**** BEGIN LOGGING AT Tue Apr 13 02:59:56 2021 Apr 13 11:22:25 AP_XD was added by: AP_XD Apr 13 14:07:45 hello Apr 13 14:08:06 i am in a weird situationn for halium, my device goes into some weird power reset mode Apr 13 14:08:19 but there is something in /tmp/recovery.log is that useful? Apr 13 14:11:43 any clues anyone? device is samsung galaxy tab 2 Apr 13 14:12:09 @mighty17 [but there is something in /tmp/recovery.log is that useful?], Might be Apr 13 14:12:15 what do you get? Apr 13 14:16:15 nah thats twrp stuff Apr 13 14:16:19 didnt help Apr 13 14:32:57 @mighty17 [didnt help], Does it get stuck in a bootloop? Apr 13 14:37:43 I think you should first compile hybris-recovery Apr 13 14:40:12 hybris-recovery? Apr 13 14:40:19 @mighty17 [didn"> no it goes into power reset mode Apr 13 14:46:41 @mighty17> @mighty17 [didn" [no it goes into power reset mode], Do you still have your build directory setup? Apr 13 14:47:00 yeah Apr 13 14:47:19 @mighty17 [yeah], Breakfast your device and then run `mka hybris-recovery` Apr 13 14:47:59 yeah did that Apr 13 14:48:09 it should build an image called hybris-recovery.img Apr 13 14:48:17 Now flash that with heimdall Apr 13 14:48:23 and then go to recovery Apr 13 14:48:26 flash it? Apr 13 14:48:38 as in flash to boot or recovery? Apr 13 14:48:45 recovery Apr 13 14:48:54 flash to recovery Apr 13 14:49:12 uhhh im scared Apr 13 14:49:21 @mighty17 [uhhh im scared], don't worry Apr 13 14:49:25 as recovery is the only way to come back Apr 13 14:49:29 you can always go back to download mode Apr 13 14:49:38 don't you have download mode? Apr 13 14:49:42 it's a samsung Apr 13 14:49:46 yeah i do Apr 13 14:49:49 lel Apr 13 14:49:57 you can always go back Apr 13 14:50:01 nothing to lose Apr 13 14:50:08 hybirs boot works, as in i can ssh into it Apr 13 14:50:11 but halium boot doesnt Apr 13 14:50:25 i mean i can port sfos as well to it and that worked Apr 13 14:50:27 did you do the kernel edits? Apr 13 14:50:37 as halium-boot never boots for me Apr 13 14:50:43 nope Apr 13 14:50:49 kernel edits? Apr 13 14:51:02 @mighty17 [nope], Well that's self-explanatory Apr 13 14:51:25 do u mean the config or the kernel files Apr 13 14:51:33 @mighty17 [do u mean the config or the kernel files], the config script Apr 13 14:51:38 i have changed the configs as halium wiki says Apr 13 14:51:50 `./halium/halium-boot/check-kernel-config` Apr 13 14:51:57 Did you run this? Apr 13 14:52:08 yup Apr 13 14:52:21 did you set console to /dev/tty0? Apr 13 14:52:31 yup Apr 13 14:52:50 then I don't have any other ideas Apr 13 14:53:01 It also doesn't work for me Apr 13 14:53:10 though hybris-boot works Apr 13 14:53:39 Did you add the samsung fix tho? Apr 13 14:53:46 do u also have a samsung? Apr 13 14:53:59 @mighty17 [yup], yes Apr 13 14:54:00 TheVancedGamer also what to do after hybris-recovery? Apr 13 14:54:25 @mighty17 [TheVancedGamer also what to do after hybris-recovery?], Just flash it with heimdal Apr 13 14:54:45 @mighty17 [do u also have a samsung?], a lot of samsung's actually Apr 13 14:54:48 yeah done Apr 13 14:55:00 now force reboot to recovery Apr 13 14:55:10 i am in recovery Apr 13 14:55:12 oh wait Apr 13 14:55:54 well its a blank screen Apr 13 14:55:54 what? Apr 13 14:56:00 @mighty17 [well its a blank screen], Again? Apr 13 14:56:08 keep holding the button combo Apr 13 14:56:12 for hybris-recovery yes Apr 13 14:56:26 @mighty17 [for hybris-recovery yes], Also did you flash the rootfs? Apr 13 14:56:36 na Apr 13 14:56:43 oh derp Apr 13 14:56:55 i can atleast get dmesg now Apr 13 14:57:37 and dmesg is saying bad news Apr 13 14:57:46 @mighty17 [and dmesg is saying bad news], oh buddy Apr 13 14:57:54 https://pastebin.ubuntu.com/p/S9T6zX684G/ Apr 13 14:58:19 * mighty17 < https://matrix.org/_matrix/media/r0/download/matrix.org/LmXcOKoBCBIWjcvtuqcJuzFg/message.txt > Apr 13 14:58:22 fun Apr 13 14:59:40 yay Apr 13 14:59:45 you have an OMAP chip? Apr 13 15:00:08 @mighty17 [fun], Hello? Apr 13 15:00:15 but what about halium boot? i mean i cannot get ubuntu touch without that right? Apr 13 15:00:21 you have an OMA"> yup Apr 13 15:00:23 omap4430 Apr 13 15:01:34 (Photo, 391x84) https://irc.ubports.com/840LLkaE.png bad signs Apr 13 15:01:58 indeed Apr 13 15:02:18 the kernel seems to be fine as it runs on android Apr 13 15:02:25 yes Apr 13 15:02:32 what is your exact model number? Apr 13 15:02:53 p3100 Apr 13 15:03:01 tab s2? Apr 13 15:03:29 tab 2 Apr 13 15:03:37 ok Apr 13 15:03:39 not tab s2 Apr 13 15:03:43 its also called espresso3g Apr 13 15:04:04 https://github.com/Unlegacy-Android/android_kernel_ti_omap4/tree/3.4/common/ Apr 13 15:05:06 i should use 3.0 kernel? and LineageOS sources? Apr 13 15:05:51 @mighty17 [i should use 3.0 kernel? and LineageOS sources?], What device and vendor tree? Apr 13 15:06:05 https://github.com/LineageOS/android_kernel_samsung_espresso10 Apr 13 15:06:22 https://github.com/LineageOS/android_device_samsung_espresso3g Apr 13 15:07:17 It's RIL as well? Apr 13 15:07:22 Like sim supported? Apr 13 15:07:23 yeah Apr 13 15:07:28 hm Apr 13 15:07:42 but RIL doesnt matter as much Apr 13 15:08:46 right of the bat there is a huge difference in files to cm-13.0 device tree Apr 13 15:08:47 * https://github.com/LineageOS/android_kernel_ti_omap4 Apr 13 15:09:08 cm13? halium doesnt support android 6 :( Apr 13 15:09:14 There's wayyy too less files Apr 13 15:09:58 what tree u looking at? Apr 13 15:10:09 the old android 6 trees Apr 13 15:10:17 the old trees are ported to new trees Apr 13 15:10:21 hmm Apr 13 15:10:26 drivers are missing Apr 13 15:10:42 So it can't initialize screen Apr 13 15:10:52 the old android"> or wait, i have latest sources Apr 13 15:11:07 https://github.com/Unlegacy-Android/android_kernel_ti_omap4/tree/3.4/common/ Apr 13 15:12:49 i use these sources as they are newer than LineageOS @TheVancedGamer Apr 13 15:12:57 i see Apr 13 15:13:21 they have everything Apr 13 15:13:26 from device tree to vendor files Apr 13 15:13:30 run this Apr 13 15:13:54 first cd into device dir Apr 13 15:14:13 like device/samsung Apr 13 15:14:27 and then run ls Apr 13 15:14:32 and tell what you see Apr 13 15:14:53 i know linux xD Apr 13 15:15:01 i know Apr 13 15:15:11 what's the folder name? Apr 13 15:15:12 `espresso espresso3g` Apr 13 15:15:22 so there's two right? Apr 13 15:15:29 yo[ Apr 13 15:15:45 yup* Apr 13 15:15:48 nuke both folders Apr 13 15:16:00 uh Apr 13 15:16:29 ohk Apr 13 15:16:39 now what sources do u want me to use Apr 13 15:16:44 now do: Apr 13 15:16:55 espresso3g device tree is small as it depends on espresso device tree* Apr 13 15:17:20 `git clone https://github.com/Chirag-Galani/android_device_samsung_p3100 -b kitkat expresso3g Apr 13 15:17:25 just a test Apr 13 15:17:35 kitkat????????????? Apr 13 15:17:48 yes Apr 13 15:17:57 done Apr 13 15:18:03 cloned? Apr 13 15:18:13 yup Apr 13 15:18:25 now rebuild, reflash and retest Apr 13 15:18:36 but now lunch or breakfast? Apr 13 15:19:28 none of those work Apr 13 15:21:13 What does it say? Apr 13 15:21:52 I remembered something Apr 13 15:21:56 * mighty17 < https://matrix.org/_matrix/media/r0/download/matrix.org/JqZSVvZpJzNApsMGZIeZJVnk/message.txt > Apr 13 15:22:17 Go to device dir Apr 13 15:22:26 Inside the expresso3g folder Apr 13 15:22:36 And change every cm to lineage Apr 13 15:22:38 kk Apr 13 15:22:56 kk Apr 13 15:23:07 Then change every cm to lineage in the Board Config.mk Apr 13 15:23:32 