ChanServ changed the topic of #linux-msm to:
marvin24_ has joined #linux-msm
marvin24 has quit [Ping timeout: 480 seconds]
jhovold has joined #linux-msm
jhovold has quit [Ping timeout: 480 seconds]
enok_ has quit [Remote host closed the connection]
enok has joined #linux-msm
enok has quit [Remote host closed the connection]
enok has joined #linux-msm
<aka_[m]> So it seems I need some changes for bengal
<aka_[m]> tho this one appears to be handled already:... (full message at https://matrix.org/_matrix/media/r0/download/matrix.org/tsWpcLAYGFZvpltgGjxgAjLy)
<aka_[m]> ok, im lost in qcom
<aka_[m]> code removes vdd-io-bias frtom sdhci-msm driver, yet flag is still there on dts
lumag_ has quit [Ping timeout: 480 seconds]
<Mis012[m]> device tree bindings are ABI
<aka_[m]> <Mis012[m]> "device tree bindings are ABI" <- theory
<aka_[m]> konradybcio: don't you wanna dig in that bongol tablet?
<aka_[m]> this thing appears here to not detect sdcard
<aka_[m]> with or without gpios
<aka_[m]> reinserting doesnt change anything
<konradybcio> I will not until lenovo releases sources
<konradybcio> which may never happen..
<aka_[m]> pain >_>
<aka_[m]> imma try #define DEBUG and see if it outputs msgs
<aka_[m]> konradybcio: scared of some custom brusn or?
<aka_[m]> hexagonwin had kernel source for j606 but without NVT-ts
<konradybcio> Yeah there is a drop from early 2021
<konradybcio> But my hw revision is not included
<aka_[m]> ichernev:
<aka_[m]> 🤨
<aka_[m]> konradybcio any comment?
<konradybcio> The order with this one was not really important or something, I don't remember exactly what it was..
<aka_[m]> ok so i have schematics open
<aka_[m]> L22 is wired in ok way
<aka_[m]> 🤨
<aka_[m]> L9A_1P8
<aka_[m]> DETECT SWITH VCC
<aka_[m]> no idea what exactly is that 3.2.16.0147_nonpth thing
<aka_[m]> konradybcio:
<aka_[m]> [ 0.641811] mmc1: new ultra high speed SDR104 SDHC card at address 0001
<konradybcio> nice
<aka_[m]> VDDIO was a little bit too low
<aka_[m]> bad thing ssh doesnt let me connect
<aka_[m]> after pmos_continue_boot
<aka_[m]> EXT4-fs (mmcblk1p2): mounted filesystem with ordered data mode. Quota mode: none.
<aka_[m]> tho i don't see any stuff starting later
<aka_[m]> only iommu keeping throwing context faults at me
<aka_[m]> it does more of them after i decided to add clk_ignore_unused
<aka_[m]> bengal and qcm2290 needs some quirks for gcc and something about always having parent on before reparenting
<ichernev[m]> So you put more volts than DS to get it going? What gives?
<ichernev[m]> So your device is not virtual a/b or you just plan to wipe them or there is already support in pmos?
<aka_[m]> <ichernev[m]> "So you put more volts than DS to..." <- nah at first i was kanging from nile for two regulators because it was throwing unsupported range.
<aka_[m]> <ichernev[m]> "So your device is not virtual a..." <- its virtual i assume because there is /super
<aka_[m]> now i should jump and rebase for rc5 i think
<aka_[m]> vmmc was lowered to be max 2.7 to use 1.8V signaling(?), and vqmmc can use downstream values
Danct12 has quit [Quit: Quitting]
Danct12 has joined #linux-msm
<aka_[m]> konradybcio: dpu still doesnt probe after rebasing >_>
<konradybcio> <aka_[m]> "konradybcio: dpu still doesnt..." <- Unlikely, but check if the compatibles exist, if there are no typos etc
<konradybcio> Also recheck configs and make sure things that it depends on probe
<konradybcio> Also without a panel driver, dpu wont do much
<aka_[m]> after removing clk ignore unused it cries about dispcc
<aka_[m]> i just dont understand fuckery behind dispcc
<aka_[m]> qcom does not pass clocks via dt
<aka_[m]> konradybcio: doesnt it make on downstream use global name lookup and end with GPLL0 from gcc?
<aka_[m]> in theory i don't like fact MUX value is 4 and it has different "parent_names"
<aka_[m]> its not like you can connect multiple clocks into one mux input or im too dumb to imagine it
<aka_[m]> or every rcg have its own mux ?
<aka_[m]> uh
<konradybcio> Gpll0 has an output to dispcc
<konradybcio> So that the entire dispcc(*) can be gated with a single clock
<konradybcio> Though it admittedly isn't the intended usecase 🤪
<aka_[m]> konradybcio: now im lost
<aka_[m]> i have changed one parent to go to gpll0 like it says on downstream driver compared to scuba yet same thing
<aka_[m]> 5f00000.clock-controller platform: wait for supplier dsi-phy@5e94400
lumag_ has joined #linux-msm
<aka_[m]> Marijn: doesn't that sound like this:
<aka_[m]> ?
