This is the mail archive of the ecos-discuss@sources.redhat.com mailing list for the eCos project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]
Other format: [Raw text]

Re: FreeBSD-Stack not responding


hi,

quoting http://sources.redhat.com/ml/ecos-discuss/2002-07/msg00425.html

> On Sat, 2002-07-27 at 15:03, Wolfgang Heppner wrote:
> > Hello,
> >
> > I recently switched from the Open- to the FreeBSD stack (latest CVS
> > version)
> > When i try to send Data to my target (e.g. ascii strings via a telnet
> > client on Linux and/or Windows), the data is lost in the stack (the
> > "read" call in my application never returns) - even when a connection
> > could be established and my application could send a "Prompt" to the
> > telnet client! With the OpenBSD Stack this works fine.
> >
> > I figured out that packets are dropped in the cyg_tcp_input function
> > because of a bad checksum.
> >
> > I got some warnings during the eCos-built:
> > implicit declarations of function 'cyg_in_cksum_hdr' in ip_input.c:332
> > implicit declarations of function 'cyg_in_cksum_hdr' and
> > 'cyg_in_cksum_skip'
> > in ip_output and
> > implicit declarations of function 'cyg_in_pseudo' in tcp_input:405
> > (among some others)
> 
> These are just warnings and should be harmless.
> 
> >
> > Has anybode experienced similar problems with the FreeBSD stack?
> > Any ideas?
> >
> > Thanks in advance
> > Wolfgang
> >
> > PS: I built the eCos library with the new_net template and removed libm
> > and dns support.
> 
> What's the target?

Similar problems happens to me.
Target is a PPC8xx custom one with a derived SCC driver.
(the "old" stack is working fine)

DHCP at boot time is not working.
its seems i have the same behaviour described in
http://sources.redhat.com/ml/ecos-discuss/2002-01/msg00418.html
(Re: DHCP server xid)

Testing the ping_test with a static IP address
results in 

PING server 10.16.2.20
64 bytes from 10.16.2.20: icmp_seq=0, time=0ms
recvfrom: Operation timed out
64 bytes from 10.16.2.20: icmp_seq=2, time=0ms
recvfrom: Operation timed out
64 bytes from 10.16.2.20: icmp_seq=4, time=0ms
 :
 .

ethereal shows the echo requests and replys.
it seems an ICMP echo reply with a payload of
(N mod 4 != 0) is not handled correctly.
( 64 is working, 65, 66, 67 isnt working, 68 is working ....).

BTW: a TCP connect() is returning -1. :-/
starting to track it down...

any ideas?, regards Christoph

-- 

-- 
Before posting, please read the FAQ: http://sources.redhat.com/fom/ecos
and search the list archive: http://sources.redhat.com/ml/ecos-discuss


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]