| Zaba | vkoskiv: yeah was super nice to meet up, thanks a lot :) I'm home now but hopefully we can catch up another time | 00:00 |
|---|---|---|
| - mtm (QUIT: Ping timeout: 252 seconds) (~textual@c-71-228-84-213.hsd1.fl.comcast.net) | 02:04 | |
| + mtm (~textual@c-71-228-84-213.hsd1.fl.comcast.net) | 02:06 | |
| - cobra (QUIT: Ping timeout: 252 seconds) (~cobra@user/Cobra) | 02:11 | |
| + cobra (~cobra@user/Cobra) | 02:23 | |
| - jn (QUIT: Ping timeout: 252 seconds) (~quassel@user/jn/x-3390946) | 03:39 | |
| - 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 | |
| - colinsane (QUIT: Quit: bye) (~colinunin@97-113-150-69.tukw.qwest.net) | 04:30 | |
| + colinsane (~colinunin@97-113-150-69.tukw.qwest.net) | 04:32 | |
| - mark_ (QUIT: Ping timeout: 276 seconds) (~mjw@gnu.wildebeest.org) | 04:41 | |
| + jn (~quassel@2001-4dd3-e14e-0-20d-b9ff-fe49-15fc.ipv6dyn.netcologne.de) | 05:36 | |
| - jn (QUIT: Changing host) (~quassel@2001-4dd3-e14e-0-20d-b9ff-fe49-15fc.ipv6dyn.netcologne.de) | 05:36 | |
| + jn (~quassel@user/jn/x-3390946) | 05:36 | |
| - nsc (QUIT: Ping timeout: 255 seconds) (~nicolas@i5C74DCE3.versanet.de) | 07:09 | |
| - cobra (QUIT: Quit: ZNC 1.8.2 - https://znc.in) (~cobra@user/Cobra) | 08:03 | |
| + cobra (~cobra@user/Cobra) | 08:08 | |
| - GNUmoon (QUIT: Ping timeout: 260 seconds) (~GNUmoon@gateway/tor-sasl/gnumoon) | 12:09 | |
| + GNUmoon (~GNUmoon@gateway/tor-sasl/gnumoon) | 12:21 | |
| + nsc (~nicolas@i5C74DCE3.versanet.de) | 12:23 | |
| - _justin_kelly (QUIT: Quit: The Lounge - https://thelounge.chat) (~justinkel@user/justin-kelly/x-6011154) | 13:06 | |
| + _justin_kelly (~justinkel@user/justin-kelly/x-6011154) | 13:15 | |
| + mark_ (~mjw@gnu.wildebeest.org) | 13:35 | |
| - mtm (QUIT: Ping timeout: 260 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 | |
| Twodisbetter | minute: I appreciate all of your efforts and the MNT's efforts very much! I hope that you'll be able to square the suspend issue on the Pocket Reform soon. As soon as suspend is possible I'll be dailying it. It is such a great devicve. Once again I really love the design! Knocked it out of the park again! | 15:53 |
| Twodisbetter | for those not familiar with baseball lingo, knocking it out of the park means you hit a homerun. A homerun is an automatic point as if you ran all of the bases and made it to home plate. A homerun is one of the best things you can get in baseball. So really it is just a way of saying your really did a great job. | 16:06 |
| * mark_ -> mjw | 16:52 | |
| jn | Twodisbetter: as a german (there's no baseball here, not even as a significant cultural import), thank you for the explanation :) | 18:16 |
| + blast007` (~blast@user/blast007) | 18:42 | |
| - blast007 (QUIT: Killed (NickServ (GHOST command used by blast007`))) (~blast@user/blast007) | 18:46 | |
| * blast007` -> blast007 | 18:46 | |
| - ptrc (QUIT: *.net *.split) (~ptrc@ptrc.gay) | 18:47 | |
| - inhji1 (QUIT: *.net *.split) (~inhji@2a01:4f8:1c0c:42e4::1) | 18:47 | |
| - n_to (QUIT: *.net *.split) (~n_to@2a03:4000:6:3662:24b1:57ff:fec6:76c1) | 18:47 | |
| + ptrc (~ptrc@ptrc.gay) | 18:48 | |
| + inhji1 (~inhji@2a01:4f8:1c0c:42e4::1) | 18:48 | |
| + n_to (~n_to@2a03:4000:6:3662:24b1:57ff:fec6:76c1) | 18:48 | |
| - TadeusTaD (QUIT: Quit: Stay Cheeki Breeki) (tadeustad@user/TadeusTaD) | 19:21 | |
| + TadeusTaD (tadeustad@psifactor.pl) | 19:22 | |
| - TadeusTaD (QUIT: Changing host) (tadeustad@psifactor.pl) | 19:22 | |
| + TadeusTaD (tadeustad@user/TadeusTaD) | 19:22 | |
| - TadeusTaD (QUIT: Remote host closed the connection) (tadeustad@user/TadeusTaD) | 19:25 | |
| + TadeusTaD (tadeustad@psifactor.pl) | 19:47 | |
| - TadeusTaD (QUIT: Changing host) (tadeustad@psifactor.pl) | 19:47 | |
| + TadeusTaD (tadeustad@user/TadeusTaD) | 19:47 | |
| - TadeusTaD (QUIT: Remote host closed the connection) (tadeustad@user/TadeusTaD) | 19:48 | |
| + TadeusTaD (tadeustad@psifactor.pl) | 19:55 | |
