Luotao Fu wrote: > Hi, > > On Wed, Nov 04, 2009 at 09:13:03AM +0100, Wolfgang Grandegger wrote: >> Luotao Fu wrote: >>> Hi Wolfgang, >>> >>> On Tue, Nov 03, 2009 at 07:54:21PM +0100, Wolfgang Grandegger wrote: >>>> Wolfgang Grandegger wrote: >>>>> Hi Fu, >>>>> >>>>> Luotao Fu wrote: >>>>>> Hi, >>>>>> in following the new patch as discussed yesterday. I add a XXX marked >>>>>> comment to this. Works all right on my setup. Please review. >>>>> I will try it on my setup later today or tomorrow. >>>> I just started to do some tests, sorry for the delay. When I stop the >>>> device without any traffic, I get the message: >>>> >>>> # ./ip link set can2 up type can bitrate 125000 >>>> # ifconfig can2 down >>> should the ifconfig down not come before setting the bitrate? >> The device was not yet started (up). >> > > ? I'm a little confused, just for the record: You called > ifconfig can2 down > before you brought up the interface?
No, because the module was just loaded. But even if I do a "ifconfig can2 down", it will fail. >>>> mpc52xx_can f0000980.can: SLPRQ timeout reached, failed to enter sleep >>>> mode! >>>> >>>> I have a MPC5200 v1.2. I assume that you do not see that behavior on >>>> your test board!? >>>> >>> Yes I do. You just ran into the ominous hardware bug we discussed and >>> the patch is for. I replaced the error return value with a warning here. >>> This is exactly the fix we need so that the device won't get frozen. >> You mean, an "ifconfig canX up" followed by a "ifconfig canX down" will >> always through the warning above? If yes, a simple mdelay() would do the >> jobs as well. >> > > No. The slprq will eventually fail if there're irregular activities on > the bus, like a opposite with wrong baudrate keep resending its stuffs. On my board, it always fails! Even without any traffic on the bus. Wolfgang. _______________________________________________ Socketcan-users mailing list [email protected] https://lists.berlios.de/mailman/listinfo/socketcan-users
