WO2015035870A1

WO2015035870A1 - Multiple cpu scheduling method and device

Info

Publication number: WO2015035870A1
Application number: PCT/CN2014/085818
Authority: WO
Inventors: 罗治凡; 彭钰; 陈幼珍
Original assignee: 华为终端有限公司
Priority date: 2013-09-13
Filing date: 2014-09-03
Publication date: 2015-03-19
Also published as: CN104460934A

Abstract

A method and a device for scheduling of multiple CPUs, the method comprising: obtaining one graphics thread from among a number M of graphics threads, and one CPU from among a number N of operating CPUs, M being an integer greater than zero and N being an integer greater than zero; binding the one graphics thread so obtained to the one CPU so obtained for processing. With no impact upon system performance, the resources of each CPU are thus fully utilized while overall power consumption of the multiple CPUs is decreased.

Description

Multi-CPU scheduling method and device

This application claims the Chinese Patent Application No. 20131041955, filed on Sep. 13, 2013, the entire disclosure of which is hereby incorporated by reference.

Technical field

The present invention relates to communication technologies, and in particular, to a multi-CPU scheduling method and apparatus.

Background technique

With the continuous development of CPU technology of the central processing unit, the application of multi-processor technology, for example, Symmetrical Multi-Processing (SMP) technology, is becoming more and more popular. At present, many terminals adopt multi-processor technology. Among the threads processed by the existing terminal multi-processor, the image drawing thread is mainly used as a multi-processor of the terminal, which has a great influence on the power consumption of the multi-processor.

In the actual situation, one drawing thread may jump on different CPUs, which may cause the drawing thread to be affected by other threads on each CPU, making scheduling uneven, which may increase the overall power consumption of the multiprocessor.

Summary of the invention

The embodiment of the invention provides a multi-CPU scheduling method, so that a drawing thread is processed on a running CPU, which can reduce multi-CPU power consumption.

A first aspect of the embodiments of the present invention provides a multi-CPU scheduling method, including:

Obtaining one of the M drawing threads and one of the N running CPUs, where M is an integer greater than 0, and N is an integer greater than 0;

Binding one of the M drawing threads to one of the N running CPUs for processing.

With reference to the first aspect, in the first possible implementation, the other M-1 drawing threads of the M drawing threads and the other N-1 CPUs of the N running CPUs are acquired;

When the M-1 is smaller than the N-1, the one-to-one binding of the M-1 drawing threads is processed on the M-1 running CPUs; or

When the M-1 is equal to the N-1, the one-to-one binding of the M-1 drawing threads is processed on the N-1 running CPUs; or

When the M-1 is greater than the N-1, the one-to-one binding of the N-1 drawing threads is processed on the N-1 running CPUs, and the MN drawing threads are bound to the Processing on MN CPUs on N-1 running CPUs.

With reference to the first aspect and the first possible implementation of the first aspect, in a second possible implementation manner,

Obtaining one of the M drawing threads whose CPU usage is greater than a threshold and/or other M-1 drawing threads of the M drawing threads.

With reference to the first possible implementation of the first aspect, and the second possible implementation of the first aspect, in a third possible implementation, the method further includes:

Obtaining a drawing time of the M drawing threads, and comparing the M drawing times to obtain a maximum drawing time thereof;

The maximum drawing time is compared with a time threshold, and the frequency of the N running CPUs is adjusted according to the comparison result.

In conjunction with the third possible implementation of the first aspect, in a fourth possible implementation, the comparing the maximum drawing time with a time threshold includes:

If the maximum drawing time is less than the time threshold, obtaining an estimated drawing time;

The drawing time is after the frequency of the N running CPUs is decreased, the drawing thread Drawing time

The estimated drawing time is compared with a screen refresh time, and if the estimated drawing time is less than the screen refresh time, the frequency of the N running CPUs is decreased.

In conjunction with the third possible implementation of the first aspect, in a fifth possible implementation, the comparing the maximum mapping time with a time threshold includes:

If the maximum drawing time is greater than the time threshold, comparing the maximum drawing time with a screen refresh time;

If the maximum drawing time is greater than the screen refresh time, the frequency of the N running CPUs is increased.

A second aspect of the embodiments of the present invention provides a multi-CPU scheduling apparatus, including:

The obtaining module is configured to acquire one of the M drawing threads and one of the N running CPUs, where M is an integer greater than 0, and N is an integer greater than 0;

And a binding module, configured to bind one of the M drawing threads to one CPU of the N running CPUs.

With reference to the second aspect, in a first possible implementation, the acquiring module is configured to acquire other M-1 drawing threads of M drawing threads and other N-1 CPUs of N running CPUs;

The binding module is configured to:

When the M-1 is greater than the N-1, the one-to-one binding of the N-1 drawing threads is processed on the N-1 running CPUs, and the MN drawing threads are bound to the N-1 running CPU M-N CPU processing.

With reference to the second aspect, and the first possible implementation manner of the second aspect, in a second possible implementation manner, the acquiring module is configured to: acquire, among the M drawing threads whose CPU usage is greater than a threshold One drawing thread and/or other M-1 drawing threads of the M drawing threads.

With reference to the first possible implementation of the second aspect and the second possible implementation of the second aspect, in a third possible implementation manner, the method further includes:

a maximum drawing time obtaining module, configured to acquire a drawing time of the M drawing threads, compare the M drawing times, and obtain a maximum drawing time thereof;

And a comparison module, configured to compare the maximum drawing time with a time threshold, and adjust a frequency of the N running CPUs according to the comparison result.

With reference to the third possible implementation of the second aspect, in a fourth possible implementation, the comparing module is configured to acquire an estimated drawing time if the maximum drawing time is less than the time threshold;

The estimated drawing time is a drawing time of the drawing thread after the frequency of the N running CPUs is decreased;

The comparing module is configured to compare the estimated drawing time with a screen refresh time;

And an adjustment module, configured to reduce a frequency of the N running CPUs if the estimated drawing time is less than the screen refresh time.

In conjunction with the third possible implementation of the second aspect, in a fifth possible implementation, the comparing module is configured to: when the maximum drawing time is greater than the time threshold, Screen refresh time comparison;

The adjusting module is configured to increase a frequency of the N running CPUs if the maximum drawing time is greater than the screen refresh time.

The multi-CPU scheduling method provided by the embodiment of the present invention binds one of the M drawing threads to the one by acquiring one of the M drawing threads and one of the N running CPUs. Processing on one of the N running CPUs. Under the premise of not affecting the performance of the system, the resources of each CPU are fully utilized, which can prevent the drawing thread from jumping between CPUs and reduce the power consumption of multiple CPUs.

DRAWINGS

In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, a brief description of the drawings used in the embodiments or the prior art description will be briefly described below. Obviously, the drawings in the following description It is a certain embodiment of the present invention, and other drawings can be obtained from those skilled in the art without any inventive labor.

1 is a schematic flowchart of a multi-CPU scheduling method according to Embodiment 1 of the present invention;

2a is a schematic diagram of a multi-CPU scheduling in the prior art;

2b is a schematic diagram showing the effects of the first embodiment and the second embodiment of the multi-CPU scheduling method according to the present invention;

3 is a schematic flowchart of a multi-CPU scheduling method according to Embodiment 1 and Embodiment 2;

4 is a schematic diagram of CPU usage of a multi-CPU scheduling method according to Embodiment 2 of the present invention;

FIG. 5 is a schematic flowchart of a multi-CPU frequency modulation method according to Embodiment 3 of the present invention; FIG.

