Hi all, So I have been bumping into problems when using my SuperSpeedPlus-capable UAS drives/enclosures (with bridge chips from different vendors, namely ASMedia and Realtek). I don't exactly remember when I first noticed the problem because at that time I only had one cheap (Intel Alder Lake) laptop from a Chinese vendor (Chuwi) with Gen 2 ports (one A + one C) and I thought maybe they were fake or something, but it has probably been a year or longer. And now I am also experiencing it on a desktop with an XHCI card with two C ports (The controller is ASMedia ASM3142, I think). Basically the problem is that the device would seem to hang and/or reset, often (if not always) for just *either* reading or writing. Occasionally the same device (on the same host/port) will work perfectly fine. By that I mean it *feels like* whether the problem will show up is determined at the time the device is connected / "initialized" -- if things happen to have "gone right" at that stage, it will keep on working fine (until e.g. physical reconnection / driver reinitialization). (But it's not entirely conclusive whether that is really the case.) I don't think it is because any of the devices / ports / cables are broken or bad in any sense, because the problem occurs literally on *entirely different sets* of them (including host controller / ports). I remember it was first noticed on a few UAS-SATA enclosures that had an ASMedia ASM225CM (which is USB 3.x Gen 2) bridge, and the issue could be observed when the drive is connected to either the A port or the C port of my Intel laptop. And now I can observe similar problems on a few UAS-NVMe enclosures (which have either an ASMedia or a Realtek bridge). (And because of the SATA case, in which a WD Green SSD was inside, I don't think it has anything to do with insufficient / unstable port power or so either.) The only way I can workaround this problem is to use a Gen 1 / SuperSpeed cable. So far I haven't had any problems with any of my devices whenever they are working at SuperSpeed. I also have an AMD laptop with only Gen 1 ports, and they work perfectly fine on that host as well. I've also tried to disable UAS by using the "u quirk", but the problem remains, so it doesn't seem to have anything to do with that. So has anyone ever experienced / heard of similar issues? Anything I could try / test further in order to isolate the cause? Is there any existing XHCI driver quirk or so that could be relevant? P.S. Both machines have Arch Linux (not any derivative) on them, so I have been rolling through the year(s) with pretty-recent (and pretty-upstream, AFAIK) kernels. No LTS or anything. Regards, Tom