| - TadeusTaD (QUIT: Changing host) (tadeustad@psifactor.pl) | 19:55 | |
| + TadeusTaD (tadeustad@user/TadeusTaD) | 19:55 | |
| + gustav28 (~gustav@c-1d52205c.019-141-67626730.bbcust.telenor.se) | 20:13 | |
| Twodisbetter | jn: Mein Geschenen! | 20:24 |
| Twodisbetter | Or my pleasure! | 20:25 |
| jn | you probably mean "gern geschehen" (literally "[I'm] glad it happened", semantically equvalent to "my pleasure") | 20:30 |
| Twodisbetter | Yeah, ich kann Deutsch aber es ist nicht perfekt. Noch nicht. Thanks for the correction! | 20:37 |
| jn | you'll get there :) | 20:38 |
| Twodisbetter | yeah, I live in Germany, but at work English is used so it slows down progress | 21:16 |
| Twodisbetter | But I am married to a German and my kids all speak it, so I'll get there for sure. | 21:16 |
| minute | played some dhewm3 with hd texture and lighting mods on my rk3588 reform. fun! | 21:43 |
| josch | minute: will the imx8mplus load u-boot only from emmc and never from sd-card? | 21:46 |
| josch | i'm wondering whether running "reform-flash-uboot emmc" on the pocket was ever tested (there is a bug in reform-tools which prevents it from working right now) and whether this could soft-brick the pocket | 21:47 |
| minute | josch: imx8mplus can only boot from emmc yes | 21:51 |
| minute | josch: i have always (in our factory script) manually flashed uboot on the emmc boot partition | 21:52 |
| josch | minute: on the forums there is somebody who patched reform-flash-uboot to make it work on the pocket and can now not boot it anymore after flashing latest u-boot to it | 21:52 |
| josch | minute: you probably didn't use reform-flash-uboot to do that? | 21:52 |
| minute | josch: no | 21:52 |
| josch | hm :/ | 21:53 |
| minute | josch: just dd | 21:53 |
| josch | i guess i might not fix this bug too soon | 21:53 |
| minute | i don't know why someone would flash uboot on the imx8mplus | 21:53 |
| minute | it's already there :D | 21:53 |
| josch | minute: in reform-check i output a message when u-boot is not the latest version | 21:53 |
| josch | minute: the goal with that message is to get people away from old u-boot because of teh features that were only added later | 21:54 |
| josch | minute: remember the things we talked in the past that we cannot do (like u-boot-menu) because we cannot know whether or not the user has a modern-enugh u-boot? | 21:54 |
| josch | minute: if reform-tools should disallow flashing u-boot for some platforms, then i'll implement that next | 21:55 |
| minute | josch: yeah, maybe for imx8mplus it should be done only by people who _really_ want to do that and know what they're doing | 21:58 |
| josch | ah this was the last time we talked about this: https://mntre.com/reform-irc-logs/2024-03-19.log.html#t21:50:47 | 21:58 |
| minute | i just posted a reply here https://community.mnt.re/t/flashing-emmc-bootloader-unbricking-imx8mp/2287/10 | 21:59 |
| josch | okay, in reform-tools we have encoded for the imx8mplus: UBOOT_OFFSET=33792 | 22:01 |
| josch | but in march you wrote: "EMMC offset is 0 if it is a boot partition or 32K if it is a user partition :D" | 22:01 |
| josch | so the offset should actually be zero? | 22:01 |
| ^alex | oh hey we know andy piper | 22:02 |
| josch | i shall also set EMMC_BOOT=falso so that this cannot happen again unless you "know what you are doing" | 22:02 |
| josch | minute: am i reading this correctly, that the correct u-boot offset would be 0? | 22:03 |
| minute | josch: yes, 0 | 22:03 |
| josch | sorry for this :( | 22:04 |
| minute | josch: on emmc. the other offset is for sd card i think, which doesn't work anyway on that module because of efuses or sth | 22:04 |
| minute | josch: no problem | 22:04 |
| minute | with that adafruit cable it should be a bit less painful to recover | 22:05 |
| josch | i'm going to correct this in the next reform-tools version. It's probably not urgent because another bug in reform-flash-uboot makes flashing u-boot on the pocket impossible unless you change the script. | 22:05 |
