| - mtm (QUIT: Ping timeout: 260 seconds) (~textual@c-71-228-84-213.hsd1.fl.comcast.net) | 02:03 | |
| + jacobk (~quassel@47-186-105-237.dlls.tx.frontiernet.net) | 02:03 | |
| + mtm (~textual@c-71-228-84-213.hsd1.fl.comcast.net) | 02:05 | |
| + staticbunny (~staticbun@76-223-253-78.lightspeed.frokca.sbcglobal.net) | 02:15 | |
| - paperManu (QUIT: Ping timeout: 272 seconds) (~paperManu@198.16.214.40) | 03:26 | |
| - op_4 (QUIT: Remote host closed the connection) (~tslil@user/op-4/x-9116473) | 04:05 | |
| + op_4 (~tslil@user/op-4/x-9116473) | 04:05 | |
| - mjw (QUIT: Ping timeout: 255 seconds) (~mjw@gnu.wildebeest.org) | 04:07 | |
| - staticbunny (QUIT: Ping timeout: 245 seconds) (~staticbun@76-223-253-78.lightspeed.frokca.sbcglobal.net) | 05:16 | |
| - aloo_shu (QUIT: Ping timeout: 255 seconds) (~aloo_shu@90.166.193.0) | 05:31 | |
| + staticbunny (~staticbun@76-223-253-78.lightspeed.frokca.sbcglobal.net) | 05:35 | |
| + aloo_shu (~aloo_shu@90.166.193.0) | 05:39 | |
| - staticbunny (QUIT: Ping timeout: 276 seconds) (~staticbun@76-223-253-78.lightspeed.frokca.sbcglobal.net) | 06:16 | |
| - mesaoptimizer (QUIT: Ping timeout: 265 seconds) (~mesaoptim@user/PapuaHardyNet) | 08:54 | |
| + gustav28 (~gustav@c-4135524e.019-141-67626730.bbcust.telenor.se) | 10:02 | |
| + mesaoptimizer (~mesaoptim@user/PapuaHardyNet) | 10:31 | |
| - jacobk (QUIT: Ping timeout: 248 seconds) (~quassel@47-186-105-237.dlls.tx.frontiernet.net) | 11:04 | |
| + jacobk (~quassel@47-186-105-237.dlls.tx.frontiernet.net) | 11:05 | |
| _hramrach | hm, Fn-Backspace is Delete, would be probably more useful if it were F11 | 13:18 |
|---|---|---|
| _hramrach | hm, there is F11 on Q and F13 on E but W sends something that is not recognized | 13:19 |
| minute | _hramrach: yeah sth is wrong with F12 | 13:45 |
| minute | probably wrong scancode? | 13:45 |
| _hramrach | yes, probably. Or some wrong translation table somewhere | 14:00 |
| - mtm (QUIT: Ping timeout: 252 seconds) (~textual@c-71-228-84-213.hsd1.fl.comcast.net) | 14:04 | |
| + mtm (~textual@c-71-228-84-213.hsd1.fl.comcast.net) | 14:06 | |
| + paperManu (~paperManu@198.16.214.40) | 14:10 | |
| + akira (~akira@2a01:599:a2b:5753:8d8a:bcf0:667c:3c30) | 14:35 | |
| + mjw (~mjw@gnu.wildebeest.org) | 14:56 | |
| + amospalla (~jordi@user/amospalla) | 15:21 | |
| - xktr_ (QUIT: Ping timeout: 248 seconds) (~xktr@user/xktr) | 15:31 | |
| + xktr (~xktr@user/xktr) | 15:38 | |
| - akira (QUIT: Read error: Connection reset by peer) (~akira@2a01:599:a2b:5753:8d8a:bcf0:667c:3c30) | 15:59 | |
| + akira (~akira@37.4.230.225) | 15:59 | |
| amospalla | Hi, any hint about gpu power consumption tweaks on Pocket? | 16:02 |
| minute | amospalla: not yet, but we need to look into it, esp gpu clocking | 16:34 |
| minute | amospalla: running chromium with hardware accel off leads to a faster & lower energy experience at least :D | 16:34 |
| minute | (on imx8mp) | 16:35 |
| amospalla | minute: I read this on the forums, but my workflow is firefox based. I'll live with that until I upgrade my soc. | 16:37 |
| amospalla | But I was interested in something to lower the power consumption, to include in my pseudo-suspend script. | 16:37 |
| amospalla | Ah well, you already answered that :D | 16:39 |
| amospalla | So right now there is not anything I can do in user space as I understand. | 16:40 |
| amospalla | Thank you. | 16:40 |
| _hramrach | how does software-only rendering work for video playback? | 17:32 |
| _hramrach | my general impression is that usually it's worthwhile to offload decoding to GPU so long as it can actually decode the video format but did not try to do any benchmarking on imx8mp | 17:34 |
| + murph_nj (~murph@pool-100-1-80-189.nwrknj.fios.verizon.net) | 17:50 | |
| - murphnj (QUIT: Ping timeout: 248 seconds) (~murph@user/murphnj) | 17:52 | |
| + murphnj (~murph@user/murphnj) | 17:55 | |
| - murph_nj (QUIT: Ping timeout: 264 seconds) (~murph@pool-100-1-80-189.nwrknj.fios.verizon.net) | 17:57 | |