FIG. 6 is a schematic diagram showing the effect of a feasible implementation manner of a multi-CPU scheduling method in Embodiment 2 and Embodiment 3 according to the present invention;

FIG. 7 is a schematic diagram showing a comparison of a feasible implementation manner of a multi-CPU scheduling method according to Embodiment 2 and Embodiment 3 of the present invention;

8 is a schematic structural diagram of a multi-CPU scheduling apparatus according to Embodiment 4 of the present invention;

9 is a schematic structural diagram of a multi-CPU scheduling apparatus according to Embodiment 5 of the present invention;

FIG. 10 is a schematic structural diagram of a terminal according to Embodiment 6 of the present invention.

detailed description

The technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the drawings in the embodiments of the present invention. It is a partial embodiment of the invention, and not all of the embodiments. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative efforts are within the scope of the present invention.

With the development of terminal devices, the CPU of the terminal device needs to run more and more programs. The process of running the program is called a process, and a process can include one or more threads, where the thread is an operating system capable of scheduling operations. The smallest unit is the actual operating unit in the process. A standard thread consists of a thread ID, a current instruction pointer, a register set, and a stack. One thread can create and revoke another thread, and multiple threads in the same process can execute concurrently. Each program has at least one thread. If the program has only one thread, the thread is the program itself. Among the many threads processed by the CPU, the drawing thread related to image processing has a very significant influence on the CPU frequency in the terminal device, and the human eye has a visual persistence. If the frame rate is higher than 60 fps, the picture is considered to be coherent. Most of the CPUs that process graphics refresh the screen 60 times per second, then the screen refresh time, that is, the screen refresh time is 1÷60=16.6ms, so the CPU of the terminal device processes the drawing thread each time. As long as the image is drawn before the screen refresh time of 16.6 ms, the user will not feel stuck when viewing the image. The drawing time of the drawing thread is mainly determined by the CPU frequency. Most CPUs of the terminal device will have a higher CPU frequency than the actual working. Therefore, in a terminal device with multiple CPUs, the CPU that needs to process the drawing thread is required. Said that it requires a large frequency to complete the processing of the drawing thread, and for the CPU that does not process the drawing thread, the other threads processed by the CPU do not have high CPU frequency requirements, so the drawing thread is not processed. The CPU will have a lot of free time. Various embodiments of the present invention provide a multi-CPU scheduling method and apparatus based on the characteristics of the above-described drawing thread. The technical solutions of the present invention are described below through various embodiments below.

1 is a schematic flowchart of a multi-CPU scheduling method according to Embodiment 1 of the present invention. It should be noted that the executor of the first embodiment may be any multi-processor, such as Symmetrical Multi-Processing (Symmetrical Multi-Processing). :SMP), and terminal devices that can process images, such as smartphones, tablets, and the like. Specifically, for a terminal device having multiple processors, the execution entity of the multi-CPU scheduling method may be one CPU of the plurality of CPUs of the terminal device, and may be determined according to a CPU scheduling program. For example, in most CPU operating systems, CPU scheduling is performed according to the priority of the process, the CPU affinity of the process, etc. The CPU scheduler executes the CPU scheduling process at the time of execution, and some operating systems bind the CPU scheduling process to CPU_0. On the other hand, some operating system default CPU scheduling processes can be operated on all CPUs. Among them, the CPU affinity can bind the thread to a specified CPU in the multi-CPU for as long as possible, so that the thread does not jump between the CPUs. One of the CPU scheduling processes is based on the CPU affinity of the process. Specifically, the CPU affinity setting of the process can be implemented by using the CPU affinity mask, and the Process Identifier (PID) of a process is used to bind the process corresponding to the PID to a specific process. On the CPU, if the process is not on the specific CPU at this time, the process is migrated to the specific CPU, and the operation of binding the process to the CPU according to the CPU affinity is implemented. For example, if the CPU affinity of the CPU scheduling process is set to CPU_0 according to the PID, then the CPU scheduling process only operates on CPU_0; if the CPU affinity of the CPU scheduling process defaults to all CPUs, then the CPU scheduling process is also the same. Other threads participate in scheduling, and migration operations can be performed on all CPUs. The multi-CPU scheduling method in the following embodiments of the present invention is implemented by the CPU scheduling process, and the terminal device also assigns a PID to it. The initial CPU affinity of the CPU scheduling process defaults to all CPUs. Further, the embodiment related to CPU frequency adjustment involved in the following embodiments of the present invention is implemented by a CPU frequency modulation program, which is a CPU frequency modulation process when executed, and the CPU frequency modulation process also participates in CPU scheduling. So executed in the CPU In the process, the CPU frequency modulation process also performs operations on the corresponding CPU according to the scheduling result of the CPU scheduling process. If the CPU frequency modulation process sets its own CPU affinity to CPU_0, then the CPU scheduling process will dispatch the CPU frequency modulation process to CPU_0, that is, the CPU frequency modulation process is processed by CPU_0.

Referring to FIG. 1, a multi-CPU scheduling method includes:

Step 100: Obtain one of the M drawing threads and one of the N running CPUs, where M is an integer greater than 0, and N is an integer greater than 0.

It should be noted that, as described above, when a drawing program is executed, it is a drawing process, and the actual operating unit of the drawing process is a drawing thread. For a thread processed by the CPU, when a thread calls a terminal device to store and draw. When the related function is used, it can be determined that the thread is a drawing thread. Further, the number M of drawing threads is obtained by acquiring the PID of the drawing thread. For the number of running CPUs N, the number of running CPUs N can be obtained through a preset program, and the specific program is not limited herein. Among them, the running CPU can be the CPU that is processing the thread or the CPU that allows the processing thread. For example, for a terminal with four CPUs (CPU_0, CPU_1, CPU_2, CPU_3), the terminal can be set by the operating system to specify the number of CPUs to be specifically operated. For example, CPU_2 and CPU_3 are not working, that is, the terminal is not processed. The process that the terminal needs to process, and only the CPU_0, CPU_1 to process the process, for this terminal, the running CPU is CPU_0, CPU_1, the number of which is 2.

Step 101: Bind one of the M drawing threads to one CPU of the N running CPUs for processing.

The solution for binding one of the M drawing threads to one of the N running CPUs may be implemented by setting a CPU affinity of the drawing thread. Specifically, the drawing thread may be bound to a specific CPU by using the PID of the drawing thread. If the drawing thread is not on the specific CPU at this time, the drawing thread is migrated to the specific CPU, and the CPU is implemented according to the CPU. Affinity binds a process to an operation on the CPU. For example, if the CPU affinity of the drawing thread is set to CPU_0 according to the PID of one of the M drawing threads, the drawing thread operates only on CPU_0.

Further, the process of binding one of the M drawing threads to one of the N running CPUs is described below by using a specific example, for example, when the number of drawing threads is 2 , that is, there are drawing thread 1 and drawing thread 2, and the number of running CPUs is 4, including CPU_0, CPU_1, CPU_2, CPU_3, and the drawing thread 1 is bound by setting the CPU affinity of drawing thread 1 to CPU_0. On CPU_0, a scheme is implemented in which one of the M drawing threads is bound to one of the N running CPUs. This ensures that other non-drawing threads will not affect CPU_0 processing drawing thread 1, and drawing thread 1 will not migrate between different CPUs, thus reducing the extra due to migration of drawing thread 1 without binding between different CPUs. Power consumption.

