Vladimir Cotfas wrote: > Jan, > >> You were asking me to make the delay timeouts configurable. Problem is >> that >>> there's more than one and I think they should be left as-is. >> If you look closer, you'll find only one effective timeout: that for the >> outer loop. >> >>> I've had >>> problems with the rt_eepro100 timeout and I had to extend it. >> That's why it has to be configurable. The defaults are unacceptable for >> normal use as they can bring down the whole system (latency-wise) if the >> hardware runs into troubles. >> > I am happy with the delays as they are -- the driver works just fine for us > [but again I am not a RT guy]. The only potential problem may be in > e100_exec_cmd as far as I can see. > > This is the patch I am submitting. It is against e100.c of Linux 2.6.29. I > attached the patch and the "before" and "after" files (compressed). > > Porting notes: > a) most of the eth_tool functions (not used) were removed; > b) we use an RTAI thread in lieu of the NAPI delayed-work scheme; > c) we use and rtdm signal + schedule_work for TX timeout/lockups; > d) we disabled the LED blink timer; > e) we kept the "watchdog" timer as a Linux timer > f) we intend to use an external timer interrupt to drive the e100 driver in > lieu of its own IRQ, hence USE_RTAI_TASKS; > g) the RX budget is 10 packets -- just an empirical value which can be > changed via a module param. >
Thanks again for your work, it will serve as a valuable foundation for an upstream merge. Once time permits, I will start the remaining work. Could you sent me your last rt_e100.c version attached? This patch was wrapped once again. Jan
signature.asc
Description: OpenPGP digital signature
------------------------------------------------------------------------------
_______________________________________________ RTnet-users mailing list RTnet-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/rtnet-users
