JP2011237972A - Data processing device and data processing program - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a data processing device capable of reducing a delay in a succeeding process caused by data buffering in a storage.SOLUTION: A main controller 2 queues data writing requests for each of a series of a plurality of band data in order in a writing side thread 31a after notifying a reading side thread 32a of a file identifier in a HDD 4 through the writing side thread 31a. Then, the main controller 2 makes a HDD controller 14 write the series of the plurality of the band data in the order of the queued writing requests using a file writing thread 31b. When the reading side thread 32a receives the file identifier, the main controller 2 queues data reading requests for each of the series of the plurality of the band data in order in the reading side thread 32a and then makes the HDD controller 14 read the series of the plurality of the band data in the order of the queued reading requests using a file reading thread 32b.

Description

  The present invention relates to a data processing apparatus and a data processing program.

  Generally, an image forming apparatus such as a copying machine or a multifunction machine has a built-in data processing device that performs various types of image processing. Such a data processing device, when buffering image data, converts image data for one page into a file in a storage such as a hard disk drive (HDD) by a writing side program (for example, an image processing program). After the storage, the file identifier of the file is notified to a reading side program (for example, a subsequent image processing program). The reading side program reads the image data from the file in the storage.

  In addition, when writing data to the HDD, a FIFO (First-In First-Out) is provided as a buffer, and data is written to the FIFO by one thread, and data is written from the FIFO to the HDD by another thread. (See, for example, Patent Document 1).

JP 2007-17996 A

  As described above, normally, after data of a predetermined unit (in the above case, page) is stored in the storage, the data is read and read sequentially in order to read the data. For this reason, before the data writing of a predetermined unit of data is completed, the data cannot be read, and a delay occurs in the processing after buffering.

  The present invention has been made in view of the above problems, and an object of the present invention is to obtain a data processing device and a data processing program capable of reducing a processing delay caused by data buffering to a storage.

  In order to solve the above problems, the present invention is configured as follows.

  A data processing apparatus according to the present invention includes a storage controller that writes a series of data blocks constituting a predetermined unit of data to a storage, reads the series of data blocks from the storage, and a processor that controls the storage controller. . The processor then generates (i) a writer process and a reader process, (ii) a first thread and a third thread in the writer process, and (iii) a second thread and in the reader process After the fourth thread is generated and (iv) the storage area identifier in the storage is notified from the first thread to the second thread, the first thread sequentially issues the data write requests for each of the series of data blocks. (V) using the third thread, causing the storage controller to write a series of data blocks in the order of the queued data write request, and (vi) identifying the identifier with the second thread. Upon receipt, the data read requests for each of the series of data blocks are queued in order, and (vii) With threads, in the order of queuing data read request to read out a series of data blocks to the storage controller.

  As a result, before the data is written to the storage, the identifier of the storage area for the buffer is passed to the reading side process, and data reading can be started by the reading side process independently of the writing side process. Data can be read even in the middle. Therefore, the processing delay after data buffering to the storage can be reduced. Further, since the program for writing to the storage operates in another thread, the data writing is performed asynchronously, and the writing side program can execute another process without waiting for the data writing to be completed. . Further, since the program for reading from the storage operates in another thread, the data reading is performed asynchronously, and the reading side program can execute another process without waiting for the completion of data reading.

  In addition to the data processing apparatus described above, the data processing apparatus according to the present invention may be configured as follows. In this case, the data processing apparatus includes a memory that stores a series of a plurality of data blocks. The series of data blocks described above is a plurality of band data constituting one page of image data in one document. The memory stores page node data indicating each page in the document. Further, the plurality of band data is a link list provided for each page in the document, and the page node data of a certain page has a link to the head of the plurality of band data for the page.

  Thereby, since the band data of a plurality of pages is made into a separate link list for each page, the band data to be processed can be found in a short time from the band data of the plurality of pages.

