeated as many times as specified by the TCR each time an activation factor occurs.

  After repeating the operation as many times as specified by CR, the CPU is requested to generate an interrupt that has caused the activation.

Repeat mode When the TMD [1: 0] bits are “01”, the transfer mode is the repeat mode. When the DIR bit is cleared to “0”, the transfer source address is set as a repeat area. When the DIR bit is set to “1”, the transfer destination address is set as a repeat area. In the repeat mode, TCRH is used as a transfer count register and TCRL is used as a transfer count holding register. The size of the repeat area is specified by TCRH. Before the start of data transfer, the same value is set as the initial value for TCRH and TCRL.

  In the repeat mode, data is transferred once from the transfer source address indicated by SAR to the transfer destination address indicated by DAR by one activation. After the data transfer is finished, the SAR is incremented, decremented or fixed according to the SM [1: 0] bits. The DAR is incremented, decremented or fixed according to the DM [1: 0] bits. Thereafter, TCR (TCRH) is decremented. In the repeat mode, the data transfer and the register operation described above are repeated as many times as specified by the TCR (TCRH and TCRL) each time an activation factor is generated.

  When the designated number of times of data transfer is completed, TCRH becomes “0”. After that, based on the contents held in the TCRL, some or all of the initial setting values of the SAR, DAR, and TCRH are recovered.

  When TCRH becomes “0”, the contents of TCRL are transferred to TCRH of the transfer counter register. Thereby, TCRH is restored to the initial value. In addition, SAR or DAR is restored to the initial setting value by the following calculation.

When the DIR bit is set to “1”, the operation OP1 shown in Expression (1) is performed, and the SAR is restored to the initial setting value.

When the DIR bit is cleared to “0”, the operation OP2 shown in Expression (2) is performed, and the DAR is restored to the initial setting value.

  In order to perform the above calculation, the contents of the TCRL are input to the ALU 35 via the internal bus 36 for the SAR or DAR specified as the repeat area. The ALU 35 initializes the SAR or DAR by performing the calculation shown in the formula (1) or the formula (2).

Block Transfer Mode When the TMD [1: 0] bits are “10”, the transfer mode is the block transfer mode. When the DIR bit is cleared to “0”, the transfer source address is set as a block area. When the DIR bit is set to “1”, the transfer destination address is set as a block area. In the block transfer mode, SAR is used as a transfer source address specification register, DAR as a transfer destination address specification register, TCRH as a block size count register, TCRL as a block size holding register, and BTCR as a block transfer count register.

  In the block transfer mode, data corresponding to the block size is transferred from the transfer source address indicated by SAR to the transfer destination address indicated by DAR by one activation. For each data transfer, the SAR is incremented, decremented or fixed depending on the SM [1: 0] bits. The DAR is incremented, decremented or fixed according to the DM [1: 0] bits. Thereafter, TCR (TCRH) is decremented.

  When the designated number of times of data transfer is completed, TCRH becomes “0”. In this case, as in the repeat mode, some or all of the initial setting values of SAR, DAR, and TCRH are recovered based on the contents held in the TCRL.

  When the data transfer in the block area is completed, the BTCR is decremented. In the block transfer mode, the data transfer in the block area described above is repeated the number of times designated by the BTCR (until BTCR becomes “0”) every time an activation factor occurs.

  FIG. 4 shows a state transition diagram of DTC3. The state transition shown in FIG. 4 is mainly implemented in the data transfer control block (DTCCNT) 31 of DTC3.

  When the activation request signal DTCREQ is given from the INT2, the DTC3 transits to the VR state. After the transition to the VR state, the DTC 3 reads the start address of the data transfer information stored in the vector area (data transfer information start address arrangement area) based on the vector address generated by the VG 33 according to the vector number DTCVEC. The DTC 3 stores the read contents in the VAR 34.

  Next, DTC 3 transitions to the IR state. After the transition to the IR state, the DTC 3 reads the data transfer information stored in the data transfer information arrangement area according to the head address read in the VR state. At this time, the DTC 3 reads MR, DR, SAR1, SAR2, DAR, and CR.

  In the state where the TLU bit is cleared to “0” and the normal mode data transfer is performed, the DTC 3 transits to the SR state, the DW state, and the IW state.

  In the SR state, the DTC 3 reads the content of the transfer source address according to the content of the SAR2. The DTC 3 increments the SAR 2 according to the contents of the MR.

  In the DW state, the DTC 3 writes the read contents to the transfer destination address according to the contents of the DAR. The DTC 3 increments the DAR, decrements the CR, and the like according to the contents of the MR.

  In the IW state, DTC 3 writes MR, SAR1, SAR2, DAR, and CR back to the data transfer information allocation area according to the contents of VAR34. However, in the IW state, it is possible to prevent writing of data that has not been updated among MR, DR, SAR, DAR, and CR. For example, when the DAR is fixed, the DAR is not written.

  The DTC 3 repeats the above-described SR state and DW state operations as many times as specified by the TCR when performing data transfer in the block transfer mode in accordance with the contents of the MR.

  After that, when the chain operation is not performed, the DTC 3 clears the interrupt factor flag or the DTE bit that becomes the activation factor according to the contents of CR, ends the operation, and returns to the stopped state.

  When performing a chain operation, the DTC 3 returns from the IW state to the IR state and can perform another data transfer.

  On the other hand, the DTC 3 transitions from the IR state to the DR state when performing data transfer in the normal mode when the TLU bit is cleared to “1” (when referring to the table).

