From nobody Fri Dec 19 20:55:28 2025 Received: from bali.collaboradmins.com (bali.collaboradmins.com [148.251.105.195]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by smtp.subspace.kernel.org (Postfix) with ESMTPS id A85951547D2; Wed, 15 Oct 2025 15:30:29 +0000 (UTC) Authentication-Results: smtp.subspace.kernel.org; arc=none smtp.client-ip=148.251.105.195 ARC-Seal: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1760542231; cv=none; b=i1ARhRdIDN0o0TUJbKCjmWLPsgfVJEqtisxMHpGExnb9Rmg0Jzt+UhhC4LbtQmLgB7oXuNIu8aKaVytEdnK4lP+L2xDnZql615dssmh1TdWM7vB9DkGmvkNigpK+5TyJ+oHPfX/jsc1OItaAdKBybLwPk6pu9o40MMPWO+1O4Sc= ARC-Message-Signature: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1760542231; c=relaxed/simple; bh=sQcBV/ZWlVLGeAwivyndjv3Kv+k/mZS823qVLbIeFmg=; h=From:To:Cc:Subject:Date:Message-ID:In-Reply-To:References: MIME-Version:Content-Type; b=Z1Umb48aCU96uQ/dzRLsN8OnpgX1sBbu3+T2MBoPMukLi2GWqscf9efcac0MYyyLWQtOs7E1sxZiqNP1hKwHSwvRBrnWalWkH0IROxdg1+8NnXCsdCK62EgP5XjyN+cyE1Jf2tcVzkwogTS+Kl8t6pZqS/YXV0l8Sf5ZwPJYpK8= ARC-Authentication-Results: i=1; smtp.subspace.kernel.org; dmarc=pass (p=none dis=none) header.from=collabora.com; spf=pass smtp.mailfrom=collabora.com; dkim=pass (2048-bit key) header.d=collabora.com header.i=@collabora.com header.b=VHdkIRj7; arc=none smtp.client-ip=148.251.105.195 Authentication-Results: smtp.subspace.kernel.org; dmarc=pass (p=none dis=none) header.from=collabora.com Authentication-Results: smtp.subspace.kernel.org; spf=pass smtp.mailfrom=collabora.com Authentication-Results: smtp.subspace.kernel.org; dkim=pass (2048-bit key) header.d=collabora.com header.i=@collabora.com header.b="VHdkIRj7" DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=collabora.com; s=mail; t=1760542227; bh=sQcBV/ZWlVLGeAwivyndjv3Kv+k/mZS823qVLbIeFmg=; h=From:To:Cc:Subject:Date:In-Reply-To:References:From; b=VHdkIRj7sg4VwSgsQanwXYtbK/UcvTe7EmUkEqqxHkXOqGzYygtHQ7zYKzTzN0BW0 DVlMfVVHuUVa4r3IHy86HYMx7KYMhmY6osKmaU4PMNbo+WlYSZefsYM7zb639o0aA/ ywmDIaSM8jVfaDD9GVr7KGmxiAMWOMnqZJI4O5IeZq7EcOlIu749ykkS7wXJ2xxBD4 Qb3umxLKhtjCGU/3anoP0khorYvpn7lF4YwEktiPC00IvTmS5qwmS1vf0pjGegIhmI cxfqel3TgrUMeNW//bU37J3wRhpV55lJlZy1Qeq8HVTySPa4FIJHJRhrHnwvayxp17 yopEYsaDRwSIA== Received: from debian-rockchip-rock5b-rk3588.. (unknown [IPv6:2a01:e0a:5e3:6100:826d:bc07:e98c:84a]) (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: loicmolinari) by bali.collaboradmins.com (Postfix) with ESMTPSA id 0BD0E17E1340; Wed, 15 Oct 2025 17:30:27 +0200 (CEST) From: =?UTF-8?q?Lo=C3=AFc=20Molinari?= To: Maarten Lankhorst , Maxime Ripard , Thomas Zimmermann , David Airlie , Simona Vetter , Jani Nikula , Joonas Lahtinen , Rodrigo Vivi , Tvrtko Ursulin , Boris Brezillon , Rob Herring , Steven Price , Liviu Dudau , Melissa Wen , =?UTF-8?q?Ma=C3=ADra=20Canal?= , Hugh Dickins , Baolin Wang , Andrew Morton , =?UTF-8?q?Lo=C3=AFc=20Molinari?= , Al Viro , =?UTF-8?q?Miko=C5=82aj=20Wasiak?= , Christian Brauner , Nitin Gote , Andi Shyti , Jonathan Corbet , Christopher Healy , Matthew Wilcox , Bagas Sanjaya Cc: linux-kernel@vger.kernel.org, dri-devel@lists.freedesktop.org, intel-gfx@lists.freedesktop.org, linux-mm@kvack.org, linux-doc@vger.kernel.org, kernel@collabora.com Subject: [PATCH v4 01/13] drm/shmem-helper: Simplify page offset calculation in fault handler Date: Wed, 15 Oct 2025 17:30:05 +0200 Message-ID: <20251015153018.43735-2-loic.molinari@collabora.com> X-Mailer: git-send-email 2.47.3 In-Reply-To: <20251015153018.43735-1-loic.molinari@collabora.com> References: <20251015153018.43735-1-loic.molinari@collabora.com> Precedence: bulk X-Mailing-List: linux-kernel@vger.kernel.org List-Id: List-Subscribe: List-Unsubscribe: MIME-Version: 1.0 Content-Type: text/plain; charset="utf-8" Content-Transfer-Encoding: quoted-printable For a fault at address addr, the page offset is page_offset =3D (vmf->address - vma->vm_start) >> PAGE_SHIFT =3D ((addr & PAGE_MASK) - vma->vm_start) >> PAGE_SHIFT =3D (addr - vma->vm_start) >> PAGE_SHIFT Since the faulty logical page offset based on VMA is vmf->pgoff =3D vma->vm_pgoff + ((addr - vma->vm_start) >> PAGE_SHIFT) We can slightly simplify the calculation using page_offset =3D vmf->pgoff - vma->vm_pgoff Signed-off-by: Lo=C3=AFc Molinari --- drivers/gpu/drm/drm_gem_shmem_helper.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/drivers/gpu/drm/drm_gem_shmem_helper.c b/drivers/gpu/drm/drm_g= em_shmem_helper.c index dc94a27710e5..be89be1c804c 100644 --- a/drivers/gpu/drm/drm_gem_shmem_helper.c +++ b/drivers/gpu/drm/drm_gem_shmem_helper.c @@ -577,8 +577,8 @@ static vm_fault_t drm_gem_shmem_fault(struct vm_fault *= vmf) struct page *page; pgoff_t page_offset; =20 - /* We don't use vmf->pgoff since that has the fake offset */ - page_offset =3D (vmf->address - vma->vm_start) >> PAGE_SHIFT; + /* Offset to faulty address in the VMA (without the fake offset). */ + page_offset =3D vmf->pgoff - vma->vm_pgoff; =20 dma_resv_lock(shmem->base.resv, NULL); =20 --=20 2.47.3 From nobody Fri Dec 19 20:55:28 2025 Received: from bali.collaboradmins.com (bali.collaboradmins.com [148.251.105.195]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by smtp.subspace.kernel.org (Postfix) with ESMTPS id 79F3A302CC7; Wed, 15 Oct 2025 15:30:32 +0000 (UTC) Authentication-Results: smtp.subspace.kernel.org; arc=none smtp.client-ip=148.251.105.195 ARC-Seal: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1760542235; cv=none; b=FgknN1GZx7gQBuojlnZS5d1ooaB5DUPXRcSQB3t99ntQTb74mfDLe8e7TghxvPSR7at2dPphyff37NaLLvNFnEWS7ojweJ97Gcf9HHvdmgQcCYlttrvIxw9WIomX3t3xvuLyV+mEjRs9RkEW0xcXD+2yjwS1LYGtZe2AAEIcjXU= ARC-Message-Signature: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1760542235; c=relaxed/simple; bh=KQV7Ou7tPtnCCwHJMiVJOpNVyDdq8Ql41g0y4SHmXds=; h=From:To:Cc:Subject:Date:Message-ID:In-Reply-To:References: MIME-Version:Content-Type; b=fJuJJXQtmLzOyl4k4PHHUS1rfY+tdqjQGxvCjslthVtnbWGvf5tK6Rb5a9OCkdTtFnSzqileuifr7URLY/zc/cYpejAEq47L68+aeW38DeBE4PFKbtuO0YgY/IT+g8kvfZzr8QiMSrYkQluhRLV5m4S9leXxLhZ4IKFaQ4QFn5c= ARC-Authentication-Results: i=1; smtp.subspace.kernel.org; dmarc=pass (p=none dis=none) header.from=collabora.com; spf=pass smtp.mailfrom=collabora.com; dkim=pass (2048-bit key) header.d=collabora.com header.i=@collabora.com header.b=eHAmHCJq; arc=none smtp.client-ip=148.251.105.195 Authentication-Results: smtp.subspace.kernel.org; dmarc=pass (p=none dis=none) header.from=collabora.com Authentication-Results: smtp.subspace.kernel.org; spf=pass smtp.mailfrom=collabora.com Authentication-Results: smtp.subspace.kernel.org; dkim=pass (2048-bit key) header.d=collabora.com header.i=@collabora.com header.b="eHAmHCJq" DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=collabora.com; s=mail; t=1760542230; bh=KQV7Ou7tPtnCCwHJMiVJOpNVyDdq8Ql41g0y4SHmXds=; h=From:To:Cc:Subject:Date:In-Reply-To:References:From; b=eHAmHCJq9CHmEpVjvgz1qigTjIVJFFK9ZhUOKlaOXjDfZD2eF49vJtX3YAl+J4qG5 dEOcBvyK4cfnxwjrhL2CZslgUJNpEnIhuJY67F8Ok+Kt1pUTHwIsQzThMUEY59EDLx 9W7FHK5G5o2bk8opmpCRQU57tGjObB4fdsOpm6EabsOEkP6r0nIrjiJcLnSuD/NmiM hHVcEqg/Wy4+yOf5+6exKsv8odBVsInQoQGmsORy1pXP8E0xwImciHf65R8gewSSXy n3N/G7Phsqm286vuMG05Z/YZT/xKLbzRzl8UaI/cP6lqItS/4OG6gMSgngdmNgYg+q VKp6Xyv4Rq+Jw== Received: from debian-rockchip-rock5b-rk3588.. (unknown [IPv6:2a01:e0a:5e3:6100:826d:bc07:e98c:84a]) (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: loicmolinari) by bali.collaboradmins.com (Postfix) with ESMTPSA id E5BF917E00A3; Wed, 15 Oct 2025 17:30:29 +0200 (CEST) From: =?UTF-8?q?Lo=C3=AFc=20Molinari?= To: Maarten Lankhorst , Maxime Ripard , Thomas Zimmermann , David Airlie , Simona Vetter , Jani Nikula , Joonas Lahtinen , Rodrigo Vivi , Tvrtko Ursulin , Boris Brezillon , Rob Herring , Steven Price , Liviu Dudau , Melissa Wen , =?UTF-8?q?Ma=C3=ADra=20Canal?= , Hugh Dickins , Baolin Wang , Andrew Morton , =?UTF-8?q?Lo=C3=AFc=20Molinari?= , Al Viro , =?UTF-8?q?Miko=C5=82aj=20Wasiak?= , Christian Brauner , Nitin Gote , Andi Shyti , Jonathan Corbet , Christopher Healy , Matthew Wilcox , Bagas Sanjaya Cc: linux-kernel@vger.kernel.org, dri-devel@lists.freedesktop.org, intel-gfx@lists.freedesktop.org, linux-mm@kvack.org, linux-doc@vger.kernel.org, kernel@collabora.com Subject: [PATCH v4 02/13] drm/shmem-helper: Implement map_pages fault-around handler Date: Wed, 15 Oct 2025 17:30:06 +0200 Message-ID: <20251015153018.43735-3-loic.molinari@collabora.com> X-Mailer: git-send-email 2.47.3 In-Reply-To: <20251015153018.43735-1-loic.molinari@collabora.com> References: <20251015153018.43735-1-loic.molinari@collabora.com> Precedence: bulk X-Mailing-List: linux-kernel@vger.kernel.org List-Id: List-Subscribe: List-Unsubscribe: MIME-Version: 1.0 Content-Type: text/plain; charset="utf-8" Content-Transfer-Encoding: quoted-printable This gives the mm subsystem the ability to increase fault handling performance by proposing the insertion of a range of pages around the faulty address in a single batch. v4: - Implement map_pages instead of huge_fault Signed-off-by: Lo=C3=AFc Molinari --- drivers/gpu/drm/drm_gem_shmem_helper.c | 53 +++++++++++++++++++++++--- 1 file changed, 48 insertions(+), 5 deletions(-) diff --git a/drivers/gpu/drm/drm_gem_shmem_helper.c b/drivers/gpu/drm/drm_g= em_shmem_helper.c index be89be1c804c..e151262332f9 100644 --- a/drivers/gpu/drm/drm_gem_shmem_helper.c +++ b/drivers/gpu/drm/drm_gem_shmem_helper.c @@ -567,24 +567,66 @@ int drm_gem_shmem_dumb_create(struct drm_file *file, = struct drm_device *dev, } EXPORT_SYMBOL_GPL(drm_gem_shmem_dumb_create); =20 -static vm_fault_t drm_gem_shmem_fault(struct vm_fault *vmf) +static bool drm_gem_shmem_fault_is_valid(struct drm_gem_object *obj, + pgoff_t pgoff) +{ + struct drm_gem_shmem_object *shmem =3D to_drm_gem_shmem_obj(obj); + + if (drm_WARN_ON_ONCE(obj->dev, !shmem->pages) || + pgoff >=3D (obj->size >> PAGE_SHIFT) || + shmem->madv < 0) + return false; + + return true; +} + +static vm_fault_t drm_gem_shmem_map_pages(struct vm_fault *vmf, + pgoff_t start_pgoff, + pgoff_t end_pgoff) { struct vm_area_struct *vma =3D vmf->vma; struct drm_gem_object *obj =3D vma->vm_private_data; struct drm_gem_shmem_object *shmem =3D to_drm_gem_shmem_obj(obj); - loff_t num_pages =3D obj->size >> PAGE_SHIFT; + unsigned long addr, pfn; vm_fault_t ret; + + start_pgoff -=3D vma->vm_pgoff; + end_pgoff -=3D vma->vm_pgoff; + addr =3D vma->vm_start + (start_pgoff << PAGE_SHIFT); + + dma_resv_lock(shmem->base.resv, NULL); + + if (unlikely(!drm_gem_shmem_fault_is_valid(obj, start_pgoff))) { + ret =3D VM_FAULT_SIGBUS; + } else { + /* Map a range of pages around the faulty address. */ + do { + pfn =3D page_to_pfn(shmem->pages[start_pgoff]); + ret =3D vmf_insert_pfn(vma, addr, pfn); + addr +=3D PAGE_SIZE; + } while (++start_pgoff <=3D end_pgoff && ret =3D=3D VM_FAULT_NOPAGE); + } + + dma_resv_unlock(shmem->base.resv); + + return ret; +} + +static vm_fault_t drm_gem_shmem_fault(struct vm_fault *vmf) +{ + struct vm_area_struct *vma =3D vmf->vma; + struct drm_gem_object *obj =3D vma->vm_private_data; + struct drm_gem_shmem_object *shmem =3D to_drm_gem_shmem_obj(obj); struct page *page; pgoff_t page_offset; + vm_fault_t ret; =20 /* Offset to faulty address in the VMA (without the fake offset). */ page_offset =3D vmf->pgoff - vma->vm_pgoff; =20 dma_resv_lock(shmem->base.resv, NULL); =20 - if (page_offset >=3D num_pages || - drm_WARN_ON_ONCE(obj->dev, !shmem->pages) || - shmem->madv < 0) { + if (unlikely(!drm_gem_shmem_fault_is_valid(obj, page_offset))) { ret =3D VM_FAULT_SIGBUS; } else { page =3D shmem->pages[page_offset]; @@ -632,6 +674,7 @@ static void drm_gem_shmem_vm_close(struct vm_area_struc= t *vma) } =20 const struct vm_operations_struct drm_gem_shmem_vm_ops =3D { + .map_pages =3D drm_gem_shmem_map_pages, .fault =3D drm_gem_shmem_fault, .open =3D drm_gem_shmem_vm_open, .close =3D drm_gem_shmem_vm_close, --=20 2.47.3 From nobody Fri Dec 19 20:55:28 2025 Received: from bali.collaboradmins.com (bali.collaboradmins.com [148.251.105.195]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by smtp.subspace.kernel.org (Postfix) with ESMTPS id 0A79B1547D2; Wed, 15 Oct 2025 15:30:34 +0000 (UTC) Authentication-Results: smtp.subspace.kernel.org; arc=none smtp.client-ip=148.251.105.195 ARC-Seal: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1760542237; cv=none; b=uocceaw0zex7hRbOKQpLbRWbScDRItpCcKsjlZrReJNJfQRKEesLHkVHvNKglQcsKODFwBtuxCMGGLBY//4zIURDA34iGtTUn6gwacq4xGZrMI3hTVsb0bwnykiZPjuHphT6hLzpn5yzqzkflrm6f9SxFuJWFxl7DaLiaGRSWSs= ARC-Message-Signature: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1760542237; c=relaxed/simple; bh=i4jm8rfdS6G2Q04+yn738KVCd/HzsGzB/QnSS+UslAY=; h=From:To:Cc:Subject:Date:Message-ID:In-Reply-To:References: MIME-Version:Content-Type; b=jqs5uAcZzjze364u9Kgegi+/3ptYtS+vH5ZIdaestm/aeWifpHZU/OF0EHmxfiSlg62HgS6AK5H6vF8bYOGk6rRtDpCdOtVQ2HqHM1JVtExgenXFhYjNG+IybEQSQy/qc1Mc0tpSGHwJZGrOaiLKi6HKhTbNLD6jrZgzNIz1U7Y= ARC-Authentication-Results: i=1; smtp.subspace.kernel.org; dmarc=pass (p=none dis=none) header.from=collabora.com; spf=pass smtp.mailfrom=collabora.com; dkim=pass (2048-bit key) header.d=collabora.com header.i=@collabora.com header.b=bbJ9H8kS; arc=none smtp.client-ip=148.251.105.195 Authentication-Results: smtp.subspace.kernel.org; dmarc=pass (p=none dis=none) header.from=collabora.com Authentication-Results: smtp.subspace.kernel.org; spf=pass smtp.mailfrom=collabora.com Authentication-Results: smtp.subspace.kernel.org; dkim=pass (2048-bit key) header.d=collabora.com header.i=@collabora.com header.b="bbJ9H8kS" DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=collabora.com; s=mail; t=1760542233; bh=i4jm8rfdS6G2Q04+yn738KVCd/HzsGzB/QnSS+UslAY=; h=From:To:Cc:Subject:Date:In-Reply-To:References:From; b=bbJ9H8kSS1f9ZoeWKjtxRJDn57F4aP9klFG+XuHkXvOM8yZFcdQTWI5cb2BEuBEk1 gVlMJJSTxWdv7EeLK6TeJ7u5ga9M4xvTanGsth/ibKUIo0KZC3bn7JO+7mQYiswrYR pkac4KqbcFUdapTzUYAgNaWFReRqspgUU8FjvxSyQ5H14ybi4wPK/XEgakwCCCjVM4 Kj3PtLfZOyhbkx1CKLrHBeYNCKoCZCdQKzmNbkwp61DBH5Ok3vzetZkd8rGCiKiE0p inqYp/LBPgAunLivVYd7YRz17mgsyxLAww+B6EVZZ/EllLRF1uf5lGlsRpq8/qrnQ6 OtowYlUAfOltg== Received: from debian-rockchip-rock5b-rk3588.. (unknown [IPv6:2a01:e0a:5e3:6100:826d:bc07:e98c:84a]) (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: loicmolinari) by bali.collaboradmins.com (Postfix) with ESMTPSA id 7DF5917E055D; Wed, 15 Oct 2025 17:30:32 +0200 (CEST) From: =?UTF-8?q?Lo=C3=AFc=20Molinari?