| - bkeys1 (QUIT: Ping timeout: 240 seconds) (~Thunderbi@172.56.71.36) | 00:02 | |
| + bkeys (~Thunderbi@172.56.71.36) | 00:05 | |
| - pomel0 (QUIT: Ping timeout: 246 seconds) (~pomel0@user/pomel0) | 00:34 | |
| + pomel0 (~pomel0@user/pomel0) | 00:35 | |
| - pomel0 (QUIT: Read error: Connection reset by peer) (~pomel0@user/pomel0) | 00:46 | |
| + pomel0 (~pomel0@user/pomel0) | 00:51 | |
| josch | minute: we have the same problem with a311d u-boot as we have with ls1028a uboot i'm afraid :( | 01:05 |
|---|---|---|
| josch | a311d u-boot gets flashed at an offset of 512 bytes | 01:05 |
| minute | maybe f_ knows more? | 01:05 |
| josch | (which is even less than the old ls1028a offset with 4096 bytes) | 01:05 |
| minute | wait, a311d also doesn't use the emmc boot partitions? | 01:06 |
| josch | minute: thank you for pinging -- i'm currently working through https://source.mnt.re/bugs/bugs/-/work_items/50 and now i'm happy i'm using the efi images and not the default images or otherwise i wouldn't have noticed this :/ | 01:06 |
| josch | a311d uses sd-card if emmc is zeroed | 01:06 |
| josch | if you want to drop support, then shipping the a311d sd-card images makes no sense anymore and then a311d u-boot has to always be on emmc | 01:08 |
| josch | because yes, in contrast to rk3588, it is put onto boot0 and not onto the main drive | 01:09 |
| josch | the reason that reform-system-images is building a imx8m+ image (even though imx8m+ never reads u-boot from sd-card) is because the first version of u-boot that got flashed to pocket reforms does not support the reform-system-any image | 01:10 |
| minute | ok so what's the gpt offset again from the beginning of sd? | 01:12 |
| josch | guid partition table starts right after MBR at 512 bytes | 01:13 |
| josch | and then takes about 16kB of space | 01:14 |
| minute | ah, thx | 01:14 |
| josch | the first byte that is usable after guid partition table is byte 17408 | 01:15 |
| kop316 | minute: if i may ask, i put in a warranty request a week ago and didn't hear anything back. If I need to put in an order and pay for a new display I will, I just want to know. I also didn't see anything about the charger, and I think you wanted to expedite it | 01:15 |
| josch | (if i'm reading wikipedia correctly) | 01:15 |
| minute | kop316: sorry, i wasn't notified about the request yet, weird that you didn't hear back. will put this in my todo now | 01:17 |
| kop316 | minute: no worries! things happen | 01:18 |
| - chomwitt (QUIT: Ping timeout: 245 seconds) (~chomwitt@2a02:85f:9a67:8b00:1ac0:4dff:fedb:a3f1) | 01:19 | |
| minute | josch: i can try to read the reference manual tomorrow to see if there is a secondary boot offset. or maybe f_ will be able to say something in the meantime. | 01:22 |
| josch | minute: there is no rush -- i also have to find out how to tell amlogic-boot-fip about different offsets | 01:24 |
| josch | minute: and thank you :) | 01:24 |
| josch | and i can also ask in #linux-amlogic | 01:24 |
| josch | minute: i could ping you again if i fail to figure it out myself | 01:25 |
| josch | i think chance are higher that i find help with a311d elsewhere than it was with ls1028a :) | 01:25 |
| minute | josch: alright! | 02:04 |
| - vagrantc (QUIT: Quit: leaving) (~vagrant@2600:3c01:e000:21:7:77:0:50) | 02:20 | |
| - paperManu (QUIT: Ping timeout: 240 seconds) (~paperManu@107.159.15.124) | 02:35 | |
| - timide (QUIT: Ping timeout: 246 seconds) (~timide@user/timide) | 02:49 | |
| + timide (~timide@user/timide) | 02:58 | |
| - paperManu_ (QUIT: Ping timeout: 246 seconds) (~paperManu@107.159.15.124) | 03:10 | |
| - spew (QUIT: Quit: nyaa~) (~spew@user/spew) | 03:10 | |
| - bkeys (QUIT: Read error: Connection reset by peer) (~Thunderbi@172.56.71.36) | 03:24 | |
| - pomel0 (QUIT: Remote host closed the connection) (~pomel0@user/pomel0) | 04:16 | |