  In the DR state, the DTC 3 reads input data for referring to the table. After that, as described above, the state transits to the SR state, the DW state, and the IW state. In the SR state, a required table is referenced (read) based on the input data and the contents of SAR2. After that, it is the same as the normal data transfer mode.

  In the present embodiment, the operation is performed by selecting a branch mode among the table reference modes. In the branch mode, when the TLU bit is “1” and the BRM bit is “1”, the table reference mode is the branch mode. FIG. 5 shows data transfer in the branch mode.

  The CPU 1 writes required data transfer information in a predetermined transfer information address area. As table data, an offset of the head address of data transfer information to be executed following the operation according to the data transfer information is prepared. In this state, when an interrupt factor occurs and DTC 3 is activated, the data transfer information head address is read from the corresponding vector area (VR state).

  Based on the read address, the data transfer information INF11 is read from the data transfer information arrangement area (IR state). The data transfer information INF11 at this time is MR, SAR1, and SAR2.

  Data having a data size designated by ISz [1: 0] bits is read from the address designated by SAR1 (DR state). The data read here is referred to as determination data. The determination data is an offset of the start address of the data transfer information that is subsequently executed as a chain operation.

  The ALU 35 multiplies the read determination data by 1, 2, or 4 according to the transfer data size (table data size) specified by the Sz [1: 0] bit, or shifts it by 0, 1, or 2 bits. Then, an offset value of the table is generated. The shift amount can be specified by SF [2: 0] bits as necessary. Then, the offset value of the table is added to SAR2 and used as the transfer source address.

  Read transfer data from the transfer source address (SR state). Note that the read transfer data is not written (DW state).

  Thereafter, without updating the SAR2 and writing to the address of the original data transfer information INF11, the VAR corresponding to MR, SAR1, and SAR2 read in the IR state is incremented, and the processing by the data transfer information INF11 is completed. (IW state).

  Next, the content read from the table is multiplied by 16, or the result of 4-bit shift is added to VAR corresponding to the next address of the data transfer information INF11. Thereby, the head address of the data transfer information that is activated as a chain operation is generated. Note that multiplying the table by 16 corresponds to the fact that the data transfer information is 4 long words (ie, 16 bytes).

  Next, the data transfer information INF22 is read based on the top address of the data transfer information activated by the generated chain operation, and data transfer is performed.

  In the branch mode, for example, a command is input via a communication means such as a serial communication interface (SCI), and the DTC 3 is activated by a reception completion interrupt. By referring to the table using this command as determination data, data transfer information corresponding to the table reference result can be obtained. Subsequently, data transfer (chain operation) is performed based on the generated data transfer information. According to this, the data transfer information can be designated by the parameter (offset stored in the table) corresponding to the command. As a result, the transfer source address and the transfer destination address can be flexibly selected in accordance with the parameters.

  FIG. 6 shows an example of designation of data transfer information in the branch mode. In the branch mode, a necessary number of sets of data transfer information are prepared in advance in order to switch data transfer information used for data transfer in the chain operation according to the determination data.

  The data transfer information INF11 is specified by a vector. After completion of the first data transfer by the data transfer information INF11, the data transfer information INF11 is written to the address stored in the VAR, the address is incremented, and the next address storing the data transfer information INF11 is held in the VAR. .

  In the branch mode, since the data transfer information INF11 is not updated, the actual write is not performed, and only the address (VAR) is incremented within the DTC 3.

  The ALU 35 adds the VAR and the determination data multiplied by 16 to obtain the head address of the data transfer information INF22.

  Note that VAR may be added to the determination data without multiplying the determination data by 16. According to this, even when the data transfer information has a variable length, it can be handled by setting the determination data.

  The same data or data transfer information INF12 can be used for a plurality of commands or parameters. For example, for a plurality of commands to be entrusted to the CPU, the no operation mode is designated in the data transfer information INF12, the DTE bit is cleared to “0”, and an interrupt is requested to the CPU.

  For example, when “0” is set in the parameter corresponding to the status request command, the data transfer information INF12 in the area continuous to the data transfer information INF11 is transferred to the predetermined transmission data register of the predetermined status register. To set. A response to the command can be realized only by the DTC without requiring any CPU processing.

  As described above, according to this configuration, the control corresponding to the control can be realized by obtaining the output data corresponding to the input data with reference to the table. Further, by using the table, it is possible to minimize the dedicated hardware required for performing control corresponding to the control. In addition, by specifying valid bits of input data (logical product with DR and shift), the amount of information in the table can be reduced.

  Also, according to this configuration, a normal data transfer function (normal data transfer) function corresponding to an event and a data transfer function (table reference) function for referring to a table can be realized by a single piece of hardware. These functions can be switched by data transfer information. That is, it is sufficient to provide dedicated hardware as much as necessary for one data transfer. Therefore, by adding functions, the amount of data transfer information to be used increases, and as a result, even if the hardware scale increases, an increase in the physical scale of the entire semiconductor device (microcomputer) can be suppressed. In other words, the normal data transfer function and the data transfer function for referring to the table can be realized by a single piece of hardware, so that an increase in the physical scale of the semiconductor device (microcomputer) can be suppressed. Also, the hardware configuration can be simplified by sharing the interrupt controller, DTC vector, and the like.

  In this configuration, it is possible to specify that the CR is invalid. As a result, apart from performing a predetermined number of data transfers, control according to the state of the semiconductor device (microcomputer) can be executed any number of times. As a result, the data transfer information can be shortened, and the DTC operation can be speeded up and the memory utilization efficiency can be improved. Also, by disabling CR, the data transfer information is not updated, that is, it is not necessary to write back the data transfer information.

