fossxplorer Posted September 5, 2017 Posted September 5, 2017 17 minutes ago, balbes150 said: Update TEST images kernel 4.13.0-next_20170905. In these images there is USB support for S912. Awesome! Will try this tonight already! Another question: every time i upgrade, i loose the full setup i have on my SD card? Now that i've installed Docker, LXC etc i'd like to avoid it.
balbes150 Posted September 5, 2017 Posted September 5, 2017 11 minutes ago, fossxplorer said: Another question: every time i upgrade, i loose the full setup i have on my SD card? Now that i've installed Docker, LXC etc i'd like to avoid it. To enable USB you can only replace dtb file. 1
debianxfce Posted September 5, 2017 Posted September 5, 2017 1 hour ago, Shimon said: Nah, why waste so much space? The most efficient way to do it looks like that: $ zcat Whatever_image_img.gz | sudo dd of=/dev/sdX bs=4M && sync For xz compressed images use xzcat obviously. What space, cp is simpler command than dd. This is more more efficient than yours: xzcat Debian_testing_4.13.0_xfce.img.xz > /dev/sdX
fossxplorer Posted September 5, 2017 Posted September 5, 2017 I can't get your updated image @balbes150 since Yandex Disk seems to have issues
fossxplorer Posted September 5, 2017 Posted September 5, 2017 If possible, please upload the latest image here: https://my.owndrive.com/index.php/s/vNynzw9pk5fIYhT
debianxfce Posted September 7, 2017 Posted September 7, 2017 On 05/09/2017 at 9:47 PM, fossxplorer said: I can't get your updated image @balbes150 since Yandex Disk seems to have issues I can acces this: https://yadi.sk/d/kZHc0uQE34z5jR Reboot your modem/router. I hope you have a reliable internet connection too, you can test with Linux distributions that offer checksum files.
debianxfce Posted September 7, 2017 Posted September 7, 2017 1 hour ago, balbes150 said: You mean kernel 4.13...There is no 3.14 kernel for today in your yandex folder. Jerome brunet audio patches for S912 does have only spdif support. Have someone tested spdif audio patches?
balbes150 Posted September 7, 2017 Posted September 7, 2017 Thank you for your attention. Yes, I mean 4.13.0. In images enabled sound via HDMI to s912 (and s905x). Work sound tested on a Tronsmart Vega S96 (s912). Sound works through HDMI on test files from the image. Once again - you need to have 1920x1080 resolution (you can go into desktop settings in the tools menu and select directly in the working system needs a permit). 1
debianxfce Posted September 7, 2017 Posted September 7, 2017 I added patches from here https://github.com/150balbes/meta-meson/tree/master/recipes-kernel/linux/linux-yocto-meson64-4.13. The command aplay -l lists only spdif device, no hdmi audio. What is required to enable hdmi audio for a s912 device?
balbes150 Posted September 7, 2017 Posted September 7, 2017 Need a new kernel + dtb. You installed a new kernel and dtb ? After installation you need to copy the new version to "dtb.img".
debianxfce Posted September 7, 2017 Posted September 7, 2017 10 minutes ago, balbes150 said: Need a new kernel + dtb. You installed a new kernel and dtb ? After installation you need to copy the new version to "dtb.img". I added patches to 4.13-rc1 kernel where was usb and spdif patches already. Added missing clocks and hdmi audio patches. Copied new a dtb.img file of course. 0005-clk-meson-gxbb-aoclk-Add-CEC-32k-clock.patch was missing, implementing now...
balbes150 Posted September 7, 2017 Posted September 7, 2017 For proper operation of the patches, they should be used (in exact order as listed in source directory) and apply the kernel sources. See all options in my last git Armbian Assembly. By the way, You can get your own version of Armbian images (to edit the list of packages, change settings, etc.).
