02:50 <eiln> > 522: Does not support 1 plane for Skye WARP

02:50 <eiln> no luck

06:39 <marcan> chadmed: any chance it's an audio format issue? can you identify any specific signal threshold where it clips?

06:40 <marcan> fun fact: i2s and left-justified audio formats are identical except for a one-bit shift in the msb, which means LJ going into an i2s-configured codec will wrap around nastily at 1/2 full scale

06:41 <marcan> eiln: no 1-plane mode is unfortunate, but it is what it is. means we need to fix up libcamera / whatever else doesn't support it

06:42 <marcan> thankfully desktop linux is moving towards that and PipeWire as abstractions so most modern apps should be fixable in just one or two global places

06:42 <marcan> and besides multiplanar is a better format anyway

06:57 <jannau> pitty, that would have increased compatibility simply by doing what most other HW does

06:58 <jannau> a little unexpected that the sensor or isp does only 4:2:0

06:59 <jannau> eiln: thanks for looking

11:08 <chadmed> marcan: a sine wave at 160 hz will cause hard digital clipping at 65% going straight into the woofers

11:08 <chadmed> an A4 sine wave does not

11:08 <chadmed> furthermore the 160hz sine wave contains very audible harmonics so the digital side of the chip seems to be saturated

11:09 <chadmed> as i said last night the quirks povik added to sound/ivsense-improvements seem to do *something* but i didnt get around to doing much testing on that branch

11:10 <sven> huh, maybe i shoud've based my work on asahi instead of torvald's tree. lots of conflicts in the pcie stuff

11:10 <sven> guess that means only the new usb3 code for now :D

11:12 <jannau> yes, pre-brought up ports (by u-boot) and t602x* support are not yet submitted upstream

11:13 <sven> make sense. let me just split off the new usb3 work for now then so that people can test at first

11:13 <sven> and then thunderbolt afterwards

11:17 <chadmed> i should note that it doesnt matter if the sine wave is being generated at 0 dBFS or less, its definitely a codec thing

12:57 <arix> sv3n: do you mind posting at least wrong-based version of usb4 stuff? It'd be curious to see that while it probably won't be testable by non-professionals

12:59 <arix> Almost no blog posts from you (and very little from Janne), while you both truly do amazing job with display/io stack

12:59 <jannau> sigh

13:02 <sven> 🙃

13:06 <jeisom> I'd love to see a blog post about the TB3/usb stuff AFTER you have gotten it all worked out and merged for testing at least. Would be interesting from a challeges of RE that and getting it working reliably.

13:08 <sven> sure

13:13 <marcan> chadmed: could it be some kind of interaction with mechanical resonance? if it's frequency dependent it feels weird that it'd be strictly a codec issue

13:14 <marcan> like at certain frequencies the voltage ends up higher and hits the rails or something

13:14 <marcan> (I'm not terribly familiar with the phase response/etc of typical speakers but it feels there could be something going on there)

13:15 <marcan> if codec quirks improve it though, then maybe not

13:16 <chadmed> its certainly possible and its something i can and will test once the visense stuff is merged

13:16 <chadmed> though the fact that we start getting harmonics in the output of a pure sine wave just before we turn the volume up enough to get the noise (and the fact that its a hard cut in) suggests that the digital side of the amp is saturated to me

13:16 <chadmed> ill upload some videos hang on

13:17 <chadmed> OH thats right macos is also prone to this!

13:17 <chadmed> when i was running frequency sweeps in macos for fun ages ago i hit it

13:29 <chadmed> https://chadmed.au/res/buzzing.mp4

14:14 <sven> maz: did you ever see the pass-through usb connection of the central scrutinizer falling back to full speed only? it doesn't seem to want to do high speed anymore for me

14:14 <sven> (and full speed isn't supported in m1n1 right now I think)

14:20 <sven> can't seem to get it back to high speed even with different cables right now :(

14:21 <sven> unfortunately I also don't have anything fast enough to actually look at what's going on with D+/D-

14:21 <sven> I guess the usb2 chirp doesn't reach the other side for some reason

16:43 <maz> sven: odd. I'm pretty sure the impedance matching of the USB lines is... iffy, but it always worked on the 3 machines I use that directly boot over the USB pass-through.

