#
✕ nick:j_ey
#asahi · 2021-09-23 07:41
j_eymaz: :-)
#asahi · 2021-09-23 07:32
j_eymaz: and the in-memory state thing, I started to try to remove, but decided it was better to get a first version on the mailing list first
#asahi · 2021-09-23 07:29
j_eymaz: thanks, I assumed it had to be unique
#asahi · 2021-09-22 22:30
j_ey(you said to use GPIO, but theres 2 (or more) pinctrl devices)
#asahi · 2021-09-22 22:30
j_eymaz: re: pctl->irq_chip.name, how does that work with name clashes?
#asahi · 2021-09-22 16:23
j_eymaz: awesome
#asahi · 2021-09-22 16:22
j_eymaz: I think I'll need a bit of help with "You really should have one context per mux" (not right now though)
#asahi · 2021-09-22 15:47
j_eyalyssa: yeah, I should have done that *right* after I cherry-picked. I realised too late. "I'll just do it before I send it out" I thought.. and then forgot :)
#asahi · 2021-09-21 22:42
j_ey(I mean have drivers/gpio/apple.c and drivers/pinctrl/apple.c)
#asahi · 2021-09-21 22:41
j_eynormally you would want to split gpio and pinctrl I think, but theyre all in the same hw block on the m1
#asahi · 2021-09-21 22:41
j_eymuxing is well. muxing
#asahi · 2021-09-21 22:40
j_eypinconf is about electrical settings (voltage etc), that we dont support
#asahi · 2021-09-21 22:40
j_eypinctrl includes pinmux, pinctrl is the 'overall' name for it. there's also pinconf
#asahi · 2021-09-21 22:37
j_eyalyssa: note2 is for you :P
#asahi · 2021-09-21 22:34
j_eyjust making it clear I wasnt being paid to look at it :P
#asahi · 2021-09-21 22:31
j_eywell I finally sent it out https://lore.kernel.org/linux-gpio/20210921222956.40719-1-joey.gouly@arm.com/T/#t
#asahi · 2021-09-21 16:59
j_eymarcan: nothing else really looks similar to that samsung spi
#asahi · 2021-09-21 16:45
j_eymarcan: im happy to test with keyboard, since the other obvious spi thing is the nor flash.. and well.. we dunno about that yet :P
#asahi · 2021-09-21 16:38
j_eysven: lol, high praise on that patch.
#asahi · 2021-09-21 16:38
j_eymarcan: I have tried doing some random changes and stuff in this driver.. but if you want to redo it completely, I think that might be better
#asahi · 2021-09-21 16:37
j_eymarcan: I watch a Ben Eater video on spi (https://www.youtube.com/watch?v=MCi7dCBhVpQ) but thats about as far as I got
#asahi · 2021-09-21 16:35
j_eyI'll try that in a bit
#asahi · 2021-09-21 16:32
j_eyoh right
#asahi · 2021-09-21 16:31
j_eymarcan: can you get the data for spi3? what you pasted was spi1, want to make sure it's the same
#asahi · 2021-09-21 16:28
j_eymarcan: thanks for all the quick fixes today!
#asahi · 2021-09-21 16:17
j_eyalways-on isn't needed either
#asahi · 2021-09-21 16:17
j_eyremoves always-on
#asahi · 2021-09-21 16:15
j_eyyup, that works!
#asahi · 2021-09-21 16:15
j_eyat least in the old way, I think refclk was the topmost level of the chain of clocks, so it makes snese maybe
#asahi · 2021-09-21 16:14
j_eyok, I didnt retry that after fixing the power issues
#asahi · 2021-09-21 16:12
j_eykettenis: the irqs are working, i can hit keys and see 'corrupted spi packet'!
#asahi · 2021-09-21 16:12
j_eymarcan: first attempt at converting to pstate driver https://paste.gg/p/anonymous/f864f096973f4f298ca0919d160e050a
#asahi · 2021-09-21 16:10
j_eyyes
#asahi · 2021-09-21 16:08
j_eyok getting closer, no SError now, but all the SPI packets are corrupted
#asahi · 2021-09-21 16:06
j_ey(I made another clock, I just realised my issue might be that I messed up the power domains, so double checking)
#asahi · 2021-09-21 16:05
j_eymarcan: well is it dyanmic or fixed? :P
#asahi · 2021-09-21 16:04
j_eyim confused what I should put in the dt for that then
#asahi · 2021-09-21 16:04
j_eyso spi isnt in there
#asahi · 2021-09-21 15:57
j_ey(it doesnt)
#asahi · 2021-09-21 15:56
j_eygoing to try with refclk first, just to see if that works :p
#asahi · 2021-09-21 15:56
j_eymarcan: boots now. I have to figure out a clock for spi though
#asahi · 2021-09-21 15:48
j_eyI just rebased, lol, guess I'll fetch and do it again
#asahi · 2021-09-21 15:47
j_eynot something I intentionally selected
#asahi · 2021-09-21 15:45
j_eymarcan: so do you have CONFIG_OF_DYNAMIC? wondering why you didnt see it
#asahi · 2021-09-21 15:43
j_eyWARN also causes the brk though, I do have CONFIG_OF_DYNAMIC, I will try lower()!
#asahi · 2021-09-21 14:59
j_eystuff like this makes WARN behaviour confusing: https://github.com/torvalds/linux/blob/master/drivers/pinctrl/core.c#L1711
#asahi · 2021-09-21 14:56
j_eyfrom https://github.com/torvalds/linux/blob/master/arch/arm64/include/asm/bug.h#L22 https://github.com/torvalds/linux/blob/master/arch/arm64/include/asm/asm-bug.h#L37 and https://elixir.bootlin.com/linux/v5.15-rc2/source/include/asm-generic/bug.h#L97 it ""obviously"" does come to a brk.
#asahi · 2021-09-21 14:43
j_eyIm trying to understand why I hit that, a WARN shouldnt cause it.. (I've seen this with other WARNS)
#asahi · 2021-09-21 14:34
j_eyI dont know why those lead to brks for me
#asahi · 2021-09-21 14:34
j_eyrefcount_warn_saturate.. I thought it might be a WARN

plus ›