Posts by BeneDicT

Paul69 I will give it a test and see. I use my box as a home server + mediacenter so I leave it running 24/7. I don't use the included remote (which is garbage BTW) as CEC is fully functioning with my TV (Samsung KU55) and AVR (Denon X1200W).

I now have a M96X Plus and its working great; complete with gigabit LAN.

In my opinion, and I suspect it will be an unpopular one around here, as it currently stands a second hand Apple TV is the best 1080/4K streaming device for services like Netflix, Stan etc. It's hands down the most widely supported. Roku is second best if your looking lower in a budget.

Where these boxes exceed is running Libreelec for local content + home server duties. It would nice to have all things in one place but thats just not where the market is right now. I use my linux box as a NAS, webserver and various other bits and pieces, retro gaming and running Kodi. It's great! I just don't expect it to be the be all and end all of home streaming solutions.

N

Thanks for the suggestion but I feel the minix, despite the better heat sink, isn’t worth twice the asking if competitors. Always planned to put some thermal past and if necessary. The m96x or the t95z seem to fit the bill though they aren’t as mainstream as MECOOL

wrxtasy would you be able to recommend a S912 device that has actual real gbit lan. Having been burnt once I'm doubting the claims from these sellers.

Thanks mate. I'll have to return the box, misleading advertisement on the sellers part. I'm guessing they don't realize they are wrong.FYI to any aussie buyers on there, these are the misleading items:

2018 2+16GB DDR4 Android 7.1 Amlogic S912 Octa Core Smart TV BOX 4K Media Movies | eBay

Cheers

Ok, further digging and it seems that the device does infact include hard ware supporting gigabit lan.

Looking deeper into /sys/dev

c9410000.ethernet has a modalias property of Camlogic, _gxbb-rgmii-dwmac. Reading networking - Clarification on Ethernet, MII, SGMII, RGMII and PHY - Stack Overflow

rgmii stands for reduced gigabit media interface.

Further google-fu indicates that c941000.ethernet is gbit:

"
Arch Linux 4.11.0-rc1-1-ARCH (ttyAML0)


alarm login: [   11.389324] meson8b-dwmac c9410000.ethernet eth0: Link is Up - 1Gbps/Full - flow control rx/tx

"

Another piece of the puzzle is that I find the following using gbit build trees:

[ 0.555597@0] eth0: No PHY found

Obviously, no issue when using 100mbit bt.

I'm a tad out of depth but would someone more knowledgeable in this area chime in? Assuming it is a build tree issue, is there a possibility of getting a tree made for the new 1gbit Lan DDR4 S912 boxes that have hit the market in the last few months.

Thanks for all the help guys.

I posted this in the other "Chose S9xx" thread but recieved no reply. Perhaps this is a more suitable place to post. Any comments? A lot of sellers are claiming gigabit LAN on this latest DDR4 generation of S912 devices. I'm trying to figure out if the sellers are telling fibs or the gigabit device tree's aren't working for some reason. Any tips?

Hi Gents,

Funnily enough wrxtasy thats basically what I have ended up opting for. In addition to what you said below, in Australia at least, the streaming app support situation for anything other than Apple TV and Telstra TV (Roku) is a joke.

I ended up pulling the trigger on a Mecool M8S Pro (DDR4 version) off ebay from an Australian "ish" seller. The tech specs claimed it included Gbit LAN and I even wrote them to double check. However after flashing and loading up 8.2.2.3 with the device tree including gbit drivers no eth0 was to be found. My question is this: is there some way (bar pulling the lid of the thing - I may decide to return it) to query whether there is in-fact a GMAC chip onboard? I loaded up the 100 Mbit device tree and it seems to be working. My understanding is that that device tree simply enables the S912 onboard ethernet and therefore isn't a conclusive proof that they are telling a fib.

Otherwise the box is working fantastic. HDMI-CEC, H265 etc are all working a dream. Highly recommended.

