docs: drop the usbh enum-timeout handoff - its fix is PR #3862etmtrace-rp2350

1 files changed, 0 insertions, 33 deletions

diff --git a/docs/superpowers/followup/pr3851-usbh-enum-timeout.md b/docs/superpowers/followup/pr3851-usbh-enum-timeout.md
deleted file mode 100644
index 5dc7b1d35..000000000
--- a/docs/superpowers/followup/pr3851-usbh-enum-timeout.md
+++ /dev/null
@@ -1,33 +0,0 @@
-# usbh: enumeration control-transfer timeout
-
-Split out of PR #3851 (found while debugging the rig's pico PIO-USB fixture, 2026-08-27).
-
-## Established
-
-- `src/host/usbh.c` (`TODO timeout_ms is not supported yet`, ~line 873) — enumeration
-  control transfers have no timeout.
-- Observed wedge on real hardware (rig ci.lan, raspberry_pi_pico PIO-USB host,
-  GDB autopsy in ci.lan:`~/dbg-pico/`): a fixture device (048d:04d2) ACKs the SETUP of
-  the first 8-byte GET_DESCRIPTOR and never answers the EP0 IN data stage. State
-  freezes at `_usbh_data.enumerating_daddr = 0`, `ctrl_xfer_info.stage = DATA`,
-  `failed_count = 0`; the PIO-USB ep pool keeps `daddr 0 ep 0x80 has_transfer=1`
-  armed forever.
-- Consequence: ALL enumeration stops — including the parent hub's interrupt-EP
-  status polling — so one dead device blanks every other port. Core stays healthy
-  (main loop runs).
-- Only the NAK-forever case wedges: Pico-PIO-USB already errors out after
-  `TRANSACTION_MAX_RETRY` for a *silent* device.
-
-## Remains
-
-- Add a timeout to enumeration control transfers (or a watchdog on
-  `enumerating_daddr`): on expiry, fail the enumeration, free dev0/EP0, and let the
-  hub continue polling. Respect the existing retry/error path so a normal slow
-  device is not broken.
-- Coverage: needs a HIL-style repro (device that ACKs SETUP then NAKs IN forever);
-  consider a unit test on the usbh control state machine if the timeout is tick-based.
-
-## Why split
-
-PR #3851 is a board+clock PR; this is a host-stack robustness fix with its own
-review surface and test needs.