Zitat von Jan Kiszka <[EMAIL PROTECTED]>: > Bernhard Pfund wrote: >> All, >> >> I have a kernel module sending and receiving UDP packets at 10kHz, same >> socket. To make it work (ENOBUFS in the first place) I had to increase >> socket_rtskbs to 2048 and now I get the following dmesg:
I can actually correct myself here, socket_rtskbs = 512 is enough for the moment (very small packets only). >> FATAL: rt_e1000 ran into tail close to head situation! >> hard_start_xmit returned 1 >> Packet send failed! Errno 11 (that's my sender task) >> >> Several packets are transmitted and received before the error. > > What does "several" roughly means? Note that the TX ring size appears to > be 256 for the e1000 (according to quick code inspection). So, if > "several" in the same dimension, you are getting no IRQs from the > adapter to release buffers. Also check against the IRQ stats that RTAI > may provide. Several means ~2000 (1800 - 2100) cycles. > If this doesn't help, you will have to dig deeper, probably with the > help of tracers (printk instrumentation may blow up your logs and > destroy the timing). I c, no problem if it helps. >> >> My configuration is: >> >> Linux : 2.6.26.6 >> Ipipe : hal-linux-2.6.26-x86-2.0-15.patch >> RTnet : latest svn >> RTAI : latest magma >> >> Any hints? >> >> Bernhard >> > > Jan Thanks! Bernhard ------------------------------------------------------------------------- This SF.Net email is sponsored by the Moblin Your Move Developer's challenge Build the coolest Linux based applications with Moblin SDK & win great prizes Grand prize is a trip for two to an Open Source event anywhere in the world http://moblin-contest.org/redirect.php?banner_id=100&url=/ _______________________________________________ RTnet-users mailing list RTnet-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/rtnet-users