debianxfce Posted September 7, 2017 Posted September 7, 2017 I tested yours Debian Xfce image, aplay -l list only spdif device and pulseaudio control shows only analog output. Install wicd and xfce4-terminal instead networkmanager and xterminal. Use the xfce whisker menu. Your distro image is about 1.5 GB bigger than mine, so strip some packages. Your 86 patches are confusing, it is easier to copy whole source files from authors. Using 200q dtb.
fossxplorer Posted September 7, 2017 Posted September 7, 2017 Is anyone running Docker aarch64/arm64v8 images on this? I'm able to run armhf/arm32v7 images, but all the arm64v8 images crash. According to 'uname' it's 64-bits kernel so i wonder what the issue is with the Docker.
balbes150 Posted September 8, 2017 Posted September 8, 2017 11 hours ago, debianxfce said: Install wicd and xfce4-terminal instead networkmanager and xterminal. Use the xfce whisker menu. Why ? 11 hours ago, debianxfce said: Your distro image is about 1.5 GB bigger than mine, so strip some packages. If You are not satisfied with something in my images, gather your option. All source code freely available. 11 hours ago, debianxfce said: Your 86 patches are confusing, it is easier to copy whole source files from authors. I do not give You to use other patches ?
debianxfce Posted September 9, 2017 Posted September 9, 2017 I2S is connected to the HDMI audio in the S9xx SOC. Other S9xx devices do have the hdmi audio support working but not meson-GXM devices officially: https://github.com/superna9999/linux/tree/linux-4.13-le-amlogic-gx/arch/arm64/boot/dts/amlogic
fossxplorer Posted September 10, 2017 Posted September 10, 2017 On 7/19/2017 at 1:46 AM, gnthibault said: ycomp: when your are logged as root, just run the install script in /home/root as suggested by debianxfce. My question is related to vnc, I always get grey window when connecting to the tightvnc server. I use the latest 3.14 Ubuntu Mate (16.04) desktop image, but I am not connecting my box to a display. Can someone give me an example of xstartup file that works ? What bothers me for the moment is that log does not show error, bu still nothing but a grey screen ... Edit, here is the solution (yaaayyy!): Setup password with vncpasswd /etc/vnc/vncpasswd Then add the following lines to /etc/lightdm/lightdm.conf : [VNCServer] enabled=true command=/usr/bin/Xvnc -rfbauth /etc/vnc/vncpasswd port=5900 width=1280 height=1024 depth=24 I'm also testing tightvncserver and there are issues with LXDE. Mate-session works, but it can't automatically resize screen size. Also, when i used the box as a vncviewer, the resizing doesn't work, i get a black area at bottom for instance when i choose full screen. I'd like to use one of the boxes as a vnc/spice client and another as a server
balbes150 Posted September 11, 2017 Posted September 11, 2017 11 hours ago, fossxplorer said: I'm also testing tightvncserver and there are issues with LXDE. Mate-session works, but it can't automatically resize screen size. Also, when i used the box as a vncviewer, the resizing doesn't work, i get a black area at bottom for instance when i choose full screen. What kernel is used ?
balbes150 Posted September 12, 2017 Posted September 12, 2017 Kernel 4.13 is under development, so many features have changed and the old soft yet will not work correctly.
Jens Bauer Posted September 12, 2017 Posted September 12, 2017 On 4/5/2017 at 11:50 AM, kingul said: @gnthibault I have 2GB of ram with 32GB rom. (Yoka TV KB2) First time I tried with nand-sata-install, but it was a bad move, because bootloader was rewritten, I had to restore official image of my device. After a little search I found /root/install.sh script and it worked perfectly. I'm interested in an update (as I might want to purchase a Yoka KB2 with 3GB DDR4 RAM): Did you solve the static IP problem completely, or do you need to use a dynamic IP still ? How do the latest images work with the Yoka KB2 ?
