From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org X-Spam-Level: X-Spam-Status: No, score=-5.0 required=3.0 tests=BAYES_00,DKIMWL_WL_HIGH, DKIM_SIGNED,DKIM_VALID,DKIM_VALID_AU,HEADER_FROM_DIFFERENT_DOMAINS, MAILING_LIST_MULTI,SPF_HELO_NONE,SPF_PASS,URIBL_BLOCKED autolearn=no autolearn_force=no version=3.4.0 Received: from mail.kernel.org (mail.kernel.org [198.145.29.99]) by smtp.lore.kernel.org (Postfix) with ESMTP id 636ADC4741F for ; Sat, 26 Sep 2020 15:25:47 +0000 (UTC) Received: from vger.kernel.org (vger.kernel.org [23.128.96.18]) by mail.kernel.org (Postfix) with ESMTP id 1F6C72168B for ; Sat, 26 Sep 2020 15:25:47 +0000 (UTC) Authentication-Results: mail.kernel.org; dkim=pass (1024-bit key) header.d=chromium.org header.i=@chromium.org header.b="IgNtiCGT" Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1729634AbgIZPZq (ORCPT ); Sat, 26 Sep 2020 11:25:46 -0400 Received: from lindbergh.monkeyblade.net ([23.128.96.19]:59066 "EHLO lindbergh.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1725208AbgIZPZq (ORCPT ); Sat, 26 Sep 2020 11:25:46 -0400 Received: from mail-ej1-x643.google.com (mail-ej1-x643.google.com [IPv6:2a00:1450:4864:20::643]) by lindbergh.monkeyblade.net (Postfix) with ESMTPS id CABECC0613CE for ; Sat, 26 Sep 2020 08:25:45 -0700 (PDT) Received: by mail-ej1-x643.google.com with SMTP id nw23so2596790ejb.4 for ; Sat, 26 Sep 2020 08:25:45 -0700 (PDT) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=chromium.org; s=google; h=mime-version:references:in-reply-to:from:date:message-id:subject:to :cc; bh=NyijGOIcOg0IQ8xJ7tRK9WXb8VbwFr7joS3ngRxTFtY=; b=IgNtiCGTIRw+9Jq9+OugAEUgOtWjA8eNb9N0lGKIZhSzq8+Nq6i8p6dAPgDSw3WCu5 5zLCQcy1nGd484dN8MEuuVSYHqucVtHjEfU81Ws4/lYX/HOp4JbMrZzXbrOql6BtdQAy +hEnD/Pr6tT/viPh9YkxxEEc7eplb0NwJB92c= X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20161025; h=x-gm-message-state:mime-version:references:in-reply-to:from:date :message-id:subject:to:cc; bh=NyijGOIcOg0IQ8xJ7tRK9WXb8VbwFr7joS3ngRxTFtY=; b=Yu68ISbJD4bsmwNpeLpec2nDpoQ7EMFcoE8pL+HzXmZr1yQyvGhkhac5gwxfvdn2sN U0OWbdSOesU2ZFYwfE4bo64HRpD98uP/flpVX/Nw5NzZ+Fo1p9OVxyvVTOph4R0x8p7B vUatr75ZUHDuZqKeddPVA2ds9kNavmOhqbEGT5mZbJe3bY/Y04tkn9OMP1lxm5aASGua +6R2/f1F1tqqLUs0BnppEzr/r/pEzUnz2/9vFWMGAlTZYpUmi4Hhc9HFNeKqTc/sqEli 2sWJPqod1oXgvZabwt6nuR49ksG5qhc10jGXNOk/WpXak8CwBRVuDmqh7jdDsn72Oo5k RtcA== X-Gm-Message-State: AOAM533Vn/8gMYn7T9W1mpZzZa6VtZGPzdQZflLVbAtweUeLdTBXCefD za6bMTArHgT9v3HPYWv/RJrhdV+lkaIS8w== X-Google-Smtp-Source: ABdhPJyhSvYrDdDDk3Z9CsZlRdpuUpKnsjFKfAMrIIyFeG1fPnHf1IghQgtPx8z9osZ5GpqlxKu9/A== X-Received: by 2002:a17:906:874f:: with SMTP id hj15mr8220792ejb.539.1601133944396; Sat, 26 Sep 2020 08:25:44 -0700 (PDT) Received: from mail-wm1-f48.google.com (mail-wm1-f48.google.com. [209.85.128.48]) by smtp.gmail.com with ESMTPSA id y9sm4617979edo.37.2020.09.26.08.25.43 for (version=TLS1_3 cipher=TLS_AES_128_GCM_SHA256 bits=128/128); Sat, 26 Sep 2020 08:25:44 -0700 (PDT) Received: by mail-wm1-f48.google.com with SMTP id b79so2330294wmb.4 for ; Sat, 26 Sep 2020 08:25:43 -0700 (PDT) X-Received: by 2002:a7b:c925:: with SMTP id h5mr2968012wml.28.1601133943201; Sat, 26 Sep 2020 08:25:43 -0700 (PDT) MIME-Version: 1.0 References: <20200915155122.1768241-1-hch@lst.de> <20200915155122.1768241-18-hch@lst.de> <20200925184622.GB3607091@chromium.org> <20200926141428.GB10379@lst.de> In-Reply-To: <20200926141428.GB10379@lst.de> From: Tomasz Figa Date: Sat, 26 Sep 2020 17:25:27 +0200 X-Gmail-Original-Message-ID: Message-ID: Subject: Re: [PATCH 17/18] dma-iommu: implement ->alloc_noncoherent To: Christoph Hellwig Cc: Mauro Carvalho Chehab , Thomas Bogendoerfer , "James E.J. Bottomley" , Joonyoung Shim , Seung-Woo Kim , Ben Skeggs , Marek Szyprowski , Matt Porter , "list@263.net:IOMMU DRIVERS , Joerg Roedel ," , Stefan Richter , linux1394-devel@lists.sourceforge.net, Linux Doc Mailing List , Linux Kernel Mailing List , Linux Media Mailing List , "list@263.net:IOMMU DRIVERS , Joerg Roedel ," , linux-mips@vger.kernel.org, linux-parisc@vger.kernel.org, linux-samsung-soc , nouveau@lists.freedesktop.org, netdev@vger.kernel.org, linux-scsi@vger.kernel.org, linux-mm@kvack.org, alsa-devel@alsa-project.org Content-Type: text/plain; charset="UTF-8" Precedence: bulk List-ID: X-Mailing-List: netdev@vger.kernel.org On Sat, Sep 26, 2020 at 4:14 PM Christoph Hellwig wrote: > > On Fri, Sep 25, 2020 at 06:46:22PM +0000, Tomasz Figa wrote: > > > +static void *iommu_dma_alloc_noncoherent(struct device *dev, size_t size, > > > + dma_addr_t *handle, enum dma_data_direction dir, gfp_t gfp) > > > +{ > > > + if (!gfpflags_allow_blocking(gfp)) { > > > + struct page *page; > > > + > > > + page = dma_common_alloc_pages(dev, size, handle, dir, gfp); > > > + if (!page) > > > + return NULL; > > > + return page_address(page); > > > + } > > > + > > > + return iommu_dma_alloc_remap(dev, size, handle, gfp | __GFP_ZERO, > > > + PAGE_KERNEL, 0); > > > > iommu_dma_alloc_remap() makes use of the DMA_ATTR_ALLOC_SINGLE_PAGES attribute > > to optimize the allocations for devices which don't care about how contiguous > > the backing memory is. Do you think we could add an attrs argument to this > > function and pass it there? > > > > As ARM is being moved to the common iommu-dma layer as well, we'll probably > > make use of the argument to support the DMA_ATTR_NO_KERNEL_MAPPING attribute to > > conserve the vmalloc area. > > We could probably at it. However I wonder why this is something the > drivers should care about. Isn't this really something that should > be a kernel-wide policy for a given system? There are IOMMUs out there which support huge pages and those can benefit *some* hardware depending on what kind of accesses they perform, possibly on a per-buffer basis. At the same time, order > 0 allocations can be expensive, significantly affecting allocation latency, so for devices which don't care about huge pages anyone would prefer simple single-page allocations. Currently the drivers know the best on whether the hardware they drive would care. There are some decision factors listed in the documentation [1]. I can imagine cases where drivers could not be the best to decide about this - for example, the workload could vary depending on the userspace or a product decision regarding the performance vs allocation latency, but we haven't seen such cases in practice yet. [1] https://www.kernel.org/doc/html/latest/core-api/dma-attributes.html?highlight=dma_attr_alloc_single_pages#dma-attr-alloc-single-pages Best regards, Tomasz