Posts by jim_p

    I changed the vm's vga to "vmsvga", enabled 3d acceleration and maxed its video memory to 128mb. The best thing it managed to do is to show libreelec's splash screen, followed by 4-5 failed attempts to start xorg and then a reboot.
    What is worse, while the vm could not start xorg, it was keeping the mouse captured, so I had no way to close it. I switched to a tty and force-closed it with

    Code
    vboxmanage controlvm LibreELEC poweroff

    Le 12 on virtualbox: does not work, many issues, would not recommend, 1 star :P

    As for the original issue that I opened the thread, it seems to still be present as you can see in the screenshots below.

    lspci shows that libreelec/linux recognizes video card as VMWare SVGA II adapter.

    The lspci output above says that it is a virtualbox vga. So, if there is no driver for it in the supplied ova file, this means it is incompatible with virtualbox. If this is really the case, I would like it to be mentioned somewhere, like the generic image states "no nvidia support" in parentheses.

    As for the hijacking, I opened this thread a year ago and my issue was that le could not run properly on vmware... because it had no free space to do anything and, as the dev said, it was an issue with the ova file. I remembered it when I decided to try le 12 in a vm, but I was hoping it was fixed, because the pull request was indeed merged. I am now downloading gparted to check its size.

    I learned about brcm80211 and its family of drivers (brcmfmac and brcmsmac) a few days ago, because I thought it was either b43 or wl for the broadcoms. I guess I am the lucky one whose cards are supported by neither of the brcm80211 family. And if I recall correctly, I had asked in the past for le to include b43 instead of wl.

    My problem is how this (wireless card on an 15 year old laptop I use with an old x64 version of le, so older kernel)

    Code
    04:00.0 Network controller [0280]: Broadcom Inc. and subsidiaries BCM4312 802.11b/g LP-PHY [14e4:4315] (rev 01)
           Subsystem: Broadcom Inc. and subsidiaries BCM4312 802.11b/g LP-PHY [14e4:04b5]
           Kernel driver in use: wl
           Kernel modules: wl


    runs better than this (wireless card on a 9-10 year old laptop I set up a week ago with debian unstable and 6.9.8 today)

    Code
    02:00.0 Network controller [0280]: Broadcom Inc. and subsidiaries BCM43228 802.11a/b/g/n [14e4:4359]
           Subsystem: Dell Device [1028:0014]
           Kernel driver in use: bcma-pci-bridge
           Kernel modules: bcma


    On this laptop, which I will use until the end of summer probably, I have to select between a poor performing b43 or a wl that gives a kernel panic on boot. Last update for the wl driver, for fixing compatibilities or builds with the kernels was on March 4th 2024, so it is pretty fresh.

    As for the rpi... don't get me started. A few years ago, when I first got my hands on a 4gb rpi4, I had commented on the lack of a purely arm64 os from the rpi foundation and how dhcpcd is insufficient. Oh what a blasphemy that was for the fanboys! Fast forward to today and to rpi5, there is finally an arm64 image, although not fully 64bit (it has 32bit userspace as someone said) and dhcpcd was ditched in favor of network manager.

    Are you sure? Because I am pretty sure even le 12 uses wl for the broadcom chips,

    My issue it is not libreelec related though. The kernel panic brought back memories from that tvbox and I said "omg please do not have the same chip with the tvbox"! The only hardware I have used extensively and has a broadcom wifi chipset is my rpi3b, and its wifi is... bad, plain and simple.

    Please excuse my semi-offtopic. May I ask what wifi chip my tvbox had? Please don't say broadcom :D

    Excluding my rpi3b, this is the first time I come accross a broadcom wifi chip (bcm43228). It peforms like trash with b43, with the download speeds maxing at 300kb/s. It performs perfectly with wl, but it gets a kernel panic on boot.

    Regular x64 distro installation.

    No, it is about x64, the generic-legacy img and my old laptop. Sorry for not mentioning it.

    The le 11 issue with wifi is/was platform independant though. I faced it on the nightlies of le 11 from May to September of 2022 on my rpi3b+, until I gave up completely on le there.

    Thank you for the new release. I just downloaded it and I am testing it now.

    As it seems, it still has the same issue with wifi that le 11 had, because iwd and connman could not work well together. Instead of rebooting until it works, which is what I did on le 11, this time I tried disabling and re-enabling it from le settings. The first time it started after 2 or 3 tries. This time I have already tried 9 or 10, but it still does not come up, so I connected via ethernet to ssh to it, get logs etc.

    So, what should I include on the new thread I am about to open?

    Sorry for resurrecting this thread, but it is the most relevant one to my issue.

    I have this gamepad

    Code
    [  305.188133@0] Indeed it is in host mode hprt0 = 00041901
    [  305.390771@0] usb 2-1: New USB device found, idVendor=0810, idProduct=0001
    [  305.390787@0] usb 2-1: New USB device strings: Mfr=0, Product=2, SerialNumber=0
    [  305.390794@0] usb 2-1: Product:  USB Gamepad
    [  305.421210@0] input:  USB Gamepad           as /devices/lm0/usb2/2-1/2-1:1.0/input/input2
    [  305.433236@0] pantherlord 0003:0810:0001.0001: input,hidraw0: USB HID v1.10 Joystick [ USB Gamepad          ] on usb-lm0-1/input0
    [  305.433304@0] pantherlord 0003:0810:0001.0001: Force feedback for PantherLord/GreenAsia devices by Anssi Hannula <[email protected]>


    this udev rule

    Code
    SUBSYSTEM=="input", ATTRS{name}==" USB Gamepad          ", KERNEL=="event*", MODE="0666", ENV{LIBINPUT_IGNORE_DEVICE}="1"
    SUBSYSTEM=="input", ATTRS{name}==" USB Gamepad          ", KERNEL=="event*", MODE="0666", ENV{ID_INPUT_JOYSTICK}="1"


    and I want it to work on dtech's build of le 9.2.8.x for amlogic devices. I have also tested it on le 9.2.6 and 10.0.4 on x64, which are the only ones I have available on my end.

    Nothing moves, nothing reacts to button presses. The best I have made it do was to react on some button presses when I am under the input settings. Pressing some buttons there poped up a dialog to configure the pad, I tried it, but again no reaction to button presses. If it helps, the behavior is the same regardless if I set the joystick driver to linux, udev or none.

    Errr... probably not. Here is the dmesg.
    https://paste.libreelec.tv/unbiased-cattle.log

    In case I did not mention it before, I would like to add that the tvbox becomes visible on the network on the second boot and I don't know when it started. The routine is: I disconnect it from the tv, get it close to my router and then wait for it to show up under my router's connected devices so as to ssh to it and update it. When the boot takes more than like 2 minutes, I unplug it and plug it again, and it then shows up.

    In other news, yesterday I failed to install its stock firmware. That is why I asked for the wireless chip, Good thing it still boots from the sd.

    Thanks. The annoying thing on my end is that lately I have to boot the tvbox, wait for like 5 minutes for it to show up on my router's connected clients or respond to a ping and, if it does not, then power it off, remove the sd and power it on again (and repeat it as many times I have to) until it connects properly.

    When I am updating it, I disconnect it from the tv, I get it close to my router so that the wifi signal is good. If I leave it on the tv, even the ssh section will be laggy.