The multi-CPU scheduling method provided by the embodiment of the present invention obtains one of the M drawing threads and one of the N running CPUs, where M is an integer greater than 0, and N is an integer greater than 0. One of the M drawing threads is bound to be processed on one of the N running CPUs. Under the premise of not affecting the performance of the system, avoiding the migration of the drawing thread on multiple running CPUs causes additional power consumption, thereby reducing the overall power consumption of the multiple CPUs.

Further, after performing corresponding binding processing for acquiring one of the M drawing threads, or performing corresponding binding processing for acquiring one of the M drawing threads for the foregoing, The remaining M-1 drawing threads perform similar binding processing. The following describes the binding method of the remaining M-1 drawing threads through the second embodiment:

First, get the other M-1 drawing threads of the M drawing threads and the other N-1 CPUs of the N running CPUs. Comparing M-1 with N-1, according to the comparison result, it can be divided into at least three cases to perform specific operations:

Case 1: When the M-1 is smaller than the N-1, the one-to-one binding of the M-1 drawing threads is processed on the M-1 running CPUs; or

Case 2: when the M-1 is equal to the N-1, the one-to-one binding of the M-1 drawing threads is processed on the N-1 running CPUs; or

Case 3: When the M-1 is greater than the N-1, the N-1 drawing threads are bound one to one. Processing on the N-1 running CPUs, and M-N drawing threads are bound to the M-N CPUs on the N-1 running CPUs for processing.

Further, for case 1, for example, the number of drawing threads is 2, that is, there are drawing thread 1 and drawing thread 2, and the number of running CPUs is 4, including CPU_0, CPU_1, CPU_2, CPU_3, when one of the drawing threads 1 has Bind to CPU_0, at this time, M-1=1, N-1=3. Obviously, if M-1 is smaller than N-1, the drawing thread 2 is bound to any one of CPU_1, CPU_2, and CPU_3, and If the drawing thread 2 is bound to CPU_1, preferably, the CPU affinity of other non-drawing threads is set to CPU_2 or CPU_3, that is, other non-drawing threads can be processed on CPU_2 or CPU_3, which has the advantage of avoiding other non- The drawing thread is migrated to CPU_1 to avoid affecting CPU_1 processing drawing thread 2, thereby reducing the CPU power consumption of multiple running groups as a whole.

For case 2, for example, the number of drawing threads is 4, that is, there are drawing thread 1, drawing thread 2, drawing thread 3, drawing thread 4, and the number of running CPUs is 4, including CPU_0, CPU_1, CPU_2, CPU_3, among them. A drawing thread 1 has been bound to CPU_0, at this time M-1=3, N-1=3, obviously M-1 is equal to N-1, then the drawing thread 2 is bound to CPU_1, and the drawing thread 3 is tied. Set on CPU_2, bind drawing thread 4 to CPU_3, so that all drawing threads are bound one-to-one on all running CPUs, so that the power consumption of each running CPU is balanced, thus reducing overall The power consumption of a running CPU, while other non-drawing threads are processed on the corresponding CPU according to their initial CPU affinity, or can also be migrated between different CPUs, which is not limited herein.

For case three, for example, the number of drawing threads is 4, that is, there are drawing thread 1, drawing thread 2, drawing thread 3, drawing thread 4, and the number of running CPUs is 2, including CPU_0, CPU_1, when one of the drawing threads 1 Already bound to CPU_0, at this time M-1=3, N-1=1, obviously M-1 is greater than N-1, then drawing any one of drawing thread 2, drawing thread 3, drawing thread 4 The thread is bound to CPU_1, for example, the drawing thread 2 is bound to CPU_1, and the drawing thread 3, the drawing thread 4, and other non-drawing threads are processed on the corresponding CPU according to their initial CPU affinity, here Not limited.

Please refer to FIG. 2a. FIG. 2a is a schematic diagram of a multi-CPU scheduling in the prior art, as shown in FIG. 2a. In the technology, the terminal device migrates the threads between multiple CPUs through thread scheduling to achieve equalization of the CPU usage. However, the prior art thread migration scheme cannot effectively achieve the CPU usage balance, and when the CPU is When the workload is heavy, multiple migration threads will increase the CPU workload and CPU cache utilization will decrease.

2b is a schematic diagram of the effect of the first embodiment and the second embodiment of the multi-CPU scheduling method provided by the present invention. As shown in FIG. 2b, the M drawing threads are obtained according to the number M of drawing threads and the number of CPUs running N. One of the drawing thread bindings is processed on one of the N running CPUs. When the terminal device performs thread scheduling, the drawing thread does not migrate between multiple CPUs. Further, other M-1 drawing threads of M drawing threads and other N-1 CPUs of N running CPUs are acquired. And bind the other M-1 drawing threads to implement the binding scheme for the M drawing threads. For example, the thread numbers in Figure 2b are: 1, 2, 3, 4 of four drawing threads. After the thread scheduling, the four drawing threads are not migrated, so that the drawing threads that have a greater influence on the CPU frequency can be evenly distributed on different CPUs, ensuring that one CPU is not bound to other CPUs. The effect of the drawing thread, for each drawing thread bound on the CPU, the CPU is more efficient at processing the drawing thread bound to it. Further, after each CPU performs the binding processing on the drawing thread by using the solution of the embodiment of the present invention, the overall frequency requirement of the multi-CPU of the terminal device can be down-regulated, thereby reducing the power consumption of the terminal device.

Preferably, one possible determining manner of the number M of drawing threads in the first embodiment and the second embodiment is: acquiring one of the M drawing threads whose CPU usage is greater than a threshold and/or the M The other M-1 drawing threads of the drawing thread. For example, if the current multi-CPU processes a total of 7 drawing threads, and the number of drawing threads whose CPU usage is greater than the threshold is 4, then M is 4.

Further, FIG. 3 is a schematic flowchart diagram of the multi-CPU scheduling method according to the first embodiment and the second embodiment. When the terminal device constantly updates the screen, the drawing thread becomes heavy, and the condition for deciding whether or not to perform the CPU frequency adjustment can be a heavy task. Therefore, the thread cumbersomeness can be determined by obtaining the CPU usage of the drawing thread. When the CPU usage of the drawing thread is changed from the CPU usage threshold (abbreviation: threshold) to less than the CPU usage threshold (referred to as: If the threshold is greater than the CPU usage threshold, the number of drawing threads whose CPU usage is greater than the threshold and the number of running CPUs are obtained. A preferred implementation manner is as follows: when the CPU usage of the drawing thread is changed from the CPU usage threshold to the CPU usage threshold, or the CPU usage threshold is greater than the CPU usage threshold. An identifier is added to the drawing thread in which the above two cases occur, and the number M of drawing threads larger than the threshold is determined by the identifier. Moreover, the threshold is a preset decimal number. Since the CPU model and operating system used by different terminal devices are different, the specific value of the threshold can be set according to different CPU models and operating systems. limited.

