Hi,ESP_Angus wrote:Hi Pratik,
Thanks for posting the code and the log. It's not clear to me from the log that you have a memory leak. If I look at the same log line in your code each time around the loop, there's no long-term trend of less free RAM:
In fact, the final log line shows more free heap (230872) than the initial line (230864).Code: Select all
I (23662) main: system_get_free_heap_size after accept =230864 I (24452) main: system_get_free_heap_size after accept =230648 I (25592) main: system_get_free_heap_size after accept =230440 I (879372) main: system_get_free_heap_size after accept =229896 I (892582) main: system_get_free_heap_size after accept =229352 I (925662) main: system_get_free_heap_size after accept =230524 I (925802) main: system_get_free_heap_size after accept =230872
This kind of minor variation over time is due to the TCP/IP & WiFi library activity. Some of it may be explained by TCP sockets in the TIME_WAIT state, as I mentioned above, but the WiFi library may also allocate temporary buffers for various uses.
If you see a long term trend of free memory reducing, that is a memory leak.
Ok. Yes.we can say it as short term memory leak which is going to free after sometime.
I think Espressif Systems has fixed that memory leak issue from LWIP netconn code and also released it under master tree.