13:35 <maz> sven: re ATCPHY: the DT changes are making u-boot pretty unhappy ("Bus usb@702280000: Port not available."). I'll try and have a look.

14:40 <maz> sven: duh, resets. I've bodged it.

15:57 <maz> sven: we're in business. I have the feeling that there is some odd orientation issues (flipping cables made things work...)

15:58 <j`ey> run the debian test!

16:00 <maz> j`ey: it *is* running!

16:02 <maz> j`ey: about 30 second *slower* than the Mini.

16:03 <j`ey> huh

16:13 <maz> I can gain *some* performance back by pinning the guest to a single cluster, but this is still about 22 seconds behind at best (about 15% slower).

16:14 <maz> this is still bloody fast, but I guess physics gets in the way of absolute performance.

16:26 <sven> huh, the orientation should be correct. i know about two other issues though: sometimes it won't find the device when it's connected during boot; and it takes ~5 seconds to find newly connected devices

16:27 <sven> 676748389f5 should also already be in torvald's tree, without that it won't be able to tell the orientation

16:28 <maz> sven: yeah, I've seen the delay, but there definitely was something with one of my USB/SATA adapter. it wasn't detected, and I was trying to understand why it was failing. in the end, I flipped the cable and the device showed up.

16:29 <sven> weird

16:29 <sven> which adapter is that? all my usb 3 devices work in both orientations

16:30 <sven> hrm, i guess it could also be some difference they introduced in the max/pro/ultra

16:30 <maz> Bus 012 Device 002: ID 152d:0578 JMicron Technology Corp. / JMicron USA Technology Corp. JMS578 SATA 6Gb/s

16:31 <maz> I'll try and reproduce it once I'm done doing silly thing with the new toy...

16:31 <maz> I'll push the DT patches somewhere so that you and jannau can merge them into what you already have.

16:32 <maz> ah, it could also be that the fuse values are not entirely correct for Ultra (I lifted the Max settings).

16:32 <sven> sounds good, i found an adapter with that chip for like 10 eur on amazon. if you can reproduce it i'll just buy that one and see what happens here

16:33 <sven> true, i don't really know what half of these fuses do anyway

16:34 <maz> sven: don't buy anything until I've confirmed it... (and worse case, I can send it to you).

16:35 <sven> sure :)

16:46 <maz> sven: it's more subtle than that. it depends on the *cable*. one of them (rumoured to be a USB-3.2 Gen-2 cable) triggers the problem.

16:46 <maz> another (non-descript) cable works just fine.

16:46 <sven> oh fun

16:48 <maz> and I got the enclosure wrong. the one I trigger this on is:

16:48 <maz> Bus 006 Device 002: ID 174c:235c ASMedia Technology Inc. Ugreen Storage Device

16:48 <maz> (too many storage devices)

18:46 <kettenis> jannau, sven: regarding that u-boot patch that's still missing upstream; we could work around the issue in the DT by dropping the dr_mode property (since 'otg' is the default for the Linux kernel

18:47 <kettenis> but hopefully it will be picked up soon