Received-SPF: pass (google.com: domain of linux-kernel+bounces-153750-linux.lists.archive=gmail.com@vger.kernel.org designates 2604:1380:40f1:3f00::1 as permitted sender) client-ip=2604:1380:40f1:3f00::1;
Date: Mon, 22 Apr 2024 10:49:21 -0600
From: Keith Busch <kbusch@kernel.org>
To: Sean Anderson <sean.anderson@linux.dev>
Cc: Jens Axboe <axboe@kernel.dk>, Christoph Hellwig <hch@lst.de>,
	Sagi Grimberg <sagi@grimberg.me>, linux-nvme@lists.infradead.org,
	linux-kernel@vger.kernel.org, stable@vger.kernel.org
Subject: Re: [PATCH] nvme-pci: Add quirk for broken MSIs
Message-ID: <ZiaVEfdUXO97eWzV@kbusch-mbp.dhcp.thefacebook.com>
References: <20240422162822.3539156-1-sean.anderson@linux.dev>
Precedence: bulk
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <20240422162822.3539156-1-sean.anderson@linux.dev>

On Mon, Apr 22, 2024 at 12:28:23PM -0400, Sean Anderson wrote:
> Sandisk SN530 NVMe drives have broken MSIs. On systems without MSI-X
> support, all commands time out resulting in the following message:
> 
> nvme nvme0: I/O tag 12 (100c) QID 0 timeout, completion polled
> 
> These timeouts cause the boot to take an excessively-long time (over 20
> minutes) while the initial command queue is flushed.
> 
> Address this by adding a quirk for drives with buggy MSIs. The lspci
> output for this device (recorded on a system with MSI-X support) is:

Based on your description, the patch looks good. This will fallback to
legacy emulated pin interrupts, and that's better than timeout polling,
but will still appear sluggish compared to MSI's. Is there an errata
from the vendor on this? I'm just curious if the bug is at the Device ID
level, and not something we could constrain to a particular model or
firmware revision. 
 
> 02:00.0 Non-Volatile memory controller: Sandisk Corp Device 5008 (rev 01) (prog-if 02 [NVM Express])
> 	Subsystem: Sandisk Corp Device 5008
> 	Flags: bus master, fast devsel, latency 0, IRQ 16, NUMA node 0
> 	Memory at f7e00000 (64-bit, non-prefetchable) [size=16K]
> 	Memory at f7e04000 (64-bit, non-prefetchable) [size=256]
> 	Capabilities: [80] Power Management version 3
> 	Capabilities: [90] MSI: Enable- Count=1/32 Maskable- 64bit+
> 	Capabilities: [b0] MSI-X: Enable+ Count=17 Masked-

Interesting, the MSI capability does look weird here. I've never seen
MSI-x count smaller than the MSI's. As long as both work, though, I
think nvme would actually prefer whichever is bigger!