tuxd3v_ | c0rnelius, does you use rng-tools using hardware {t,P}RNG | 15:49 |
---|---|---|
tuxd3v_ | I can't on h3 | 15:49 |
tuxd3v_ | it doesn't have TRNG | 15:50 |
tuxd3v_ | but have PRNS | 15:50 |
tuxd3v_ | sorry | 15:50 |
tuxd3v_ | PRNG | 15:50 |
tuxd3v_ | but sun8i-ss has PRNg support :/ | 16:12 |
tuxd3v_ | so it shouold workd | 16:13 |
tuxd3v_ | should* | 16:13 |
tuxd3v_ | well I go with /dev/urandom :/ | 16:59 |
tuxd3v_ | sun8i-ss should provide at least a PRNG | 17:00 |
tuxd3v_ | c0rnelius, h3 is pin to pin compatible with h5, onlr the mmc card ip is diferentr :) | 17:02 |
c0rnelius | Yeah I know there isn't much diff. Besides one being 32 and the other 64bit. | 18:02 |
c0rnelius | Currently messing around with dual booting using extlinux on the pi4. | 18:08 |
tuxd3v_ | nice :) | 18:09 |
c0rnelius | So far I got it working but it does take manual setup. | 18:09 |
c0rnelius | Only problem is because of the nature of u-boot I can't seem to load a 32bit kernel on a 64bit system. | 18:11 |
c0rnelius | But could have a Devuan desktop on USB and Debian on the SD... I guess if I built it for armhf could put LibreELEC on USB and whatever on SD. | 18:12 |
c0rnelius | Yesterday I bought Odroid N2+ with everything but the kitchen sink included off one of the Armbian boys. So that should be fun to play with next week when it arrives. | 18:16 |
tuxd3v_ | ho you got new board, nice to hear that :) | 18:32 |
tuxd3v_ | I am waiting rk3588 boards.. | 18:32 |
tuxd3v_ | the chip never gets out :( | 18:32 |
c0rnelius | Yeah. I've been waiting to order the N2+ forever and he happened to be trying to sell his. So it worked out. Comes with PSU, 32GB eMMC, case, fan, heatsink, wifi dongle... Pretty much everything I thinks? | 18:35 |
c0rnelius | Only a $130 | 18:35 |
tuxd3v_ | wow, you brought it from a armbian guy? | 18:36 |
c0rnelius | Mhmm | 18:36 |
c0rnelius | One of the guys who is doing the desktop stuff for them. | 18:37 |
c0rnelius | I guess he moved on to messing with the NanoPC-T4 now and doesn't want it anymore. | 18:37 |
c0rnelius | Plus they want me to fix mainline u-boot for them on the N2/+. :) | 18:38 |
c0rnelius | Which I'm pretty sure I already know why its not working for them, but I need the board first before I can do the PR. | 18:38 |
c0rnelius | Amlogic u-boot creates two binaries, one specific to MMC and one for SD. It looks to me that they are flashing the wrong binary and doing so the wrong way. chewitt from LibreELEC says the second binary "u-boot.bin.sd.bin" is suppose to work for both MMC/SD but I need to test that first. | 18:42 |
tuxd3v_ | nanopi NEO plus2, is another device were I notice the problem, but there are a lot.. | 18:47 |
c0rnelius | Yeah that's the one I have. | 18:48 |
tuxd3v_ | in bananapi m2 zero, it sometimes work sometimes not...it will be my next check, using CONFIG_SERIAL_8250_16550A_VARIANTS in a kernel for it, so see if it solves, but I am sure it will.. | 18:48 |
c0rnelius | Good board. | 18:49 |
tuxd3v_ | the neo plus2 is a nice board indeed | 18:49 |
tuxd3v_ | it has lots of options on it.. | 18:49 |
c0rnelius | The onboard wifi/bluetooth chip sux but besides that it's pretty solid. | 18:49 |
tuxd3v_ | I got amazed when I looked at the schematics the other day :) | 18:49 |