  According to this configuration, the DTC can execute data transfer and table reference instead of the CPU. As a result, the frequency of requesting an interrupt to the CPU can be reduced, and the period during which the CPU is in a low power consumption state can be extended. Further, it is possible to eliminate the exception processing, stack save / restore operation, and execution of the return instruction to be executed by the CPU in interrupt processing, which can contribute to the simplification of the program and the efficiency of the system. In addition, the time from the occurrence of an event to the execution of a required operation can be shortened, and so-called responsiveness can be improved. Furthermore, the DTC has a smaller logical scale than the CPU and can perform processing at high speed. Thereby, the power consumption of the semiconductor device (microcomputer) can be reduced by setting the CPU in the low power consumption state during the DTC operation.

Further, the data transfer device described in this embodiment is not particularly limited in usage, and can store data transfer information in a general-purpose storage device such as a RAM. Such a data transfer device has the following advantages.
(1) The number of data transfers can be increased. In a system having data transfer information as dedicated hardware of the data transfer apparatus itself, such as a so-called DMA (Direct Memory Access) controller, the number of data transfers is limited by the installed hardware. On the other hand, since the data transfer device stores the data transfer information in a RAM that is not limited to the intended use, it is easy to increase the number of data transfers, and it is possible to cope with various usage methods of the user.
(2) As in the so-called DMA controller, it is difficult to deal with many interrupts or events in a method in which the selection of the activation factor is specified by the control register of the data transfer device itself and the microcomputer interface. On the other hand, the data transfer device can select whether to request an interrupt from the CPU or to request a data transfer from the DTC using an interrupt factor. Thus, data transfer can be performed for many interrupts or event occurrences.
(3) It is possible to increase the number of data transfers to be executed at one activation such as a chain operation, and to realize a function by combining different data transfers.
(4) The configuration of the data transfer information can be changed. In addition, the data constituting the data transfer information can be increased or decreased.
(5) The dedicated hardware only needs to have the minimum necessary (one time) data transfer. Therefore, even if the function of the MCU is added and the hardware scale increases, an increase in the physical scale of the entire MCU can be suppressed.
(6) Since there is no control register or the like as dedicated hardware of the data transfer apparatus itself, it is not necessary to consider complicated operating conditions such as competition with writing from the CPU. Therefore, it can contribute to suppressing an increase in the physical scale of the MCU.

Embodiment 2
A microcomputer according to the second embodiment will be described. In the present embodiment, the operation of the MCU 100 when the table reference mode is the shift mode will be described. FIG. 7 shows data transfer in the shift mode. When the TLU bit is “1” and the SFT bit is “1”, the table reference mode is the shift mode.

  The CPU 1 writes required data transfer information in a predetermined transfer information address area. As table data, data to be written with the data transfer information (also referred to as output data), for example, a speed parameter is prepared. In this state, when an interrupt factor is generated and DTC 3 is activated, the data transfer information head address is read from the corresponding vector area (VR state).

  Based on the read address, the data transfer information INF21 is read from the data transfer information arrangement area (IR state). In the shift mode, the data transfer information INF21 is MR, SAR1, SAR2, and DAR.

  Data having a data size designated by ISz [1: 0] bits is read from the address designated by SAR1 (DR state). Here, the read data is referred to as shift data.

  The ALU 35 doubles the read shift data according to the transfer data size (table data size) specified by the Sz bits [1: 0], or shifts it by 0, 1, 2 bits. Perform the operation. The shift amount can be specified by SF [2: 0] bits as necessary. Then, the operation result is added to SAR2 and used as a transfer source address. DTC3 sets the addition result as the contents of SAR2.

  Transfer data is read from the transfer source address specified by SAR2 (SR state), and written to the transfer destination address specified by DAR (DW state). The transfer destination address is, for example, a timer.

  Thereafter, the data transfer information that has been updated, specifically, SAR2 is written to the original address (IW state).

  FIG. 8 shows an example of a method for specifying a table used in the shift mode. For example, assume that speed data is given to a timer that drives a motor in control of data that specifies the speed of the motor. The speed parameter is calculated in advance and stored in the ROM as a table. The address is designated by SAR2. This speed parameter is, for example, 16 bits, and the speed increases in ascending order.

  The word size is designated by Sz [1: 0] bits, the input data (shift data) is doubled and added to SAR2. In the initial setting stage, SAR2 designates parameter 0.

  When it becomes necessary to drive the motor, a predetermined speed parameter, for example, parameter m is transferred to the timer based on the shift data. If the motor status is monitored by another timer or analog input and it is judged that acceleration is sufficient, negative shift data, that is, a slower speed parameter (parameter n) is obtained and transferred to the timer. To do.

  In this example, for example, the state of the motor controlled by the microcomputer is input by an A / D converter (corresponding to the analog module 9). The DTC 3 can be activated using this input as a conversion completion interrupt. Then, the conversion result of the A / D converter can be used as shift data, and speed data (output data) can be given to a timer that drives the motor. Since the speed data corresponding to the motor state is calculated in advance and stored in the ROM as a table, it is not necessary to perform calculation in the actual operation state, and the table may be referred to.

  As described above, by using the shift mode, data transfer can be performed by feeding back the state of the microcomputer to be controlled.

