Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::1:20 as permitted sender) client-ip=2620:137:e000::1:20;
Date:   Wed, 11 Jan 2023 05:54:02 +0100
From:   Christoph Hellwig <hch@lst.de>
To:     Hector Martin <marcan@marcan.st>
Cc:     Keith Busch <kbusch@kernel.org>, Jens Axboe <axboe@fb.com>,
        Christoph Hellwig <hch@lst.de>,
        Sagi Grimberg <sagi@grimberg.me>,
        Eric Curtin <ecurtin@redhat.com>, Janne Grunau <j@jannau.net>,
        Sven Peter <sven@svenpeter.dev>,
        Alyssa Rosenzweig <alyssa@rosenzweig.io>,
        asahi@lists.linux.dev, linux-arm-kernel@lists.infradead.org,
        linux-nvme@lists.infradead.org, linux-kernel@vger.kernel.org
Subject: Re: [PATCH 1/2] nvme-apple: Do not try to shut down the controller
 twice
Message-ID: <20230111045402.GB15520@lst.de>
References: <20230111043614.27087-1-marcan@marcan.st> <20230111043614.27087-2-marcan@marcan.st>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <20230111043614.27087-2-marcan@marcan.st>
User-Agent: Mutt/1.5.17 (2007-11-01)
Precedence: bulk

On Wed, Jan 11, 2023 at 01:36:13PM +0900, Hector Martin wrote:
> The blamed commit stopped explicitly disabling the controller when we do
> a controlled shutdown, but apple_nvme_reset_work was only checking for
> the disable bit before deciding to issue another disable. Check for the
> shutdown state too, to avoid breakage.
> 
> This issue does not affect nvme-pci, since it only issues controller
> shutdowns when the system is actually shutting down anyway.

There's a few other places where nvme-pci does a shutdown like
probe/reset failure and most notably and mostly notably various
power management scenarios.

What path is causing a problem here for nvme-apple?  I fear we're
missing some highler level check here and getting further out of
sync.