| minute | haha i'm reporting bugs in the panfrost channel and they're like "kernel 6.9 is EOL" | 22:09 |
| minute | but i'm like, that's the newest thing on debian...... | 22:09 |
| josch | 6.10 is in experimental but has other regressions | 22:09 |
| minute | meh :D | 22:09 |
| josch | i left a message on the forums explaining the mess surrounding reform-flash-uboot on the imx8mplus... i hope it helps a bit, sorry | 22:12 |
| josch | both fixes are pushed to the staging branch in https://source.mnt.re/reform/reform-tools/-/merge_requests/79 | 22:14 |
| minute | josch: many thanks! | 22:14 |
| - gustav28 (QUIT: Quit: Quit) (~gustav@c-1d52205c.019-141-67626730.bbcust.telenor.se) | 22:15 | |
| josch | meh... machste dreckig machste sauber (you make a mess, you clean it up) | 22:19 |
| josch | is there anything else you can think of that is on my TODO list in the near future? | 22:19 |
| josch | next week there is no kita and my partner is working so i'll be mostly afk | 22:19 |
| minute | josch: system image for next would be cool, but yeah i still need to push a dts fix tomorrow | 22:22 |
| josch | that should be simple to do -- i'm looking forward to pushing that :) | 22:22 |
| josch | we can just have a mini reform-tools release which also gets the pocket emmc u-boot fix out together with the reform next machines config | 22:23 |
| josch | one thing to think about is whether the number of system images starts getting out of hand with the introduction of the reform next | 22:25 |
| josch | unless the reform next should only have the rk3588 and not a311d for example? | 22:25 |
| minute | josch: only rk3588 for now | 22:30 |
| minute | josch: but yeah, later we could think about a different approach of avoiding all the duplication among system images... not sure if feasible | 22:31 |
| josch | for reform.debian.net i thought about not letting the webserver serve plain files but use a cgi-script which concatenates the /boot partition which is variable with the static root partition | 22:32 |
| josch | another way would be to offer multiple /boot images and one root image as files and then tell the user to just dd two things instead of one -- but i don't like shifting the complexity to the user much | 22:32 |
| josch | and then there is how Debian arm64 netboot installer images are supposed to be manually dd-ed together before flashing them to the disk (also static rootfs and variable boot): http://ftp.debian.org/debian/dists/bookworm/main/installer-arm64/current/images/netboot/SD-card-images/ | 22:34 |
| ^alex | we heard from the old sysadmins at our college that the homegrown imager they used before switching to redhat kickstart was effectively a netbooted `cat $IMAGE | sed $STUFF | dd of=$DRIVE` | 22:37 |
| ^alex | only the tiniest bit cursed | 22:37 |
| josch | i we knew we had modern u-boot on all platforms we could do something very similar because the major difference is which device tree to load | 22:41 |
| josch | the sed magic works well if it operated on a fat32 partition where no magic is going on when files get changed :) | 22:42 |
| josch | ^alex: i don't think the sed option is that cursed... it should be possible to modify an ext4 inode content of a symlink that way if it is crafted carefully and then the /boot partition just has to contain all the dtbs we care about... hrm... | 22:45 |
| josch | i currently do not have a second computer at hand -- would it be easy for somebody to share their printenv output from u-boot? | 22:59 |
| - Gooberpatrol66 (QUIT: Read error: Connection reset by peer) (~Gooberpat@user/gooberpatrol66) | 23:06 | |
| jfred | Interesting, I just tried launching Roll20 on my Pocket Reform for a D&D session I'm in and Firefox crashed O.o | 23:23 |
Generated by irclog2html.py 2.17.2 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!