Jens Bauer Posted September 12, 2017 Posted September 12, 2017 On 9/4/2017 at 10:38 AM, fossxplorer said: EDIT: have to admit i feel like a kid now after getting this to work! Love the 3G RAM on this box+ 1GbE. It's using 6-7W when i run stress-ng --cpu 8 --io 2 --vm 1 --vm-bytes 1G --timeout 300s. Really nice! Quite nice indeed - and a very low power usage. -How fast do the cores run ... 8 x 1 GHz or 4 x 2 GHz + 4 x 1 GHz ?
fossxplorer Posted September 12, 2017 Posted September 12, 2017 Yep When all cores enabled, it's 4 x 1.512Ghz + 4 x 1Ggz according to lshw -c cpu. root@amlogic:~# cat /sys/devices/system/cpu/cpu0/cpufreq/scaling_available_frequencies 100000 250000 500000 667000 1000000 1200000 1512000 root@amlogic:~# I've only seen big cores at 1512Mhz even when i disable the small 4 cores. How can we get 2Ghz?
Jens Bauer Posted September 13, 2017 Posted September 13, 2017 17 hours ago, fossxplorer said: Yep When all cores enabled, it's 4 x 1.512Ghz + 4 x 1Ggz according to lshw -c cpu. This information is indeed very useful! 4 x 1.5 + 4 x 1.0 (10) ... against 4 x 2 GHz (8) . Remembering there's a small loss of speed due to things that requires single-threading/tasking ... That's close to 4x2 against 8 x 1; but if compiling/building software, the speed gain will probably be more noticable. -But I do believe that there *is* a performance gain; or perhaps a 'responsiveness gain'. (Eg. the PC I'm using at the moment sometimes stalls while moving the windows, typing or when moving the mouse. After several seconds (up to a minute), the window wil move or the mouse arrow will move). Having multiple cores can in those cases make the computer/device more responsive. 17 hours ago, fossxplorer said: root@amlogic:~# cat /sys/devices/system/cpu/cpu0/cpufreq/scaling_available_frequencies 100000 250000 500000 667000 1000000 1200000 1512000 root@amlogic:~# I've only seen big cores at 1512Mhz even when i disable the small 4 cores. How can we get 2Ghz? That I don't know, but I remember reading a while back, that it was only possible to run at 4x2 GHz or 8 x 1 GHz. This means that someone must have done something to improve it. I've been battling with myself about whether I should purchase a bunch of MiQi devices for a build-farm or a bunch of Yoka-boxes (I will be purchasing a MacchiatoBIN as build-master either way).
tkaiser Posted September 14, 2017 Posted September 14, 2017 On 13.9.2017 at 1:05 AM, fossxplorer said: I've only seen big cores at 1512Mhz even when i disable the small 4 cores. How can we get 2Ghz? Why do you trust even in these 1512MHz? Care to remember what Willy Tarreau discovered when trying to use Amlogic SoCs for his build farms: http://wiki.ant-computing.com/Choosing_a_processor_for_a_build_farm#Very_interesting_discoveries_about_cheating_vendors_-_2016.2F07.2F09 Your consumption numbers also look suspicious but the tool you used is pretty lightweight. When I tested with a NanoPi M3 last year (octa-core A53 with up to 1.4 GHz) I would've to search for another PSU to power the board when allowing the 1400 MHz since consumption was too high for my 5V/2.5A 5V/2A PSU (IIRC with cpuburn-a53 1300 MHz was the maximum). You could try this NEON optimized Linpack here: https://github.com/ehoutsma/StabilityTester to get a clue. Maybe stabilitytester.sh already runs out of the box (trying to walk through available cpufreq/dvfs OPP) but at least you can check for dependencies and how to call xhpl64. Testing individually by sending CPU cores 0-3 or 4-7 offline is an option too Edit: NanoPi M3 tests back then (it was even worse and with the Linpack not even running at 1GHz on all 8 cores was possible): https://forum.armbian.com/index.php?/topic/1285-nanopi-m3-cheap-8-core-35/&do=findComment&comment=14697
fossxplorer Posted September 14, 2017 Posted September 14, 2017 @Jens Bauer Sure, i'd have loved to put a bigger heatsink and a small fan, if that helps to achieve higher speeds. But now as @tkaiser mentions it, i can recall the Amlogic cheating revelations i read at CNX-soft thoug! Here is what i got from StabilityTester: Done testing stability by using both ondemand and performance scaling governors: Frequency: 667000 Voltage: 3300000 Success: 1 Result: 0.0048034 Frequency: 1000000 Voltage: 3300000 Success: 1 Result: 0.0048034 Frequency: 1200000 Voltage: 3300000 Success: 1 Result: 0.0048034 Although the thermal readings doesn't work on this box: Testing frequency 1200000 Soc temp:cat: /sys/class/thermal/thermal_zone0/temp: No such file or directory Soc temp: CPU Freq: 1200000 CPU Core: 3300000 cat: /sys/class/thermal/thermal_zone0/temp: No such file or directory Soc temp: CPU Freq: 1200000 CPU Core: 3300000 cat: /sys/class/thermal/thermal_zone0/temp: No such file or directory Anyway, i guess i'm also subject to their cheating since i can get max 1.2Ghz even with all small cores disabled.