= To: Maarten Lankhorst , Maxime Ripard , Thomas Zimmermann , David Airlie , Simona Vetter , Jani Nikula , Joonas Lahtinen , Rodrigo Vivi , Tvrtko Ursulin , Boris Brezillon , Rob Herring , Steven Price , Liviu Dudau , Melissa Wen , =?UTF-8?q?Ma=C3=ADra=20Canal?= , Hugh Dickins , Baolin Wang , Andrew Morton , =?UTF-8?q?Lo=C3=AFc=20Molinari?= , Al Viro , =?UTF-8?q?Miko=C5=82aj=20Wasiak?= , Christian Brauner , Nitin Gote , Andi Shyti , Jonathan Corbet , Christopher Healy , Matthew Wilcox , Bagas Sanjaya Cc: linux-kernel@vger.kernel.org, dri-devel@lists.freedesktop.org, intel-gfx@lists.freedesktop.org, linux-mm@kvack.org, linux-doc@vger.kernel.org, kernel@collabora.com Subject: [PATCH v4 03/13] drm/shmem-helper: Map huge pages in fault handlers Date: Wed, 15 Oct 2025 17:30:07 +0200 Message-ID: <20251015153018.43735-4-loic.molinari@collabora.com> X-Mailer: git-send-email 2.47.3 In-Reply-To: <20251015153018.43735-1-loic.molinari@collabora.com> References: <20251015153018.43735-1-loic.molinari@collabora.com> Precedence: bulk X-Mailing-List: linux-kernel@vger.kernel.org List-Id: List-Subscribe: List-Unsubscribe: MIME-Version: 1.0 Content-Type: text/plain; charset="utf-8" Content-Transfer-Encoding: quoted-printable Check whether the starting address in the fault-around handler or the faulty address in the fault handler is part of a huge page in order to attempt a PMD sized PFN insertion into the VMA. On builds with CONFIG_TRANSPARENT_HUGEPAGE enabled, if the mmap() user address is PMD size aligned, if the GEM object is backed by shmem buffers on mount points setting the 'huge=3D' option and if the shmem backing store manages to allocate a huge folio, CPU mapping would then benefit from significantly increased memcpy() performance. When these conditions are met on a system with 2 MiB huge pages, an aligned copy of 2 MiB would raise a single page fault instead of 4096. v4: - Implement map_pages instead of huge_fault Signed-off-by: Lo=C3=AFc Molinari --- drivers/gpu/drm/drm_gem_shmem_helper.c | 57 +++++++++++++++++++++----- 1 file changed, 46 insertions(+), 11 deletions(-) diff --git a/drivers/gpu/drm/drm_gem_shmem_helper.c b/drivers/gpu/drm/drm_g= em_shmem_helper.c index e151262332f9..5d11bea573fe 100644 --- a/drivers/gpu/drm/drm_gem_shmem_helper.c +++ b/drivers/gpu/drm/drm_gem_shmem_helper.c @@ -580,6 +580,26 @@ static bool drm_gem_shmem_fault_is_valid(struct drm_ge= m_object *obj, return true; } =20 +static bool drm_gem_shmem_map_pmd(struct vm_fault *vmf, unsigned long addr, + struct page *page) +{ +#ifdef CONFIG_ARCH_SUPPORTS_PMD_PFNMAP + unsigned long pfn =3D page_to_pfn(page); + unsigned long paddr =3D pfn << PAGE_SHIFT; + bool aligned =3D (addr & ~PMD_MASK) =3D=3D (paddr & ~PMD_MASK); + + if (aligned && + pmd_none(*vmf->pmd) && + folio_test_pmd_mappable(page_folio(page))) { + pfn &=3D PMD_MASK >> PAGE_SHIFT; + if (vmf_insert_pfn_pmd(vmf, pfn, false) =3D=3D VM_FAULT_NOPAGE) + return true; + } +#endif + + return false; +} + static vm_fault_t drm_gem_shmem_map_pages(struct vm_fault *vmf, pgoff_t start_pgoff, pgoff_t end_pgoff) @@ -587,6 +607,7 @@ static vm_fault_t drm_gem_shmem_map_pages(struct vm_fau= lt *vmf, struct vm_area_struct *vma =3D vmf->vma; struct drm_gem_object *obj =3D vma->vm_private_data; struct drm_gem_shmem_object *shmem =3D to_drm_gem_shmem_obj(obj); + struct page **pages =3D shmem->pages; unsigned long addr, pfn; vm_fault_t ret; =20 @@ -598,15 +619,22 @@ static vm_fault_t drm_gem_shmem_map_pages(struct vm_f= ault *vmf, =20 if (unlikely(!drm_gem_shmem_fault_is_valid(obj, start_pgoff))) { ret =3D VM_FAULT_SIGBUS; - } else { - /* Map a range of pages around the faulty address. */ - do { - pfn =3D page_to_pfn(shmem->pages[start_pgoff]); - ret =3D vmf_insert_pfn(vma, addr, pfn); - addr +=3D PAGE_SIZE; - } while (++start_pgoff <=3D end_pgoff && ret =3D=3D VM_FAULT_NOPAGE); + goto out; } =20 + if (drm_gem_shmem_map_pmd(vmf, addr, pages[start_pgoff])) { + ret =3D VM_FAULT_NOPAGE; + goto out; + } + + /* Map a range of pages around the faulty address. */ + do { + pfn =3D page_to_pfn(pages[start_pgoff]); + ret =3D vmf_insert_pfn(vma, addr, pfn); + addr +=3D PAGE_SIZE; + } while (++start_pgoff <=3D end_pgoff && ret =3D=3D VM_FAULT_NOPAGE); + + out: dma_resv_unlock(shmem->base.resv); =20 return ret; @@ -617,8 +645,9 @@ static vm_fault_t drm_gem_shmem_fault(struct vm_fault *= vmf) struct vm_area_struct *vma =3D vmf->vma; struct drm_gem_object *obj =3D vma->vm_private_data; struct drm_gem_shmem_object *shmem =3D to_drm_gem_shmem_obj(obj); - struct page *page; + struct page **pages =3D shmem->pages; pgoff_t page_offset; + unsigned long pfn; vm_fault_t ret; =20 /* Offset to faulty address in the VMA (without the fake offset). */ @@ -628,12 +657,18 @@ static vm_fault_t drm_gem_shmem_fault(struct vm_fault= *vmf) =20 if (unlikely(!drm_gem_shmem_fault_is_valid(obj, page_offset))) { ret =3D VM_FAULT_SIGBUS; - } else { - page =3D shmem->pages[page_offset]; + goto out; + } =20 - ret =3D vmf_insert_pfn(vma, vmf->address, page_to_pfn(page)); + if (drm_gem_shmem_map_pmd(vmf, vmf->address, pages[page_offset])) { + ret =3D VM_FAULT_NOPAGE; + goto out; } =20 + pfn =3D page_to_pfn(pages[page_offset]); + ret =3D vmf_insert_pfn(vma, vmf->address, pfn); + + out: dma_resv_unlock(shmem->base.resv); =20 return ret; --=20 2.47.3 From nobody Fri Dec 19 20:55:28 2025 Received: from bali.collaboradmins.com (bali.collaboradmins.com [148.251.105.195]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by smtp.subspace.kernel.org (Postfix) with ESMTPS id 1365A306499; Wed, 15 Oct 2025 15:30:37 +0000 (UTC) Authentication-Results: smtp.subspace.kernel.org; arc=none smtp.client-ip=148.251.105.195 ARC-Seal: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1760542239; cv=none; b=FSeobUqoCGPFa+bLu8mUBogDZmPO93dOlUe162h88zdrX3KfKqp4g6UnuvlhvtAP9yiU5aLCqvki5Lqgypfx6JE6FfHVLO0dvOQjVlMeiIzfujQpI4TBaYsGpo7bCAGGdS5B069uDlICyobHUSZq/eSpdMdPjzYGaR8JzCb5Mhs= ARC-Message-Signature: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1760542239; c=relaxed/simple; bh=XKuTiQL7iYB4W4XaqtHtHTqK+/eQHDdkS3/taRaSYOk=; h=From:To:Cc:Subject:Date:Message-ID:In-Reply-To:References: MIME-Version:Content-Type; b=Gy84LzrZeAhJ/mufzbo3SJIXP/R//naeFr4M5U9JLPRQEkLj5N+f51PCA4RkWKtF31GkEUYLtMrhd+7FNCeYgCnNl79V4xFUmN85QZs8RWXDtmyChylCMs5WPPNXqR7kjsqcOrtOCkbwGKg4RepjSUdrI+Y17y1ZGzcrWQ3Ze34= ARC-Authentication-Results: i=1; smtp.subspace.kernel.org; dmarc=pass (p=none dis=none) header.from=collabora.com; spf=pass smtp.mailfrom=collabora.com; dkim=pass (2048-bit key) header.d=collabora.com header.i=@collabora.com header.b=mVbq0vwh; arc=none smtp.client-ip=148.251.105.195 Authentication-Results: smtp.subspace.kernel.org; dmarc=pass (p=none dis=none) header.from=collabora.com Authentication-Results: smtp.subspace.kernel.org; spf=pass smtp.mailfrom=collabora.com Authentication-Results: smtp.subspace.kernel.org; dkim=pass (2048-bit key) header.d=collabora.com header.i=@collabora.com header.b="mVbq0vwh" DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=collabora.com; s=mail; t=1760542235; bh=XKuTiQL7iYB4W4XaqtHtHTqK+/eQHDdkS3/taRaSYOk=; h=From:To:Cc:Subject:Date:In-Reply-To:References:From; b=mVbq0vwhzXLTQmUr3cUXRSZv5u2Vjx188EMpQ8VT6qFZ3eXbEBOdLcSgpE4iSf+8n 1lw9mZq31by5uV9nU1P+PNu8VprXKM7PjGMt/m6mBKK3agI8ucX6d6HI4n7UurURP/ OagbXDcLrFQ4vM2Iu/e8uZ3LyXw9bm+pl6qY1XLbWY0wqi/pUEPcoFPFdGE+OO01oz e0uHRmOhs75evuLxLYQlaW5pGeJLPr5QI1IivQdGtaK8qGrMbj5apZvRqG7912CeaY tahOhS1PzIkfg7LWNMn5BCsdIg4THp/VRL0Cu6TVWc3D6jzcND1CxtwQdAdnnYPaL5 fZ3BlTlTJKT1w== Received: from debian-rockchip-rock5b-rk3588.. (unknown [IPv6:2a01:e0a:5e3:6100:826d:bc07:e98c:84a]) (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: loicmolinari) by bali.collaboradmins.com (Postfix) with ESMTPSA id AF63317E1305; Wed, 15 Oct 2025 17:30:34 +0200 (CEST) From: =?UTF-8?q?Lo=C3=AFc=20Molinari?= To: Maarten Lankhorst , Maxime Ripard , Thomas Zimmermann , David Airlie , Simona Vetter , Jani Nikula , Joonas Lahtinen , Rodrigo Vivi , Tvrtko Ursulin , Boris Brezillon , Rob Herring , Steven Price , Liviu Dudau , Melissa Wen , =?UTF-8?q?Ma=C3=ADra=20Canal?= , Hugh Dickins , Baolin Wang , Andrew Morton , =?UTF-8?q?Lo=C3=AFc=20Molinari?= , Al Viro , =?UTF-8?q?Miko=C5=82aj=20Wasiak?= , Christian Brauner , Nitin Gote , Andi Shyti , Jonathan Corbet , Christopher Healy , Matthew Wilcox , Bagas Sanjaya Cc: linux-kernel@vger.kernel.org, dri-devel@lists.freedesktop.org, intel-gfx@lists.freedesktop.org, linux-mm@kvack.org, linux-doc@vger.kernel.org, kernel@collabora.com Subject: [PATCH v4 04/13] drm/gem: Introduce drm_gem_get_unmapped_area() fop Date: Wed, 15 Oct 2025 17:30:08 +0200 Message-ID: <20251015153018.43735-5-loic.molinari@collabora.com> X-Mailer: git-send-email 2.47.3 In-Reply-To: <20251015153018.43735-1-loic.molinari@collabora.com> References: <20251015153018.43735-1-loic.molinari@collabora.com> Precedence: bulk X-Mailing-List: linux-kernel@vger.kernel.org List-Id: List-Subscribe: List-Unsubscribe: MIME-Version: 1.0 Content-Type: text/plain; charset="utf-8" Content-Transfer-Encoding: quoted-printable mmap() calls on the DRM file pointer currently always end up using mm_get_unmapped_area() to get a free mapping region. On builds with CONFIG_TRANSPARENT_HUGEPAGE enabled, this isn't ideal for GEM objects backed by shmem buffers on mount points setting the 'huge=3D' option because it can't correctly figure out the potentially huge address alignment required. This commit introduces the drm_gem_get_unmapped_area() function which is meant to be used as a get_unmapped_area file operation on the DRM file pointer to lookup GEM objects based on their fake offsets and get a properly aligned region by calling shmem_get_unmapped_area() with the right file pointer. If a GEM object isn't available at the given offset or if the caller isn't granted access to it, the function falls back to mm_get_unmapped_area(). This also makes drm_gem_get_unmapped_area() part of the default GEM file operations so that all the DRM drivers can benefit from more efficient mappings thanks to the huge page fault handler introduced in previous commit 'drm/shmem-helper: Add huge page fault handler'. The shmem_get_unmapped_area() function needs to be exported so that it can be used from the DRM subsystem. v3: - add missing include: 'linux/sched/mm.h' - forward to shmem layer in builds with CONFIG_TRANSPARENT_HUGEPAGE=3Dn Signed-off-by: Lo=C3=AFc Molinari --- drivers/gpu/drm/drm_gem.c | 107 ++++++++++++++++++++++++++++++-------- include/drm/drm_gem.h | 4 ++ mm/shmem.c | 1 + 3 files changed, 90 insertions(+), 22 deletions(-) diff --git a/drivers/gpu/drm/drm_gem.c b/drivers/gpu/drm/drm_gem.c index a1a9c828938b..a98d5744cc6c 100644 --- a/drivers/gpu/drm/drm_gem.c +++ b/drivers/gpu/drm/drm_gem.c @@ -36,6 +36,7 @@ #include #include #include +#include #include #include #include @@ -1187,36 +1188,27 @@ int drm_gem_mmap_obj(struct drm_gem_object *obj, un= signed long obj_size, } EXPORT_SYMBOL(drm_gem_mmap_obj); =20 -/** - * drm_gem_mmap - memory map routine for GEM objects - * @filp: DRM file pointer - * @vma: VMA for the area to be mapped - * - * If a driver supports GEM object mapping, mmap calls on the DRM file - * descriptor will end up here. - * - * Look up the GEM object based on the offset passed in (vma->vm_pgoff will - * contain the fake offset we created when the GTT map ioctl was called on - * the object) and map it with a call to drm_gem_mmap_obj(). - * - * If the caller is not granted access to the buffer object, the mmap will= fail - * with EACCES. Please see the vma manager for more information. +/* + * Look up a GEM object in offset space based on the exact start address. = The + * caller must be granted access to the object. Returns a GEM object on su= ccess + * or a negative error code on failure. The returned GEM object needs to be + * released with drm_gem_object_put(). */ -int drm_gem_mmap(struct file *filp, struct vm_area_struct *vma) +static struct drm_gem_object * +drm_gem_object_lookup_from_offset(struct file *filp, unsigned long start, + unsigned long pages) { struct drm_file *priv =3D filp->private_data; struct drm_device *dev =3D priv->minor->dev; struct drm_gem_object *obj =3D NULL; struct drm_vma_offset_node *node; - int ret; =20 if (drm_dev_is_unplugged(dev)) - return -ENODEV; + return ERR_PTR(-ENODEV); =20 drm_vma_offset_lock_lookup(dev->vma_offset_manager); node =3D drm_vma_offset_exact_lookup_locked(dev->vma_offset_manager, - vma->vm_pgoff, - vma_pages(vma)); + start, pages); if (likely(node)) { obj =3D container_of(node, struct drm_gem_object, vma_node); /* @@ -1235,14 +1227,85 @@ int drm_gem_mmap(struct file *filp, struct vm_area_= struct *vma) drm_vma_offset_unlock_lookup(dev->vma_offset_manager); =20 if (!obj) - return -EINVAL; + return ERR_PTR(-EINVAL); =20 if (!drm_vma_node_is_allowed(node, priv)) { drm_gem_object_put(obj); - return -EACCES; + return ERR_PTR(-EACCES); } =20 - ret =3D drm_gem_mmap_obj(obj, drm_vma_node_size(node) << PAGE_SHIFT, + return obj; +} + +/** + * drm_gem_get_unmapped_area - get memory mapping region routine for GEM o= bjects + * @filp: DRM file pointer + * @uaddr: User address hint + * @len: Mapping length + * @pgoff: Offset (in pages) + * @flags: Mapping flags + * + * If a driver supports GEM object mapping, before ending up in drm_gem_mm= ap(), + * mmap calls on the DRM file descriptor will first try to find a free lin= ear + * address space large enough for a mapping. Since GEM objects are backed = by + * shmem buffers, this should preferably be handled by the shmem virtual m= emory + * filesystem which can appropriately align addresses to huge page sizes w= hen + * needed. + * + * Look up the GEM object based on the offset passed in (vma->vm_pgoff will + * contain the fake offset we created) and call shmem_get_unmapped_area() = with + * the right file pointer. + * + * If a GEM object is not available at the given offset or if the caller i= s not + * granted access to it, fall back to mm_get_unmapped_area(). + */ +unsigned long drm_gem_get_unmapped_area(struct file *filp, unsigned long u= addr, + unsigned long len, unsigned long pgoff, + unsigned long flags) +{ + struct drm_gem_object *obj; + unsigned long ret; + + obj =3D drm_gem_object_lookup_from_offset(filp, pgoff, len >> PAGE_SHIFT); + if (IS_ERR(obj)) + return mm_get_unmapped_area(current->mm, filp, uaddr, len, 0, + flags); + + ret =3D shmem_get_unmapped_area(obj->filp, uaddr, len, 0, flags); + + drm_gem_object_put(obj); + + return ret; +} +EXPORT_SYMBOL(drm_gem_get_unmapped_area); + +/** + * drm_gem_mmap - memory map routine for GEM objects + * @filp: DRM file pointer + * @vma: VMA for the area to be mapped + * + * If a driver supports GEM object mapping, mmap calls on the DRM file + * descriptor will end up here. + * + * Look up the GEM object based on the offset passed in (vma->vm_pgoff will + * contain the fake offset we created) and map it with a call to + * drm_gem_mmap_obj(). + * + * If the caller is not granted access to the buffer object, the mmap will= fail + * with EACCES. Please see the vma manager for more information. + */ +int drm_gem_mmap(struct file *filp, struct vm_area_struct *vma) +{ + struct drm_gem_object *obj; + int ret; + + obj =3D drm_gem_object_lookup_from_offset(filp, vma->vm_pgoff, + vma_pages(vma)); + if (IS_ERR(obj)) + return PTR_ERR(obj); + + ret =3D drm_gem_mmap_obj(obj, + drm_vma_node_size(&obj->vma_node) << PAGE_SHIFT, vma); =20 drm_gem_object_put(obj); diff --git a/include/drm/drm_gem.h b/include/drm/drm_gem.h index 8d48d2af2649..7c8bd67d087c 100644 --- a/include/drm/drm_gem.h +++ b/include/drm/drm_gem.h @@ -469,6 +469,7 @@ struct drm_gem_object { .poll =3D drm_poll,\ .read =3D drm_read,\ .llseek =3D noop_llseek,\ + .get_unmapped_area =3D drm_gem_get_unmapped_area,\ .mmap =3D drm_gem_mmap, \ .fop_flags =3D FOP_UNSIGNED_OFFSET =20 @@ -506,6 +507,9 @@ void drm_gem_vm_close(struct vm_area_struct *vma); int drm_gem_mmap_obj(struct drm_gem_object *obj, unsigned long obj_size, struct vm_area_struct *vma); int drm_gem_mmap(struct file *filp, struct vm_area_struct *vma); +unsigned long drm_gem_get_unmapped_area(struct file *filp, unsigned long u= addr, + unsigned long len, unsigned long pgoff, + unsigned long flags); =20 /** * drm_gem_object_get - acquire a GEM buffer object reference diff --git a/mm/shmem.c b/mm/shmem.c index b9081b817d28..612218fc95cb 100644 --- a/mm/shmem.c +++ b/mm/shmem.c @@ -2851,6 +2851,7 @@ unsigned long shmem_get_unmapped_area(struct file *fi= le, return addr; return inflated_addr; } +EXPORT_SYMBOL_GPL(shmem_get_unmapped_area); =20 #ifdef CONFIG_NUMA static int shmem_set_policy(struct vm_area_struct *vma, struct mempolicy *= mpol) --=20 2.47.3 From nobody Fri Dec 19 20:55:28 2025 Received: from bali.collaboradmins.com (bali.collaboradmins.com [148.251.105.195]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by smtp.subspace.kernel.org (Postfix) with ESMTPS id 8E9BC308F26; Wed, 15 Oct 2025 15:30:39 +0000 (UTC) Authentication-Results: smtp.subspace.kernel.org; arc=none smtp.client-ip=148.251.105.195 ARC-Seal: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1760542241; cv=none; b=jH/9UzdXvIfaKOdDW/kL9bMeeVruLa4Aj8gM7keCJrw5BiixeAKVxbqtzf8RB9I4imanVPQ+1NwY3QihzCgwRwHuAwZvIUWUv0Y+nJ2/rp+NnzRI7+CG1uLcwewCLBc0TtVhLJjJy5w66tLD5ajujf9+wpxdIkTpoUE/iUjQI40= ARC-Message-Signature: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1760542241; c=relaxed/simple; bh=2YyHnJ7MG9Sk+eB8X8EQ2czQs/R8vZqVR26EnmlFgQo=; h=From:To:Cc:Subject:Date:Message-ID:In-Reply-To:References: MIME-Version:Content-Type; b=omm7Qx70qNPiRc3stV4heUmk2XVHoowcCWeVKql5CA2xe312f+1zrV6mpqqmyVIdAM5/91GfLB5Z/ZpDlWNwBgk9VCWPfou/kvBECidlWmk+3jmMZCFXDlDaVRgpSDHWCAcd5A3BaNiuPGAonZZzHckWK3hMa5UO+oHKQNBZ8B8= ARC-Authentication-Results: i=1; smtp.subspace.kernel.org; dmarc=pass (p=none dis=none) header.from=collabora.com; spf=pass smtp.mailfrom=collabora.com; dkim=pass (2048-bit key) header.d=collabora.com header.i=@collabora.com header.b=anvXxSpt; arc=none smtp.client-ip=148.251.105.195 Authentication-Results: smtp.subspace.kernel.org; dmarc=pass (p=none dis=none) header.from=collabora.com Authentication-Results: smtp.subspace.kernel.org; spf=pass smtp.mailfrom=collabora.com Authentication-Results: smtp.subspace.kernel.org; dkim=pass (2048-bit key) header.d=collabora.com header.i=@collabora.com header.b="anvXxSpt" DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=collabora.com; s=mail; t=1760542237; bh=2YyHnJ7MG9Sk+eB8X8EQ2czQs/R8vZqVR26EnmlFgQo=; h=From:To:Cc:Subject:Date:In-Reply-To:References:From; b=anvXxSptCiIgOuEsSH5DjAHWDBZLh4/OWy4y/sLeOha9IsxpqBMOWJtzGnyVjvGr4 WQqx8qzJeyzJ7puOnBg7dq6o/RuIDKgWjnBYWcECEJtrlf6nX8RZWB/8dEIL7LZDJc L9HBPWOekFbsGbnW9LvpukRcrPBU2bHB5IPo9SUsO2ycN6l0lC3btFYNbH+Q7CbSDx ezu6N8IZPuOQ6ZC3Y7velZQ80GGHxahF0nrYtgMUixxqU2LG8Cd0z8pnoPxrR5kXru 7a+U5qeZ1EMEFPAfuEaHddCtLHMgLPsYXnftfOaedjeszqK5brUqS8SWeTgNMPnnqX vO7qbk4bCp1ag== Received: from debian-rockchip-rock5b-rk3588.. (unknown [IPv6:2a01:e0a:5e3:6100:826d:bc07:e98c:84a]) (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: loicmolinari) by bali.collaboradmins.com (Postfix) with ESMTPSA id 0D01E17E0CA3; Wed, 15 Oct 2025 17:30:37 +0200 (CEST) From: =?UTF-8?q?Lo=C3=AFc=20Molinari?= To: Maarten Lankhorst , Maxime Ripard , Thomas Zimmermann , David Airlie , Simona Vetter , Jani Nikula , Joonas Lahtinen , Rodrigo Vivi , Tvrtko Ursulin , Boris Brezillon , Rob Herring , Steven Price , Liviu Dudau , Melissa Wen , =?UTF-8?q?Ma=C3=ADra=20Canal?= , Hugh Dickins , Baolin Wang , Andrew Morton , =?UTF-8?q?Lo=C3=AFc=20Molinari?= , Al Viro , =?UTF-8?q?Miko=C5=82aj=20Wasiak?= , Christian Brauner , Nitin Gote , Andi Shyti , Jonathan Corbet , Christopher Healy , Matthew Wilcox , Bagas Sanjaya Cc: linux-kernel@vger.kernel.org, dri-devel@lists.freedesktop.org, intel-gfx@lists.freedesktop.org, linux-mm@kvack.org, linux-doc@vger.kernel.org, kernel@collabora.com Subject: [PATCH v4 05/13] drm/gem: Add huge tmpfs mount point helper Date: Wed, 15 Oct 2025 17:30:09 +0200 Message-ID: <20251015153018.43735-6-loic.molinari@collabora.com> X-Mailer: git-send-email 2.47.3 In-Reply-To: <20251015153018.43735-1-loic.molinari@collabora.com> References: <20251015153018.43735-1-loic.molinari@collabora.com> Precedence: bulk X-Mailing-List: linux-kernel@vger.kernel.org List-Id: List-Subscribe: List-Unsubscribe: MIME-Version: 1.0 Content-Type: text/plain; charset="utf-8" Content-Transfer-Encoding: quoted-printable Add the drm_gem_huge_mnt_create() helper to avoid code duplication in the i915, V3D, Panfrost and Panthor drivers. It creates and mounts a dedicated huge tmpfs mountpoint, for the lifetime of a DRM device, used at GEM object initialization. The next commits will port drivers to this helper. v3: - store huge tmpfs mountpoint in drm_device v4: - return 0 in builds with CONFIG_TRANSPARENT_HUGEPAGE=3Dn - return 0 when huge_mnt already exists Signed-off-by: Lo=C3=AFc Molinari --- drivers/gpu/drm/drm_gem.c | 58 +++++++++++++++++++++++++++++++++++++++ include/drm/drm_device.h | 11 ++++++++ include/drm/drm_gem.h | 1 + 3 files changed, 70 insertions(+) diff --git a/drivers/gpu/drm/drm_gem.c b/drivers/gpu/drm/drm_gem.c index a98d5744cc6c..db8c0a217add 100644 --- a/drivers/gpu/drm/drm_gem.c +++ b/drivers/gpu/drm/drm_gem.c @@ -29,6 +29,7 @@ #include #include #include +#include #include #include #include @@ -82,6 +83,63 @@ * up at a later date, and as our interface with shmfs for memory allocati= on. */ =20 +static void drm_gem_huge_mnt_free(struct drm_device *dev, void *data) +{ + drm_WARN_ON(dev, dev->huge_mnt =3D=3D NULL); + + kern_unmount(dev->huge_mnt); + dev->huge_mnt =3D NULL; +} + +/** + * drm_gem_huge_mnt_create - Create, mount and use a huge tmpfs mountpoint + * @dev: drm_device a huge tmpfs mountpoint should be used with + * @value: huge tmpfs mount option value + * + * This function creates and mounts a dedicated huge tmpfs mountpoint for = the + * lifetime of the drm device @dev which is used at GEM object initializat= ion + * with drm_gem_object_init(). + * + * The most common option value @value is "within_size" which only allocat= es + * huge pages if the page will be fully within the GEM object size. "alway= s", + * "advise" and "never" are supported too but the latter would just create= a + * mountpoint similar to the default one (`shm_mnt`). See shmemfs and + * Transparent Hugepage for more information. + * + * Returns: + * 0 on success or a negative error code on failure. + */ +int drm_gem_huge_mnt_create(struct drm_device *dev, const char *value) +{ + struct file_system_type *type; + struct fs_context *fc; + int ret; + + if (!IS_ENABLED(CONFIG_TRANSPARENT_HUGEPAGE)) + return 0; + if (unlikely(dev->huge_mnt)) + return 0; + + type =3D get_fs_type("tmpfs"); + if (unlikely(!type)) + return -EOPNOTSUPP; + fc =3D fs_context_for_mount(type, SB_KERNMOUNT); + if (IS_ERR(fc)) + return PTR_ERR(fc); + ret =3D vfs_parse_fs_string(fc, "source", "tmpfs"); + if (unlikely(ret)) + return -ENOPARAM; + ret =3D vfs_parse_fs_string(fc, "huge", value); + if (unlikely(ret)) + return -ENOPARAM; + + dev->huge_mnt =3D fc_mount_longterm(fc); + put_fs_context(fc); + + return drmm_add_action_or_reset(dev, drm_gem_huge_mnt_free, NULL); +} +EXPORT_SYMBOL_GPL(drm_gem_huge_mnt_create); + static void drm_gem_init_release(struct drm_device *dev, void *ptr) { diff --git a/include/drm/drm_device.h b/include/drm/drm_device.h index 778b2cca6c49..352e3db402d7 100644 --- a/include/drm/drm_device.h +++ b/include/drm/drm_device.h @@ -3,6 +3,7 @@ =20 #include #include +#include #include #include #include @@ -168,6 +169,16 @@ struct drm_device { */ struct drm_master *master; =20 + /** + * @huge_mnt: + * + * Huge tmpfs mountpoint used at GEM object initialization + * drm_gem_object_init(). Drivers can call drm_gem_huge_mnt_create() to + * create a huge tmfps mountpoint. The default tmpfs mountpoint + * (`shm_mnt`) is used if NULL. + */ + struct vfsmount *huge_mnt; + /** * @driver_features: per-device driver features * diff --git a/include/drm/drm_gem.h b/include/drm/drm_gem.h index 7c8bd67d087c..7285a62d9afc 100644 --- a/include/drm/drm_gem.h +++ b/include/drm/drm_gem.h @@ -492,6 +492,7 @@ struct drm_gem_object { DRM_GEM_FOPS,\ } =20 +int drm_gem_huge_mnt_create(struct drm_device *dev, const char *value); void drm_gem_object_release(struct drm_gem_object *obj); void drm_gem_object_free(struct kref *kref); int drm_gem_object_init(struct drm_device *dev, --=20 2.47.3 From nobody Fri Dec 19 20:55:28 2025 Received: from bali.collaboradmins.com (bali.collaboradmins.com [148.251.105.195]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by smtp.subspace.kernel.org (Postfix) with ESMTPS id 23CC6309DB0; Wed, 15 Oct 2025 15:30:41 +0000 (UTC) Authentication-Results: smtp.subspace.kernel.org; arc=none smtp.client-ip=148.251.105.195 ARC-Seal: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1760542244; cv=none; b=RxlzoIUpthSZ1Uab2O70wR8l5VjBQhQZoL77wAaW5AIrLl2dOxsEtF7Z2F6s5yoDWu0oh8X86EY8D1yD99RhgRJjotajeb4tbyp3EUanG1y3B3jImS2B112JFjafr1xg3w/6jaGyeaNuwmMkLQmJS9Qmvga/TM9x8aicwkgk+98= ARC-Message-Signature: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1760542244; c=relaxed/simple; bh=NNGZxqXvdZMXNrsqK35o/grwPvXGY5zrGR9jS5Fvnhs=; h=From:To:Cc:Subject:Date:Message-ID:In-Reply-To:References: MIME-Version:Content-Type; b=Gy2V9rxQx24hcGjUH8ID6whX3Pt9dNzQjvgoCGKIO6005GUVvmfffe+2Ti3wiLM1rRaRKROxvBOCRtmonBrNsxpEFg0oVn6IbXUauLfTCFSPB/rYHbtDX86MepfLkMI3ppaLqs38LRdHxXRGV3x7zMkRgXBJzmNNqukpPIu1tVg= ARC-Authentication-Results: i=1; smtp.subspace.kernel.org; dmarc=pass (p=none dis=none) header.from=collabora.com; spf=pass smtp.mailfrom=collabora.com; dkim=pass (2048-bit key) header.d=collabora.com header.i=@collabora.com header.b=C7YyY+pF; arc=none smtp.client-ip=148.251.105.195 Authentication-Results: smtp.subspace.kernel.org; dmarc=pass (p=none dis=none) header.from=collabora.com Authentication-Results: smtp.subspace.kernel.org; spf=pass smtp.mailfrom=collabora.com Authentication-Results: smtp.subspace.kernel.org; dkim=pass (2048-bit key) header.d=collabora.com header.i=@collabora.com header.b="C7YyY+pF" DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=collabora.com; s=mail; t=1760542240; bh=NNGZxqXvdZMXNrsqK35o/grwPvXGY5zrGR9jS5Fvnhs=; h=From:To:Cc:Subject:Date:In-Reply-To:References:From; b=C7YyY+pF+WioRCOzFcpU5hYsJASu/Uwd/V4lI6X7oQeUH5Jfv/iu42d8KI4RpAW3I AiO0dOC47FAIlNc843ATbqQg8U2Y3EbiZu5O1skmd2qD22pADXnlY5gtB3wgIQ6x7A tRk8/u+U6u9hBcKFIAjsBZLT+4LklAdNCEdjDCZcsFmBe8aEYvvgbGB5Wp9snt0j3F XYVLAnDJkqaVjh8E0SP8zd15KTkAiTU9EfXpmUIndLCuvrUXcLWt4yrwKXWANdu+Rz ejp4UyEmTUqodCdMhIB8YfcM2uRn9aLqBh3XJYz7iIY0VjmfiqUSIsU7VzRs687sSQ U1SonWE2VRXEg== Received: from debian-rockchip-rock5b-rk3588.. (unknown [IPv6:2a01:e0a:5e3:6100:826d:bc07:e98c:84a]) (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: loicmolinari) by bali.collaboradmins.com (Postfix) with ESMTPSA id 9154817E00A3; Wed, 15 Oct 2025 17:30:39 +0200 (CEST) From: =?UTF-8?q?Lo=C3=AFc=20Molinari?