1. Field of the Invention
The present invention relates to a functional device such as an FPGA (Field Programmable Gate Arrays) which carries out a necessary function by programming. In particular, the present invention relates to a functional device, a function maintaining method and a function maintaining program which overcome a fault such as a software error and can achieve maintenance of a continuous function.
2. Description of the Related Art
As is well known, the FPGA is an integrated circuit device in which circuit information inside the device can be written electrically by a program given from an outside, and which carries out a necessary circuit function by that circuit information.
Such an FPGA has a RAM (Random-Access Memory) for the purpose of maintaining the circuit information, and, for this RAM, an SRAM (Static RAM) is used. In the SRAM, the quantity of an individual electric charge is becoming small with the low voltage of a supply voltage and the minuteness of an integrated circuit in recent years. Because of this, there is some fear that a software error occurs. If the circuit information which has been written is changed by some electrical cause such as a software error, abnormality occurs in circuit behavior of the FPGA. In order to make a normal function recover, it is necessary to stop a system and to correct memory contents of the SRAM provided inside the FPGA. That is, it is necessary to rewrite the circuit information.
By the way, as patent documents which disclose in regard to the FPGA and the recovery of its function, the Japanese Laid Open Publications No. H09-62528 and No. H08-44581 are in existence.
As described above, in case in which the circuit information written in the FPGA is changed by some electrical cause such as a software error, the correction of the memory contents of the SRAM or the rewriting of the SRAM becomes necessary in order to recover a function of the FPGA. However, in such processing, a temporary stoppage and a restart of the system must be performed, for example. This is troublesome.
By the way, in the publication No. H09-62528, enabling self-recovery during service or during use after forwarding a functional device is disclosed. That is, as a functional device, for example, functional components and packaged devices are provided over a plurality of systems in parallel. One system out of those systems is used, the other system is made into a backup system at the time of a fault, and the backup system executes a function as proxy when a fault occurs in the system under use. Further, the publication No. H09-62528 discloses that an integrated circuit device is used as the functional device, and an FPGA is exemplified as the integrated circuit device. In this FPGA, the detection and decision of fault occurrence are disclosed, and, further, the recovery of a function executed based on that is also disclosed.
However, in the self-recovering device disclosed in the publication No. H09-62528, a functional component or a packaged device which presents a functional fault can not carry out self-recovery in essence. Because of this, when the backup system is used, the system which has presented a fault can not be used as a backup system. Further, in case in which only one backup system is provided, the state of non-redundant operation without a backup system is caused at the time of the occurrence of a fault. Furthermore, also in case in which a plurality of backup systems are in existence, the backup systems are reduced in number if a fault occurs. It is disclosed that prompt repair is needed in order to avoid such a state, and therefore imperfection of providing the backup system is indicated.
In addition, this self-recovering device has a functional circuit part, a fault detection part, an internal memory, a definition processing part and so on. The internal memory stores a plurality of pieces of layout pattern definition information which defines a function for the purpose of realizing a function of the functional circuit part. Based on the discriminated result of a fault area from the fault detection part, the definition processing part re-defines the functional circuit part in accordance with definition information not including a divided area, in which a fault is generated, from the layout pattern definition information stored in the internal memory. By such a constitution, the self-recovering device realizes a functional recovery by self-restoration and nothing more.
On the other hand, the publication No. H08-44581 discloses on an information processing device with a self-repairing function. This publication only discloses that the information processing device is constituted by providing a plurality of FPGAs and thereby the FPGAs are re-constituted based on the occurrence of a fault.
The present invention relates to a functional device such as an FPGA carrying out a necessary function by programming, and an object of the present invention is to provide a functional device, a function maintaining method and a function maintaining program which can maintain a function continuously.
Another object of the present invention is to prevent the stop of behavior against the occurrence of a failure such as a software error.
Still another object of the present invention is to recover a function against the occurrence of a failure such as a software error as a normal function is continued.
In order to attain the above objects, a first featured matter of a functional device according to the present invention is that a plurality of function parts are provided and a function is maintained by switching from a function part, in which a failure is occurring, to a function part under stand-by. That is, the plurality of function parts, a failure detection part and a switching part are provided. The function parts are a constitution which carries out a necessary function by circuit information written therein. By a function part being formed by the plurality of function parts, a function part under operation and a function part under stand-by can be set. The failure detection part detects the occurrence of a failure of the function part under operation out of the function parts. Then, the switching part switches from the function part, in which a failure is occurring, to the function part under stand-by based on the detection of the failure by the failure detection part. That is, the function part in which the failure is occurring is made to stand by, and the function part which is in a stand-by state is made to operate. Therefore, it is possible to continuously maintain the function without stopping the operation of a system due to the failure occurring in the function part. Because of this, the reliability of the system is heightened by the maintenance of the function.
In order to attain the above objects, a second featured matter of a function device according to the present invention is to make a function of the function part with the failure recover. That is, giving a constitution which has a function recovery part recovering the function of the above-mentioned function part with the failure, it is possible to make the above-mentioned function part with the failure recover by the function recovery part and to make this function part providing a normal function stand by for next operation, by removing the function part with the failure from an active system. This enables the function device to maintain a continuous and normal function.
