SJA1105 stops counting packets, and stop sending frame after 30sec to 2 minutes

[sberthiaume2000]

Hi,

we are using an MCU which is connected to the SJA1105 via MII.
We also have 3 other Ethernet link connected to the SJA1105 (2 via MII, and 1 via RMII)

I've used the configuration tool in python,a nd used the example sja1105QS_simple.py (cause we are using the SJA1105Q), I've changed the port and speed according to ours needs (all 100Mbps)

mii_mode_parameters.append(
{
"xMII_MODE[0]": MII,
"PHY_MAC[0]": MAC_MODE,
"xMII_MODE[1]": MII,
"PHY_MAC[1]": MAC_MODE,
"xMII_MODE[2]": RMII,
"PHY_MAC[2]": MAC_MODE,
"xMII_MODE[3]": MII,
"PHY_MAC[3]": PHY_MODE,
"xMII_MODE[4]": MII,
"PHY_MAC[4]": MAC_MODE,
})

Using the MCU I can update the switch configuration using the SPI, I get a Valid configuration once pushed.

My MCU is connected on PHY_MAC[3], so it is in MII and PHY_MODE, since my MCU will act as a MAC.

Now, on reboot we are able to PING the MCU for approximately 2 minutes, then, all OUTGOING packet from the MCU are totally ignored by the switch (no packet counts, and haven't seen any errors)

we can reboot the device, and same thing will happen again.
We have check the MII signals between the MCU and the switch, they are all looking good (we send packets every 2 seconds, so we see action on signals every 2 seconds)... 

I would have expected either a valid packet count increase or some errors....

I have attach the .py configuration.. like I said, did not modified much..

If anyone has ideas....

Thanks

[abdullahzafar]

Hi,

Did you try using a different port to interface with MCU? or did you try RMII for interfacing?

Did you face the same issue? 

BR,

Abdullah

[sberthiaume2000]

Here's the switch port 3 stats 

Current Docking RX packet:
NB RX Broadcast : 0x3
NB RX Multicast : 0x0
NB RX Bytes 64 : 0x1
NB RX Bytes 65-127 : 0x0
NB RX Bytes 128-255 : 0x0
NB RX Bytes 256-511 : 0x0
NB RX Bytes 512-1023 : 0x0
NB RX Bytes 1024-1518 : 0x2

Drop RX untagged : 0x0
Drop RX single inner tag : 0x0
Drop RX single outer tag : 0x0
Drop RX double tag : 0x0
Drop RX illegal dble tag : 0x0
Drop RX drop empty route : 0x0
Drop RX drop MAC notlearn: 0x0

High Level port status:
Not reacheable : 0x0
Egress Disabled : 0x0
Part drop : 0x0
Queue Full : 0x0
Queue Level_0 : 0x0
Queue Level_HWM_0 : 0x1
Queue Level_1 : 0x0
Queue Level_HWM_1 : 0x0
Queue Level_2 : 0x0
Queue Level_HWM_2 : 0x0
Queue Level_3 : 0x0
Queue Level_HWM_3 : 0x0
Queue Level_4 : 0x0
Queue Level_HWM_4 : 0x0
Queue Level_5 : 0x0
Queue Level_HWM_5 : 0x0
Queue Level_6 : 0x0
Queue Level_HWM_6 : 0x0
Queue Level_7 : 0xe
Queue Level_HWM_7 : 0x0
High Level port statistics:
TX BYTE : 0x1804
TX BYTESH : 0x0
TX FRAME : 0x29
TX FRAMESH : 0x0

RX BYTE : 0x86c
RX BYTESH : 0x0
RX FRAME : 0x3
RX FRAMESH : 0x0

Policy Error : 0x0
Crit Policy Error : 0x0
VLAN not found : 0x0
RX CRC error : 0x0
Size error : 0x0
Unreleased : 0x0
VLAN error : 0x0
N664ERR : 0x0

Port status:
MAC-level diag counter : 0x10000
MAC-level diag flags : 0x0


TX Stats still increments, but all RX stats are stuck...

[sberthiaume2000]

Forgot to add, we are able to ping in between other PHY (lan-2-lan) with no problem, it is only the MCU link we have issues with...