umm dude Apr 13 15:23:43 u realise that i will need to use a different kernel for this right Apr 13 15:23:51 @mighty17 [u realise that i will need to use a different kernel for this right], Just trying Apr 13 15:23:56 You might need to Apr 13 15:23:59 and the 4.4 device wont work Apr 13 15:24:00 tree* Apr 13 15:24:04 But just testing Apr 13 15:24:19 we need this https://github.com/Unlegacy-Android/android_device_samsung_espresso3g/blob/aosp-6.0/vendorsetup.sh Apr 13 15:24:27 for lunch/breakfast to work Apr 13 15:24:50 You can try then Apr 13 15:25:04 But it somehow doesn't recognize the lcd device Apr 13 15:25:15 Drivers aren't there Apr 13 15:25:23 uhh Apr 13 15:25:30 so i will try 3.0 lineage kernel then Apr 13 15:25:58 @mighty17 [so i will try 3.0 lineage kernel then], 3.0 has loads of bugs Apr 13 15:26:19 there is 3.4 then for it? Apr 13 15:26:20 Just retry again Apr 13 15:26:22 no other kernel available Apr 13 15:26:28 @mighty17 [no other kernel available], No other?? Apr 13 15:26:35 I just saw 3.4 right now Apr 13 15:26:36 @mighty17 [so i"> 3.0.101? i mean thats the eol Apr 13 15:26:46 @mighty17 [no o"> nope all kernels are 3.0.x and 3.4.x Apr 13 15:26:53 It is suppprted Apr 13 15:26:57 [Edit] It is suppported Apr 13 15:27:01 But buggsss Apr 13 15:27:34 but i have no other option Apr 13 15:27:46 I just saw 3.4 for it Apr 13 15:28:06 i was using 3.4 this whole time dude Apr 13 15:28:11 ... Apr 13 15:28:28 Hm Apr 13 15:28:33 Try 3.0 then Apr 13 15:28:37 Maybe you'll be lucky Apr 13 15:28:41 Waaiiitt Apr 13 15:28:49 Flash any rom on your tab Apr 13 15:28:52 Wuick Apr 13 15:28:56 [Edit] Quick Apr 13 15:29:06 ok Apr 13 15:29:07 We can get lcd driver from it Apr 13 15:29:08 wait Apr 13 15:29:16 cant twrp give it anyways Apr 13 15:29:28 @mighty17 [cant twrp give it anyways], If you've wiped system, no Apr 13 15:30:02 sadness Apr 13 15:30:17 i will try 3.0 kernel Apr 13 15:30:25 frick Apr 13 15:30:53 @mighty17 [frick], What? Apr 13 15:31:04 mount -o bind '/data/media/0' 'sdcard' process ended with error 255 Apr 13 15:31:43 @mighty17 [mount -o bind '/data/media/0' 'sdcard' process ended with error 255], You have data on there? Apr 13 15:31:51 yeah some roms Apr 13 15:31:52 Like important data on the tab? Apr 13 15:32:01 @mighty17 [yeah some roms], It's gone now Apr 13 15:32:07 i mean i can nuke it but i would like a backup Apr 13 15:32:17 no it isnt Apr 13 15:32:22 @mighty17 [i mean i can nuke it but i would like a backup], Well try again Apr 13 15:32:23 twrp still shows them Apr 13 15:32:44 @mighty17 [twrp still shows them], It's having issue bind-mounting stuff Apr 13 15:32:51 nuked it Apr 13 15:32:56 @mighty17 [nuked it], Wow Apr 13 15:33:04 had backup on pc Apr 13 15:33:08 Anywho then Apr 13 15:33:29 will try 3.0 and give the results Apr 13 15:33:34 hybris-recovery right? Apr 13 15:34:17 @mighty17 [hybris-recovery right?], Yess Apr 13 15:34:20 Actually Apr 13 15:34:30 I might do a combined device tree for you Apr 13 15:34:36 ?? Apr 13 15:34:43 Like the expresso and the expresso3g device trees together Apr 13 15:35:00 Because it doesn't look like the expresso3g tree is calling the expresso one Apr 13 15:35:28 I mean that isn't necessary Apr 13 15:35:46 Because it does"> It does, I have tried with just espresso3g tree and it fails Apr 13 15:36:09 @mighty17> Because it does" [It does, I have tried with just espresso3g tree and it fails], Hmm Apr 13 15:36:14 That's weird Apr 13 15:36:22 Just try with expresso then Apr 13 15:36:28 Don't use expresso3g Apr 13 15:36:53 Hmkay also will be using LineageOS sources from now just so it's clear Apr 13 15:38:30 @mighty17 [Hmkay also will be using LineageOS sources from now just so it's