USB connection timeouts when attempting to attach a Canon SELPHY CP1200

An example of the failure. This repeats endlessly:

04-06-2021 21:29:35:   ===============================
04-06-2021 21:29:35: + Bus 003 Device 015: added SELPHY CP1200
04-06-2021 21:29:35:   Device info:
04-06-2021 21:29:35:     Ident:         04a9-32b1-ZT15120100000091-Canon-SELPHY-CP1200
04-06-2021 21:29:35:     Manufacturer:  Canon
04-06-2021 21:29:35:     Product:       SELPHY CP1200
04-06-2021 21:29:35:     SerialNumber:  ZT15120100000091
04-06-2021 21:29:35:     MfgAndProduct: Canon SELPHY CP1200
04-06-2021 21:29:35:
04-06-2021 21:29:35:   Device quirks:
04-06-2021 21:29:35:
04-06-2021 21:29:35:   USB stack parameters
04-06-2021 21:29:35:     authorized_default:          -1
04-06-2021 21:29:35:     autosuspend:                 2
04-06-2021 21:29:35:     blinkenlights:               N
04-06-2021 21:29:35:     initial_descriptor_timeout:  5000
04-06-2021 21:29:35:     nousb:                       N
04-06-2021 21:29:35:     old_scheme_first:            N
04-06-2021 21:29:35:     quirks:
04-06-2021 21:29:35:     usbfs_memory_mb:             16
04-06-2021 21:29:35:     usbfs_snoop:                 N
04-06-2021 21:29:35:     usbfs_snoop_max:             65536
04-06-2021 21:29:35:     use_both_schemes:            Y
04-06-2021 21:29:35:
04-06-2021 21:29:35:   USB interfaces:
04-06-2021 21:29:35:     Config Interface Alt Class SubClass Proto
04-06-2021 21:29:35:        1       0      0   7      1       2
04-06-2021 21:29:35: *      1       0      1   7      1       4
04-06-2021 21:29:35:        1       1      0   255    0       0
04-06-2021 21:29:35: *      1       1      1   7      1       4
04-06-2021 21:29:35:
04-06-2021 21:29:35:   USB[0]: open: Bus 003 Device 015 Interface 0 Alt 1
04-06-2021 21:29:35:   USB[1]: open: Bus 003 Device 015 Interface 1 Alt 1
04-06-2021 21:29:35: > IPP request:
04-06-2021 21:29:35: > {
04-06-2021 21:29:35: >     VERSION 2.0
04-06-2021 21:29:35: >     OPERATION Get-Printer-Attributes
04-06-2021 21:29:35: >
04-06-2021 21:29:35: >     GROUP operation-attributes-tag
04-06-2021 21:29:35: >     ATTR "attributes-charset" charset: utf-8
04-06-2021 21:29:35: >     ATTR "attributes-natural-language" naturalLanguage: en-US
04-06-2021 21:29:35: >     ATTR "printer-uri" uri: http://localhost:60000/ipp/print
04-06-2021 21:29:35: >     ATTR "requested-attributes" keyword: all
04-06-2021 21:29:35: > }
04-06-2021 21:29:35:
04-06-2021 21:29:35: > HTTP[038]: POST http://localhost:60000/ipp/print
04-06-2021 21:29:35: > HTTP[038]: request body: got 151 bytes; closed
04-06-2021 21:29:35: > HTTP[038]: body is small (151 bytes), prefetched before sending
04-06-2021 21:29:35: > HTTP[038]: HTTP request header:
04-06-2021 21:29:35: >   POST /ipp/print HTTP/1.1
04-06-2021 21:29:35: >   Host: localhost:60000
04-06-2021 21:29:35: >   User-Agent: ipp-usb
04-06-2021 21:29:35: >   Content-Length: 151
04-06-2021 21:29:35: >   Content-Type: application/ipp
04-06-2021 21:29:35: >
04-06-2021 21:29:35:   USB[0]: connection allocated, 1 in use: a-- ---
04-06-2021 21:29:35:   HTTP[038]: connection 0 allocated
04-06-2021 21:29:35: > USB[0]: write: wanted 275 sent 275 total 275
04-06-2021 21:29:40: < USB[0]: read: wanted 4096 got 0 total 0
04-06-2021 21:29:40: ! USB[0]: recv: libusb_bulk_transfer: Operation timed out
04-06-2021 21:29:40: ! HTTP[038]: libusb_bulk_transfer: Operation timed out
04-06-2021 21:29:40:   USB[0]: connection released, 0 in use: --- ---
04-06-2021 21:29:40: ! IPP: HTTP: Post "http://localhost:60000/ipp/print": libusb_bulk_transfer: Operation timed out
04-06-2021 21:29:40: - Bus 003 Device 015: resetting SELPHY CP1200
04-06-2021 21:29:40:   USB[0]: closed
04-06-2021 21:29:40:   USB[1]: closed
04-06-2021 21:29:40: - Bus 003 Device 015: removed SELPHY CP1200

I'm using ipp-usb 0.9.19, on a Fedora 34 box.

It appears the printer exports two interface/configurations that claim to support ipp-over-usb, but ipp-usb only ever attempts to bind to the first...

EDIT: I mean to say we never attempt to use USB[1], even when USB[0] is failing. That seems like an oversight!

The issue names no files or tests; start by tracing the USB interface selection and timeout/reset path represented by USB[0] and USB[1] in the attached log. Done means the device can fall back to the second IPP-over-USB interface when the first times out, with the behavior covered by an appropriate test.