Message-ID: <7584abe7-0c3f-4022-b510-c2a57fd167bb@ursulin.net>
Date: Mon, 20 Oct 2025 10:10:33 +0100
From: Tvrtko Ursulin <tursulin@...ulin.net>
To: Loïc Molinari <loic.molinari@...labora.com>,
 Maarten Lankhorst <maarten.lankhorst@...ux.intel.com>,
 Maxime Ripard <mripard@...nel.org>, Thomas Zimmermann <tzimmermann@...e.de>,
 David Airlie <airlied@...il.com>, Simona Vetter <simona@...ll.ch>,
 Jani Nikula <jani.nikula@...ux.intel.com>,
 Joonas Lahtinen <joonas.lahtinen@...ux.intel.com>,
 Rodrigo Vivi <rodrigo.vivi@...el.com>,
 Boris Brezillon <boris.brezillon@...labora.com>,
 Rob Herring <robh@...nel.org>, Steven Price <steven.price@....com>,
 Liviu Dudau <liviu.dudau@....com>, Melissa Wen <mwen@...lia.com>,
 Maíra Canal <mcanal@...lia.com>,
 Hugh Dickins <hughd@...gle.com>, Baolin Wang
 <baolin.wang@...ux.alibaba.com>, Andrew Morton <akpm@...ux-foundation.org>,
 Al Viro <viro@...iv.linux.org.uk>, Mikołaj Wasiak
 <mikolaj.wasiak@...el.com>, Christian Brauner <brauner@...nel.org>,
 Nitin Gote <nitin.r.gote@...el.com>, Andi Shyti
 <andi.shyti@...ux.intel.com>, Jonathan Corbet <corbet@....net>,
 Christopher Healy <healych@...zon.com>, Matthew Wilcox
 <willy@...radead.org>, Bagas Sanjaya <bagasdotme@...il.com>
Cc: linux-kernel@...r.kernel.org, dri-devel@...ts.freedesktop.org,
 intel-gfx@...ts.freedesktop.org, linux-mm@...ck.org,
 linux-doc@...r.kernel.org, kernel@...labora.com
Subject: Re: [PATCH v4 05/13] drm/gem: Add huge tmpfs mount point helper


On 15/10/2025 16:30, Loïc Molinari wrote:
> 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=n
> - return 0 when huge_mnt already exists
> 
> Signed-off-by: Loïc Molinari <loic.molinari@...labora.com>
> ---
>   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 <linux/export.h>
>   #include <linux/file.h>
>   #include <linux/fs.h>
> +#include <linux/fs_context.h>
>   #include <linux/iosys-map.h>
>   #include <linux/mem_encrypt.h>
>   #include <linux/mm.h>
> @@ -82,6 +83,63 @@
>    * up at a later date, and as our interface with shmfs for memory allocation.
>    */
>   
> +static void drm_gem_huge_mnt_free(struct drm_device *dev, void *data)
> +{
> +	drm_WARN_ON(dev, dev->huge_mnt == NULL);

I don't see a benefit of adding this check but maybe I am missing something.

> +
> +	kern_unmount(dev->huge_mnt);
> +	dev->huge_mnt = NULL;

Ditto - device is going away, no? So why bother clearing the pointer?

Also, is the compiler smart enough to not compile or complain this 
function is unused in the !CONFIG_TRANSPARENT_HUGEPAGE case?

> +}
> +
> +/**
> + * 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 initialization
> + * with drm_gem_object_init().
> + *
> + * The most common option value @value is "within_size" which only allocates
> + * huge pages if the page will be fully within the GEM object size. "always",
> + * "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;

Is there a specific reason why the !CONFIG_TRANSPARENT_HUGEPAGE path is 
not implemented in the header as a static inline? That would enable 
those builds to avoid the pointless function in text and function call 
in the drivers.

> +	if (unlikely(dev->huge_mnt))
> +		return 0;

Any special reason why it is allowed to call it multiple times with success?

> +
> +	type = get_fs_type("tmpfs");
> +	if (unlikely(!type))
> +		return -EOPNOTSUPP;
> +	fc = fs_context_for_mount(type, SB_KERNMOUNT);
> +	if (IS_ERR(fc))
> +		return PTR_ERR(fc);
> +	ret = vfs_parse_fs_string(fc, "source", "tmpfs");
> +	if (unlikely(ret))
> +		return -ENOPARAM;
> +	ret = vfs_parse_fs_string(fc, "huge", value);
> +	if (unlikely(ret))
> +		return -ENOPARAM;
> +
> +	dev->huge_mnt = 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 @@
>   
>   #include <linux/list.h>
>   #include <linux/kref.h>
> +#include <linux/mount.h>
>   #include <linux/mutex.h>
>   #include <linux/idr.h>
>   #include <linux/sched.h>
> @@ -168,6 +169,16 @@ struct drm_device {
>   	 */
>   	struct drm_master *master;
>   
> +	/**
> +	 * @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;

Maybe it would be nice to hide this in the !CONFIG_TRANSPARENT_HUGEPAGE 
case? A bit ugly to add an ifdef but it is also a bit questionable to 
force the member on everyone.

Regards,

Tvrtko

> +
>   	/**
>   	 * @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,\
>   	}
>   
> +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,

Powered by blists - more mailing lists