In order to attain the above objects, a third featured matter of a functional device according to the present invention has a plurality of function parts, and is to maintain a function by periodically removing a function part under operation from an active system and by making a function part under stand-by return to the active system. That is, the plurality of function parts, a switching timing detection part and a switching part are provided. As described before, the function parts are a constitution which carries out a necessary function by circuit information written therein, and by a function part being formed by the plurality of function parts, a function part under operating and a function part under stand-by can be set. The switching timing detection part detects switching timing for the purpose of switching from the function part under operation to the function part under stand-by out of these function parts. Further, based on the above-mentioned switching timing which is detected by the switching timing detection part, the switching part makes the function part under operation stand by, and makes the above-mentioned function part under stand-by return to the active system. Therefore, since the operation of a system is executed by switching the plurality of function parts without stopping the operation of the system, a continuous function can be maintained. By this, the improvement of the reliability of the system is given as compared with the continuous use of a single function part.
In order to attain the above objects, the functional device of the present invention may also be constituted so that the functional device has a rewriting part which rewrites the above-mentioned circuit information into the function part under stand-by.
In order to attain the above objects, a functional device of the present invention may also be constituted so that the functional device has a plurality of function parts, a failure detection part, a switching timing detection part and a switching part. The function parts carry out a necessary function by circuit information written therein. The failure detection part detects a failure of a function part under operation out of these function parts. The switching timing detection part detects switching timing for the/purpose of switching from the function part under operation to a function part under stand-by out of the above-mentioned function parts. The switching part makes the above-mentioned function part under operation stand by and makes a function part under stand-by operate in case in which the above-mentioned switching timing comes, and switches from the function part with the failure to a function part under stand-by by detection of the failure.
In order to attain the above objects, a functional device of the present invention may also be constituted so that the functional device has a plurality of function parts, an input data maintaining part, an input switching part, an output switching part and an output data maintaining part. The plurality of function parts carry out a necessary function by circuit information written therein. The input data maintaining part maintains an input data which is applied to the above-mentioned function parts. The input switching part switches input of the above-mentioned input data, which is maintained in the input data maintaining part, toward the above-mentioned function parts. The output switching part switches taking-out of an output data which is taken out from the above-mentioned function parts. The output data maintaining part maintains the above-mentioned output data which is taken out from the function parts.
In order to attain the above objects, the functional device of the present invention may also be constituted so that a failure detection part which detects a failure of the above-mentioned function part under operation is provided, and so that either the above-mentioned input switching part or the above-mentioned output switching part is switched or both of them are switched based on detection of the failure of the failure detection part.
Further, in order to attain the above objects, a function maintaining method of the present invention is a constitution which includes processing that selects at least one function part from a plurality of function parts carrying out a necessary function by circuit information written therein and makes it operate, processing that detects a failure of the above-mentioned function part under operation, and processing that switches from the function part, in which the failure is occurring, to a function part under stand-by base on detection of the above-mentioned failure. This function maintaining method corresponds to the first featured matter of the functional device of the present invention described before. Also by such processing, it is possible to continuously maintain the function without stopping the operation of a system due to the failure occurring in a function part. Because of this, the reliability of the system is heightened by the maintenance of the function.
In order to attain the above objects, the function maintaining method of the present invention may also be constituted so that the method includes processing that recovers a function of the above-mentioned function part in which the failure is occurring.
In order to attain the above objects, a function maintaining method of the present invention may also be constituted so that the method includes processing that selects at least one function part from a plurality of function parts carrying out a necessary function by circuit information written therein and makes it operate, processing that detects switching timing for the purpose of switching from the function part under operation to a function part under stand-by out of these function parts, and processing that makes the above-mentioned function part under operation stand by and makes a function part under stand-by operate based on the above-mentioned switching timing.
In order to attain the above objects, a function maintaining method of the present invention may also be constituted so that the method has processing that selects at least one function part from a plurality of function parts carrying out a necessary function by circuit information written therein and makes it operate, processing that detects a failure of the above-mentioned function part under operation, processing that detects switching timing for the purpose of switching from the above-mentioned function part under operation to a function part under stand-by, and processing that makes the above-mentioned function part under operation stand by and makes a function part under stand-by operate in case in which the above-mentioned switching timing comes, and switches from the function part, in which the failure is occurring, to a function part under stand-by based on detection of the above-mentioned failure.
In order to attain the above objects, a function maintaining method of the present invention may also be constituted so that the method includes processing that selects at least one function part from a plurality of function parts carrying out a necessary function by circuit information written therein and makes it operate, processing that maintains an input data which is given to the above-mentioned function part, and processing that maintains an output data which is taken out from the above-mentioned function part.
Further, in order to attain the above objects, a function maintaining program of the present invention is a program which makes a computer execute the following steps, and is a constitution which includes a step that selects at least one function part from a plurality of function parts carrying out a necessary function by circuit information written therein and makes it operate, a step that takes in information representative of a failure which occurs in the above-mentioned function part under operation, and a step that switches from the function part, in which the failure is occurring, to a function part under stand-by. Also by such a program, it is possible to continuously maintain the function without stopping the operation of a system due to the failure occurring in a function part. Because of this, the reliability of the system is heightened by the maintenance of the function.
In order to attain the above objects, the function maintaining program of the present invention may also be constituted so that the program includes a step that makes a function of the above-mentioned function part with the failure recover.
In order to attain the above objects, a function maintaining program of the present invention is a program which makes a computer execute the following steps, and may also be constituted so that the program includes a step that selects at least one function part from a plurality of function parts carrying out a necessary function by circuit information written therein and makes it operate, a step that takes in switching timing for the purpose of switching from the function part under operation to a function part under stand-by out of these function parts, and a step that makes the above-mentioned function part under operation stand by and makes a function part under stand-by operate based on the above-mentioned switching timing.
