[Top][All Lists]
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [lwip-users] Picking appropriate sizes for the different mboxes
From: |
Simon Goldschmidt |
Subject: |
Re: [lwip-users] Picking appropriate sizes for the different mboxes |
Date: |
Thu, 16 Feb 2012 15:43:32 +0100 |
Mason <address@hidden> wrote:
> > On my platform, the Ethernet driver manages a queue of incoming
> > frames (there can be max RX_BUF_COUNT outstanding frames).
> > Typically, RX_BUF_COUNT is between 100 and 200.
> >
> > Should TCPIP_MBOX_SIZE have a simple relation to RX_BUF_COUNT?
I could imagine setting TCPIP_MBOX_SIZE to "Number of API tasks" + RX_BUF_COUNT
then. On the other hand, some of the RX_BUF_COUNT buffers will probably be
buffere *after* being passed to tcpip_thread (by applications or by TCP OOS
buffer), so setting TCPIP_MBOX_SIZE that high is probably not necessary. Keep
in mind also that you need enough items in MEMP_TCPIP_MSG_INPKT.
> > The packets queued in TCPIP_MBOX are subsequently dispatched to
> > the appropriate {UDP,TCP}_RECVMBOX?
Not necessarily. They are not queued when they are not passed to a netconn or
socket (i.e. using the raw (callback) API or packet is dropped).
> > If the RAW API is disabled, DEFAULT_RAW_RECVMBOX_SIZE can be 0, right?
Don't confuse the raw (callback) API with raw netconns/sockets! The latter
indicates a connection doesn't use UDP/TCP but its own transport-layer protocol.
> > Is there a wiki, or text document covering this material?
Hmm, I don't know, but contributing documentation is always welcome!
> For this error, one can
> #define MEMP_NUM_SYS_TIMEOUT 5
> in lwipopts.h
>
> Perhaps the default value should be raised?
> Or is this the desired outcome?
I've just fixed the checks that they are not executed when MEMP uses MEM.
Simon
--
Empfehlen Sie GMX DSL Ihren Freunden und Bekannten und wir
belohnen Sie mit bis zu 50,- Euro! https://freundschaftswerbung.gmx.de