  The data processing apparatus according to the present invention may be as follows in addition to any of the data processing apparatuses described above. In this case, the processor uses the third thread to identify the start positions of the plurality of band data with the page node data, and traces the link list to write the plurality of band data to the storage controller.

  Thereby, it is possible to find out the band data to be written from the band data of a plurality of pages in a short time.

  The data processing apparatus according to the present invention may be as follows in addition to any of the data processing apparatuses described above. In this case, the processor secures a storage area for a plurality of band data when storing a plurality of band data for one page in the memory, and releases the storage area when the writing of the page is completed. If there is a shortage, when a plurality of band data storage areas for other pages are released, a plurality of band data storage areas are secured.

  Thus, the band data storage area of the page for which processing has been completed is immediately released and can be used as the band data storage area of the subsequent page.

  The data processing apparatus according to the present invention may be as follows in addition to any of the data processing apparatuses described above. In this case, the processor (viii) stores the data block in the memory in the third thread, (ix) issues an event each time a data write request for the data block is queued in the first thread, and (x) storage When the data block is completely written to the memory block, the third thread erases the data block from the memory, and (xi) when executing the queued data read request in sequence, if there is a data block in the memory, The fourth thread reads the data block from the memory, and if there is no data block in the memory, the fourth thread is used to cause the storage controller to read the data block from the storage and (xii) read from the memory at a predetermined timing. Queue data write requests for issued data blocks To remove et al.

  As a result, it is possible to avoid a processing delay due to a data write request remaining in the queue despite the data block being read from the memory.

  The data processing apparatus according to the present invention may be as follows in addition to any of the data processing apparatuses described above. In this case, the processor deletes the data write request for the data block read from the memory from the queue when the job is completed.

  The data processing apparatus according to the present invention may be as follows in addition to any of the data processing apparatuses described above. In this case, the series of data blocks described above is written in a file in the storage, and the identifier of the storage area is the identifier of the file.

  As a result, the buffer storage area can be managed by the file system, and the data in the storage from each process can be easily accessed.

  According to the present invention, in the data processing apparatus, it is possible to reduce a delay in subsequent processing due to data buffering to the storage.

FIG. 1 is a block diagram showing a configuration of a data processing apparatus according to an embodiment of the present invention. FIG. 2 is a diagram showing a configuration of data handled by the data processing circuit in FIG. FIG. 3 is a diagram showing an example of data handled by the data processing circuit in FIG. FIG. 4 is a block diagram illustrating the main controller in the first embodiment. FIG. 5 is a flowchart for explaining the operation of the write-side process of the main controller in the first embodiment. FIG. 6 is a flowchart for explaining the operation of the reading-side process of the main controller in the first embodiment. FIG. 7 is a diagram showing a shared memory area in the queue system in FIG. FIG. 8 is a flowchart for explaining the operation of the write-side process of the main controller in the second embodiment. FIG. 9 is a flowchart for explaining the operation of the reading-side process of the main controller in the second embodiment. FIG. 10 is a diagram for explaining reading of band data from the RAM in the second embodiment.

  Hereinafter, embodiments of the present invention will be described with reference to the drawings.

Embodiment 1 FIG.

  FIG. 1 is a block diagram showing a configuration of a data processing apparatus according to an embodiment of the present invention. The data processing apparatus shown in FIG. 1 includes a data processing circuit 1, a main controller 2, a RAM (Random Access Memory) 3, a hard disk drive (HDD) 4, a scanner 5, a print engine control circuit 6, a storage device 7, and printing on paper. A print engine (not shown) to perform.

  The data processing circuit 1 is a device that processes image data from the scanner 5 or the like in band units or page units and outputs the processed image data to the HDD 4 or the print engine control circuit 6.

  The main controller 2 is a device that supplies commands to each unit in the data processing circuit 1. The main controller 2 supplies a command to each part in the data processing circuit 1 via the bus interface circuit 11 and the bus 12. The main controller 2 is a processor such as a CPU (Central Processing Unit) and outputs a command according to a data processing program 8 stored in the storage device 7.

