| + mark_ (~mark@gnu.wildebeest.org) | 00:10 | |
| - mjw (QUIT: Ping timeout: 250 seconds) (~mark@gnu.wildebeest.org) | 00:12 | |
| - S0rin (QUIT: Ping timeout: 276 seconds) (~S0rin@user/s0rin) | 00:25 | |
| + S0rin (~S0rin@user/s0rin) | 00:27 | |
| - mtm (QUIT: Ping timeout: 276 seconds) (~mtm@c-71-228-84-213.hsd1.fl.comcast.net) | 02:03 | |
| bkeys | minute: That is great news! When do you plan to ship it back? | 02:09 |
|---|---|---|
| bkeys | I sent them $15 for metal sides to go with it and asked to have y'all ship it with my Reform to save on shipping | 02:09 |
| - vagrantc (QUIT: Quit: leaving) (~vagrant@2600:3c01:e000:21:7:77:0:20) | 02:11 | |
| bkeys | And anri_paul said that they would see into it if the repair was a warranty issue or not. I have no clue what your profit margins are but they can't be that big. I am fine with whatever you think is fair, I am grateful that the laptop exists in the first place | 02:12 |
| + mtm (~mtm@71.228.84.213) | 04:01 | |
| + reformer (~reformer@softboy.mntmn.com) | 04:22 | |
| + cwebber (~user@user/cwebber) | 04:24 | |
| + jackhill (~jackhill@kalessin.dragonsnail.net) | 04:25 | |
| - S0rin (QUIT: Ping timeout: 260 seconds) (~S0rin@user/s0rin) | 06:19 | |
| + S0rin (~S0rin@user/s0rin) | 06:19 | |
| - frank2- (QUIT: Quit: ZNC 1.8.2 - https://znc.in) (~frank2@2a05:f480:2000:1d0d:8527:b4ad:1361:a9ad) | 06:54 | |
| + frank2 (~frank2@juicy.frank2.net) | 06:54 | |
| + bgs (~bgs@212-85-160-171.dynamic.telemach.net) | 07:03 | |
| - bgs (QUIT: Remote host closed the connection) (~bgs@212-85-160-171.dynamic.telemach.net) | 07:22 | |
| + chomwitt (~chomwitt@2a02:587:dc10:8200:1ac0:4dff:fedb:a3f1) | 07:36 | |
| - ajr (QUIT: Quit: WeeChat 3.7) (~ajr@user/ajr) | 07:53 | |
| * mark_ -> mjw | 09:21 | |
| - mjw (QUIT: Killed (NickServ (GHOST command used by wielaard!~mjw_@2001:1c06:2488:1400:9e5c:8eff:fe8f:a440))) (~mark@gnu.wildebeest.org) | 10:01 | |
| * wielaard -> mjw | 10:01 | |
| + mark_ (~mark@gnu.wildebeest.org) | 10:01 | |
| - qwer (QUIT: Read error: Connection reset by peer) (~qwer@37.48.27.27) | 10:40 | |
| + qwer (~qwer@37-48-27-27.nat.epc.tmcz.cz) | 10:46 | |
| - qwer (QUIT: Ping timeout: 240 seconds) (~qwer@37-48-27-27.nat.epc.tmcz.cz) | 11:31 | |
| - mtm (QUIT: Ping timeout: 260 seconds) (~mtm@c-71-228-84-213.hsd1.fl.comcast.net) | 14:03 | |
| - ex-parrot (QUIT: Ping timeout: 260 seconds) (~fincham@user/ex-parrot) | 14:49 | |
| + ex-parrot (~fincham@user/ex-parrot) | 14:57 | |
| - S0rin (QUIT: Ping timeout: 240 seconds) (~S0rin@user/s0rin) | 15:16 | |
| + S0rin (~S0rin@user/s0rin) | 15:17 | |
| - Nulo (QUIT: Read error: Connection reset by peer) (~Nulo@user/nulo) | 15:24 | |
| + Nulo (~Nulo@user/nulo) | 15:24 | |
| - ex-parrot (QUIT: Quit: _b) (~fincham@user/ex-parrot) | 15:34 | |
| + ex-parrot (~fincham@user/ex-parrot) | 15:34 | |
| + rafostar[m] (~rafostarm@2001:470:69fc:105::2:52da) | 15:38 | |
| + mtm (~mtm@c-71-228-84-213.hsd1.fl.comcast.net) | 16:09 | |
| minute | what would you prefer as a high perf soc for reform, nvidia jetson xavier nx or rockchip rk3588? | 16:40 |
| bkeys | minute: Whichever one has potential UEFI (edk2) support? | 16:43 |
| minute | bkeys: that's more important to you than performance, drivers etc? | 16:43 |
| bkeys | Well no, but I don't see the two as mutually exclusive | 16:43 |