| + bkeys (~Thunderbi@98.19.131.193) | 04:30 | |
| - bkeys (QUIT: Quit: With every step we take, danger will follow closely) (~Thunderbi@98.19.131.193) | 04:42 | |
| + bkeys1 (~Thunderbi@h193.131.19.98.dynamic.ip.windstream.net) | 04:42 | |
| * bkeys1 -> bkeys | 04:45 | |
| + pomel0 (~pomel0@user/pomel0) | 05:04 | |
| - bkeys (QUIT: Read error: Connection reset by peer) (~Thunderbi@h193.131.19.98.dynamic.ip.windstream.net) | 05:08 | |
| - aelius (QUIT: Ping timeout: 246 seconds) (~aelius@user/aelius) | 06:02 | |
| + libercv (~libercv@user/libercv) | 06:49 | |
| - libercv (QUIT: Quit: Konversation terminated!) (~libercv@user/libercv) | 06:56 | |
| + chomwitt (~chomwitt@2a02:85f:9a67:8b00:1ac0:4dff:fedb:a3f1) | 07:17 | |
| + libercv (~libercv@user/libercv) | 07:22 | |
| - chomwitt (QUIT: Read error: Connection reset by peer) (~chomwitt@2a02:85f:9a67:8b00:1ac0:4dff:fedb:a3f1) | 07:29 | |
| f_ | minute: there isn't | 07:39 |
| - nybble (QUIT: Ping timeout: 256 seconds) (~sprang@user/csprng) | 07:39 | |
| f_ | but you can use boot0/boot1 | 07:39 |
| gordon1 | josch: i also noticed that my balls started to wear off (sorry) | 08:33 |
| gordon1 | the little bearing ones, i wonder if that's because of pigment used in POM ball | 08:34 |
| gordon1 | if it is carbon black it can have some mild abrasive properties maybe? | 08:36 |
| josch | f_: Thank you! Now I have to think whether I want to special-case A311D and build a special MBR image for it or not... hrm... | 08:54 |
| + FirefoxDeHuk (~FirefoxDe@user/FirefoxDeHuk) | 08:57 | |
| + liber (~libercv@88.148.112.158) | 09:02 | |
| - libercv (QUIT: Ping timeout: 240 seconds) (~libercv@user/libercv) | 09:04 | |
| - FirefoxDeHuk (QUIT: Quit: Client closed) (~FirefoxDe@user/FirefoxDeHuk) | 09:06 | |
| - liber (QUIT: Quit: Konversation terminated!) (~libercv@88.148.112.158) | 09:08 | |
| + jordi (~jordi@84.123.7.93.dyn.user.ono.com) | 09:31 | |
| + mjw (~mjw@gnu.wildebeest.org) | 10:01 | |
| josch | minute: i think i will just add a --disk-label={gpt,mbr} option to mkimage.sh which defaults to gpt and the a311d image is built with --disk-label=mbr | 10:32 |
| f_ | josch: you can use the boot0/boot1 partitions though | 10:36 |
| josch | f_: we are using them | 10:37 |
| f_ | Then you can use gpt? Unless you absolutely want a bootloader on the SD card | 10:37 |
| josch | f_: but what do i do when i want to experiment with gxlimg which i uploaded to Debian unstable recently and i build a new bootloader image -- do i flash it to my eMMC and risk soft-bricking my device? Or do i instead zero-out my boot0/boot1 partitions and then let a311d load my experimental bootloader from sd-card? | 10:38 |
| f_ | Right, then MBR it is for SD card | 10:39 |
| josch | i know i can recover from this soft-bricking but i find it easier to just flash stuff to sd-cards when experimenting with new images | 10:39 |
| f_ | same :) | 10:39 |
| josch | and then i want "create a bootable image with MBR" be a first-class citizen in our tooling and not something which i have to spend time hacking into the scripts every time i need it | 10:40 |
| f_ | that said, if all you want is the bootloader, no matter what the SD has, you can just dd said bootloader without dealing with partitions? | 10:40 |
| josch | f_: but then it will load whatever i have on eMMC and flashing eMMC with my latest disk images is harder than flashing an sd-card with them | 10:41 |
| f_ | ah ok | 10:41 |
| josch | i was running into this right now because i was updating u-boot (and the bootloaders before it) for all our five platforms | 10:42 |
| josch | at the same time we have an experimental MR which switches from MBR to GPT | 10:42 |
| josch | so i want to test both things at the same time | 10:42 |
| josch | and it's only because this was possible that allowed me to find this issue with the a311d offset which conflicts with the guid table | 10:42 |