clear], Oki Apr 13 15:39:10 Thanks for the help 🙂 Apr 13 21:39:40 Hey @erfanoabdi or @NotKit, got any devices with the qcacld3.0 driver? Apr 13 21:39:51 How do they wake from suspend on a packet? Apr 13 21:40:21 Or... Do the devices that contain them just never go to sleep? Apr 13 21:40:29 Seems like the radios disassociate from WiFi whenever the host device decides to sleep. Apr 13 21:41:23 I don't know about qcacld3.0, but on MTK it is needed to tell the driver we're going to suspend Apr 13 21:41:31 How? Apr 13 21:41:36 (which we don't implement properly yet) Apr 13 21:41:50 Hrm. How would it be done? Apr 13 21:42:00 https://github.com/mer-hybris/sailfish-connman-plugin-suspend/blob/master/src/suspendplugin.c Apr 13 21:42:27 I started to hack around https://github.com/ubports/repowerd/commit/183f5a43760ff1032c64248ed747b34d8d1718c1, but didn't get it right Apr 13 21:42:29 Does that cause the device to wake on packets? Apr 13 21:42:40 Or does it just prevent dissociation during suspend? Apr 13 21:42:42 at least that allows it to suspend Apr 13 21:42:51 currently Volla doesn't suspend properly Apr 13 21:43:07 it just has large enough battery Apr 13 21:43:20 Hmm. Does that seem common on new devices? Apr 13 21:43:33 I notice that Active Display is pretty popular on Android devices these days. Apr 13 21:43:42 Is suspend just not fashionable any more? Apr 13 21:43:46 no, devices still go to suspend on Android Apr 13 21:44:04 Do they do it with active display? Apr 13 21:44:12 no idea about that Apr 13 21:44:20 hrm Apr 13 21:44:32 How could we find out more? Apr 13 21:44:41 Where would you look, rather? Apr 13 21:45:52 Android userspace Wi-Fi HAL probably does something on suspend Apr 13 21:47:37 @phhusson maybe you have an idea about how Qualcomm does it? Apr 13 21:50:41 @UniversalSuperBox [Do they do it with active display?], active display is only on devices with OLED i guess? so 99.99% of the screen is "off" still Apr 13 21:51:08 But which parts of the SOC are on? Apr 13 21:51:17 This is puzzle. Apr 13 21:51:48 still minimal i think Apr 13 21:52:11 but i don't know technical details of qcom SoCs Apr 13 21:52:18 it doesnt need much to update a clock Apr 13 21:52:37 especially not 3ghx octa-core Apr 13 21:52:40 No, but what is more important is how does it enter that state and what does that mean Apr 13 21:52:44 [Edit] especially not 3ghz octa-core Apr 13 21:53:17 Ubuntu Touch on android9 devices is woefully inadequate at power saving. How do we do better without sacrificing its experience? This is one piece. Apr 13 21:56:14 @UniversalSuperBox https://github.com/mer-hybris/sailfish-connman-plugin-suspend/commit/6e9d866b2fa94e101ffca7084b8230e413a048cf Apr 13 21:56:40 I think it was needed on Pro1 SFOS port too Apr 13 21:59:27 @NotKit [Android userspace Wi-Fi HAL probably does something on suspend], Uh maybe it's the power hal that does stuff like that (you don't use it right?) Apr 13 22:00:14 It's possible, but I don't think so. Erfan has created a power HAL binding for repowerd for suspend, but it didn't seem to fiddle with it. Apr 13 22:02:31 Unless there's something that ties everything together and says "WLAN HAL, we're going to sleep; Display HAL, we're going to sleep"... Apr 13 22:04:32 Or some other hook system that we can look at to find the secret sauce. It's just gotta be one sysfs write away... Apr 13 22:05:24 https://gitlab.com/mt6797/vendor/-/blob/ALPS-MP-M0.MP9-V1.78.1_NB6797_6C_M/mediatek/proprietary/hardware/connectivity/wlan/wpa_supplicant_8_lib/mediatek_driver_cmd_nl80211.c#L2160 Apr 13 22:05:41 MediaTek caught SETSUSPENDMODE from wpa_supplicant Apr 13 22:05:48 ew **** ENDING LOGGING AT Wed Apr 14 02:59:56 2021