Embodiment 3
A microcomputer according to the third embodiment will be described. In the present embodiment, the operation of the MCU 100 when the table reference mode is the offset mode will be described. FIG. 9 shows data transfer in the offset mode. When the TLU bit is “1” and the OFM bit is “1”, the table reference mode is the offset mode.

  First, as in the shift mode, the state transits to the VR state and the IR state. The data transfer information INF31 in the offset mode is MR, DR, SAR1, SAR2, and DAR.

  Data is read from the address designated by SAR1 (DR state). Here, the read data is referred to as offset data.

  In DR, mask data designating valid bits of the read offset data is held. For example, when the offset data is 32 bits and the upper 8 bits and the lower 4 bits are ignored, “00FFFFF0” is set as the mask data. Since the upper 8 bits are ignored, sign extension should be performed. Since the lower 4 bits are ignored, 4-bit right shift is performed once. Then, the left shift is performed by 0, 1, 2 bits according to the size specified by the Sz [1: 0] bits. The result after the end of the left shift is added to SAR2, and the addition result is used as the transfer source address. DTC3 sets the addition result as the contents of SAR2.

  Transfer data is read from the transfer source address specified by SAR2 (SR state), and written to the transfer destination address specified by DAR (DW state).

  Thereafter, the process is terminated (IW state) without updating the SAR2, that is, without writing the updated SAR2 in the data transfer information to the original address.

  FIG. 10 shows an example of a method for specifying a table used in the offset mode. The structure of the table can be the same as in the shift mode.

  SAR2 holds the reference address of the table. In FIG. 10, the center of the address area is the reference address (parameter m), but another address such as the head address may be used as the reference address. Output data (parameter n) is obtained by adding an offset based on the input data (offset data) to the reference address.

  By specifying valid bits of input data (logical product with DR), a necessary table can be reduced.

Embodiment 4
A microcomputer according to the fourth embodiment will be described. In the present embodiment, the operation of the MCU 100 when performing PID control using the table reference mode will be described. In the present embodiment, proportional mode (P), integral mode (I), and differential mode (D) are used as the table reference mode. Then, PID control is realized by adding the data transfer results in the proportional mode, the integral mode and the differential mode in the data transfer in the block addition mode which is one of the transfer modes. Hereinafter, data transfer in each table reference mode and block addition mode will be described.

Proportional mode The proportional mode is explained. FIG. 11 shows data transfer in the proportional mode. When the TLU bit is “1” and the PRM bit is “1”, the table reference mode is the proportional mode.

  First, as in the shift mode, the state transits to the VR state and the IR state. The data transfer information INF41 in the proportional mode is MR, DR, SAR1, SAR2, and DAR.

  Data is read from the address designated by SAR1 (DR state). Here, the read data is referred to as current value data.

  Target value data is held in DR. Then, a deviation between the current value data and the target value data is obtained. Then, the left shift is performed by 0, 1, 2 bits according to the size specified by the Sz [1: 0] bits. The result after the end of the left shift is added to SAR2, and the addition result is used as the transfer source address. DTC3 sets the addition result as the contents of SAR2.

  Thereafter, as in the offset mode, the state transits to the SR state and the DW state.

  Thereafter, the process is terminated (IW state) without updating the SAR2, that is, without writing the updated SAR2 in the data transfer information to the original address.

  The current value data can be obtained by the conversion result of the A / D converter (corresponding to the analog module 9) or the phase count of the timer. The DTC 3 can be started by a conversion end interrupt of the A / D converter, an interval timer (corresponding to the timer 7) interrupt, or the like.

  In the proportional mode, output data corresponding to the deviation between the current value of the controlled object that is input data and the target value can be obtained by referring to the table. For example, proportional control can be realized in which the output data is increased if the deviation is large and the output data is decreased if the deviation is small.

Differential Mode Next, the differential mode will be described. FIG. 12 shows data transfer in the differential mode. When the TLU bit is “1” and the DRM bit is “1”, the table reference mode is the differential mode.

  First, as in the shift mode, the state transits to the VR state and the IR state. The differential mode data transfer information INF51 is MR, DR, SAR1, SAR2, and DAR.

  Current value data is read from the address designated by SAR1 (DR state).

  The previous (old) current value data is held in the DR, and the new current value data read is compared with the previous current value data. This subtraction result corresponds to a differential value (speed). Then, the left shift is performed by 0, 1, 2 bits according to the size specified by the Sz [1: 0] bits. The result after the end of the left shift is added to SAR2, and the addition result is used as the transfer source address. DTC3 sets the addition result as the contents of SAR2. The contents of DR are replaced with the read new current value data.

  Thereafter, as in the offset mode, the state transits to the SR state and the DW state.

  After that, the updated data transfer information, specifically, the DR (the new current value data read) is written to the original address (IW state).

  According to the differential mode, output data corresponding to two changes (differential values) of the current value of the controlled object that is input data can be obtained by referring to the table. For example, differential control can be realized such that the output data is small if the change is large, and the output data is large if the change is small.

Integration Mode Next, the integration mode will be described. FIG. 13 shows data transfer in the integration mode. The TLU bit is set to “1”. In the integration mode, two data transfers combined by a chain operation are performed. The TMD [1: 0] bit is set to “01” (repeat mode), and the NXTE1 bit is set to “1” (chain operation).

  First, the first data transfer is performed. The data transfer information INF61 at this time is MR, DR, SAR1, DAR, and CR. At this time, the DRE bit is set to “1”.

  As in the shift mode, the state transits to the VR state and the IR state.

