From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from mga09.intel.com (mga09.intel.com. [134.134.136.24]) by gmr-mx.google.com with ESMTPS id f13si733257iog.3.2021.05.21.08.22.17 for (version=TLS1_2 cipher=ECDHE-ECDSA-AES128-GCM-SHA256 bits=128/128); Fri, 21 May 2021 08:22:17 -0700 (PDT) Subject: Re: [RFC] ntb/plx: Add support for PLX hardware References: <20210520222323.104901-1-jkirsher@ixsystems.com> <533937f2-7d41-7bfc-9323-e67c8d4878c9@deltatee.com> From: Dave Jiang Message-ID: <2275dc4e-fea1-dafd-c256-0f6f32f9e3fc@intel.com> Date: Fri, 21 May 2021 08:22:14 -0700 MIME-Version: 1.0 In-Reply-To: <533937f2-7d41-7bfc-9323-e67c8d4878c9@deltatee.com> Content-Type: text/plain; charset=utf-8; format=flowed Content-Transfer-Encoding: 7bit Content-Language: en-US To: Logan Gunthorpe , Jeff Kirsher , jdmason@kudzu.us, allenbh@gmail.com Cc: linux-ntb@googlegroups.com, Alexander Motin List-ID: On 5/20/2021 6:13 PM, Logan Gunthorpe wrote: > > > On 2021-05-20 6:17 p.m., Jeff Kirsher wrote: >> Is this something new that all NTB drivers are expected to adopt? Just >> asking because I used the Intel and AMD drivers as examples and they >> are still using pci_iomap(), etc... > I can't speak to expectations, I'm not the gatekeeper, only a previewer > passing by. > > I believe they are newer than the AMD and Intel drivers and there are > new ones added here and there. I think generally, kernel wide, they're a > better choice. You'll find your code will be a lot cleaner, smaller and > easier to understand if you adopt them. The Intel driver is pre devm allocators. Maybe one of these days when I have time I'll clean that up and covert it. But I agree with Logan that things will be cleaner and smaller with devm_ calls. > Logan