I'm still fighting with my problem. Now I know what is going wrong, but I cannot find the reason. Probably somebody can help me. I attached a Wireshark log. You can see that communication stops after the confirmation from the PC for the frames that transfer the file style.css. All seems to be correct in this log, but my device doesn't send the next frames after this confirmation. I debugged my software and I see that an assert is generated in the file tcp_in.c in the function tcp_receive(), line 1020:
LWIP_ASSERT("pcb->snd_queuelen >= pbuf_clen(next->p)", (pcb->snd_queuelen >= pbuf_clen(next->p)));
Here I can see that pcb->snd_queuelen is 2, but pbuf_clen(next->p) returns big value. See the screenshot bellow:
Here I can see that the chain of the buffer is incorrect. Except the first buffer, the "next" pointer has a valid address, but the pointer buffer is empty (also the length of the data is zero).
I cannot understand why this happens. I will continue debugging, but if you have an idea what could be wrong, please, give me a hint.
Simeon
If you want the server to accept 6 simultaneous connections, you have to provide the stack with room for at least 6 pcbs, memory for the 6 connection structures, perhaps enough pbufs for prompt response...
There's been a long time I'm not with the server, but I don't remember having a built-in limit.
Perhaps if you read the docs, enable debugging (and probably stats), take note of what is happening, and post it, one of us could have a clue to your specific issue.
_______________________________________________
lwip-users mailing list
address@hidden
https://lists.nongnu.org/mailman/listinfo/lwip-users
-- Simeon Trifonov Head of department "Development" AMK Drives and Controls Ltd. Bulgaria / Gabrovo Phone: +359 (0)66 819125 Fax: +359 (0)66 819101 E-Mail: address@hidden Web: www.amk-drives.bg