In order to attain the above objects, a function maintaining program of the present invention is a program which makes a computer execute the following steps, and may also be constituted so that the program includes a step that selects at least one function part from a plurality of function parts carrying out a necessary function by circuit information written therein and makes it operate, a step that takes in information representative of a failure which occurs in the above-mentioned function part under operation, a step that detects switching timing for the purpose of switching from the above-mentioned function part under operation to a function part under stand-by, and a step that makes the above-mentioned function part under operation stand by and makes a function part under stand-by operate in case in which the above-mentioned switching timing comes, and switches from the function part, in which the failure is occurring, to a function part under stand-by based on detection of the above-mentioned failure.
As described above, in the present invention, at least one out of a plurality of function parts which carry out a necessary circuit function by rewriting circuit information is made into an active system, and the other or the others are made into a stand-by system. Then, the function parts are switched to an active system or a stand-by system periodically, or, in case in which a failure occurs in a function part of the active system, a function part of the stand-by system is made into the active system. Further, the function of a function part which is shifted to the stand-by system is guaranteed or recovered by the execution of configuration processing. Because of this, the continuous maintenance of a function becomes possible. By this, the present invention can contribute to the improvement of the reliability of a system, and is useful.
Furthermore, enumerating the featured matters and advantages of the present invention, these are as in the following.
(1) A plurality of function parts which execute a necessary circuit function by writing of circuit information are provided, and switching from a function part with a failure to a function part under stand-by is performed. Because of this, the present invention can continuously maintain the function without stopping a system due to the occurrence of a failure, and the reliability of the system can be improved.
(2) Giving a constitution which makes the function of a function part with a failure recover, it is possible to maintain a function part under stand-by in a normal state. By this, the continuation of functional maintenance is improved, and the reliability of a system can be heightened.
(3) The present invention has a plurality of function parts which carry out a necessary circuit function by writing of circuit information, and switches periodically from a function part under operation to a function part under stand-by. Because of this, it is possible to continuously maintain a function, and it is possible to improve the reliability of a system.
The foregoing and other objects, features and attendant advantages of the present invention will be appreciated as the same become better understood by means of the following description and accompanying drawings wherein:
A first embodiment of the present invention is explained by referring to
In this functional device 2, as a plurality of function parts which carry out a necessary circuit function by programming, a first FPGA (Field Programmable Gate Arrays) 40 and a second FPGA (Field Programmable Gate Arrays) 41 are provided. Here, the FPGAs 40 and 41 are a device able to rewrite circuit information electrically, and are a programmable integrated circuit device which carries out a necessary circuit function by the circuit information written therein. That is, each of the FPGAs 40 and 41 has a RAM (Random-Access Memory) 5 as a function storage part able to write the circuit information, for example. Further, similarly to the FPGAs 40 and 41, the functional device 2 serving as a system including the FPGAs 40 and 41 is also constituted as an integrated circuit device. That is, this functional device 2 is a doubled constitution of FPGAs, which has the two FPGAs 40 and 41. By making one of these FPGAs 40 and 41 operate and making the other stand by, periodically, and by making an FPGA under stand-by operate instead of an FPGA under operation in case in which a failure occurs in the FPGA under operation, the functional device 2 realizes a continuous function. In this embodiment, for the sake of convenience of explanation, the FPGA 40 is represented as a 0 (zero) system, and the FPGA 41 is represented as a 1 (one) system. An FPGA which is in a relation of co-operation with a system including the functional device 2 is represented as an active (ACT) system or as “under operation”, and an FPGA which is out of the relation of co-operation and is stood by is represented as a stand-by (SBY) system or as “under stand-by”.
Therefore, this functional device 2 has a constitution which switches alternately periodically one of the FPGAs 40 and 41 to the active system and the other to the stand-by system, and a constitution which continues a function by switching the active system with a failure to a stand-by state and by switching the FPGA 40 or the FPGA 41 under stand-by to the active system in case in which the failure occurs in the FPGA 40 or the FPGA 41 of the active system.
As a constitution which switches periodically one of the FPGAs 40 and 41 to effectiveness and the other to invalidity, a periodic switching circuit 6 serving as a switching timing detection part which detects switching timing of the FPGAs 40 and 41 is provided. This periodic switching circuit 6 receives a clock signal CK from a timer circuit 8, detects the switching timing by timer operation, and generates a switching signal SW. As for the switching signal SW, since the FPGAs 40 and 41 are alternately switched, the periodic switching circuit 6 generates a switching signal SW01 in case of switching from the FPGA 40 under operation to the FPGA 41 under stand-by (switching of 0→1 system), and generates a switching signal SW10 in case of switching from the FPGA 41 under operation to the FPGA 40 under stand-by (switching of 1→0 system)
These switching signals SW01 and SW10 are given to a CPU (Central Processing Unit) 12 of a control part 10, the FPGAs 40 and 41 and switching circuits 14 and 16, and are representative of generation of the switching timing. Along with this, the switching signals SW01 and SW10 act as a reset signal to the FPGAs 40 and 41, and act as a switching signal to the switching circuits 14 and 16. The control part 10 has the CPU 12, a ROM (Read-Only Memory) 18, a RAM (Random-Access Memory) 20 and so on, and is a computer which executes information processing. The control part 10 controls sequential operation of all systems of the functional device 2. In this case, taking-in of the switching signals SW01 and SW10 taken in the CPU 12 and switching states of the switching circuits 14 and 16 are stored in the RAM 20. In the ROM 18, a program for the purpose of executing information processing like this is stored.
