feat(midi2): opt-in app responder for UMP Stream

[sauloverissimo]

Summary

Small, opt-in change, quick to review. Touches only the MIDI 2.0 device class
(legacy MIDI 1.0 driver untouched) and is byte-identical to current behavior
when the flag is off.

Adds CFG_TUD_MIDI2_USER_RESPONDER (default off). When set, the driver leaves
UMP Stream messages in the RX FIFO so the application can answer Endpoint /
Function Block discovery and Stream Configuration itself, for endpoints whose
topology is dynamic (e.g. a bridge whose function blocks track the devices
plugged into it). Built-in responder is unchanged when off.

Test plan

• Builds with the flag off (default) and on (RP2350 midi2_device example).
• Used on hardware by an ESP32-P4 USB MIDI bridge to expose dynamic FB names/groups.

Ref #3571

[Copilot]

Pull request overview

Adds an opt-in build-time switch to the MIDI 2.0 device class that lets applications take over UMP Stream (MT=0xF) discovery/configuration handling, while preserving current driver behavior when the option is disabled.

Changes:

• Introduces CFG_TUD_MIDI2_USER_RESPONDER (default 0) configuration define.
• When enabled, disables the built-in UMP Stream responder and leaves UMP Stream messages in the RX FIFO for the application to read/handle.

Reviewed changes

Copilot reviewed 2 out of 2 changed files in this pull request and generated 2 comments.

File	Description
src/class/midi/midi2_device.h	Adds the new opt-in configuration macro default.
src/class/midi/midi2_device.c	Gates the built-in UMP Stream negotiation responder and skips consuming stream messages when user-responder mode is enabled.

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

[chatgpt-codex-connector]

💡 Codex Review

Here are some automated review suggestions for this pull request.

Reviewed commit: 5f8a734576

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

• Open a pull request for review
• Mark a draft as ready
• Comment "@codex review".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".

[chatgpt-codex-connector]

Preserve negotiation state in user responder mode

With CFG_TUD_MIDI2_USER_RESPONDER=1, this branch returns before the only code path that handles STREAM_CONFIG_REQUEST; since there is also no public setter for p_midi->protocol or p_midi->negotiated, an application that accepts a host's MIDI 1.0 Config Request and sends its own Config Notify will still see tud_midi2_n_protocol() report the reset default MIDI2 and tud_midi2_n_negotiated() stay false. Please either keep parsing config requests for state or provide a way for the user responder to update these getters.

Useful? React with 👍 / 👎.

[sauloverissimo]

Updated to address the automated review: added tud_midi2_n_set_protocol() so the application can keep tud_midi2_n_protocol / tud_midi2_n_negotiated accurate when it owns UMP Stream negotiation, plus header/.c notes on draining the RX FIFO. Builds with the flag off (default) and on.

[github-actions]

Hardware-in-the-loop (HIL) Test Report

hfp.json

✅ 52 passed · ❌ 0 failed · ⚪ 0 skipped · blank not run

Board	cdc_dual_ports	cdc_msc	dfu	cdc_msc_throughput	audio_test_freertos	dfu_runtime	cdc_msc_freertos	hid_boot_interface	msc_dual_lun	hid_generic_inout	printer_to_cdc	midi_test	mtp
stm32l412nucleo	✅	✅	✅	✅ CDC 628k/403k MSC 825k/767k	✅	✅	✅	✅	✅	✅	✅	✅	✅
stm32f746disco	✅	✅	✅	✅ CDC 13.7M/10M MSC 15.4M/28.8M	✅	✅	✅	✅	✅	✅	✅	✅	✅
stm32f746disco-DMA	✅	✅	✅	✅ CDC 13.4M/9.2M MSC 18M/29.8M	✅	✅	✅	✅	✅	✅	✅	✅	✅
lpcxpresso43s67	✅	✅	✅	✅ CDC 12.6M/11.8M MSC 29.6M/32.5M	✅	✅	✅	✅	✅	✅	✅	✅	✅

