I am using LWIP stack for local TCP communication on a Marvell MW300 microcontroller, a TCP server is listening for incoming connections and processing commands.
All works OK, however only for a short period of time (about 15-20 minutes). After that, it seems like the PCB stops accepting connections, despite that according to its state it is still listening. Memory usage is stable the whole testing period.
Here is an overview of the state of the LWIP stack right after it stopped working (stats_display("")):
[1010.140004] : LINK
[1010.140052] : xmit: 1324
[1010.141696] : recv: 3075
[1010.144213] : fw: 0
[1010.146295] : drop: 431
[1010.148725] : chkerr: 0
[1010.151157] : lenerr: 0
[1010.153587] : memerr: 0
[1010.156016] : rterr: 0
[1010.158361] : proterr: 0
[1010.160877] : opterr: 0
[1010.163308] : err: 3
[1010.165477] : cachehit: 0
[1010.167996] :
ETHARP
[1010.170253] : xmit: 11
[1010.172596] : recv: 261
[1010.175025] : fw: 0
[1010.177109] : drop: 0
[1010.179366] : chkerr: 0
[1010.181796] : lenerr: 0
[1010.184228] : memerr: 0
[1010.186658] : rterr: 0
[1010.189001] : proterr: 0
[1010.191517] : opterr: 0
[1010.193948] : err: 0
[1010.196119] : cachehit: 960
[1010.198809] :
IP_FRAG
[1010.201153] : xmit: 0
[1010.203409] : recv: 0
[1010.205668] : fw: 0
[1010.207750] : drop: 0
[1010.210007] : chkerr: 0
[1010.212438] : lenerr: 0
[1010.214868] : memerr: 0
[1010.217297] : rterr: 0
[1010.219641] : proterr: 0
[1010.222159] : opterr: 0
[1010.224589] : err: 0
[1010.226760] : cachehit: 0
[1010.229277] :
IP
[1010.231185] : xmit: 1285
[1010.233704] : recv: 2814
[1010.236221] : fw: 0
[1010.238303] : drop: 0
[1010.240559] : chkerr: 0
[1010.242990] : lenerr: 0
[1010.245421] : memerr: 0
[1010.247852] : rterr: 0
[1010.250195] : proterr: 0
[1010.252711] : opterr: 0
[1010.255142] : err: 0
[1010.257313] : cachehit: 0
[1010.259829] :
IGMP
[1010.261914] : xmit: 5
[1010.264170] : recv: 10
[1010.266512] : drop: 0
[1010.268771] : chkerr: 0
[1010.271200] : lenerr: 0
[1010.273630] : memerr: 0
[1010.276062] : proterr: 0
[1010.278578] : rx_v1: 0
[1010.280922] : rx_group: 0
[1010.283525] : rx_general: 0
[1010.286305] : rx_report: 10
[1010.289082] : tx_join: 2
[1010.291599] : tx_leave: 0
[1010.294202] : tx_report: 3
[1010.296893] :
ICMP
[1010.298978] : xmit: 0
[1010.301233] : recv: 0
[1010.303489] : fw: 0
[1010.305574] : drop: 0
[1010.307830] : chkerr: 0
[1010.310262] : lenerr: 0
[1010.312692] : memerr: 0
[1010.315121] : rterr: 0
[1010.317465] : proterr: 0
[1010.319982] : opterr: 0
[1010.322413] : err: 0
[1010.324581] : cachehit: 0
[1010.327101] :
UDP
[1010.329095] : xmit: 3
[1010.331352] : recv: 867
[1010.333783] : fw: 0
[1010.335868] : drop: 429
[1010.338297] : chkerr: 0
[1010.340727] : lenerr: 0
[1010.343158] : memerr: 0
[1010.345589] : rterr: 0
[1010.347931] : proterr: 0
[1010.350448] : opterr: 0
[1010.352878] : err: 0
[1010.355050] : cachehit: 0
[1010.357566] :
TCP
[1010.359563] : xmit: 1004
[1010.362080] : recv: 1937
[1010.364596] : fw: 0
[1010.366681] : drop: 0
[1010.368937] : chkerr: 0
