[lwip-users] bug in tcp_close_shutdown?

[Tomas Hruby]

Hi,

when I enable TCP_DEBUG_PCB_LISTS I trigger LWIP_ASSERT("TCP_RMV: pcbs
!= NULL", *(pcbs) != NULL); in TCP_RMV when I call tcp_close on a
tcp_pcb which was only created by tcp_new and closed immediately
afterwards. No other action performed on the pcb in the meantime. The
problem is that the pcb is in the CLOSED state and TCP_RMV(&tcp_bound_pcbs, pcb)
is called in tcp_close_shutdown. But there is no bound socket at the
moment. I experience the problem in 1.4.0.rc1 but a quick look at rc2
suggest that the problem is still there. Is it save just not to call
the TCP_RMV? Or call it only in certain situations?

Cheers, T.