kernel: FPM Pool 1: invalid token 0x21950000 freed

[Kabouter Plop]

After updating to 386.4 not a single error...really hope it will stay like that..
But couldn't hurt to try to crimp a new connector if these errors start again.

Thx for the info

You can just tell if the order is 1 of these 2 its fine

If its 568B for example on 1 end it should be on other end as well unless 2 of same devices are connected i believe.
I bassicly had white orange and orange swapped in place on 1 end on my modem, if the one who terminated it did not check it then its quite likely to have issue like that, because anyone that knows how to use cable tester would find out its correct or not.

 

[rawdmon]

I came across this thread and just wanted to shed a bit more light on these log messages...

"FPM Pool 0: invalid token 0x003ff000 freed"

FPM stands for "Frame Pool Manager". I suspect that the associated Broadcom module (BCM_FPM which depends on BCM_RDP and BCM94908) is being used to create pools in memory where network frames are stored as sort of a buffer while undergoing some kind of processing.

I suspect that these "invalid token" messages may just mean that data in one of the pools was found to no longer be relevant and is therefore discarded.

I'm not having any problems with my router at all and I'm still seeing these in my logs. These are likely part of normal operation and in most cases are probably just a red herring that people are focusing on because they can get quite noisy in the log, so when people are having issues they'll definitely notice these if they are happening and probably think that they are the cause. In my opinion seeing a flood of these in the log is generally going to be a symptom rather than the cause of an issue.

I've noticed that I see a fair number of these any time I do something that causes AiMesh nodes to need to reconnect to the main router. I'm thinking that maybe having so many connections abruptly terminate causes the frames in the pools to become no longer relevant (invalid) and that they then get discarded to free up space in the pools. This is all just a theory since there isn't really any documentation on this stuff but it would make sense.

Just to add, I was having some weird instability problems where my main router (a GT-AX11000 serving 5 AiMesh nodes and 35+ devices, so it's fairly busy) would freeze up for a while or reboot on it's own but it just turned out to be thermal related. I bought a cheap USB fan by AC Infinity off Amazon and placed it on top of the router which brought the CPU temp down by 20+ degrees (celsius) and I haven't had any stability problems since. Just something to look at if you are seeing issues with stability.