  The RAM 3 is a memory for temporarily storing band data and the like in the processing by the data processing circuit 1. As the RAM 3, for example, one or more DRAMs (Dynamic RAM) are used.

  The HDD 4 is a storage device for temporarily storing a series of a plurality of band data constituting one page of image data.

  The scanner 5 is an apparatus that optically reads an image of a document and outputs image data obtained by the image reading.

  The print engine control circuit 6 is a circuit that supplies print image data to the print engine and controls the print engine to execute printing.

  The storage device 7 is a non-volatile storage device such as a flash memory, and stores various programs and data used by the main controller 2. Here, the data processing program 8 is stored in the storage device 7.

  Here, the configuration of the data processing circuit 1 will be described.

  The data processing circuit 1 includes a bus interface circuit 11, a bus 12, a memory interface 13, an HDD controller 14, a data acquisition circuit 15, an image processing circuit 16, a raster processing circuit 17, a halftone processing circuit 18, and a data output circuit 19. . The data processing circuit 1 is realized as an ASIC (Application Specific Integrated Circuit).

  The bus interface circuit 11 is a circuit that connects the main controller 2 to the bus 12.

  The memory interface 13 is a circuit that transmits and receives data and commands between the RAM 3 and the bus 12.

  The HDD controller 14 is a circuit that reads / writes data from / to the HDD 4 in response to a command from the main controller 2.

  The data acquisition circuit 15 is a circuit that sequentially acquires image data as band data from the scanner 5 in accordance with a command from the main controller 2 and stores it in the RAM 3 or the HDD 4. The data acquisition circuit 15 has a DMA (Direct Memory Access) controller. For example, the band data is data for 128 lines, and there is band data for 58 bands per page.

  FIG. 2 is a diagram showing a data structure handled by the data processing circuit 1. One document is composed of one or a plurality of pages. A page is composed of a predetermined number of bands. Data for one document is supplied to the main controller 2 by one job.

  The document structure is represented by a linked list of nodes. One or more documents are represented by a linked list of document node data. A link list of page node data for one or a plurality of pages constituting the document is connected to the document node data of a document as a branch. Furthermore, a link list of band data for a plurality of bands constituting the page is connected to the page node data of a page as a branch. Therefore, the data structure is a three-layer structure including document node data, page node data, and band data.

  FIG. 3 is a diagram illustrating an example of data handled by the data processing circuit 1. As shown in FIG. 3, this data structure has a three-layer structure of a document link list, a page link list, and a band link list. In this example, the document link list includes three document node data. The document node data of document # 1 includes link information to the page link list, and the page link list includes page node data of two pages. The document node data of document # 2 includes link information to the page link list, and the page link list includes page node data of one page. The document node data of document # 3 includes link information to the page link list, and the page link list includes page node data of three pages. Each page node data includes link information to the band link list. Since the number of bands in one page is constant, the size of the band link list is constant. The document node data includes document identification information (such as an identification number), link information for linking with subsequent document node data, and link information of a page link list. The page node data includes page identification information (such as an identification number), link information for connection with subsequent page node data, and link information of a band link list. The band link list has a predetermined number of band data. The band data includes image data for one band and link information of subsequent band data.

  Therefore, by tracing the above-described link from the document, page, and band numbers, data of a certain band on a certain page in the document can be specified and read out.

  The image processing circuit 16 reads band data of the image data from the RAM 3 or the HDD 4 in response to a command from the main controller 2, and performs predetermined image processing (image enlargement, image reduction, color conversion, etc.) on the band data. The band data after processing is stored in the RAM 3 or the HDD 4. The image processing circuit 16 has a DMA controller.

  The raster processing circuit 17 is a circuit that reads band data of image data from the RAM 3 or the HDD 4, performs rasterization on the band data, and generates raster data. The raster data is stored in the HDD 4 for each band and read out to the RAM 3 when the halftone process is executed.

