Hi Chris, Chris Verges wrote: > Hi Wolfgang, > > I've upgraded to v2.6.32 and am building the entire SocketCAN subsystem > into the kernel (monolithic). Our electrical engineer has confirmed > that this is the proper initialization command for the can0 interface: > > ip link set can0 \ > type can \ > tq 90 \ > prop-seg 1 \ > phase-seg1 5 \ > phase-seg2 4 \ > sjw 2 > > This results in the following: > > # ip link set can0 type can tq 90 prop-seg 1 phase-seg1 5 phase-seg2 4 > sjw 2 > at91_can at91_can: writing AT91_BR: 0x00081043 > > # ip -d -s link show can0 > 2: can0: <NOARP,ECHO> mtu 16 qdisc pfifo_fast state DOWN qlen 1000 > link/can > can state STOPPED restart-ms 0 > bitrate 1003313 sample-point 0.636 > tq 90 prop-seg 1 phase-seg1 5 phase-seg2 4 sjw 2 > : tseg1 4..16 tseg2 2..8 sjw 1..4 brp 2..128 brp-inc 1 > clock 99328000 > re-started bus-errors arbit-lost error-warn error-pass bus-off > 0 0 0 0 0 0 > RX: bytes packets errors dropped overrun mcast > 0 0 0 0 0 0 > TX: bytes packets errors dropped carrier collsns > 0 0 0 0 0 0 > > Unfortunately, we still see cansend hang: > > # cansend can0 123# > BUG: soft lockup - CPU#0 stuck for 61s! [cansend:1424] > Modules linked in: > > Pid: 1424, comm: cansend > CPU: 0 Not tainted (2.6.32 #1) > PC is at at91_poll+0x294/0x4d8 > LR is at net_rx_action+0xa0/0x218 > pc : [<c0199440>] lr : [<c0218ba8>] psr: 00000013 > sp : c3979d98 ip : c3979dd8 fp : 4026e000 > r10: 0000000c r9 : c03ae1f0 r8 : 00000000 > r7 : 4426e000 r6 : c39be2c0 r5 : 0000000c r4 : c39be000 > r3 : 04000000 r2 : c3978000 r1 : 0000000c r0 : 00000000 > Flags: nzcv IRQs on FIQs on Mode SVC_32 ISA ARM Segment user > Control: 0005317f Table: 23b4c000 DAC: 00000015 > > There definitely seems to be interaction with the at91_poll() function > in this "hang" time. I'm wondering why the driver isn't failing > gracefully. I'll start diving into the code, but hopefully someone else > with more experience on the at91_can driver will have a quicker insight.
Hm, the hang is not OK, of course, and might be due to a software issue. Nevertheless, to you see any message going out to the wire? Wolfgang. > Thanks, > Chris > _______________________________________________ > Socketcan-users mailing list > [email protected] > https://lists.berlios.de/mailman/listinfo/socketcan-users > > _______________________________________________ Socketcan-users mailing list [email protected] https://lists.berlios.de/mailman/listinfo/socketcan-users
