Hi, Thanks a lot. The said issue occurs after I run iperf for say 4/5 hours. I mean there is no specific time when it occurs. My guess is, since my tcp_tmr() is not protected, probably some other thread was getting executed in between tcp_tmr.
Not sure. I will try to protect tcp_tmr and see if it improves the performance. regards Anirudha ________________________________ From: Kieran Mansley <kie...@recoil.org> To: Mailing list for lwIP users <lwip-users@nongnu.org> Sent: Thursday, 7 July 2011 16:27:04 Subject: Re: [lwip-users] (no subject) On Thu, 2011-07-07 at 18:50 +0800, Anirudha Sarangi wrote: > Just cross-checked. There is a mistake. With my present timer setup, > tcp_tmr is > getting called every 350 msec instead of every 250 msec. Do you think > it could > be causing my issue ? No, the difference between every 350ms and every 250ms would probably be unnoticable in TCP behaviour (your retransmissions would just start a few fractions of a second later). > Then as I understand, once I am in tcp_tmr, I should disable all > other > interrupts so that there could be no corruption. Will that be fine? That's one way. I prefer to arrange it such that the interrupts do not call into lwIP, but instead queue work to be processed by the main loop. You should also worry about application threads calling in, unless you only call lwIP API functions from the context that is calling tcp_tmr() in which case they will be naturally excluded. Kieran _______________________________________________ lwip-users mailing list lwip-users@nongnu.org https://lists.nongnu.org/mailman/listinfo/lwip-users
_______________________________________________ lwip-users mailing list lwip-users@nongnu.org https://lists.nongnu.org/mailman/listinfo/lwip-users