| josch | so to be able to find bugs like this in the future, i'd like to keep this feature | 10:43 |
| josch | and it looks simple enough to support both mbr and gpt using sfdisk in our scripts | 10:43 |
| f_ | sure thing | 10:43 |
| josch | also TIL: if you pass sector-size: 512 to sfdisk, the resulting disk images will not be sparse anymore | 10:44 |
| josch | (we used parted before but had to switch to sfdisk because parted is unable to set the partition uuid and without being able to set it, the resulting disk images are not bit-by-bit reproudcible) | 10:45 |
| + XYZ_ (~XYZ@89-24-49-185.nat.epc.tmcz.cz) | 10:47 | |
| minute | josch: i see, then we're stuck with mbr/gpt duality there | 10:55 |
| minute | f_: thank you for confirming! | 10:55 |
| f_ | yw! | 10:56 |
| f_ | I hope someone begs amlogic to fix their offsets | 10:56 |
| f_ | ... a boy can dream... | 10:56 |
| gordon1 | minute: okay, i detected some ghost activity on my classic reform today, so sometimes when you turn it off and on quite quickly it refuses to boot and gets stuck somewhere in u-boot, i decided to plug in usb-c console to see what's happening, and it is stuck in u-boot boot menu and something is typing AT<LF>, i was pretty sure that it is something on my laptop messing with it and trying to init | 11:16 |
| gordon1 | it as a modem or something, so i plugged it in one of my boards that has pretty much nothing but picocom and it still did that, clues? | 11:16 |
| gordon1 | i'll check if it happens lest often with usb-c serial disconnected with dip switches | 11:17 |
| gordon1 | is there a convenient place where i can tap into CPU UART on mb3.0? | 11:18 |
| - mjw (QUIT: Ping timeout: 240 seconds) (~mjw@gnu.wildebeest.org) | 12:14 | |
| * Guest966 -> mjw | 12:14 | |
| - mjw (QUIT: Quit: Leaving) (~mjw@2001:1c06:2486:a800:a09a:fc1c:5a8:e74d) | 12:23 | |
| + gidzit (~gidzit@gidzit.org) | 12:28 | |
| - manis (QUIT: Ping timeout: 252 seconds) (01a66df340@185.72.67.185) | 12:29 | |
| + paperManu (~paperManu@107.159.15.124) | 12:47 | |
| + manis (01a66df340@185.72.67.185) | 12:48 | |
| josch | f_: even if they fix it, that will only happen in new silicon as the bootloader is in some read-only memory, no? | 12:50 |
| josch | minute: we do not have to be stuck in duality *if* we mandate that on a311d only booting from emmc is supported as it is the case with imx8m+ | 12:50 |
| - pomel0 (QUIT: Ping timeout: 240 seconds) (~pomel0@user/pomel0) | 12:58 | |
| + pomel0 (~pomel0@user/pomel0) | 12:58 | |
| + gustav25 (~gustav@c-78-82-54-48.bbcust.telenor.se) | 13:02 | |
| + mjw (~mjw@2001:1c06:2486:a800:a09a:fc1c:5a8:e74d) | 13:09 | |
| + paperManu_ (~paperManu@107.159.15.124) | 13:10 | |
| - Ar|stote|is (QUIT: Quit: No Ping reply in 180 seconds.) (~linx@149.210.3.83) | 13:18 | |
| f_ | josch: yes | 13:22 |
| + Ar|stote|is (~linx@149.210.3.83) | 13:24 | |
| + jahkosha (~jahkosha@user/jahkosha) | 13:45 | |
| - jah (QUIT: Ping timeout: 260 seconds) (~jahkosha@user/jahkosha) | 13:45 | |
| * jahkosha -> jah | 13:45 | |
| - pomel0 (QUIT: Read error: Connection reset by peer) (~pomel0@user/pomel0) | 14:35 | |
| + pomel0 (~pomel0@user/pomel0) | 14:36 | |
| - pomel0 (QUIT: Read error: Connection reset by peer) (~pomel0@user/pomel0) | 14:37 | |
| + pomel0 (~pomel0@user/pomel0) | 14:37 | |
| - paperManu_ (QUIT: Quit: WeeChat 4.7.2) (~paperManu@107.159.15.124) | 14:37 | |
| - jordi (QUIT: Ping timeout: 252 seconds) (~jordi@84.123.7.93.dyn.user.ono.com) | 14:38 | |
| minute | josch: where exactly is reform-desktop-full and reform-desktop-minimal built? i can't figure this out in reform-tools, which is specified as the source repo | 15:03 |
| minute | josch: i wanted to know if we still ship thunar, as we're editing the handbook | 15:03 |
| - pomel0 (QUIT: Ping timeout: 240 seconds) (~pomel0@user/pomel0) | 15:05 | |
| + pomel0 (~pomel0@user/pomel0) | 15:05 | |
| - pomel0 (QUIT: Ping timeout: 246 seconds) (~pomel0@user/pomel0) | 16:08 | |
| + paperManu_ (~paperManu@107.159.15.124) | 16:08 | |
| + pomel0 (~pomel0@user/pomel0) | 16:08 | |
| + jordi (~jordi@84.78.250.152) | 16:09 | |
| - jordi (QUIT: Read error: Connection reset by peer) (~jordi@84.78.250.152) | 16:39 | |
| + jordi (~jordi@84.78.250.152) | 16:39 | |
| + jordi__ (~jordi@84.78.250.152) | 16:44 | |
| - jordi (QUIT: Read error: Connection reset by peer) (~jordi@84.78.250.152) | 16:44 | |
| josch | minute: what ends up in the MNT debian repository comes from here: https://source.mnt.re/reform/reform-debian-packages/-/blob/main/reform-tools/debian/control?ref_type=heads | 16:47 |
| josch | (so yes, thunar is in there) | 16:47 |
| josch | minute: this will also tell you: apt-cache show reform-desktop-full reform-desktop-minimal | 16:48 |
| josch | but if you want to throw thunar out, that is the file you want to throw it out of :) | 16:52 |
| + nybble (~sprang@user/csprng) | 17:17 | |
| - pomel0 (QUIT: Ping timeout: 246 seconds) (~pomel0@user/pomel0) | 17:26 | |
| - jordi__ (QUIT: Ping timeout: 240 seconds) (~jordi@84.78.250.152) | 17:27 | |
| + spew (~spew@user/spew) | 17:32 | |
| + spew_ (~spew@user/spew) | 17:40 | |
| - spew (QUIT: Ping timeout: 255 seconds) (~spew@user/spew) | 17:40 | |
| * spew_ -> spew | 17:41 | |
| minute | josch: ah so weird that i couldn't find this with rgrep | 17:51 |
| minute | josch: ah that's because debian/control is not "upstreamed" to MNT :D could you do that, or is there a downside? | 17:52 |
| - gidzit (QUIT: Ping timeout: 264 seconds) (~gidzit@gidzit.org) | 17:52 | |
| minute | josch: (and i don't wanna throw thunar out yet, just wanted to know if it's still there) | 17:58 |
| - spew (QUIT: Read error: Connection reset by peer) (~spew@user/spew) | 18:10 | |
| josch | minute: it is not upstreamed? how much more upstream than being in reform-debian-packages is there? | 18:11 |
| josch | or do you mean debian by upstream? | 18:12 |
| josch | techincally, debian is upstream because reform-debian-packages:reform-tools/debian/control is overwriting what is in debian | 18:12 |
| minute | josch: ah sorry. i checked reform-tools m)) | 18:12 |
| + spew (~spew@user/spew) | 18:13 | |
| minute | minute@minute-i9:~/src/mref/reform-debian-packages/reform-tools$ rgrep thunar | 18:13 |
| minute | debian/control: thunar, | 18:13 |
| josch | this reminds me that it's still on my todo list to write down all of this... | 18:13 |
| minute | josch: got it! sorry for wasting your time | 18:13 |
| josch | no time wasted | 18:13 |
| josch | i take it as another motivation to sit down and write docs :D | 18:13 |
| + bkeys (~Thunderbi@172.56.67.146) | 18:31 | |
| - bkeys (QUIT: Read error: Connection reset by peer) (~Thunderbi@172.56.67.146) | 18:35 | |
| - spew (QUIT: Quit: nyaa~) (~spew@user/spew) | 18:36 | |
| + spew (~spew@user/spew) | 18:37 | |
| + bkeys (~Thunderbi@98.19.131.193) | 18:41 | |
| + bkeys1 (~Thunderbi@98.19.131.193) | 18:46 | |
| - bkeys (QUIT: Read error: Connection reset by peer) (~Thunderbi@98.19.131.193) | 18:47 | |
| * bkeys1 -> bkeys | 18:47 | |
| - bkeys (QUIT: Ping timeout: 240 seconds) (~Thunderbi@98.19.131.193) | 18:53 | |
| - colinsane (QUIT: Quit: bye) (~colinunin@97-113-73-113.tukw.qwest.net) | 18:56 | |
| + wielaard (~mjw@gnu.wildebeest.org) | 19:13 | |
| + colinsane (~colinunin@97-113-73-113.tukw.qwest.net) | 19:18 | |
| + pomel0 (~pomel0@user/pomel0) | 19:29 | |
| minute | writing from my MNT Reform Next again!!12 | 19:29 |
| - spew (QUIT: Quit: WeeChat 4.7.2) (~spew@user/spew) | 19:30 | |
| cwebber | minute: yay! | 19:35 |
| minute | it's getting pretty good | 19:47 |
| + aelius (~aelius@user/aelius) | 19:55 | |
| josch | i love that the software improvements that you have in the pipeline for the next (graphics during u-boot/barebox) and s2idle/suspend *also* automatically affect the existing platforms and make them better :) | 19:58 |
| jfred | agreed! <3 | 20:10 |
| minute | josch: for sure! | 20:13 |
| - mjw (QUIT: Killed (tantalum.libera.chat (Nickname regained by services))) (~mjw@2001:1c06:2486:a800:a09a:fc1c:5a8:e74d) | 20:23 | |
| * wielaard -> mjw | 20:23 | |
| + Guest2700 (~mjw@2001:1c06:2486:a800:a09a:fc1c:5a8:e74d) | 20:23 | |
| + Guest1056 (~user@p200300fc5703ea00645c142b5dbec2e7.dip0.t-ipconnect.de) | 21:30 | |
| wickedshell | I wanted to share that while I keep looking at the RCore updates for my pocket reform to speed up compile times (and maybe deal with cursed wifi) I've been developing an autopilot log analysis tool out of anger at the exsisting python tooling. My tool can be more responsive on the iMX8MP then the exsisting python tools are on my desktop 7950X (let alone the same tools on the pocket). | 22:09 |
| wickedshell | It's been a nice reminder of just how damn fast computers are, and that if you develop for it/and are aware the iMX8MP is *just* fine | 22:09 |
| minute | yeah a lot of stuff is unoptimized :D | 22:09 |
| wickedshell | The crazy thing to me is I can't say this is really opimized. It's a concern and a restriction, but even just doing the obvious thing can be done fast if you are tracking the full stack and avoid obvious bad ideas. | 22:11 |
| wickedshell | (I do want to play with more optimization, so far I've limited myself on the SIMD side to just playing with portable_simd in rust nightly, and even that is only lightly dropped in]) | 22:11 |
| wickedshell | Anyways, I just wanted to note that you can still get a lot of really nice performance out of it, and seeing it just work so damn well on the hardware makes me really happy, so thanks again for the experience enabling! | 22:14 |
| - gustav25 (QUIT: Quit: Quit) (~gustav@c-78-82-54-48.bbcust.telenor.se) | 22:15 | |
| minute | wickedshell: i'm very glad to hear it! | 22:17 |
| josch | wickedshell: i feel the rk3588 is too much power for the reform -- i mean i'm "dealing" with a311d as my main computer so i don't need much and rk3588 performance is nuts compared to a311d and i struggle to really "make use" of it in the small form factor of the pocket reform | 22:19 |
| josch | so usually i have the pocket reform with a 80 mm fan behind it so that rk3588 can compute things at full speed (like encode videos) | 22:20 |
| - pomel0 (QUIT: Read error: Connection reset by peer) (~pomel0@user/pomel0) | 22:26 | |
| josch | ucch MNT Reform Next Update on crowdsupply | 22:26 |
| + pomel0 (~pomel0@user/pomel0) | 22:26 | |
| josch | *uuuh even | 22:26 |
| - Guest1056 (QUIT: Ping timeout: 244 seconds) (~user@p200300fc5703ea00645c142b5dbec2e7.dip0.t-ipconnect.de) | 23:05 | |
| minute | oh :3 i didn't even read that one yet | 23:07 |
| minute | ah i see | 23:08 |
| wickedshell | josch: the only time I really want it is when I need to do large (re)compiles of Rust code bases. The incremental build is annoying, but not bad. But when I have to wait for a large dependency chain to recompile it's painful, that's the only thing I've really wanted it for. | 23:45 |
| josch | would be nice if rust had something like shared libraries | 23:47 |
| wickedshell | Yeah, althoughy I have liked cross compiling for some windows uses and just having a nice (mostly?) static exe that I've been able to distribute to users | 23:49 |
| wickedshell | Accesskit related dependencies are really some of the largest dependency trees for me, but I'm loathe to just turn them off. At the same time I've never actually used the accessibility tools that plug into it, so it might be a very broken dependency anyways... It's on my to solve list | 23:51 |
| wickedshell | (I'd like the features to work rather then just disabling them) | 23:51 |
Generated by irclog2html.py 2.17.3 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!