Cheers

PS already tried lshw, nothing particularly informative came up

So further last, a lot of sellers are claiming gigabit LAN ... M96x ...T95z and so on. They all can’t be wrong?

I tried the 2gb gig device tree no success. Should I try the other 2 gb gbit options to see if they work?

Quote from BeneDicT

Hi Gents,

Funnily enough wrxtasy thats basically what I have ended up opting for. In addition to what you said below, in Australia at least, the streaming app support situation for anything other than Apple TV and Telstra TV (Roku) is a joke.

I ended up pulling the trigger on a Mecool M8S Pro (DDR4 version) off ebay from an Australian "ish" seller. The tech specs claimed it included Gbit LAN and I even wrote them to double check. However after flashing and loading up 8.2.2.3 with the device tree including gbit drivers no eth0 was to be found. My question is this: is there some way (bar pulling the lid of the thing - I may decide to return it) to query whether there is in-fact a GMAC chip onboard? I loaded up the 100 Mbit device tree and it seems to be working. My understanding is that that device tree simply enables the S912 onboard ethernet and therefore isn't a conclusive proof that they are telling a fib.

Otherwise the box is working fantastic. HDMI-CEC, H265 etc are all working a dream. Highly recommended.

Cheers

PS already tried lshw, nothing particularly informative came up

Display More

Hi Gents,

Funnily enough wrxtasy thats basically what I have ended up opting for. In addition to what you said below, in Australia at least, the streaming app support situation for anything other than Apple TV and Telstra TV (Roku) is a joke.

I ended up pulling the trigger on a Mecool M8S Pro (DDR4 version) off ebay from an Australian "ish" seller. The tech specs claimed it included Gbit LAN and I even wrote them to double check. However after flashing and loading up 8.2.2.3 with the device tree including gbit drivers no eth0 was to be found. My question is this: is there some way (bar pulling the lid of the thing - I may decide to return it) to query whether there is in-fact a GMAC chip onboard? I loaded up the 100 Mbit device tree and it seems to be working. My understanding is that that device tree simply enables the S912 onboard ethernet and therefore isn't a conclusive proof that they are telling a fib.

Otherwise the box is working fantastic. HDMI-CEC, H265 etc are all working a dream. Highly recommended.

Cheers

PS already tried lshw, nothing particularly informative came up

Quote from wrxtasy

LE on the S912 uses what I will call hybrid Android/Linux GPU drivers. This is what draws the Kodi graphic user interface and all those overlays you see on top of video during playback.

There are NO optimised Linux GPU video drivers for the S912 available at all, no matter what the Linux OS used.

It's my understanding kszaq has "robbed" CPU processing cycles to get this hybrid GPU driver to draw the Kodi graphic user interface smoothly without user interface picture tearing that was seen in early versions of LE on S912's.

LE Kodi Leia when streaming ANY copy protected video content like Netflix has to CPU Software decode those h.264 video streams. Max. 720p at the moment because ffmpeg CPU Software decoding and playback is not yet optimised on any AMLogic LibreELEC platforms.

LE on the S912's simply cannot playback CPU Software decoded 720p h.264 smoothly due to the hybrid GPU drivers used.

All Hardware decoded video playback is unaffected by LE on the S912 when using Hybrid GPU video drivers. A co-processor called the VPU handles hardware video decoding.

CPU Software decoded video playback is affected. There really is not much of that on S9xx platforms unless you are using the as previously mentioned Kodi Leia Netflix addon.

The only Gigabit Ethernet DIY S912 option I know of that can dual boot to run LibreELEC and an unofficial Android TV ROM that will give you 1080p Netflix is the MINIX U9. DD+ 5.1 Audio from Netflix on the U9 is a mess.

