**** BEGIN LOGGING AT Sun Jul 31 02:59:56 2022 Jul 31 08:05:16 Morning! Jul 31 09:41:15 Morning! Jul 31 10:21:43 Morning Jul 31 10:53:48 Herrie: for TP mainline, I think the first thing to get up and running is the usb connection Jul 31 10:54:07 Tofe: Yeah that would help for sure Jul 31 10:54:22 Maybe we can piggyback some defconfig options from hammerhead? Jul 31 10:54:28 I think it works there? Jul 31 10:55:15 I don't recall our very last status, actually Jul 31 10:55:19 but should work yes Jul 31 10:55:25 I see in initrd: Cannot find device "usb0" Jul 31 10:55:32 so something basic is missing Jul 31 10:58:15 we do have "usbcore: registered new device driver usb" so the base is there Jul 31 11:01:59 I see CONFIG_USB_CHIPIDEA=m , maybe it should be builtin Jul 31 11:02:53 more generally, having modules isn't such a great idea right now Jul 31 11:04:01 Some of the other options are set like M too. CONFIG_USB_CONFIGFS=m etc Jul 31 11:04:11 Maybe change quite some of those m values to y simply? Jul 31 11:11:59 I just did that, now rebuilding kernel Jul 31 11:22:42 nope, still no usb0 it seems Jul 31 11:37:54 Tofe: According to the Topaz3G there's a ISP1763 USB controller Jul 31 11:39:30 We might need this flag as well? https://github.com/torvalds/linux/commit/5c81cfb6a60a7abbfe7520fa13cabdad094d3d39 Jul 31 11:39:50 The 5.15 defconfig has CONFIG_USB_ISP1301=y from what I can tell Jul 31 11:39:55 It might be somehow compatible as well Jul 31 11:41:18 so it wouldn't be chipidea ? Jul 31 11:42:28 Tofe: Hard to say Jul 31 11:42:41 There's also a HS USB Jul 31 11:42:59 Which is connected to Micro USB and MAX8903 charger Jul 31 11:43:10 let me try and find this usb in the dts Jul 31 11:43:17 HS USB is part of the APQ8060 Jul 31 11:45:18 mmh I don't find any usb description, apart for some power regulators Jul 31 11:46:17 was looking for something like this: https://elixir.bootlin.com/linux/v4.9/source/arch/arm/boot/dts/qcom-apq8064.dtsi#L852 Jul 31 11:47:50 Yes we'd need to have something like that in DTB for sure for it to work Jul 31 11:49:59 I guess like this? https://github.com/mbd98/hp-touchpad-linux/blob/bc754423af6d851766973b35f9a8890e38478ffb/arch/arm/boot/dts/qcom-apq8060-tenderloin.dts#L660-L675 Jul 31 11:51:08 I guess this whole commit might be useful somehow? https://github.com/flto/linux/commit/943c90b96c2b7f29c7bab2ebb5e7b3b5a7ba96ff Jul 31 11:52:01 https://github.com/flto/linux/commit/943c90b96c2b7f29c7bab2ebb5e7b3b5a7ba96ff#diff-89bd340318fe1aa8809c5ff52d442e48275afdacf3c7d7aa2994c93d370a7cf2R661 this is what we want Jul 31 11:52:21 but yes, the whole commit is great Jul 31 11:52:57 It could be some of the bindings have been renamed in 5.19 though Jul 31 11:53:13 I've seen some renaming going on between 5.10 and 5.19 while following Megi's kernel stuff Jul 31 11:53:15 For PP(P) Jul 31 11:53:37 But I guess that would show up quickly enough while compiling Jul 31 11:53:44 let's hope so Jul 31 11:54:18 DTB validation would fail then Jul 31 12:14:10 ok, first try for the merge is done... let's see if that builds Jul 31 12:17:52 Or at least that's my expectation Jul 31 12:27:55 yep: /media/LuneOS/testing/webos-ports/tmp-glibc/work-shared/tenderloin/kernel-source/arch/arm/boot/dts/qcom-apq8060-tenderloin.dts:892.2-18: ERROR (duplicate_property_names): /soc/gsbi@19800000:status: Duplicate property name Jul 31 12:30:30 Well the error message is clear enough :D Jul 31 15:06:09 Herrie: so, I'll be trying this: https://github.com/Tofee/shr-linux/commit/e56402695aed2fdec38c549bc57eb22a6f69b325 Jul 31 15:07:36 Tofe: OK Jul 31 15:07:48 mdp4 is display related by the looks of it? Jul 31 15:09:49 yes, but my TP got stuck when intiializing display, it seems Jul 31 15:14:13 I'll deactivate the panel for now, and try to get usb working Jul 31 15:26:01 damn, I