Referring to FIG. 3, a scheme for implementing one-to-one binding on each CPU for processing each drawing thread according to the CPU usage of the drawing thread is described in detail, wherein, for a terminal device having multiple processors, more The execution body of the CPU scheduling method may be one of a plurality of CPUs of the terminal device, and may be determined according to a CPU scheduler. For a step 100 in FIG. 1, one possible implementation manner includes:

Step 100a: If the CPU usage of the drawing thread changes from less than the threshold to greater than the threshold, or if the CPU usage of the drawing thread changes from greater than the threshold to less than the threshold, acquiring the M drawing threads whose CPU usage is greater than the threshold One drawing thread and/or other M-1 drawing threads of the M drawing threads and N running CPUs.

Specifically, calculating the CPU usage of all the drawing threads; comparing the CPU usage of all the drawing threads with the threshold, when the CPU usage of the drawing thread changes from less than the threshold to greater than the threshold, or the CPU usage of the drawing thread When the trigger condition is changed from the threshold value to the threshold value, the operation of acquiring the number M of the drawing threads whose CPU usage is greater than the threshold value and the number of running CPUs N is obtained, and further, the M drawing threads have been After a drawing thread binding is processed on one of the N running CPUs, the obtained M is decremented by one.

Preferably, if the CPU usage of the K drawing threads changes from greater than the threshold to less than the threshold, the number of drawing threads M is reduced by K, and the drawing thread binding strategy is determined according to the MK drawing threads; or, if K drawing threads are If the CPU usage changes from less than the threshold to greater than the threshold, the number of drawing threads is increased by M. K, determine the drawing thread binding strategy according to M+K drawing threads. It should be noted that the K value is an integer greater than or equal to zero, and if the CPU usage of the K drawing threads changes from greater than the threshold to less than the threshold, the K value must be less than or equal to an integer of the M value.

Further, the embodiment of the present invention expresses the heavyness of the thread by the CPU usage of the thread. Each CPU can identify the executing thread, and calculate the execution time of the thread on the CPU for a period of time. The execution time is longer than the total time to obtain the CPU usage of the thread, namely:

CPU usage of thread = execution time / total time

It should be noted that the total time is the sum of the times of n clock cycles of the CPU, where n is an integer greater than zero; and the threshold in the above is a preset decimal number, which is used for CPU usage with one thread. For example, the threshold can be set to 0.6. When the CPU usage of a drawing thread is calculated to be 0.7, it is obvious that the CPU usage of the drawing thread is greater than the threshold, indicating that the drawing thread is heavy. Similarly, when the CPU usage of a thread is less than the threshold, it indicates that the thread is less heavy.

4 is a schematic diagram of CPU usage of a multi-CPU scheduling method according to Embodiment 2 of the present invention. As shown in FIG. 4, taking two running CPUs (CPU_0, CPU_1) as an example, CPU_1 and CPU_1 run thread 1 and thread. 2 and thread 3, wherein the CPU usage of the thread 1 during the time T (n clock cycles) is calculated, and the terminal device is uniformly based on a CPU, and then:

CPU usage of thread 1 = (t1 + t2 + t3) / T

Wherein, since the thread 1 may migrate between the two CPUs, it can be seen from FIG. 4 that t1 and t3 represent the execution time of the thread 1 on the CPU_0, and t3 represents the execution time of the thread 1 on the migration to the CPU_1. The CPU usage calculation method of the drawing thread is similar. For example: If the threshold is 0.3. There are 4 drawing threads, wherein: the CPU usage of drawing thread 1 is 0.8, the CPU usage of drawing thread 2 is 0.35, the CPU usage of drawing thread 3 is 0.3, and the CPU usage of drawing thread 4 is 0.1. Therefore, the CPU usage of two

drawing threads

1 and 2 exceeds the threshold of 0.3, and then the value of M is 2.

In addition, when the number of drawing threads whose CPU usage changes are acquired in real time, the following may occur: a. Create a thread or delete a thread, then judge the created new thread or delete the created thread. Whether the thread is a drawing thread, and if so, the value of M is updated; b. When the number of drawing threads is dynamically calculated, if the CPU usage of the drawing thread whose previous CPU usage has not changed has changed, then if the drawing thread of the drawing thread changes If the occupancy rate changes from greater than the threshold to less than the threshold, then M-1, if the occupancy rate of the drawing thread changes from less than the threshold to greater than the threshold, then M+1; if the number of running CPUs changes, reacquire N .

Referring to FIG. 3, after step 101 in FIG. 1, the method further includes:

Step 102: Obtain other M-1 drawing threads of the M drawing threads and other N-1 CPUs of the N running CPUs, and determine a drawing thread binding policy according to the M-1 and the N-1.

Specifically, referring to the second embodiment of the present invention, the drawing thread binding policy may include at least the following three cases:

Case 1: When the M-1 is smaller than the N-1, the one-to-one binding of the M-1 drawing threads is processed on the M-1 running CPUs. In addition, preferably, the non-drawing thread may be set to be processed by N-M running CPUs on N-M running CPUs.

Case 2: When the M-1 is equal to the N-1, the one-to-one binding of the M-1 drawing threads is processed on the N-1 running CPUs. Additionally, preferably, the non-drawing thread may be in an initial affinity state.

Case 3: When the M-1 is greater than the N-1, the one-to-one binding of the N-1 drawing threads is processed on the N-1 running CPUs, and the MN drawing threads are bound. Processing on MN CPUs on the N-1 running CPUs. Additionally, preferably, the non-drawing thread may be in an initial affinity state.

Step 103: Bind the M-1 drawing threads one-to-one to the N-1 CPUs for processing according to a drawing thread binding policy.

Specifically, in the first embodiment, after one of the M drawing threads is bound to one of the N running CPUs, according to the situation one, the second case and the situation in step 101a. Third, the M-1 drawing threads are bound one-to-one on the N-1 CPUs for processing.

Preferably, referring to FIG. 3, after step 103, the method further includes:

Step 104: Obtain one of the M drawing threads whose CPU usage is greater than a threshold and/or other M-1 drawing threads of the M drawing threads.

Step 105: Obtain a drawing time of the M drawing threads, and compare the M drawing times to obtain a maximum drawing time.

Specifically, when the drawing thread of the CPU process is drawn, the drawing thread will call the drawing function, and the time required for the drawing function operation is the drawing time. A feasible way to obtain the drawing time is to draw a point at the beginning and end of the drawing function, obtain the start and end time, and then calculate the time difference between the end of the drawing and the start to get the drawing time.

Step 106: Compare the maximum drawing time with a time threshold.

Specifically, the time threshold can be configured as needed, such as time threshold = screen refresh time x 70%. If the maximum drawing time is less than the time threshold, step 107 is performed; if the maximum drawing time is greater than the time threshold, step 109 is performed.

Step 107: Obtain an estimated drawing time.

Specifically, the estimated drawing time is the drawing time of the drawing thread after the frequency of the N running CPUs is lowered by one level. Among them, the estimated drawing time can be obtained by the following formula (1):

Estimated plot time = maximum plot time × current CPU frequency / small first-level CPU frequency (1)

It should be noted that, as shown by the formula (1), the estimated drawing time is calculated by the formula (1) for the maximum drawing time of the M drawing times, and the estimated drawing time is the estimated drawing thread having the maximum drawing time. 1 Plot time after reducing the frequency of the CPU running at one level. In addition, reducing the frequency of the CPU operating at one level is to reduce the frequency as a whole for the frequencies of the N operating CPUs. For example, before reducing the running CPU frequency, the N running CPU frequencies are all 800MHZ. After reducing the CPU frequency of the first-level operation, the N running CPU frequencies are all 600MHZ.

