Hello Thomas,
I'm currently rewriting the mcp25xxfd CAN driver and observing a strange
effect on the mcp2517fd. I'll retest thi on the mcp2518fd later.
The test is to send CAN-2.0 frames at 500 kbit/s with a gap of ~100ms
with dlc=1 and incrementing CAN-ID and first data byte from can0 to
can1. Both interfaces are on the same host, both are mcp2517fd.
After some time I short CAN-H and CAN-L for several seconds, until this
problem occurs.
This is the log, ignore the fact that the CAN frames are shown as RX'ed
before TX'ed.
CANID DLC data[0]
> (2019-12-09 19:09:04.583745) can1 RX - - 047 [1] 47
> (2019-12-09 19:09:04.583766) can0 TX - - 047 [1] 47
> (2019-12-09 19:09:04.683866) can1 RX - - 048 [1] 48
> (2019-12-09 19:09:04.683885) can0 TX - - 048 [1] 48
> (2019-12-09 19:09:05.384905) can1 RX - - 048 [1] 48
> (2019-12-09 19:09:05.485030) can1 RX - - 049 [1] 49
The problem is, that can1 receives the message with CANID 048 two times.
Looking at the TEF Objects, it shows that the id=048 CAN frame is
recorded twice as send. TEF Object 0x0 and TEF Object 0x1 both have
id=0x48 and Sequence Number=0x48.
> [ 125.607608] TEF Object: 0x00 (0x400)
> [ 125.611174] id = 0x00000048
> [ 125.615190] flags = 0x00009001
> [ 125.619194] ts = 0x344e4c79
> [ 125.623195] SEQ = 0x000048 Sequence
> [ 125.627904]
> [ 125.629390] TEF Object: 0x01 (0x40c) priv-HEAD chip-TAIL priv-TAIL priv-FIFO-full
> [ 125.637227] id = 0x00000048
> [ 125.641229] flags = 0x00009001
> [ 125.645244] ts = 0x35fa2cba
> [ 125.649249] SEQ = 0x000048 Sequence
> [ 125.653959]
> [ 125.655444] TEF Object: 0x02 (0x418)
> [ 125.659011] id = 0x00000049
> [ 125.663012] flags = 0x00009201
> [ 125.667027] ts = 0x36374a5b
> [ 125.671029] SEQ = 0x000049 Sequence
> [ 125.675738]
> [ 125.677223] TEF Object: 0x03 (0x424)
> [ 125.680789] id = 0x00000043
> [ 125.684805] flags = 0x00008601
> [ 125.688810] ts = 0x331cb9e4
> [ 125.692811] SEQ = 0x000043 Sequence
> [ 125.697543]
> [ 125.699029] TEF Object: 0x04 (0x430)
> [ 125.702596] id = 0x00000044
> [ 125.706615] flags = 0x00008801
> [ 125.710620] ts = 0x3359d7d5
> [ 125.714636] SEQ = 0x000044 Sequence
> [ 125.719331]
> [ 125.720814] TEF Object: 0x05 (0x43c)
> [ 125.724396] id = 0x00000045
> [ 125.728401] flags = 0x00008a01
> [ 125.732401] ts = 0x3396f4d6
> [ 125.736417] SEQ = 0x000045 Sequence
> [ 125.741112]
> [ 125.742595] TEF Object: 0x06 (0x448)
> [ 125.746177] id = 0x00000046
> [ 125.750182] flags = 0x00008c01
> [ 125.754196] ts = 0x33d41227
> [ 125.758198] SEQ = 0x000046 Sequence
> [ 125.762892]
> [ 125.764391] TEF Object: 0x07 (0x454)
> [ 125.767961] id = 0x00000047
> [ 125.771962] flags = 0x00008e01
> [ 125.775977] ts = 0x34112f78
> [ 125.779979] SEQ = 0x000047 Sequence
In the TX Object 0x0 the id=0x48 is already overwritten by 0x50, but in
TX Object 0x1, we see id=0x49.
