[Intel-gfx] [PATCH 02/23] drm/i915: introduce simple gemfs

[Joonas Lahtinen]

On Mon, 2017-08-21 at 19:34 +0100, Matthew Auld wrote:
> Not a fully blown gemfs, just our very own tmpfs kernel mount. Doing so
> moves us away from the shmemfs shm_mnt, and gives us the much needed
> flexibility to do things like set our own mount options, namely huge=
> which should allow us to enable the use of transparent-huge-pages for
> our shmem backed objects.
> 
> v2: various improvements suggested by Joonas
> 
> v3: move gemfs instance to i915.mm and simplify now that we have
> file_setup_with_mnt
> 
> Signed-off-by: Matthew Auld <matthew.auld at intel.com>
> Cc: Joonas Lahtinen <joonas.lahtinen at linux.intel.com>
> Cc: Chris Wilson <chris at chris-wilson.co.uk>
> Cc: Dave Hansen <dave.hansen at intel.com>
> Cc: Kirill A. Shutemov <kirill at shutemov.name>
> Cc: Hugh Dickins <hughd at google.com>
> Cc: linux-mm at kvack.org

<SNIP>

> @@ -4288,6 +4289,25 @@ static const struct drm_i915_gem_object_ops i915_gem_object_ops = {
>  	.pwrite = i915_gem_object_pwrite_gtt,
>  };
>  
> +static int i915_gem_object_create_shmem(struct drm_device *dev,
> +					struct drm_gem_object *obj,
> +					size_t size)
> +{
> +	struct drm_i915_private *i915 = to_i915(dev);
> +	struct file *filp;
> +
> +	drm_gem_private_object_init(dev, obj, size);
> +
> +	filp = shmem_file_setup_with_mnt(i915->mm.gemfs, "i915", size,
> +					 VM_NORESERVE);

Can you double-check that /proc/meminfo is unaffected by this change?
If we stop appearing under "Shemem:" we maybe need to maybe highlight
this somewhere (at least in commit message).

<SNIP>

> +int i915_gemfs_init(struct drm_i915_private *i915)
> +{
> +	struct file_system_type *type;
> +	struct vfsmount *gemfs;
> +
> +	type = get_fs_type("tmpfs");
> +	if (!type)
> +		return -ENODEV;
> +
> +	gemfs = kern_mount(type);
> +	if (IS_ERR(gemfs))
> +		return PTR_ERR(gemfs);

By occasionally checking that "i915->mm.gemfs" might be NULL, could we
continue without our own gemfs mount and just lose the additional
features? Or is it not worth the hassle?

Anyway, this is:

Reviewed-by: Joonas Lahtinen <joonas.lahtinen at linux.intel.com>

Regards, Joonas
-- 
Joonas Lahtinen
Open Source Technology Center
Intel Corporation