From nobody Thu Dec 18 16:33:48 2025 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 vger.kernel.org (vger.kernel.org [23.128.96.18]) by smtp.lore.kernel.org (Postfix) with ESMTP id C5D5FC83F13 for ; Sun, 27 Aug 2023 17:57:04 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S230187AbjH0R4j (ORCPT ); Sun, 27 Aug 2023 13:56:39 -0400 Received: from lindbergh.monkeyblade.net ([23.128.96.19]:50820 "EHLO lindbergh.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S230016AbjH0R4Q (ORCPT ); Sun, 27 Aug 2023 13:56:16 -0400 Received: from madras.collabora.co.uk (madras.collabora.co.uk [46.235.227.172]) by lindbergh.monkeyblade.net (Postfix) with ESMTPS id 5707D102 for ; Sun, 27 Aug 2023 10:56:14 -0700 (PDT) Received: from workpc.. (109-252-153-31.dynamic.spd-mgts.ru [109.252.153.31]) (using TLSv1.3 with cipher TLS_AES_256_GCM_SHA384 (256/256 bits) key-exchange X25519 server-signature RSA-PSS (4096 bits) server-digest SHA256) (No client certificate requested) (Authenticated sender: dmitry.osipenko) by madras.collabora.co.uk (Postfix) with ESMTPSA id 9602766072AC; Sun, 27 Aug 2023 18:56:11 +0100 (BST) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=collabora.com; s=mail; t=1693158973; bh=QUE15l49Vy2Mm2iNdtx68DoK3cUPhFFXaVfUeuwXxjE=; h=From:To:Cc:Subject:Date:In-Reply-To:References:From; b=IK+L/f4YJfT49LgtwI608Dhd+wPJIBCTmNsceuOKea9Vpb4zkwnYSdhbMt9onTEbH a+Ru9PBapO9uFP4dkAK6pn029m2qMvse+eoKU0PHxrCP5yDCN1/MT2352UvL0gP1u7 0cnlyPhZss7fdMcgFxt3VcqcMM1ClaxArzLb/+vncM+NM0JotORI07RTQBDpoUCWVz Mfm9cM2zAy8dgHBIYoWapBd+j6jyXtiiJTLkxYBFrvD86OAVA46YauYz4YmfHSqgOq ZsfJxKL8Nrg48m6iJfs/6L6ksE6AVQkeDIQIfBw4Rj5oYNO/0/O7RUqLocQ3nN3Bzn MVP1VY9KPrUFw== From: Dmitry Osipenko To: David Airlie , Gerd Hoffmann , Gurchetan Singh , Chia-I Wu , Daniel Vetter , Maarten Lankhorst , Maxime Ripard , Thomas Zimmermann , =?UTF-8?q?Christian=20K=C3=B6nig?= , Qiang Yu , Steven Price , Boris Brezillon , Emma Anholt , Melissa Wen , Will Deacon , Peter Zijlstra , Boqun Feng , Mark Rutland Cc: dri-devel@lists.freedesktop.org, linux-kernel@vger.kernel.org, kernel@collabora.com, virtualization@lists.linux-foundation.org, intel-gfx@lists.freedesktop.org Subject: [PATCH v15 13/23] drm/shmem-helper: Use kref for pages_use_count Date: Sun, 27 Aug 2023 20:54:39 +0300 Message-ID: <20230827175449.1766701-14-dmitry.osipenko@collabora.com> X-Mailer: git-send-email 2.41.0 In-Reply-To: <20230827175449.1766701-1-dmitry.osipenko@collabora.com> References: <20230827175449.1766701-1-dmitry.osipenko@collabora.com> MIME-Version: 1.0 Content-Transfer-Encoding: quoted-printable Precedence: bulk List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Type: text/plain; charset="utf-8" Use atomic kref helper for pages_use_count to optimize pin/unpin functions by skipping reservation locking while GEM's pin refcount > 1. Suggested-by: Boris Brezillon Signed-off-by: Dmitry Osipenko --- drivers/gpu/drm/drm_gem_shmem_helper.c | 48 ++++++++++++++----------- drivers/gpu/drm/lima/lima_gem.c | 2 +- drivers/gpu/drm/panfrost/panfrost_mmu.c | 2 +- include/drm/drm_gem_shmem_helper.h | 2 +- 4 files changed, 30 insertions(+), 24 deletions(-) diff --git a/drivers/gpu/drm/drm_gem_shmem_helper.c b/drivers/gpu/drm/drm_g= em_shmem_helper.c index 1a7e5c332fd8..5a2e37b3e51d 100644 --- a/drivers/gpu/drm/drm_gem_shmem_helper.c +++ b/drivers/gpu/drm/drm_gem_shmem_helper.c @@ -155,7 +155,7 @@ void drm_gem_shmem_free(struct drm_gem_shmem_object *sh= mem) if (shmem->got_sgt) drm_gem_shmem_put_pages_locked(shmem); =20 - drm_WARN_ON(obj->dev, shmem->pages_use_count); + drm_WARN_ON(obj->dev, kref_read(&shmem->pages_use_count)); =20 dma_resv_unlock(shmem->base.resv); } @@ -172,14 +172,13 @@ static int drm_gem_shmem_get_pages_locked(struct drm_= gem_shmem_object *shmem) =20 dma_resv_assert_held(shmem->base.resv); =20 - if (shmem->pages_use_count++ > 0) + if (kref_get_unless_zero(&shmem->pages_use_count)) return 0; =20 pages =3D drm_gem_get_pages(obj); if (IS_ERR(pages)) { drm_dbg_kms(obj->dev, "Failed to get pages (%ld)\n", PTR_ERR(pages)); - shmem->pages_use_count =3D 0; return PTR_ERR(pages); } =20 @@ -195,26 +194,20 @@ static int drm_gem_shmem_get_pages_locked(struct drm_= gem_shmem_object *shmem) =20 shmem->pages =3D pages; =20 + kref_init(&shmem->pages_use_count); + return 0; } =20 -/* - * drm_gem_shmem_put_pages_locked - Decrease use count on the backing page= s for a shmem GEM object - * @shmem: shmem GEM object - * - * This function decreases the use count and puts the backing pages when u= se drops to zero. - */ -void drm_gem_shmem_put_pages_locked(struct drm_gem_shmem_object *shmem) -{ - struct drm_gem_object *obj =3D &shmem->base; - - dma_resv_assert_held(shmem->base.resv); =20 - if (drm_WARN_ON_ONCE(obj->dev, !shmem->pages_use_count)) - return; +static void drm_gem_shmem_kref_release_pages(struct kref *kref) +{ + struct drm_gem_shmem_object *shmem; + struct drm_gem_object *obj; =20 - if (--shmem->pages_use_count > 0) - return; + shmem =3D container_of(kref, struct drm_gem_shmem_object, + pages_use_count); + obj =3D &shmem->base; =20 #ifdef CONFIG_X86 if (shmem->map_wc) @@ -226,6 +219,19 @@ void drm_gem_shmem_put_pages_locked(struct drm_gem_shm= em_object *shmem) shmem->pages_mark_accessed_on_put); shmem->pages =3D NULL; } + +/* + * drm_gem_shmem_put_pages_locked - Decrease use count on the backing page= s for a shmem GEM object + * @shmem: shmem GEM object + * + * This function decreases the use count and puts the backing pages when u= se drops to zero. + */ +void drm_gem_shmem_put_pages_locked(struct drm_gem_shmem_object *shmem) +{ + dma_resv_assert_held(shmem->base.resv); + + kref_put(&shmem->pages_use_count, drm_gem_shmem_kref_release_pages); +} EXPORT_SYMBOL_GPL(drm_gem_shmem_put_pages_locked); =20 static int drm_gem_shmem_pin_locked(struct drm_gem_shmem_object *shmem) @@ -556,8 +562,8 @@ static void drm_gem_shmem_vm_open(struct vm_area_struct= *vma) * mmap'd, vm_open() just grabs an additional reference for the new * mm the vma is getting copied into (ie. on fork()). */ - if (!drm_WARN_ON_ONCE(obj->dev, !shmem->pages_use_count)) - shmem->pages_use_count++; + drm_WARN_ON_ONCE(obj->dev, + !kref_get_unless_zero(&shmem->pages_use_count)); =20 dma_resv_unlock(shmem->base.resv); =20 @@ -638,7 +644,7 @@ void drm_gem_shmem_print_info(const struct drm_gem_shme= m_object *shmem, if (shmem->base.import_attach) return; =20 - drm_printf_indent(p, indent, "pages_use_count=3D%u\n", shmem->pages_use_c= ount); + drm_printf_indent(p, indent, "pages_use_count=3D%u\n", kref_read(&shmem->= pages_use_count)); drm_printf_indent(p, indent, "vmap_use_count=3D%u\n", shmem->vmap_use_cou= nt); drm_printf_indent(p, indent, "vaddr=3D%p\n", shmem->vaddr); } diff --git a/drivers/gpu/drm/lima/lima_gem.c b/drivers/gpu/drm/lima/lima_ge= m.c index 7d74c71f5558..a5f015d188cd 100644 --- a/drivers/gpu/drm/lima/lima_gem.c +++ b/drivers/gpu/drm/lima/lima_gem.c @@ -47,7 +47,7 @@ int lima_heap_alloc(struct lima_bo *bo, struct lima_vm *v= m) } =20 bo->base.pages =3D pages; - bo->base.pages_use_count =3D 1; + kref_init(&bo->base.pages_use_count); =20 mapping_set_unevictable(mapping); } diff --git a/drivers/gpu/drm/panfrost/panfrost_mmu.c b/drivers/gpu/drm/panf= rost/panfrost_mmu.c index 7771769f0ce0..c9ac9d361864 100644 --- a/drivers/gpu/drm/panfrost/panfrost_mmu.c +++ b/drivers/gpu/drm/panfrost/panfrost_mmu.c @@ -487,7 +487,7 @@ static int panfrost_mmu_map_fault_addr(struct panfrost_= device *pfdev, int as, goto err_unlock; } bo->base.pages =3D pages; - bo->base.pages_use_count =3D 1; + kref_init(&bo->base.pages_use_count); } else { pages =3D bo->base.pages; if (pages[page_offset]) { diff --git a/include/drm/drm_gem_shmem_helper.h b/include/drm/drm_gem_shmem= _helper.h index afb7cd671e2a..a5a3c193cc8f 100644 --- a/include/drm/drm_gem_shmem_helper.h +++ b/include/drm/drm_gem_shmem_helper.h @@ -37,7 +37,7 @@ struct drm_gem_shmem_object { * Reference count on the pages table. * The pages are put when the count reaches zero. */ - unsigned int pages_use_count; + struct kref pages_use_count; =20 /** * @pages_pin_count: --=20 2.41.0