Step 108: Compare the estimated drawing time with the screen refresh time. If the estimated drawing time is less than the screen refresh time, reduce the frequency of the N running CPUs.

If the estimated drawing time is still less than the screen refresh time, the operating CPU frequency is lowered. Here, after lowering the first-level CPU frequency, referring to the above steps 105 to 108, the maximum drawing time is continued at the time threshold. For comparison, the CPU frequency adjustment is not required until the time threshold <maximum drawing time <screen refresh time is satisfied. The following steps 107 and 108 are described according to specific scenarios:

Scene 1, for example, by comparing the drawing times of M drawing threads, the maximum drawing time is 11ms, at which time the CPU frequencies of N running are 800MHZ, and the CPU frequency of the smaller one is 600MHZ (requires description, N runs) The CPU frequency is still at 800 MHz. The estimated plot time calculated according to formula (1) is about 15 ms. Referring to the above, the screen refresh time Tr is 16.6 ms. According to the comparison of step 108, the estimated plot time is 15 ms. The screen refresh time Tr16.6ms reduces the frequency of N running CPUs by one level. The frequency of the N running CPUs is lowered from 800 MHz to 600 MHz. When the CPU frequency is 600 MHz, the drawing time is reacquired, and the operations of the above steps 105 to 108 are repeated until the time threshold <maximum drawing time <screen refresh time is satisfied, and the CPU frequency adjustment is not required.

Scenario 2, for example, by comparing the drawing times of M drawing threads, the maximum drawing time is 14 ms, at which time the CPU frequencies of N running are 800 MHz, and the CPU frequency of the smaller one is 600 MHz (required, N runs) The CPU frequency is still at 800 MHz. The estimated plot time calculated according to formula (1) is about 18 ms. Referring to the above, the screen refresh time Tr is 16.6 ms. At this time, the plot time > screen refresh time Tr16 is estimated. 6ms, obviously, if the CPU frequency is lowered by one level at this time, the image file displayed by the terminal device will be stuck. Then even if the current maximum drawing time of 14ms is less than the time threshold, the N CPU frequencies can not be adjusted to 600MHZ, but remain 800MHZ.

Step 109: Compare the maximum drawing time with the screen refresh time. If the maximum drawing time is greater than the screen refresh time, the frequency of the N running CPUs is increased.

Preferably, if the maximum drawing time is greater than the screen refresh time, in order to avoid the image jamming caused by the maximum drawing time being greater than the screen refresh time, the N running CPUs may be The frequency is increased by one level, so that the drawing time can be reduced at a higher CPU frequency, and the condition that the drawing time is less than the screen refresh time is satisfied, thereby avoiding the occurrence of image jamming. In addition, the embodiment related to CPU frequency adjustment involved in this embodiment may be implemented by a CPU frequency modulation program.

Referring to FIG. 5, the frequency modulation scheme of the multiple CPUs in the foregoing embodiment 2 can be performed independently, and the frequency modulation scheme of the multiple CPUs in the above is not required after the binding of the M graphics threads. FIG. 5 is provided in Embodiment 3 of the present invention. A schematic diagram of a multi-CPU frequency modulation method, wherein the multi-CPU frequency modulation method can be executed by one CPU of a plurality of CPUs. As shown in FIG. 5, the multi-CPU frequency modulation method includes the following steps:

Step 200: Obtain drawing time of M drawing threads, and compare M drawing times to obtain a maximum drawing time, where M is an integer greater than 0.

Step 201: Compare the maximum drawing time with a time threshold.

Specifically, the time threshold can be configured as needed, such as time threshold = screen refresh time x 70%. If the maximum drawing time is less than the time threshold, step 202 is performed; if the maximum drawing time is greater than the time threshold, step 204 is performed.

Step 202: Obtain an estimated drawing time.

Specifically, the method for obtaining the estimated drawing time is referred to the second embodiment, and details are not described herein again.

Step 203: Compare the estimated drawing time with the screen refresh time. If the estimated drawing time is less than the screen refresh time, reduce the frequency of the N running CPUs.

In step 203, if the estimated drawing time is less than the screen refresh time, the frequency of the N running CPUs is lowered, so that when the user is guaranteed to view the image, there is no feeling of stuttering, and the power consumption of the multiple CPUs can also be reduced.

Step 204: Compare the maximum drawing time with the screen refresh time. If the maximum drawing time is greater than the screen refresh time, increase the frequency of the N running CPUs.

Step 204 is adopted. If the maximum drawing time is greater than the screen refresh time, the frequency of the N running CPUs is increased, so that the user does not feel stuck when viewing the image.

It is to be noted that FIG. 6 is a schematic diagram showing the effect of a feasible implementation manner of the multi-CPU scheduling method in the second embodiment and the third embodiment, and FIG. 7 is a second embodiment of the present invention. A comparison diagram of a feasible implementation manner of the multi-CPU scheduling method in the third embodiment, referring to FIG. 6, where T1 is the drawing time of one drawing thread, and T2 is the screen refresh time of the CPU processing the drawing thread, by adopting the present According to the CPU frequency adjustment scheme of the embodiment of the invention, it can be clearly found that after the adjustment, the difference between T2 and T1 is significantly shortened, that is, the utilization rate of the CPU frequency resource is significantly improved. Referring to FIG. 7, for the implementation of the present invention. In the CPU of the second embodiment and the third embodiment, the CPU frequency is as shown in the waveform of the CPU frequency (solid line) for which the CPU frequency is not adjusted, and the CPU frequency waveform (dotted line) after the CPU frequency adjustment processing is obviously flatter. The resource utilization of the CPU frequency is also more sufficient, and the CPU frequency resource is not wasted. After adopting the CPU frequency adjustment scheme in the second embodiment and the third embodiment provided by the present invention, the frequency of the CPU after the CPU frequency adjustment processing is not going to be high or low, and most of the time is kept at a lower frequency, and the CPU frequency adjustment processing is performed. Does not affect the refresh frame rate of the screen. It can be seen from these two diagrams that the screen refresh CPU frequency modulation scheme can reduce the idle time of the CPU without affecting the system performance (the screen refresh rate is unchanged), make full use of the CPU resources, and enable the CPU to work more often. At lower frequencies.

It should be noted that, in each of the above embodiments and each possible implementation manner, the number M of drawing threads may be the number of drawing threads whose CPU usage is greater than the CPU occupancy threshold in all drawing threads.

FIG. 8 is a schematic structural diagram of a multi-CPU scheduling apparatus according to Embodiment 4 of the present invention. As shown in FIG. 8, the multi-CPU scheduling apparatus includes: an obtaining module 10 and a binding module 11.

The obtaining module 10 is configured to acquire one of the M drawing threads and one of the N running CPUs, where M is an integer greater than 0, and N is an integer greater than 0.

Specifically, the possible implementation manners of obtaining the number M of the drawing threads and the number of running CPUs N have been described in detail in the first embodiment, and are not described herein again.