= To: Maarten Lankhorst , Maxime Ripard , Thomas Zimmermann , David Airlie , Simona Vetter , Jani Nikula , Joonas Lahtinen , Rodrigo Vivi , Tvrtko Ursulin , Boris Brezillon , Rob Herring , Steven Price , Liviu Dudau , Melissa Wen , =?UTF-8?q?Ma=C3=ADra=20Canal?= , Hugh Dickins , Baolin Wang , Andrew Morton , =?UTF-8?q?Lo=C3=AFc=20Molinari?= , Al Viro , =?UTF-8?q?Miko=C5=82aj=20Wasiak?= , Christian Brauner , Nitin Gote , Andi Shyti , Jonathan Corbet , Christopher Healy , Matthew Wilcox , Bagas Sanjaya Cc: linux-kernel@vger.kernel.org, dri-devel@lists.freedesktop.org, intel-gfx@lists.freedesktop.org, linux-mm@kvack.org, linux-doc@vger.kernel.org, kernel@collabora.com Subject: [PATCH v4 06/13] drm/i915: Use huge tmpfs mount point helper Date: Wed, 15 Oct 2025 17:30:10 +0200 Message-ID: <20251015153018.43735-7-loic.molinari@collabora.com> X-Mailer: git-send-email 2.47.3 In-Reply-To: <20251015153018.43735-1-loic.molinari@collabora.com> References: <20251015153018.43735-1-loic.molinari@collabora.com> Precedence: bulk X-Mailing-List: linux-kernel@vger.kernel.org List-Id: List-Subscribe: List-Unsubscribe: MIME-Version: 1.0 Content-Type: text/plain; charset="utf-8" Content-Transfer-Encoding: quoted-printable Make use of the new drm_gem_huge_mnt_create() helper to avoid code duplication. Now that it's just a few lines long, the single function in i915_gemfs.c is moved into v3d_gem_shmem.c. v3: - use huge tmpfs mountpoint in drm_device - move i915_gemfs.c into i915_gem_shmem.c v4: - clean up mountpoint creation error handling Signed-off-by: Lo=C3=AFc Molinari --- drivers/gpu/drm/i915/Makefile | 3 +- drivers/gpu/drm/i915/gem/i915_gem_shmem.c | 47 +++++++++---- drivers/gpu/drm/i915/gem/i915_gemfs.c | 69 ------------------- drivers/gpu/drm/i915/gem/i915_gemfs.h | 14 ---- .../gpu/drm/i915/gem/selftests/huge_pages.c | 10 +-- drivers/gpu/drm/i915/i915_drv.h | 5 -- 6 files changed, 40 insertions(+), 108 deletions(-) delete mode 100644 drivers/gpu/drm/i915/gem/i915_gemfs.c delete mode 100644 drivers/gpu/drm/i915/gem/i915_gemfs.h diff --git a/drivers/gpu/drm/i915/Makefile b/drivers/gpu/drm/i915/Makefile index e58c0c158b3a..e22393a7cf6f 100644 --- a/drivers/gpu/drm/i915/Makefile +++ b/drivers/gpu/drm/i915/Makefile @@ -163,8 +163,7 @@ gem-y +=3D \ gem/i915_gem_ttm_move.o \ gem/i915_gem_ttm_pm.o \ gem/i915_gem_userptr.o \ - gem/i915_gem_wait.o \ - gem/i915_gemfs.o + gem/i915_gem_wait.o i915-y +=3D \ $(gem-y) \ i915_active.o \ diff --git a/drivers/gpu/drm/i915/gem/i915_gem_shmem.c b/drivers/gpu/drm/i9= 15/gem/i915_gem_shmem.c index b9dae15c1d16..a72d98f1202f 100644 --- a/drivers/gpu/drm/i915/gem/i915_gem_shmem.c +++ b/drivers/gpu/drm/i915/gem/i915_gem_shmem.c @@ -9,14 +9,15 @@ #include =20 #include +#include =20 #include "gem/i915_gem_region.h" #include "i915_drv.h" #include "i915_gem_object.h" #include "i915_gem_tiling.h" -#include "i915_gemfs.h" #include "i915_scatterlist.h" #include "i915_trace.h" +#include "i915_utils.h" =20 /* * Move folios to appropriate lru and release the batch, decrementing the @@ -506,9 +507,9 @@ static int __create_shmem(struct drm_i915_private *i915, if (BITS_PER_LONG =3D=3D 64 && size > MAX_LFS_FILESIZE) return -E2BIG; =20 - if (i915->mm.gemfs) - filp =3D shmem_file_setup_with_mnt(i915->mm.gemfs, "i915", size, - flags); + if (i915->drm.huge_mnt) + filp =3D shmem_file_setup_with_mnt(i915->drm.huge_mnt, "i915", + size, flags); else filp =3D shmem_file_setup("i915", size, flags); if (IS_ERR(filp)) @@ -635,21 +636,41 @@ i915_gem_object_create_shmem_from_data(struct drm_i91= 5_private *i915, =20 static int init_shmem(struct intel_memory_region *mem) { - i915_gemfs_init(mem->i915); - intel_memory_region_set_name(mem, "system"); + struct drm_i915_private *i915 =3D mem->i915; + int err; =20 - return 0; /* We have fallback to the kernel mnt if gemfs init failed. */ -} + /* + * By creating our own shmemfs mountpoint, we can pass in + * mount flags that better match our usecase. + * + * One example, although it is probably better with a per-file + * control, is selecting huge page allocations ("huge=3Dwithin_size"). + * However, we only do so on platforms which benefit from it, or to + * offset the overhead of iommu lookups, where with latter it is a net + * win even on platforms which would otherwise see some performance + * regressions such a slow reads issue on Broadwell and Skylake. + */ =20 -static int release_shmem(struct intel_memory_region *mem) -{ - i915_gemfs_fini(mem->i915); - return 0; + if (GRAPHICS_VER(i915) < 11 && !i915_vtd_active(i915)) + goto no_thp; + + err =3D drm_gem_huge_mnt_create(&i915->drm, "within_size"); + if (i915->drm.huge_mnt) + drm_info(&i915->drm, "Using Transparent Hugepages\n"); + else if (err) + drm_notice(&i915->drm, + "Transparent Hugepage support is recommended for optimal performance= %s\n", + GRAPHICS_VER(i915) >=3D 11 ? " on this platform!" : + " when IOMMU is enabled!"); + + no_thp: + intel_memory_region_set_name(mem, "system"); + + return 0; /* We have fallback to the kernel mnt if huge mnt failed. */ } =20 static const struct intel_memory_region_ops shmem_region_ops =3D { .init =3D init_shmem, - .release =3D release_shmem, .init_object =3D shmem_object_init, }; =20 diff --git a/drivers/gpu/drm/i915/gem/i915_gemfs.c b/drivers/gpu/drm/i915/g= em/i915_gemfs.c deleted file mode 100644 index 8f13ec4ff0d0..000000000000 --- a/drivers/gpu/drm/i915/gem/i915_gemfs.c +++ /dev/null @@ -1,69 +0,0 @@ -// SPDX-License-Identifier: MIT -/* - * Copyright =C2=A9 2017 Intel Corporation - */ - -#include -#include -#include - -#include "i915_drv.h" -#include "i915_gemfs.h" -#include "i915_utils.h" - -void i915_gemfs_init(struct drm_i915_private *i915) -{ - struct file_system_type *type; - struct fs_context *fc; - struct vfsmount *gemfs; - int ret; - - /* - * By creating our own shmemfs mountpoint, we can pass in - * mount flags that better match our usecase. - * - * One example, although it is probably better with a per-file - * control, is selecting huge page allocations ("huge=3Dwithin_size"). - * However, we only do so on platforms which benefit from it, or to - * offset the overhead of iommu lookups, where with latter it is a net - * win even on platforms which would otherwise see some performance - * regressions such a slow reads issue on Broadwell and Skylake. - */ - - if (GRAPHICS_VER(i915) < 11 && !i915_vtd_active(i915)) - return; - - if (!IS_ENABLED(CONFIG_TRANSPARENT_HUGEPAGE)) - goto err; - - type =3D get_fs_type("tmpfs"); - if (!type) - goto err; - - fc =3D fs_context_for_mount(type, SB_KERNMOUNT); - if (IS_ERR(fc)) - goto err; - ret =3D vfs_parse_fs_string(fc, "source", "tmpfs"); - if (!ret) - ret =3D vfs_parse_fs_string(fc, "huge", "within_size"); - if (!ret) - gemfs =3D fc_mount_longterm(fc); - put_fs_context(fc); - if (ret) - goto err; - - i915->mm.gemfs =3D gemfs; - drm_info(&i915->drm, "Using Transparent Hugepages\n"); - return; - -err: - drm_notice(&i915->drm, - "Transparent Hugepage support is recommended for optimal performance%= s\n", - GRAPHICS_VER(i915) >=3D 11 ? " on this platform!" : - " when IOMMU is enabled!"); -} - -void i915_gemfs_fini(struct drm_i915_private *i915) -{ - kern_unmount(i915->mm.gemfs); -} diff --git a/drivers/gpu/drm/i915/gem/i915_gemfs.h b/drivers/gpu/drm/i915/g= em/i915_gemfs.h deleted file mode 100644 index 16d4333c9a4e..000000000000 --- a/drivers/gpu/drm/i915/gem/i915_gemfs.h +++ /dev/null @@ -1,14 +0,0 @@ -/* SPDX-License-Identifier: MIT */ -/* - * Copyright =C2=A9 2017 Intel Corporation - */ - -#ifndef __I915_GEMFS_H__ -#define __I915_GEMFS_H__ - -struct drm_i915_private; - -void i915_gemfs_init(struct drm_i915_private *i915); -void i915_gemfs_fini(struct drm_i915_private *i915); - -#endif diff --git a/drivers/gpu/drm/i915/gem/selftests/huge_pages.c b/drivers/gpu/= drm/i915/gem/selftests/huge_pages.c index bd08605a1611..b41a38af63fd 100644 --- a/drivers/gpu/drm/i915/gem/selftests/huge_pages.c +++ b/drivers/gpu/drm/i915/gem/selftests/huge_pages.c @@ -1316,7 +1316,7 @@ typedef struct drm_i915_gem_object * =20 static inline bool igt_can_allocate_thp(struct drm_i915_private *i915) { - return i915->mm.gemfs && has_transparent_hugepage(); + return i915->drm.huge_mnt && has_transparent_hugepage(); } =20 static struct drm_i915_gem_object * @@ -1761,7 +1761,7 @@ static int igt_tmpfs_fallback(void *arg) struct drm_i915_private *i915 =3D arg; struct i915_address_space *vm; struct i915_gem_context *ctx; - struct vfsmount *gemfs =3D i915->mm.gemfs; + struct vfsmount *huge_mnt =3D i915->drm.huge_mnt; struct drm_i915_gem_object *obj; struct i915_vma *vma; struct file *file; @@ -1782,10 +1782,10 @@ static int igt_tmpfs_fallback(void *arg) /* * Make sure that we don't burst into a ball of flames upon falling back * to tmpfs, which we rely on if on the off-chance we encounter a failure - * when setting up gemfs. + * when setting up a huge mountpoint. */ =20 - i915->mm.gemfs =3D NULL; + i915->drm.huge_mnt =3D NULL; =20 obj =3D i915_gem_object_create_shmem(i915, PAGE_SIZE); if (IS_ERR(obj)) { @@ -1819,7 +1819,7 @@ static int igt_tmpfs_fallback(void *arg) out_put: i915_gem_object_put(obj); out_restore: - i915->mm.gemfs =3D gemfs; + i915->drm.huge_mnt =3D huge_mnt; =20 i915_vm_put(vm); out: diff --git a/drivers/gpu/drm/i915/i915_drv.h b/drivers/gpu/drm/i915/i915_dr= v.h index 6a768aad8edd..1bfee23e64a3 100644 --- a/drivers/gpu/drm/i915/i915_drv.h +++ b/drivers/gpu/drm/i915/i915_drv.h @@ -141,11 +141,6 @@ struct i915_gem_mm { */ atomic_t free_count; =20 - /** - * tmpfs instance used for shmem backed objects - */ - struct vfsmount *gemfs; - struct intel_memory_region *regions[INTEL_REGION_UNKNOWN]; =20 struct notifier_block oom_notifier; --=20 2.47.3 From nobody Fri Dec 19 20:55:28 2025 Received: from bali.collaboradmins.com (bali.collaboradmins.com [148.251.105.195]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by smtp.subspace.kernel.org (Postfix) with ESMTPS id 5077030BF66; Wed, 15 Oct 2025 15:30:44 +0000 (UTC) Authentication-Results: smtp.subspace.kernel.org; arc=none smtp.client-ip=148.251.105.195 ARC-Seal: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1760542246; cv=none; b=EATPMFMR6x9lFJbzuxgIaePsWWpixhIt0oe7IEY90G6vBaaBdtoaiAhNlOBqjd5/mLnFtMJ3WkSxJDVcymPTACU72phPmwi17njZ1OyGuOda2TNUbEvF9vtXyqFVW58/Ud2/j7JHHPLJzz//Oi2fw/UNxIOJV9wAZun0zHmVse0= ARC-Message-Signature: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1760542246; c=relaxed/simple; bh=mm3M5IWbHJSSC/HZMhCT0yYxpiSW+pNm8+0hCzyQoFk=; h=From:To:Cc:Subject:Date:Message-ID:In-Reply-To:References: MIME-Version:Content-Type; b=rTequtAlK/oAb8BONpSHP3rk3Q/2al8zMTlQ6uJb86aWgHeIepEDjVhRXlFQoET9t86QCU02s/KaRbvXd3Z1RgC34+K04Eo79KY6eUYFIuI4OabGC9r+TNSpv66TUe5/9rGqlgK/8o8JF3gybu/vMQ+dWDXeGK/oxdJez3W6lYs= ARC-Authentication-Results: i=1; smtp.subspace.kernel.org; dmarc=pass (p=none dis=none) header.from=collabora.com; spf=pass smtp.mailfrom=collabora.com; dkim=pass (2048-bit key) header.d=collabora.com header.i=@collabora.com header.b=OwIMJ0Dc; arc=none smtp.client-ip=148.251.105.195 Authentication-Results: smtp.subspace.kernel.org; dmarc=pass (p=none dis=none) header.from=collabora.com Authentication-Results: smtp.subspace.kernel.org; spf=pass smtp.mailfrom=collabora.com Authentication-Results: smtp.subspace.kernel.org; dkim=pass (2048-bit key) header.d=collabora.com header.i=@collabora.com header.b="OwIMJ0Dc" DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=collabora.com; s=mail; t=1760542242; bh=mm3M5IWbHJSSC/HZMhCT0yYxpiSW+pNm8+0hCzyQoFk=; h=From:To:Cc:Subject:Date:In-Reply-To:References:From; b=OwIMJ0Dcwtvomqe64Excosp0itSDMxWufpTakTNXPYLWib4FeWjtjeHWY7xm2UG60 xQlXW7kjLVStAfwwNr1TxLukaUhqqDQemjDzo17aKbEnXg2131uldcKncMzWzpTrs/ UoLw4Htf/fDKDkShypj5obO7tQ75u+i9+t6diLHoeroxFI5E1iyOaJVAY9C8yh8c/X VYyEaVp2SiGKL9OLN7JlZ9A9ag2sjBPj8rsphH9fOKTYcD6si6hDv1+eCVNiOo+r2t b17o16qp5C/yJpMDwmvJ1BtuwyQtLJbcQfN6yyL9ZT/Wj/Adu+vj+4lwNyAV0o+g3Y JRWWBQPK+3b9w== Received: from debian-rockchip-rock5b-rk3588.. (unknown [IPv6:2a01:e0a:5e3:6100:826d:bc07:e98c:84a]) (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: loicmolinari) by bali.collaboradmins.com (Postfix) with ESMTPSA id C1FFE17E055D; Wed, 15 Oct 2025 17:30:41 +0200 (CEST) From: =?UTF-8?q?Lo=C3=AFc=20Molinari?= To: Maarten Lankhorst , Maxime Ripard , Thomas Zimmermann , David Airlie , Simona Vetter , Jani Nikula , Joonas Lahtinen , Rodrigo Vivi , Tvrtko Ursulin , Boris Brezillon , Rob Herring , Steven Price , Liviu Dudau , Melissa Wen , =?UTF-8?q?Ma=C3=ADra=20Canal?= , Hugh Dickins , Baolin Wang , Andrew Morton , =?UTF-8?q?Lo=C3=AFc=20Molinari?= , Al Viro , =?UTF-8?q?Miko=C5=82aj=20Wasiak?