CN102855194A - Data storage method and storer - Google Patents
Data storage method and storer Download PDFInfo
- Publication number
- CN102855194A CN102855194A CN2012102810428A CN201210281042A CN102855194A CN 102855194 A CN102855194 A CN 102855194A CN 2012102810428 A CN2012102810428 A CN 2012102810428A CN 201210281042 A CN201210281042 A CN 201210281042A CN 102855194 A CN102855194 A CN 102855194A
- Authority
- CN
- China
- Prior art keywords
- data
- subclauses
- clauses
- compose buffer
- merging
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Images
Landscapes
- Memory System (AREA)
- Memory System Of A Hierarchy Structure (AREA)
Abstract
The invention discloses a data storage method and a storer. The method comprises the following steps of: receiving data source data to be stored, wherein the data source data carries source address information; according to a preset selection mode, selecting a first stage of write buffer corresponding to the data source data; comparing the source address information with the current first address information in the first stage of write buffer; and if the source address information is matched with the current first address information, writing the data source data into the first stage of write buffer, otherwise, transferring data in the first stage of write buffer into a corresponding combined item in a second stage of write buffer, then writing the data source data into the first stage of write buffer, and updating the first address information to the source address information. By the embodiment, the speed of the storer for processing the written data can be increased, so that the data processing efficiency is improved.
Description
Technical field
The present invention relates to technical field of data processing, more particularly, relate to a kind of date storage method and storer.
Background technology
Storer is used for store data and program, is the important composition parts of computing machine.The multi-level hierarchy of storer is so that storage operation needs first data to be put on the low one-level memory hierarchy i.e.: compose buffer (write buffer).In the prior art, a kind of implementation of compose buffer is: merge compose buffer (combining write buffer).
For merging compose buffer, when compose buffer was sky, data and the address that will write compose buffer were put in the compose buffer, and write operation namely finishes; If the address is then checked in existing data and the address that writes first in the compose buffer, the address of judging new data to be written whether with buffer zone in a certain item address of data with existing be complementary.If coupling, new data just merges in the data field of matching addresses item correspondence, and therefore, this process is called " writing merging ", and coupling refers to that the address is continuous herein.
Table 1 has adopted writes merging, four groups of data storages on the continuation address is merged in the merging clauses and subclauses of compose buffer.Wherein, write address comprises: 100,108,116 and 124, and these four addresses are continuous address.
| Write address | V | The data field | V | The data field | V | The data field | V | The |
| 100 | 1 | Data [100] | 1 | Data [108] | 1 | Data [116] | 1 | Data [124] |
| 0 | 0 | 0 | 0 | |||||
| 0 | 0 | 0 | 0 | |||||
| 0 | 0 | 0 | 0 |
Table 1
Can find, comprise a plurality of data fields in every merging clauses and subclauses of merging compose buffer, because not necessarily the address is continuous for the data that write, so merging clauses and subclauses (entry), each produces easily the empty word that occurs owing to there not being data to write, i.e. empty data field.Shown in following table 2, wherein the write address of data [100] and data [108] is 100, and the write address of data [172] is 172, therefore, the address of these two groups of data is discontinuous, is writing when merging so that exist null data area to cause the storage space utilization factor low in the buffer zone according to existing scheme.
| Write address | V | The data field | V | The data field | V | The data field | V | The |
| 100 | 1 | Data [100] | 1 | Data [108] | 0 | 0 | ||
| 172 | 1 | Data [172] | 0 | 0 | 0 | |||
| 0 | 0 | 0 | 0 | |||||
| 0 | 0 | 0 | 0 |
Table 2
In addition, when data are write in a plurality of data source applications of appearance, the existing compose buffer that merges once can only be processed the data that a data source writes, and the application that writes of this data source replied, the data source of application before can only waiting for that write of other data source write data and just can be obtained replying after complete.Therefore, the processing speed of storer is slow, and treatment effeciency is low.
Summary of the invention
In view of this, the embodiment of the invention provides a kind of date storage method and storer, can improve storer to the processing speed of data writing, and then improves data-handling efficiency.
The embodiment of the invention provides a kind of date storage method, comprising:
Receive data source data to be stored, carry source address information in the described data source data;
According to the selection mode that sets in advance, select the first order compose buffer corresponding with described data source data;
Current the first address information in described source address information and the described first order compose buffer is compared; If the two coupling is then write described data source data and is incorporated in the described first order compose buffer;
Otherwise, the data in the described first order compose buffer are transferred in the corresponding merging clauses and subclauses of second level compose buffer, described data source data is write be incorporated into described first order compose buffer again, described the first address information is updated to source address information.