For 1080p copy protected video playback, like Netflix when using Android you need a box with Widevine L1 DRM and HDCP 1.x

In the end it's a lot easier just buying something like a Plug N Play - Apple TV 4K that has really nice tvOS Firmware Apps support features for auto frame rate and auto dynamic range matching for smooth video playback with the correct SDR and 4K HDR colorspace outputs. Such Firmware Apps support is better than anything Android is offering. Impressive powerful hardware too.

Then use a complimentary - AMLogic S9xx box for LibreELEC Kodi Krypton with all the trimmings.

Display More

Hi Mate,

Considering buying this unit. Can you tell me how it has worked out for you? Can you confirm it has Gbit LAN?

Cheers

I was all ready to pull the trigger on a s912 based device but saw wrxtasy comment down the bottom. I don't intend to use the device for netflix but have plenty of h264 720 and 1080p content. Based upon the sticky I thought this would play OK.

From below this appears to not be the case?

Thanks!

Quote from wrxtasy

MINIX U1 vs ODROID C2.

U1 = excellent Wireless A2 lite mini keyboard remote (one of my favorites), AML class leading MIMO, dual antennas AC WiFi, excellent thermal engineering due to a Bloody great big heat sink in a quality hardware case.

C2 = best HDMI CEC control of all the AML Boxes, still no guarantee it works properly with all TV's / AVR's simply due to manufacturers stuffing about with their own custom CEC tweaks. Honestly you are better off using a Wireless remote anyway for Kodi responsiveness like the MINIX A2 lite remote or the aftermarket Xiaomi Mi Box remote.

C2 can also be Overclocked safely due to it's nice big heatsink, you can Overclock the CPU, GPU and also the RAM speed. Fastest device in the non HDR AML arsenal when used with eMMC storage and a wireless remote.

If you want a Wireless remote that can also turn the box OFF / ON using Infra Red buy the MINIX U1. Infra Red has to be used for OFF / ON. It's IR signal is hard coded into the devices boot loader.

Personally I would just leave these devices on 24/7. They use stuff all power anyway and have good thermal heat disappapation.

The other option which I use is power the ODROID C2 from a USB > barrel connector power cord, plug that into the TV's USB2/3 port. C2 boots up into LE from eMMC in 9 seconds when TV power is turned on. Not to be used this way if Hard drives are connected to C2. Even WiFi will struggle due simply to WiFi power draw requirements.

Use a CSR 4.0 Bluetooth USB dongle, a Xiaomi Wireless minimal remote, Overclock and you end up with a snappy setup.

*** No AML S912 box will do LE software decoding and smooth playback of 720p H264 properly for Netflix due to the relatively slow (vs S905/S905X) hybrid Linux/android GPU drivers used.

If users have a S912 MECOOL OR MINIX U9 with L1 DRM and HDCP 1.x, dual boot into a unofficial Android TV for 1080p Netflix. That is the only option.

Display More

Quote from DaVu

moved thread on demand of the user

Thanks

Further info.

The system also refuses to reboot after issues arise with mounting.

Output of dmesg is as follows:

