Wince memory leak problem

lbj412

Wince memory leak problem [Copy link]

The project requires the use of an RFID card reader under wince. This program queries every 200 milliseconds whether an RFID card is swiped (initiates an invitation). Now, there is a memory leak during the execution of the program. It is very regular. It can be located that a certain line of code that queries whether an RFID card is swiped causes the leak. This line of code is a call to an SDK function. Since I have the source code of the SDK function, I found that this function should not cause a memory leak. Now I wonder if the memory leak is caused by the RFID card reader driver? PS: The leaked memory can be recovered when the program exits, but once the program is re-executed, the system's memory usage increases a lot, exceeding the value before the previous execution and shutdown. —— I don't know if this can be confirmed as a driver memory leak?

ruibin_shu

Fatigue test to see if there is memory failure and the application cannot run normally. The memory recycling mechanism of CE is a little different from that of PC. I guess it is just memory fragmentation.

5532

After the program runs for a long time, it will prompt that the system memory is seriously insufficient.

wbatjd

It should be caused by virtual memory not being released----------------------------- You can call the API to query the size of the available virtual memory in the current slot to determine the problem. I have encountered similar problems before, and this is how I determined it----------------------------- You can check the API online. If you can't find it, I will find it for you this afternoon.

ygwilliam

But once the program was restarted, the system's memory usage suddenly increased a lot, exceeding the value before the program was shut down. --------I remember that this was not accurate. It took him a while to react.

chym168

Thank you for your warm-hearted friend. If the virtual memory is not released, how can I solve it?

Quote the reply of guopeixin on the 3rd floor: It should be caused by the virtual memory not being released. You can call the API to query the size of the available virtual memory in the current slot to determine the problem. I have encountered similar problems before, and this is how I determined it. You can check the API online. If you can't find it, I will find it for you this afternoon.

simple_head

Does it take a long time? It took me a while, but it still didn't work.

Quote from the reply of mxm1986 on the 4th floor: But once the program is re-executed, the system's memory usage suddenly increases a lot, exceeding the value before the previous execution and shutdown. --------I remember that this is not accurate. . It will take him a while to react.

redmonster

Isn't virtual memory automatically released by the system?

wang1jin

Quote from m_jy163 on the 5th floor: You are such a caring friend, thank you. If the virtual memory is really not released, how can we solve it? Quote from guopeixin on the 3rd floor: It should be caused by the virtual memory not being released----------------------------- You can call the API to query the size of the available virtual memory in the current slot to determine the problem. I have encountered similar problems before, and this is how I determined it----------------------------- You can check the API online. If you can't find it, I will find it for you this afternoon.

It happened in the 5.0 project before. Because each slot has 32mb of space, we will cause 64kb of space to be lost at a time. After many operations on the upper-level AP, the current slot space will be used out. Just release it.

stone_111

Make a mark

xiaoyuanhb09

How to release it? Haha. I am a novice, I hope you can give me some advice.

Quote the reply from guopeixin on the 8th floor: Quote the reply from m_jy163 on the 5th floor: You are a really enthusiastic friend, thank you. If the virtual memory is really not released, how to solve it? Quote the reply from guopeixin on the 3rd floor: It should be caused by the virtual memory not being released----------------------------- You can call the API to query the size of the available virtual memory in the current slot to determine the problem. I have encountered similar problems before, and this is how I determined it----------------------------- You can check the API online. If you can't find it, I will find it for you in the afternoon. It appeared in the 5.0 project before. Because each slot is 32mb of space, we will cause 64kb of space to be lost at a time. After many operations on the upper-level AP, the current slot space will be used out. Just release it.

stdio_h_0

Will memory leaks cause system crashes? My board crashes and the screen remains frozen, but dnw does not show any debugging information

LQpoliceofficer

Quote from m_jy163 on the 10th floor: How to release it? Haha. I am a novice, so I hope you can give me some advice. Quote from guopeixin on the 8th floor: Quote from m_jy163 on the 5th floor: You are such a caring friend, thank you. If it is really the virtual memory that has not been released, how can we solve it? Quote from guopeixin on the 3rd floor: It should be caused by the virtual memory not being released----------------------------- You can call the API to query the size of the available virtual memory in the current slot to determine the problem. I have encountered similar problems before, and that is how I determined it----------------------------- You can check the API online to change it. If you can't find it, I will find it for you in the 5.0 project this afternoon. Because each slot is 32mb of space, we will cause the loss of 64kb of space at a time. After many operations on the upper-level AP, the current slot space will be used out. Just release it.

First find the location of the leak and release it directly.

coldra

UP

IceA

Quote from the reply of the original poster m_jy163: The project requires the use of an RFID card reader under WinCE. This program queries every 200 milliseconds whether an RFID card is swiped (initiates an invitation). Now, there is a memory leak during the execution of the program. It is very regular. It can be located that a certain line of code that queries whether an RFID card is swiped causes the leak. This line of code is a call to an SDK function. Since I have the source code of the SDK function, I found that this function should not cause a memory leak. Now I suspect that the memory leak is caused by the RFID card reader driver? PS: The leaked memory can be recovered when the program exits, but once the program is re-executed, the system's memory usage increases a lot, exceeding the value before the previous execution and shutdown. - I don't know if this can be confirmed as a driver memory leak?

Is there a buffer clearing operation function in the driver? If so, call it, or initialize the call parameters before the function to see if the memory will increase. Of course, this is to test whether the memory is generated by the input parameters.

yutm

http://topic.eeworld.net/u/20091111/16/8f1cde89-22c6-467e-a11e-8d2e5f020885.html?68418 You will find the result if you follow this post. You can also google it yourself. How to release the memory applied for, for example, after new, it will be followed by delete, so there should be a one-to-one correspondence.

snowmoonsad

Quote from stoneqiang's reply on the 15th floor: http://topic.eeworld.net/u/20091111/16/8f1cde89-22c6-467e-a11e-8d2e5f020885.html?68418 You will find the result if you follow this post. You can also google it yourself. How to release the memory applied for, for example, after new, it will be followed by delete, one by one

I like this~~~ Hehe

xiangzi

When the program runs for a long time, try to reduce dynamic allocation. Even if you release it after allocation, it will still be the same.

lhuahua

Help

aphonline

Using CodeSnitch