  The halftone processing circuit 18 is a circuit that reads raster data for each band from the RAM 3, executes halftone processing, and stores the data after halftone processing in the RAM 3.

  The data output circuit 19 is a circuit that reads out the data after halftone processing from the RAM 3 and supplies the data to the print engine control circuit 6.

  In the data processing circuit 1, image processing such as rotation and enlargement / reduction is performed on the image data acquired from the scanner 5 as necessary. Thereafter, print image data is generated by halftoning and output to the print engine control circuit 6. During these data processing, the RAM 3 is used for data buffering in band units, and the HDD 4 is used for data buffering in page units.

  Next, data writing and data reading operations to the HDD 4 by the data processing apparatus will be described.

  The main controller 2 includes a thread in which a shared library program for writing band data to a file operates, and a shared library for reading band data from a file, separately from a writing thread and a reading thread in which the application 23 operates. Create a thread on which the program runs. As a result, the data processing apparatus executes band data writing and reading processing asynchronously with the processing of the application 23.

  FIG. 4 is a block diagram for explaining the operation of the main controller 2 in the first embodiment. FIG. 5 is a flowchart for explaining the operation of the writing process 21 of the main controller 2 in the first embodiment. FIG. 6 is a flowchart for explaining the operation of the reading-side process 22 of the main controller 2 in the first embodiment.

  In the main controller 2, an operating system that supports multi-process and multi-thread is operating. The main controller 2 executes the data processing program 8 and first generates a writing side process 21 and a reading side process 22. The main controller 2 executes an application 23 for buffering data in the HDD 4 using the writing side process 21 and the reading side process 22. Of the application 23, the writing side program 41 is executed by the writing side thread 31 a in the writing side process 21. The main controller 2 can cause the HDD controller 14 to write band data to the HDD 4 by executing the writing side program 41. On the other hand, in the application 23, the reading side program 51 is executed by the reading side thread 32 b in the process 22. The main controller 2 can cause the HDD controller 14 to read the band data from the HDD 4 by executing the reading side program 51.

  Then, the main controller 2 generates a file writing thread 31b in the writing side process 21, and executes the shared library program 71 for writing data to the file by the file writing thread 31b. The main controller 2 generates a file reading thread 32b in the reading process 22, and executes a shared library program 81 for reading band data from the file by the file reading thread 32b. Further, the main controller 2 queues data write requests and data read requests generated by the application 23 as access requests to the queue system 91, executes the file write program 71 and the file read program 81, and sends them to the queue system 91. The band data is written and read based on the queued access request. Thereby, the data processing device executes writing and reading of band data asynchronously with the processing of the application 23.

  When the main controller 2 receives the document data, the main controller 2 stores the data in the RAM 3 in a three-layer structure as shown in FIG. At this time, for each document, one document node data is added to the document link list, and the same number of page node data as the number of pages included in the document is added to the page link list. Furthermore, a band link list storage area is newly secured for each page, and the band link list is stored in the storage area. When the main controller 2 stores a plurality of band data for one page in the RAM 3, it adds the page node data of the page to the page link list and secures a storage area for the plurality of band data. When completed, the storage area is released and the page node data is deleted from the page link list. If the remaining amount of RAM 3 is insufficient when the band data storage area is secured, the main controller 2 waits until the band data storage area for the other page is released. A storage area for the band data is secured.

  The main controller 2 notifies the file identifier of the file to which the band data is written to the reading side program 51 of the reading side process 22 by executing the writing side program 41 in the writing side thread 31a (step S31). S21). The file identifier is, for example, a file name. The main controller 2 calls the shared library of the file system 61 with the API 42 according to the writing side program 41, opens the file with the file identifier, and generates a file handle (step S22).

