First crash ever

[hivetech]

Hi all
I'm running pfSense since 2 or 3 years now and usually I had uptimes from one release to another (Kudos to the devs).
I'm also a pfSense Manual Subscriber and helped a little bit with german translations.
Now I had a unexpected Crash two days ago, and because I'm was on Holidays, I was not able to submit crashlogs til today, 16th August till aroud 22:00 GMT+2.
My IP is: 185.63.x.x
I had a look at it before I uploaded it but I was not able to see an error, please help me out if you like.
When I submitted the crashlog it was deleted after or I am just too stupid to find it again.

Best regards
Dario

[jimp]

I don't see a crash from today from that address. I see one from the 8th and it was a libreadline.so error that isn't really a crash, but an expected byproduct of a snapshot update around that time.

If you have IPv6, it may have submitted the crash over IPv6 instead.

[hivetech]

IPv6 address is: 2a04:f340:9ff….....
thx for your fast response Jim

[jimp]

There was a crash from that address.

db:0:kdb.enter.default>  bt
Tracing pid 69828 tid 100940 td 0xfffff80257318000
m_tag_delete_chain() at m_tag_delete_chain+0xa3/frame 0xfffffe02361fd800
m_freem() at m_freem+0x63/frame 0xfffffe02361fd820
sbflush_internal() at sbflush_internal+0x20/frame 0xfffffe02361fd840
sbdestroy() at sbdestroy+0x12/frame 0xfffffe02361fd860
sofree() at sofree+0x18e/frame 0xfffffe02361fd890
soclose() at soclose+0x34f/frame 0xfffffe02361fd8d0
_fdrop() at _fdrop+0x29/frame 0xfffffe02361fd8f0
closef() at closef+0x21e/frame 0xfffffe02361fd980
fdescfree() at fdescfree+0x4f9/frame 0xfffffe02361fda30
exit1() at exit1+0x576/frame 0xfffffe02361fdac0
sys_sys_exit() at sys_sys_exit+0xe/frame 0xfffffe02361fdad0
amd64_syscall() at amd64_syscall+0x40f/frame 0xfffffe02361fdbf0
Xfast_syscall() at Xfast_syscall+0xfb/frame 0xfffffe02361fdbf0
--- syscall (1, FreeBSD ELF64, sys_sys_exit), rip = 0x800b1c18a, rsp = 0x7fffdf9fa5a8, rbp = 0x7fffdf9fa5c0 ---

Fatal trap 9: general protection fault while in kernel mode
cpuid = 3; apic id = 06
instruction pointer	= 0x20:0xffffffff80b82f23
stack pointer	        = 0x28:0xfffffe02361fd780
frame pointer	        = 0x28:0xfffffe02361fd800
code segment		= base 0x0, limit 0xfffff, type 0x1b
			= DPL 0, pres 1, long 1, def32 0, gran 1
processor eflags	= interrupt enabled, resume, IOPL = 0
current process		= 69828 (dpinger)

It looks similar to https://redmine.pfsense.org/issues/4689 or https://redmine.pfsense.org/issues/7149

You may want to edit /boot/loader.conf.local and put in this:

hw.igb.num_queues="1"

And then reboot.

[hivetech]

Hi Jim

I've added the mentionend line and rebooted the device.
But regarding the fact, that it was my first crash after nearly 3 years, I think this wouldn't happen so fast again.
Is there a way to "provocate" that bug, or should I just wait til 2.4?
Thanks anyway for your help.

[jimp]

Wait for 2.4, or depending on the cause, 2.4.1. If it's stable with hw.igb.num_queues="1" there is no harm in running that way for now, except perhaps with lower throughput depending on the load.