On 03/08/2017 19:12, Ian Buckley wrote: > Paolo, > Thats a fundamental H/W limitation of the current FPGA design. Long > story short, it would have been inefficient to support an MTU slightly > larger than a 2^N size. It’s also a pretty conservative number in > terms of a an MTU that works for 99% of Hosts and network gear. > -Ian
Hello Ian, thank you for your reply. What you say make perfect sense. From an FPGA, let me say hw, point of view to overcome the 8000 frame size limitation (if we can talk of a limitation) it is likely to require a buffer 16384 in size instead of 8192 and this is an inefficient use of the FPGA resources indeed. However, just for research purpose, I want to ask you if you know what buffers I need to increase. To be plain, I was able to get, playing with X300_10GE_DATA_FRAME_MAX_SIZE value and recv_frame_size arg an MTU as large as 8132, but I would like to reach the 9710 mtu limit on my linux box. Thank you again. Paolo _______________________________________________ USRP-users mailing list USRP-users@lists.ettus.com http://lists.ettus.com/mailman/listinfo/usrp-users_lists.ettus.com