  After the file is opened, the main controller 2 calls the API 42 of the file system 61 according to the write-side program 41 by the write-side thread 31a, and sequentially issues the band data data write requests to the file corresponding to the file identifier in the queue system. Queue to 91. In response to the data write request, the main controller 2 supplies a write command for each band data to the file write thread 31b according to the write side program 41 (step S23). In the write command, the identification information of the band data to be written (that is, the identification information of the document including the band data, the identification information of the page including the band data, and the identification information of the band data in the page) ) Is specified. The main controller 2 repeats the process of step S23 until queuing data write requests for all band data (step S24).

  FIG. 7 is a diagram showing a shared memory area in the queue system 91 in FIG. This shared memory area is allocated on the RAM 3 and its physical address is mapped to the logical address of the writing process 21 and the reading process 22 so that each process can access the shared memory area. In the queue system 91, the shared memory area includes a document list area, a page list area, an FD list area, and a queue area.

  In the document list area, document node data including band data to be written or read is stored by API 42 or API 52. The node data of the document has link information to the node data of the pages constituting the document.

  The page list area stores node data of pages including band data to be written or read by API 42 or API 52. The node data of the page includes link information to the band node data constituting the page and link information to the file descriptor.

  In the band list area, node data of a band to be written or read is stored by the API 42 or API 52. The node data of the band has the ID and serial number of the document and page to which the band belongs, and further has identification information (serial number and the like) of the band data in the file.

  In the FD list area, a file descriptor of a file corresponding to the page is written for each page by API 42 or API 52.

  In the queue area, an access request is written for each band data by API 42 or API 52. The access request includes an access type (write or read) and identification information of target band data. Therefore, a data write request and a data read request are mixed in one queue.

  Access to the queue system 91 is controlled exclusively by a semaphore, and cannot be simultaneously accessed from two or more of the writing thread 31a, the file writing thread 31b, the reading thread 32a, and the file reading thread 32b. Yes.

  On the other hand, the main controller 2 searches the access request in the queue by the file writing thread 31b according to the file writing program 71 (step S31). When the main controller 2 finds a data write request for the band data for which a write command has been issued in the queue (step S32), the main controller 2 determines the band data and write destination specified by the data write request as the queue system 91. Identifies from the structure data and file descriptor written in The main controller 2 reads the band data from the RAM 3 to the HDD controller 14 by DMA transfer from the file writing thread 31b.

  At this time, the document and page to which the band data to be written belongs are specified from the data write request. First, the document node data of the document to which the band data to be written belongs is specified and connected to the document node data. In the page link list, page node data of a page to which the band data to be written belongs is specified. Then, the band data to be written is specified in the band link list connected to the page node data.

  Then, the HDD controller 14 writes the read band data to the specified writing destination (step S33).

  When the data writing of one band data is completed, the main controller 2 deletes the data writing request of the band data from the queue according to the file writing program 71 by the file writing thread 31b (step S34), and the API 42 A data write completion notice of the band data is supplied to (step S35).

  As described above, the queued data write access request is processed asynchronously by executing the file write program 71 of the file write thread 31 b different from the application 23.

  When the data write completion notification is supplied to the API 42, the main controller 2 issues an event through the API 42 (step S26). The main controller 2 erases the band data for which data writing has been completed from the RAM 3 in accordance with the write side program 41 by the write side thread 31a (step S27).

  When the writing of all the band data for one page is completed (step S28), the main controller 2 calls the shared library of the file system 61 using the API 42 according to the writing program 41, and sets the file identifier. The corresponding file is closed (step S29).

  On the other hand, the main controller 2 executes the reading side program 51 by the reading side thread 32a. Here, the reading side program 51 operates asynchronously with the writing side. When the main controller 2 receives a file identifier from the writing thread 31a in accordance with the reading program 51 in the reading thread 32a (step S41), the main controller 2 opens a file corresponding to the file identifier and generates a file handle (step S41). S42).