| bkeys | Being able to run my distro of choice is important to me | 16:43 |
| minute | ok | 16:44 |
| bkeys | Does xavier and rk3588 not have mainline kernel support? | 16:45 |
| minute | rk3588 does not | 16:47 |
| minute | and the open gpu driver is very far behind the blob one | 16:47 |
| minute | according to https://forum.radxa.com/t/rock-5b-debug-party-invitation/10483 | 16:48 |
| minute | the rk3588 cpu perf is pretty spectacular though | 16:48 |
| bkeys | It's just my 2 cents but I would be willing to pay a pretty penny for a decent performance riscv SoM for the Reform | 16:48 |
| bkeys | https://github.com/antmicro/arvsom | 16:49 |
| bkeys | Something like this for the Reform would be awesome | 16:49 |
| minute | that's not performant at all though | 16:49 |
| minute | but yeah | 16:49 |
| minute | risc-v will be interesting at some point | 16:50 |
| bkeys | Yeah riscv lacks anything with decent horsepower; but a ppc64le SoM would also be cool, although that is probably not practical at all | 16:50 |
| minute | yeah, and ppc community is tiny | 16:50 |
| minute | i'm pretty sure there will be a cm4 compatible starfive risc-v thing next year | 16:51 |
| bkeys | As for me I am fine living with imx8 on a laptop for the time being; although if somehow there came out a SoM that would let me use the Reform as my desktop as well that would be really cool | 16:51 |
| minute | yeah, that's what i want, too | 16:52 |
| josch | what is the use-case for more performance than the imx8 or even the ls1028a? personally i'd prefer that as much as possible is upstreamed into the respective projects and the platform is really stable and reliable. | 16:52 |
| minute | josch: for that avenue, we'll have imx8mplus as the next step (i just have to assemble and test it) | 16:53 |
| minute | josch: for my coworker, reform is too slow for customer support using gitlab web application | 16:53 |
| minute | the javascript perf isn't enough | 16:53 |
| josch | damn javascript is at it again XD | 16:54 |
| minute | yeah, especially bad because js is single threaded | 16:54 |
| minute | i guess 8mplus will give us 20-30% more | 16:56 |
| josch | to be fair, javascript also brought my old intel laptop to its knees. | 16:56 |
| minute | but jetson xavier nx can give us roughly triple the single core perf | 16:56 |
| bkeys | And at least eventually there would be an officially supported driver | 16:57 |
| minute | bkeys: there is ESXi-Arm for jetson nx | 16:58 |
| bkeys | I think adapter support for anything cm4 compatible would be nice | 16:58 |
| bkeys | What is ESXi-Arm? | 16:58 |
| minute | ah, there is tianocore https://nullr0ute.com/2021/12/fedora-on-nvidia-jetson-xavier/ | 16:59 |
| minute | bkeys: i do have a CM4 adapter here, but will continue the bringup only next week. | 17:00 |
| minute | (finishing reform motherboard 2.5 at the moment) | 17:00 |
| minute | mainline linux support also doesn't look so bad. | 17:01 |
| bkeys | If a cm4 adapter existed where I could put any cm4 compatible SoM in I would buy that in a heartbeat | 17:02 |
| minute | well, we'll sell it soon. i just need to fix some display and ethernet things. | 17:04 |
| josch | how does it work with the remaining available height for the heatsink if there is the cm4 adapter board in the middle? | 17:06 |
