[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <3191de59-16bb-4c54-87c3-110a0285ac13@bjorling.me>
Date: Wed, 9 Oct 2024 15:09:18 +0200
From: Matias Bjørling <m@...rling.me>
To: Christoph Hellwig <hch@....de>
Cc: kbusch@...nel.org, dlemoal@...nel.org, cassel@...nel.org,
linux-nvme@...ts.infradead.org, linux-block@...r.kernel.org,
linux-kernel@...r.kernel.org, Matias Bjørling
<matias.bjorling@....com>
Subject: Re: [PATCH 2/2] nvme: add rotational support
On 09-10-2024 09:48, Christoph Hellwig wrote:
> On Tue, Oct 08, 2024 at 04:55:03PM +0200, Matias Bjørling wrote:
>> + if (info->is_rotational)
>> + lim.features |= BLK_FEAT_ROTATIONAL | BLK_FEAT_ADD_RANDOM;
>
> Entropy from block devices is pretty useless. The only reason we still
> keep it for SCSI is because of retro-computing platforms without a proper
> platform hardware RNG. NVMe HDDs reall should not show up in those kinds
> of environments. Also without a add_disk_randomness in the nvme I/O
> completion handler this won't actually do anything.
>
Thanks for the details. I'll remove it in the next revision.
Powered by blists - more mailing lists