The binding module 11 is configured to bind one of the M drawing threads to one CPU of the N running CPUs.

Specifically, the binding module 11 implements a scheme of binding the drawing threads one-to-one on each CPU for processing by setting the CPU affinity of the drawing thread. The CPU affinity has been described in detail in the first embodiment, and details are not described herein again.

The multi-CPU scheduling method provided by the embodiment of the present invention acquires the number of drawing threads and the number of running CPUs by the acquiring module, and then the binding module according to the number of drawing threads and the number of running CPUs, when the drawing threads When the number is less than or equal to the number of running CPUs, each drawing thread is bound one-to-one on each running CPU for processing. Under the premise of not affecting the performance of the system, the resources of each CPU are fully utilized, so that the multi-CPU of the terminal device works at a lower frequency as a whole, thereby reducing power consumption.

Preferably, the obtaining module 10 is further configured to acquire other M-1 drawing threads of the M drawing threads and other N-1 CPUs of the N running CPUs.

Referring to the second embodiment, the binding module 11 is configured to bind the M-1 drawing threads one-to-one to the M-1 running CPUs when the M-1 is smaller than the N-1. Processed on; or

Further, the obtaining module 10 is further configured to: acquire one of the M drawing threads whose CPU usage is greater than a threshold, and/or other M-1 drawing threads of the M drawing threads.

Preferably, after completing the binding of one of the M drawing threads, and/or the binding of the other M-1 drawing threads of the M drawing threads, a multi-CPU frequency modulation scheme may also be performed, and the The solution may perform the frequency adjustment of the N running CPUs on the premise that the binding module 11 binds one of the drawing threads and/or the M-1 drawing threads in the M, and may also refer to the manner of the third embodiment above. Specifically, in order to perform the multi-CPU frequency modulation scheme, referring to FIG. 9, on the basis of FIG. 8, FIG. 9 is a schematic structural diagram of a multi-CPU scheduling apparatus according to Embodiment 5 of the present invention, as shown in FIG. The CPU scheduling device further includes: a maximum drawing time acquiring module 12, a comparing module 13, and an adjusting module 14.

a maximum drawing time obtaining module 12, configured to acquire a drawing time of the M drawing threads, and The M plot times are compared to obtain the maximum plot time.

Specifically, in the step 102 of the foregoing embodiment 2, how to obtain the drawing time of the drawing thread and how to obtain the maximum drawing time are described in detail, and details are not described herein again.

The comparison module 13 is configured to compare the maximum drawing time with a time threshold.

Further, the comparison module 13 is configured to compare the maximum drawing time with a time threshold, and adjust the frequency of the N running CPUs according to the comparison result.

Specifically, the time threshold is described in detail in step 103 of Embodiment 2 above, and the estimated drawing time is obtained by the above formula (1), and details are not described herein again.

Further, the comparing module 13 is further configured to: if the maximum drawing time is less than the time threshold, obtain an estimated drawing time;

Estimate the plot time for the drawing time of the drawing thread after the frequency of the N running CPUs is reduced. Among them, the frequency of N running CPUs can be reduced by one level.

The comparison module 13 is configured to compare the estimated drawing time with a screen refresh time.

The adjustment module 14 is configured to reduce the frequency of the N running CPUs if the estimated drawing time is less than the screen refresh time.

It should be noted that, to reduce the frequency of the CPU, the frequency may be reduced overall for the frequencies of the N CPUs.

Preferably, the comparison module 13 is further configured to compare the maximum drawing time with the screen refresh time if the maximum drawing time is greater than the time threshold.

The adjustment module 14 is configured to increase the frequency of the N running CPUs if the maximum drawing time is greater than the screen refresh time. For example, the frequency of N running CPUs can be increased by one level.

If the maximum drawing time is greater than the screen refresh time, in order to avoid image jamming caused by the maximum drawing time being greater than the screen refresh time, the frequency of the N running CPUs can be increased by one level, thereby making the CPU higher. At the frequency, the drawing time can be reduced to meet the condition that the drawing time is less than the screen refresh time, thereby avoiding the occurrence of image jamming.

Embodiment 6 of the present invention provides a terminal, which can be used to perform Embodiment 1 to Embodiment 3 of the present invention. The method described in the above. FIG. 10 is a schematic structural diagram of a terminal according to Embodiment 6 of the present invention, and FIG. 10 is a schematic structural diagram of a terminal 500 according to Embodiment 6 of the present invention.

The terminal may be a terminal device including a mobile phone, a tablet computer, a PDA (Personal Digital Assistant), a POS (Point of Sales), an in-vehicle computer, etc., and the terminal is a mobile phone as an example, and FIG. 10 shows A block diagram of a partial structure of a mobile phone 500 related to a terminal provided by an embodiment of the present invention. Referring to FIG. 10, the mobile phone 500 includes an RF (Radio Frequency) circuit 510, a memory 520, an input unit 530, a display unit 540, a sensor 550, an audio circuit 560, a WiFi (Wireless Fidelity) module 570, and a processor 580. And power supply 590 and other components. It can be understood by those skilled in the art that the structure of the mobile phone shown in FIG. 10 is only an example of implementation, and does not constitute a limitation on the mobile phone, and may include more or less components than those illustrated, or combine some components, or Different parts are arranged.

The components of the mobile phone 500 will be specifically described below with reference to FIG. 10:

The RF circuit 510 can be used for receiving and transmitting signals during and after the transmission or reception of information, in particular, after receiving the downlink information of the base station, and processing it to the processor 580; in addition, transmitting the uplink data to the base station. Generally, RF circuits include, but are not limited to, an antenna, at least one amplifier, a transceiver, a coupler, an LNA (Low Noise Amplifier), a duplexer, and the like. In addition, RF circuitry 510 can also communicate with the network and other devices via wireless communication. The wireless communication may use any communication standard or protocol, including but not limited to GSM (Global System of Mobile communication), GPRS (General Packet Radio Service), CDMA (Code Division Multiple Access). , Code Division Multiple Access), WCDMA (Wideband Code Division Multiple Access), LTE (Long Term Evolution), e-mail, SMS (Short Messaging Service), and the like.

The memory 520 can be used to store software programs, and the processor 580 executes various functional applications and data processing of the mobile phone 500 by running software programs stored in the memory 520. The memory 520 may mainly include a storage program area and a storage data area, wherein the storage program area may store an operating system, an application required for at least one function (such as a sound playing function, an image playing function, etc.), and the like; The data area can store data (such as audio data, phone book, etc.) created according to the use of the mobile phone 500. Moreover, memory 520 can include high speed random access memory, and can also include non-volatile memory, such as at least one magnetic disk storage device, flash memory device, or other volatile solid state storage device.

The input unit 530 can be configured to receive input numeric or character information and to generate key signal inputs related to user settings and function control of the handset 500. Specifically, the input unit 530 may include a touch panel 531 and other input devices 532. The touch panel 531, also referred to as a touch screen, can collect touch operations on or near the user (such as the user using a finger, a stylus, or the like on the touch panel 531 or near the touch panel 531. Operation), and drive the corresponding connecting device according to a preset program. Optionally, the touch panel 531 can include two parts: a touch detection device and a touch controller. Wherein, the touch detection device detects the touch orientation of the user, and detects a signal brought by the touch operation, and transmits the signal to the touch controller; the touch controller receives the touch information from the touch detection device, converts the touch information into contact coordinates, and sends the touch information. The processor 580 is provided and can receive commands from the processor 580 and execute them. In addition, the touch panel 531 can be implemented in various types such as resistive, capacitive, infrared, and surface acoustic waves. In addition to the touch panel 531, the input unit 530 may also include other input devices 532. Specifically, other input devices 532 may include, but are not limited to, one or more of a physical keyboard, function keys (such as volume control buttons, switch buttons, etc.), trackballs, mice, joysticks, and the like.