16:44 <maz> has it worked before? or is it the first time you're trying to make use of it?

17:32 <sven> it has worked before. last time I used it I had to use an additional hub and blamed the raspberry pi it was connected to though

17:35 <maz> I'm definitely going through a hub too (too many machines connected to the console server). given that the version you have has no active compotents between the two ends of the D+/D- lines, it must be related to the dodgy impedance of the tracks... :-/

17:35 <maz> components*

17:36 <sven> :/

17:36 <sven> oh well. back to direct connection to my MacBook for now

17:37 <sven> I’ll take a look at the connectors under a microscope tomorrow, could also be iffy solder joints I guess

17:38 <maz> could you inspect the solder pads on both ends? I can also post you a more recent version, to work out whether it is something that magically improved.

17:38 <sven> is that pcb 2 or 4 layers btw?

17:38 <maz> 2 layers only.

17:38 <sven> I think you can do usb 2 on two layers but it’s more tricky iirc

17:39 <sven> I’ll check the solder joints tomorrow

17:39 <maz> OK, thanks. I'll see if I can learn how to route these signals in a better way.

17:40 <sven> I’m far from an expert in signal integrity and pcb so I’m probably not the right person to listen to ;)

17:40 <sven> *pcbs

17:41 <maz> hey, you raised the issue. you're the expecrt for now! :D

17:41 <maz> expert*

17:44 <sven> uh oh

17:45 <sven> https://www.altium.com/live-conference/altiumlive-2018-annual-pcb-design-summit/sessions/extreme-importance-pc-board-stack-up that talk is probably like 90% of what I know :D

17:48 <maz> oh, interesting. thanks for the pointer. I always wondered how much I could get away with not caring too much about that, but I guess I now know... ;-)

17:49 <maz> FTR, I just booted a machine with the USB pass-through directly attached to the server, and it went into full speed without any issue.

17:49 <maz> s/full/high/

17:50 <maz> (i.e. it booted fine).

17:50 <maz> no hub involved. but a sample of one doesn't mean much.

17:50 <sven> yeah, it could just be a bad solder joint on my board after all

17:50 <sven> with all the crappy usb 2 cables out there those signals can probably take a lot of abuse

17:51 <maz> I've had a number of quality issues with JLCPCB -- one batch of CS (v3) was a total dud...

17:53 <sven> wow.. 7$ for a 5 4 layer boards according to their quote tool

17:54 <sven> not surprised they might have quality issues at that price

17:56 <maz> yup. the level of QC is... terrible. on that batch, I have components that are literally off by a mm... they just give you a coupon to respin it.

19:46 <zerdox> hi everyone. not sure, if this is the right place to ask, but i got no response in #asahi-alt. i moved to 6.4.0 and updated mesa in order to get new gpu drivers, but everything broke and now wayland doesn't work without software rendering. i found out that some modules like appledrm and apple_dcb are not loaded due to "No such device" error. zcat on config.gz gives me =m on both DRM_APPLE and DRM_ASAHI. i always assume that i did something wrong on my side

19:46 <zerdox> but i already spent a few days trying to get my GPU work. and yes, i did reload, uname gives me 6.4.0 and modules appear in /lib/modules

19:54 <jannau> zerdox: this is not the place, let's continue in #asahi-alt. please upload your kernel (outout of dmesg after boot) log somewhere

23:20 <amarioguy> hrm seems like when i install the m1n1 custom boot object on a ventura partition, doesn't seem to set up the boot time framebuffer correctly? (even w/ a display connected it keeps using the fake framebuffer) (note that i do have a sonoma partition on this mac so the first stage iboot is the sonoma one)

23:39 <jannau> on which device? m1 mac mini / m1* studio?

23:44 <jannau> amarioguy: ^^^

23:48 <jannau> hmm, doesn't that sound more like if m1n1's display init code fails

23:51 <jannau> if possible provide serial log of the display init (either by using the HW serial or a 1st stage m1n1 which skips display init)

23:52 <jannau> that way it's possible to look at the display init under HV

23:53 <amarioguy> jannau: m1 ultra studio

23:54 <amarioguy> getting that log, one sec