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 Received: from gabe.freedesktop.org (gabe.freedesktop.org [131.252.210.177]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by smtp.lore.kernel.org (Postfix) with ESMTPS id DCD0A10F9972 for ; Wed, 8 Apr 2026 20:15:50 +0000 (UTC) Received: from gabe.freedesktop.org (localhost [127.0.0.1]) by gabe.freedesktop.org (Postfix) with ESMTP id 8DC2610E260; Wed, 8 Apr 2026 20:15:46 +0000 (UTC) Authentication-Results: gabe.freedesktop.org; dkim=pass (2048-bit key; unprotected) header.d=intel.com header.i=@intel.com header.b="boT5w+8c"; dkim-atps=neutral Received: from mgamail.intel.com (mgamail.intel.com [198.175.65.10]) by gabe.freedesktop.org (Postfix) with ESMTPS id 9BCE510E0AB; Wed, 8 Apr 2026 20:15:44 +0000 (UTC) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=intel.com; i=@intel.com; q=dns/txt; s=Intel; t=1775679345; x=1807215345; h=from:to:cc:subject:date:message-id:in-reply-to: references:mime-version:content-transfer-encoding; bh=CJ7mip8ySPRr8YXN1vu1UPVHbFgshw6EcihFI3kF6oU=; b=boT5w+8ciejSee6cCdGj/+YbKPX8U+g42e+VeMak/6ye8pdcM/JUhmH1 PtukZ4jUq9rC26onT2WxIh+ss7zmBc9S7KBodqvnpqesYSsT6Wss6pj0E LLTqz7owG+a00NvR9p/bXNMCoBhkv9gCkmOEC4u8leK5/qj6dwV+S6QZK 4fmdSZUoilMUR86x5RLDUnp1giMQw2ZF9Ez/H9nsHMXcOym5sp5bPNQg+ bcH+gsidtEmcIe8lO6fSdtT0xUhi5bkJYMzIYIZFpsBVOfMMOPstToiH9 Hp4ifSwMBRbMXe1GBBgoNfM+b05vG7femE/C1LbwJ3UBkgtP6cZ0NUjI5 A==; X-CSE-ConnectionGUID: 2Eg4tT2pR5iXGwco87ph6Q== X-CSE-MsgGUID: PO47lqHqRi6EmQNuKq/eaw== X-IronPort-AV: E=McAfee;i="6800,10657,11753"; a="94063723" X-IronPort-AV: E=Sophos;i="6.23,168,1770624000"; d="scan'208";a="94063723" Received: from orviesa009.jf.intel.com ([10.64.159.149]) by orvoesa102.jf.intel.com with ESMTP/TLS/ECDHE-RSA-AES256-GCM-SHA384; 08 Apr 2026 13:15:44 -0700 X-CSE-ConnectionGUID: z/k5EwzrTYG2j0W+dPCW6g== X-CSE-MsgGUID: rpDOiYQwQHOUzPxw2qiZKw== X-ExtLoop1: 1 X-IronPort-AV: E=Sophos;i="6.23,168,1770624000"; d="scan'208";a="228451013" Received: from gsse-cloud1.jf.intel.com ([10.54.39.91]) by orviesa009-auth.jf.intel.com with ESMTP/TLS/ECDHE-RSA-AES256-GCM-SHA384; 08 Apr 2026 13:15:44 -0700 From: Matthew Brost To: intel-xe@lists.freedesktop.org, dri-devel@lists.freedesktop.org, francois.dugast@intel.com Cc: =?UTF-8?q?Thomas=20Hellstr=C3=B6m?= Subject: [PATCH v6 1/5] drm/gpusvm: Use dma-map IOVA alloc, link, and sync API in GPU SVM Date: Wed, 8 Apr 2026 13:15:33 -0700 Message-Id: <20260408201537.3580549-2-matthew.brost@intel.com> X-Mailer: git-send-email 2.34.1 In-Reply-To: <20260408201537.3580549-1-matthew.brost@intel.com> References: <20260408201537.3580549-1-matthew.brost@intel.com> MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit X-BeenThere: intel-xe@lists.freedesktop.org X-Mailman-Version: 2.1.29 Precedence: list List-Id: Intel Xe graphics driver List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Errors-To: intel-xe-bounces@lists.freedesktop.org Sender: "Intel-xe" The dma-map IOVA alloc, link, and sync APIs perform significantly better than dma-map / dma-unmap, as they avoid costly IOMMU synchronizations. This difference is especially noticeable when mapping a 2MB region in 4KB pages. Use the IOVA alloc, link, and sync APIs for GPU SVM, which create DMA mappings between the CPU and GPU. Signed-off-by: Matthew Brost Reviewed-by: Thomas Hellström --- drivers/gpu/drm/drm_gpusvm.c | 55 ++++++++++++++++++++++++++++++------ include/drm/drm_gpusvm.h | 5 ++++ 2 files changed, 52 insertions(+), 8 deletions(-) diff --git a/drivers/gpu/drm/drm_gpusvm.c b/drivers/gpu/drm/drm_gpusvm.c index 7993e85c0566..00744e3f72a7 100644 --- a/drivers/gpu/drm/drm_gpusvm.c +++ b/drivers/gpu/drm/drm_gpusvm.c @@ -1144,11 +1144,19 @@ static void __drm_gpusvm_unmap_pages(struct drm_gpusvm *gpusvm, struct drm_gpusvm_pages_flags flags = { .