fossxplorer Posted September 14, 2017 Posted September 14, 2017 After a reboot of the box: root@amlogic:~# for i in {0..7};do cat /sys/devices/system/cpu/cpu$i/cpufreq/scaling_governor;done cat: /sys/devices/system/cpu/cpu0/cpufreq/scaling_governor: No such file or directory cat: /sys/devices/system/cpu/cpu1/cpufreq/scaling_governor: No such file or directory cat: /sys/devices/system/cpu/cpu2/cpufreq/scaling_governor: No such file or directory cat: /sys/devices/system/cpu/cpu3/cpufreq/scaling_governor: No such file or directory cat: /sys/devices/system/cpu/cpu4/cpufreq/scaling_governor: No such file or directory cat: /sys/devices/system/cpu/cpu5/cpufreq/scaling_governor: No such file or directory cat: /sys/devices/system/cpu/cpu6/cpufreq/scaling_governor: No such file or directory cat: /sys/devices/system/cpu/cpu7/cpufreq/scaling_governor: No such file or directory I can't see cpufreq module to load, hmm. EDIT: root@amlogic:~# ll /lib/modules/4.13.0-rc7-next-20170901-amlogics905x/kernel/drivers/cpufreq/cppc_cpufreq.ko -rw-r--r-- 1 root root 7840 Sep 1 17:26 /lib/modules/4.13.0-rc7-next-20170901-amlogics905x/kernel/drivers/cpufreq/cppc_cpufreq.ko root@amlogic:~# modprobe cppc_cpufreq modprobe: ERROR: could not insert 'cppc_cpufreq': No such device EDIT2: Seems some of the files are messed up by apt-get update or install of other files; root@amlogic:~# ll -th /boot/ total 47M-rwxr-xr-x 1 root root 5.0M Sep 12 21:33 initrd.img-4.13.0-rc7-next-20170901-amlogics905x* -rwxr-xr-x 1 root root 5.0M Sep 12 21:33 uInitrd* -rwxr-xr-x 1 root root 25K Sep 7 10:10 dtb.img* drwxr-xr-x 23 root root 4.0K Sep 6 13:06 ../ -rwxr-xr-x 1 root root 1.6K Sep 1 17:30 armbian_first_run.txt* -rwxr-xr-x 1 root root 718 Sep 1 17:30 aml_autoscript* -rwxr-xr-x 1 root root 651 Sep 1 17:30 aml_autoscript.zip* -rwxr-xr-x 1 root root 1.6K Sep 1 17:30 s905_autoscript* -rwxr-xr-x 1 root root 1.5K Sep 1 17:30 s905_autoscript.cmd* -rwxr-xr-x 1 root root 38K Sep 1 17:30 boot.bmp* -rwxr-xr-x 1 root root 4.8K Sep 1 17:30 boot-desktop.png* -rwxr-xr-x 1 root root 1.1K Sep 1 17:30 hdmi.sh* drwxr-xr-x 2 root root 4.0K Sep 1 17:29 dtb/ drwxr-xr-x 2 root root 4.0K Sep 1 17:29 dtb-4.13.0-rc7-next-20170901-amlogics905x/ -rwxr-xr-x 1 root root 23K Sep 1 17:26 old_dtb.img* -rwxr-xr-x 1 root root 143K Sep 1 17:26 config-4.13.0-rc7-next-20170901-amlogics905x* -rwxr-xr-x 1 root root 3.5M Sep 1 17:26 System.map-4.13.0-rc7-next-20170901-amlogics905x* -rwxr-xr-x 1 root root 17M Sep 1 17:26 vmlinuz-4.13.0-rc7-next-20170901-amlogics905x* -rwxr-xr-x 1 root root 17M Sep 1 17:26 zImage* drwxr-xr-x 4 root root 16K Jan 1 1970 ./ EDIT3: i replaced the whole /boot with latest release + copied back dtb.img. Problem solved. I think one of the package dependency messed up with uInitrd and initrd.img-4.13.0-rc7-next-20170901-amlogics905x.
