From ee002b1db6407d8e36e790313f0393b2b7e025dc Mon Sep 17 00:00:00 2001
From: Jacob Dahl <37091262+dakejahl@users.noreply.github.com>
Date: Wed, 20 May 2026 21:08:53 -0400
Subject: [PATCH] fix(tools/uploader): handle USB CDC reconnect race during
 reboot (#27419)

After sending reboot-to-bootloader, the PX4 USB CDC node briefly
disappears while the bootloader re-enumerates. Reopening the serial
port can land on a half-broken descriptor and the next tcdrain()
raises termios.error (5, 'Input/output error'). That bare OSError
escaped every retry layer and crashed the uploader, even though a
manual re-run would succeed once enumeration settled.

Convert OSError/SerialException from flush() and reset_buffers() into
the module's ConnectionError, matching how send()/recv() already
behave, and let the identify retry loops in _try_identify also catch
ConnectionError so a single transient I/O hiccup doesn't abort the
upload.

Signed-off-by: Jacob Dahl <dahl.jakejacob@gmail.com>
---
 Tools/px4_uploader.py | 29 ++++++++++++++++++++++++-----
 1 file changed, 24 insertions(+), 5 deletions(-)

diff --git a/Tools/px4_uploader.py b/Tools/px4_uploader.py
index a6eb3fa604..c4dcfc754f 100755
--- a/Tools/px4_uploader.py
+++ b/Tools/px4_uploader.py
@@ -520,14 +520,31 @@ class SerialTransport:
 
     def flush(self) -> None:
         """Flush output buffer."""
-        if self._port is not None:
+        if self._port is None:
+            return
+        # tcdrain() can raise termios.error (a bare OSError) if the device
+        # disappeared mid-flush — common right after a reboot-to-bootloader
+        # when the USB CDC node is being torn down and re-enumerated.
+        try:
             self._port.flush()
+        except (OSError, serial.SerialException) as e:
+            raise ConnectionError(
+                f"Flush failed: {e}", port=self.port_name, operation="flush"
+            )
 
     def reset_buffers(self) -> None:
         """Reset input and output buffers."""
-        if self._port is not None:
+        if self._port is None:
+            return
+        try:
             self._port.reset_input_buffer()
             self._port.reset_output_buffer()
+        except (OSError, serial.SerialException) as e:
+            raise ConnectionError(
+                f"Buffer reset failed: {e}",
+                port=self.port_name,
+                operation="reset_buffers",
+            )
 
     def set_baudrate(self, baudrate: int) -> None:
         """Change baud rate.
@@ -1737,7 +1754,7 @@ class Uploader:
                 port=transport.port_name,
             )
             return True
-        except (ProtocolError, TimeoutError):
+        except (ProtocolError, TimeoutError, ConnectionError):
             pass
 
         # Try rebooting at each baud rate
@@ -1802,8 +1819,10 @@ class Uploader:
                         port=transport.port_name,
                     )
                     return True
-                except (ProtocolError, TimeoutError):
-                    # Board may still be rebooting, wait a bit and retry
+                except (ProtocolError, TimeoutError, ConnectionError):
+                    # Board may still be rebooting, wait a bit and retry.
+                    # ConnectionError covers the USB CDC node briefly going
+                    # away as the bootloader re-enumerates.
                     time.sleep(0.3)
 
         return False