tinyusb.json

✅ 200 passed · ❌ 138 failed · ⚪ 11 skipped · blank not run

Board	cdc_dual_ports	cdc_msc	dfu	cdc_msc_throughput	audio_test_freertos	dfu_runtime	cdc_msc_freertos	hid_boot_interface	msc_dual_lun	hid_generic_inout	printer_to_cdc	midi_test	mtp	host_info_to_device_cdc	cdc_msc_hid	msc_file_explorer	msc_file_explorer_freertos	device_info	hid_composite_freertos
ek_tm4c123gxl	✅	✅	✅	✅ CDC 849k/1M MSC 825k/941k	✅	✅	✅	✅	✅	✅	✅	✅	✅
espressif_p4_function_ev					✅		✅										rd 409KB/s	✅	✅
espressif_p4_function_ev-DMA					✅		✅										rd 409KB/s	✅	✅
espressif_s3_devkitm					✅		✅										rd 409KB/s	✅	✅
espressif_s3_devkitm-DMA					✅		✅										rd 409KB/s	✅	✅
feather_nrf52840_express	❌	❌	❌	❌	❌	❌	❌	❌	❌	❌	❌	❌	❌
max32666fthr	❌	❌	❌	❌	❌	❌	❌	❌	❌	❌	❌	❌	❌
metro_m4_express	❌	❌	❌	❌		❌	❌	❌	❌	❌	❌	❌	❌	❌
mimxrt1015_evk	❌	❌	❌	❌	❌	❌	❌	❌	❌	❌	❌	❌	❌
mimxrt1064_evk	❌	❌	❌	❌	❌	❌	❌	❌	❌	❌	❌	❌	❌	❌	❌	❌	❌	❌
lpcxpresso11u37	❌	❌	❌	❌	❌	❌	❌	❌	❌	❌	❌	❌	❌
ra4m1_ek	❌	❌	❌	❌	⚪	❌	⚪	❌	❌	❌	❌	❌	❌
raspberry_pi_pico	❌	❌	❌	❌	❌	❌	❌	❌	❌	❌	❌	❌	❌	❌	❌	❌	❌	❌
raspberry_pi_pico_w															❌	❌	❌	❌
raspberry_pi_pico2															❌	❌	❌	❌
adafruit_fruit_jam	❌	❌	❌	❌	❌	❌	❌	❌	❌	❌	❌	❌	❌	❌	❌	❌	❌	❌
stm32f072disco	✅	✅	✅	✅ CDC 466k/333k MSC 617k/499k	✅	✅	✅	✅	✅	✅	✅	✅	✅
stm32f407disco	⚪	✅	✅	✅ CDC 921k/496k MSC 830k/916k	✅	✅	✅	✅	✅	✅	✅	✅	✅
stm32f723disco	✅	✅	✅	✅ CDC 738k/776k MSC 704k/921k	✅	✅	✅	✅	✅	✅	✅	✅	✅		✅	rd 18078KB/s	rd 4096KB/s	✅
stm32f723disco-DMA	✅	✅	✅	✅ CDC 922k/786k MSC 921k/986k	✅	✅	✅	✅	✅	✅	✅	✅	✅		✅	rd 19418KB/s	rd 4096KB/s	✅
stm32h743nucleo	✅	✅	✅	✅ CDC 838k/864k MSC 941k/1M	✅	✅	✅	✅	✅	✅	✅	✅	✅
stm32h743nucleo-DMA	✅	✅	✅	✅ CDC 137k/640k MSC 908k/669k	✅	✅	✅	✅	✅	✅	✅	✅	✅
stm32g0b1nucleo	✅	✅	✅	✅ CDC 488k/485k MSC 564k/538k	✅	✅	✅	✅	✅	✅	✅	✅	✅
stm32l476disco	✅	✅	✅	✅ CDC 509k/506k MSC 551k/907k	✅	✅	✅	✅	✅	✅	✅	✅	✅
stm32u083nucleo	✅	✅	✅	✅ CDC 370k/479k MSC 636k/542k	✅	✅	✅	✅	✅	✅	✅	✅	✅
nanoch32v203-fsdev	✅	✅	✅	✅ CDC 754k/181k MSC 775k/824k	⚪	✅	⚪	✅	✅	✅	✅	✅	✅
nanoch32v203-usbfs	✅	✅	✅	✅ CDC 663k/684k MSC 537k/970k	⚪	✅	⚪	✅	✅	✅	✅	✅	✅
ch32v103r_r1_1v0	✅	✅	✅		⚪	✅	⚪	✅	✅	✅	✅	✅	✅
ch582m_evt	✅	✅	✅	✅ CDC 231k/219k MSC 434k/463k	⚪	✅	⚪	✅	✅	✅	✅	✅	✅

[github-actions]

MemBrowse Memory Report

No memory changes detected across 2404 targets. View Project Dashboard →

[HiFiPhile]

I'm thinking about adding a weak callback here, with the return value we can differentiate between Handled, Negotiated to V1 , Negotiated to V2 , etc.

Maybe the user only want to override a one specific response instead of implement full negotiation in app.

[sauloverissimo]

Hi @HiFiPhile! Are you thinking of a per-message hook at the stream dispatch, where the return value decides fallthrough vs handled? So the app could intercept just FB Discovery, say, and let the built-in responder take the rest.

[HiFiPhile]

Yes exactely.

[sauloverissimo]

Closing this in favor of #3738. While I was working through the function block changes there, I ended up adding a per-message callback (tud_midi2_stream_msg_cb, returning PASS / HANDLED / NEGOTIATED_*) that does what the CFG_TUD_MIDI2_USER_RESPONDER flag here does, but per message and without a compile-time switch. That made this one redundant.