Wetek 2 random kernel freeze

  • Hardware: WETEK Play2
    Software Version: 7.95.3
    Version: NAND

    Hi all,

    I have random freeze while reading videos on 7.90.10 and 7.95.3. I've only tried MKV files with audio passthrough on S/PDIF .

    It seems to be a kernel issue : the box does not ping anymore and pushing the power button has no effect.

    There is absolutely no log in the (serial) console.

    The console seems to be read-only, Is there a way to use the serial port with gdb or something to have a stacktrace ?

    Any idea ?

    Olivier

  • Update.

    I've finally spot the bug, it's an issue with the Ethernet (wired) driver. The box lost the network, and because of NFS, it freeze all the system most of the time.

    Sometimes, the GUI of not frozen, and in that case, I can disable/enable the wired connection and everything goes well (for 30 minutes).

    I can reproduce the bug just by running iperf (without any NFS share mounted, nor playing a video), but It take a random time (from 10 minutes to 4 hours ...).

    A good way to reproduce this, is to run a movie over NFS and if it does not crash in 40 minutes, start an iperf test.

    There absolutely no log with a journalctl -kr (video stopped at 23:43, start an iperf test, the crash occured 3-5 minutes later)

    -- Logs begin at Thu 2015-01-01 01:00:11 CET, end at Sat 2017-02-25 23:56:11 CET. --
    Feb 25 23:56:04 LibreELEC kernel: IPv6: ADDRCONF(NETDEV_CHANGE): eth0: link becomes ready
    Feb 25 23:56:04 LibreELEC kernel: libphy: stmmac-0:00 - Link is Up - 1000/Full
    Feb 25 23:56:01 LibreELEC kernel: scancode = 0x001c, maptable = 0,code:0x00000000
    Feb 25 23:56:01 LibreELEC kernel: remote:
    Feb 25 23:56:01 LibreELEC kernel: remote: release ircode = 0x47,
    Feb 25 23:56:01 LibreELEC kernel: IPv6: ADDRCONF(NETDEV_UP): eth0: link is not ready
    Feb 25 23:56:01 LibreELEC kernel: scancode = 0x001c,maptable = 0,code:0xb8475f5e
    Feb 25 23:56:01 LibreELEC kernel: remote:
    Feb 25 23:56:01 LibreELEC kernel: remote: press ircode = 0x47,
    [more scancode ...] <--- reenabling the network
    0x0067,maptable = 0,code:0xaf505f5e
    Feb 25 23:55:44 LibreELEC kernel: remote:
    Feb 25 23:55:44 LibreELEC kernel: remote: press ircode = 0x50,
    Feb 25 23:45:04 LibreELEC kernel: aml_snd_card: audio speaker off
    Feb 25 23:44:49 LibreELEC kernel: aml_snd_m8_card aml_m8_snd.45: i2s/958 same source
    Feb 25 23:44:49 LibreELEC kernel: hdmitx: audio: no update
    Feb 25 23:44:49 LibreELEC kernel: hdmitx: audio: aout notify size 32
    Feb 25 23:44:49 LibreELEC kernel: hdmitx: audio: aout notify rate 44100
    Feb 25 23:44:49 LibreELEC kernel: aml_audio_hw: IEC958 PCM32
    Feb 25 23:44:49 LibreELEC kernel: aml_spdif_dai: share the same clock
    Feb 25 23:44:49 LibreELEC kernel: aml_spdif_dai: aml_hw_iec958_init,runtime->rate=44100, runtime->channels=2, same source mode(1)
    Feb 25 23:44:49 LibreELEC kernel: aml_snd_m8_card aml_m8_snd.45: IEC958 playback disable
    Feb 25 23:44:49 LibreELEC kernel: aml_snd_m8_card aml_m8_snd.45: I2S playback disable
    Feb 25 23:43:49 LibreELEC kernel: scancode = 0x00a6, maptable = 0,code:0x00000000
    Feb 25 23:43:49 LibreELEC kernel: remote:
    Feb 25 23:43:49 LibreELEC kernel: remote: release ircode = 0x2b,
    [Stopping a video, then launch an iperf test]

  • It seems to be this bug : thread-4988.html

    Because the patch is not on 8.0, I've compile the latest master libreelec-8.0 branch, and after a night of iperf was not able to trigger the crash.

    I'll continue my monologue if the crash occurs again ...