Attempting to re-derive `SizeOf` causes panic and kernel errors

Prerequisites

Please put an X between the brackets as you perform the following steps:

• Check that your issue is not already filed:
  https://github.com/leanprover/lean4/issues
• Reduce the issue to a minimal, self-contained, reproducible test case.
  Avoid dependencies to Mathlib or Batteries.
• Test your test case against the latest nightly release, for example on
  https://live.lean-lang.org/#project=lean-nightly
  (You can also use the settings there to switch to “Lean nightly”)

Description

Attempting to derive the class SizeOf for a type for which it already exists causes multiple panics and kernel errors.

Context

This issue arose when writing a test for #9371 which attempted to run all registered deriving handlers on a given type.

Steps to Reproduce

Consider the following:

structure Foo where  -- PANIC: duplicate normalized declaration name Foo._sizeOf_1 vs. Foo._sizeOf_1 (+ 5 more)
  deriving SizeOf  -- error: (kernel) constant has already been declared 'Foo._sizeOf_1' (+ 2 more)

inductive Enum where
  | a | b

deriving instance SizeOf for Enum  -- 8 similar panics, 4 similar kernel errors

Expected behavior: Most likely, an error message indicates that this instance cannot be rederived.

Actual behavior: A large number of panics and kernel errors appear, as shown above.

Versions

Lean 4.23.0-nightly-2025-07-21

Additional Information

This is similar to #8258 inasmuch as it is a form of "failing to check environment when auto-generating constant" error.

Impact

Add 👍 to issues you consider important. If others are impacted by this issue, please ask them to add 👍 to it.

Start with the minimal reproducer in the issue: derive SizeOf for Foo and then for Enum, and run it against the latest Lean nightly. Inspect the deriving and auto-generated declaration path implicated by the duplicate Foo._sizeOf_1 errors. Done means the case reports a clear re-derivation error instead of panics or kernel errors.