Preferably, in the described corresponding merging clauses and subclauses that data in the described first order compose buffer are transferred to second level compose buffer, comprising:
The address information that described the first address information is corresponding with each bar merging clauses and subclauses in the compose buffer of the described second level compares; If have the first merging clauses and subclauses that are complementary with described the first address information in the compose buffer of the described second level, then the data in the described first order compose buffer write described first and merge clauses and subclauses; Otherwise, according to presetting rule, data in the described first order compose buffer are transferred to data field in the compose buffer of the described second level in the empty merging clauses and subclauses, and address information corresponding to these merging clauses and subclauses in the compose buffer of the described second level is updated to described the first address information.
Preferably, the data in the described first order compose buffer are transferred to data field in the compose buffer of the described second level in the empty merging clauses and subclauses, comprise:
With the data in the described first order compose buffer be transferred in the compose buffer of the described second level numbering minimum or numbering is maximum and the data field in the empty merging clauses and subclauses, wherein, respectively merge clauses and subclauses in the compose buffer of the described second level and be assigned respectively continuous numbering.
Preferably, described method also comprises:
Write in the merging clauses and subclauses that are incorporated into the data board unit data that have in the compose buffer of the described second level in the merging clauses and subclauses of same identification information and output, wherein, described identification information is used for identifying the merging clauses and subclauses of described second level compose buffer, and the identification information of the continuous merging clauses and subclauses in address is identical.
Preferably, the data that have in the compose buffer of the described second level in the merging clauses and subclauses of same identification information are write in the merging clauses and subclauses that are incorporated into the data board unit, being comprised:
The data that have in the compose buffer of the described second level in the merging clauses and subclauses of same identification information are sorted according to the precedence of address information, according to mode from right to left, write in the merging clauses and subclauses that are incorporated into described data board unit.
A kind of storer comprises:
The data receiver unit is used for receiving data source data to be stored, carries source address information in the described data source data;
Selected cell is used for selecting the first order compose buffer corresponding with described data source data according to the selection mode that sets in advance;
The first address comparing unit is used for current first address information of described source address information and described first order compose buffer is compared, if the two coupling then triggers first and writes the merging control module; Otherwise, trigger second and write the merging control module;
Described first writes the merging control module, is incorporated into described first order compose buffer for described data source data is write;
Second writes the merging control module, be used for the data in the described first order compose buffer are transferred to the corresponding merging clauses and subclauses of second level compose buffer, again described data source data is write and be incorporated into described first order compose buffer, described the first address information is updated to source address information.
Preferably, described second writes the merging control module, comprising:
The second address is subelement relatively, being used for that each bar of described the first address information and described second level compose buffer is merged address information corresponding to clauses and subclauses compares, if exist first of matching addresses to merge clauses and subclauses in the compose buffer of the described second level, then trigger the first control subelement; Otherwise, trigger the second control subelement;
Described the first control subelement is used for that the data in the described first order compose buffer are write described first and merges clauses and subclauses;
Described the second control subelement, be used for according to presetting rule, data in the described first order compose buffer are transferred to data field in the compose buffer of the described second level in the empty merging clauses and subclauses, and address information corresponding to these merging clauses and subclauses in the compose buffer of the described second level is updated to described the first address information.
Preferably, also comprise:
The numbering allocation units are used to respectively to merge clauses and subclauses in the compose buffer of the described second level and distribute respectively continuous number information;
Then described the second control subelement specifically is used for the data in the described first order compose buffer are transferred to described second level compose buffer numbering minimum or numbering maximum and data field in the empty merging clauses and subclauses.
Preferably, also comprise:
The identification information allocation units are used to the continuous identical identification information of merging clauses and subclauses distribution in address in the compose buffer of the described second level;
The data board control module is write in the merging clauses and subclauses that are incorporated into the data board unit for the data in the merging clauses and subclauses that described second level compose buffer had same identification information;
Output control unit is used for the data output with the merging clauses and subclauses of described data board unit.
Preferably, described data board control module, sort according to the precedence of address information by the data that will have in the compose buffer of the described second level in the merging clauses and subclauses of same identification information, according to mode from right to left, write in the merging clauses and subclauses that are incorporated into described data board unit.
In the embodiment of the invention, the compose buffer refinement of storer is divided into a plurality of first order compose buffers corresponding with data source and a second level compose buffer, because corresponding first order compose buffer of each data source, therefore, the application of data source write store can obtain the Quick-return of storer, and storer can significantly improve processing speed and the treatment effeciency of data writing;
In addition, by at second level compose buffer by the number that merges clauses and subclauses specifically is set, realize the thin width that it merges clauses and subclauses of dividing, increase the dirigibility that clauses and subclauses merge in the compose buffer of the second level, avoid owing to every merging clauses and subclauses produce the low defective of storage space utilization factor that empty word causes easily.