can't get it to work again Jul 31 15:34:43 You might need this? https://github.com/mbd98/hp-touchpad-linux/commit/ff35130647ee7c7bb592781290ba0cd6ce21e589 Jul 31 16:09:15 could be but the dts is a bit hybrid right now, so hard to tell Jul 31 18:14:04 another day of slow charging didn't help my TP and disassembling it didn't help as well yet (other than loosing some clips, so it's easier to disassembly now) Jul 31 18:30:15 JaMa: Yeah opening it without breaking clips is close to impossible based on my experience Jul 31 18:40:42 Herrie: it would seem the issue was https://github.com/Tofee/shr-linux/blob/tenderloin/5.19/mainline/arch/arm/boot/dts/qcom-apq8060-tenderloin.dts#L858 , maybe a conflict with my serial cable ? Jul 31 18:41:30 I'm now reactivating stuff slowly, to be sure it's only that Jul 31 18:46:02 Tofe: Not unlikely it somehow interferes Jul 31 18:46:14 I've always tried either USB or serial in the past Jul 31 18:51:04 ok, it boots again Jul 31 18:51:38 still no usb... Jul 31 18:53:23 Here's the paste of the bootup: https://paste2.org/sjxXDMYU Jul 31 18:56:12 There's some mention of usbcore in there Jul 31 18:57:45 https://github.com/Tofee/shr-linux/commits/tenderloin/5.19/mainline here's the current status Jul 31 18:58:23 also, the screen now clears out Jul 31 18:58:51 Tofe: Seems we might need this flag? https://www.kernelconfig.io/CONFIG_MSM_GCC_8660 Jul 31 18:59:28 ah, well, yes, I think we need it Jul 31 18:59:37 don't we have it already ? that would be a shame :p Jul 31 19:00:16 Seems not... Jul 31 19:00:17 CONFIG_MSM_GCC_8660=y from serial console, so it's there Jul 31 19:01:44 https://paste2.org/e0DkFypO here's what kernel drivers are loaded Jul 31 19:01:53 we can see gcc-msm8660 Jul 31 19:03:03 That's quite osme, but no USB ones Jul 31 19:04:09 flto's dts says that usb should be taken in charge by "qcom,ci-hdrc" driver Jul 31 19:04:20 and we see ci_hdrc Jul 31 19:08:26 Tofe: Found this as well for same chipset: https://github.com/peremen/linux-apq8064-mainline/commits/work-8960-usb Jul 31 19:09:33 the dts change looks very similar to what we have Jul 31 19:10:42 ah, he put compatible = "qcom,usb-hs-phy-msm8960" Jul 31 19:10:50 instead of our qcom,usb-hs-phy-apq8064 Jul 31 19:10:59 interesting, actually Jul 31 19:13:06 all this is handled by drivers/phy/qualcomm/phy-qcom-usb-hs.c, it seems Jul 31 19:13:51 leading to CONFIG_PHY_QCOM_USB_HS Jul 31 19:14:39 CONFIG_PHY_QCOM_USB_HS=y damn Jul 31 19:17:28 Well msm8x60 and apq8064 should be very similar Jul 31 19:17:47 From what I understood apq8064 lacks modem where others have modem Jul 31 19:18:10 For the rest they should be close to identical at least that was my understanding Jul 31 19:22:26 well I don't really see the difference regarding usb here :/ Jul 31 19:29:29 Ah wait I'm mixing up 8064 with 8060 Jul 31 19:32:27 Then again 8064 seems successor in 2013 of 2012 8060 Jul 31 19:37:29 ok I give to today... my commits are on my repo, in case you want to try something Jul 31 19:37:40 s/to/up for/ Jul 31 19:39:51 Tofe: OK, do you have your defconfig also? Jul 31 19:39:59 I'll be giving it some go later I guess Jul 31 19:40:20 sure, one moment Jul 31 19:42:41 https://github.com/Tofee/meta-smartphone-1/commit/d17c655af63cce22f9f3f1d95078970287d3454a Jul 31 19:42:57 I should probably rebase my meta-smartphone Jul 31 19:43:39 done https://github.com/shr-distribution/meta-smartphone/compare/kirkstone...Tofee:meta-smartphone-1:kirkstone/mainline-tenderloin?expand=1 Jul 31 21:27:40 Tofe: Thnx Jul 31 21:32:40 Tofe: We might need the avdd33 regulator. Comment there says: "regulator-always-on; // has to be on for USB to work" in https://github.com/flto/linux/commit/943c90b96c2b7f29c7bab2ebb5e7b3b5a7ba96ff **** BEGIN LOGGING AT Sun Jul 31 22:15:57 2022 **** ENDING LOGGING AT Mon Aug 01 02:59:57 2022