> [ 125.802466] TX Object: 0x00 (0x460)
> [ 125.805960] id = 0x00000050
> [ 125.809963] flags = 0x0000a001
> [ 125.813979] SEQ_MCP2517FD = 0x000050 Sequence (MCP2517)
> [ 125.819544] SEQ_MCP2518FD = 0x000050 Sequence (MCP2518)
> [ 125.825144] data = 50
> [ 125.828453]
> [ 125.829938] TX Object: 0x01 (0x470) priv-HEAD priv-TAIL priv-FIFO-full
> [ 125.836748] id = 0x00000049
> [ 125.840752] flags = 0x00009201
> [ 125.844772] SEQ_MCP2517FD = 0x000049 Sequence (MCP2517)
> [ 125.850340] SEQ_MCP2518FD = 0x000049 Sequence (MCP2518)
> [ 125.855920] data = 49
> [ 125.859234]
> [ 125.860719] TX Object: 0x02 (0x480) chip-HEAD chip-TAIL chip-FIFO-empty
> [ 125.867602] id = 0x0000004a
> [ 125.871607] flags = 0x00009401
> [ 125.875624] SEQ_MCP2517FD = 0x00004a Sequence (MCP2517)
> [ 125.881189] SEQ_MCP2518FD = 0x00004a Sequence (MCP2518)
> [ 125.886769] data = 4a
> [ 125.890083]
> [ 125.891567] TX Object: 0x03 (0x490)
> [ 125.895062] id = 0x0000004b
> [ 125.899063] flags = 0x00009601
> [ 125.903064] SEQ_MCP2517FD = 0x00004b Sequence (MCP2517)
> [ 125.908644] SEQ_MCP2518FD = 0x00004b Sequence (MCP2518)
> [ 125.914227] data = 4b
> [ 125.917536]
> [ 125.919019] TX Object: 0x04 (0x4a0)
> [ 125.922499] id = 0x0000004c
> [ 125.926512] flags = 0x00009801
> [ 125.930514] SEQ_MCP2517FD = 0x00004c Sequence (MCP2517)
> [ 125.936093] SEQ_MCP2518FD = 0x00004c Sequence (MCP2518)
> [ 125.941657] data = 4c
> [ 125.944979]
> [ 125.946463] TX Object: 0x05 (0x4b0)
> [ 125.949943] id = 0x0000004d
> [ 125.953979] flags = 0x00009a01
> [ 125.957984] SEQ_MCP2517FD = 0x00004d Sequence (MCP2517)
> [ 125.963549] SEQ_MCP2518FD = 0x00004d Sequence (MCP2518)
> [ 125.969130] data = 4d
> [ 125.972444]
> [ 125.973928] TX Object: 0x06 (0x4c0)
> [ 125.977426] id = 0x0000004e
> [ 125.981427] flags = 0x00009c01
> [ 125.985443] SEQ_MCP2517FD = 0x00004e Sequence (MCP2517)
> [ 125.991010] SEQ_MCP2518FD = 0x00004e Sequence (MCP2518)
> [ 125.996597] data = 4e
> [ 125.999906]
> [ 126.001389] TX Object: 0x07 (0x4d0)
> [ 126.004886] id = 0x0000004f
> [ 126.008890] flags = 0x00009e01
> [ 126.012891] SEQ_MCP2517FD = 0x00004f Sequence (MCP2517)
> [ 126.018471] SEQ_MCP2518FD = 0x00004f Sequence (MCP2518)
> [ 126.024054] data = 4f
Is this behaviour known, or even documented somewhere?
Feel free to call me if you need more details.
regards,
Marc
--
Pengutronix e.K. | Marc Kleine-Budde |
Embedded Linux | https://www.pengutronix.de |
Vertretung West/Dortmund | Phone: +49-231-2826-924 |
Amtsgericht Hildesheim, HRA 2686 | Fax: +49-5121-206917-5555 |
Attachment:
signature.asc
Description: OpenPGP digital signature