Description of drawings
In order to be illustrated more clearly in the technical scheme of the embodiment of the invention, the below will do to introduce simply to the accompanying drawing of required use in embodiment or the description of the Prior Art, apparently, accompanying drawing in the following describes only is some embodiments of the present invention, for those of ordinary skills, under the prerequisite of not paying creative work, can also obtain according to these accompanying drawings other accompanying drawing.
A kind of date storage method schematic flow sheet that Fig. 1 provides for the embodiment of the invention;
The another kind of date storage method schematic flow sheet that Fig. 2 provides for the embodiment of the invention;
Fig. 3 is the structural representation of first order compose buffer in the embodiment of the invention;
Fig. 4 is structural representation Fig. 4 who merges clauses and subclauses in the compose buffer of the second level in the embodiment of the invention;
Fig. 5 is the synoptic diagram when in the embodiment of the invention first order compose buffer being write second level compose buffer;
Fig. 6 is the synoptic diagram when in the embodiment of the invention data in the compose buffer of the second level being write the next stage storer;
Fig. 7 is a kind of memory construction synoptic diagram provided by the invention;
Fig. 8 is another kind of memory construction synoptic diagram provided by the invention;
Fig. 9 is another memory construction synoptic diagram provided by the invention.
Embodiment
Below in conjunction with the accompanying drawing in the embodiment of the invention, the technical scheme in the embodiment of the invention is clearly and completely described, obviously, described embodiment only is the present invention's part embodiment, rather than whole embodiment.Based on the embodiment among the present invention, those of ordinary skills belong to the scope of protection of the invention not making the every other embodiment that obtains under the creative work prerequisite.
At first a kind of date storage method provided by the invention is described, described storer is divided into a plurality of first order compose buffers and a second level compose buffer, described first order compose buffer comprises that unique one merges clauses and subclauses, described second level compose buffer comprises that many merge clauses and subclauses, as shown in Figure 1, described method comprises:
In this step, storer at the corresponding levels is write in a plurality of data source applications;
The selection mode that step 102, basis set in advance is selected the first order compose buffer corresponding with described data source data;
Among this embodiment, the corresponding first order compose buffer of each data source, therefore, the first order compose buffer that data source can be write direct corresponding with this data source; Wherein, the selection mode that sets in advance refers to exist between data source and the first order compose buffer one to one relation;
In this step, in the first order compose buffer of source address information that data source data is carried and corresponding described data source, merge the current address information (i.e. the first address information) of clauses and subclauses when not mating, need first the existing data in this first order compose buffer to be transferred in the corresponding merging clauses and subclauses of second level compose buffer, data source data is write be incorporated into this first order compose buffer again.The source address information of carrying with data source data simultaneously, is upgraded the first current address information of this first order compose buffer.
In the embodiment of the invention, the compose buffer refinement of storer is divided into a plurality of first order compose buffers corresponding with data source and a second level compose buffer, because the corresponding first order compose buffer of each data source is so the application of data source write store can obtain the Quick-return of storer.In addition, by at second level compose buffer by the number that merges clauses and subclauses specifically is set, realize the thin width that it merges clauses and subclauses of dividing, increase the dirigibility that clauses and subclauses merge in the compose buffer of the second level, avoid owing to every merging clauses and subclauses produce the low defective of storage space utilization factor that empty word causes easily.
In above-described embodiment, in the described corresponding merging clauses and subclauses that data in the first order compose buffer of the described data source of correspondence are transferred to second level compose buffer, comprising:
The address information that described the first address information is corresponding with each bar merging clauses and subclauses in the compose buffer of the described second level compares; If have the first merging clauses and subclauses that are complementary with described the first address information in the compose buffer of the described second level, then the data in the described first order compose buffer write described first and merge clauses and subclauses; Otherwise, according to presetting rule, data in the described first order compose buffer are transferred to data field in the compose buffer of the described second level in the empty merging clauses and subclauses, and address information corresponding to these merging clauses and subclauses in the compose buffer of the described second level is updated to described the first address information.
Need to prove that address information is complementary and can comprises two kinds of situations: address information is identical, and perhaps, address information is continuous herein.After the data of a plurality of data sources write respectively in the corresponding first order compose buffer, the data in a plurality of first order compose buffers also can sort successively and write second level compose buffer, and write merging in the compose buffer of the second level.During implementation, can be the identification information that each first order compose buffer distributes unique this buffer zone of sign, for example number, can distribute respectively continuous numbering for a plurality of first order compose buffers, thus, can according to the numbering successively order of ordering, the data in a plurality of first order compose buffers be write second level compose buffer successively.
In above-described embodiment, described data in the described first order compose buffer are write data field in the compose buffer of the described second level in the empty merging clauses and subclauses, this implementation can for:
With the data in the described first order compose buffer be transferred in the compose buffer of the described second level numbering minimum or numbering is maximum and the data field in the empty merging clauses and subclauses, wherein, respectively merge clauses and subclauses in the compose buffer of the described second level and be assigned respectively continuous numbering.
