<edrex>
<samueldr "and indeed, fully reproducible u"> what triggers it? just plugging into a particular PS?
aleph- is now known as Batou
<samueldr>
fusb302 driver loading
<samueldr>
the SBC is powered form type-c without alternative sources
<samueldr>
I *assume* something assumes type-c+fusb302 means cutting the power for a moment won't be an issue
<samueldr>
see, there's a battery
<samueldr>
except on that device!
<edrex>
but this just started happening one day? kernel upgrade?
<samueldr>
it's uh... a bit more involved than that
<samueldr>
I believe the dtb didn't describe things properly at an earlier point
ryantrinkle has quit [Ping timeout: 252 seconds]
<samueldr>
or I'd have to check
<samueldr>
maybe I already had the module blacklisted with kernel 5.5
<samueldr>
I didn't always have an RTC clock battery in it
<samueldr>
I had trouble getting one locally :)
<samueldr>
(stores always were out)
<samueldr>
talking with you I recall that enabling type-c would cut its power
<samueldr>
so same issue
<samueldr>
60% sure it's nothing new, that I'm the first one to try on mainline + rtc clock battery
<edrex>
if removing the RTC battery resets the error state, not having one in would certainly have an effect :D
ryantrinkle has joined #nixos-aarch64
<edrex>
oh so the boards ship with an empty battery slot
<edrex>
what does it do for power if the fusb302 module isn't loaded? does the usbc sync still do its thing in some default mode?
<edrex>
sink rather
<samueldr>
no shipper from china will ship a coin cell battery without heavy paperwork because it's a lithium battery
<edrex>
ah right
<samueldr>
good q
<samueldr>
I don't know enough about fusb302 to understand it
<edrex>
i say USBC when I mean PD
<edrex>
gotta break that habit
<samueldr>
but IIRC some of those PD power chips have firmware storage
<samueldr>
and maybe it's fine with whatever firmware it has in it
<samueldr>
(a person involved with the board conception linked to the datasheet, and yeah it's "programmable")
<samueldr>
(not sure though what it means in reality)
<edrex>
you're able to power off a standard 5V supply with an A to C cable then, but it goes undervoltage if you try to compile stuff?
figgyc has quit [Quit: No Ping reply in 180 seconds.]
figgyc has joined #nixos-aarch64
<edrex>
<samueldr "(not sure though what it means i"> looks like it has a bunch of i2c register switches you can poke at, i wonder if that's all they mean
<samueldr>
could be
<samueldr>
I don't know, haven't tried solely on 5V
<samueldr>
might be fine if it's only compiling
<samueldr>
not sure how much the NVMe interface needs
<samueldr>
and it's not really the 5V I'm concerned with
<samueldr>
but the power
<samueldr>
I think type-c (yes type-c) changed the max power to 3A
<edrex>
> VBUS and GND provide 5 V up to 900 mA, in accordance with the USB 3.1 specification. A specific USB-C mode may also be entered, where 5 V at either 1.5 A or 3 A is provided.[56] A third alternative is to establish a Power Delivery contract.
<{^_^}>
error: syntax error, unexpected ',', expecting ')', at (string):494:38
<edrex>
{^_^}: you shouldn't go eval'ing strange codes
<clever>
edrex: usb-c PD is an active digital protocol over one of the CC pins
<clever>
it can do various things, including requesting a higher voltage on VBUS, and deciding to route things like hdmi or displayport over the usb3 pairs
<edrex>
the situation is not confusing at all. Not even when you throw in Quick Charge
<clever>
usb-c also has a passive protocol on the same CC pins, using pull resistors
<edrex>
but I'm at least happy i can charge my laptop and my phone off the same cord, after all the pain
<clever>
the passive protocol is 5v only
<edrex>
so that's the max 3A one?
<hexa->
it's a bit annoying that > triggers evals, when it has been used for quoting since maybe forever(TM)