[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <20221028165116.GC9077@asutoshd-linux1.qualcomm.com>
Date: Fri, 28 Oct 2022 09:51:16 -0700
From: Asutosh Das <quic_asutoshd@...cinc.com>
To: Bart Van Assche <bvanassche@....org>
CC: <quic_cang@...cinc.com>, <martin.petersen@...cle.com>,
<linux-scsi@...r.kernel.org>, <quic_nguyenb@...cinc.com>,
<quic_xiaosenh@...cinc.com>, <stanley.chu@...iatek.com>,
<eddie.huang@...iatek.com>, <daejun7.park@...sung.com>,
<avri.altman@....com>, <mani@...nel.org>, <beanhuo@...ron.com>,
<quic_richardp@...cinc.com>, <linux-arm-msm@...r.kernel.org>,
Alim Akhtar <alim.akhtar@...sung.com>,
"James E.J. Bottomley" <jejb@...ux.ibm.com>,
"open list" <linux-kernel@...r.kernel.org>
Subject: Re: [PATCH v3 03/17] ufs: core: Introduce Multi-circular queue
capability
On Thu, Oct 27 2022 at 14:10 -0700, Bart Van Assche wrote:
>On 10/20/22 11:03, Asutosh Das wrote:
>>+
>>+ /*
>>+ * This capability allows the host controller driver to turn on/off
>>+ * MCQ mode. MCQ mode may be used to increase performance.
>>+ */
>>+ UFSHCD_CAP_MCQ_EN = 1 << 12,
>> };
>
>I prefer that the above flag would only be introduced after a need
>appears to disable MCQ, e.g. discovery of a host controller that is
>too broken to enable MCQ.
>
I think this flag is useful during chip bring-ups &/or debugs.
Considering that this is a fairly new hardware block even though the HC supports
MCQ there could be times to keep it disabled.
>Since UFSHCI 4.0 controllers must support UFSHCI 3.0, shouldn't users
>have a way to disable MCQ, e.g. via a kernel module parameter?
>
I can add a kernel module parameter to disable MCQ.
I think adding it to host driver (e.g. ufs-qcom) would be good.
Please let me know if you have a better place in mind.
>Thanks,
>
>Bart.
Powered by blists - more mailing lists