Re: panic in usb_detach_device / device_printf
- In reply to: Warner Losh : "Re: panic in usb_detach_device / device_printf"
 - Go to: [ bottom of page ] [ top of archives ] [ this month ]
 
Date: Sun, 11 May 2025 14:49:15 UTC
On Sat, 10 May 2025, Warner Losh wrote: > OK. Sorry for the top post. > > I need to recreate this because this data is already freed and corrupted. > > So is this panic in bhyve? What's the bhyve configuration you are using? > You say this happens as you move the xhci pass through device back to > the host. Does the same thing happen if you instead devctl detach the xhciX > the device instead? It looks like you have a da0 device on the usb bus, any > others? "Seems xhci -> ppt" So this is on the base system moving the XHCI to ppt to be used in bhyve, otherwise it wouldn't detach all the USB devices but create them. I booted a non-FreeBSD guest. The da0 is created by a blank adapter in the ugen0.7: <Card Reader Realtek Semiconductor Corp.> at usbus0, cfg=0 md=HOST spd=SUPER (5.0Gbps) pwr=ON (200mA) <Generic- SD/MMC 1.00> at scbus1 target 0 lun 0 (da0,pass1) There's no card there. But it's been like this for years; even before the reader was supported and I've done this regularly. > I need a way to reproduce it, and I kinda get what you're doing, but step > by step instructions would be way better... And it may be a week before > I get to it: my daughter is graduating next Friday, so my whole routine and > schedule is off. # pciconf -l pci0:0:20:0 xhci0@pci0:0:20:0: class=0x0c0330 rev=0x21 hdr=0x00 vendor=0x8086 device=0x9d2f subvendor=0x17aa subdevice=0x225d I cannot reproduce it anymore. history is gone sadly. I do not know what I typed anymore. My bhyve.sh has these as comments so I likely just pasted them when bhyve refused to start because of no ppt0 not being there already. # devctl detach pci0:0:20:0 || true # devctl set driver pci0:0:20:0 ppt || true /bz -- Bjoern A. Zeeb r15:7