Among this embodiment, for a plurality of merging clauses and subclauses in the compose buffer of the described second level are assigned respectively continuous numbering, according to numbering, can conveniently identify the specific clauses and subclauses that merge in the compose buffer of the described second level.
In another embodiment, as shown in Figure 2, described date storage method is comprising that above-mentioned steps 101, can also may further comprise the steps outside the described content to step 105:
During implementation, the data that have in the compose buffer of the described second level in the merging clauses and subclauses of same identification information are write in the merging clauses and subclauses that are incorporated into the data board unit, its implementation can for:
The data that have in the compose buffer of the described second level in the merging clauses and subclauses of same identification information are sorted according to the precedence of address information, according to mode from right to left, write in the merging clauses and subclauses that are incorporated into described data board unit.
Among this embodiment, when writing single-level memory, merged at second level compose buffer on the basis of clauses and subclauses and carried out remerging the second time merging clauses and subclauses by the data board unit, write the next stage storer by the transmission of memory bus, reduce the number of times of writing single-level memory, taken full advantage of memory bus.
For the ease of the understanding to above-mentioned matching algorithm, below by concrete example overall plan is described in detail explanation.
As shown in Figure 3, structural representation for first order compose buffer in the embodiment of the invention, first order compose buffer comprises that unique one merges clauses and subclauses, every the data field that merges clauses and subclauses can comprise eight 32 (bit) word (word, 1word=32bit), each byte (byte, 1byte=8bit) has one to write dirty position " D ", writes dirty position " D " and represents data writing of corresponding byte.Every merges the write address (8-word align address) that clauses and subclauses have an Eight characters alignment, and this write address is identifying the address information of corresponding data writing.
As shown in Figure 4, be structural representation that merges clauses and subclauses in the second level compose buffer in the embodiment of the invention, can comprise in the compose buffer of the second level that many merge clauses and subclauses, the concrete merging clauses and subclauses that have 16 same structures.Every the data field that merges clauses and subclauses can comprise eight 32 words, and each byte has one to write dirty position " D ".In addition, can distribute a significance bit " V(Value) " for per four words, whether this significance bit " V " is used for indicating four words and writes, and writes fashionablely when four words, and the V value is 1; Otherwise the V value is 0.Every merges clauses and subclauses and shares the address (8-word align address) of an Eight characters alignment and one four mark (tag).Whole second level compose buffer has one four write pointer, one four read pointer.
Be understandable that, in the embodiment of the invention, in the compose buffer of the second level each merges in the clauses and subclauses and comprises a data field, the technical scheme that comprises a plurality of data fields with the merging clauses and subclauses that have a certain write address in the prior art is compared, the width that in the compose buffer of the second level each merges clauses and subclauses reduces, therefore, the probability that occurs empty word in the compose buffer of the second level in each merging clauses and subclauses can correspondingly reduce.
As seen, by merging the width of clauses and subclauses in the compose buffer of the thin division second level, increase the dirigibility that clauses and subclauses merge in the compose buffer of the second level, avoid owing to every merging clauses and subclauses produce the low defective of storage space utilization factor that empty word causes easily.
In the embodiment of the invention, when data source is write storer at the corresponding levels, at first write first order compose buffer, if data source address and first order compose buffer address are complementary, then write and merge in the first order compose buffer.If does not mate data source address and first order compose buffer address, then the data of first order compose buffer to be write in the compose buffer of the second level first, and then data source data write in the first order compose buffer, use simultaneously " write address " of data source to upgrade current " write address " of first order compose buffer.
First order compose buffer is when writing second level compose buffer, first with " write address " of first order compose buffer simultaneously with second level compose buffer in 16 " write address " relatively, if " write address " of first order compose buffer a certain " write address " item with second level compose buffer is complementary, then the data with first order compose buffer write data field corresponding to this clauses and subclauses address; If " write address " of first order compose buffer 16 " write address " item with second level compose buffer does not mate, then the data of first order compose buffer being written to 16, to merge numbering in the clauses and subclauses minimum or numbering is maximum and value significance bit " V " is in 0 the merging clauses and subclauses, upgrades simultaneously this and merge objective and be designated as current write pointer value.And then write pointer value added 1.
The below describes the process that first order compose buffer is write second level compose buffer in detail.As shown in Figure 5, write the instance graph of second level compose buffer for first order compose buffer.Every " write address " that merges clauses and subclauses is the address of an Eight characters alignment in the compose buffer of the second level.When the address of the 32 words alignment that merges clauses and subclauses was identical, their mark was identical, so the merging clauses and subclauses of same tag should have at most four merging clauses and subclauses.Be numbered 0,4,5,7 merging clauses and subclauses, their mark all is 0.Be numbered 1,3,9 merging clauses and subclauses, their mark all is 1.Be numbered 2,6,8 merging clauses and subclauses, their mark all is 2.
