Marco Jakobs wrote:
OK ... good to know. :-) But i see no advantage to rewrite it from the
netconn API to the socket API, as everything works fine
No, just wantet to make this clear for anyone having to decide between
the 2 APIs.
So the calculation of TCP_SND_QUEULEN will result in 8. Does this mean
that TCP may use 8 out of 16 (MEMP_NUM_PBUF) buffers?
Not really. TCP_SND_QUEUELEN == 8 says that at maximum, 8 pbufs will be
queued for sending per TCP pcb. However, (up to CVS HEAD) every call to
netconn_write creates 1 pbuf (PBUF_RAM when telling the stack to copy
the data) or 2 pbufs (one PBUF_RAM and one PBUF_REF, when not telling
the stack to copy the data) at minimum. Only the PBUF_REF pbufs come
from the pool specified by MEMP_NUM_PBUF. The PBUF_RAM pbufs are
allocated on the heap, the size of which is defined by MEM_SIZE.
Simon
_______________________________________________
lwip-users mailing list
[email protected]
http://lists.nongnu.org/mailman/listinfo/lwip-users