= , Christian Brauner , Nitin Gote , Andi Shyti , Jonathan Corbet , Christopher Healy , Matthew Wilcox , Bagas Sanjaya Cc: linux-kernel@vger.kernel.org, dri-devel@lists.freedesktop.org, intel-gfx@lists.freedesktop.org, linux-mm@kvack.org, linux-doc@vger.kernel.org, kernel@collabora.com Subject: [PATCH v4 07/13] drm/v3d: Use huge tmpfs mount point helper Date: Wed, 15 Oct 2025 17:30:11 +0200 Message-ID: <20251015153018.43735-8-loic.molinari@collabora.com> X-Mailer: git-send-email 2.47.3 In-Reply-To: <20251015153018.43735-1-loic.molinari@collabora.com> References: <20251015153018.43735-1-loic.molinari@collabora.com> Precedence: bulk X-Mailing-List: linux-kernel@vger.kernel.org List-Id: List-Subscribe: List-Unsubscribe: MIME-Version: 1.0 Content-Type: text/plain; charset="utf-8" Content-Transfer-Encoding: quoted-printable Make use of the new drm_gem_huge_mnt_create() helper to avoid code duplication. Now that it's just a few lines long, the single function in v3d_gemfs.c is moved into v3d_gem.c. v3: - use huge tmpfs mountpoint in drm_device - move v3d_gemfs.c into v3d_gem.c v4: - clean up mountpoint creation error handling Signed-off-by: Lo=C3=AFc Molinari --- drivers/gpu/drm/v3d/Makefile | 3 +- drivers/gpu/drm/v3d/v3d_bo.c | 5 ++- drivers/gpu/drm/v3d/v3d_drv.c | 2 +- drivers/gpu/drm/v3d/v3d_drv.h | 11 +----- drivers/gpu/drm/v3d/v3d_gem.c | 27 +++++++++++++-- drivers/gpu/drm/v3d/v3d_gemfs.c | 60 --------------------------------- 6 files changed, 30 insertions(+), 78 deletions(-) delete mode 100644 drivers/gpu/drm/v3d/v3d_gemfs.c diff --git a/drivers/gpu/drm/v3d/Makefile b/drivers/gpu/drm/v3d/Makefile index fcf710926057..b7d673f1153b 100644 --- a/drivers/gpu/drm/v3d/Makefile +++ b/drivers/gpu/drm/v3d/Makefile @@ -13,8 +13,7 @@ v3d-y :=3D \ v3d_trace_points.o \ v3d_sched.o \ v3d_sysfs.o \ - v3d_submit.o \ - v3d_gemfs.o + v3d_submit.o =20 v3d-$(CONFIG_DEBUG_FS) +=3D v3d_debugfs.o =20 diff --git a/drivers/gpu/drm/v3d/v3d_bo.c b/drivers/gpu/drm/v3d/v3d_bo.c index c41476ddde68..6b9909bfce82 100644 --- a/drivers/gpu/drm/v3d/v3d_bo.c +++ b/drivers/gpu/drm/v3d/v3d_bo.c @@ -112,7 +112,7 @@ v3d_bo_create_finish(struct drm_gem_object *obj) if (IS_ERR(sgt)) return PTR_ERR(sgt); =20 - if (!v3d->gemfs) + if (!obj->dev->huge_mnt) align =3D SZ_4K; else if (obj->size >=3D SZ_1M) align =3D SZ_1M; @@ -148,12 +148,11 @@ struct v3d_bo *v3d_bo_create(struct drm_device *dev, = struct drm_file *file_priv, size_t unaligned_size) { struct drm_gem_shmem_object *shmem_obj; - struct v3d_dev *v3d =3D to_v3d_dev(dev); struct v3d_bo *bo; int ret; =20 shmem_obj =3D drm_gem_shmem_create_with_mnt(dev, unaligned_size, - v3d->gemfs); + dev->huge_mnt); if (IS_ERR(shmem_obj)) return ERR_CAST(shmem_obj); bo =3D to_v3d_bo(&shmem_obj->base); diff --git a/drivers/gpu/drm/v3d/v3d_drv.c b/drivers/gpu/drm/v3d/v3d_drv.c index c5a3bbbc74c5..19ec0ea7f38e 100644 --- a/drivers/gpu/drm/v3d/v3d_drv.c +++ b/drivers/gpu/drm/v3d/v3d_drv.c @@ -106,7 +106,7 @@ static int v3d_get_param_ioctl(struct drm_device *dev, = void *data, args->value =3D v3d->perfmon_info.max_counters; return 0; case DRM_V3D_PARAM_SUPPORTS_SUPER_PAGES: - args->value =3D !!v3d->gemfs; + args->value =3D !!dev->huge_mnt; return 0; case DRM_V3D_PARAM_GLOBAL_RESET_COUNTER: mutex_lock(&v3d->reset_lock); diff --git a/drivers/gpu/drm/v3d/v3d_drv.h b/drivers/gpu/drm/v3d/v3d_drv.h index 1884686985b8..99a39329bb85 100644 --- a/drivers/gpu/drm/v3d/v3d_drv.h +++ b/drivers/gpu/drm/v3d/v3d_drv.h @@ -158,11 +158,6 @@ struct v3d_dev { struct drm_mm mm; spinlock_t mm_lock; =20 - /* - * tmpfs instance used for shmem backed objects - */ - struct vfsmount *gemfs; - struct work_struct overflow_mem_work; =20 struct v3d_queue_state queue[V3D_MAX_QUEUES]; @@ -569,6 +564,7 @@ extern const struct dma_fence_ops v3d_fence_ops; struct dma_fence *v3d_fence_create(struct v3d_dev *v3d, enum v3d_queue q); =20 /* v3d_gem.c */ +extern bool super_pages; int v3d_gem_init(struct drm_device *dev); void v3d_gem_destroy(struct drm_device *dev); void v3d_reset_sms(struct v3d_dev *v3d); @@ -576,11 +572,6 @@ void v3d_reset(struct v3d_dev *v3d); void v3d_invalidate_caches(struct v3d_dev *v3d); void v3d_clean_caches(struct v3d_dev *v3d); =20 -/* v3d_gemfs.c */ -extern bool super_pages; -void v3d_gemfs_init(struct v3d_dev *v3d); -void v3d_gemfs_fini(struct v3d_dev *v3d); - /* v3d_submit.c */ void v3d_job_cleanup(struct v3d_job *job); void v3d_job_put(struct v3d_job *job); diff --git a/drivers/gpu/drm/v3d/v3d_gem.c b/drivers/gpu/drm/v3d/v3d_gem.c index bb110d35f749..635ff0fabe7e 100644 --- a/drivers/gpu/drm/v3d/v3d_gem.c +++ b/drivers/gpu/drm/v3d/v3d_gem.c @@ -258,6 +258,30 @@ v3d_invalidate_caches(struct v3d_dev *v3d) v3d_invalidate_slices(v3d, 0); } =20 +static void +v3d_huge_mnt_init(struct v3d_dev *v3d) +{ + int err =3D 0; + + /* + * By using a huge shmemfs mountpoint when the user wants to + * enable Super Pages, we can pass in mount flags that better + * match our usecase. + */ + + if (super_pages) + err =3D drm_gem_huge_mnt_create(&v3d->drm, "within_size"); + + if (v3d->drm.huge_mnt) + drm_info(&v3d->drm, "Using Transparent Hugepages\n"); + else if (err) + drm_warn(&v3d->drm, "Can't use Transparent Hugepages (%d)\n", + err); + else + drm_notice(&v3d->drm, + "Transparent Hugepage support is recommended for optimal performance= on this platform!\n"); +} + int v3d_gem_init(struct drm_device *dev) { @@ -309,7 +333,7 @@ v3d_gem_init(struct drm_device *dev) v3d_init_hw_state(v3d); v3d_mmu_set_page_table(v3d); =20 - v3d_gemfs_init(v3d); + v3d_huge_mnt_init(v3d); =20 ret =3D v3d_sched_init(v3d); if (ret) { @@ -329,7 +353,6 @@ v3d_gem_destroy(struct drm_device *dev) enum v3d_queue q; =20 v3d_sched_fini(v3d); - v3d_gemfs_fini(v3d); =20 /* Waiting for jobs to finish would need to be done before * unregistering V3D. diff --git a/drivers/gpu/drm/v3d/v3d_gemfs.c b/drivers/gpu/drm/v3d/v3d_gemf= s.c deleted file mode 100644 index c1a30166c099..000000000000 --- a/drivers/gpu/drm/v3d/v3d_gemfs.c +++ /dev/null @@ -1,60 +0,0 @@ -// SPDX-License-Identifier: GPL-2.0+ -/* Copyright (C) 2024 Raspberry Pi */ - -#include -#include -#include - -#include "v3d_drv.h" - -void v3d_gemfs_init(struct v3d_dev *v3d) -{ - struct file_system_type *type; - struct fs_context *fc; - struct vfsmount *gemfs; - int ret; - - /* - * By creating our own shmemfs mountpoint, we can pass in - * mount flags that better match our usecase. However, we - * only do so on platforms which benefit from it. - */ - if (!IS_ENABLED(CONFIG_TRANSPARENT_HUGEPAGE)) - goto err; - - /* The user doesn't want to enable Super Pages */ - if (!super_pages) - goto err; - - type =3D get_fs_type("tmpfs"); - if (!type) - goto err; - - fc =3D fs_context_for_mount(type, SB_KERNMOUNT); - if (IS_ERR(fc)) - goto err; - ret =3D vfs_parse_fs_string(fc, "source", "tmpfs"); - if (!ret) - ret =3D vfs_parse_fs_string(fc, "huge", "within_size"); - if (!ret) - gemfs =3D fc_mount_longterm(fc); - put_fs_context(fc); - if (ret) - goto err; - - v3d->gemfs =3D gemfs; - drm_info(&v3d->drm, "Using Transparent Hugepages\n"); - - return; - -err: - v3d->gemfs =3D NULL; - drm_notice(&v3d->drm, - "Transparent Hugepage support is recommended for optimal performance = on this platform!\n"); -} - -void v3d_gemfs_fini(struct v3d_dev *v3d) -{ - if (v3d->gemfs) - kern_unmount(v3d->gemfs); -} --=20 2.47.3 From nobody Fri Dec 19 20:55:28 2025 Received: from bali.collaboradmins.com (bali.collaboradmins.com [148.251.105.195]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by smtp.subspace.kernel.org (Postfix) with ESMTPS id 8E50730C37E; Wed, 15 Oct 2025 15:30:46 +0000 (UTC) Authentication-Results: smtp.subspace.kernel.org; arc=none smtp.client-ip=148.251.105.195 ARC-Seal: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1760542248; cv=none; b=WlHjSRLIsU7aJEJbVfffBaRjdjva9ATcGTp4PgIUn2xasYZRltqYICLvPH327K7kLcEWvCbcZ3i/Xuxovu6zAxnokGti4qmQP1jtVS7raprVcdeO9Sr7RaJuDUsXc4W/+Tfy94Bh8nSKr883uhZyPJL25FWLNqSaKfo21teRPm8= ARC-Message-Signature: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1760542248; c=relaxed/simple; bh=UPaCf2ud0ScDDITxAmGIS91OLHTTJaUwZwLkZIUbeUs=; h=From:To:Cc:Subject:Date:Message-ID:In-Reply-To:References: MIME-Version:Content-Type; b=GDlB9R6duMDs92H1DpvkAm3DIAW+eoPyNImKdUMEaeww0ucAITciOPwLCfdMm9f/PSwhEwCcka15HcjVgJhwnpKMXaVO4H3gEx+TOvGCYAD643ejtDh0sxNLn2npEyFwye7TCBqu7TbE28mRIFBN/Q4ZeE2fBMX3x/E1XqI0mbE= ARC-Authentication-Results: i=1; smtp.subspace.kernel.org; dmarc=pass (p=none dis=none) header.from=collabora.com; spf=pass smtp.mailfrom=collabora.com; dkim=pass (2048-bit key) header.d=collabora.com header.i=@collabora.com header.b=jeeeN8Bx; arc=none smtp.client-ip=148.251.105.195 Authentication-Results: smtp.subspace.kernel.org; dmarc=pass (p=none dis=none) header.from=collabora.com Authentication-Results: smtp.subspace.kernel.org; spf=pass smtp.mailfrom=collabora.com Authentication-Results: smtp.subspace.kernel.org; dkim=pass (2048-bit key) header.d=collabora.com header.i=@collabora.com header.b="jeeeN8Bx" DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=collabora.com; s=mail; t=1760542245; bh=UPaCf2ud0ScDDITxAmGIS91OLHTTJaUwZwLkZIUbeUs=; h=From:To:Cc:Subject:Date:In-Reply-To:References:From; b=jeeeN8BxcXaTgnHKIpiYrn+CKNf3vzZAwdLoR0Jbg0cLukwhxxryk1mNoaHnmNpLR tJi2gd9VdiUhAVVPlo52Kli1RsxYYbfe8Jhe+s303YZRmAOsQcgPgQGLB8eqdLvO0N wo037Pz+BKczYBMKpsPgLSgoxRj4rPWV2Q6l2/4Ye6bbBwR+6+dWa2NtojiXOZgQcs wVZpQRvNzPtq2jy0E5uDXIFv2IZ3jiljNtn006VVEx2yz7INFAeIBkzXrfr6Svv8xe ArQ4AZpx1jwe2iPIK6a1wtLoE3BkoMuXQB94OH0CXBi+RCki/6rz83FkeunpUZMgD0 3YtRzPYkV6Qhw== Received: from debian-rockchip-rock5b-rk3588.. (unknown [IPv6:2a01:e0a:5e3:6100:826d:bc07:e98c:84a]) (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: loicmolinari) by bali.collaboradmins.com (Postfix) with ESMTPSA id 28D5317E1340; Wed, 15 Oct 2025 17:30:44 +0200 (CEST) From: =?UTF-8?q?Lo=C3=AFc=20Molinari?= To: Maarten Lankhorst , Maxime Ripard , Thomas Zimmermann , David Airlie , Simona Vetter , Jani Nikula , Joonas Lahtinen , Rodrigo Vivi , Tvrtko Ursulin , Boris Brezillon , Rob Herring , Steven Price , Liviu Dudau , Melissa Wen , =?UTF-8?q?Ma=C3=ADra=20Canal?= , Hugh Dickins , Baolin Wang , Andrew Morton , =?UTF-8?q?Lo=C3=AFc=20Molinari?= , Al Viro , =?UTF-8?q?Miko=C5=82aj=20Wasiak?= , Christian Brauner , Nitin Gote , Andi Shyti , Jonathan Corbet , Christopher Healy , Matthew Wilcox , Bagas Sanjaya Cc: linux-kernel@vger.kernel.org, dri-devel@lists.freedesktop.org, intel-gfx@lists.freedesktop.org, linux-mm@kvack.org, linux-doc@vger.kernel.org, kernel@collabora.com Subject: [PATCH v4 08/13] drm/v3d: Fix builds with CONFIG_TRANSPARENT_HUGEPAGE=n Date: Wed, 15 Oct 2025 17:30:12 +0200 Message-ID: <20251015153018.43735-9-loic.molinari@collabora.com> X-Mailer: git-send-email 2.47.3 In-Reply-To: <20251015153018.43735-1-loic.molinari@collabora.com> References: <20251015153018.43735-1-loic.molinari@collabora.com> Precedence: bulk X-Mailing-List: linux-kernel@vger.kernel.org List-Id: List-Subscribe: List-Unsubscribe: MIME-Version: 1.0 Content-Type: text/plain; charset="utf-8" Content-Transfer-Encoding: quoted-printable Don't declare "super_pages" on builds with CONFIG_TRANSPARENT_HUGEPAGE disabled to prevent build error: ERROR: modpost: "super_pages" [drivers/gpu/drm/v3d/v3d.ko] undefined! Signed-off-by: Lo=C3=AFc Molinari --- drivers/gpu/drm/v3d/v3d_drv.h | 2 ++ drivers/gpu/drm/v3d/v3d_gem.c | 2 ++ 2 files changed, 4 insertions(+) diff --git a/drivers/gpu/drm/v3d/v3d_drv.h b/drivers/gpu/drm/v3d/v3d_drv.h index 99a39329bb85..481502104391 100644 --- a/drivers/gpu/drm/v3d/v3d_drv.h +++ b/drivers/gpu/drm/v3d/v3d_drv.h @@ -564,7 +564,9 @@ extern const struct dma_fence_ops v3d_fence_ops; struct dma_fence *v3d_fence_create(struct v3d_dev *v3d, enum v3d_queue q); =20 /* v3d_gem.c */ +#ifdef CONFIG_TRANSPARENT_HUGEPAGE extern bool super_pages; +#endif int v3d_gem_init(struct drm_device *dev); void v3d_gem_destroy(struct drm_device *dev); void v3d_reset_sms(struct v3d_dev *v3d); diff --git a/drivers/gpu/drm/v3d/v3d_gem.c b/drivers/gpu/drm/v3d/v3d_gem.c index 635ff0fabe7e..0039063eb8b2 100644 --- a/drivers/gpu/drm/v3d/v3d_gem.c +++ b/drivers/gpu/drm/v3d/v3d_gem.c @@ -269,7 +269,9 @@ v3d_huge_mnt_init(struct v3d_dev *v3d) * match our usecase. */ =20 +#ifdef CONFIG_TRANSPARENT_HUGEPAGE if (super_pages) +#endif err =3D drm_gem_huge_mnt_create(&v3d->drm, "within_size"); =20 if (v3d->drm.huge_mnt) --=20 2.47.3 From nobody Fri Dec 19 20:55:28 2025 Received: from bali.collaboradmins.com (bali.collaboradmins.com [148.251.105.195]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by smtp.subspace.kernel.org (Postfix) with ESMTPS id A32B53128DD; Wed, 15 Oct 2025 15:30:48 +0000 (UTC) Authentication-Results: smtp.subspace.kernel.org; arc=none smtp.client-ip=148.251.105.195 ARC-Seal: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1760542250; cv=none; b=CSBqhKUbRRr01PrbBQGmtUz8qLADmc+400Ocb9n5zFUnP69s9seF88E+ganZlAQEYno4D+59jrAaYNAgXYewod0qTKorw/xeczzVnurel5RNurieGEVOSvLXv0UVh5LQRZMby8FYQPNnLMNFcAV9JNMVFrEFAnHWfUMC25glgq0= ARC-Message-Signature: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1760542250; c=relaxed/simple; bh=kTIWND+a8wk3F9Zm2bolmicWjJFPvKRfO75cSh1AbVs=; h=From:To:Cc:Subject:Date:Message-ID:In-Reply-To:References: MIME-Version:Content-Type; b=DoXEJo+RqaT4UQNb2Sj3awBZjJC1HY9oxX8Gbr6AisWFj3DLUP/yevpXLVsKhkXFIzwlSfO1qyq93Ixf0JkvPfxy+5vEsI8slDsdiYvdOVZ8lCYAXR8qWMJ2fqLAK8LapGgPnaq+VetcKPrakrFONkUvxFQqUgdPm+XkCRL8jyo= ARC-Authentication-Results: i=1; smtp.subspace.kernel.org; dmarc=pass (p=none dis=none) header.from=collabora.com; spf=pass smtp.mailfrom=collabora.com; dkim=pass (2048-bit key) header.d=collabora.com header.i=@collabora.com header.b=Jb7T6g+a; arc=none smtp.client-ip=148.251.105.195 Authentication-Results: smtp.subspace.kernel.org; dmarc=pass (p=none dis=none) header.from=collabora.com Authentication-Results: smtp.subspace.kernel.org; spf=pass smtp.mailfrom=collabora.com Authentication-Results: smtp.subspace.kernel.org; dkim=pass (2048-bit key) header.d=collabora.com header.i=@collabora.com header.b="Jb7T6g+a" DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=collabora.com; s=mail; t=1760542247; bh=kTIWND+a8wk3F9Zm2bolmicWjJFPvKRfO75cSh1AbVs=; h=From:To:Cc:Subject:Date:In-Reply-To:References:From; b=Jb7T6g+a7jJR0+/+0RjYJYg1nRDwKC+EqcrbP++h/QraoRIQhetWUz3k2CiFiODJt ZQtdJJISr0C+YHGXq6TOgGq/+y58aH+D3DsSHP12WVxMbvmbJ7NwtxPIaTFIErrsdS d2nMTeMP0iEfFUuSTJN2g8w/pBVlz+M2XgRh7pttovOr2e7/WzIaG3EsUyj+0t6acA xVA/n8So/OPdPHcNoA1XlYT97OH60glRJvTGIKfmvuWgIj/e+XOvPL8/YhS6SZ17IB lwSJkj3+nQD+XAzFO2UiAkTF2DAZF+IzXrodsJ0nQmJWJsHtnlL5gBTd6mhvkMDnYX ih0wlMgAC3Xjw== Received: from debian-rockchip-rock5b-rk3588.. (unknown [IPv6:2a01:e0a:5e3:6100:826d:bc07:e98c:84a]) (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: loicmolinari) by bali.collaboradmins.com (Postfix) with ESMTPSA id 5794217E1305; Wed, 15 Oct 2025 17:30:46 +0200 (CEST) From: =?UTF-8?q?Lo=C3=AFc=20Molinari?