<aka_[m]> i think i should just stick to FB backend and try to bring rest hardware till smarter heads gets done with rest
<aka_[m]> appears i had wrong understanding of -next branches lol, gonna rebase on rc5 tomorrow instead of current next
<steev> current -next makes me a sad panda
<Mis012[m]> steev: ikr, ugly Rust in there I hears
<Mis012[m]> *heard
<Mis012[m]> steev: not sure if you caught it, but I've missed decoding some segements and it looks like something very secureboot related is getting blown :(
<Mis012[m]> I guess ugly Rust is a pleonasm...
<steev> Mis012[m]: yeah i saw, but i'm not sure what it means, since secureboot is off
<steev> or you're saying it'll only work with secureboot on?
<Mis012[m]> steev: is secureboot off?
<steev> yes
<Mis012[m]> interesting
<steev> can't boot linux with it on
<Mis012[m]> ...
<Mis012[m]> -_-
<steev> or at least, can't use a dt with it on
<Mis012[m]> not UEFI secureboot
<Mis012[m]> nobody cares about that
<steev> oh
<steev> that is the only secureboot i am familiar with :(
<Mis012[m]> I'm talking about the "can I replace uefi with u-boot" secureboot
<Mis012[m]> also known as "do I own the device" secureboot
<Mis012[m]> usually the answer is "haha no"
<steev> that is likely haha no, but i wouldn't want to anyway? i don't hate u-boot, but i definitely prefer uefi
<aka_[m]> unless you get poorwinner
<aka_[m]> Mis012:
<Mis012[m]> steev: are you a vendor?
<Mis012[m]> otherwise no reason to prefer UEFI
<Mis012[m]> oops, used reply
<steev> Mis012[m]: i'm a distribution. one iso for many > one image per device
<Mis012[m]> u-boot literally has a hush shell script to parse grub.cfg iirc
<Mis012[m]> also, nice sane foss bootchain > ugly proprietary crap
<Mis012[m]> what's next, advocating for ACPI?
<steev> i'm not advocating for anything? i simply said one image that works across many devices is better than one image per device
<Mis012[m]> that can be arranged with u-boot
<steev> except it can't if u-boot isn't read from the same address across all devices
<Mis012[m]> ?
<steev> in the android world, you may have that already
<steev> in the sbc world, we do not
<Mis012[m]> you can't touch uefi as an owner, why do you need to touch u-boot as a distribution
<Mis012[m]> u-boot loads Linux, not the other way around...?
<steev> because not all devices have spi with u-boot on them?
<steev> it's not rocket surgery
<Mis012[m]> doesn't need to be SPI?
<Mis012[m]> you can boot the u-boot from sdcard if you wish so
<Mis012[m]> you can boot it over USB with EDL
<steev> well considering not all sbc even have storage....
<steev> you kinda have to write it on sdcard if you want it to be used?
<steev> is this really a foreign concept?
<Mis012[m]> how does uefi vs u-boot change anything wrt the sdacrd
<Mis012[m]> *sdcard
<steev> i'm saying u-boot isn't superior just because *handwavy*
<Mis012[m]> it's superior because it's a saner approach and because it's foss
<steev> foss doesn't make it immediately superior
<Mis012[m]> it does
<steev> it means someone CAN make changes, but that doesn't mean someone WILL
<Mis012[m]> u-boot is saner and it's foss
<steev> okie dokie
<Mis012[m]> what does eewefi have going for it
<steev> let me guess, you also call it microshit or micro$hit
<steev> is it really still 1998?
<Mis012[m]> I call it "can't see any good reason to touch this with a ten foot pole"
<steev> so then don't
<steev> your reasons for things, aren't everyone's reasons
<Mis012[m]> M$ (stands for MicroSoft) being less openly hostile to Linux is clearly bad for Linux Desktop as far as I can tell, if that's what the 1998 thing refers to
<Mis012[m]> unless you could WSL as Linux Desktop, but then you're objectively wrong
<steev> i'm glad you're the expert on what is and isn't objectively wrong, and i will seek out your guidance any time i need to make a decision
<Mis012[m]> I'm not saying I'm an expert on everything, but I'd say using WSL not counting as using Linux as your Desktop operating system is fair?
<steev> do you consider someone using a vm to be using a desktop?
<Mis012[m]> I consider them to be using the desktop that is hosting the VM
<steev> do you often see people running around saying they use linux as their desktop because they are using wsl? because i've never seen anyone say that
<Mis012[m]> well... I've seen a lot of allusions to it being basically equivalent
<steev> the 1998 thing was more about edgy teens calling it micro$shit or M$, instead of just saying the name, as if it's cool and hip to call things improperly.
<Mis012[m]> it's an identity thing
<Mis012[m]> afaict
<Mis012[m]> and also, if you hang out around people with sane opinions on M$, you kinda get used to writing at as M$
<Mis012[m]> it spreads
<Mis012[m]> and I don't see any reason not to write it as such, so /shrug
<Mis012[m]> interestingly Rust is already literally called after something that invades a pristine material and deteriorates it, so they're safe from this problem
<Mis012[m]> and I guess eewefi also gets ahead of my inner temptation of telling people what I think about uefi
<Mis012[m]> well, in any case, no wonder ownership is getting eroded when "why should I care" is what most of the population is going to say
<steev> if it makes you happy mi$o
<konradybcio> Speaking of linux-msm...
<konradybcio> aka_: yes ive seen that many times, it usually fixes itself, I have no idea why and how
<konradybcio> as I said, rc3 works for me