__flags = svm_pages->flags.__flags, }; + bool use_iova = dma_use_iova(&svm_pages->state); + + if (use_iova) { + dma_iova_unlink(dev, &svm_pages->state, 0, + svm_pages->state_offset, + svm_pages->dma_addr[0].dir, 0); + dma_iova_free(dev, &svm_pages->state); + } for (i = 0, j = 0; i < npages; j++) { struct drm_pagemap_addr *addr = &svm_pages->dma_addr[j]; - if (addr->proto == DRM_INTERCONNECT_SYSTEM) + if (!use_iova && addr->proto == DRM_INTERCONNECT_SYSTEM) dma_unmap_page(dev, addr->addr, PAGE_SIZE << addr->order, @@ -1413,6 +1421,7 @@ int drm_gpusvm_get_pages(struct drm_gpusvm *gpusvm, struct drm_gpusvm_pages_flags flags; enum dma_data_direction dma_dir = ctx->read_only ? DMA_TO_DEVICE : DMA_BIDIRECTIONAL; + struct dma_iova_state *state = &svm_pages->state; retry: if (time_after(jiffies, timeout)) @@ -1451,6 +1460,9 @@ int drm_gpusvm_get_pages(struct drm_gpusvm *gpusvm, if (err) goto err_free; + *state = (struct dma_iova_state){}; + svm_pages->state_offset = 0; + map_pages: /* * Perform all dma mappings under the notifier lock to not @@ -1544,13 +1556,33 @@ int drm_gpusvm_get_pages(struct drm_gpusvm *gpusvm, goto err_unmap; } - addr = dma_map_page(gpusvm->drm->dev, - page, 0, - PAGE_SIZE << order, - dma_dir); - if (dma_mapping_error(gpusvm->drm->dev, addr)) { - err = -EFAULT; - goto err_unmap; + if (!i) + dma_iova_try_alloc(gpusvm->drm->dev, state, + npages * PAGE_SIZE >= + HPAGE_PMD_SIZE ? + HPAGE_PMD_SIZE : 0, + npages * PAGE_SIZE); + + if (dma_use_iova(state)) { + err = dma_iova_link(gpusvm->drm->dev, state, + hmm_pfn_to_phys(pfns[i]), + svm_pages->state_offset, + PAGE_SIZE << order, + dma_dir, 0); + if (err) + goto err_unmap; + + addr = state->addr + svm_pages->state_offset; + svm_pages->state_offset += PAGE_SIZE << order; + } else { + addr = dma_map_page(gpusvm->drm->dev, + page, 0, + PAGE_SIZE << order, + dma_dir); + if (dma_mapping_error(gpusvm->drm->dev, addr)) { + err = -EFAULT; + goto err_unmap; + } } svm_pages->dma_addr[j] = drm_pagemap_addr_encode @@ -1562,6 +1594,13 @@ int drm_gpusvm_get_pages(struct drm_gpusvm *gpusvm, flags.has_dma_mapping = true; } + if (dma_use_iova(state)) { + err = dma_iova_sync(gpusvm->drm->dev, state, 0, + svm_pages->state_offset); + if (err) + goto err_unmap; + } + if (pagemap) { flags.has_devmem_pages = true; drm_pagemap_get(dpagemap); diff --git a/include/drm/drm_gpusvm.h b/include/drm/drm_gpusvm.h index 2578ac92a8d4..cd94bb2ee6ee 100644 --- a/include/drm/drm_gpusvm.h +++ b/include/drm/drm_gpusvm.h @@ -6,6 +6,7 @@ #ifndef __DRM_GPUSVM_H__ #define __DRM_GPUSVM_H__ +#include #include #include #include @@ -136,6 +137,8 @@ struct drm_gpusvm_pages_flags { * @dma_addr: Device address array * @dpagemap: The struct drm_pagemap of the device pages we're dma-mapping. * Note this is assuming only one drm_pagemap per range is allowed. + * @state: DMA IOVA state for mapping. + * @state_offset: DMA IOVA offset for mapping. * @notifier_seq: Notifier sequence number of the range's pages * @flags: Flags for range * @flags.migrate_devmem: Flag indicating whether the range can be migrated to device memory @@ -147,6 +150,8 @@ struct drm_gpusvm_pages_flags { struct drm_gpusvm_pages { struct drm_pagemap_addr *dma_addr; struct drm_pagemap *dpagemap; + struct dma_iova_state state; + unsigned long state_offset; unsigned long notifier_seq; struct drm_gpusvm_pages_flags flags; }; -- 2.34.1