= To: Maarten Lankhorst , Maxime Ripard , Thomas Zimmermann , David Airlie , Simona Vetter , Jani Nikula , Joonas Lahtinen , Rodrigo Vivi , Tvrtko Ursulin , Boris Brezillon , Rob Herring , Steven Price , Liviu Dudau , Melissa Wen , =?UTF-8?q?Ma=C3=ADra=20Canal?= , Hugh Dickins , Baolin Wang , Andrew Morton , =?UTF-8?q?Lo=C3=AFc=20Molinari?= , Al Viro , =?UTF-8?q?Miko=C5=82aj=20Wasiak?= , Christian Brauner , Nitin Gote , Andi Shyti , Jonathan Corbet , Christopher Healy , Matthew Wilcox , Bagas Sanjaya Cc: linux-kernel@vger.kernel.org, dri-devel@lists.freedesktop.org, intel-gfx@lists.freedesktop.org, linux-mm@kvack.org, linux-doc@vger.kernel.org, kernel@collabora.com Subject: [PATCH v4 09/13] drm/gem: Get rid of *_with_mnt helpers Date: Wed, 15 Oct 2025 17:30:13 +0200 Message-ID: <20251015153018.43735-10-loic.molinari@collabora.com> X-Mailer: git-send-email 2.47.3 In-Reply-To: <20251015153018.43735-1-loic.molinari@collabora.com> References: <20251015153018.43735-1-loic.molinari@collabora.com> Precedence: bulk X-Mailing-List: linux-kernel@vger.kernel.org List-Id: List-Subscribe: List-Unsubscribe: MIME-Version: 1.0 Content-Type: text/plain; charset="utf-8" Content-Transfer-Encoding: quoted-printable drm_gem_object_init_with_mnt() and drm_gem_shmem_create_with_mnt() can be removed now that the drivers use drm_gem_huge_mnt_create(). Signed-off-by: Lo=C3=AFc Molinari Reviewed-by: Boris Brezillon --- drivers/gpu/drm/drm_gem.c | 34 ++++++----------------- drivers/gpu/drm/drm_gem_shmem_helper.c | 38 ++++++-------------------- drivers/gpu/drm/v3d/v3d_bo.c | 3 +- include/drm/drm_gem.h | 3 -- include/drm/drm_gem_shmem_helper.h | 3 -- 5 files changed, 17 insertions(+), 64 deletions(-) diff --git a/drivers/gpu/drm/drm_gem.c b/drivers/gpu/drm/drm_gem.c index db8c0a217add..d8fa789cf330 100644 --- a/drivers/gpu/drm/drm_gem.c +++ b/drivers/gpu/drm/drm_gem.c @@ -172,28 +172,26 @@ drm_gem_init(struct drm_device *dev) } =20 /** - * drm_gem_object_init_with_mnt - initialize an allocated shmem-backed GEM - * object in a given shmfs mountpoint + * drm_gem_object_init - initialize an allocated shmem-backed GEM object * * @dev: drm_device the object should be initialized for * @obj: drm_gem_object to initialize * @size: object size - * @gemfs: tmpfs mount where the GEM object will be created. If NULL, use - * the usual tmpfs mountpoint (`shm_mnt`). * * Initialize an already allocated GEM object of the specified size with - * shmfs backing store. + * shmfs backing store. A huge mountpoint can be used by calling + * drm_gem_huge_mnt_create() beforehand. */ -int drm_gem_object_init_with_mnt(struct drm_device *dev, - struct drm_gem_object *obj, size_t size, - struct vfsmount *gemfs) +int drm_gem_object_init(struct drm_device *dev, struct drm_gem_object *obj, + size_t size) { struct file *filp; =20 drm_gem_private_object_init(dev, obj, size); =20 - if (gemfs) - filp =3D shmem_file_setup_with_mnt(gemfs, "drm mm object", size, + if (dev->huge_mnt) + filp =3D shmem_file_setup_with_mnt(dev->huge_mnt, + "drm mm object", size, VM_NORESERVE); else filp =3D shmem_file_setup("drm mm object", size, VM_NORESERVE); @@ -205,22 +203,6 @@ int drm_gem_object_init_with_mnt(struct drm_device *de= v, =20 return 0; } -EXPORT_SYMBOL(drm_gem_object_init_with_mnt); - -/** - * drm_gem_object_init - initialize an allocated shmem-backed GEM object - * @dev: drm_device the object should be initialized for - * @obj: drm_gem_object to initialize - * @size: object size - * - * Initialize an already allocated GEM object of the specified size with - * shmfs backing store. - */ -int drm_gem_object_init(struct drm_device *dev, struct drm_gem_object *obj, - size_t size) -{ - return drm_gem_object_init_with_mnt(dev, obj, size, NULL); -} EXPORT_SYMBOL(drm_gem_object_init); =20 /** diff --git a/drivers/gpu/drm/drm_gem_shmem_helper.c b/drivers/gpu/drm/drm_g= em_shmem_helper.c index 5d11bea573fe..33f2ecc5f31f 100644 --- a/drivers/gpu/drm/drm_gem_shmem_helper.c +++ b/drivers/gpu/drm/drm_gem_shmem_helper.c @@ -50,7 +50,7 @@ static const struct drm_gem_object_funcs drm_gem_shmem_fu= ncs =3D { }; =20 static int __drm_gem_shmem_init(struct drm_device *dev, struct drm_gem_shm= em_object *shmem, - size_t size, bool private, struct vfsmount *gemfs) + size_t size, bool private) { struct drm_gem_object *obj =3D &shmem->base; int ret =3D 0; @@ -62,7 +62,7 @@ static int __drm_gem_shmem_init(struct drm_device *dev, s= truct drm_gem_shmem_obj drm_gem_private_object_init(dev, obj, size); shmem->map_wc =3D false; /* dma-buf mappings use always writecombine */ } else { - ret =3D drm_gem_object_init_with_mnt(dev, obj, size, gemfs); + ret =3D drm_gem_object_init(dev, obj, size); } if (ret) { drm_gem_private_object_fini(obj); @@ -103,13 +103,12 @@ static int __drm_gem_shmem_init(struct drm_device *de= v, struct drm_gem_shmem_obj */ int drm_gem_shmem_init(struct drm_device *dev, struct drm_gem_shmem_object= *shmem, size_t size) { - return __drm_gem_shmem_init(dev, shmem, size, false, NULL); + return __drm_gem_shmem_init(dev, shmem, size, false); } EXPORT_SYMBOL_GPL(drm_gem_shmem_init); =20 static struct drm_gem_shmem_object * -__drm_gem_shmem_create(struct drm_device *dev, size_t size, bool private, - struct vfsmount *gemfs) +__drm_gem_shmem_create(struct drm_device *dev, size_t size, bool private) { struct drm_gem_shmem_object *shmem; struct drm_gem_object *obj; @@ -129,7 +128,7 @@ __drm_gem_shmem_create(struct drm_device *dev, size_t s= ize, bool private, obj =3D &shmem->base; } =20 - ret =3D __drm_gem_shmem_init(dev, shmem, size, private, gemfs); + ret =3D __drm_gem_shmem_init(dev, shmem, size, private); if (ret) { kfree(obj); return ERR_PTR(ret); @@ -150,31 +149,10 @@ __drm_gem_shmem_create(struct drm_device *dev, size_t= size, bool private, */ struct drm_gem_shmem_object *drm_gem_shmem_create(struct drm_device *dev, = size_t size) { - return __drm_gem_shmem_create(dev, size, false, NULL); + return __drm_gem_shmem_create(dev, size, false); } EXPORT_SYMBOL_GPL(drm_gem_shmem_create); =20 -/** - * drm_gem_shmem_create_with_mnt - Allocate an object with the given size = in a - * given mountpoint - * @dev: DRM device - * @size: Size of the object to allocate - * @gemfs: tmpfs mount where the GEM object will be created - * - * This function creates a shmem GEM object in a given tmpfs mountpoint. - * - * Returns: - * A struct drm_gem_shmem_object * on success or an ERR_PTR()-encoded nega= tive - * error code on failure. - */ -struct drm_gem_shmem_object *drm_gem_shmem_create_with_mnt(struct drm_devi= ce *dev, - size_t size, - struct vfsmount *gemfs) -{ - return __drm_gem_shmem_create(dev, size, false, gemfs); -} -EXPORT_SYMBOL_GPL(drm_gem_shmem_create_with_mnt); - /** * drm_gem_shmem_release - Release resources associated with a shmem GEM o= bject. * @shmem: shmem GEM object @@ -902,7 +880,7 @@ drm_gem_shmem_prime_import_sg_table(struct drm_device *= dev, size_t size =3D PAGE_ALIGN(attach->dmabuf->size); struct drm_gem_shmem_object *shmem; =20 - shmem =3D __drm_gem_shmem_create(dev, size, true, NULL); + shmem =3D __drm_gem_shmem_create(dev, size, true); if (IS_ERR(shmem)) return ERR_CAST(shmem); =20 @@ -950,7 +928,7 @@ struct drm_gem_object *drm_gem_shmem_prime_import_no_ma= p(struct drm_device *dev, =20 size =3D PAGE_ALIGN(attach->dmabuf->size); =20 - shmem =3D __drm_gem_shmem_create(dev, size, true, NULL); + shmem =3D __drm_gem_shmem_create(dev, size, true); if (IS_ERR(shmem)) { ret =3D PTR_ERR(shmem); goto fail_detach; diff --git a/drivers/gpu/drm/v3d/v3d_bo.c b/drivers/gpu/drm/v3d/v3d_bo.c index 6b9909bfce82..ef9e9c15da3c 100644 --- a/drivers/gpu/drm/v3d/v3d_bo.c +++ b/drivers/gpu/drm/v3d/v3d_bo.c @@ -151,8 +151,7 @@ struct v3d_bo *v3d_bo_create(struct drm_device *dev, st= ruct drm_file *file_priv, struct v3d_bo *bo; int ret; =20 - shmem_obj =3D drm_gem_shmem_create_with_mnt(dev, unaligned_size, - dev->huge_mnt); + shmem_obj =3D drm_gem_shmem_create(dev, unaligned_size); if (IS_ERR(shmem_obj)) return ERR_CAST(shmem_obj); bo =3D to_v3d_bo(&shmem_obj->base); diff --git a/include/drm/drm_gem.h b/include/drm/drm_gem.h index 7285a62d9afc..42d9066b1785 100644 --- a/include/drm/drm_gem.h +++ b/include/drm/drm_gem.h @@ -497,9 +497,6 @@ void drm_gem_object_release(struct drm_gem_object *obj); void drm_gem_object_free(struct kref *kref); int drm_gem_object_init(struct drm_device *dev, struct drm_gem_object *obj, size_t size); -int drm_gem_object_init_with_mnt(struct drm_device *dev, - struct drm_gem_object *obj, size_t size, - struct vfsmount *gemfs); void drm_gem_private_object_init(struct drm_device *dev, struct drm_gem_object *obj, size_t size); void drm_gem_private_object_fini(struct drm_gem_object *obj); diff --git a/include/drm/drm_gem_shmem_helper.h b/include/drm/drm_gem_shmem= _helper.h index 589f7bfe7506..6b6478f5ca24 100644 --- a/include/drm/drm_gem_shmem_helper.h +++ b/include/drm/drm_gem_shmem_helper.h @@ -109,9 +109,6 @@ struct drm_gem_shmem_object { =20 int drm_gem_shmem_init(struct drm_device *dev, struct drm_gem_shmem_object= *shmem, size_t size); struct drm_gem_shmem_object *drm_gem_shmem_create(struct drm_device *dev, = size_t size); -struct drm_gem_shmem_object *drm_gem_shmem_create_with_mnt(struct drm_devi= ce *dev, - size_t size, - struct vfsmount *gemfs); void drm_gem_shmem_release(struct drm_gem_shmem_object *shmem); void drm_gem_shmem_free(struct drm_gem_shmem_object *shmem); =20 --=20 2.47.3 From nobody Fri Dec 19 20:55:28 2025 Received: from bali.collaboradmins.com (bali.collaboradmins.com [148.251.105.195]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by smtp.subspace.kernel.org (Postfix) with ESMTPS id C2C9A31B83D; Wed, 15 Oct 2025 15:30:50 +0000 (UTC) Authentication-Results: smtp.subspace.kernel.org; arc=none smtp.client-ip=148.251.105.195 ARC-Seal: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1760542252; cv=none; b=k6i73GoBUvqAI6t/e+r/+ndmONRI2bJC6tV7Oh6/PkYpaP4TRd7I11tunRlBwd4JZLfOiq9JzVzaGc4M62jGpVb7PnnfPHBtvegQYrjDRF9dugRKYOexltHJzGov8fmdTlt9qED1/qv5uXtZRY9wXRXPGZpIpyJa6/VTrj8HWUs= ARC-Message-Signature: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1760542252; c=relaxed/simple; bh=+cl/GzgcT1Fb2lvuYJSy48Zz4ImIu8RmCkKrQxlGrLk=; h=From:To:Cc:Subject:Date:Message-ID:In-Reply-To:References: MIME-Version:Content-Type; b=PzSls0G7F2o6d4nhysSM3KI0G0uCi8CstpGH5z2PlmEuakL3Xb8C9FuDHuARd8Ucsp4yw3QSaPG4dfXWV+jFrdmvHtNqW86xuVWOxkiucTQ0b+q/8oyKiBpogJNx0eT6isIVAw/JNWuwNDv9uC/8G9Rd1uYcj/wssBXofpUStRI= ARC-Authentication-Results: i=1; smtp.subspace.kernel.org; dmarc=pass (p=none dis=none) header.from=collabora.com; spf=pass smtp.mailfrom=collabora.com; dkim=pass (2048-bit key) header.d=collabora.com header.i=@collabora.com header.b=Sko1fOkK; arc=none smtp.client-ip=148.251.105.195 Authentication-Results: smtp.subspace.kernel.org; dmarc=pass (p=none dis=none) header.from=collabora.com Authentication-Results: smtp.subspace.kernel.org; spf=pass smtp.mailfrom=collabora.com Authentication-Results: smtp.subspace.kernel.org; dkim=pass (2048-bit key) header.d=collabora.com header.i=@collabora.com header.b="Sko1fOkK" DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=collabora.com; s=mail; t=1760542249; bh=+cl/GzgcT1Fb2lvuYJSy48Zz4ImIu8RmCkKrQxlGrLk=; h=From:To:Cc:Subject:Date:In-Reply-To:References:From; b=Sko1fOkKwhU5QJtjbTLNt357vjSNq/kUG6zxztjtADWLI2V+bQ5AvA0/dSCH2xTko X3S86Z2JAt2MIs2p36/0a9E7PVnXYqs6fb9MsE7FWooHUJs0M/N3NfuWGFvY8/hWpS KTbcapTiE+EutYTrsx5nq+RIPWbBmwHrY8TZvyIvOl81FyMqp+9CctxsiXRCSu3vcY yGSQDHCJsEAGzQP6sYRtQuY8opTNhbznPw6+zboNbvl2c7RXtn3nL8wBGmdloIS0la PP9NnWVDzKP3ME/CaXzDE/TodgUfp5v5L5dn0Fs3a5uQvgSdhRIG9NZUvmPywNEW1b 02Yua7S1vc4YQ== Received: from debian-rockchip-rock5b-rk3588.. (unknown [IPv6:2a01:e0a:5e3:6100:826d:bc07:e98c:84a]) (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: loicmolinari) by bali.collaboradmins.com (Postfix) with ESMTPSA id 6A18C17E00A3; Wed, 15 Oct 2025 17:30:48 +0200 (CEST) From: =?UTF-8?q?Lo=C3=AFc=20Molinari?= To: Maarten Lankhorst , Maxime Ripard , Thomas Zimmermann , David Airlie , Simona Vetter , Jani Nikula , Joonas Lahtinen , Rodrigo Vivi , Tvrtko Ursulin , Boris Brezillon , Rob Herring , Steven Price , Liviu Dudau , Melissa Wen , =?UTF-8?q?Ma=C3=ADra=20Canal?