Jens Bauer Posted September 15, 2017 Posted September 15, 2017 On 9/14/2017 at 3:10 PM, fossxplorer said: Anyway, i guess i'm also subject to their cheating since i can get max 1.2Ghz even with all small cores disabled. That is so annoying. -But please make sure that it actually is the CPU that needs to be blamed and not the RAM-speed. I'd like to see a comparison of a TV-box having DDR4 RAM installed against one that has DDR3 RAM installed - I wonder if it makes any difference at all. (Since this is a S912 thread I've restricted myself from comparing to other CPU vendors, but it seems that Amlogic is not the most interesting vendor anymore).
tkaiser Posted September 16, 2017 Posted September 16, 2017 9 hours ago, Jens Bauer said: I'd like to see a comparison of a TV-box having DDR4 RAM installed against one that has DDR3 RAM installed - I wonder if it makes any difference at all. Well, I'm surprised why people expect $anything from TV boxes. These things aren't for Linux enthusiasts but for clueless people that only buy numbers (8 CPU cores vs. 4 cores --> 8! 2 GB DRAM vs. 3 GB DRAM --> 3! $30 vs. $28 --> 28!). They have to look performant by specs while they have to be as cheap as possible for their target audience to buy them. Why does someone think both can work at the same time? This means manufacturers don't have to take care about performance but only about cluelessness (use 8 cores even if it's totally useless since the competitors also use 8 cores now so you are not able to sell faster boxes with less CPU cores anymore in the markets where they make money, put recycled DRAM on the PCB and combine it with an u-boot that dynamically tunes DRAM clockspeed down until the box doesn't immediately crash so you as manufaturer can afford putting 3 GB crappy/slow DRAM on board since customers would never buy your box with 2 GB faster DRAM) And then this kind of 'cheating' also happens everywhere. Don't know exactly how it works with Amlogic (since I consider all their SoCs not worth a look, might change if a dev board relying on Amlogic A113D appears) but I know how it's done with Allwinner and there it's easy: the 'board support packages' (BSP) TV box vendors get contains a kernel with a list of cpufreq operation points (that's the entire 'up to x.x GHz' thing) that does not need to have any relationship with reality (Allwinner sets 1536 MHz here for A64 as an example). Then there are other settings defining how voltage and clockspeed relates (DVFS) and then there's another settings defining thermal behaviour. The result is that the TV box will be marketed as 'up to 1.5GHz', the real settings already prevent everything above 1152 MHz and due to shitty thermal design of most boxes the SoC won't even see those 1.1GHz when running under constant load since already throttled down to less than 1 GHz. And their target audience is happy since in Android's CPU-Z and stupid benchmarks like Geekbench '4 cores @ 1536 MHz' is listed and that's all they care about, IMO interesting read: https://www.cnx-software.com/2017/09/12/factory-price-of-some-tv-boxes-and-accessories/#comments 1
Recommended Posts