acooks / tn40xx-driver

Linux driver for tn40xx from Tehuti Networks
71 stars 50 forks source link

tn40xx 0000:07:00.0 enp7s0: fw 0xe [ 644.481235] BUG: unable to handle page fault for address: fffff64d83ac3c00 [ 644.481247] #PF: supervisor read access in kernel mode #64

Open afrojewelz opened 8 months ago

afrojewelz commented 8 months ago

[Uploading dmesg.txt…]() image

root@asrb55z:~# lspci -nnk -s 07:00 07:00.0 Ethernet controller [0200]: Tehuti Networks Ltd. TN9710P 10GBase-T/NBASE-T Ethernet Adapter [1fc9:4027] Subsystem: Tehuti Networks Ltd. Ethernet Adapter [1fc9:3015] Kernel driver in use: tn40xx image

use dkms add -m tn40xx/006 && dkms install -m tn40xx/006 on latest proxmox which is kernel 6.5.11-7-pve

the compiled module can be load,but whenever memory is occupied enough or running stress mem will result BUG unable to handle page fault for address: xxxx then RIP my applications. although it's not fatal but it's causes unstable to sustain file service i had onboard 2.5G realtek NIC when i switch to this one ,i can run stress memory whenever i want and stable which won't result RIP or BUG or crash. so if any great man can look into codes and try to debug them. btw my phy is MV88X3310.so i add hdr manually.

already ruled out the overclock issue , even return to default clk still occurred memleak event with module tn40xx loaded while memory stressed

acooks commented 8 months ago

You could try the potential fix in commit 493dcc7fc80ac41ad40dfe3feccf735102944801

afrojewelz commented 8 months ago

image

when i only apply this line "if (likely(db->rptr->len > 0))" it's still triggers page fault to a page that kernel can't alloc. i could try rest of patches

afrojewelz commented 8 months ago

nope, after applied two changed files tn40xx.c & tn40xx.h in dev branch which was update 3weeks ago.still triggers page crashed. image it won't run normal until me unload the module and reboot with clean kernel without tn40xx.ko