Situation shown in Fig. 5 arrow (1), " write address " item of first order compose buffer is 140, when 16 " write address " item with second level compose buffer compares, be complementary with the merging clauses and subclauses that are numbered 6, then the data of first order compose buffer can be write and be numbered 6 data field corresponding to merging clauses and subclauses in the compose buffer of the second level.
Situation shown in Fig. 5 arrow (2), wherein italics represents that first order compose buffer writes that to be numbered 9 merging clauses and subclauses before the compose buffer of the second level every for empty, " write address " item of first order compose buffer is 180, when 16 " write address " item with second level compose buffer compares, find that not merging clauses and subclauses is complementary, but " write address " item 180 and coding 1, " write address " of 3 merging clauses and subclauses is the alignment of 32 words, so it is in 0 the merging clauses and subclauses that the data of first order compose buffer are written to 16 values that merge the minimum merging clauses and subclauses of numbering in the clauses and subclauses and significance bit " V ", namely be numbered in 9 the merging clauses and subclauses, " mark " that will be numbered simultaneously in 9 the merging clauses and subclauses puts 1, and " write address " puts 180.
Situation shown in Fig. 5 arrow (3), wherein "-" expression first order compose buffer writes that to merge clauses and subclauses before the compose buffer of the second level every for empty, " write address " item of first order compose buffer is 196, when 16 " write address " item with second level compose buffer compares, find that not merging clauses and subclauses is complementary, and and 16 " write address " also non-32 words alignment that merge clauses and subclauses, therefore, it is in 0 the merging clauses and subclauses that the data of first order compose buffer are written to 16 values that merge the minimum merging clauses and subclauses of numbering in the clauses and subclauses and significance bit " V ", namely be numbered in 10 the merging clauses and subclauses, simultaneously " mark " item is put 3, the value that is current write pointer is 3, " write address " item puts 196, again write pointer value is added 1.
Among this embodiment, the numerical range of write pointer is 0 ~ 15, and write pointer value is 0 during initialization, and the value of write pointer is that current 16 values that merge " mark " maximum in the clauses and subclauses add 1.When the value of write pointer was 15, the value that increases again 1 write pointer was returned as 0.
The numerical range of mark is 0 ~ 15, and the value of initialization tense marker is 0, and when the value of mark was 15, the value that increases again 1 mark was returned as 0.
In the time the data in the compose buffer of the second level need to being write the next stage storer, the data in the corresponding merging clauses and subclauses in the compose buffer of the second level need to be write and be incorporated in the data board unit.As shown in Figure 6, the value of current read pointer is 0, compare with 16 marks that merge clauses and subclauses, with 16 merge the numbering that is labeled as 0 merging clauses and subclauses in the clauses and subclauses according to the value ordering that merges " write address " in the clauses and subclauses according to ascending, put into the data board unit with mode from right to left.
The data board unit comprises that unique one merges clauses and subclauses, the maximum data width that the data width of these mergings clauses and subclauses can merge according to merging clauses and subclauses in the compose buffer of the described second level and setting.When memory bus is supported once writing out of 32 words, then the data of data board record can be at most the data of 32 words.
Among this embodiment, the numerical range of read pointer is 0 ~ 15, and the read pointer value is 0 during initialization, and when the value of read pointer was 15, the value that increases again 1 read pointer was returned as 0.
It is pointed out that in the embodiment of the invention that the figure place of the word that the merging clauses and subclauses can comprise at most can be set according to practical application, to this, the embodiment of the invention is not specifically limited.
Corresponding above-mentioned date storage method embodiment, the present invention also provides a kind of storer, described storer is divided into a plurality of first order compose buffers and a second level compose buffer, described first order compose buffer comprises that unique one merges clauses and subclauses, described second level compose buffer comprises that many merge clauses and subclauses, as shown in Figure 7, described storer can comprise:
The first address comparing unit 703 is used for current first address information of described source address information and described first order compose buffer is compared, if the two coupling then triggers first and writes merging control module 704; Otherwise, trigger second and write merging control module 705;
Described first writes merging control module 704, is incorporated into described first order compose buffer for described data source data is write;
Second writes merging control module 705, be used for the data in the described first order compose buffer are transferred to the corresponding merging clauses and subclauses of second level compose buffer, again described data source data is write and be incorporated into described first order compose buffer, described the first address information is updated to source address information.
In the embodiment of the invention, the compose buffer refinement of storer is divided into a plurality of first order compose buffers corresponding with data source and a second level compose buffer, because the corresponding first order compose buffer of each data source is so the application of data source write store can obtain the Quick-return of storer; In addition, by at second level compose buffer by the number that merges clauses and subclauses specifically is set, realize the thin width that it merges clauses and subclauses of dividing, increase the dirigibility that clauses and subclauses merge in the compose buffer of the second level, avoid owing to every merging clauses and subclauses produce the low defective of storage space utilization factor that empty word causes easily.
