Regular Crashes, no hardware issues

ThaChillera

I've now had 5 total crashes, about one day a week twice a day. No hardware issues that I can find, ran memtest for multiple passes, prime95, fsck, no SMART errors.

Hardware is PCEngine's APU2e0, with an msata ssd.

Any advice would be really appreciated, this is my first experience with pfsense, and I really like it and would like to keep using it :)

Crash reports:

stephenw10

Hmm, you appear to have at least 3 different crashes there.

1st and 5th are similar:

db:0:kdb.enter.default>  bt
Tracing pid 0 tid 100030 td 0xfffff800041d9740
kdb_enter() at kdb_enter+0x37/frame 0xfffffe0000432b80
vpanic() at vpanic+0x197/frame 0xfffffe0000432bd0
panic() at panic+0x43/frame 0xfffffe0000432c30
trap_fatal() at trap_fatal+0x391/frame 0xfffffe0000432c90
trap_pfault() at trap_pfault+0x4f/frame 0xfffffe0000432ce0
trap() at trap+0x286/frame 0xfffffe0000432df0
calltrap() at calltrap+0x8/frame 0xfffffe0000432df0
--- trap 0xc, rip = 0xffffffff8109623b, rsp = 0xfffffe0000432ec0, rbp = 0xfffffe00004330f0 ---
pf_test() at pf_test+0x19db/frame 0xfffffe00004330f0
pf_check_in() at pf_check_in+0x1d/frame 0xfffffe0000433110
pfil_run_hooks() at pfil_run_hooks+0xa1/frame 0xfffffe00004331b0
ip_input() at ip_input+0x475/frame 0xfffffe0000433260
netisr_dispatch_src() at netisr_dispatch_src+0xca/frame 0xfffffe00004332b0
ng_iface_rcvdata() at ng_iface_rcvdata+0x131/frame 0xfffffe00004332f0
ng_apply_item() at ng_apply_item+0x8c/frame 0xfffffe0000433380

2nd and 4th are the same crash:

db:0:kdb.enter.default>  bt
Tracing pid 11 tid 100003 td 0xfffff800041ac000
kdb_enter() at kdb_enter+0x37/frame 0xfffffe00183767e0
vpanic() at vpanic+0x197/frame 0xfffffe0018376830
panic() at panic+0x43/frame 0xfffffe0018376890
trap_fatal() at trap_fatal+0x391/frame 0xfffffe00183768f0
trap() at trap+0x67/frame 0xfffffe0018376a00
calltrap() at calltrap+0x8/frame 0xfffffe0018376a00
--- trap 0x1, rip = 0xffffffff80daa10e, rsp = 0xfffffe0018376ad0, rbp = 0xfffffe0018376af0 ---
mi_switch() at mi_switch+0xfe/frame 0xfffffe0018376af0
sched_idletd() at sched_idletd+0x63/frame 0xfffffe0018376bb0
fork_exit() at fork_exit+0x7e/frame 0xfffffe0018376bf0
fork_trampoline() at fork_trampoline+0xe/frame 0xfffffe0018376bf0
--- trap 0, rip = 0, rsp = 0, rbp = 0 ---

3rd is different:

panic: Unregistered use of FPU in kernel
cpuid = 0
time = 1644003790
KDB: enter: panic

That seems more like hardware or at least something introduced by a hardware issue.

There are some known issues in some Coreboot versions on the APU. What version are you using?

Steve

ThaChillera

I'm running 1.14.0.6, the recommended version according to https://pcengines.github.io/.

Do you have any other hardware tests you can recommend to prove/rule out hardware issues?

stephenw10

Mmm, you might want to try one of the new versions. There's a whole thread detailing other users experiences:
https://forum.netgate.com/topic/95148/pc-engines-apu2-experiences

Coreboot versions are discussed at length there.

What packages are you running? Any unusual config?

Steve

ThaChillera

@stephenw10 Sorry for the long delay, busy & confused.

So the version number I mentioned was the SeaBios version, the actual bios version is 4.14.0.6. But as I mentioned it is the latest version recommended by PCEngines. I also haven't found any mention of random reboots in the linked thread, but its quite long so maybe I wasn't searching for the correct keywords.

No custom packages were installed, any configuration was through the PFSense webinterface, so I wouldn't expect any weirdness. the only notable thing is that I use PPPoE for internet access, which might be uncommon and thuis badly tested, but I doubt that that is the issue.

stephenw10

PPPoE is quite common, I also doubt that's the issue.

It looks at least partly like hardware given the randomness. Perhaps your hardware testing os just not hitting it for some reason?

Have you updated to 2.6 or 22.01 to test there?