| bkeys | Cause I think even with rpi cm4 module I could get edk2 and mainline Fedora finally supports rpi; even if my GPU performance wouldn't be as good I'd have Vulkan | 17:06 |
| minute | josch: there will be a slimmer and potentially wider heatsink, but otoh the rpi cm4 generates less heat | 17:06 |
| josch | ACTION is still waiting for the transcend 220S nvme ssd which will hopefull fix the suspend issues... | 17:12 |
| minute | :0 | 17:15 |
| josch | shipment is stuck with DHL for over a week and the online shop now sent a second package via UPS express | 17:16 |
| minute | josch: oof :/ | 17:30 |
| + vagrantc (~vagrant@2600:3c01:e000:21:7:77:0:20) | 18:17 | |
| mtm | minute: is it possible to reconfigure parts of the FPGA in the RKX7 on the fly after the initial config is loaded from the flash? | 18:47 |
| sigrid | 16:40 < minute> what would you prefer as a high perf soc for reform, nvidia jetson xavier nx or rockchip rk3588? | 19:08 |
| sigrid | whichever has a better documentation to implement support for | 19:08 |
| sigrid | does nvidia one even have some kind of reference manual? all I can find is "this is how you run Linux on this thing" pdfs | 19:11 |
| + Boostisbetter (4a410829d7@irc.cheogram.com) | 19:24 | |
| - mjw (QUIT: Killed (NickServ (GHOST command used by mark_!~mark@gnu.wildebeest.org))) (~mjw_@2001:1c06:2488:1400:9e5c:8eff:fe8f:a440) | 19:28 | |
| * mark_ -> mjw | 19:28 | |
| + wielaard (~mjw_@2001:1c06:2488:1400:9e5c:8eff:fe8f:a440) | 19:29 | |
| + ajr (~ajr@user/ajr) | 19:32 | |
| - vkoskiv (QUIT: Ping timeout: 260 seconds) (~vkoskiv@89-166-62-97.bb.dnainternet.fi) | 19:33 | |
| - sigrid (QUIT: Remote host closed the connection) (~sigrid@ftrv.se) | 19:38 | |
| + sigrid (~sigrid@ftrv.se) | 19:42 | |
| + vkoskiv (~vkoskiv@89-166-62-97.bb.dnainternet.fi) | 19:46 | |
| Boostisbetter | hello again MNT crew | 20:09 |
| vkoskiv | Boostisbetter: Hello! | 20:21 |
| vkoskiv | minute: I saw an article claim that the chip shortage may be starting to reverse. Have you seen lead times decrease? | 20:22 |
| minute | Boostisbetter: welcome back, i was already wondering where you went! | 20:43 |
| minute | sigrid: well, i only know that there is a secret manual for rk3588 somewhere. | 20:43 |
| mtm | minute: congrats on getting the new MB update completed | 20:44 |
| minute | mtm: thanks! | 20:44 |
| minute | here's a render of motherboard 2.5 https://mastodon.social/@mntmn/109241624940403405 | 20:44 |
| minute | rk3588 technical reference manual "part 2" (pdf, 56 MB) https://dl.radxa.com/rock5/hw/datasheet/Rockchip%20RK3588%20TRM%20V1.0-Part2%2020220309.pdf | 20:45 |
| minute | ah, they killed the link. | 20:45 |
| sigrid | is it https://dl.radxa.com/rock3/docs/hw/datasheet/Rockchip%20RK3568%20TRM%20Part2%20V1.1-20210301.pdf ? | 20:46 |
| sigrid | oh no, it's 68... | 20:46 |
| minute | yeah, no | 20:46 |
| minute | it was floating around for a while, i guess someone will have a copy. | 20:47 |
| minute | meanwhile, xavier TRM _is_ available from nvidia | 20:48 |
| minute | i just downloaded it, 8288 pages | 20:48 |
| mtm | where can I find the equivalent of this: https://github.com/orangecrab-fpga/orangecrab-examples/blob/main/verilog/orangecrab_r0.2.pcf for the RKX7? I've been poking around the git repo but I'm not finding it | 20:49 |
| minute | mtm: what's a pcf? | 20:49 |