Wherein, described second writes the merging control module, can comprise:
The second address is subelement relatively, being used for that each bar of described the first address information and described second level compose buffer is merged address information corresponding to clauses and subclauses compares, if exist first of matching addresses to merge clauses and subclauses in the compose buffer of the described second level, then trigger the first control subelement; Otherwise, trigger the second control subelement;
Described the first control subelement is used for that the data in the described first order compose buffer are write described first and merges clauses and subclauses;
Described the second control subelement, be used for according to presetting rule, data in the described first order compose buffer are transferred to data field in the compose buffer of the described second level in the empty merging clauses and subclauses, and address information corresponding to these merging clauses and subclauses in the compose buffer of the described second level is updated to described the first address information.
In another kind of embodiment provided by the invention, as shown in Figure 8, storer can also comprise:
Then described second control subelement, be used for the data in the described first order compose buffer are transferred to described second level compose buffer numbering minimum or numbering maximum and data field in the empty merging clauses and subclauses.
In another embodiment provided by the invention, as shown in Figure 9, storer can also comprise:
Identification information allocation units 707 are used to the continuous identical identification information of merging clauses and subclauses distribution in address in the compose buffer of the described second level;
Data board control module 708 is write in the merging clauses and subclauses that are incorporated into the data board unit for the data in the merging clauses and subclauses that described second level compose buffer had same identification information;
Implementation extremely, described data board unit can comprise that unique one merges clauses and subclauses, the maximum data width that the data width of these mergings clauses and subclauses can merge according to merging clauses and subclauses in the compose buffer of the described second level and setting.
In above-described embodiment, described data board control module, sort according to the precedence of address information by the data that will have in the compose buffer of the described second level in the merging clauses and subclauses of same identification information, according to mode from right to left, write in the merging clauses and subclauses that are incorporated into described data board unit.
For device embodiment, because it is substantially corresponding to embodiment of the method, so describe fairly simplely, relevant part gets final product referring to the part explanation of embodiment of the method.Device embodiment described above only is schematic, wherein said unit as the separating component explanation can or can not be physically to separate also, the parts that show as the unit can be or can not be physical locations also, namely can be positioned at a place, perhaps also can be distributed on a plurality of network element.Can select according to the actual needs wherein some or all of module to realize the purpose of present embodiment scheme.Those of ordinary skills namely can understand and implement in the situation of not paying creative work.
One of ordinary skill in the art will appreciate that all or part of flow process that realizes in above-described embodiment method, to come the relevant hardware of instruction to finish by computer program, described program can be stored in the computer read/write memory medium, this program can comprise the flow process such as the embodiment of above-mentioned each side method when carrying out.Wherein, described storage medium can be magnetic disc, CD, read-only store-memory body (Read-Only Memory, ROM) or store-memory body (Random Access Memory, RAM) etc. at random.
To the above-mentioned explanation of the disclosed embodiments, make this area professional and technical personnel can realize or use the present invention.Multiple modification to these embodiment will be apparent concerning those skilled in the art, and General Principle as defined herein can in the situation of the spirit or scope that do not break away from the embodiment of the invention, realize in other embodiments.Therefore, the embodiment of the invention will can not be restricted to these embodiment shown in this article, but will meet the widest scope consistent with principle disclosed herein and features of novelty.
Claims (10)
1. a date storage method is characterized in that, comprising:
Receive data source data to be stored, carry source address information in the described data source data;
According to the selection mode that sets in advance, select the first order compose buffer corresponding with described data source data;
Current the first address information in described source address information and the described first order compose buffer is compared; If the two coupling is then write described data source data and is incorporated in the described first order compose buffer;
Otherwise, the data in the described first order compose buffer are transferred in the corresponding merging clauses and subclauses of second level compose buffer, described data source data is write be incorporated into described first order compose buffer again, described the first address information is updated to source address information.
2. date storage method according to claim 1 is characterized in that, in the described corresponding merging clauses and subclauses that data in the described first order compose buffer are transferred to second level compose buffer, comprising:
The address information that described the first address information is corresponding with each bar merging clauses and subclauses in the compose buffer of the described second level compares; If have the first merging clauses and subclauses that are complementary with described the first address information in the compose buffer of the described second level, then the data in the described first order compose buffer write described first and merge clauses and subclauses; Otherwise, according to presetting rule, data in the described first order compose buffer are transferred to data field in the compose buffer of the described second level in the empty merging clauses and subclauses, and address information corresponding to these merging clauses and subclauses in the compose buffer of the described second level is updated to described the first address information.
