02:17 <HdkR> Does Snapdragon have any stats from its memory controller about memory BW load? Tegra has something like that which is super handy

05:21 <steev> clover[m]: okay tagged and pushed 6.0.7 (i hope) - new change is adding in the bwmon stuff so better memory bandwidth probably, additionally is mani's thermal patches (with some mainline things pulled in to support v4 since v5 hasn't been posted yet)

05:22 <steev> actually it's still pushing

05:33 <steev> now it should be good and ugh... apparently i pushed every single tag and not just the lenovo-x13s-6.0.7 one

11:45 <hexdump0815> aceridus: just fyi, i'm currently trying to get the ufs storage working on the galaxy book go based on the sm7125 info

11:48 <hexdump0815> aceridus: in case i get that somewhere my plan would be to look if adding the extra usb properties in the aspire dts will help to get usb working more generic

11:49 <hexdump0815> aceridus: regarding the memory regions for gpu etc. i think we should just try to start with the aspire dts ones as they are working and i guess woa sc7180 devices maybe do not differ that much in that regard (just a guess)

12:30 <Dylanger> <EricCurtin[m]> "If you want to give this a go, I..." <- I'm keen, I'll wait til you've finished holidays then I'll look at installing Fedora on it, seeing as though Debian runs _very_ smoothly it shouldn't be too difficult

12:31 <Dylanger> If Fedora supported Depthcharge that'd be awesome

12:31 <Dylanger> People could install Fedora on most aarch64 laptops, the seemingly random 34MB boot limit is annoying

13:15 <EricCurtin[m]> Most aarch64 laptops would be the dream, but we need to encourage vendors to follow "upstream first" to achieve this, you and everyone else who wants/needs this should try and send this feedback to them like Matt Miller stated recently on twitter https://twitter.com/mattdm/status/1588597287101632514 ... It's not trivial to get everything working... Chromebooks make it easier as Google force vendors to upstream most of the stack...

13:15 <EricCurtin[m]> The reason x86/AMD machines "just work" is because those companies put effort into upstreaming (and other related issues like agreeing on standards, etc.)

16:12 <hexdump0815> aceridus: jenneron[m]: i tried to add ufs support for the galaxy book go and here is how far i got: https://pastebin.com/raw/hnmPVpYi - so not working yet ...

16:13 <jenneron[m]> weird dt binding

16:14 <hexdump0815> some questions come up for me: is the problem some still missing regulators vdd-hba-supply and vccq-supply, is it maybe wrong values in values supplied in drivers/phy/qualcomm/phy-qcom-qmp-ufs.c (diff is at the end of the above pastebin) or something else wrong or missing maybe?

16:15 <jenneron[m]> isn't 517 deffered probe?

16:15 <hexdump0815> jenneron[m]: you mean i did something wrong with it - which could definitely be the case ... yes its deferred

16:16 <jenneron[m]> deffered probe means missing dependency

16:16 <jenneron[m]> e.g. when you reference to disabled node

16:17 <hexdump0815> then i looked at the usb ports and tried to find the proer regulators for the (currently commented out) usb entries taken from travmurav[m] aspire dts, but in the dsdt.dsl (link on top of pastebin) i see two usb sections with different regulators

16:17 <hexdump0815> ah - ok, so maybe something else is required for ufs to work?

16:18 <jenneron[m]> yes, some dt binding or some kernel config

16:18 <jenneron[m]> also, it may happen when a dependency fails to probe

16:19 <hexdump0815> last question: if i find multiple regulators in dsdt.dsl and for instance three defined in the usb entry - how du i find out which is which, i.e. vdd-supply, vdda-pll-supply, vdda-phy-dpdm-supply ?

16:26 <hexdump0815> for reference: full dmesg (with some i2c spam removed): https://pastebin.com/raw/LVD0EfU6

16:28 <hexdump0815> i think the reason why usb works might be that its initialized by the bios if a device is connected at boot and left on when searching for potential usb boot devices - if no device is connected its not initialized and some regulators for usb are wrong in the kernel so that usb will not come up if not initialized properly beforehand already by the bios - but its just a guess

16:40 <jenneron[m]> "how do i find out which is which" -> by voltage

16:43 <hexdump0815> ah - ok - thanks