[<prev] [next>] [day] [month] [year] [list]
Message-ID: <f55c36c5-0b8f-4b7a-a0b9-b11320781a4f@acm.org>
Date: Tue, 20 Feb 2024 10:19:45 -0800
From: Bart Van Assche <bvanassche@....org>
To: Rohit Ner <rohitner@...gle.com>
Cc: Can Guo <quic_cang@...cinc.com>, Bean Huo <beanhuo@...ron.com>,
Stanley Chu <stanley.chu@...iatek.com>,
"Martin K. Petersen" <martin.petersen@...cle.com>,
Jaegeuk Kim <jaegeuk@...nel.org>, Avri Altman <avri.altman@....com>,
linux-scsi@...r.kernel.org, linux-kernel@...r.kernel.org
Subject: Re: [PATCH] scsi: ufs: core: Fix setup_xfer_req invocation
On 2/20/24 09:58, Rohit Ner wrote:
> Do you propose to remove this callback altogether? This callback should
> either support both transfer modes or none.
The only UFSHCI controller I know of that needs this callback is the Exynos
UFSHCI controller. As far as I know there are Exynos UFSHCI controllers that
support UFSHCI 3.0 but UFSHCI 4.0 Exynos controllers are not yet available.
Standard compliant controllers should not use the .setup_xfer_req() callback.
As you may know invoking that callback means performing an indirect function
call. Indirect function calls are slower than direct calls, especially if
speculative execution vulnerability security mitigations are enabled.
Bart.
Powered by blists - more mailing lists