Further, the switching circuit 14 is provided at a data input side of the FPGAs 40 and 41, and the switching circuit 16 is provided at a data output side of these. The switching circuits 14 and 16 constitute first and second switching parts which switch one of the FPGAs 40 and 41 to the active system and switch the other to the stand-by system. The switching circuit 14 switches an input data Din to the FPGA 40 or the FPGA 41 with arrival of the switching signal SW01 or SW10 as a trigger. For example, if the switching signal SW01 arrives, the FPGA 40 of the active system comes into the stand-by system, and the FPGA 41 of the stand-by system is to be switched to the active system. On the other hand, if the switching signal SW10 arrives, the FPGA 41 of the active system comes into the stand-by system, and the FPGA 40 of the stand-by system is to be switched to the active system. Further, the switching circuit 16 switches the taking-out of an output data Dout with arrival of the switching signal SW01 or SW10 as a trigger. For example, if the switching signal SW01 arrives, the FPGA 40 of the active system comes into the stand-by system, and the FPGA 41 of the stand-by system is to be switched to the active system. On the other hand, if the switching signal SW10 arrives, the FPGA 41 of the active system comes into the stand-by system, and the FPGA 40 of the stand-by system is to be switched to the active system. Because of this, the input data Din and the taking-out of the output data Dout are to be switched at the same time with the arrival of the switching signal SW01 or SW10 as a trigger. By this, either the FPGA 40 or the FPGA 41 is switched to the active system or the stand-by system, the input data Din is given to the FPGA 40 or 41 of the active system, and the taking-out of the output data Dout corresponding to the input data Din is carried out.
Further, at a forward stage of the switching circuit 14, a data maintaining circuit 22 which maintains the input data Din is provided. At a backward stage of the switching circuit 16, a data maintaining circuit 24 which maintains the output data Dout is provided. These data maintaining circuits 22 and 24 maintain a corresponding data for a predetermined time, and update the maintained data by the lapse of time. That is, in the data maintaining circuit 22, the input data Din which is given to the FPGA 40 or 41 of the active system through the switching circuit 14 is maintained. In the data maintaining circuit 24, the output data Dout which is taken out from the FPGA 40 or 41 of the active system through the switching circuit 16 is maintained. By data maintenance like this, it is possible to prevent a lack of the input data Din or the output data Dout due to switching of the FPGAs 40 and 41 described before.
Further, as a constitution which performs switching with failure occurrence of the FPGA 40 or 41 as a trigger, a 0-system failure detection circuit 26 serving as a failure detection part which detects the occurrence of a failure is provided in regard to the FPGA 40, and a 1-system failure detection circuit 28 serving as a failure detection part which detects the occurrence of a failure is provided in regard to the FPGA 41. The 0-system failure detection circuit 26 detects a failure concerning the FPGA 40 of the active system. For example, the 0-system failure detection circuit 26 detects operating abnormality caused by some electrical cause, for example, a software error of circuit information written in the FPGA 40, and generates a failure detection signal. Similarly, the 1-system failure detection circuit 28 detects a failure concerning the FPGA 41 of the active system. For example, the 1-system failure detection circuit 28 detects operating abnormality caused by some electrical cause, for example, a software error of circuit information written in the FPGA 41, and generates a failure detection signal.
The failure detection signal generated by the 0-system failure detection circuit 26 or the 1-system failure detection circuit 28 is given to the CPU 12, the FPGAs 40 and 41 and the switching circuits 14 and 16, and is representative of failure occurrence. Along with this, the failure detection signal acts as a reset signal to the FPGAs 40 and 41, and acts as a switching signal to the switching circuits 14 and 16. That is, in case in which the 0-system failure detection circuit 26 detects a failure, the FPGA 40 is reset and is switched from the active system to the stand-by system. Along with this, by switching the switching circuits 14 and 16, the input data Din is given to the FPGA 41 of the stand-by system, and its output data Dout is taken out. On the other hand, in case in which the 1-system failure detection circuit 28 detects a failure, the FPGA 41 is reset and is switched from the active system to the stand-by system. Similarly, by switching the switching circuits 14 and 16, the input data Din is given to the FPGA 40 of the stand-by system, and its output data Dout is taken out.
Furthermore, as a storage part which stores circuit information to be written in each of the FPGAs 40 and 41, a Configuration ROM 30 (hereinafter represented as “ROM 30”) is provided. This ROM 30 is used for the initialization of each of the FPGAs 40 and 41. Along with this, the ROM 30 constitutes a function recovery part on a point that the ROM 30 recovers a function of the FPGA 40 or 41 in which a failure is occurring, and constitutes a rewriting part on a point that the ROM 30 rewrites the circuit information. This embodiment is a constitution in which rewriting of the circuit information is performed from the ROM 30 to the FPGA 40 or 41 of the stand-by system by failure occurrence or periodic switching.