[1010.371369] : lenerr: 0
[1010.373799] : memerr: 0
[1010.376228] : rterr: 0
[1010.378573] : proterr: 0
[1010.381090] : opterr: 0
[1010.383520] : err: 0
[1010.385689] : cachehit: 0
[1010.388208] :
MEM HEAP
[1010.390636] : avail: 34760
[1010.393327] : used: 112
[1010.395757] : max: 2972
[1010.398188] : err: 0
[1010.400274] :
MEM RAW_PCB
[1010.402964] : avail: 2
[1010.405308] : used: 0
[1010.407564] : max: 0
[1010.409734] : err: 0
[1010.411817] :
MEM UDP_PCB
[1010.414508] : avail: 8
[1010.416851] : used: 4
[1010.419108] : max: 4
[1010.421277] : err: 0
[1010.423362] :
MEM TCP_PCB
[1010.426051] : avail: 8
[1010.428396] : used: 0
[1010.430652] : max: 4
[1010.432823] : err: 0
[1010.434906] :
MEM TCP_PCB_LISTEN
[1010.438203] : avail: 4
[1010.440548] : used: 1
[1010.442803] : max: 1
[1010.444974] : err: 0
[1010.447058] :
MEM TCP_SEG
[1010.449748] : avail: 12
[1010.452180] : used: 0
[1010.454436] : max: 3
[1010.456605] : err: 0
[1010.458689] :
MEM REASSDATA
[1010.461552] : avail: 16
[1010.463983] : used: 0
[1010.466240] : max: 0
[1010.468410] : err: 0
[1010.470493] :
MEM FRAG_PBUF
[1010.473358] : avail: 15
[1010.475789] : used: 0
[1010.478046] : max: 0
[1010.480214] : err: 0
[1010.482299] :
MEM NETBUF
[1010.484901] : avail: 16
[1010.487332] : used: 0
[1010.489589] : max: 3
[1010.491760] : err: 0
[1010.493843] :
MEM NETCONN
[1010.496534] : avail: 18
[1010.498964] : used: 2
[1010.501220] : max: 2
[1010.503391] : err: 0
[1010.505474] :
MEM TCPIP_MSG_API
[1010.508685] : avail: 8
[1010.511029] : used: 0
[1010.513286] : max: 1
[1010.515457] : err: 0
[1010.517539] :
MEM TCPIP_MSG_INPKT
[1010.520924] : avail: 16
[1010.523356] : used: 0
[1010.525612] : max: 5
[1010.527780] : err: 0
[1010.529865] :
MEM IGMP_GROUP
[1010.532817] : avail: 8
[1010.535160] : used: 4
[1010.537417] : max: 4
[1010.539586] : err: 0
[1010.541669] :
MEM SYS_TIMEOUT
[1010.544708] : avail: 12
[1010.547137] : used: 2
[1010.549394] : max: 4
[1010.551565] : err: 0
[1010.553648] :
MEM NETDB
[1010.556165] : avail: 1
[1010.558508] : used: 0
[1010.560764] : max: 0
[1010.562935] : err: 0
[1010.565019] :
MEM PBUF_REF/ROM
[1010.568143] : avail: 10
[1010.570574] : used: 0
[1010.572830] : max: 0
[1010.575002] : err: 0
[1010.577084] :
MEM PBUF_POOL
[1010.579948] : avail: 20
[1010.582378] : used: 0
[1010.584636] : max: 6
[1010.586805] : err: 0
[1010.588890] :
SYS
[1010.590886] : sem.used: 4
[1010.593575] : sem.max: 5
[1010.596267] : sem.err: 0
[1010.598957] : mutex.used: 0
[1010.601734] : mutex.max: 0
[1010.604513] : mutex.err: 0
[1010.607290] : mbox.used: 3
[1010.610068] : mbox.max: 3
[1010.612844] : mbox.err: 0
[1010.615624] :
Here is an overview of memory usage of the Marvell chip (os_dump_mem_stats()):
Heap size ---------------------- : 160784
Free size ---------------------- : 120920
Peak Heap Usage since bootup --- : 43136
Total allocations -------------- : 85
Failed allocations ------------- : 0
Min overhead per allocation ---- : 16
Biggest free block available now : 118408
What could be the problem here?
Thanks!