178.127551] usb 1-1.3: USB disconnect, device number 5
178.127588] usb 1-1.3.1: USB disconnect, device number 6
178.142273] usb 1-1.3.2: USB disconnect, device number 7
178.142312] usb 1-1.3.2.1: USB disconnect, device number 10
178.161232] usb 1-1.3.2.4: USB disconnect, device number 11
178.167096] sd 4:0:0:0: [sde] Synchronizing SCSI cache
178.177254] sd 4:0:0:0: [sde] Synchronize Cache(10) failed: Result: hostbyte=0x01 driverbyte=0x00
178.181471] usb 1-1.3.3: USB disconnect, device number 8
178.192690] blk_update_request: I/O error, dev sdb, sector 0
178.205814] sd 1:0:0:0: [sdb] Synchronizing SCSI cache
178.208175] sd 1:0:0:0: [sdb] Synchronize Cache(10) failed: Result: hostbyte=0x01 driverbyte=0x00
178.226434] usb 1-1.3.4: USB disconnect, device number 9
179.565538] Buffer I/O error on dev sdc1, logical block 121667584, lost sync page write
179.565572] JBD2: Error -5 detected when updating journal superblock for sdc1-8.
179.565581] Aborting journal on device sdc1-8.
179.565599] Buffer I/O error on dev sdc1, logical block 121667584, lost sync page write
179.565613] JBD2: Error -5 detected when updating journal superblock for sdc1-8.
179.611232] EXT4-fs error (device sdb1): ext4_find_entry:1451: inode #2: comm udevil: reading directory lblock 0
179.653484] Buffer I/O error on dev sdb1, logical block 91258880, lost sync page write
179.653517] JBD2: Error -5 detected when updating journal superblock for sdb1-8.
179.653526] Aborting journal on device sdb1-8.
179.653542] Buffer I/O error on dev sdb1, logical block 91258880, lost sync page write
179.653554] JBD2: Error -5 detected when updating journal superblock for sdb1-8.
198.976165] systemd-journald[191]: Failed to send WATCHDOG=1 notification message: Connection refused
[hr]
Further information:

Cannot access via Samba and also cannot reboot, error messages in dmesg as follows:

2621.208159] systemd-journald[175]: Failed to send WATCHDOG=1 notification message: Connection refused
2715.141342] systemd-journald[175]: Failed to send WATCHDOG=1 notification message: Transport endpoint is not connected
2801.208673] systemd-journald[175]: Failed to send WATCHDOG=1 notification message: Transport endpoint is not connected
2921.208614] systemd-journald[175]: Failed to send WATCHDOG=1 notification message: Transport endpoint is not connected
2981.207907] systemd-journald[175]: Failed to send WATCHDOG=1 notification message: Transport endpoint is not connected
3092.179589] systemd-journald[175]: Failed to send WATCHDOG=1 notification message: Transport endpoint is not connected
3161.217009] systemd-journald[175]: Failed to send WATCHDOG=1 notification message: Transport endpoint is not connected
3281.217365] systemd-journald[175]: Failed to send WATCHDOG=1 notification message: Transport endpoint is not connected
3341.216399] systemd-journald[175]: Failed to send WATCHDOG=1 notification message: Transport endpoint is not connected
3447.818410] systemd-journald[175]: Failed to send WATCHDOG=1 notification message: Transport endpoint is not connected
3521.216981] systemd-journald[175]: Failed to send WATCHDOG=1 notification message: Transport endpoint is not connected
3641.217050] systemd-journald[175]: Failed to send WATCHDOG=1 notification message: Transport endpoint is not connected
3701.216483] systemd-journald[175]: Failed to send WATCHDOG=1 notification message: Transport endpoint is not connected
3821.216364] systemd-journald[175]: Failed to send WATCHDOG=1 notification message: Transport endpoint is not connected
3881.216929] systemd-journald[175]: Failed to send WATCHDOG=1 notification message: Transport endpoint is not connected
4001.216939] systemd-journald[175]: Failed to send WATCHDOG=1 notification message: Transport endpoint is not connected
4061.216325] systemd-journald[175]: Failed to send WATCHDOG=1 notification message: Transport endpoint is not connected
4181.225176] systemd-journald[175]: Failed to send WATCHDOG=1 notification message: Transport endpoint is not connected
4241.225740] systemd-journald[175]: Failed to send WATCHDOG=1 notification message: Transport endpoint is not connected
4338.606149] systemd-journald[175]: Failed to send WATCHDOG=1 notification message: Transport endpoint is not connected
4421.225167] systemd-journald[175]: Failed to send WATCHDOG=1 notification message: Transport endpoint is not connected
4541.225088] systemd-journald[175]: Failed to send WATCHDOG=1 notification message: Transport endpoint is not connected
4601.225687] systemd-journald[175]: Failed to send WATCHDOG=1 notification message: Transport endpoint is not connected
4692.034970] systemd-journald[175]: Failed to send WATCHDOG=1 notification message: Transport endpoint is not connected
4781.225044] systemd-journald[175]: Failed to send WATCHDOG=1 notification message: Transport endpoint is not connected
4882.550687] systemd-journald[175]: Failed to send WATCHDOG=1 notification message: Transport endpoint is not connected
4956.378693] systemd-journald[175]: Failed to send WATCHDOG=1 notification message: Transport endpoint is not connected
[hr]
Debug.log output