  Current value data is read from the address designated by SAR1 (SR state).

  Target value data is held in DR, and subtraction from the read current value data is performed. Then, the subtraction result is written to an address designated by the DAR (DW state). This writing is performed by the same operation as in the repeat mode. As a result, deviation data (deviation n to deviation (n + 3)) between the current value and the target value is accumulated for the number specified by the TCRH of CR (four in FIG. 13).

  Thereafter, the updated data transfer information, specifically, the DAR and CR are written to the original address (IW state).

  Next, the second data transfer is performed. The data transfer information INF62 at this time is MR, SAR1, SAR2, DAR, and CR. At this time, the INM bit is set to “1”.

  Data transfer information INF62 is read from the data transfer information arrangement area (IR state). The data transfer information INF62 is read from an address continuous to the data transfer information INF61.

  As in the block transfer mode, the deviation data between the current value and the target value is read from the address specified by SAR1 by the number of times specified by TCRL while incrementing or decrementing SAR1 and decrementing TCRH (DR state). . When TCRH becomes “0”, the contents of SAR1 and TCRH are returned to the initial set values. At this time, the data size is specified by ISz [1: 0] bits.

  The read deviation data is added inside the DTC 3, and division or right shift is performed according to the added number. The division or the right shift may be automatically performed according to the number of added data (contents of TCRL), or may be designated by SF [2: 0] bits. The result of division or right shifting is shifted left by 0, 1, 2 bits according to the size specified by Sz [1: 0] bits. The result of the left shift is added to SAR2 and used as the transfer source address.

  Transfer data is read from the transfer source address designated by SAR2 (SR state). Then, the read data is written to the transfer destination address designated by DAR (DW state).

  Thereafter, the process is terminated (IW state) without updating the SAR2, that is, without writing the updated SAR2 in the data transfer information to the original address.

  Note that the area that is written by the first data transfer by the data transfer information INF61 and is read by the second data transfer by the data transfer information INF62 is initialized before the operation is started.

  According to the integration mode, output data corresponding to the accumulation of a certain time of deviation between the current value of the controlled object and the target value as input data can be obtained by referring to the table. For example, it is possible to perform integration control such that the output data is increased if the accumulated value is large, and the output data is decreased if the accumulated value is small.

  Further, the subtraction with the target value data set in DR is not performed in the first data transfer by the data transfer information INF61 (DRE = 0), but is performed in the second data transfer by the data transfer information INF62 (DRE = 1). You may do it.

Block Addition Mode Next, the block addition mode, which is one of the data transfer modes, will be described. FIG. 14 shows data transfer when the transfer mode is the block addition mode. In the block addition mode, the TLU bit is cleared to “0” (normal data transfer function), and the TMD1 and TMD0 bits are set to “11” (block addition mode).

  When the DTC 3 is activated, the data transfer information head address is read (VR state), and the data transfer information is read from the data transfer information arrangement area (IR state). Data transfer information in the block addition mode is MR, SAR2, DAR, and CR.

  Similarly to the block transfer mode, SAR2 is incremented or decremented from the address designated by SAR2, and TCRH is decremented, and the number of times designated by TCRL (three times in FIG. 14) is read (SR state). When TCRH becomes 0, the contents of SAR2 and TCRH are returned to the initial set values. At this time, the data size is specified by Sz [1: 0] bits.

  The read data is added inside the DTC 3 and written to the transfer destination address designated by the DAR (DW state).

  Thereafter, the process is terminated (IW state) without updating the SAR2, that is, without writing the updated SAR2 in the data transfer information to the original address.

  The transfer destination addresses in the proportional mode, integral mode, and differential mode can be made continuous addresses in the RAM, which are superimposed in the block addition mode, and written to a required timer that performs motor control and the like. For example, the DTC is started by the interval timer, and the proportional mode, the integration mode, the differentiation mode, and the block addition mode are continuously executed in a chain operation (the NXTE1 bit other than the last block addition mode is set to 1). be able to. PID control corresponding to input data can be realized.

Embodiment 5
A microcomputer according to the fifth embodiment will be described. In the present embodiment, a modification of the branch mode according to the first embodiment will be described. FIG. 15 shows data transfer in a modification example of the branch mode. In this modification, the presence / absence of a chain operation is selected by the most significant bit (MSB) of the data prepared in the table.

  The data transfer information INF71 is MR, DR, SAR1, and SAR2. The DR is set to “00000000” in the initial setting or in the resetting process after the series of processes such as CR is “0”.

  Read shift data of the data size specified by the ISz [1: 0] bit from the address specified by SAR1 (DR state). A transfer source address is generated based on the read shift data and SAR2.

  Transfer data is read from the transfer source address designated by SAR2, the read data is stored in DR, and DR bit 31 is set to "1" as a status indicating that the table has been read (SR state). Note that transfer data is not written (DW state).

  The updated data transfer information, specifically, the DR is written to the original address (IW state).

  A case where the MSB of the read table data is set to “1” will be described. From the data stored in the DR, which is the read table data (ignoring bit 31), and the VAR, the head address of the data transfer information activated by the chain operation is generated as in the first embodiment. Data transfer information INF72 corresponding to the generated head address is read, and data transfer is performed based on this.

  Next, a case where the MSB of the read table data is cleared to “0” will be described. In this case, the process is terminated without performing the chain operation.

