usbconfig: allow to get usb vendors and products from the usb.ids database

[bapt]

Summary

Allow to get usb vendors and products from the usb.ids databse.

I plan to add a share/misc/usb_vendors like we do for pciconf

Prefer port version as updated more often

The code is strongly based on pciconf one

[bapt]

Forgot to say:

before

ugen0.1: <XHCI root HUB 0x8086> at usbus0, cfg=0 md=HOST spd=SUPER (5.0Gbps) pwr=SAVE (0mA)
ugen0.2: <Lenovo EasyCamera Generic> at usbus0, cfg=0 md=HOST spd=HIGH (480Mbps) pwr=ON (500mA)
ugen0.3: <USB2.0 Hub vendor 0x05e3> at usbus0, cfg=0 md=HOST spd=HIGH (480Mbps) pwr=SAVE (100mA)
ugen0.4: <product 0x3004 vendor 0x0cf3> at usbus0, cfg=0 md=HOST spd=FULL (12Mbps) pwr=ON (100mA)
ugen0.5: <USB2.0-CRW Generic> at usbus0, cfg=0 md=HOST spd=HIGH (480Mbps) pwr=ON (500mA)

After:

ugen0.1: <XHCI root HUB 0x8086> at usbus0, cfg=0 md=HOST spd=SUPER (5.0Gbps) pwr=SAVE (0mA)
ugen0.2: <Lenovo EasyCamera Acer, Inc> at usbus0, cfg=0 md=HOST spd=HIGH (480Mbps) pwr=ON (500mA)
ugen0.3: <4-port hub Genesys Logic, Inc.> at usbus0, cfg=0 md=HOST spd=HIGH (480Mbps) pwr=SAVE (100mA)
ugen0.4: <AR3012 Bluetooth 4.0 Atheros Communications, Inc.> at usbus0, cfg=0 md=HOST spd=FULL (12Mbps) pwr=ON (100mA)
ugen0.5: <RTS5129 Card Reader Controller Realtek Semiconductor Corp.> at usbus0, cfg=0 md=HOST spd=HIGH (480Mbps) pwr=ON (500mA)

[• hselasky]

lib/libusb/libusb20_ugen20.c

53 ↗

(On Diff #7969)

Maybe putting the vendor info only is enough?

How about using STAILQ_XXX() instead of TAILQ_XXX() to save a pointer per entry?

406 ↗

(On Diff #7969)

Hi,

Use this way:

while ((vi = TAILQ_FIRST(uvendors) ) != NULL) {
TAILQ_REMOVE(...);

}

430 ↗

(On Diff #7969)

Loading the product and vendors for every USB access would be overkill. The table is quite big. I would make it a utility function:

int
libusb20_str_vendor(uint16_t, const char **);

int
libusb20_str_product(uint16_t, const char **);

And only load when those two functions are called.

--HPS

[imp]

I generally like it. TAILQ vs STAILQ I'm agnostic on, though I don't see a need to traverse the list backwards ever...

[• hselasky]

Ping - any updates?

[emaste]

This would be nice to have

[bapt]

rework entirely to not clutter libusb and only be in the frontend

[imp]

Just the one question... otherwise it looks good

usr.sbin/usbconfig/dump.c

349

Is conputing the address 1 before the buf well defined?

[emaste]

usr.sbin/usbconfig/dump.c

331–337

the priority is slightly odd e.g. if you don't set USB_VENDOR_DATABASE you get the /usr/local/file (assuming it exists), but if you set USB_VENDOR_DATABASE to a nonexistent file you get the non-ports version. In practice it probably doesn't matter too much, but it is a little odd.

[bapt]

address @emaste comments

[bapt]

Address @imp case

[bapt]

provided the last messages looks scary I am not giving up on this, just I don't want to be my own reviewer