In such a constitution, the periodic switching circuit 6 forms the switching signal SW01 or SW10 based on the clock signal CK from the timer circuit 8. The switching circuits 14 and 16 are switched by the switching signal SW01 or SW10. Along with this, by the switching signal SW01, the FPGA 40 is reset and comes into the stand-by system, and the FPGA 41 is also switched to the active system. On the other hand, by the switching signal SW10, the FPGA 41 is reset and comes into the stand-by system, and the FPGA 40 is also switched to the active system. As a result, the FPGAs 40 and 41 are switched to the active system or the stand-by system selectively. This switching is executed every predetermined time which is set to the switching signals SW01 and SW10, for example, every twelve hours.
For example, assuming that the FPGA 40 is in the active system and the FPGA 41 is in the stand-by system, the input data Din is given to the FPGA 40, and the output data Dout is taken out from the FPGA 40. In this case, the input data Din is maintained in the data maintaining circuit 22 and is given to the FPGA 40 through the switching circuit 14, and the output data Dout of the FPGA 40 is maintained in the data maintaining circuit 24 through the switching circuit 16 and is output. Similarly, assuming that the FPGA 41 is in the active system and the FPGA 40 is in the stand-by system, the input data Din is given to the FPGA 41, and the output data Dout is taken out from the FPGA 41. In this case, the input data Din is maintained in the data maintaining circuit 22 and is given to the FPGA 41 through the switching circuit 14, and the output data Dout of the FPGA 41 is maintained in the data maintaining circuit 24 through the switching circuit 16 and is output.
Then, the circuit information is given to the FPGA 40 or the FPGA 41, which is shifted to the stand-by system, from the ROM 30, and, after the rewriting of the circuit information is performed as configuration processing, its stand-by state is maintained.
Further, in addition to the above-mentioned periodic switching, if a failure occurs in the FPGA 40 under operation and this failure is detected by the 0-system failure detection circuit 26, the switching signal is output from the 0-system failure detection circuit 26. By this switching signal, the FPGA 40 is reset, and the switching circuits 14 and 16 are also switched to the FPGA 41 side. As a result, the FPGA 40 comes into the stand-by system, and the FPGA 41 comes into the active system. By this, the input data Din is given to the FPGA 41, and the output data Dout is taken out from this FPGA 41.
On the other hand, if a failure occurs in the FPGA 41 under operation and this failure is detected by the 1-system failure detection circuit 28, the switching signal is output from the 1-system failure detection circuit 28. By this switching signal, the FPGA 41 is reset, and the switching circuits 14 and 16 are also switched to the FPGA 40 side. As a result, the FPGA 41 comes into the stand-by system, and the FPGA 40 comes into the active system. By this, the input data Din is given to the FPGA 40, and the output data Dout is taken out from this FPGA 40.
Then, since the failure is occurring, the circuit information is given to the FPGA 40 or the FPGA 41, which is shifted to the stand-by system, from the ROM 30. By this, the writing of the circuit information is performed as the configuration processing, and, after recovery of the function, its stand-by state is maintained.
Like this, the FPGAs 40 and 41 are provided as a plurality of function parts, and one of these is switched to the active system and the other is switched to the stand-by system, periodically. Further, the configuration processing is given to the FPGA 40 or 41 of the stand-by system. Furthermore, the occurrence of a failure in the FPGA 40 or 41 of the active system is observed, the FPGA 40 or 41 which is in the failure is switched to the stand-by system, and the configuration processing is given to the FPGA 40 or 41 of the stand-by system and it is made to stand by. Because of this, the function is continuously maintained, and the reliability of the system is heightened.
Next, with this functional device 2 as a premise, a function maintaining method and a function maintaining program are explained by referring to
By impressing a power source, all systems of the functional device 2 are shifted to an operating state, and, as initialization processing of the FPGAs 40 and 41, configuration processing of the FPGA 40 is executed (a step S1). As a result of this processing, the FPGA 40 comes into the active system (a step S2). Receiving a shift to the active system of the FPGA 40, configuration processing of the FPGA 41 is executed (a step S3). As a result of this processing, the FPGA 41 comes into the stand-by system (a step S4). In the configuration processing of the FPGAs 40 and 41, circuit information is given to the FPGAs 40 and 41 from the ROM 30, and the circuit information which is read from the ROM 30 is written in the internal RAM 5.
After the above-mentioned configuration processing, as operation processing of the FPGAs 40 and 41, either one of the FPGAs 40 and 41 is set to active system and the other is set to the stand-by system by the switching signal SW01 or SW10 of the periodic switching circuit 6. As a result of this, the input data Din is given to the FPGA 40 or the FPGA 41 under operation, and the output data Dout is taken out from the FPGA 40 or the FPGA 41 under operation.
In connection with these FPGAs 40 and 41 which are in the active system or in the stand-by system, whether or not a failure is occurring is decided by the outputs of the 0-system failure detection circuit 26 and the 1-system failure detection circuit 28. That is, whether or not a failure is occurring in the FPGA 40 is decided (a step S5), and, in case in which the failure is not occurring in the FPGA 40, whether or not a failure is occurring in the FPGA 41 is decided (a step S6).
In case in which a failure is occurring in neither of the FPGAs 40 and 41, whether or not periodic switching timing for the purpose of making one of the FPGAs 40 and 41 the active system and making the other the stand-by system comes is decided (a step S7). In case in which this periodic switching timing is not generated, the processing of the step S5 through the step S7 is repeated until the switching timing comes.