TWVY

Hey,

Did some further digging and have definitely ruled the problem down to udev.

Checking dmesg its clear that the system is picking up the hub and attached hard drives no problem.

Currently if I run udevadm trigger --action=add the hard drives are succesfully mounted and seem to work.

My guess is a bug of some sort has crept into the automount script either on the LE side or on the udev side. A complicating factor may be that I'm also running a slower/older RPi with a comparatively high number of HDD's (5).

Anyway, let me know if I can provide logs or any further information to help track this down.

Neil
[hr]
On Second thought I suspect this should be moved to Bug Reports?

Also please see output of lsusb

Bus 001 Device 009: ID 1058:10a2 Western Digital Technologies, Inc. Elements SE Portable (WDBPCK)
Bus 001 Device 008: ID 0bc2:3000 Seagate RSS LLC FreeAgent Desktop
Bus 001 Device 011: ID 0bc2:2300 Seagate RSS LLC Expansion Portable
Bus 001 Device 010: ID 1058:1042 Western Digital Technologies, Inc. Elements SE Portable (WDBPCK)
Bus 001 Device 007: ID 05e3:0608 Genesys Logic, Inc. Hub
Bus 001 Device 006: ID 1058:1042 Western Digital Technologies, Inc. Elements SE Portable (WDBPCK)
Bus 001 Device 005: ID 05e3:0608 Genesys Logic, Inc. Hub
Bus 001 Device 004: ID 148f:5370 Ralink Technology, Corp. RT5370 Wireless Adapter
Bus 001 Device 003: ID 0424:ec00 Standard Microsystems Corp. SMSC9512/9514 Fast Ethernet Adapter
Bus 001 Device 002: ID 0424:9512 Standard Microsystems Corp. SMC9512/9514 USB Hub
Bus 001 Device 001: ID 1d6b:0002 Linux Foundation 2.0 root hub

Hi Guys,

Firstly let me say really appreciate all your work. Been a longterm OE and now LE user. Recently migrated over to 7.0.2 from OE 6.0.3. It was a very smooth process all in all.
I was saddened to see the split in the community but hopefully everyone is happier this way!

My issue is related to mounting of my USB hard drives which previously worked without problem under various flavours of OE.

If I turn off the USB hub and boot the Rpi then turn on the USB hub everything mounts fine. However after a various period of time hard drives 'disappear' so that when I select an item to watch in the library to flags "Item no longer found, do you want to remove" warning. I haven't been able to catch the random dismounts on dmesg yet so cannot provide logs for those. My guess is it is due to power fluctuations which cause the hard drives to un-mount.

However I do not believe this is the central issue which I believe is due to a failure to mount or 're-mount' the drive. If I turn of the USB hub once the pi is booted everything dismounts. However if I turn it back on, no hard drives are mounted. I also then find that when I attempt to reboot the pi it times out and the only way I can reboot it is to manually pull the power. This isn't something I've come across before.

Ive attached the dmesg below. At 170 you can see where I turned off the USB hub and at 200 you can see where I turned it back on. Interestingly there is no mention of EXT-FS4 saying it has mounted the drives (as can be found earlier during boot up).

Anyway your input is greatly appreciated.

Regards,
B

dmesg: LibreELEC:~ # dmesg [ 0.000000] Booting Linux on physical CPU 0x0 [ 0.00 - Pastebin.com