= , Hugh Dickins , Baolin Wang , Andrew Morton , =?UTF-8?q?Lo=C3=AFc=20Molinari?= , Al Viro , =?UTF-8?q?Miko=C5=82aj=20Wasiak?= , Christian Brauner , Nitin Gote , Andi Shyti , Jonathan Corbet , Christopher Healy , Matthew Wilcox , Bagas Sanjaya Cc: linux-kernel@vger.kernel.org, dri-devel@lists.freedesktop.org, intel-gfx@lists.freedesktop.org, linux-mm@kvack.org, linux-doc@vger.kernel.org, kernel@collabora.com Subject: [PATCH v4 10/13] drm/panthor: Introduce huge tmpfs mount point option Date: Wed, 15 Oct 2025 17:30:14 +0200 Message-ID: <20251015153018.43735-11-loic.molinari@collabora.com> X-Mailer: git-send-email 2.47.3 In-Reply-To: <20251015153018.43735-1-loic.molinari@collabora.com> References: <20251015153018.43735-1-loic.molinari@collabora.com> Precedence: bulk X-Mailing-List: linux-kernel@vger.kernel.org List-Id: List-Subscribe: List-Unsubscribe: MIME-Version: 1.0 Content-Type: text/plain; charset="utf-8" Content-Transfer-Encoding: quoted-printable Introduce the 'panthor.transparent_hugepage' boolean module parameter (false by default). When the parameter is set to true, a new tmpfs mount point is created and mounted using the 'huge=3Dwithin_size' option. It's then used at GEM object creation instead of the default 'shm_mnt' mount point in order to enable Transparent Hugepage (THP) for the object (without having to rely on a system wide parameter). v3: - use huge tmpfs mountpoint in drm_device v4: - fix builds with CONFIG_TRANSPARENT_HUGEPAGE=3Dn - clean up mountpoint creation error handling - print negative error value Signed-off-by: Lo=C3=AFc Molinari Reviewed-by: Boris Brezillon --- drivers/gpu/drm/panthor/panthor_device.c | 3 +++ drivers/gpu/drm/panthor/panthor_drv.c | 7 +++++++ drivers/gpu/drm/panthor/panthor_drv.h | 11 +++++++++++ drivers/gpu/drm/panthor/panthor_gem.c | 19 +++++++++++++++++++ drivers/gpu/drm/panthor/panthor_gem.h | 2 ++ 5 files changed, 42 insertions(+) create mode 100644 drivers/gpu/drm/panthor/panthor_drv.h diff --git a/drivers/gpu/drm/panthor/panthor_device.c b/drivers/gpu/drm/pan= thor/panthor_device.c index c7033d82cef5..4ced0bef22e8 100644 --- a/drivers/gpu/drm/panthor/panthor_device.c +++ b/drivers/gpu/drm/panthor/panthor_device.c @@ -17,6 +17,7 @@ #include "panthor_devfreq.h" #include "panthor_device.h" #include "panthor_fw.h" +#include "panthor_gem.h" #include "panthor_gpu.h" #include "panthor_hw.h" #include "panthor_mmu.h" @@ -271,6 +272,8 @@ int panthor_device_init(struct panthor_device *ptdev) if (ret) goto err_unplug_fw; =20 + panthor_gem_init(ptdev); + /* ~3 frames */ pm_runtime_set_autosuspend_delay(ptdev->base.dev, 50); pm_runtime_use_autosuspend(ptdev->base.dev); diff --git a/drivers/gpu/drm/panthor/panthor_drv.c b/drivers/gpu/drm/pantho= r/panthor_drv.c index fb4b293f17f0..9e49af1d0959 100644 --- a/drivers/gpu/drm/panthor/panthor_drv.c +++ b/drivers/gpu/drm/panthor/panthor_drv.c @@ -1556,6 +1556,7 @@ static const struct file_operations panthor_drm_drive= r_fops =3D { .read =3D drm_read, .llseek =3D noop_llseek, .mmap =3D panthor_mmap, + .get_unmapped_area =3D drm_gem_get_unmapped_area, .show_fdinfo =3D drm_show_fdinfo, .fop_flags =3D FOP_UNSIGNED_OFFSET, }; @@ -1623,6 +1624,12 @@ static const struct drm_driver panthor_drm_driver = =3D { #endif }; =20 +#ifdef CONFIG_TRANSPARENT_HUGEPAGE +bool panthor_transparent_hugepage; +module_param_named(transparent_hugepage, panthor_transparent_hugepage, boo= l, 0400); +MODULE_PARM_DESC(transparent_hugepage, "Use a dedicated tmpfs mount point = with Transparent Hugepage enabled (false =3D default)"); +#endif + static int panthor_probe(struct platform_device *pdev) { struct panthor_device *ptdev; diff --git a/drivers/gpu/drm/panthor/panthor_drv.h b/drivers/gpu/drm/pantho= r/panthor_drv.h new file mode 100644 index 000000000000..27fe9b6f77bd --- /dev/null +++ b/drivers/gpu/drm/panthor/panthor_drv.h @@ -0,0 +1,11 @@ +// SPDX-License-Identifier: GPL-2.0 or MIT +/* Copyright 2025 Amazon.com, Inc. or its affiliates */ + +#ifndef __PANTHOR_DRV_H__ +#define __PANTHOR_DRV_H__ + +#ifdef CONFIG_TRANSPARENT_HUGEPAGE +extern bool panthor_transparent_hugepage; +#endif + +#endif diff --git a/drivers/gpu/drm/panthor/panthor_gem.c b/drivers/gpu/drm/pantho= r/panthor_gem.c index 156c7a0b62a2..1a176e7e35ea 100644 --- a/drivers/gpu/drm/panthor/panthor_gem.c +++ b/drivers/gpu/drm/panthor/panthor_gem.c @@ -1,6 +1,7 @@ // SPDX-License-Identifier: GPL-2.0 or MIT /* Copyright 2019 Linaro, Ltd, Rob Herring */ /* Copyright 2023 Collabora ltd. */ +/* Copyright 2025 Amazon.com, Inc. or its affiliates */ =20 #include #include @@ -11,10 +12,28 @@ #include =20 #include "panthor_device.h" +#include "panthor_drv.h" #include "panthor_fw.h" #include "panthor_gem.h" #include "panthor_mmu.h" =20 +void panthor_gem_init(struct panthor_device *ptdev) +{ +#ifdef CONFIG_TRANSPARENT_HUGEPAGE + int err; + + if (!panthor_transparent_hugepage) + return; + + err =3D drm_gem_huge_mnt_create(&ptdev->base, "within_size"); + if (ptdev->base.huge_mnt) + drm_info(&ptdev->base, "Using Transparent Hugepage\n"); + else if (err) + drm_warn(&ptdev->base, "Can't use Transparent Hugepage (%d)\n", + err); +#endif +} + #ifdef CONFIG_DEBUG_FS static void panthor_gem_debugfs_bo_init(struct panthor_gem_object *bo) { diff --git a/drivers/gpu/drm/panthor/panthor_gem.h b/drivers/gpu/drm/pantho= r/panthor_gem.h index 80c6e24112d0..2eefe9104e5e 100644 --- a/drivers/gpu/drm/panthor/panthor_gem.h +++ b/drivers/gpu/drm/panthor/panthor_gem.h @@ -136,6 +136,8 @@ struct panthor_gem_object *to_panthor_bo(struct drm_gem= _object *obj) return container_of(to_drm_gem_shmem_obj(obj), struct panthor_gem_object,= base); } =20 +void panthor_gem_init(struct panthor_device *ptdev); + struct drm_gem_object *panthor_gem_create_object(struct drm_device *ddev, = size_t size); =20 int --=20 2.47.3 From nobody Fri Dec 19 20:55:28 2025 Received: from bali.collaboradmins.com (bali.collaboradmins.com [148.251.105.195]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by smtp.subspace.kernel.org (Postfix) with ESMTPS id 4BFEA2FE572; Wed, 15 Oct 2025 15:30:53 +0000 (UTC) Authentication-Results: smtp.subspace.kernel.org; arc=none smtp.client-ip=148.251.105.195 ARC-Seal: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1760542255; cv=none; b=ny/KkAohZmWitJiZjPLDRtjSbbOmpfA3v96vkH2T2n9Ptzy2xH1hQeB6JGW5eIw48ajVzb7qodurdiiX0Gelv+kxM8qsbGDEx0HktCVRcyxOnh7KcfU4F2HP/LAioxjfEbu0u33+u+VEtcxg4D1AYpjSk1qlevKpM35pBr+zEeI= ARC-Message-Signature: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1760542255; c=relaxed/simple; bh=ROKoonOOaPmiRlBVqml71l+74uvjDMGSuEeXXfy4kSs=; h=From:To:Cc:Subject:Date:Message-ID:In-Reply-To:References: MIME-Version:Content-Type; b=f/8pxGIitsXm/T/djZrOEwt4pTI+an941E3gRwYQy5XW/MwA+hpfXQRLvIIF8eYaAcnpAxOMEDIsuXJYuhHln7+270flE4QmfLe7KORG3COXfBhXBxrrKdoHHhb0DjkR5w+s8bbxx4O9dwaQQyvdtIhWHIb5zNiNpkn4hiLtgm4= ARC-Authentication-Results: i=1; smtp.subspace.kernel.org; dmarc=pass (p=none dis=none) header.from=collabora.com; spf=pass smtp.mailfrom=collabora.com; dkim=pass (2048-bit key) header.d=collabora.com header.i=@collabora.com header.b=oWauNRo5; arc=none smtp.client-ip=148.251.105.195 Authentication-Results: smtp.subspace.kernel.org; dmarc=pass (p=none dis=none) header.from=collabora.com Authentication-Results: smtp.subspace.kernel.org; spf=pass smtp.mailfrom=collabora.com Authentication-Results: smtp.subspace.kernel.org; dkim=pass (2048-bit key) header.d=collabora.com header.i=@collabora.com header.b="oWauNRo5" DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=collabora.com; s=mail; t=1760542251; bh=ROKoonOOaPmiRlBVqml71l+74uvjDMGSuEeXXfy4kSs=; h=From:To:Cc:Subject:Date:In-Reply-To:References:From; b=oWauNRo5O8AxCrbqN3wpQuZhqtYBS/whks9Dwc5EtVArhQT5qpEhVrSlH6wBw4uEU iKaDieQ7m7kvx0TbZ7pJDXX5+Ms8gyoCJ+mrOS6zO98Xy0LPr3eY6ZoVwRRpj/qt51 Q1Ck9KR4XbG7CyXHbj+L4+vIgms72znNIyV2+qxI9nH+F11CvHJMqvLGbCsnusFBO/ V8TY6t5nVgAuKW1sos2k9HuBavmxb6aGEV/t83si+7Q3ZeUzuGjLEeMh/R8QU44LrH cr2BxHFzhREQhUiJEFzo2Uq+ORjUrKy/qJMkAdxNVlFpILq3pSpcjE57bYZ8Ik0lO4 TejE+TXE6gekg== Received: from debian-rockchip-rock5b-rk3588.. (unknown [IPv6:2a01:e0a:5e3:6100:826d:bc07:e98c:84a]) (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: loicmolinari) by bali.collaboradmins.com (Postfix) with ESMTPSA id CBCEC17E1340; Wed, 15 Oct 2025 17:30:50 +0200 (CEST) From: =?UTF-8?q?Lo=C3=AFc=20Molinari?= To: Maarten Lankhorst , Maxime Ripard , Thomas Zimmermann , David Airlie , Simona Vetter , Jani Nikula , Joonas Lahtinen , Rodrigo Vivi , Tvrtko Ursulin , Boris Brezillon , Rob Herring , Steven Price , Liviu Dudau , Melissa Wen , =?UTF-8?q?Ma=C3=ADra=20Canal?= , Hugh Dickins , Baolin Wang , Andrew Morton , =?UTF-8?q?Lo=C3=AFc=20Molinari?= , Al Viro , =?UTF-8?q?Miko=C5=82aj=20Wasiak?= , Christian Brauner , Nitin Gote , Andi Shyti , Jonathan Corbet , Christopher Healy , Matthew Wilcox , Bagas Sanjaya Cc: linux-kernel@vger.kernel.org, dri-devel@lists.freedesktop.org, intel-gfx@lists.freedesktop.org, linux-mm@kvack.org, linux-doc@vger.kernel.org, kernel@collabora.com Subject: [PATCH v4 11/13] drm/panthor: Improve IOMMU map/unmap debugging logs Date: Wed, 15 Oct 2025 17:30:15 +0200 Message-ID: <20251015153018.43735-12-loic.molinari@collabora.com> X-Mailer: git-send-email 2.47.3 In-Reply-To: <20251015153018.43735-1-loic.molinari@collabora.com> References: <20251015153018.43735-1-loic.molinari@collabora.com> Precedence: bulk X-Mailing-List: linux-kernel@vger.kernel.org List-Id: List-Subscribe: List-Unsubscribe: MIME-Version: 1.0 Content-Type: text/plain; charset="utf-8" Content-Transfer-Encoding: quoted-printable Log the number of pages and their sizes actually mapped/unmapped by the IOMMU page table driver. Since a map/unmap op is often split in several ops depending on the underlying scatter/gather table, add the start address and the total size to the debugging logs in order to help understand which batch an op is part of. Signed-off-by: Lo=C3=AFc Molinari Reviewed-by: Boris Brezillon --- drivers/gpu/drm/panthor/panthor_mmu.c | 19 ++++++++++++++----- 1 file changed, 14 insertions(+), 5 deletions(-) diff --git a/drivers/gpu/drm/panthor/panthor_mmu.c b/drivers/gpu/drm/pantho= r/panthor_mmu.c index 6dec4354e378..aefbd83d1a75 100644 --- a/drivers/gpu/drm/panthor/panthor_mmu.c +++ b/drivers/gpu/drm/panthor/panthor_mmu.c @@ -917,10 +917,9 @@ static int panthor_vm_unmap_pages(struct panthor_vm *v= m, u64 iova, u64 size) { struct panthor_device *ptdev =3D vm->ptdev; struct io_pgtable_ops *ops =3D vm->pgtbl_ops; + u64 start_iova =3D iova; u64 offset =3D 0; =20 - drm_dbg(&ptdev->base, "unmap: as=3D%d, iova=3D%llx, len=3D%llx", vm->as.i= d, iova, size); - while (offset < size) { size_t unmapped_sz =3D 0, pgcount; size_t pgsize =3D get_pgsize(iova + offset, size - offset, &pgcount); @@ -935,6 +934,12 @@ static int panthor_vm_unmap_pages(struct panthor_vm *v= m, u64 iova, u64 size) panthor_vm_flush_range(vm, iova, offset + unmapped_sz); return -EINVAL; } + + drm_dbg(&ptdev->base, + "unmap: as=3D%d, iova=3D0x%llx, sz=3D%llu, va=3D0x%llx, pgcnt=3D%zu, pg= sz=3D%zu", + vm->as.id, start_iova, size, iova + offset, + unmapped_sz / pgsize, pgsize); + offset +=3D unmapped_sz; } =20 @@ -950,6 +955,7 @@ panthor_vm_map_pages(struct panthor_vm *vm, u64 iova, i= nt prot, struct scatterlist *sgl; struct io_pgtable_ops *ops =3D vm->pgtbl_ops; u64 start_iova =3D iova; + u64 start_size =3D size; int ret; =20 if (!size) @@ -969,15 +975,18 @@ panthor_vm_map_pages(struct panthor_vm *vm, u64 iova,= int prot, len =3D min_t(size_t, len, size); size -=3D len; =20 - drm_dbg(&ptdev->base, "map: as=3D%d, iova=3D%llx, paddr=3D%pad, len=3D%z= x", - vm->as.id, iova, &paddr, len); - while (len) { size_t pgcount, mapped =3D 0; size_t pgsize =3D get_pgsize(iova | paddr, len, &pgcount); =20 ret =3D ops->map_pages(ops, iova, paddr, pgsize, pgcount, prot, GFP_KERNEL, &mapped); + + drm_dbg(&ptdev->base, + "map: as=3D%d, iova=3D0x%llx, sz=3D%llu, va=3D0x%llx, pa=3D%pad, pgcnt= =3D%zu, pgsz=3D%zu", + vm->as.id, start_iova, start_size, iova, &paddr, + mapped / pgsize, pgsize); + iova +=3D mapped; paddr +=3D mapped; len -=3D mapped; --=20 2.47.3 From nobody Fri Dec 19 20:55:28 2025 Received: from bali.collaboradmins.com (bali.collaboradmins.com [148.251.105.195]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by smtp.subspace.kernel.org (Postfix) with ESMTPS id 561E9324B11; Wed, 15 Oct 2025 15:30:55 +0000 (UTC) Authentication-Results: smtp.subspace.kernel.org; arc=none smtp.client-ip=148.251.105.195 ARC-Seal: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1760542257; cv=none; b=gaVmttjI9vCu2w5fl8mnJ1F0fukr00Bk2lDyP4VCJ0YiWxFoEmEUh8cHtBHfh3tmt4ft6iCYUycaXzQ3IiXTTTF2wRCNPDaVgLSGqQHuvKNsu/y6hg2TNsDxFRao5LkaEFyKzSSlTfanwV1nvu0xXn6BQ/BCjulKjA32DVF/cPA= ARC-Message-Signature: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1760542257; c=relaxed/simple; bh=0MFrZ+8U4pjX98vlkZE+Xc8sfj14rbqqp40IJzqK240=; h=From:To:Cc:Subject:Date:Message-ID:In-Reply-To:References: MIME-Version:Content-Type; b=WZLWjjzF8RmHHAVR2vGjjf8AveRg6SQkppNCluLekoXl0ZNA4L3LPwH99795nMh0C6rt7AoR0US5DO9AtPCluYc2cXeSZbJXC9BqaYHlwMmO0G3PfJVNp0fk05QybhCB9bkm1UJwoc4KImqt6dMIyl4J8CQrLdu0i/VSSVN+8vc= ARC-Authentication-Results: i=1; smtp.subspace.kernel.org; dmarc=pass (p=none dis=none) header.from=collabora.com; spf=pass smtp.mailfrom=collabora.com; dkim=pass (2048-bit key) header.d=collabora.com header.i=@collabora.com header.b=lYblDiWR; arc=none smtp.client-ip=148.251.105.195 Authentication-Results: smtp.subspace.kernel.org; dmarc=pass (p=none dis=none) header.from=collabora.com Authentication-Results: smtp.subspace.kernel.org; spf=pass smtp.mailfrom=collabora.com Authentication-Results: smtp.subspace.kernel.org; dkim=pass (2048-bit key) header.d=collabora.com header.i=@collabora.com header.b="lYblDiWR" DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=collabora.com; s=mail; t=1760542253; bh=0MFrZ+8U4pjX98vlkZE+Xc8sfj14rbqqp40IJzqK240=; h=From:To:Cc:Subject:Date:In-Reply-To:References:From; b=lYblDiWRP2ZiZoD1N4aTSKmGyPD8NqQsvlDE6s5Xv71Fc107aPTTmCxNq3C3EXCQI T7axrrnR44h4ihvOHtcBWq8KQyLh9qFiqlET7mTrGOLfLDQfX2EvN/omg4Hj0+JC+F 5UF7nNzyszK+HGUMrAUNOYXf3wYrRTqMkpjYOhTvFWCrNaYawcy/W1EGQqmgLbNmVN dTeCt6KkQYXmQ8/RMkOLHCOm6SCQJhRZmTkxAmqwW9VKcN8DrwiYR+zdIpAdAR4kyC cmDvAkm4f7Nx1qjXNMtadMvzp9ZmwR6fYPnRkZ6LJQIncpMt82ldwTMIqA2Ipym35V u+t1CWvPyMP2g== Received: from debian-rockchip-rock5b-rk3588.. (unknown [IPv6:2a01:e0a:5e3:6100:826d:bc07:e98c:84a]) (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: loicmolinari) by bali.collaboradmins.com (Postfix) with ESMTPSA id 0F4C317E055D; Wed, 15 Oct 2025 17:30:53 +0200 (CEST) From: =?UTF-8?q?Lo=C3=AFc=20Molinari?