The display unit 540 can be used to display information input by the user or information provided to the user and various menus of the mobile phone 500. The display unit 540 can include a display panel 541. Alternatively, the display panel 541 can be configured in the form of an LCD (Liquid Crystal Display), an OLED (Organic Light-Emitting Diode), or the like. Further, the touch panel 531 can cover the display panel 541. When the touch panel 531 detects a touch operation on or near it, the touch panel 531 transmits to the processor 580 to determine the type of the touch event, and then the processor 580 according to the touch event. The type provides a corresponding visual output on display panel 541. Although in FIG. 10, the touch panel 531 and the display panel 541 are two independent components to implement the input and input functions of the mobile phone 500, in some embodiments, the touch panel 531 can be integrated with the display panel 541. And realize the input of the mobile phone 500 and Output function.

The handset 500 can also include at least one type of sensor 550, such as a light sensor, motion sensor, and other sensors. Specifically, the light sensor may include an ambient light sensor and a proximity sensor, wherein the ambient light sensor may adjust the brightness of the display panel 541 according to the brightness of the ambient light, and the proximity sensor may close the display panel 541 when the mobile phone 500 moves to the ear. / or backlight. As a kind of motion sensor, the accelerometer sensor can detect the magnitude of acceleration in all directions (usually three axes). When it is stationary, it can detect the magnitude and direction of gravity. It can be used to identify the gesture of the mobile phone (such as horizontal and vertical screen switching, related Game, magnetometer attitude calibration), vibration recognition related functions (such as pedometer, tapping), etc. As for the mobile phone 500 can also be configured with gyroscopes, barometers, hygrometers, thermometers, infrared sensors and other sensors, here Let me repeat.

Audio circuit 560, speaker 561, and microphone 562 can provide an audio interface between the user and handset 500. The audio circuit 560 can transmit the converted electrical data of the received audio data to the speaker 561, and convert it into a sound signal output by the speaker 561. On the other hand, the microphone 562 converts the collected sound signal into an electrical signal, and the audio circuit 560 is used by the audio circuit 560. After receiving, it is converted into audio data, and then processed by the audio data output processor 580, sent to the other mobile phone via the RF circuit 510, or outputted to the memory 520 for further processing.

WiFi is a short-range wireless transmission technology, and the mobile phone 500 can help users to send and receive emails, browse web pages, and access streaming media through the WiFi module 570, which provides wireless broadband Internet access for users. Although FIG. 10 shows the WiFi module 570, it can be understood that it does not belong to the essential configuration of the mobile phone 500, and may be omitted as needed within the scope of not changing the essence of the invention.

Processor 580 is the control center of handset 500, which connects various portions of the entire handset using various interfaces and lines, by running or executing software programs and/or modules stored in memory 520, and recalling data stored in memory 520, The various functions and processing data of the mobile phone 500 are performed to perform overall monitoring of the mobile phone. Optionally, the processor 580 may include one or more processing units; preferably, the processor 580 may integrate an application processor and a modem processor, where the application processes The main processing system, user interface and applications, etc., the modem processor mainly deals with wireless communication. It will be appreciated that the above described modem processor may also not be integrated into the processor 580.

The handset 500 also includes a power source 590 (such as a battery) that supplies power to the various components. Preferably, the power source can be logically coupled to the processor 580 via a power management system to manage functions such as charging, discharging, and power management through the power management system.

Although not shown, the mobile phone 500 may further include a camera, a Bluetooth module, and the like, and details are not described herein again.

In the embodiment of the present invention, the terminal includes a processor 580 having the following functions:

The processor 580 is configured to acquire one of the M drawing threads and one of the N running CPUs, where M is an integer greater than 0, and N is an integer greater than 0.

Specifically, it should be noted that, for a multi-processor terminal, the processor 580 is one of the multi-processors. The possible implementation manners of obtaining the number M of the drawing threads and the number of running CPUs N have been described in detail in the first embodiment, and are not described herein again.

The processor 580 is configured to bind one of the M drawing threads to one CPU of the N running CPUs.

Specifically, the processor 580 implements a scheme of binding one of the M drawing threads to one of the N running CPUs by setting a CPU affinity of the drawing thread. The CPU affinity has been described in detail in the first embodiment, and details are not described herein again.

In the embodiment of the present invention, the memory 520 included in the terminal has the following functions: a CPU scheduler, a CPU frequency modulation program, and a drawing-related function stored in the present invention, in preparation for the processor to run the CPU scheduling process, and the CPU. Called when the FM process and the drawing thread.

The multi-CPU scheduling method provided by the embodiment of the present invention acquires the number of drawing threads and the number of running CPUs by the processor, and then the number of drawing threads according to the number of drawing threads and the number of running CPUs by the processor. When the number of CPUs is less than or equal to each other, each drawing thread is bound one-to-one on each CPU for processing. Under the premise of not affecting the performance of the system, the resources of each CPU are fully utilized, so that the multi-CPU of the terminal device works at a lower frequency as a whole, thereby reducing power consumption.

Preferably, the processor 580 is further configured to acquire other M-1 drawing threads of the M drawing threads. And other N-1 CPUs of N running CPUs.

Further, the processor 580 is specifically configured to:

Further, after completing binding of one drawing thread of the M drawing threads, and/or binding of other M-1 drawing threads of the M drawing threads, a multi-CPU frequency modulation scheme may also be performed, and The solution may perform the frequency adjustment of the N running CPUs on the premise that the processor 580 binds one of the drawing threads and/or the M-1 drawing threads in the M, and may also implement the method in the third embodiment. Specifically, the processor 580 is configured to acquire one of the M drawing threads whose CPU usage is greater than a threshold, and/or other M-1 drawing threads of the M drawing threads, and obtain the The drawing time of the M drawing threads, and comparing the M drawing times, and obtaining the maximum drawing time among them.

Specifically, it should be noted that, in the foregoing Embodiment 2, the CPU usage and the threshold, how to obtain the drawing time of the drawing thread, and how to obtain the maximum drawing time have been described in detail, and details are not described herein again.

The processor 580 is configured to compare the maximum drawing time with a time threshold, and adjust a frequency of the N running CPUs according to the comparison result.

Further, the processor 580 is specifically configured to: if the maximum drawing time is less than the time threshold, obtain an estimated drawing time.

Estimating the plot time can be a plot time for the plot thread after the frequency of the N running CPUs is reduced by one.

The processor 580 is further configured to compare the estimated drawing time with the screen refresh time.

The processor 580 is configured to reduce the frequency of the N running CPUs if the estimated drawing time is less than the screen refresh time.

It should be noted that reducing the frequency of the running CPU is to reduce the frequency of the entire frequency of the N operating CPUs.

Preferably, the processor 580 is further configured to compare the maximum drawing time with the screen refresh time if the maximum drawing time is greater than the time threshold.