  Next, when DTC3 is activated by the same activation factor and data transfer information INF71 is read, the TLU bit and the SFM bit are set to “1”, and the DR bit 31 is set to “1”. If not, the data transfer is not performed, and the chain operation is started based on the data stored in the DR and the VAR as described above. Similarly to the above, the data transfer information INF72 is read, and based on this, data transfer is performed.

  For example, when a command is input via a communication means such as a serial communication interface (SCI), a DTC is activated by a reception completion interrupt, and when this is referred to as a determination data, a data transfer corresponding to the command is immediately executed. In some cases, it may be necessary to wait for the next reception completion interrupt.

  If the data transfer corresponding to the command can be executed immediately, the MSB of the table data can be set to “1” and the data transfer corresponding to the command can be executed by the chain operation.

  When it is necessary to wait for the next reception completion interrupt, the MSB of the table data is cleared to “0”, and the table data is stored in the DR at the first DTC activation, and the chain operation, that is, the data transfer is not performed. Exit once. When the DTC 3 is activated for the second time, the data transfer information INF31 is read, and the TLU bit, SFM bit, and DR bit 31 are all set to “1”. Next data transfer information INF32 is read, and based on this, data transfer is performed.

  When the DTC is activated for the second time, the same operation can be performed regardless of the MSB of the table data.

  In addition to storing the read table data, the next data transfer information start address of the chain operation may be stored in the DR. The status indicating that the table has been read may use DR as well as an appropriate bit of MR.

FIG. 16 shows a state transition diagram of DTC in a modification of the branch mode.
If the TLU bit, the SFM bit, and the DR bit 31 are all set to “1” in the IR state, the state transits to the IR state again, and the next data transfer information is read.

  As described above, according to this configuration, when the table reference mode is the branch mode, it is possible to determine whether or not to execute the chain operation based on the information included in the read data.

Embodiment 6
A microcomputer according to the sixth embodiment will be described. In the present embodiment, the operation of the MCU 100 when the table reference mode is the address calculation mode will be described. FIG. 17 shows data transfer in the address calculation mode. When the TLU bit is “1” and the AEM bit is “1”, the table reference mode is the address calculation mode.

  When the DTC 3 is activated, the data transfer information head address is read from the corresponding vector area (VR state).

  Based on the read address, the data transfer information INF81 is read from the data transfer information arrangement area (IR state). The data transfer information INF81 at this time is MR, SAR1, SAR2, DAR, and CR.

  Data having a data size designated by ISz [1: 0] bits is read from the address designated by SAR1 (DR state). Here, the read data is referred to as address calculation data.

  The read address calculation data is calculated by the ALU 35 by 1, 2, 4 times or shifting by 0, 1, 2 bits according to the transfer data size specified by the Sz [1: 0] bit. The calculation result is temporarily held inside the DTC 3 (for example, DR).

  Transfer data is read from the transfer source address specified by SAR2 (SR state), and written to the transfer destination address specified by DAR (DW state). Then, the CR is decremented.

  When the DIR bit is “0”, the shifted address operation data is added to or subtracted from SAR2 instead of the increment or decrement specified by the SM [1: 0] bit, and the result is stored in SAR2.

  When the DIR bit is “1”, the shifted address operation data is added to or subtracted from the DAR instead of the increment or decrement specified by the DM [1: 0] bit, and the result is stored in the DAR.

  After that, the updated data transfer information INF81, specifically, SAR2 or DAR and CR is written to the original address (IW state).

  The above data transfer is repeated as many times as specified by CR (TCRH).

  FIG. 18 shows an example of how to specify a table used in the address calculation mode. When data Di is transferred in the i-th data transfer, the parameter Pi is read from the address calculation data (table). Based on the parameter Pi, an operation is performed on the transfer source address (SAR2) or the transfer destination address (DAR) after the data Di is transferred.

  When the data transfer of the data Di + 1 is performed in the next i + 1-th data transfer, the addresses of the data Di and the data Di + 1 are not continuous, and an interval is inserted based on the parameter Pi.

  This interval can be specified by a parameter included in the address calculation data (table) for each data transfer. Therefore, the DTC 3 can continuously transfer data arranged at irregular intervals by referring to the table.

Embodiment 7
A microcomputer according to the seventh embodiment will be described. In the present embodiment, a specific example of INT2 of MCU 100 will be described. FIG. 19 shows a block diagram of the interrupt controller (INT) 2.

  The INT 2 includes an interrupt / DTC determination circuit 51, a DTC permission register (also referred to as DTER) 52, a priority determination circuit 53, a latch circuit 54, and a decoder circuit 55.

  There are two types of interrupt factors of the MCU 100, an internal interrupt and an external interrupt, each having an interrupt factor flag. The interrupt factor flag is set to “1” when the timer, communication, and analog function blocks enter a predetermined state. For example, the external interrupt factor flag is set to “1” when the external interrupt input terminal becomes a predetermined level or when a predetermined signal change occurs. The interrupt factor flag is cleared to “0” by the write operation of the CPU 1 and is also cleared to “0” when the data transfer by the DTC 3 is completed.

  The output of each bit of the interrupt factor flag is input to the interrupt permission circuit. The interrupt permission circuit further receives the contents of the interrupt permission register, that is, the interrupt permission bit. The interrupt permission register is a register readable / writable by the CPU 1 and selects whether to permit or prohibit the corresponding interrupt.

  If the interrupt factor flag is set to “1” and the interrupt permission bit is set to “1”, an interrupt is requested. The output of the interrupt permission circuit, which is an interrupt request, is input to the interrupt / DTC determination circuit 51.