If the switching timing comes, whether or not a failure is occurring in both of the systems of the FPGAs 40 and 41 is decided (a step S8). Also in case in which a failure is occurring in either one of the FPGAs 40 and 41, the processing of the step S8 is executed. Then, in case in which the failure is occurring in both of the systems, processing is returned to the step S1, and the processing described before is executed again.
In case in which a failure is occurring in neither of the FPGAs 40 and 41, whether at present the FPGA 40 is under operation or the FPGA 41 is under operation is decided (a step S9). By this decision processing, in case in which at present the FPGA 40 is in the active system, the FPGA 41 is switched to the active system (a step S10). At this time, the FPGA 40 which is in the active system is switched to the stand-by system. To the FPGA 40 shifted to the stand-by system, the configuration processing is executed (a step S11). After that, the processing is returned to the step S5. In this case, also in case in which a failure occurs in the FPGA 40 (the step S5) and it is shifted to the stand-by system, the configuration processing is executed to the FPGA 40, similarly. Further, as a result of this decision processing (the step S9), in case in which at present the FPGA 41 is in the active system, the FPGA 40 is switched to the active system (a step S12). At this time, the FPGA 41 which is in the active system is switched to the stand-by system. To the FPGA 41 shifted to the stand-by system, the configuration processing is executed (a step S13). After that, the processing is returned to the step S5. In this case, also in case in which a failure occurs in the FPGA 41 (the step S6) and it is shifted to the stand-by system, it goes without saying that the configuration processing is executed to the FPGA 41 similarly.
Next, in connection with the function maintaining method or the function maintaining program, its fundamental operation is explained by referring to
If a power source is impressed, the FPGA 40 comes into the active system (ACT) after the configuration processing to the FPGA 40. With this shift to the active system of the FPGA 40, the configuration processing of the FPGA 41 is executed, and the FPGA 41 comes into the stand-by system (SBY).
If operation like this is performed continuously and a predetermined time passes, the periodic switching is executed. As a result of this, the FPGA 41 under stand-by comes into the active system (ACT). The FPGA 40 which is in the active system is shifted to the stand-by system (SBY), and its stand-by state is maintained after the configuration processing. Further, if the predetermined time passes from this state, the periodic switching is executed. By this, the FPGA 40 under stand-by comes into the active system, and the FPGA 41 which is in the active system is shifted to the stand-by system. After that, the configuration processing is given, and its stand-by state is maintained. Such operation is executed until the power source is released.
Next, in connection with the function maintaining method or the function maintaining program, operation corresponding to failure occurrence before the periodic switching is explained by referring to
As shown in
Further, as shown in
Next, in connection with the function maintaining method or the function maintaining program, operation corresponding to failure occurrence after the periodic switching from the 0 system to the 1 system is explained by referring to
As shown in
As shown in
Next, in connection with the function maintaining method or the function maintaining program, operation corresponding to failure occurrence after the periodic switching from the 1 system to the 0 system is explained by referring to
As shown in
As shown in
Next, in connection with the function maintaining method or the function maintaining program, operation corresponding to failure occurrence in both systems of the 0 system and the 1 system is explained by referring to
The basic operation after the impression of the power source is as described before. In this case, for example, on the supposition that the FPGA 40 is the active system and the FPGA 41 is the stand-by system, if a failure occurs in both of the 0 system and the 1 system, the operation of the FPGA 40 which is under operation is stopped, and, after the configuration processing, this FPGA 40 is set to the active system. After this, the configuration processing is given to the FPGA 41 which is under stand-by, and its stand-by state is maintained again. Further, if the predetermined time passes from this state, the periodic switching is executed, the FPGA 41 under stand-by comes into the active system, and the FPGA 40 which is the active system comes into the stand-by system after the configuration processing. Similarly, such operation is executed until the power source is released.
Operating sequence like this is explained by referring to
This sequence includes each state of a state C1 of power-impression (PW-ON) and reset, a stand-by state C2 of the FPGA 40, an operating state C3 of the FPGA 40, a stand-by state C4 of the FPGA 41 and an operating state C5 of the FPGA 41.
Now, if the power source is impressed (PW-ON), the configuration of the FPGA 41 is practiced after practice of the configuration of the FPGA 40, and the FPGA 40 shifts from this power-impression to the stand-by state C2 (a step S101). In the stand-by state C2 of the FPGA 40, the configuration processing of the FPGA 40 is executed. Further, the FPGA 40 shifts from the stand-by state C2 to the operating state C3 (a step S102). In this operating state C3, the processing after the power-impression, the processing of the time of the periodic switching, the processing of the time of a failure in the FPGA 41 and the processing of the time of failures in both the systems are executed. In case in which the FPGA 40 shifts from the operating state C3 to the stand-by state C2 (a step S103), the configuration processing of the time of the recovery of a failure is executed.
Further, a shift from the operating state C3 of the FPGA 40 to the stand-by state C4 of the FPGA 41 (a step S104) is a case of the failure occurrence in the FPGA 40 or a case of the periodic switching. In the stand-by state C4 of the FPGA 41, the configuration processing of the FPGA 41 is executed. Further, the FPGA 41 shifts from the stand-by state C4 to the operating state C5 (a step S105). In this operating state C5, the processing after the power-impression, the processing of the time of the periodic switching, the processing of the time of a failure in the FPGA 40 and the processing of the time of failures in both the systems are executed. In case in which the FPGA 41 shifts from the operating state C5 to the stand-by state C4 (a step S106), the configuration processing of the time of the recovery of a failure is executed. A shift from the operating state C5 of the FPGA 41 to the stand-by state C2 of the FPGA 40 (a step S107) is a case of the failure occurrence in the FPGA 40 or a case of the periodic switching.