tuxd3v_ | You need to work better the dts, and I am 100% sure it will work out-of-the-box, always after that.. | 18:50 |
tuxd3v_ | CONFIG_SERIAL_8250_16550A_VARIANTS, ends being a work around, as email was sent to a lot of locations, I also received it, by apritzel, asking why that commit was made.. | 18:51 |
tuxd3v_ | he is asking why, and asking for more insight of the problem.. | 18:52 |
c0rnelius | Yeah. I mean I the wifi works, its just 2.4GHz which in my house isn't the best. Bluetooth is a whole other thing. | 18:52 |
tuxd3v_ | because to gain 20ms in a virtual machine boot up, it ended cracking the uart devices, and so bluetooth is off in vitually any device :( | 18:53 |
c0rnelius | Only seen work like twice? | 18:53 |
tuxd3v_ | yes because that patch cut the timings needed for the uarts | 18:53 |
tuxd3v_ | the uarts ends timeout | 18:53 |
tuxd3v_ | WIFI is another history,because its sdio( the same interface as emmc,sdcard ) | 18:54 |
c0rnelius | Since I currently use it to run weechat I don't even bother with it anymore. I use a rtl88XXAU dongle on it, as that can get 5GHz. | 18:55 |
c0rnelius | But I know the onboard wifi works fine. | 18:56 |
tuxd3v_ | yeah WIFI 5Ghx, I don't know if the kernel already supports both modes? | 18:56 |
c0rnelius | That board only supports 2.4 | 18:56 |
tuxd3v_ | does you need a patch for it to work on 2.4Ghz and 5Ghz? | 18:56 |
tuxd3v_ | ok that was my idea | 18:56 |
c0rnelius | the NEO3/4 looks nice too | 18:57 |
c0rnelius | Those are rockchip though | 18:57 |
c0rnelius | rk3328/3399 | 18:57 |
c0rnelius | They have usb3 :) | 18:57 |
c0rnelius | I bet they get hot as all hell though being in that small form factor. | 18:59 |
tuxd3v_ | I have nanipi R2s its a rk3328 and yes it get hot as hell, even without load.. | 19:03 |
tuxd3v_ | I needed to buy the alumunium case, and even then its around 50-60 degress in the summer | 19:04 |
c0rnelius | Those are nuts. I have a Renegade and Rock64 which uses the same SoC and without a heatsink and fan they are kinda useless. | 19:05 |
tuxd3v_ | it has 1 downside, it has a rtl8153b and it is a usb3-to-ethernet controller.. | 19:05 |
tuxd3v_ | becaue of that it doesn't have tx buffers, and needs to be done on the kernel side of usb3 | 19:06 |
tuxd3v_ | and it resets a lot of times, if pushed to the extreme | 19:06 |
tuxd3v_ | I tested it around 1 month | 19:06 |
c0rnelius | Probs cuz of the heat | 19:06 |
tuxd3v_ | heavy testing it | 19:06 |
c0rnelius | USB will reset if it gets to hot. | 19:07 |
tuxd3v_ | I believe its also the drivers, but I can't rule out the heat indeed.. | 19:07 |
c0rnelius | The board is trying to save its self :) | 19:07 |
tuxd3v_ | the driver has a very bad decision ... when it resets it atributes a random maccadress do the device.. I had to craft there a static maccadress.. | 19:07 |
tuxd3v_ | but the resets I couldn't solve.. not a new driver is out | 19:08 |
tuxd3v_ | 2.0.15 I believe | 19:08 |
c0rnelius | Put a fan on it and place it under heavy load and see what happens? | 19:08 |
tuxd3v_ | but I read that some users had problems with it, reseting so maybe the problem persists :/ | 19:09 |
tuxd3v_ | in that regard I think that Alwinner devices are a lot better, a lot less heat | 19:09 |
tuxd3v_ | but they lack usb3 :/ | 19:09 |
c0rnelius | Sorta yeah... They are good as they are easy to work with, but they also overheat really bad. | 19:10 |
