[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <20210323083749.r272grolxozf3w2f@mpHalley.local>
Date: Tue, 23 Mar 2021 09:37:49 +0100
From: Javier González <javier.gonz@...sung.com>
To: Christoph Hellwig <hch@....de>
CC: Keith Busch <kbusch@...nel.org>,
Dmitry Monakhov <dmtrmonakhov@...dex-team.ru>,
<linux-kernel@...r.kernel.org>, <linux-nvme@...ts.infradead.org>,
<Chaitanya.Kulkarni@....com>
Subject: Re: [PATCH 1/1] nvme-pci: add the DISABLE_WRITE_ZEROES quirk for a
Samsung PM1725a
On 11.03.2021 11:47, Christoph Hellwig wrote:
>On Wed, Mar 10, 2021 at 12:00:30PM -0800, Keith Busch wrote:
>> On Wed, Mar 10, 2021 at 02:41:10PM +0100, Christoph Hellwig wrote:
>> > On Wed, Mar 10, 2021 at 02:21:56PM +0100, Christoph Hellwig wrote:
>> > > Can you try this patch instead?
>> > >
>> > > http://lists.infradead.org/pipermail/linux-nvme/2021-February/023183.html
>> >
>> > Actually, please try the patch below instead, it looks like our existing
>> > logic messes up the units:
>>
>> This seems like a good opportunity to incorporate TP4040 for non-MDTS
>> command limits. I sent a proposal here
>>
>> http://lists.infradead.org/pipermail/linux-nvme/2021-March/023522.html
>>
>> And it defaults to your suggestion if the controller doesn't implement
>> the capability.
>
>I think TP4040 is good and useful, but I defintively want the pure
>fix get in first.
Quick question. It seems like the current quirk simply disables
write-zeroes. Would you be open for a quirk that aligns with MDTS for
models that implemented it this way before TP4040?
Powered by blists - more mailing lists