  Further, the contents of the DTC permission register (also referred to as DTER) 52 are input to the interrupt / DTC determination circuit 51. The interrupt / DTC determination circuit 51 selects whether to activate the DTC 3 or to allow the CPU 1 to interrupt when an interrupt is requested.

  When the bit corresponding to the interrupt factor of the DTC permission register 52 is set to “1”, the interrupt / DTC determination circuit 51 requests activation of the DTC 3 and does not request an interrupt from the CPU 1. If the bit corresponding to the interrupt factor in the DTC permission register 52 is cleared to “0”, the interrupt / DTC determination circuit 51 requests an interrupt from the CPU 1 and does not request activation of the DTC 3. The interrupt / DTC determination circuit 51 outputs an activation request to the priority order determination circuit 53.

  The priority determination circuit 53 determines the priority for each of the interrupt request from the CPU 1 and the activation request from the DTC 3 when a plurality of interrupt requests are generated. When determining the priority order of the activation request of the CPU 1, the mask level can be determined together. The determination of the priority of the activation request of the CPU 1 is controlled according to the priority register, the interrupt mask level, and the like.

  The priority determination circuit 53 selects the activation request having the highest priority and generates a vector number. When the activation request for the CPU 1 is selected, the priority determination circuit 53 generates an interrupt request CPUINT and a vector number CPUVEC for the CPU 1. When the activation request for DTC3 is selected, the activation request signal DTCREQ for DTC3 and the vector number DTCVEC are generated. The activation request signal DTCREQ is input to DTC3. The vector number DTCVEC is input to the latch circuit 54.

  The latch circuit 54 receives a DTC operation start signal and a DTC operation end signal from the DTC 3. That is, when the DTC 3 starts operation, the DTC operation signal becomes active, and the latch circuit 54 latches or holds the vector number DTCVEC. When the data transfer of DTC3 is completed and the DTC operation end signal is activated, the latch by the latch circuit 54 is released.

  The vector number DTCVEC and the DTC operation end signal are input to the decoder circuit 55. As a result, the factor clear signal for the corresponding interrupt factor flag is activated, and the corresponding interrupt factor flag or DTE bit is cleared.

  When the DTC 3 is activated by a required interrupt factor, the CPU 1 writes data transfer information and the like in advance to a required address, and also sets an interrupt permission bit corresponding to the interrupt factor and a DTE corresponding to the interrupt factor of the DTC permission register 52. The bit is set to “1”.

  When the interrupt factor flag is set to “1”, the DTC 3 is activated. In a state where the DTC 3 is executing a predetermined data transfer, the DTC 3 clears the interrupt factor flag to “0” for each data transfer. At this time, no interruption is requested to the CPU 1.

  When the predetermined data transfer is completed, the DTC 3 clears the DTE bit to “0” after the data transfer. At this time, since the interrupt factor flag is held at “1” and the DTE bit is cleared to “0”, the CPU 1 is requested to interrupt. The CPU 1 executes processing corresponding to the end of predetermined data transfer, and resets the data transfer information and the DTE bit.

Embodiment 8
A microcomputer according to the eighth embodiment will be described. In this embodiment, an example in which the above-described MCU is applied to a camera system that controls a camera will be described. FIG. 20 shows a block diagram of a camera system 1000 including an MCU.

  The camera system 1000 includes a camera body MCU1001, a camera lens MCU1002, a focus motor 1003, and an encoder 1004 that detects the position of the focus motor. Among these, for the camera lens MCU 1002, any of the MCUs described in the first to seventh embodiments is used.

  The camera lens MCU 1002 includes a focus motor driving timer F (also referred to as TF), an encoder input timer D (also referred to as TD), and a timer I having an interval timer function (also referred to as TI). These timer F, timer D, and timer I correspond to the timer 7, respectively. The camera lens MCU 1002 includes a ROM 4 (not shown) that stores a table for the focus motor. Also, switches such as auto-focus permission and anti-shake permission are input via the input / output port (I / O) 10.

  The camera lens MCU 1002 receives commands and associated data from the camera body MCU 1001 via one line communication means such as a serial communication interface (SCI, corresponding to the communication module 8). The DTC 3 is activated by this reception completion interrupt.

  By setting the DTC3 table reference mode to the branch mode, the command input from the communication means is analyzed, and the required data transfer is performed based on the command.

  A case where an input command (for example, a status request command) requests a return of a predetermined state of the lens (for example, the state of a camera shake correction permission switch) will be described. In this case, data transmission is performed by transferring a predetermined functional block (for example, the contents of the input / output port (I / O)) of the camera lens MCU 1001 to the SCI transmission data register by the data transfer information corresponding to the branch mode. It can be performed. Then, by performing a chain operation on the command, the CPU processing is unnecessary and the required operation can be completed immediately.

  If the input command is a focusing command, the target position is continuously received. When the next reception completion interrupt occurs, the target position is set to the proportional mode DR and the timer I starts the DTC. Then, the required focus motor drive can be performed with reference to the table.

  In performing the above operation, the CPU 1 can be in a low power consumption state such as a sleep or standby mode. This can contribute to lower power consumption of the system. Since the DTC 3 has a smaller logical scale than the CPU and can perform processing at high speed, the effect of reducing power consumption can be enhanced. In a camera system driven by a battery, low power consumption is particularly important.

  In the case of a zoom lens, it may be necessary to change the focus motor drive control depending on the zoom position. In this case, a plurality of tables are prepared, and a table to be used is received from the camera body MCU 1001 as a focusing instruction, a target position, and the like. Then, SAR2 can be set to an address corresponding to the table to be used.

