kernel crash when use MC7430 modem

[Bjørn Mork]

dailijin <dailijin126 at 126.com> writes:

> My unit use MC7430 modem based on kernel 4.4.0 version, after run over
> all night, I found the system automatically reboot one time due to
> kernel crash. if there is anyone encounter this same issue.
>
>
> Though this issue only happen one time, but I still want to fix the
> potential bug, welcome any suggestion.
>
>
> the detail log as below.
>
>
> [25478.065408] skbuff: skb_under_panic: text:ffffffff8164407b len:-65296 put:-65471 head:ffff88023d47e740 data:ffff88013d48e73f tail:0xef end:0x640 dev:wwan12
> [25478.065459] ------------[ cut here ]------------
> [25478.065460] Kernel BUG at ffffffff81656eff [verbose debug info unavailable]
> [25478.065462] invalid opcode: 0000 [#1] SMP 
> [25478.065464] Modules linked in: drbg ansi_cprng ctr ccm GobiSerial(O) vfb arc4 rt2800usb rt2x00usb rt2800lib rt2x00lib mac80211 ftdi_sio cfg80211 crc_ccitt ipt_REJECT nf_reject_ipv4 xt_tcpudp xt_recent xt_conntrack iptable_filter ipt_MASQUERADE nf_nat_masquerade_ipv4 iptable_nat nf_conntrack_ipv4 nf_defrag_ipv4 nf_nat_ipv4 nf_nat nf_conntrack ip_tables x_tables x86_pkg_temp_thermal kvm_intel snd_hda_codec_hdmi kvm irqbypass snd_hda_codec_realtek snd_hda_codec_generic huawei_cdc_ncm aesni_intel aes_x86_64 lrw gf128mul glue_helper ablk_helper cryptd option(O) usb_wwan usbserial snd_hda_intel snd_pcsp snd_hda_codec snd_hwdep snd_hda_core blackmagic_io(PO) snd_pcm snd_timer cdc_ncm snd soundcore usblp i2c_i801 cdc_acm(O) shpchp i2c_hid tpm_tis tpm acpi_pad qmi_wwan(O) cdc_wdm(O) usbnet(O) mii coretemp
> [25478.065490]  usb_storage e1000e(O) ptp i915 pps_core intel_gtt i2c_algo_bit drm_kms_helper(O) syscopyarea sysfillrect sysimgblt fb_sys_fops drm(O) agpgart video sunrpc
> [25478.065499] CPU: 2 PID: 0 Comm: swapper/2 Tainted: P           O    4.4.0 #1


So that is an old and outdated kernel, with modified qmi_wwan, cdc_wdm
and usbnet source.  Not exactly a the best starting point for debugging
such issues...

But anyway, I wonder if you are hitting this issue which was recently
fixed by Kristian: https://patchwork.kernel.org/patch/10046605/

Whoch of course isn't backorted to v4.4, since the bug wasn't introduced
there.  But I guess you have backported the original bug yourself?  You
know that you are then responsible for backporting any fixes to it, too?




Bjørn