| + murph_nj (~murph@pool-100-1-80-189.nwrknj.fios.verizon.net) | 18:00 | |
| - murphnj (QUIT: Ping timeout: 264 seconds) (~murph@user/murphnj) | 18:00 | |
| + murph__ (~murph@pool-100-1-80-189.nwrknj.fios.verizon.net) | 18:02 | |
| - murph__ (QUIT: Remote host closed the connection) (~murph@pool-100-1-80-189.nwrknj.fios.verizon.net) | 18:02 | |
| - murph_nj (QUIT: Ping timeout: 244 seconds) (~murph@pool-100-1-80-189.nwrknj.fios.verizon.net) | 18:04 | |
| - mjw (QUIT: Ping timeout: 246 seconds) (~mjw@gnu.wildebeest.org) | 18:36 | |
| minute | _hramrach: there's no gpu video decode at all in no scenario on these platforms. vpu decode, outside of browsers, in some players. also gpu rendering in chromium is not strictly coupled to gpu compositing. so, video playback can still be gpu scaled | 18:43 |
| + mjw (~mjw@gnu.wildebeest.org) | 18:53 | |
| amospalla | Regarding GPU frequency on Pocket. Using sway and foot terminal, `fd / -xdev` takes: 6,75 seconds @ 1000Mhz, 8,3 seconds @ 200Mhz. Probably, if one is spending all the time between terminals it can be a good idea to stay on 200Mhz. I don't know how many current this saves though. | 19:06 |
| Twodisbetter | amospalla: The script with the gpu downclocking seems to work. I could tell that the display lid cooled down. I mean it wasn't cold but it was musch cooler than when running. Difference of 55-60C to 35-45C. | 19:09 |
| amospalla | Twodisbetter: that are good news! | 19:09 |
| Twodisbetter | for those trying to mitigate wifi reception as a result of heat build mighth find psuspend givbes the wireless chips a bit of a reprieve. | 19:09 |
| amospalla | I have a spreadsheet with the current consumption and uptimes of the Pocket with the things I was adding to psuspend. | 19:10 |
| amospalla | Tomorrow I'll measure the wattage (mean) and uptime with and without this on my Pocket, it will take a couple of days though for it to deplete completely two times. | 19:11 |
| Twodisbetter | amospalla: awesome. Looking forward to hearing about your results. For your information, I just suspended again with the USB C PD unplugged, and I am noting about a .020A difference. Combined with the noticeably cooler temps, I'd say this is a win. | 19:12 |
| amospalla | Twodisbetter: do you like the notification leds? :D | 19:13 |
| amospalla | Great, 0.02A are good to save. | 19:17 |
| _hramrach | technically it's VPU decode but the option in the player wouls still say GPU | 19:17 |
| _hramrach | I tried to selectively disable only the rendering on GPU but do not really see any significant difference either way, at least performance wise | 19:18 |
| Twodisbetter | amospalla: I sawe that in the script not sure what it is supposed to do though. | 19:23 |
| Twodisbetter | amospalla: I saw that in the script not sure what it is supposed to do though. | 19:23 |
| amospalla | Twodisbetter: when you suspend or resume, your leds should give you feedback with a green colour transition. | 19:23 |
| amospalla | Nothing important, I have fun doing these things. | 19:24 |
| Twodisbetter | amospalla: ahh that is what the purpose of that is. Yes I like that. | 19:24 |
| amospalla | The idea is, you know then it starts and when it ends. | 19:24 |
| amospalla | The OLED would be great way to notify things. | 19:26 |
| Twodisbetter | I agree. Always thought the OLED on the Reform and Pocket was a cool additional window into the Reform | 19:29 |
| dozens | i use the oled a lot! to track battery charge | 19:32 |
| Twodisbetter | for sure. I really like that it is a seperate system. Doesn't matter if the SoC is borked and not responding. The LPC always can do stuff. | 19:40 |
| ^alex | yeah: the SoC has a FEP in classic minicomputer style | 19:41 |
| vkoskiv | Did that apt upgrade issue get resolved? It seemingly borked my /boot somehow | 19:42 |
| - akira (QUIT: Quit: Quit) (~akira@37.4.230.225) | 20:01 | |
| + colinsane (~colinunin@97-113-148-88.tukw.qwest.net) | 20:45 | |
