Re: [lwip-users] PPPoS: ppp session disconnect if hit bad fcs for LCP packet

[Patrick Klos]

On 7/10/2017 5:23 AM, Axel Lin wrote:

I got a dump of the data when bad fcs happened. (This time is "IP"

rather than "LCP")
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

I think the complete packet is:

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

the bad fcs is a7c5
Is there any thing wrong in above received data?

Those 2 streams are from different packets. 

Have you already "un-escaped" the packet(s)?  If not, then the first packet is clearly not being parsed properly.  Look at the "7e21450001ae..." bytes (where I've split them below):

2145000028b56a4000ee06847834c111ea0a0002420050de7add4c7bd91377a8bd50107980ef410000279f  7e21450001aeb56b4000ee0682f134c111ea0a00024....

That is clearly the start of a new IP packet. 

And given the length of the first packet (from the IP header), you may have lost some bytes from the first packet?

Patrick Klos
Klos Technologies, Inc.