From patchwork Tue Nov 26 17:46:10 2024 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 8bit X-Patchwork-Submitter: Matthew Brost X-Patchwork-Id: 13886289 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 E6CE5D65536 for ; Tue, 26 Nov 2024 17:45:48 +0000 (UTC) Received: from gabe.freedesktop.org (localhost [127.0.0.1]) by gabe.freedesktop.org (Postfix) with ESMTP id 5B2DC10E986; Tue, 26 Nov 2024 17:45:43 +0000 (UTC) Authentication-Results: gabe.freedesktop.org; dkim=pass (2048-bit key; unprotected) header.d=intel.com header.i=@intel.com header.b="bBZvyIcx"; dkim-atps=neutral Received: from mgamail.intel.com (mgamail.intel.com [198.175.65.19]) by gabe.freedesktop.org (Postfix) with ESMTPS id 3B95D899A3; Tue, 26 Nov 2024 17:45:40 +0000 (UTC) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=intel.com; i=@intel.com; q=dns/txt; s=Intel; t=1732643140; x=1764179140; h=from:to:cc:subject:date:message-id:in-reply-to: references:mime-version:content-transfer-encoding; bh=uXHvGs7ZY7oeRDzrXB7ZrV6kkTbfhCu/sX9pLtNgwWs=; b=bBZvyIcxtE/i28pnKrvZ9HE62N8WMo5wND/G9AkpSg1uHz45oaRE+agi eY9hmuhcRqwncEMWM39/soGtMZVN29Q6Td7VMUwSI1czReoo1jRoG0SFf XqYNf6fTXHclU+RIH2CsyjcSGnOUMXSILZTuc0rQ9uVBfwqB59oVjCuDU 1kZvix9nNX4pIK3WKOTE4g8CEIlSl4kJ0b8g0S6bHVGmpYjS+bN4kaboX M5ia0clrS/4/bkO4jJSpZle2z49n1uq4LpQKCf8ywX+LkU3pQgLHoQvcz QKohyaXgNCno75Shspd1nOVPv43rhgSCFNixfzMVOvK5cumNegnAoZRmb w==; X-CSE-ConnectionGUID: O0MWp5phSpeAFvwBeyzIgA== X-CSE-MsgGUID: 59DAEKqWSZSR5yqntCyMkA== X-IronPort-AV: E=McAfee;i="6700,10204,11268"; a="32676913" X-IronPort-AV: E=Sophos;i="6.12,186,1728975600"; d="scan'208";a="32676913" Received: from fmviesa009.fm.intel.com ([10.60.135.149]) by orvoesa111.jf.intel.com with ESMTP/TLS/ECDHE-RSA-AES256-GCM-SHA384; 26 Nov 2024 09:45:40 -0800 X-CSE-ConnectionGUID: Vm2VkaoVTo+AHtVVJYOviQ== X-CSE-MsgGUID: swNb9RGwTH+/ZDuPKbcqNQ== X-ExtLoop1: 1 X-IronPort-AV: E=Sophos;i="6.12,186,1728975600"; d="scan'208";a="92152580" Received: from lstrano-desk.jf.intel.com ([10.54.39.91]) by fmviesa009-auth.fm.intel.com with ESMTP/TLS/ECDHE-RSA-AES256-GCM-SHA384; 26 Nov 2024 09:45:40 -0800 From: Matthew Brost To: intel-xe@lists.freedesktop.org, dri-devel@lists.freedesktop.org Cc: christian.koenig@amd.com, thomas.hellstrom@linux.intel.com Subject: [PATCH v7 3/8] drm/xe: Add xe_ttm_access_memory Date: Tue, 26 Nov 2024 09:46:10 -0800 Message-Id: <20241126174615.2665852-4-matthew.brost@intel.com> X-Mailer: git-send-email 2.34.1 In-Reply-To: <20241126174615.2665852-1-matthew.brost@intel.com> References: <20241126174615.2665852-1-matthew.brost@intel.com> MIME-Version: 1.0 X-BeenThere: dri-devel@lists.freedesktop.org X-Mailman-Version: 2.1.29 Precedence: list List-Id: Direct Rendering Infrastructure - Development List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Errors-To: dri-devel-bounces@lists.freedesktop.org Sender: "dri-devel" Non-contiguous VRAM cannot easily be mapped in TTM nor can non-visible VRAM easily be accessed. Add xe_ttm_access_memory which hooks into ttm_bo_access to access such memory. v4: - Assert memory access rather than taking RPM ref (Thomas / Auld) - Fix warning on xe_res_cursor.h for non-zero offset (Mika) Reported-by: Christoph Manszewski Suggested-by: Thomas Hellström Signed-off-by: Matthew Brost Reviewed-by: Matthew Auld --- drivers/gpu/drm/xe/xe_bo.c | 59 ++++++++++++++++++++++++++++++++++++-- 1 file changed, 56 insertions(+), 3 deletions(-) diff --git a/drivers/gpu/drm/xe/xe_bo.c b/drivers/gpu/drm/xe/xe_bo.c index aaf54131c89e..9de28798ad1c 100644 --- a/drivers/gpu/drm/xe/xe_bo.c +++ b/drivers/gpu/drm/xe/xe_bo.c @@ -442,6 +442,14 @@ static void xe_ttm_tt_destroy(struct ttm_device *ttm_dev, struct ttm_tt *tt) kfree(tt); } +static bool xe_ttm_resource_visible(struct ttm_resource *mem) +{ + struct xe_ttm_vram_mgr_resource *vres = + to_xe_ttm_vram_mgr_resource(mem); + + return vres->used_visible_size == mem->size; +} + static int xe_ttm_io_mem_reserve(struct ttm_device *bdev, struct ttm_resource *mem) { @@ -453,11 +461,9 @@ static int xe_ttm_io_mem_reserve(struct ttm_device *bdev, return 0; case XE_PL_VRAM0: case XE_PL_VRAM1: { - struct xe_ttm_vram_mgr_resource *vres = - to_xe_ttm_vram_mgr_resource(mem); struct xe_mem_region *vram = res_to_mem_region(mem); - if (vres->used_visible_size < mem->size) + if (!xe_ttm_resource_visible(mem)) return -EINVAL; mem->bus.offset = mem->start << PAGE_SHIFT; @@ -1121,6 +1127,52 @@ static void xe_ttm_bo_swap_notify(struct ttm_buffer_object *ttm_bo) } } +static int xe_ttm_access_memory(struct ttm_buffer_object *ttm_bo, + unsigned long offset, void *buf, int len, + int write) +{ + struct xe_bo *bo = ttm_to_xe_bo(ttm_bo); + struct xe_device *xe = ttm_to_xe_device(ttm_bo->bdev); + struct iosys_map vmap; + struct xe_res_cursor cursor; + struct xe_mem_region *vram; + int bytes_left = len; + + xe_bo_assert_held(bo); + xe_device_assert_mem_access(xe); + + if (!mem_type_is_vram(ttm_bo->resource->mem_type)) + return -EIO; + + /* FIXME: Use GPU for non-visible VRAM */ + if (!xe_ttm_resource_visible(ttm_bo->resource)) + return -EIO; + + vram = res_to_mem_region(ttm_bo->resource); + xe_res_first(ttm_bo->resource, offset & PAGE_MASK, + bo->size - (offset & PAGE_MASK), &cursor); + + do { + unsigned long page_offset = (offset & ~PAGE_MASK); + int byte_count = min((int)(PAGE_SIZE - page_offset), bytes_left); + + iosys_map_set_vaddr_iomem(&vmap, (u8 __iomem *)vram->mapping + + cursor.start); + if (write) + xe_map_memcpy_to(xe, &vmap, page_offset, buf, byte_count); + else + xe_map_memcpy_from(xe, buf, &vmap, page_offset, byte_count); + + buf += byte_count; + offset += byte_count; + bytes_left -= byte_count; + if (bytes_left) + xe_res_next(&cursor, PAGE_SIZE); + } while (bytes_left); + + return len; +} + const struct ttm_device_funcs xe_ttm_funcs = { .ttm_tt_create = xe_ttm_tt_create, .ttm_tt_populate = xe_ttm_tt_populate, @@ -1130,6 +1182,7 @@ const struct ttm_device_funcs xe_ttm_funcs = { .move = xe_bo_move, .io_mem_reserve = xe_ttm_io_mem_reserve, .io_mem_pfn = xe_ttm_io_mem_pfn, + .access_memory = xe_ttm_access_memory, .release_notify = xe_ttm_bo_release_notify, .eviction_valuable = ttm_bo_eviction_valuable, .delete_mem_notify = xe_ttm_bo_delete_mem_notify,