Re: [lwip-users] Slow release time of closed TCP PCBs

[Lou Cypher]

> I will make some new tests, with all these new hints in mind

I did a quick check, and I found *no* case yet of connection refused -- so my
last post about SYN failing is _wrong_.

Connections do work, the only errors I found right now are about the lwIP side
missing an ACK, after a FIN,ACK (and I have no idea about that, yet); could be
this type of errors I spotted previously, and wrongly(?) attributed to
memp_alloc errors

Makes sense that after a memp_alloc error starts the workaround, freeing a PCB
in TIME_WAIT state; so when working with full buffers and active debug I'll
often see the "memp_malloc: out of memory in pool TCP_PCB" message

Lou