c0rnelius | Especially the H5 | 19:10 |
tuxd3v_ | ho I didn't knew of that.. I only have h2+,h3,H6 | 19:11 |
tuxd3v_ | the H6 should be the one in nanopi R2S | 19:11 |
c0rnelius | My Tritium which is an H5 is just running Apache and has a huge heatsink on it and is currently sitting at - CPU Temp: 55°C | 19:11 |
tuxd3v_ | its formidable, but the drivers are not there yet, specially usb3 driver.. | 19:12 |
c0rnelius | If I tried to compile something on it without a fan it would stall. | 19:12 |
tuxd3v_ | I believe, that when you push the processing power they heat, even the alwinner A20 do, I can't touch it when compiling mesa lima driver for userspace | 19:13 |
tuxd3v_ | it has small heatsing.. | 19:13 |
tuxd3v_ | but without pushing them, they should mantain low temperatures.. | 19:13 |
c0rnelius | Generally, yes. | 19:14 |
tuxd3v_ | nanopi r2s is mind blowing, as it heats even without load to crazy temps.. | 19:14 |
c0rnelius | Also depends on location I suppose. | 19:14 |
tuxd3v_ | yes indeed, that also matters :) | 19:14 |
tuxd3v_ | I believe R2s heats a lot also because it has 2 internet controlers very close, the board is very small, and is closed to the heat can't excape.. | 19:15 |
c0rnelius | Well for sure. The form factor isn't ideal at all for that SoC. | 19:16 |
tuxd3v_ | but those guys are crazy.. they are charging now 48$ only for the basic board.. | 19:16 |
tuxd3v_ | https://www.friendlyarm.com/index.php?route=product/product&product_id=282 | 19:16 |
tuxd3v_ | they are rising the prices... I don't know if it has todo with pandemic.. | 19:17 |
c0rnelius | They seemed to have raised the prices on everything. | 19:17 |
tuxd3v_ | indeed | 19:17 |
c0rnelius | I think its the chip shortage. | 19:17 |
c0rnelius | this use to be way cheaper - https://www.friendlyarm.com/index.php?route=product/product&path=69&product_id=268 | 19:18 |
tuxd3v_ | yeah I read about it | 19:18 |
c0rnelius | and now you can't even buy it anymore :) | 19:19 |
tuxd3v_ | yes like my neo Air at 999$ | 19:20 |
tuxd3v_ | its mind blowing | 19:20 |
tuxd3v_ | Its a chip shortage indeed, but its related with the pandemic, that created that situation.. | 19:20 |
c0rnelius | Something like that | 19:21 |
tuxd3v_ | that are a lot of boards in standby to be produced because they can't acess the chips.. | 19:21 |
tuxd3v_ | I am waiting for the rk3588 for a year or so, and nothing.. | 19:21 |
ShorTie | it's China being ass's | 21:43 |
ShorTie | foundation was smart and moved all of that stuff to the UK | 21:43 |
c0rnelius | and the US not being smart enough to keep some of those jobs here. | 22:29 |
tuxd3v_ | ShorTie, I believe its a big problem inside of China because of the virus.. | 23:06 |
tuxd3v_ | its not being talked much, but we notice that there capacity for mass production is severely indered.. | 23:07 |
tuxd3v_ | I agree that countries should protect the jobs of their people , that is true :) | 23:07 |
tuxd3v_ | c0rnelius, your h5 has a mali 450 MP4, have you ever tried to compile mesa lima driver for it and then run: | 23:42 |
tuxd3v_ | 'vblank_mode=0 glxgears' | 23:42 |
tuxd3v_ | ? | 23:42 |
tuxd3v_ | I have Mali 400 MP2, the best results I got is ~330 fps | 23:42 |
Generated by irclog2html.py 2.17.0 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!