  Then, after the file is opened, the main controller 2 calls the API 52 of the file system 61 in accordance with the read-side program 51 by the read-side thread 32a, and sequentially issues a data read request for each band data from the file corresponding to the file identifier in the queue system. Queue to 91. In response to the data read request, the main controller 2 supplies a read command for each band data to the file read thread 32b according to the read-side program 51 (step S43). In the read command, identification information of the band data to be read (that is, identification information of the document including the band data, identification information of the page including the band data, and identification information of the band data in the page) It is specified. The main controller 2 repeats the process of step S43 until queuing data read requests for all band data (step S44).

  On the other hand, when the file read thread 32b receives the event at the completion of writing the band data (step S51), the main controller 2 searches for a data read request in the queue according to the file read program 81 (step S52). . When the main controller 2 finds an access request for the band data for which a read command has been issued in the queue (step S53), the main controller 2 calls the API 52 of the file system 61 and causes the HDD controller 14 to read the band data from the file in the HDD 4. Control (step S54). At this time, the band data and the reading source designated by the data reading request are specified from the structure data and the file descriptor written in the queue system 91. The HDD controller 14 reads the band data from the specified reading source. The main controller 2 writes the read band data in the RAM 3 by DMA transfer according to the file reading program code 81.

  When the data reading of one band data is completed, the main controller 2 deletes the data reading request of the band data from the queue in the file reading thread 32b according to the file reading program 81 (step S55), and sends the band data to the API 52. A data read completion notification is supplied (step S56).

  In this way, the band data is read in order for each event from the writing side. Then, the main controller 2 determines whether or not reading of all band data for one page has been completed for each data reading completion notification in accordance with the reading side program 51 in the reading side thread 32a (step S46). When the reading of all band data for one page is completed, the main controller 2 calls the shared library of the file system 61 using the API 52, and closes the file corresponding to the file identifier. (Step S47) Here, the HDD controller 14 erases the closed file from the HDD 4.

  As described above, according to the first embodiment, the main controller 2 (i) generates the write-side thread 31a and the file-write thread 31b in the write-side process 21, and (ii) the read-side process 22 (Iii) after writing the identifier of the file in the HDD 4 from the writing thread 31a to the reading thread 32a, the writing thread 31a generates a series of a plurality of bands. A data write request for each of the data is queued in order, and (iv) a series of a plurality of band data is sent to the HDD controller 14 in the order of the queued data write request using the file write thread 31b. (V) When the reading-side thread 32a receives the file identifier of the file, Queue data read requests for each of the data in order, and (vi) use the file read thread 32b to read a series of band data to the HDD controller 14 in the order of the queued data read requests Let

  As a result, since the program 71 for writing to the HDD 4 operates in another thread 31b, the data writing is performed asynchronously, and the writing program 41 executes another process without waiting for the completion of the data writing. can do. Further, since the program 81 for reading from the HDD 4 operates in another thread 32b, the data is read asynchronously, and the reading side program 51 can execute another process without waiting for the completion of the data reading.

  Further, according to the first embodiment, band data for one page is stored in the RAM 3 as a link list provided for each page in the document. Page node data indicating a page is stored. The page node data of a certain page has a link to the head of a plurality of band data for the page.

  Thereby, since the band data of a plurality of pages is made into a separate link list for each page, the band data to be processed can be found in a short time from the band data of the plurality of pages.

Embodiment 2. FIG.

  In the second embodiment of the present invention, an event is issued when queuing of band data writing is completed in the writing thread 31a. In the file reading thread 32b, if there is band data on the RAM 3, the band data is read from the RAM 3, and if there is no band data on the RAM 3, the band data is read from the file on the HDD 4.

  Note that the basic configuration of the data processing apparatus according to the second embodiment is the same as that of the first embodiment (FIG. 1), and a description thereof will be omitted. Since the operation of the main controller 2 in the second embodiment is different from that of the first embodiment, this point will be described below. The basic program configuration in the second embodiment is the same as that in the first embodiment (FIG. 4), but the operations by the writing thread 31a and the file reading thread 32b are the same as those in the first embodiment (FIG. 4). ) Is different.

