From patchwork Wed Nov 25 20:31:14 2020 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Ricardo Ribalda X-Patchwork-Id: 333333 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=-18.8 required=3.0 tests=BAYES_00,DKIMWL_WL_HIGH, DKIM_SIGNED,DKIM_VALID,DKIM_VALID_AU,HEADER_FROM_DIFFERENT_DOMAINS, INCLUDES_CR_TRAILER, INCLUDES_PATCH, MAILING_LIST_MULTI, SPF_HELO_NONE, SPF_PASS, URIBL_BLOCKED, USER_AGENT_GIT autolearn=unavailable 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 4562BC56201 for ; Wed, 25 Nov 2020 20:31:54 +0000 (UTC) Received: from vger.kernel.org (vger.kernel.org [23.128.96.18]) by mail.kernel.org (Postfix) with ESMTP id D9545206F7 for ; Wed, 25 Nov 2020 20:31:53 +0000 (UTC) Authentication-Results: mail.kernel.org; dkim=pass (1024-bit key) header.d=chromium.org header.i=@chromium.org header.b="BIbP6CPi" Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1730034AbgKYUbV (ORCPT ); Wed, 25 Nov 2020 15:31:21 -0500 Received: from lindbergh.monkeyblade.net ([23.128.96.19]:40144 "EHLO lindbergh.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1730015AbgKYUbV (ORCPT ); Wed, 25 Nov 2020 15:31:21 -0500 Received: from mail-wr1-x442.google.com (mail-wr1-x442.google.com [IPv6:2a00:1450:4864:20::442]) by lindbergh.monkeyblade.net (Postfix) with ESMTPS id 30DBCC0617A7 for ; Wed, 25 Nov 2020 12:31:21 -0800 (PST) Received: by mail-wr1-x442.google.com with SMTP id g14so3166658wrm.13 for ; Wed, 25 Nov 2020 12:31:21 -0800 (PST) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=chromium.org; s=google; h=from:to:cc:subject:date:message-id:mime-version :content-transfer-encoding; bh=q710+yuAXMJxe0nd7TtiMj+ZjNL+TkbdcSV2kJANPFc=; b=BIbP6CPisTVorSg5qoPf0+dLgPrcTLlTwlVp8ftZSvaFgDocUfy9W0QG0G5Ods6KYD yNuT+KUHJ3LuzVLi2aruhMW4cOqXa3HmxH07ArrA7Ka9z6jrGkuyP/FM26RjBdMrBi+s P4RXP9NBi6sM7Lf6VlkFS3ouGBO+VOYD2B60o= X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20161025; h=x-gm-message-state:from:to:cc:subject:date:message-id:mime-version :content-transfer-encoding; bh=q710+yuAXMJxe0nd7TtiMj+ZjNL+TkbdcSV2kJANPFc=; b=PD2+70IVAbPcYe/Qnpoz3u1PRsReUjC/orhTSlxcknhIr5ZbNjq4vmourYZ+KR8zeA CFIbN5E8N4EgJZpl5WSvgwxZDluRSpEdw6zaMtMTu52cLiq308iNajFCrEo419OOuLSK fAwcirzBELLujRWMzlcTdSDaGZkuRkECn8mtbWDa/wRYNpYuTjT/EzDO7XVti1GLURkO AGZCUuEjPZg5EFhETHvpHCyTVbupKtbLZzaokxPO9rAevvh7dmYjRKUWmUcbqCTMHP2n 69OudAJvwOg7/od+bdXggU6bsIDKvnGJeSYjyDRoK9T++P8URauWsjU7tyYVGOzRJGqi d/UQ== X-Gm-Message-State: AOAM530nOJ8IydO2iQD4uljnHZEIVYJEcba2OJFrOy4ELe/e0azAIJWL KyUC6951xiXyvXcyNemSwgVn4fcbD5yXlA== X-Google-Smtp-Source: ABdhPJyaT4T+JFlra8+I0dMKQzQ88jRUsULWXg1Du+OwyPFSmH7nHfduCIkjo/9lH+6AkK7sbuYwqA== X-Received: by 2002:adf:fd52:: with SMTP id h18mr6084459wrs.90.1606336279903; Wed, 25 Nov 2020 12:31:19 -0800 (PST) Received: from alco.lan ([80.71.134.83]) by smtp.gmail.com with ESMTPSA id h83sm4827088wmf.9.2020.11.25.12.31.18 (version=TLS1_3 cipher=TLS_AES_256_GCM_SHA384 bits=256/256); Wed, 25 Nov 2020 12:31:19 -0800 (PST) From: Ricardo Ribalda To: Robin Murphy , Christoph Hellwig , auro Carvalho Chehab , Marek Szyprowski , IOMMU DRIVERS , Joerg Roedel , Linux Doc Mailing List , Linux Kernel Mailing List , Linux Media Mailing List , Tomasz Figa , Sergey Senozhatsky Cc: Ricardo Ribalda Subject: [PATCH v2 5/6] media: uvcvideo: Use dma_alloc_noncontiguos API Date: Wed, 25 Nov 2020 21:31:14 +0100 Message-Id: <20201125203114.130967-1-ribalda@chromium.org> X-Mailer: git-send-email 2.29.2.454.gaff20da3a2-goog MIME-Version: 1.0 Precedence: bulk List-ID: X-Mailing-List: linux-media@vger.kernel.org On architectures where the is no coherent caching such as ARM use the dma_alloc_noncontiguos API and handle manually the cache flushing using dma_sync_sg(). With this patch on the affected architectures we can measure up to 20x performance improvement in uvc_video_copy_data_work(). Signed-off-by: Ricardo Ribalda --- v2: Thanks to Robin! Use dma_sync_sg instead of dma_sync_single drivers/media/usb/uvc/uvc_video.c | 83 ++++++++++++++++++++++++++----- drivers/media/usb/uvc/uvcvideo.h | 2 + 2 files changed, 73 insertions(+), 12 deletions(-) diff --git a/drivers/media/usb/uvc/uvc_video.c b/drivers/media/usb/uvc/uvc_video.c index a6a441d92b94..b2e6a9522999 100644 --- a/drivers/media/usb/uvc/uvc_video.c +++ b/drivers/media/usb/uvc/uvc_video.c @@ -1490,6 +1490,11 @@ static void uvc_video_encode_bulk(struct uvc_urb *uvc_urb, urb->transfer_buffer_length = stream->urb_size - len; } +static inline struct device *stream_to_dmadev(struct uvc_streaming *stream) +{ + return stream->dev->udev->bus->controller->parent; +} + static void uvc_video_complete(struct urb *urb) { struct uvc_urb *uvc_urb = urb->context; @@ -1539,6 +1544,10 @@ static void uvc_video_complete(struct urb *urb) * Process the URB headers, and optionally queue expensive memcpy tasks * to be deferred to a work queue. */ + if (uvc_urb->pages) { + dma_sync_sg_for_cpu(stream_to_dmadev(stream), uvc_urb->sgt.sgl, + uvc_urb->sgt.nents, DMA_FROM_DEVICE); + } stream->decode(uvc_urb, buf, buf_meta); /* If no async work is needed, resubmit the URB immediately. */ @@ -1566,8 +1575,16 @@ static void uvc_free_urb_buffers(struct uvc_streaming *stream) continue; #ifndef CONFIG_DMA_NONCOHERENT - usb_free_coherent(stream->dev->udev, stream->urb_size, - uvc_urb->buffer, uvc_urb->dma); + if (uvc_urb->pages) { + sg_free_table(&uvc_urb->sgt); + vunmap(uvc_urb->buffer); + dma_free_noncontiguous(stream_to_dmadev(stream), + stream->urb_size, + uvc_urb->pages, uvc_urb->dma); + } else { + usb_free_coherent(stream->dev->udev, stream->urb_size, + uvc_urb->buffer, uvc_urb->dma); + } #else