From 25a7b131b48841dcda5176eb7e4be47f77616226 Mon Sep 17 00:00:00 2001
From: Matthew Jackson <1085847+MattJackson@users.noreply.github.com>
Date: Tue, 30 Jun 2026 17:54:51 -0700
Subject: [PATCH] =?UTF-8?q?drive:=20Drive::spin=5Fcycle()=20=E2=80=94=20so?=
 =?UTF-8?q?ft=20un-wedge=20via=20START=20STOP=20UNIT=20(no=20eject)?=
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Spin the disc down (START=0, LOEJ=0) then back up (START=1) to clear the
BU40N/Initio fast-fail wedge state a run of HARDWARE_ERROR reads leaves the drive
in — the non-eject power-cycle our notes say the wedge needs. The disc stays
loaded; we never eject (slot-loading — a human eject is a product failure for an
unattended service). Validated live 2026-07-01: took the drive from
failing-every-read back to reading at MB/s.
---
 src/drive/mod.rs | 30 ++++++++++++++++++++++++++++++
 1 file changed, 30 insertions(+)

diff --git a/src/drive/mod.rs b/src/drive/mod.rs
index c86425e..3d3041b 100644
--- a/src/drive/mod.rs
+++ b/src/drive/mod.rs
@@ -51,6 +51,13 @@ pub enum DriveStatus {
 // SCSI opcodes used in drive control
 const SCSI_TEST_UNIT_READY: u8 = 0x00;
 const SCSI_START_STOP_UNIT: u8 = 0x1B;
+/// Idle time the disc sits spun-down during [`Drive::spin_cycle`] before it's
+/// spun back up — long enough for the mechanism's fast-fail wedge state to
+/// clear. Validated at 5–6 s live.
+const SPIN_DOWN_IDLE_SECS: u64 = 5;
+/// Settle time after spin-up in [`Drive::spin_cycle`] before the caller reads
+/// again, so the first post-cycle read doesn't hit a transient NOT_READY.
+const SPIN_UP_SETTLE_SECS: u64 = 10;
 const SCSI_PREVENT_ALLOW_MEDIUM_REMOVAL: u8 = 0x1E;
 const SCSI_GET_EVENT_STATUS: u8 = 0x4A;
 const SCSI_MODE_SENSE: u8 = 0x5A;
@@ -895,6 +902,29 @@ impl Drive {
         Ok(())
     }
 
+    /// Soft power-cycle the drive mechanism WITHOUT ejecting: spin the disc
+    /// down (`START STOP UNIT`, START=0, **LOEJ=0**) then back up (START=1).
+    /// This clears the BU40N/Initio fast-fail *wedge* state that a run of
+    /// `HARDWARE_ERROR` reads leaves the drive in — the non-eject equivalent of
+    /// the power-cycle our notes say the wedge needs. The disc stays loaded (the
+    /// BU40N is slot-loading; we NEVER eject to recover — a hands-on eject is a
+    /// failure for an unattended service). Validated live 2026-07-01: took the
+    /// drive from failing-every-read back to reading at MB/s.
+    pub fn spin_cycle(&mut self) -> Result<()> {
+        let stop = [SCSI_START_STOP_UNIT, 0, 0, 0, 0x00, 0]; // START=0, LOEJ=0 → spin down
+        let start = [SCSI_START_STOP_UNIT, 0, 0, 0, 0x01, 0]; // START=1, LOEJ=0 → spin up
+        let mut buf = [0u8; 0];
+        self.scsi
+            .as_mut()
+            .execute(&stop, crate::scsi::DataDirection::None, &mut buf, 30_000)?;
+        std::thread::sleep(std::time::Duration::from_secs(SPIN_DOWN_IDLE_SECS));
+        self.scsi
+            .as_mut()
+            .execute(&start, crate::scsi::DataDirection::None, &mut buf, 30_000)?;
+        std::thread::sleep(std::time::Duration::from_secs(SPIN_UP_SETTLE_SECS));
+        Ok(())
+    }
+
     pub fn scsi_execute(
         &mut self,
         cdb: &[u8],