Yes, we also hit the similar issue in the WiFi testing and the fix is ready, I will send the patch today.
Thanks Xiang On Sat, Jan 11, 2020 at 12:59 AM Gregory Nutt <spudan...@gmail.com> wrote: > > Issue received on Google group. Any insightes? Xaio Xiang? There have > been a lot of networking changes recently. > > Most issues will, like this, be received via email. Not entered into > github issues. > > Greg > > -------- Forwarded Message -------- > Subject: [nuttx] Network issues on recent master > Date: Fri, 10 Jan 2020 08:46:03 -0800 (PST) > From: Rob Voisey <r...@fenconsultants.com> > Reply-To: nu...@googlegroups.com > To: NuttX <nu...@googlegroups.com> > > > > Hi > > I'm seeing some problems on master. I haven't check against an older > version yet as it's quite a job moving between v7 and v8 but if this > doesn't ring a bell for anyone I'll try that next. I've spent a lot of > time tracing network code already. > > 1. SO_RCVTIMEO not working. > > I set SO_RCVTIMEO on a UDP socket but the recvfrom hangs indefinitely > while no data is received. It's waiting on ir_sem because nothing > runs inet_udp_eventhandler to let the timeout code run. If I comment out > the devif_event_trigger test in devif_conn_event to run the handler on > Ethernet driver polls (which it doesn't normally accept) this works > around the problem and the timeout code runs correctly. > > As an aside setsockopt appears to clear the s_options SO_RCVTIMEO bit if > a timeout is set and vice versa. That looks backwards to me, although I > don't think that bit is used directly other than by getsockopt. > > 2. TCP close not working. > > If I call close on a TCP socket it does not send a FIN | ACK and then > hangs up waiting for the response. If I close the other end it receives > the FIN | ACK which it interprets as a response and finishes closing the > socket. > > I'm guessing many people would be shouting about them if these were > really network stack bugs but I've no idea what else I might be missing. > Any thoughts appreciated. > > Thanks > > Rob > > -- > You received this message because you are subscribed to the Google > Groups "NuttX" group. > To unsubscribe from this group and stop receiving emails from it, send > an email to nuttx+unsubscr...@googlegroups.com > <mailto:nuttx+unsubscr...@googlegroups.com>. > To view this discussion on the web visit > https://groups.google.com/d/msgid/nuttx/52377393-8b7e-427a-bbcb-0213b0387f49%40googlegroups.com > <https://groups.google.com/d/msgid/nuttx/52377393-8b7e-427a-bbcb-0213b0387f49%40googlegroups.com?utm_medium=email&utm_source=footer>.