kfree(uvc_urb->buffer); #endif @@ -1577,6 +1594,56 @@ static void uvc_free_urb_buffers(struct uvc_streaming *stream) stream->urb_size = 0; } +#ifndef CONFIG_DMA_NONCOHERENT +static bool uvc_alloc_urb_buffer(struct uvc_streaming *stream, + struct uvc_urb *uvc_urb, gfp_t gfp_flags) +{ + struct device *dma_dev = dma_dev = stream_to_dmadev(stream); + + if (!dma_can_alloc_noncontiguous(dma_dev)) { + uvc_urb->buffer = usb_alloc_coherent(stream->dev->udev, + stream->urb_size, + gfp_flags | __GFP_NOWARN, + &uvc_urb->dma); + return uvc_urb->buffer != NULL; + } + + uvc_urb->pages = dma_alloc_noncontiguous(dma_dev, stream->urb_size, + &uvc_urb->dma, + gfp_flags | __GFP_NOWARN, 0); + if (!uvc_urb->pages) + return false; + + uvc_urb->buffer = vmap(uvc_urb->pages, + PAGE_ALIGN(stream->urb_size) >> PAGE_SHIFT, + VM_DMA_COHERENT, PAGE_KERNEL); + if (!uvc_urb->buffer) { + dma_free_noncontiguous(dma_dev, stream->urb_size, + uvc_urb->pages, uvc_urb->dma); + return false; + } + + if (sg_alloc_table_from_pages(&uvc_urb->sgt, uvc_urb->pages, + PAGE_ALIGN(stream->urb_size) >> PAGE_SHIFT, 0, + stream->urb_size, GFP_KERNEL)) { + vunmap(uvc_urb->buffer); + dma_free_noncontiguous(dma_dev, stream->urb_size, + uvc_urb->pages, uvc_urb->dma); + return false; + } + + return true; +} +#else +static bool uvc_alloc_urb_buffer(struct uvc_streaming *stream, + struct uvc_urb *uvc_urb, gfp_t gfp_flags) +{ + uvc_urb->buffer = kmalloc(stream->urb_size, gfp_flags | __GFP_NOWARN); + + return uvc_urb->buffer != NULL; +} +#endif + /* * Allocate transfer buffers. This function can be called with buffers * already allocated when resuming from suspend, in which case it will @@ -1607,19 +1674,11 @@ static int uvc_alloc_urb_buffers(struct uvc_streaming *stream, /* Retry allocations until one succeed. */ for (; npackets > 1; npackets /= 2) { + stream->urb_size = psize * npackets; for (i = 0; i < UVC_URBS; ++i) { struct uvc_urb *uvc_urb = &stream->uvc_urb[i]; - stream->urb_size = psize * npackets; -#ifndef CONFIG_DMA_NONCOHERENT - uvc_urb->buffer = usb_alloc_coherent( - stream->dev->udev, stream->urb_size, - gfp_flags | __GFP_NOWARN, &uvc_urb->dma); -#else - uvc_urb->buffer = - kmalloc(stream->urb_size, gfp_flags | __GFP_NOWARN); -#endif - if (!uvc_urb->buffer) { + if (!uvc_alloc_urb_buffer(stream, uvc_urb, gfp_flags)) { uvc_free_urb_buffers(stream); break; } diff --git a/drivers/media/usb/uvc/uvcvideo.h b/drivers/media/usb/uvc/uvcvideo.h index a3dfacf069c4..3e6618a2ac82 100644 --- a/drivers/media/usb/uvc/uvcvideo.h +++ b/drivers/media/usb/uvc/uvcvideo.h @@ -532,6 +532,8 @@ struct uvc_urb { char *buffer; dma_addr_t dma; + struct page **pages; + struct sg_table sgt; unsigned int async_operations; struct uvc_copy_op copy_operations[UVC_MAX_PACKETS];