Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 23.128.96.18 as permitted sender) client-ip=23.128.96.18;
Date:   Wed, 23 Jun 2021 15:24:35 -0300
From:   Jason Gunthorpe <jgg@ziepe.ca>
To:     Christian =?utf-8?B?S8O2bmln?= <ckoenig.leichtzumerken@gmail.com>
Cc:     Christian =?utf-8?B?S8O2bmln?= <christian.koenig@amd.com>,
        Oded Gabbay <oded.gabbay@gmail.com>,
        Gal Pressman <galpress@amazon.com>, sleybo@amazon.com,
        linux-rdma <linux-rdma@vger.kernel.org>,
        Oded Gabbay <ogabbay@kernel.org>,
        Christoph Hellwig <hch@lst.de>,
        Linux Kernel Mailing List <linux-kernel@vger.kernel.org>,
        dri-devel <dri-devel@lists.freedesktop.org>,
        "moderated list:DMA BUFFER SHARING FRAMEWORK" 
        <linaro-mm-sig@lists.linaro.org>,
        Doug Ledford <dledford@redhat.com>,
        Tomer Tayar <ttayar@habana.ai>,
        amd-gfx list <amd-gfx@lists.freedesktop.org>,
        Greg KH <gregkh@linuxfoundation.org>,
        Alex Deucher <alexander.deucher@amd.com>,
        Leon Romanovsky <leonro@nvidia.com>,
        "open list:DMA BUFFER SHARING FRAMEWORK" 
        <linux-media@vger.kernel.org>
Subject: Re: [Linaro-mm-sig] [PATCH v3 1/2] habanalabs: define uAPI to export
 FD for DMA-BUF
Message-ID: <20210623182435.GX1096940@ziepe.ca>
References: <d358c740-fd3a-9ecd-7001-676e2cb44ec9@gmail.com>
 <CAFCwf11h_Nj_GEdCdeTzO5jgr-Y9em+W-v_pYUfz64i5Ac25yg@mail.gmail.com>
 <20210622120142.GL1096940@ziepe.ca>
 <d497b0a2-897e-adff-295c-cf0f4ff93cb4@amd.com>
 <20210622152343.GO1096940@ziepe.ca>
 <3fabe8b7-7174-bf49-5ffe-26db30968a27@amd.com>
 <20210622154027.GS1096940@ziepe.ca>
 <09df4a03-d99c-3949-05b2-8b49c71a109e@amd.com>
 <20210622160538.GT1096940@ziepe.ca>
 <d600a638-9e55-6249-b574-0986cd5cea1e@gmail.com>
MIME-Version: 1.0
Content-Type: text/plain; charset=utf-8
Content-Disposition: inline
Content-Transfer-Encoding: 8bit
In-Reply-To: <d600a638-9e55-6249-b574-0986cd5cea1e@gmail.com>
Precedence: bulk

On Wed, Jun 23, 2021 at 10:57:35AM +0200, Christian König wrote:

> > > No it isn't. It makes devices depend on allocating struct pages for their
> > > BARs which is not necessary nor desired.
> > Which dramatically reduces the cost of establishing DMA mappings, a
> > loop of dma_map_resource() is very expensive.
> 
> Yeah, but that is perfectly ok. Our BAR allocations are either in chunks of
> at least 2MiB or only a single 4KiB page.

And very small apparently
 
> > > Allocating a struct pages has their use case, for example for exposing VRAM
> > > as memory for HMM. But that is something very specific and should not limit
> > > PCIe P2P DMA in general.
> > Sure, but that is an ideal we are far from obtaining, and nobody wants
> > to work on it prefering to do hacky hacky like this.
> > 
> > If you believe in this then remove the scatter list from dmabuf, add a
> > new set of dma_map* APIs to work on physical addresses and all the
> > other stuff needed.
> 
> Yeah, that's what I totally agree on. And I actually hoped that the new P2P
> work for PCIe would go into that direction, but that didn't materialized.

It is a lot of work and the only gain is to save a bit of memory for
struct pages. Not a very big pay off.
 
> But allocating struct pages for PCIe BARs which are essentially registers
> and not memory is much more hacky than the dma_resource_map() approach.

It doesn't really matter. The pages are in a special zone and are only
being used as handles for the BAR memory.

> By using PCIe P2P we want to avoid the round trip to the CPU when one device
> has filled the ring buffer and another device must be woken up to process
> it.

Sure, we all have these scenarios, what is inside the memory doesn't
realy matter. The mechanism is generic and the struct pages don't care
much if they point at something memory-like or at something
register-like.

They are already in big trouble because you can't portably use CPU
instructions to access them anyhow.

Jason