[Bug]: proxy minions log "Error during asyncio shutdown: The future belongs to a different loop" on Python 3.14

Description

On Python 3.14 every proxy minion logs a wall of

[ERROR   ] Error during asyncio shutdown: The future belongs to a different loop than the one specified as the loop argument

during startup. On a freshly started single proxy that is the entire log: 20 lines, all of them this. A deltaproxy with two sub-proxies produces 58.

The message says "shutdown" but these appear while the proxy is coming up, which makes it look alarming and unrelated to what is actually happening.

Root cause

salt/utils/asynchronous.py, in close():

pending_tasks = [
    task for task in asyncio.all_tasks(self.asyncio_loop) if not task.done()
]
if pending_tasks:
    for task in pending_tasks:
        task.cancel()
    gathered = asyncio.gather(*pending_tasks, return_exceptions=True)
    try:
        self.asyncio_loop.run_until_complete(gathered)

asyncio.gather has taken no loop argument since 3.10, so it resolves the loop from the calling context. close() runs outside the loop it is tearing down, and the tasks belong to self.asyncio_loop. On Python 3.14 ensure_future rejects that mismatch; earlier versions took the loop from the first future and let it through.

Instrumenting the call site confirms the loops never match:

asyncio_loop=138915023434640 current=138915031292560 same=False ntasks=1
asyncio_loop=138914893251600 current=138915031292560 same=False ntasks=1
...

The exception is caught by the broad except below and logged, so nothing crashes -- but the pending tasks are never drained, which is the work close() was doing.

Setup

Python 3.14.7, salt 3008.2, Linux. Reproduces with a single proxy minion; a deltaproxy just produces more of them.

Steps to reproduce

Start any proxy minion under Python 3.14 and read its log.

Observed (single proxy, -l error, whole log):

[ERROR   ] Error during asyncio shutdown: The future belongs to a different loop than the one specified as the loop argument
... 20 identical lines ...

Expected: no such errors, and the pending tasks actually drained.

Scoped on the same machine, so it is specific to proxies rather than salt generally:

	occurrences
salt-call --local	0
salt-master	0
single salt-proxy	20
deltaproxy, 2 sub-proxies	58
any of the above on Python 3.10	0

Versions Report

Salt Version:
           Salt: 3008.2
Python Version:
         Python: 3.14.7

Reproduces against the current 3008.x branch head.

Start in salt/utils/asynchronous.py, focusing on close() and its asyncio shutdown handling. Reproduce the startup error with a proxy minion under Python 3.14, then verify the shutdown path against the loop that owns the pending tasks. Done means proxy startup emits no asyncio shutdown errors and the pending tasks are drained.