> My best guess is that these are caused by a full xhci event ring. I'm not sure. I'm testing with a full-speed audio dongle which has two endpoints with one packet per millisecond each. Doesn't sound like much. There is nothing more on this bus, besides hubs (part of the PCIe card). (Is it possible that the hubs are causing issues due to TT or whatnot?) /: Bus 008.Port 001: Dev 001, Class=root_hub, Driver=xhci_hcd/2p, 480M |__ Port 001: Dev 002, If 0, Class=Hub, Driver=hub/4p, 480M |__ Port 002: Dev 004, If 0, Class=Audio, Driver=snd-usb-audio, 12M |__ Port 002: Dev 004, If 1, Class=Audio, Driver=snd-usb-audio, 12M |__ Port 002: Dev 004, If 2, Class=Audio, Driver=snd-usb-audio, 12M |__ Port 002: Dev 004, If 3, Class=Human Interface Device, Driver=usbhid, 12M |__ Port 002: Dev 003, If 0, Class=Hub, Driver=hub/4p, 480M /: Bus 009.Port 001: Dev 001, Class=root_hub, Driver=xhci_hcd/2p, 10000M |__ Port 001: Dev 002, If 0, Class=Hub, Driver=hub/4p, 10000M |__ Port 002: Dev 003, If 0, Class=Hub, Driver=hub/4p, 10000M I tried 6.11-rc4 and cherry-picking 28084d3fcc3c on 6.6. Neither works. An archive with short traces and corresponding dmesg logs is attached. One run on vanilla v6.11-rc4, one run with my patches (don't skip on underrun, do skip immediately on missed service if ep_trb_dma != NULL). Regards, Michal
Attachment:
traces.tgz
Description: application/compressed-tar