| minute | ah | 20:49 |
| minute | constraints file | 20:49 |
| mtm | yup | 20:49 |
| minute | well, it's generated dynamically by litex | 20:50 |
| sigrid | minute: mind giving a link? or do I need to register to download | 20:50 |
| minute | sigrid: you need to register and click "join the developer program" once, it's pretty painless | 20:50 |
| sigrid | ah ok | 20:50 |
| minute | link is here https://forums.developer.nvidia.com/t/xavier-soc-technical-reference-manual-now-available-for-download/68489 | 20:50 |
| minute | mtm: https://github.com/litex-hub/litex-boards/blob/master/litex_boards/platforms/mnt_rkx7.py | 20:51 |
| minute | mtm: https://github.com/litex-hub/litex-boards/blob/master/litex_boards/targets/mnt_rkx7.py | 20:52 |
| mtm | interesting. How would litex know the configuration of the external hardware, for example the physical pins connected to the RAM? Sorry if this is a basic question, still getting up to speed on FPGA stuff | 20:53 |
| mtm | ah | 20:53 |
| minute | mtm: see the first link (platforms) | 20:53 |
| mtm | thanks, slow typing... | 20:53 |
| mtm | perfect, just what I was looking for | 20:54 |
| bluerise | minute: just wondering, is there any ETA for the CM4 adapter? something I can use with rk35xx? iirc imx8mp is also in the pipeline? | 20:54 |
| minute | bluerise: yes, i will resume work on CM4 next week | 20:55 |
| minute | and imx8mplus adapter will be assembled tomorrow | 20:55 |
| bluerise | cool | 20:55 |
| minute | > For Xavier clocks and resets are controlled by NVIDIA provided firmware running on BPMP. This hardware is not documented, and the software documentation should be consulted for details of how these are controlled | 20:58 |
| minute | that's > Boot and Power management Processor | 20:59 |
| minute | ok, that's kinda like a boot rom, only it also has a dedicated cpu/mcu https://www.kernel.org/doc/Documentation/devicetree/bindings/firmware/nvidia%2Ctegra210-bpmp.txt | 21:01 |
| minute | oh, it's > Dual ARM Cortex-R5F cores running in delayed lock-step with 32 KiB of I-cache and 32 KiB of D-cache, both with ECC | 21:03 |
| minute | https://www.thegoodpenguin.co.uk/blog/diving-into-the-nvidia-jetson-nano-boot-process/ | 21:11 |
| sigrid | there is seems to be another arm core for audio processing as well | 21:12 |
| minute | huh! | 21:15 |
| minute | oh yeah > ADSP based on an ARM Cortex -A9 version r4p1 processor with L1 and L2 Caches | 21:15 |
| - ajr (QUIT: Ping timeout: 244 seconds) (~ajr@user/ajr) | 22:02 | |
| Boostisbetter | what up MNT people! The move is complete and I got the servers back up and running | 22:33 |
| Boostisbetter | Testing? | 22:36 |
| + chartreuse (~chartreus@S0106f0f249dfd9c3.cg.shawcable.net) | 22:36 | |
| + ajr (~ajr@user/ajr) | 22:45 | |
| minute | Boostisbetter: pocket reform is shaping up, case design almost complete | 22:48 |
| minute | next week will try to bring up imx8mplus (it currently works with the standard imx8mq) | 22:49 |
| Boostisbetter | Super awesome! Me want still big time! Glad progress has been getting made! | 22:49 |
| minute | cool cool | 22:50 |
| minute | also, custom keycap mold is expected to be ready in mid november | 22:52 |
| * tadeus_brick -> TadeusTaD | 22:58 | |
Generated by irclog2html.py 2.17.3 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!