3. date storage method according to claim 2 is characterized in that, the data in the described first order compose buffer are transferred to data field in the compose buffer of the described second level in the empty merging clauses and subclauses, comprising:
With the data in the described first order compose buffer be transferred in the compose buffer of the described second level numbering minimum or numbering is maximum and the data field in the empty merging clauses and subclauses, wherein, respectively merge clauses and subclauses in the compose buffer of the described second level and be assigned respectively continuous numbering.
4. each described date storage method is characterized in that according to claim 1-3, and described method also comprises:
Write in the merging clauses and subclauses that are incorporated into the data board unit data that have in the compose buffer of the described second level in the merging clauses and subclauses of same identification information and output, wherein, described identification information is used for identifying the merging clauses and subclauses of described second level compose buffer, and the identification information of the continuous merging clauses and subclauses in address is identical.
5. date storage method according to claim 4 is characterized in that, the data that have in the compose buffer of the described second level in the merging clauses and subclauses of same identification information are write in the merging clauses and subclauses that are incorporated into the data board unit, comprising:
The data that have in the compose buffer of the described second level in the merging clauses and subclauses of same identification information are sorted according to the precedence of address information, according to mode from right to left, write in the merging clauses and subclauses that are incorporated into described data board unit.
6. a storer is characterized in that, comprising:
The data receiver unit is used for receiving data source data to be stored, carries source address information in the described data source data;
Selected cell is used for selecting the first order compose buffer corresponding with described data source data according to the selection mode that sets in advance;
The first address comparing unit is used for current first address information of described source address information and described first order compose buffer is compared, if the two coupling then triggers first and writes the merging control module; Otherwise, trigger second and write the merging control module;
Described first writes the merging control module, is incorporated into described first order compose buffer for described data source data is write;
Second writes the merging control module, be used for the data in the described first order compose buffer are transferred to the corresponding merging clauses and subclauses of second level compose buffer, again described data source data is write and be incorporated into described first order compose buffer, described the first address information is updated to source address information.
7. storer according to claim 6 is characterized in that, described second writes the merging control module, comprising:
The second address is subelement relatively, being used for that each bar of described the first address information and described second level compose buffer is merged address information corresponding to clauses and subclauses compares, if exist first of matching addresses to merge clauses and subclauses in the compose buffer of the described second level, then trigger the first control subelement; Otherwise, trigger the second control subelement;
Described the first control subelement is used for that the data in the described first order compose buffer are write described first and merges clauses and subclauses;
Described the second control subelement, be used for according to presetting rule, data in the described first order compose buffer are transferred to data field in the compose buffer of the described second level in the empty merging clauses and subclauses, and address information corresponding to these merging clauses and subclauses in the compose buffer of the described second level is updated to described the first address information.
8. storer according to claim 7 is characterized in that, also comprises:
The numbering allocation units are used to respectively to merge clauses and subclauses in the compose buffer of the described second level and distribute respectively continuous number information;
Then described the second control subelement specifically is used for the data in the described first order compose buffer are transferred to described second level compose buffer numbering minimum or numbering maximum and data field in the empty merging clauses and subclauses.
9. each described storer is characterized in that according to claim 3-8, also comprises:
The identification information allocation units are used to the continuous identical identification information of merging clauses and subclauses distribution in address in the compose buffer of the described second level;
The data board control module is write in the merging clauses and subclauses that are incorporated into the data board unit for the data in the merging clauses and subclauses that described second level compose buffer had same identification information;
Output control unit is used for the data output with the merging clauses and subclauses of described data board unit.