The processor 580 is configured to increase the frequency of the N running CPUs if the maximum drawing time is greater than the screen refresh time.

One of ordinary skill in the art will appreciate that all or part of the steps to implement the various method embodiments described above may be accomplished by hardware associated with the program instructions. The aforementioned program can be stored in a computer readable storage medium. The program, when executed, performs the steps including the foregoing method embodiments; and the foregoing storage medium includes various media that can store program codes, such as a ROM, a RAM, a magnetic disk, or an optical disk.

Finally, it should be noted that the above embodiments are merely illustrative of the technical solutions of the present invention, and are not intended to be limiting; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art will understand that The technical solutions described in the foregoing embodiments may be modified, or some or all of the technical features may be equivalently replaced; and the modifications or substitutions do not deviate from the technical solutions of the embodiments of the present invention. range.

Claims

A multi-CPU scheduling method, comprising:

Obtaining one of the M drawing threads and one of the N running CPUs, where M is an integer greater than 0, and N is an integer greater than 0;

Binding one of the M drawing threads to one of the N running CPUs for processing.
The method according to claim 1, wherein the other M-1 drawing threads of the M drawing threads and the other N-1 CPUs of the N running CPUs are acquired;

When the M-1 is smaller than the N-1, the one-to-one binding of the M-1 drawing threads is processed on the M-1 running CPUs; or

When the M-1 is equal to the N-1, the one-to-one binding of the M-1 drawing threads is processed on the N-1 running CPUs; or

When the M-1 is greater than the N-1, the one-to-one binding of the N-1 drawing threads is processed on the N-1 running CPUs, and the MN drawing threads are bound to the Processing on MN CPUs on N-1 running CPUs.
Method according to claim 1 or 2, characterized in that

Obtaining one of the M drawing threads whose CPU usage is greater than a threshold and/or other M-1 drawing threads of the M drawing threads.
The method according to claim 2 or 3, wherein the method further comprises:

Obtaining a drawing time of the M drawing threads, and comparing the M drawing times to obtain a maximum drawing time thereof;

The maximum drawing time is compared with a time threshold, and the frequency of the N running CPUs is adjusted according to the comparison result.
The method of claim 4 wherein said comparing said maximum drawing time to a time threshold comprises:

If the maximum drawing time is less than the time threshold, obtaining an estimated drawing time;

The estimated drawing time is a drawing time of the drawing thread after the frequency of the N running CPUs is decreased;

The estimated drawing time is compared with a screen refresh time, and if the estimated drawing time is less than the screen refresh time, the frequency of the N running CPUs is decreased.
The method of claim 4 wherein said comparing said maximum drawing time to a time threshold comprises:

If the maximum drawing time is greater than the time threshold, comparing the maximum drawing time with a screen refresh time;

If the maximum drawing time is greater than the screen refresh time, the frequency of the N running CPUs is increased.
A multi-CPU scheduling device, comprising:

The obtaining module is configured to acquire one of the M drawing threads and one of the N running CPUs, where M is an integer greater than 0, and N is an integer greater than 0;

And a binding module, configured to bind one of the M drawing threads to one CPU of the N running CPUs.
The apparatus according to claim 7, wherein the obtaining module is configured to acquire other M-1 drawing threads of the M drawing threads and other N-1 CPUs of the N running CPUs;

The binding module is configured to:

When the M-1 is smaller than the N-1, the one-to-one binding of the M-1 drawing threads is processed on the M-1 running CPUs; or

When the M-1 is equal to the N-1, the one-to-one binding of the M-1 drawing threads is processed on the N-1 running CPUs; or

When the M-1 is greater than the N-1, the one-to-one binding of the N-1 drawing threads is processed on the N-1 running CPUs, and the MN drawing threads are bound to the Processing on MN CPUs on N-1 running CPUs.
The device according to claim 7 or 8, wherein the obtaining module is configured to: Obtaining one of the M drawing threads whose CPU usage is greater than a threshold and/or other M-1 drawing threads of the M drawing threads.
The device according to claim 8 or 9, further comprising:

a maximum drawing time obtaining module, configured to acquire a drawing time of the M drawing threads, compare the M drawing times, and obtain a maximum drawing time thereof;

And a comparison module, configured to compare the maximum drawing time with a time threshold, and adjust a frequency of the N running CPUs according to the comparison result.
The device according to claim 10, wherein the comparing module is configured to acquire an estimated drawing time if the maximum drawing time is less than the time threshold;

The estimated drawing time is a drawing time of the drawing thread after the frequency of the N running CPUs is decreased;

The comparing module is configured to compare the estimated drawing time with a screen refresh time;

And an adjustment module, configured to reduce a frequency of the N running CPUs if the estimated drawing time is less than the screen refresh time.
The device according to claim 10, wherein the comparing module is configured to compare the maximum drawing time with a screen refresh time if the maximum drawing time is greater than the time threshold;

The adjusting module is configured to increase a frequency of the N running CPUs if the maximum drawing time is greater than the screen refresh time.
A terminal, comprising: a plurality of processors and a memory, wherein

The memory is configured to store a program required for the first processor to operate in the plurality of processors;

The first processor runs the program of the memory, and is used to acquire one of the M drawing threads and one of the N running CPUs, where M is an integer greater than 0, and N is an integer greater than 0 And processing for binding one of the M drawing threads to one of the N running CPUs.
The terminal according to claim 13, wherein said first processor is obtained by a user Take other M-1 drawing threads of M drawing threads and other N-1 CPUs of N running CPUs;

When the M-1 is smaller than the N-1, the first processor is configured to perform one-to-one binding of the M-1 drawing threads on the M-1 running CPUs; or

When the M-1 is equal to the N-1, the first processor is configured to process the one-to-one binding of the M-1 drawing threads on the N-1 running CPUs. ;or

When the M-1 is greater than the N-1, the first processor is configured to perform one-to-one binding of N-1 drawing threads on the N-1 running CPUs, and MN drawing thread bindings are processed on MN CPUs on the N-1 running CPUs.
The terminal according to claim 13 or 14, wherein the first processor is configured to acquire one of the M drawing threads whose CPU usage is greater than a threshold and/or the M drawing The other M-1 drawing threads of the thread.
The terminal according to claim 14 or 15, wherein the first processor is configured to:

Obtaining a drawing time of the M drawing threads, and comparing the M drawing times to obtain a maximum drawing time thereof;

The maximum drawing time is compared with a time threshold, and the frequency of the N running CPUs is adjusted according to the comparison result.
The terminal according to claim 16, wherein the comparing the maximum drawing time with a time threshold comprises:

And if the maximum drawing time is less than the time threshold, the first processor is configured to obtain an estimated drawing time;

The estimated drawing time is a drawing time of the drawing thread after the frequency of the N running CPUs is decreased;

The estimated drawing time is compared with a screen refresh time, and if the estimated drawing time is less than the screen refresh time, the first processor is configured to reduce the frequency of the N running CPUs.
The terminal according to claim 16, wherein said maximum drawing time is Compare with time thresholds, including:

If the maximum drawing time is greater than the time threshold, the first processor is configured to compare the maximum drawing time with a screen refresh time;

If the maximum drawing time is greater than the screen refresh time, the first processor is configured to increase the frequency of the N running CPUs.