  FIG. 8 is a flowchart for explaining the operation of the writing process 21 of the main controller 2 according to the second embodiment. FIG. 9 is a flowchart for explaining the operation of the reading-side process 22 of the main controller 2 in the second embodiment.

  As shown in FIG. 8, in the second embodiment, the main controller 2 causes the write-side thread 31a to execute an event every time band data queuing (step S23) ends, not when band data writing is completed. Is issued (step S61).

  Then, as shown in FIG. 9, when the file read thread 32b finds an access request in the queue, the main controller 2 checks whether or not the band data to be read remains on the RAM 3 (step S71). . As a result, if the band data to be read remains on the RAM 3, the main controller 2 reads the band data from the RAM 3 according to the file reading program by the file reading thread 32 (step S72). If the band data to be read does not remain on the RAM 3, the main controller 2 controls the HDD controller to read the band data from the file in the HDD 4 according to the file read program by the file read thread. At this time, the HDD controller 14 reads band data from the HDD 4 (step S54).

  As described above, when band data to be read remains on the RAM 3, the band data is read from the RAM 3 instead of from the HDD 4. Therefore, it is not necessary to write the band data to the HDD 4 in response to a data write request for the band data. However, the data write request remains in the queue system 91. If a job (document) having such an unnecessary data write request remains, even if the job is completed on the reading side (that is, reading of the document is completed), the data write request Until the process is completed, the job is not completed on the writing side, and the subsequent processing is delayed. Therefore, the main controller 2 deletes the data write request for the band data already read from the RAM 3 from the queue at a predetermined timing by the reading-side thread 31. In this embodiment, the timing for deleting the data write request is when the job is completed on the reading side.

  FIG. 10 is a diagram for explaining reading of band data from the RAM 3 in the second embodiment. As shown in FIG. 10, the band data is stored in a predetermined shared memory area in the RAM 3, and the physical address of this shared memory area is mapped to the logical address of each process 21, 22. Each thread 31a, 31b, 32a, 32b can access the shared memory area.

  As described above, according to the second embodiment, the main controller 2 (i) stores one band data in the RAM 3 in the writing side process 21 and (ii) one in the writing side thread 31a. An event is issued each time a band data write request is queued. (Iv) When the band data write to the HDD 4 is completed, the read-side thread 32a erases the band data from the RAM 3 ( v) When sequentially executing the queued data read requests, if the band data to be read is in the RAM 3, the file read thread 32b reads the band data from the RAM 3 without reading the band data from the HDD 4, vi) If the band data is not stored in the RAM 3, the HDD controller 14 is caused to read the band data from the HDD 4.

  As a result, if band data remains in the RAM 3, the band data on the RAM 3 can be used immediately. For this reason, data can be read out in a shorter time than reading from the HDD 4, and a delay in data reading can be reduced.

  Further, according to the second embodiment, the main controller 2 deletes the data write request for the band data read from the RAM 3 from the queue at a predetermined timing.

  Thereby, it is possible to avoid a processing delay due to a data write request remaining in the queue even though the band data is read from the RAM 3 without going through the HDD 4.

  Each embodiment described above is a preferred example of the present invention, but the present invention is not limited to these, and various modifications and changes can be made without departing from the scope of the present invention. It is.

  For example, in the second embodiment, when the band data is read from the RAM 3 by the reading side thread 32a and the writing of the band data to the HDD 4 is completed on the writing side, the main controller 2 In the writing thread 31a, the band data may be deleted from the RAM 3 after the reading of the band data in the reading thread 32a is completed. For example, such processing can be realized by exclusively controlling access to the storage area of the band data.

  In each of the above embodiments, data may be compressed before data is written to the HDD 4, and data may be decompressed after data is read from the HDD 4. As a compression / decompression method, a JPEG (Joint Photographic Experts Group) method, an RLE (Run Length Encoding) method, or the like can be used.

  In each of the above embodiments, a file handle may be used as a file identifier notified from the writing side to the reading side.

