cvs commit: src/sys/netinet raw_ip.c

Bruce M. Simpson bms at
Tue Mar 20 13:23:11 UTC 2007

Gleb Smirnoff wrote:
> We've got interfaces with even larger MTU. Why shouldn't we permit sending
> larger datagrams?
It seems reasonable to be consistent in the amount of send space we 
reserve for both SOCK_RAW and SOCK_DGRAM in netinet.

I agree however that this is really the application's problem -- it 
should probe for interface MTU and set SO_SENDBUF accordingly, and this 
is what I originally told the submitter.


More information about the cvs-src mailing list