  Note that the present invention is not limited to the above-described embodiment, and can be changed as appropriate without departing from the spirit of the present invention. For example, although the data transfer information is in 32-bit units, the data transfer information can be in arbitrary bit units. The arrangement of the data transfer information can be arbitrarily changed. Specifically, the number of bits of the address register is not limited to 32 bits, and can be changed according to the address space of the CPU or microcomputer. For example, in the case of a 16 Mbyte address space, it may be 24 bits. The bit arrangement of MR may be divided and combined with 24-bit SAR and DAR. The MR bit arrangement may be arbitrary, and the SFM bit, OFM bit, PRM bit, DRM bit, INM bit, BRM bit, NOP bit, etc. may be encoded.

  In the above-described embodiment, SAR2 as hardware is used as the transfer source address register in the data transfer information during normal data transfer. However, this may be SAR1. The arrangement other than the transfer mode information included in the data transfer information can be arbitrarily changed. The transfer source address information is desirably arranged to be readable before the transfer destination address information.

  The program that operates on the CPU may be arranged in the external memory in addition to the ROM. Similarly, the memory that is the work area of the CPU is not limited to the RAM, and may be an external memory. Further, the microcomputer may not include one or both of the ROM and the RAM.

  DTC data transfer information, input data blocks, and coefficient data blocks may be stored in an external memory, although they are advantageously stored in a RAM built in the microcomputer in terms of processing speed and power consumption.

  In the above-described embodiment, the data transfer information is described as being arranged in a memory such as a RAM. However, as in a so-called DMA controller, the data transfer information may be held in the functional block as an internal I / O register and arranged in the address space.

  In the above-described embodiment, the description has been given based on a method of storing data transfer information in a storage device such as a RAM. However, although the data transfer information that can be used is limited by the installed hardware, the data transfer described in the above embodiment can be applied to a so-called DMA controller.

  In addition to the DTC, another data transfer device such as a DMA controller may be provided. The functions of the DTC and DMA controller can be combined into one function module.

  Specific circuit configurations of the DTC and the interrupt controller can be changed to various circuit configurations having functions equivalent to the functions described in the above embodiments. Details of bus operations such as BSC and buses and waits have been omitted, but these can be implemented as appropriate. The configuration of the microcomputer is also merely an example and can be changed as appropriate.

  Although the microcomputer has been described in the above embodiment, this is merely an example. The DTC described in the above-described embodiment can be applied to various devices such as a data processing device and a semiconductor integrated circuit incorporating an arithmetic unit independent of the data processing device. For example, the present invention can be applied to a semiconductor integrated circuit device centering on a digital signal processor (DSP). The present invention can be applied to at least a semiconductor integrated circuit device including a data processing device and a data transfer device.

  As mentioned above, the invention made by the present inventor has been specifically described based on the embodiments. However, the present invention is not limited to the embodiments already described, and various modifications can be made without departing from the scope of the invention. It goes without saying that it is possible.

1 Central processing unit (CPU)
2 Interrupt controller (INT)
3 Data transfer device (DTC)
4 ROM
5 RAM
6 Bus controller (BSC)
7 Timer 8 Communication module 9 Analog module 10 Input / output port (I / O)
11 Internal bus 31 DTCCNT
32 BIF
33 VG
34 VAR
35 ALU
36 Internal bus 51 Interrupt / DTC determination circuit 52 Interrupt enable register 53 Priority determination circuit 54 Latch circuit 55 Decoder circuit 100 Microcomputer (MCU)
1000 Camera system 1001 Camera body MCU
1002 Camera lens MCU
1003 Vibration isolation motor 1004 Sensor ADC A / D converter CPUVEC Vector number DTCREQ Activation request signal DTCVEC Vector number INF11, INF12, INF21, INF31, INF41, INF51, INF61, INF62, INF71, INF81, INF82, INF91 Data transfer Count register DAR Destination register DR Data register MR Mode register SAR1, SAR2 Source address register DTCREQ Data transfer request signal DTCVEC Vector number

Claims (5)

A semiconductor device having a data controller,
The data controller includes a first transfer source address register, a second transfer source address register, a transfer destination address register, and a transfer number register.
Reading first data from a first transfer source address designated by the first transfer source address register;
Generating second data by computing the first data;
Using the second transfer source address indicated by the second transfer source address register as a transfer source address, data is transferred to the transfer destination address specified by the transfer destination address register,
A new second transfer source address or a new transfer destination address is generated by adding or subtracting the second data to one of the second transfer source address and the transfer destination address,
The data transfer is repeated as many times as the value specified in the transfer count register.
Semiconductor device. The result of performing either one or both of a logical operation and a shift operation according to the data size for the first data is the second data.
The semiconductor device according to claim 1. The data controller is
Based on the data transfer information address specified in response to the activation request, data transfer information is read from the data transfer information arrangement area, and the first transfer source address and the second transfer source address included in the data transfer information And the transfer destination address are stored in the first transfer source address register, the second transfer source address register, and the transfer destination address register, respectively.
The semiconductor device according to claim 1. The data transfer information further includes mode information;
The data controller further includes a mode register for storing the mode information,
Based on the information in the mode register, the second data is added to or subtracted from either the second transfer source address or the transfer destination address.
The semiconductor device according to claim 3. The data transfer information is updated by the new second transfer source address or the new transfer destination address.
The semiconductor device according to claim 4.

Images