= To: Maarten Lankhorst , Maxime Ripard , Thomas Zimmermann , David Airlie , Simona Vetter , Jani Nikula , Joonas Lahtinen , Rodrigo Vivi , Tvrtko Ursulin , Boris Brezillon , Rob Herring , Steven Price , Liviu Dudau , Melissa Wen , =?UTF-8?q?Ma=C3=ADra=20Canal?= , Hugh Dickins , Baolin Wang , Andrew Morton , =?UTF-8?q?Lo=C3=AFc=20Molinari?= , Al Viro , =?UTF-8?q?Miko=C5=82aj=20Wasiak?= , Christian Brauner , Nitin Gote , Andi Shyti , Jonathan Corbet , Christopher Healy , Matthew Wilcox , Bagas Sanjaya Cc: linux-kernel@vger.kernel.org, dri-devel@lists.freedesktop.org, intel-gfx@lists.freedesktop.org, linux-mm@kvack.org, linux-doc@vger.kernel.org, kernel@collabora.com Subject: [PATCH v4 12/13] drm/panfrost: Introduce huge tmpfs mount point option Date: Wed, 15 Oct 2025 17:30:16 +0200 Message-ID: <20251015153018.43735-13-loic.molinari@collabora.com> X-Mailer: git-send-email 2.47.3 In-Reply-To: <20251015153018.43735-1-loic.molinari@collabora.com> References: <20251015153018.43735-1-loic.molinari@collabora.com> Precedence: bulk X-Mailing-List: linux-kernel@vger.kernel.org List-Id: List-Subscribe: List-Unsubscribe: MIME-Version: 1.0 Content-Type: text/plain; charset="utf-8" Content-Transfer-Encoding: quoted-printable Introduce the 'panfrost.transparent_hugepage' boolean module parameter (false by default). When the parameter is set to true, a new tmpfs mount point is created and mounted using the 'huge=3Dwithin_size' option. It's then used at GEM object creation instead of the default 'shm_mnt' mount point in order to enable Transparent Hugepage (THP) for the object (without having to rely on a system wide parameter). v3: - use huge tmpfs mountpoint in drm_device v4: - fix builds with CONFIG_TRANSPARENT_HUGEPAGE=3Dn - clean up mountpoint creation error handling - print negative error value Signed-off-by: Lo=C3=AFc Molinari Reviewed-by: Boris Brezillon --- drivers/gpu/drm/panfrost/panfrost_device.c | 3 +++ drivers/gpu/drm/panfrost/panfrost_drv.c | 6 ++++++ drivers/gpu/drm/panfrost/panfrost_drv.h | 11 +++++++++++ drivers/gpu/drm/panfrost/panfrost_gem.c | 19 +++++++++++++++++++ drivers/gpu/drm/panfrost/panfrost_gem.h | 2 ++ 5 files changed, 41 insertions(+) create mode 100644 drivers/gpu/drm/panfrost/panfrost_drv.h diff --git a/drivers/gpu/drm/panfrost/panfrost_device.c b/drivers/gpu/drm/p= anfrost/panfrost_device.c index 04bec27449cb..a189bfaac959 100644 --- a/drivers/gpu/drm/panfrost/panfrost_device.c +++ b/drivers/gpu/drm/panfrost/panfrost_device.c @@ -12,6 +12,7 @@ #include "panfrost_device.h" #include "panfrost_devfreq.h" #include "panfrost_features.h" +#include "panfrost_gem.h" #include "panfrost_issues.h" #include "panfrost_gpu.h" #include "panfrost_job.h" @@ -266,6 +267,8 @@ int panfrost_device_init(struct panfrost_device *pfdev) if (err) goto out_job; =20 + panfrost_gem_init(pfdev); + return 0; out_job: panfrost_job_fini(pfdev); diff --git a/drivers/gpu/drm/panfrost/panfrost_drv.c b/drivers/gpu/drm/panf= rost/panfrost_drv.c index 607a5b8448d0..dfeb6a633fa1 100644 --- a/drivers/gpu/drm/panfrost/panfrost_drv.c +++ b/drivers/gpu/drm/panfrost/panfrost_drv.c @@ -858,6 +858,12 @@ static const struct drm_driver panfrost_drm_driver =3D= { #endif }; =20 +#ifdef CONFIG_TRANSPARENT_HUGEPAGE +bool panfrost_transparent_hugepage; +module_param_named(transparent_hugepage, panfrost_transparent_hugepage, bo= ol, 0400); +MODULE_PARM_DESC(transparent_hugepage, "Use a dedicated tmpfs mount point = with Transparent Hugepage enabled (false =3D default)"); +#endif + static int panfrost_probe(struct platform_device *pdev) { struct panfrost_device *pfdev; diff --git a/drivers/gpu/drm/panfrost/panfrost_drv.h b/drivers/gpu/drm/panf= rost/panfrost_drv.h new file mode 100644 index 000000000000..60454611b6a6 --- /dev/null +++ b/drivers/gpu/drm/panfrost/panfrost_drv.h @@ -0,0 +1,11 @@ +// SPDX-License-Identifier: GPL-2.0 or MIT +/* Copyright 2025 Amazon.com, Inc. or its affiliates */ + +#ifndef __PANFROST_DRV_H__ +#define __PANFROST_DRV_H__ + +#ifdef CONFIG_TRANSPARENT_HUGEPAGE +extern bool panfrost_transparent_hugepage; +#endif + +#endif diff --git a/drivers/gpu/drm/panfrost/panfrost_gem.c b/drivers/gpu/drm/panf= rost/panfrost_gem.c index 85d6289a6eda..9b059a8b0f9f 100644 --- a/drivers/gpu/drm/panfrost/panfrost_gem.c +++ b/drivers/gpu/drm/panfrost/panfrost_gem.c @@ -1,5 +1,6 @@ // SPDX-License-Identifier: GPL-2.0 /* Copyright 2019 Linaro, Ltd, Rob Herring */ +/* Copyright 2025 Amazon.com, Inc. or its affiliates */ =20 #include #include @@ -9,9 +10,27 @@ =20 #include #include "panfrost_device.h" +#include "panfrost_drv.h" #include "panfrost_gem.h" #include "panfrost_mmu.h" =20 +void panfrost_gem_init(struct panfrost_device *pfdev) +{ +#ifdef CONFIG_TRANSPARENT_HUGEPAGE + int err; + + if (!panfrost_transparent_hugepage) + return; + + err =3D drm_gem_huge_mnt_create(pfdev->ddev, "within_size"); + if (pfdev->ddev->huge_mnt) + drm_info(pfdev->ddev, "Using Transparent Hugepage\n"); + else if (err) + drm_warn(pfdev->ddev, "Can't use Transparent Hugepage (%d)\n", + err); +#endif +} + #ifdef CONFIG_DEBUG_FS static void panfrost_gem_debugfs_bo_add(struct panfrost_device *pfdev, struct panfrost_gem_object *bo) diff --git a/drivers/gpu/drm/panfrost/panfrost_gem.h b/drivers/gpu/drm/panf= rost/panfrost_gem.h index 8de3e76f2717..1a62529ff06f 100644 --- a/drivers/gpu/drm/panfrost/panfrost_gem.h +++ b/drivers/gpu/drm/panfrost/panfrost_gem.h @@ -124,6 +124,8 @@ drm_mm_node_to_panfrost_mapping(struct drm_mm_node *nod= e) return container_of(node, struct panfrost_gem_mapping, mmnode); } =20 +void panfrost_gem_init(struct panfrost_device *pfdev); + struct drm_gem_object *panfrost_gem_create_object(struct drm_device *dev, = size_t size); =20 struct drm_gem_object * --=20 2.47.3 From nobody Fri Dec 19 20:55:28 2025 Received: from bali.collaboradmins.com (bali.collaboradmins.com [148.251.105.195]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by smtp.subspace.kernel.org (Postfix) with ESMTPS id 8CFF33054F6; Wed, 15 Oct 2025 15:30:57 +0000 (UTC) Authentication-Results: smtp.subspace.kernel.org; arc=none smtp.client-ip=148.251.105.195 ARC-Seal: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1760542259; cv=none; b=HNB9gWMHkOqMYaHdlxfNfn51za2+rxxypadUX4SbR2yxGr9MAq5fDv+mbhBcy6grdyLrT/PS7S1j4EN0DG2K5fBe8QuJcG1lAY12z0n7OOkxNNaKL4tJJYnXW70WzM5V7fRNy3lOapz1XF1SxwOJwWTM9+FiSC8FIFByeKsaOuY= ARC-Message-Signature: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1760542259; c=relaxed/simple; bh=p24aKtUnOv8fOlBIkkg8gEu+LqJuafLvKyczTp6ZjwE=; h=From:To:Cc:Subject:Date:Message-ID:In-Reply-To:References: MIME-Version:Content-Type; b=E+c/j3lteS0lp6aMqlbH/mI4tvDaXGvCNMO5XRMmDrDSbFu1zsvatwl0TxhfrSZY9Gsm5zdwXsG4ytawbBRlv8iZjZje3RQtRdAqErrnSp4qOo4cJjCZtPgaF2e6qnratLDMlTpqYNpthMQEIDaMEJlDWb/HxCC0OuaEYiu0qT0= ARC-Authentication-Results: i=1; smtp.subspace.kernel.org; dmarc=pass (p=none dis=none) header.from=collabora.com; spf=pass smtp.mailfrom=collabora.com; dkim=pass (2048-bit key) header.d=collabora.com header.i=@collabora.com header.b=MEWQfOsC; arc=none smtp.client-ip=148.251.105.195 Authentication-Results: smtp.subspace.kernel.org; dmarc=pass (p=none dis=none) header.from=collabora.com Authentication-Results: smtp.subspace.kernel.org; spf=pass smtp.mailfrom=collabora.com Authentication-Results: smtp.subspace.kernel.org; dkim=pass (2048-bit key) header.d=collabora.com header.i=@collabora.com header.b="MEWQfOsC" DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=collabora.com; s=mail; t=1760542256; bh=p24aKtUnOv8fOlBIkkg8gEu+LqJuafLvKyczTp6ZjwE=; h=From:To:Cc:Subject:Date:In-Reply-To:References:From; b=MEWQfOsCnRe6aZZhYvFIqCiIoQf3UYeJT8GZNE11MLdXGrg7qQ99gGn0Mv1U5z26U hG9wdE50d/yKpuWNzmz7/B8MOUMTpJBwb05ymIJu7pAsnJs4YPyOzkvsgCcHAftyKV KsQL+eY9RSD8kFFjMdo32EzSAaD2p4TJrhgq8N55T5GDpkoNXuWBr/NEu4lalt6bKG bUwPvfd+5x2LCuHthh8sS/qcuAbZvG8qFEneVQHgSmPf98TC4ZhRo/5gMwZXN2W83R Yd6N+hYdGkGsufOyOah/xrsePiDYufrifamK0JwB7ria3gzFXxWObkpUFVOl72Qi8O nttUt+QahpoqA== Received: from debian-rockchip-rock5b-rk3588.. (unknown [IPv6:2a01:e0a:5e3:6100:826d:bc07:e98c:84a]) (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: loicmolinari) by bali.collaboradmins.com (Postfix) with ESMTPSA id 2095217E00A3; Wed, 15 Oct 2025 17:30:55 +0200 (CEST) From: =?UTF-8?q?Lo=C3=AFc=20Molinari?= To: Maarten Lankhorst , Maxime Ripard , Thomas Zimmermann , David Airlie , Simona Vetter , Jani Nikula , Joonas Lahtinen , Rodrigo Vivi , Tvrtko Ursulin , Boris Brezillon , Rob Herring , Steven Price , Liviu Dudau , Melissa Wen , =?UTF-8?q?Ma=C3=ADra=20Canal?= , Hugh Dickins , Baolin Wang , Andrew Morton , =?UTF-8?q?Lo=C3=AFc=20Molinari?= , Al Viro , =?UTF-8?q?Miko=C5=82aj=20Wasiak?= , Christian Brauner , Nitin Gote , Andi Shyti , Jonathan Corbet , Christopher Healy , Matthew Wilcox , Bagas Sanjaya Cc: linux-kernel@vger.kernel.org, dri-devel@lists.freedesktop.org, intel-gfx@lists.freedesktop.org, linux-mm@kvack.org, linux-doc@vger.kernel.org, kernel@collabora.com Subject: [PATCH v4 13/13] Documentation/gpu/drm-mm: Add THP paragraph to GEM mapping section Date: Wed, 15 Oct 2025 17:30:17 +0200 Message-ID: <20251015153018.43735-14-loic.molinari@collabora.com> X-Mailer: git-send-email 2.47.3 In-Reply-To: <20251015153018.43735-1-loic.molinari@collabora.com> References: <20251015153018.43735-1-loic.molinari@collabora.com> Precedence: bulk X-Mailing-List: linux-kernel@vger.kernel.org List-Id: List-Subscribe: List-Unsubscribe: MIME-Version: 1.0 Content-Type: text/plain; charset="utf-8" Content-Transfer-Encoding: quoted-printable Add a paragraph to the GEM objects mapping section explaining how transparent huge pages are handled by GEM. v4: - fix wording after huge_pages handler removal Signed-off-by: Lo=C3=AFc Molinari Reviewed-by: Bagas Sanjaya --- Documentation/gpu/drm-mm.rst | 25 ++++++++++++++++++++----- 1 file changed, 20 insertions(+), 5 deletions(-) diff --git a/Documentation/gpu/drm-mm.rst b/Documentation/gpu/drm-mm.rst index d55751cad67c..3d6176adc7ca 100644 --- a/Documentation/gpu/drm-mm.rst +++ b/Documentation/gpu/drm-mm.rst @@ -283,6 +283,9 @@ made up of several fields, the more interesting ones be= ing: void (*open)(struct vm_area_struct * area); void (*close)(struct vm_area_struct * area); vm_fault_t (*fault)(struct vm_fault *vmf); + vm_fault_t (*map_pages)(struct vm_fault *vmf, + pgoff_t start_pgoff, + pgoff_t end_pgoff); }; =20 =20 @@ -290,15 +293,27 @@ The open and close operations must update the GEM obj= ect reference count. Drivers can use the drm_gem_vm_open() and drm_gem_vm_close() helper functions directly as open and close handlers. =20 -The fault operation handler is responsible for mapping individual pages -to userspace when a page fault occurs. Depending on the memory -allocation scheme, drivers can allocate pages at fault time, or can -decide to allocate memory for the GEM object at the time the object is -created. +The fault and map_pages operations are responsible for mapping pages to +userspace when a page fault occurs. Depending on the memory allocation +scheme, drivers can allocate pages at fault time, or can decide to +allocate memory for the GEM object at the time the object is created. =20 Drivers that want to map the GEM object upfront instead of handling page faults can implement their own mmap file operation handler. =20 +In order to reduce page table overhead, if the internal shmem mountpoint +"shm_mnt" is configured to use transparent huge pages (for builds with +CONFIG_TRANSPARENT_HUGEPAGE enabled) and if the shmem backing store +managed to allocate a huge page for a faulty address, the fault and +map_pages handlers will first attempt to insert that huge page into the +VMA before falling back to individual page insertion. mmap() user +address alignment for GEM objects is handled by providing a custom +get_unmapped_area file operation which forwards to the shmem backing +store. For most drivers, which don't create a huge mountpoint by default +or through a module parameter, transparent huge pages can be enabled by +either setting the "transparent_hugepage_shmem" kernel parameter or the +"/sys/kernel/mm/transparent_hugepage/shmem_enabled" sysfs knob. + For platforms without MMU the GEM core provides a helper method drm_gem_dma_get_unmapped_area(). The mmap() routines will call this to get= a proposed address for the mapping. --=20 2.47.3