Furthermore, in case in which a failure occurs in both of the systems, the configuration of the FPGA 41 is practiced after practice of the configuration of the FPGA 40 (a step S108). Then, all of the operation is stopped by a release of the power source (PW-OFF), and the sequence is returned to the state C1 (a step S109).
A second embodiment of the present invention is explained by referring to
A functional device 2 according to this embodiment is constituted with the exception of the switching due to failure occurrence from the functional device 2 (
Therefore, in this functional device 2, the periodic switching of the function of the FPGAs 40 and 41 and the periodic configuration processing of the FPGAs 40 and 41 are executed as shown in
As described before, by impressing a power source, all systems of the functional device 2 are shifted to an operating state. Then, after configuration processing of the FPGA 40 is executed as initialization processing (a step S21), the FPGA 40 which is initialized comes into the active system (a step S22). Receiving this shift to the active system, the configuration processing of the FPGA 41 is executed (a step S23), and, after that, the FPGA 41 comes into the stand-by system (a step S24). The configuration processing of these FPGAs 40 and 41 is performed by writing circuit information, which is given to FPGAs 40 and 41 from the ROM 30, in the RAM 5.
After the above-mentioned configuration processing, as operation processing of the FPGAs 40 and 41, either one of the FPGAs 40 and 41 is set to the active system and the other is set to the stand-by system by the switching signal SW01 or SW10 of the periodic switching circuit 6. As a result of this, the input data Din is given to the FPGA 40 or 41 under operation, and the output data Dout is taken out from the FPGA 40 or 41 under operation.
If the operation of a necessary function is continued by the FPGA 40 being maintained in the active system and the FPGA 41 being maintained in the stand-by system as mentioned above, a decision in regard to the coming of periodic switching timing is executed (a step S25). Then, if the switching timing comes, whether at present the FPGA 40 is under operation or the FPGA 41 is under operation is decided (a step S26). By this decision processing, in case in which at present the FPGA 40 is in the active system, the FPGA 41 is switched to the active system (a step S27). At this time, the FPGA 40 which is in the active system is switched to the stand-by system. To the FPGA 40 shifted to the stand-by system, the configuration processing is executed (a step S28). After that, the processing is returned to the step S25.
Furthermore, by the decision processing (the step S26), in case in which at present the FPGA 41 is in the active system, the FPGA 40 is switched to the active system (a step S29). To the FPGA 41 which is switched to the stand-by system, the configuration processing is executed (a step S30). After that, the processing is returned to the step S25.
If the FPGA 40 comes into the active system as mentioned above, the FPGA 41 comes into the stand-by system, and the configuration processing is executed. Further, at next timing, the FPGA 40 which is in the active system is switched to the stand-by system and the configuration processing is executed, and the FPGA 41 which is in the stand-by system comes into the active system. Since the operational switching is continuously performed every predetermined time as mentioned above, the FPGAs 40 and 41 of which the functions a reguaranteed by the configuration processing are continuously switched, and the reliability of the function is heightened.
A third embodiment of the present invention is explained by referring to
A functional device 2 according to this embodiment is constituted with the exception of a constitution of the periodic switching from the functional device 2 (
Therefore, in this functional device 2, at the time of failure occurrence, the operational switching of the FPGAs 40 and 41 is performed, and the configuration processing is also executed, as shown in
As described before, by impressing a power source, all systems of the functional device 2 are shifted to an operating state. Then, after the configuration processing of the FPGA 40 is executed as initialization processing (a step S31), the FPGA 40 which is initialized comes into the active system (a step S32). Receiving this shift to the active system, the configuration processing of the FPGA 41 is executed (a step S33), and, after that, the FPGA 41 comes into the stand-by system (a step S34). The configuration processing of these FPGAs 40 and 41 is performed by writing circuit information, which is given to the FPGAs 40 and 41 from the ROM 30, in the RAM 5.
After the above-mentioned configuration processing, as operation processing of the FPGAs 40 and 41, either one of the FPGAs 40 and 41 is set to the active system and the other is set to the stand-by system. As a result of this, the input data Din is given to the FPGA 40 or 41 under operation, and the output data Dout is taken out from the FPGA 40 or 41 under operation.
Regarding these FPGAs 40 and 41 which are in the active system or the stand-by system, whether or not a failure is occurring is decided by the outputs of the 0-system failure detection circuit 26 and the 1-system failure detection circuit 28. In this embodiment, first, whether or not a failure is occurring in the FPGA 40 is decided (a step S35). In case in which the failure is occurring in the FPGA 40, the FPGA 41 is switched to the active system (a step S36). Then, the configuration processing is executed to the FPGA 40 shifted to the stand-by system (a step S37), and the processing is returned to the step S35.
On the other hand, in case in which a failure is not occurring in the FPGA 40, whether or not a failure is occurring in the FPGA 41 is decided (a step S38). In case in which the failure is occurring in the FPGA 41, the FPGA 40 is switched to the active system (a step S39). Then, the configuration processing is executed to the FPGA 41 shifted to the stand-by system (a step S40), and the processing is returned to the step S35.
Like this, if a failure occurs in the FPGA 40 which is in the active system, the FPGA 41 comes into the active system, and the configuration processing is also executed to the FPGA 40 shifted to the stand-by system. On the other hand, if a failure occurs in the FPGA 41 which is in the active system, the FPGA 40 comes into the active system, and the configuration processing is also executed to the FPGA 41 shifted to the stand-by system. Like this, if a failure occurs, the stand-by system comes into the active system, and, along with the shift to the stand-by system, the configuration processing is performed. By this, since the function is guaranteed, the reliability of the function is heightened.
Next, the specific example of a constitution of the functional device according to the first embodiment is explained by referring to
In a functional device 2 of this embodiment, as the switching circuit 14 of the input data Din to the FPGAs 40 and 41, an AND circuits 32 and 34 are provided. The AND circuit 32 corresponds to the FPGA 40, and the AND circuit 34 corresponds to the FPGA 41. In order to selectively control the AND circuit 32 and the AND circuit 34 to a signal passing state, a NOT circuit 36 is added to one input of the AND circuit 32. Further, the input data Din is given to each of the AND circuits 32 and 34 from the data maintaining circuit 22 in common, and an output Q of a flip-flop (FF) circuit 38 is also given to them. To the FF circuit 38, the switching signal SW is given as a clock input CP from the periodic switching circuit 6 (
For the data maintaining circuit 22, a flip-flop (FF) circuit 50 is used. To this FF circuit 50, the input data Din is given to its data input D, and an output of an OR circuit 52 is given to its clock input CP. The clock signal CK and the switching signal SW from the periodic switching circuit 6 described before are given to the OR circuit 52, and an output added time-sequentially is obtained as its output. In this case, for the clock signal CK, for example, a continuous pulse signal of 100 MHz is used. Accordingly, the input data Din is maintained in the FF circuit 50, and its maintained output is given to the AND circuits 32 and 34 from its output Q.
Further, for the switching circuit 16, a 2-1 selecting circuit 54 which selects one input from two inputs and outputs it is used. The output data Dout of the FPGA 40 is given to a 0 (zero) input side of the two inputs, and the output data Dout of the FPGA 41 is given to its 1 (one) input side. To a switching input S of this selecting circuit 54, the output Q of the FF circuit 38 corresponding to the switching signal SW is given. Therefore, selecting operation of the AND circuits 32 and 34 of the switching circuit 14 and the selecting circuit 54 of the switching circuit 16 is performed synchronously with the switching signal SW. Because of this, in case in which the input data Din is applied to the FPGA 40 through the AND circuit 32, the output data Dout of the FPGA 40 is taken out from an output Y of the selecting circuit 54. On the other hand, in case in which the input data Din is applied to the FPGA 41 through the AND circuit 34, the output data Dout of the FPGA 41 is taken out from the output Y of the selecting circuit 54.
Further, for the data maintaining circuit 24, a flip-flop (FF) circuit 56 is used. To this FF circuit 56, the output data Dout is given to its data input D from the selecting circuit 54, and the output of the OR circuit 52 described before is given to its clock input CP. Accordingly, the output data Dout which is obtained from the selecting circuit 54 is maintained in the FF circuit 56, and its maintained output is taken out from its output Q.
Furthermore, the 0-system failure detection circuit 26 (
Such maintenance of an input-output data is explained by referring to
A signal width tw2 of the switching signal SW is set large as opposed to a signal width tw1 of the clock signal CK (
If this output of the OR circuit 52 is applied to the FF circuits 50 and 56, since the data maintaining operation of each of the FF circuits 50 and 56 synchronizes with rising of the clock input CP, the input data Din is maintained in the FF circuit 50 at a data maintaining interval tdh and the output data Dout is maintained in the FF circuit 56 at the data maintaining interval tdh as shown in
The functional device 2 according to the first through third embodiments can be constituted by the circuit shown in
Next, in connection with the embodiments mentioned above, modified examples are enumerated in the following.
(1) In the above-mentioned embodiments, the single ROM 30 which stores the circuit information is provided for the FPGAs 40 and 41 in common. However, as shown in
(2) In the above-mentioned embodiments, the FPGAs 40 and 41 are provided as the plurality of function parts, and one is made into the active system and the other is made into the stand-by system. However, FPGAs more than two can be provided as the plurality of function parts, and a plurality of stand-by systems may also be provided as opposed to one active system.
(3) Although in the above-mentioned embodiments the control part 10 including the CPU 12 is used for the sequential control, the present invention may also be constituted so that the control part 10 performs the switching and the configuration processing of the FPGAs 40 and 41 based on the detection of a failure.
(4) Although in the above-mentioned embodiments the control part 10 including the CPU 12 is used for the sequential control, the present invention may also be constituted so that the control part 10 performs the periodic switching and the configuration processing of the FPGAs 40 and 41.
Although the best mode for carrying out the invention, the object, the configuration and the operation and effect have been described in detail above, the invention is not limited to such embodiment for carrying out the invention, and it is a matter of course that the invention can be variously changed or modified by a person skilled in the art on the basis of a gist and split of the invention as disclosed in claims and the detailed description of the invention, and such a change or modification, and various conjectured configurations, modified examples and so forth are included in the scope of the invention, and the description of the specification and drawings are not restrictively understood.
The entire disclosure of Japanese Patent Application No. 2004-236803 including specification, claims, drawings and summary are incorporated herein by reference in its entirety.
Number | Date | Country | Kind |
---|---|---|---|
2004236803 | Aug 2004 | JP | national |