  In the above embodiments, the file system 61 is used to read / write data from / to the HDD 4, but the file system 61 may not be provided. In that case, identification information (for example, address and size) of a storage area on the HDD 4 that stores band data for one page is used instead of the file identifier.

  In each of the above embodiments, the application of the writing side process 21 and the application of the reading side process 22 may be separate and independent programs.

  In each of the above embodiments, the HDD 4 is used as a storage, but a storage device using a semiconductor memory such as an SSD (Solid State Drive) may be used.

  In each of the above embodiments, there is one process on the writing side and one on the reading side, but a plurality of similar processes may be generated on both the writing side and the reading side. Even in that case, one queue system 91 in the first and second embodiments is sufficient.

  In each of the above embodiments, the single memory controller 2 processes both the write side and the read side. However, in the processor having a plurality of processing elements, the write side and the read side are processed. You may make it process a process separately.

  In each of the above embodiments, image data divided into band data is processed. However, the present invention divides other types of data into blocks of a predetermined size, and performs processing for each block. It is applicable also to the apparatus which performs.

  The present invention can be applied to, for example, an image processing apparatus in an image forming apparatus such as a copier or a multifunction peripheral.

2 Main controller (example of processor)
3 RAM (an example of memory)
4 HDD (an example of storage)
8 Data processing program 21 Write process 22 Read process 31a Write thread (an example of a first thread)
31b File writing thread (example of third thread)
32a Reading-side thread (an example of a second thread)
32b File reading thread (an example of a fourth thread)
41 Write side program 51 Read side program

Claims (7)

A storage controller that writes a series of data blocks constituting a predetermined unit of data to the storage and reads the series of data blocks from the storage; and
A processor for controlling the storage controller;
The processor is
Create a writer process and a reader process,
Creating a first thread and a third thread in the writing process;
Generating a second thread and a fourth thread in the reading process;
After notifying the storage area identifier in the storage from the first thread to the second thread, the first thread sequentially queues data write requests for each of the series of data blocks,
Using the third thread to cause the storage controller to write the series of data blocks in the order of the queued data write requests;
When the second thread receives the identifier, it queues data read requests for each of the series of data blocks in sequence,
Using the fourth thread to cause the storage controller to read the series of data blocks in the order of the queued data read requests;
A data processing apparatus. A memory for storing the series of data blocks;
The series of a plurality of data blocks is a plurality of band data constituting image data of one page in one document,
The memory stores page node data indicating each page in the document;
The plurality of band data is a link list provided for each page in the document,
The page node data of a page has a link to the head of the plurality of band data for the page;
The data processing apparatus according to claim 1.   The processor uses the third thread to identify the head position of the plurality of band data with the page node data, and traces the link list to cause the storage controller to write the plurality of band data. The data processing apparatus according to claim 2.   The processor secures a storage area for the plurality of band data when storing the plurality of band data for one page in the memory, and releases the storage area when the writing of the page is completed. 3. The data according to claim 2, wherein when the remaining amount is insufficient, a storage area for the plurality of band data is secured when the storage area for the plurality of band data for other pages is released. Processing equipment. The processor is
Storing the data block in a memory in the third thread;
Issue an event each time the first thread queues a data write request for the data block;
When the writing of the data block to the storage is completed, the third thread erases the data block from the memory,
When executing the queued data read requests in sequence, if the data block exists in the memory, the fourth thread reads the data block from the memory, and if the data block does not exist in the memory, Using the fourth thread, the storage controller is caused to read the data block from the storage,
Deleting the data write request for the data block read from the memory at a predetermined timing from the queue;
The data processing apparatus according to claim 1.   6. The data processing apparatus according to claim 5, wherein when the job is completed, the processor deletes the data write request for the data block read from the memory from the queue. The series of data blocks is written to a file in the storage;
The identifier of the storage area is the identifier of the file;
The data processing device according to claim 1, wherein the data processing device is a data processing device.

Images