| jfred | hmm, I've been having trouble getting my yubikey to work with gpg on the pocket. this is where my lack of udev knowledge is biting me, but the yubikey seems to have the uaccess tag... | 21:23 |
| jfred | I'm used to throwing stuff in the plugdev group, less familiar with how uaccess is meant to work | 21:24 |
| minute | amospalla: 1ghz and 200mhz? sounds like noc frequency. or what are you setting? | 22:04 |
| amospalla | minute: what do you mean? | 22:10 |
| amospalla | what is noc? | 22:10 |
| amospalla | Maybe I am using the wrong path? | 22:11 |
| minute | amospalla: noc is the network-on-chip, regulates bandwidth and QoS of the IP components (who gets priority, memory throughput etc) | 22:13 |
| minute | by IP i mean blocks like CPUs, GPU, VPU, PCIe etc | 22:14 |
| minute | these are competing to access memory | 22:15 |
| - gustav28 (QUIT: Quit: Quit) (~gustav@c-4135524e.019-141-67626730.bbcust.telenor.se) | 22:15 | |
| jfred | ah... I think what's going on with my yubikey is that my gpg version and pcscd version are different (because the former is from guix) and there's a protocol mismatch | 22:23 |
| jfred | hrm | 22:23 |
| - sevan (QUIT: Ping timeout: 252 seconds) (~sevan@user/venture37) | 22:25 | |
| + sevan (~sevan@2001:470:1f1d:1d6:5a55:caff:fe24:ed4) | 22:31 | |
| jfred | minute: I'm curious, are you thinking about making a purple version of the Reform Next? It turned out really well on the Pocket | 22:39 |
| amospalla | minute: I am setting /sys/devices/platform/soc@0/32700000.interconnect/devfreq/32700000.interconnect | 22:43 |
| amospalla | My only test has been watching how a terminal emulator scrolled faster or slower by changing the frequency, nothing fancier. | 22:45 |
| amospalla | But it was clearly visible, and measureable. | 22:46 |
| amospalla | Some command scrolled in 6,75 seconds vs 8,3 seconds. I'll try tomorrow something fancier than scrolling text. | 22:47 |
| + staticbunny (~staticbun@76-223-253-78.lightspeed.frokca.sbcglobal.net) | 22:52 | |
| minute | jfred: yes | 22:52 |
| jfred | :D | 22:52 |
| minute | amospalla: ah yes, interconnect. | 22:52 |
| minute | amospalla: interconnect = NOC. i have to dig what exactly that frequency influences. but it's not the GPU directly | 22:55 |
| - staticbunny (QUIT: Ping timeout: 248 seconds) (~staticbun@76-223-253-78.lightspeed.frokca.sbcglobal.net) | 22:58 | |
| josch | vkoskiv: which apt upgrade issue? | 22:59 |
| vkoskiv | This one: https://mister-muffin.de/p/xfhP.txt | 22:59 |
| vkoskiv | This was linked when I asked about it: https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=1077678 | 23:00 |
| josch | ah the xtrx issue | 23:00 |
| minute | here one can see that the gpu runs at 800mhz: https://github.com/torvalds/linux/blob/89f5e14d05b4852db5ecdf222dc6a13edc633658/arch/arm64/boot/dts/freescale/imx8mp.dtsi#L2152 | 23:01 |
| vkoskiv | I got the flu soon after, so I really have no clue what happened next | 23:01 |
| vkoskiv | I just fired up the reform today, and no worky :D | 23:01 |
| minute | vkoskiv: do you use an xtrx card in the reform? | 23:01 |
| vkoskiv | I don't, no. | 23:02 |
| vkoskiv | But I have done some software radio stuff before, so maybe something from that caused that package to be installed | 23:02 |
| josch | since #1077678 is still unresolved, the issue seems to persist, sadly | 23:03 |
| josch | if you don't need it, maybe disable it temporarily? | 23:04 |
| vkoskiv | Yeah, I'll try to figure out a rescue thing tomorrow after work, I guess | 23:04 |
| vkoskiv | I need to re-lookup the stuff needed to mount luks, and the commands needed for chroot, and then do some apt magic to revive it | 23:05 |
| - xktr (QUIT: Ping timeout: 248 seconds) (~xktr@user/xktr) | 23:33 | |
| + xktr (~xktr@user/xktr) | 23:35 | |
| + staticbunny (~staticbun@76-223-253-78.lightspeed.frokca.sbcglobal.net) | 23:41 | |
| - staticbunny (QUIT: Ping timeout: 265 seconds) (~staticbun@76-223-253-78.lightspeed.frokca.sbcglobal.net) | 23:46 | |
Generated by irclog2html.py 2.17.2 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!