10. storer according to claim 9, it is characterized in that, described data board control module, sort according to the precedence of address information by the data that will have in the compose buffer of the described second level in the merging clauses and subclauses of same identification information, according to mode from right to left, write in the merging clauses and subclauses that are incorporated into described data board unit.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201210281042.8A CN102855194B (en) | 2012-08-08 | 2012-08-08 | Data storage method and storer |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201210281042.8A CN102855194B (en) | 2012-08-08 | 2012-08-08 | Data storage method and storer |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN102855194A true CN102855194A (en) | 2013-01-02 |
| CN102855194B CN102855194B (en) | 2015-05-13 |
Family
ID=47401795
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201210281042.8A Active CN102855194B (en) | 2012-08-08 | 2012-08-08 | Data storage method and storer |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN102855194B (en) |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2014127650A1 (en) * | 2013-02-22 | 2014-08-28 | 中兴通讯股份有限公司 | Data protection method, apparatus and device |
| CN104272274B (en) * | 2013-12-31 | 2017-06-09 | 华为技术有限公司 | Data processing method and equipment in a kind of distributed file storage system |
| CN113743569A (en) * | 2020-05-29 | 2021-12-03 | 上海新氦类脑智能科技有限公司 | Pulse signal transmission method, device and storage medium |
| CN114731282A (en) * | 2019-11-22 | 2022-07-08 | 华为技术有限公司 | Method, buffer and node for processing non-buffer write data request |
| WO2023232079A1 (en) * | 2022-05-31 | 2023-12-07 | 上海寒武纪信息科技有限公司 | Data storage method, data access method, data computing method, and related product |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20090083499A1 (en) * | 2007-09-24 | 2009-03-26 | Nvidia Corporation | Ordered Storage Structure Providing Enhanced Access to Stored Items |
| CN101739353A (en) * | 2008-11-06 | 2010-06-16 | 瑞昱半导体股份有限公司 | Adaptive buffer device and method thereof |
| CN102045258A (en) * | 2010-12-22 | 2011-05-04 | 北京星网锐捷网络技术有限公司 | Data caching management method and device |
| CN101393537B (en) * | 2008-09-18 | 2012-06-27 | 北京中星微电子有限公司 | NandFlash amortization management method |
-
2012
- 2012-08-08 CN CN201210281042.8A patent/CN102855194B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20090083499A1 (en) * | 2007-09-24 | 2009-03-26 | Nvidia Corporation | Ordered Storage Structure Providing Enhanced Access to Stored Items |
| CN101393537B (en) * | 2008-09-18 | 2012-06-27 | 北京中星微电子有限公司 | NandFlash amortization management method |
| CN101739353A (en) * | 2008-11-06 | 2010-06-16 | 瑞昱半导体股份有限公司 | Adaptive buffer device and method thereof |
| CN102045258A (en) * | 2010-12-22 | 2011-05-04 | 北京星网锐捷网络技术有限公司 | Data caching management method and device |
Cited By (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2014127650A1 (en) * | 2013-02-22 | 2014-08-28 | 中兴通讯股份有限公司 | Data protection method, apparatus and device |
| US10191670B2 (en) | 2013-02-22 | 2019-01-29 | Xi'an Zhongxing New Software Co., Ltd. | Method and device of data protection, storage equipment |
| CN104272274B (en) * | 2013-12-31 | 2017-06-09 | 华为技术有限公司 | Data processing method and equipment in a kind of distributed file storage system |
| US10127233B2 (en) | 2013-12-31 | 2018-11-13 | Huawei Technologies Co., Ltd. | Data processing method and device in distributed file storage system |
| CN114731282A (en) * | 2019-11-22 | 2022-07-08 | 华为技术有限公司 | Method, buffer and node for processing non-buffer write data request |
| CN114731282B (en) * | 2019-11-22 | 2023-06-02 | 华为技术有限公司 | Method, buffer and node for processing non-buffer write data request |
| CN113743569A (en) * | 2020-05-29 | 2021-12-03 | 上海新氦类脑智能科技有限公司 | Pulse signal transmission method, device and storage medium |
| WO2023232079A1 (en) * | 2022-05-31 | 2023-12-07 | 上海寒武纪信息科技有限公司 | Data storage method, data access method, data computing method, and related product |
Also Published As
| Publication number | Publication date |
|---|---|
| CN102855194B (en) | 2015-05-13 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN102855194B (en) | Data storage method and storer | |
| US12131779B2 (en) | Global responder signal circuitry for memory arrays | |
| CN101582052B (en) | Memory module and method for performing wear-leveling of memory module | |
| CN104636285B (en) | A kind of flash-memory storage system and read-write thereof, delet method | |
| CN101097547B (en) | Power-performance modulation in caches using a smart least recently used scheme | |
| CN102024046B (en) | Data repeatability checking method and device as well as system | |
| CN102483840B (en) | Cache memory relevance is distributed to processor core | |
| CN103383672B (en) | High-speed cache control is to reduce transaction rollback | |
| CN102428455A (en) | Data striping in a flash memory data storage device | |
| CN104813293B (en) | Use the memory management in the dirty mask space of dynamically distributes | |
| CN102023810B (en) | Method and device for writing data and redundant array of inexpensive disk | |
| CN101382873A (en) | Management method and system of storage location | |
| CN105468542A (en) | Address distribution method and device | |
| CN105684020A (en) | An order book management device in a hardware platform | |
| CN101740110A (en) | Method and device for erasure balance of Nand Flash | |
| CN103885893A (en) | Technique For Accessing Content-Addressable Memory | |
| CN103902702A (en) | Data storage system and data storage method | |
| CN107315694A (en) | A kind of buffer consistency management method and Node Controller | |
| CN102971698A (en) | Snapshot data-processing method and system, storage system and snapshot agency | |
| CN106708620A (en) | Data processing method and system | |
| CN101777061A (en) | JAVA card object management method and JAVA card | |
| CN104346458A (en) | Data storage method and device | |
| CN101419838B (en) | Method for prolonging using life of flash | |
| CN107678686B (en) | Method for realizing FT L function of flash memory based on hardware and data storage device thereof | |
| CN105260332A (en) | Method and system for orderly storing CPLD data packets |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant |