**** BEGIN LOGGING AT Mon Jul 25 02:59:56 2022 Jul 25 07:30:51 morning Jul 25 07:37:14 Morning! Jul 25 07:49:08 Morning! Jul 25 07:49:21 Tofe: You might be right about those tunes Jul 25 07:49:31 Just not so familiar with them Jul 25 07:51:05 well it's basically what cpu features we enable or not, so it's independent from halium or not Jul 25 08:00:11 OK will add it Jul 25 08:00:26 Tenderloin mainline kernel builds but doesn't boot Jul 25 08:00:28 at least not now when we append halium to TUNE_PKGARCH to distinguish them Jul 25 08:01:08 moboot seems happy with the uImage but it's quite large compared to our previous uImage so I guess it needs some work still Jul 25 08:05:02 JaMa/Tofe: https://github.com/webOS-ports/meta-webos-ports/commit/68adb12baac590f86249edace53012a8fd08d329 Jul 25 08:05:48 Herrie: lgtm Jul 25 08:06:08 Tofe: I'm really not sure how to get Tenderloin to build & boot something useful to be honest ;) Jul 25 08:06:25 But this builds and it produces an uImage, more I cannot tell about it Jul 25 08:06:35 I only flashed the uImage to /boot to test it Jul 25 08:06:53 Seems my serial cable doesn't work it seems somehow, so checking that now Jul 25 08:07:09 is it just the display ? or it really doesn't start anything ? Jul 25 08:07:50 you can do like me and replace the var/log symlink by a real directory, and create var/log/journal to make it persistent Jul 25 08:08:06 if after one boot you see content in var/log, then it's just the display Jul 25 08:08:10 Moboot showing the "usual" and then Booting... Jul 25 08:08:39 Let me first see if my serial wires are hooked up OK Jul 25 08:08:48 Could be they need to be in different order from PP(P) Jul 25 08:09:55 also, what defconfig are you using ? we probably want something slightly different for the display here Jul 25 08:10:14 and the serial console might not be active ? Jul 25 08:11:31 Tofe; Well I placed the moboot.sercon.LuneOS and moboot.sercon.webOS in /boot as per https://github.com/jcsullins/moboot/blob/fa7b8ea4d7d326df147c57e017df6deb36aabc0a/app/moboot/moboot.c#L534 Jul 25 08:11:44 And https://wiki.postmarketos.org/wiki/HP_TouchPad_(hp-tenderloin)#Serial_console Jul 25 08:13:30 Herrie: is there corresponding PR for meta-smartphone? Jul 25 08:14:16 JaMa: Yes https://github.com/shr-distribution/meta-smartphone/pull/145 Jul 25 08:15:37 Herrie: if it doesn't even work for legacy webOS, then it's the cable, somehow Jul 25 08:16:18 Tofe, seems I simply need to switch the TX and GND Jul 25 08:17:01 PPP has: Left audio to TXD, Right audio to RXD, Ground connection. Tenderloin has Tip: TX, Ring: RX, Sleeve: GND Jul 25 08:17:18 I now do get output in Putty but its all garbled for legacy webOS Jul 25 08:18:34 I'm not very good at the hardware tinkering, not sure how to tweak that Jul 25 08:19:34 Tofe, well it Jul 25 08:19:39 Tofe: Well it's easy ;) Jul 25 08:19:50 It are the left and right cable of my 3 pole connector Jul 25 08:19:54 I just switch it around :D Jul 25 08:20:04 It might help if I put 115200 instead of 115000 ;) Jul 25 08:20:25 Also, looking at your meta-smartphone PR, be careful with the init script in initrd; we have a very special one for tenderloin to activate LVM Jul 25 08:20:53 Tofe: Yeah, this all needs tweaking for sure Jul 25 08:21:00 This was merely to get things split for now Jul 25 08:21:13 It will need more work compared to Hammerhead still for sure Jul 25 08:21:18 sure Jul 25 08:21:30 I first would like to get something to build, then try to see if I can get some logs and take it from there Jul 25 08:21:37 Booting image will be way down the line I'm sure Jul 25 08:23:23 :) Jul 25 08:25:11 should I merge the PRs? :) Jul 25 08:28:22 I think it's ok, I found a little bug but nothing blocking for a merge Jul 25 08:30:35 Tofe: fixed it Jul 25 08:30:38 Force pushed Jul 25 08:30:43 Herrie: thanks! Jul 25 08:30:48 good to go then Jul 25 08:32:28 arf, JaMa is too fast :) Jul 25 08:32:54 He tends to do that ;) Jul 25 08:33:19 well, no big deal anything, we can add a little fix commit later Jul 25 08:33:24 anyway* Jul 25 08:35:23 My meta-{smartphone,webos-ports} layers are a mess, I'll just take upstream and rebuild hammerhead Jul 25 08:35:32 should be almost identical anyway Jul 25 08:36:37 strange I didn't get any notifaction from github meta-smartphone today (and it worked yesterday) Jul 25 08:37:31 JaMa: Yeah it does that sometimes... Jul 25 08:37:36 Since M$ bought them :P Jul 25 08:44:22 | install: cannot stat '/OE/build/luneos-langdale/webos-ports/tmp-glibc/work/tenderloin-webos-linux-gnueabi/systemd-machine-units/1.0-r19/tenderloin-ath6kl-module.service': No such file or directory Jul 25 08:45:45 will add that fix Jul 25 08:46:34 ah, yes, the SRC_URI thing... I didn't think of that consequence, sorry Jul 25 08:58:14 np Jul 25 10:04:58 Tofe: Made a new serial cable and this one works :D Jul 25 10:05:46 Used a 3 pin instead of 4 pin one Jul 25 10:05:55 4 pin one is OK for PP(P) but not for TP it seems Jul 25 10:07:27 And got a log: https://bpa.st/32IA Jul 25 10:09:13 At least gives clues :P Jul 25 10:41:51 well, yes, though still a bit vague ! Jul 25 10:43:07 Well I probably got something wrong with the DTB stuff Jul 25 10:50:08 If I add "CONFIG_APPENDED_DTB=y" to my defconfig it builds, without it I get: the following error about the devicetree: https://pastebin.ubuntu.com/p/ChFcG427yY/ Jul 25 10:50:19 We don't have CONFIG_APPENDED_DTB in hammerhead though Jul 25 10:51:48 https://github.com/Herrie82/linux/blob/6d76455e1dba2ef3be566326ce821abec0c709b2/tenderloin.mak#L186-L197 Jul 25 10:51:55 Might be something we need to replicate somehow Jul 25 10:51:57 Not sure really Jul 25 11:27:27 I asked some of the guys who toyed with mainline for their defconfigs and/or build scripts. Might be of use Jul 25 12:02:09 The one from Dragonboard might not be applicable to tenderloin as such Jul 25 12:03:56 Can't we just do like pmOS ? I didn't look at their recipe but there seemed to be some hints for uImage building Jul 25 12:05:30 Well Yocto does it for us somehow Jul 25 12:05:40 Well pmOS still builds the Android one ;) Jul 25 12:06:02 oh, ok, I didn't see :) Jul 25 12:06:15 At least that's my understanding Jul 25 12:07:56 "pkgdesc="HP TouchPad Evervolv kernel fork" pkgver=3.4.113 Jul 25 12:07:57 So yes Jul 25 12:08:08 So basically our kernel Jul 25 12:08:14 "pkgdesc="HP TouchPad Evervolv kernel fork" " you're correct Jul 25 12:08:47 ok, so uncharted waters it is, then Jul 25 12:09:15 Yes seems so, so if mbd98 or lopsided98 can provide some feedback that would be good ;) Jul 25 12:09:21 Saves reinventing the wheel there Jul 25 12:10:34 Seems lopsided98 is a bit busy, but he replied to my email during the weekend that he'll be sending a v2 of the dtb kernel patch "in the next months" Jul 25 12:10:43 I might have to get stuff setup myself to submit it ;) Jul 25 12:14:13 using CONFIG_APPENDED_DTB should be the way to go, if the bootloader doesn't know about dtb loading Jul 25 12:14:43 (see for instance the refusal of https://patchwork.kernel.org/project/linux-arm-kernel/patch/1433410504-11628-1-git-send-email-yangyingliang@huawei.com/ ) Jul 25 12:15:52 It's possible that fastboot images are able to handle DTB, though I agree it's a bit unlikely for hammerhead... Jul 25 12:16:14 Tofe: Well I really don't know.. Hammerhead builds zImage, we build uImage Jul 25 12:16:41 There could be some different handling Jul 25 12:17:08 It seems that for dragonboard the cpio.gz ramdisk gets picked up by the defconfig, we do a separate step there still in the kernel recipe Jul 25 12:17:53 I.e. the defconfig specifies a "CONFIG_INITRAMFS_SOURCE" Jul 25 12:18:08 I removed that, but can put CONFIG_INITRAMFS_SOURCE="initramfs-android-image-tenderloin.cpio.gz" Jul 25 12:18:14 https://github.com/shr-distribution/meta-smartphone/blob/master/meta-android/classes/kernel_android.bbclass#L22= see here for hammerhead Jul 25 12:18:24 But that should not solve the DTB issue Jul 25 12:20:44 I mean this bit: https://github.com/shr-distribution/meta-smartphone/blob/kirkstone/meta-hp/recipes-kernel/linux/linux-hp-tenderloin_git.bb#L46-L64 Jul 25 12:20:53 I think that could be done by the defconfig flag as well? Jul 25 12:21:31 maybe not for a uImage, but not sure Jul 25 12:22:03 There are some bits in kernel.bbclass Jul 25 12:23:36 https://github.com/openembedded/openembedded-core/blob/master/meta/classes/kernel.bbclass#L364-L380 and https://github.com/openembedded/openembedded-core/blob/master/meta/classes/kernel.bbclass#L304-L321 Jul 25 12:26:48 Herrie: mmh but what's the issue with initrd ? I thought the issue was with DTB Jul 25 12:27:05 Tofe: Yeah issue is DTB ;) Jul 25 12:27:13 But it could also be my cmdline or anything else really Jul 25 12:30:56 I think it's possible to append also the DTB in the multi-part main uImage Jul 25 12:31:52 something like just appending the DTB at the end here https://github.com/shr-distribution/meta-smartphone/blob/kirkstone/meta-hp/recipes-kernel/linux/linux-hp-tenderloin_git.bb#L61= Jul 25 12:32:24 (but maybe the dtb must be packed into a dedicated uboot image first... I don't know Jul 25 12:32:37 I'm also not sure really Jul 25 12:32:50 I'm examining my uImage with a hex editor, but I don't seem to see the dtb contents in there Jul 25 12:32:56 uImage old format is a bit like partitioning Jul 25 12:33:07 But it could be of course compressed or something Jul 25 12:33:26 there is a main "multi-part" uImage file, which we feed to moboot; it is composed of "mini" uimages, for the kernel and for the initramf Jul 25 12:33:28 s Jul 25 12:34:50 Herrie: would you have the output of "uboot-mkimage -h" ? just to get the list of supported uImage formats Jul 25 12:37:37 Tofe: Let me see Jul 25 12:38:08 https://stackoverflow.com/questions/70725653/add-dtb-to-u-boot-uimage-during-yocto-build here the guy just uses the defconfig option plus a basic "cat" Jul 25 12:43:11 Herrie: also regarding your error with APPEND_DTB: it's weird that the target is ".dts" and not ".dtb", maybe a typo in the recipe or so ? Jul 25 12:43:25 Well there seems to be a uImage-dtb file Jul 25 12:44:15 In the in between process Jul 25 12:44:57 Where the uImage-dtb has the dtb appended Jul 25 12:45:21 From log,do_deploy: reading kernel from /mnt/5ba5d474-0b2d-49d6-a5a6-9de20c3ac967/kirkstone/webos-ports/tmp-glibc/work/tenderloin-webos-linux-gnueabi/linux-hp-tenderloin/5.19-rc7+gitrAUTOINC+f55df88ad1_6d76455e1d-r0/linux-tenderloin-standard-build/arch/arm/boot/uImage-dtb Jul 25 12:45:38 I verified the contents and it has the DTB appended at the end of the uImage Jul 25 12:45:52 The size difference is approximately the DTB size Jul 25 12:45:55 ah, weird that the kernel build complains then Jul 25 12:46:45 Well it's all moving here a bit... Jul 25 12:46:51 So I might just clean up my build first Jul 25 12:47:44 Do you also define https://github.com/webOS-ports/meta-pine64-luneos/blob/kirkstone/conf/machine/pinephone.conf#L48= ? not sure when it's needed, for pinephone the dtb is kept separate Jul 25 12:48:43 I added that in my local build, but seems not in meta-smartphone yet Jul 25 12:49:00 Well for hammerhead we also define it in machine.conf Jul 25 12:49:33 https://github.com/shr-distribution/meta-smartphone/blob/kirkstone/meta-lg/conf/machine/hammerhead.conf#L48 Jul 25 13:11:18 Adding the KERNEL_DEVICETREE = "qcom-apq8060-tenderloin.dtb" creates the uImage-dtb in the work directory Jul 25 13:11:28 Without this, there is no uImage-dtb, only uImage Jul 25 13:12:05 I don't need the "CONFIG_APPENDED_DTB=y" it seems though now Jul 25 13:13:55 So seems something in my tree was funky Jul 25 13:14:05 I will retry the new uImage later today when home Jul 25 13:26:56 OK PR-ed this to fix the DTB issue at least: https://github.com/shr-distribution/meta-smartphone/pull/146 Jul 25 13:34:44 Tofe: Another issue might be KERNEL_EXTRA_ARGS += "LOADADDR=${UBOOT_ENTRYPOINT}" Jul 25 13:34:51 I added this, because without this it wouldn't build Jul 25 13:34:57 It could be the value there is not correct Jul 25 13:35:15 I will get our current uImage and check what it does there and outputs as values on screen Jul 25 13:35:16 ah, I'm not sure what this should be :/ Jul 25 13:35:17 And compare Jul 25 13:37:54 "If your ARM system's kernel is configured via the new (3.7+) Jul 25 13:37:55 multi-architecture support (CONFIG_ARCH_MULTIPLATFORM=y in your linux Jul 25 13:37:55 kernel .config), then it is necessary to specify a kernel load address Jul 25 13:37:55 for the uImage. This should be a hexadecimal string beginning with Jul 25 13:37:55 0x. Example setting: 0x00008000." Jul 25 13:38:11 Seems it's set to y for tenderloin, hence it has this requirement Jul 25 13:38:41 mmh okay Jul 25 13:38:57 Some discussion on that here: https://patchwork.ozlabs.org/project/buildroot/patch/1362998134-4614-1-git-send-email-julien.boibessot@free.fr/ Jul 25 13:39:05 I can remove that flag and this value from my machine.conf Jul 25 13:39:08 See what it does Jul 25 13:42:13 Seems it's needed. | multiple (or no) load addresses: | This is incompatible with uImages | Specify LOADADDR on the commandline to build an uImage Jul 25 13:44:44 but you already have UBOOT_LOADADDRESS = "41000000" right ? should be the correct value Jul 25 13:45:51 Yeah it should be Jul 25 13:45:55 I will double check when home Jul 25 13:46:12 Getting our Halium uImage and the mainline one and see what I get on screen in MoBoot Jul 25 13:47:29 Just to see if there's anything obvious there to spot Jul 25 13:47:38 The addresses should be roughly the same Jul 25 13:49:44 The uImage.webOS gives me 0x40200000 while uImage.LuneOS gives me 0x41000000 Jul 25 13:49:48 So that's not correct I guess? Jul 25 13:50:19 This is with my mainline one Jul 25 13:50:24 I need to see what the halium one shows Jul 25 15:00:19 uImage.LuneOS for halium gives me 0x40208000 Jul 25 15:37:19 Some useful feedback via mail received :D Jul 25 15:47:20 Herrie: yes, though I'm not at ease with dealing with these clocks thing Jul 25 15:47:49 but maybe we can get the display up and running, that would be a start Jul 25 15:48:16 after we get a working kernel ofc :p Jul 25 15:56:16 Tofe: Well we have the schematics and PP(P) trees to work with Jul 25 15:56:23 So that should help Jul 25 15:57:16 right Jul 25 16:57:03 Tofe: with my new uImage the same error message Jul 25 17:58:48 Tofe: https://tinypic.host/i/defconfig-diff-values.jeXMS Jul 25 17:59:36 These config values exists in both, but differ Jul 25 17:59:42 The ones in red I guess need some tweaking Jul 25 17:59:54 my new hammerhead build is now ready Jul 25 17:59:54 CONFIG_PAGE_OFFSET might be the most important one Jul 25 18:00:56 Herrie: mmh... I'd like to help you hands on with the TP, but I don't think I have an adapted serial cable Jul 25 18:01:33 Tofe: OK ;) Jul 25 18:01:47 I just have one jack for PP, and a GPIO one also for PP devboard Jul 25 18:01:48 I'm not sure which cable you have ? Jul 25 18:02:13 I just have a simple CH340 USB dongle, it's 3.3V but TP should need 2.9V, but seems to work Jul 25 18:02:26 Just connected an old headset jack cable to it ;) Jul 25 18:02:33 The PPP one might just work too Jul 25 18:02:41 Just set it to 115200 as speed :P Jul 25 18:02:54 I have that one https://pine64.com/product/pinebook-pinephone-pinetab-serial-console/ Jul 25 18:03:31 That looks similar to mine Jul 25 18:03:34 For PPP Jul 25 18:03:40 I just have a home made version Jul 25 18:03:51 Because the Pine one was sold out or too expensoive Jul 25 18:04:16 I think shipping was more vs the dongle Jul 25 18:04:19 and something like this: https://www.ebay.fr/itm/222528693981 Jul 25 18:04:30 So just ordered some of the CH340 and I had like 2 for EUR 7,50 ;) Jul 25 18:05:19 Ah that's a 5V one Jul 25 18:06:13 do you still have the reference for yours? Jul 25 18:06:25 can't find one with a jack input Jul 25 18:07:08 I have one like this: https://www.tinytronics.nl/shop/nl/communicatie-en-signalen/serieel/usb/ch340-3.3v-5v-ttl-usb-serial-port-adapter Jul 25 18:07:24 Just soldered the cables onto a connector, not pretty but works Jul 25 18:07:49 oh, ok, I don't have a solering set here :p Jul 25 18:07:56 soldering* Jul 25 18:08:09 :D Jul 25 18:08:15 Well you have jacks with screws it seems :P Jul 25 18:08:36 that's more me, yes :) Jul 25 18:08:37 https://www.onlinecamerashop.nl/eng/jack-3-5mm-audio-plug-screw-connection-to-earth.html Jul 25 18:08:45 Something like that should do the trick Jul 25 18:09:35 would the pinephone one work as-is ? or will I also need to invert two cables ? Jul 25 18:10:18 TXD = green wire (Left), RXD = red wire (Right), GND = copper wire. Jul 25 18:10:23 Tofe: You can simply try it Jul 25 18:10:53 Add the moboot.sercon.webOS and moboot.sercon.LuneOS files to /boot where it contains "yes" in the files Jul 25 18:11:21 ah yes, I forgot Jul 25 18:11:31 Then just try Jul 25 18:11:32 ok booting webOS and installing the files Jul 25 18:11:45 If you get garbage, cable is no good :P Jul 25 18:12:13 Just make sure to set the speed to 115200, since PPP uses 1500000 Jul 25 18:14:17 no output :( Jul 25 18:15:15 Ah then they would need to be switched I guess Jul 25 18:15:18 I had that too Jul 25 18:15:26 Then I switched wires and I had garbled output Jul 25 18:15:32 Then made a new one and it was OK Jul 25 18:18:06 the other adapter I have can do both 3.3 and 5V Jul 25 18:18:49 I just need a jack where I can plug these gpio things Jul 25 18:18:50 Tofe: Ah Jul 25 18:19:56 mmh it's not the right technical word Jul 25 18:20:24 Well if you have an old jack plug or cable, you can just "stick the wires in" Jul 25 18:20:28 Not great but should work Jul 25 18:20:51 I have a box with old goodies for this reason Jul 25 18:20:55 Useful on days like this Jul 25 18:20:57 let me dig in my cable treasure cache Jul 25 18:22:11 You just need to correct jack (one without microphone normally) Jul 25 18:22:22 That one has an extra ring that interferes it seems Jul 25 18:23:48 One like this: https://www.allekabels.nl/jack-stekker/42/989/jack-stekker-35mm.html Jul 25 18:25:14 yep Jul 25 18:45:19 Any luck finding the cable? Jul 25 19:09:31 well, I've done some ugly engenieering, let's see now https://imgur.com/oZfwQnN Jul 25 19:10:50 no output, let me switch wires Jul 25 19:15:07 got it! Jul 25 19:15:39 and it doesn't even catch fire Jul 25 19:31:54 Tofe: Yeah Jul 25 19:32:04 Mine is soldered but not much prettier Jul 25 19:32:28 Just put shrink wrap around it so it's "locked" Jul 25 19:34:13 https://photos.app.goo.gl/9HUTDwCqw8qwY9eN7 Jul 25 19:34:57 And my PP(P) one: https://photos.app.goo.gl/zv4FJZ7Hcy1TH9Cr7 Jul 25 20:09:42 ok, building tenderloin's kernel here Jul 25 20:35:39 Tofe: from mbd98: https://gist.github.com/mbd98/dc973ade2746b5589ab19869d2b31ca8 Jul 25 20:36:05 Some hopefully useful defconfig Jul 25 20:36:12 CONFIG_ARM_APPENDED_DTB=y ;) Jul 25 20:37:25 Tofe: This one we had already Jul 25 20:38:26 ah ok Jul 25 20:39:01 I dusted off my shortloin as well :P Jul 25 20:39:28 But there moboot is the challenge already, so will check that later Jul 25 20:39:44 Herrie: can you repaste or push somewhere your current kernel recipe ? Jul 25 20:40:01 my kernel build output is a fastboot image, so not good I guess Jul 25 20:40:51 Recipe: https://bpa.st/V2UQ Jul 25 20:41:17 And the machine.conf: https://bpa.st/MI3A Jul 25 20:41:22 Should be in the repos though Jul 25 20:46:19 Tofe: FWIW the differences between the defconfig for tenderloin and shortloin are about 15. So they're very close ;) Jul 25 20:48:10 ok, well, it means we're close :p Jul 25 20:48:23 did he also provide his build script or steps ? Jul 25 20:49:06 Tofe: No.... Jul 25 20:50:34 Tofe: Well it means that when we get tenderloin working shortloin should be fairly easy too Jul 25 20:50:45 As expected, since it's basically the mini version of the tenderloin Jul 25 20:50:56 Main difference is the display used it seems Jul 25 20:51:04 And some other NFC/BT chip Jul 25 20:51:08 At least based on defconfig Jul 25 20:55:48 But shortloin is really for later Jul 25 20:55:55 Let's do something with tenderloin first Jul 25 20:58:18 "I don't fully recall how well it worked, but hopefully it's an okay starting point!" Jul 25 21:03:57 Looking at his defconfig vs the dragonboard one, I suggest we use this one to test Jul 25 21:13:47 Getting this with mdb's defconfig: https://bpa.st/LCBQ Jul 25 21:14:49 well I'm going to try my own little uImage first Jul 25 21:15:23 ( https://bpa.st/LLRA ) Jul 25 21:16:14 Did you change something? Jul 25 21:16:39 the packing, mainly Jul 25 21:17:07 OK I'll build here too Jul 25 21:19:04 arf... "Checking uImage invalid Magic" Jul 25 21:26:25 and now "Invalid Size"...meh Jul 25 21:34:03 I think I understand why my uImage are incorrect Jul 25 21:36:20 currently we create a uImage directly for the kernel; therefore, it's not possible to modify it afterwards. But we'd like to add a DTB at the end of the linux image Jul 25 21:36:44 so I think we have to create a simple Image, add DTB, then create a uboot image out of that manually Jul 25 21:43:01 Tofe: In bootie mode you will also have the possibility to inspect stuff Jul 25 21:48:59 https://www.webos-internals.org/wiki/Bootie#Running_Bootie_commands Jul 25 22:01:47 ok, well, I'll retry tomorrow Jul 25 22:02:03 Couldn't find a uboot packing that was correct Jul 25 22:02:05 Tofe: Sleep & coffee do miracles Jul 25 22:02:10 yeah :) Jul 25 22:02:10 :D Jul 25 22:02:24 Good at lesat you have a cable and some leads :D Jul 25 22:02:44 yes, that's nice Jul 25 22:02:57 and not too difficult to test, just a lot of TP reboots Jul 25 22:03:29 Tofe: Yeah :P Jul 25 22:03:42 Here got my build environment for moboot working Jul 25 22:03:46 Can make custom moboots now :P Jul 25 22:04:21 ooh nice :) Jul 25 22:04:42 So I can hack shortloin in Jul 25 22:04:52 Seems it just needs a table with models added mainly Jul 25 22:05:09 GPIO pins for 3G models are differently setup Jul 25 22:05:18 moboot only has a model table for topaz, but not for opal Jul 25 22:05:43 I extracted lists of both board models (31 and 32 different ones) from the original uImages ;) Jul 25 22:06:40 :) Jul 25 22:07:21 DVT, EVT, PVT etc Jul 25 22:08:32 OK my own built moboot works :D **** ENDING LOGGING AT Tue Jul 26 02:59:56 2022