[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <CAF6AEGsZgjyv7r5_xWh1M9eR6+6A16bYZy9YLKAAf0Rm1iTnCQ@mail.gmail.com>
Date: Mon, 19 Dec 2022 09:10:33 -0800
From: Rob Clark <robdclark@...il.com>
To: Steven Price <steven.price@....com>
Cc: Rob Herring <robh@...nel.org>,
Tomeu Vizoso <tomeu.vizoso@...labora.com>,
Alyssa Rosenzweig <alyssa.rosenzweig@...labora.com>,
Rob Clark <robdclark@...omium.org>,
linux-kernel@...r.kernel.org, dri-devel@...ts.freedesktop.org
Subject: Re: [PATCH] drm/panfrost: Fix GEM handle creation ref-counting
On Mon, Dec 19, 2022 at 6:02 AM Steven Price <steven.price@....com> wrote:
>
> panfrost_gem_create_with_handle() previously returned a BO but with the
> only reference being from the handle, which user space could in theory
> guess and release, causing a use-after-free. Additionally if the call to
> panfrost_gem_mapping_get() in panfrost_ioctl_create_bo() failed then
> a(nother) reference on the BO was dropped.
>
> The _create_with_handle() is a problematic pattern, so ditch it and
> instead create the handle in panfrost_ioctl_create_bo(). If the call to
> panfrost_gem_mapping_get() fails then this means that user space has
> indeed gone behind our back and freed the handle. In which case just
> return an error code.
>
> Reported-by: Rob Clark <robdclark@...omium.org>
Yeah, I like getting rid of the _create_with_handle() pattern, the
only place where that pattern works is if you immediately return the
handle to userspace (and don't otherwise touch the obj)
Reviewed-by: Rob Clark <robdclark@...il.com>
> Fixes: f3ba91228e8e ("drm/panfrost: Add initial panfrost driver")
> Signed-off-by: Steven Price <steven.price@....com>
> ---
> drivers/gpu/drm/panfrost/panfrost_drv.c | 27 ++++++++++++++++---------
> drivers/gpu/drm/panfrost/panfrost_gem.c | 16 +--------------
> drivers/gpu/drm/panfrost/panfrost_gem.h | 5 +----
> 3 files changed, 20 insertions(+), 28 deletions(-)
>
> diff --git a/drivers/gpu/drm/panfrost/panfrost_drv.c b/drivers/gpu/drm/panfrost/panfrost_drv.c
> index fa619fe72086..abb0dadd8f63 100644
> --- a/drivers/gpu/drm/panfrost/panfrost_drv.c
> +++ b/drivers/gpu/drm/panfrost/panfrost_drv.c
> @@ -82,6 +82,7 @@ static int panfrost_ioctl_create_bo(struct drm_device *dev, void *data,
> struct panfrost_gem_object *bo;
> struct drm_panfrost_create_bo *args = data;
> struct panfrost_gem_mapping *mapping;
> + int ret;
>
> if (!args->size || args->pad ||
> (args->flags & ~(PANFROST_BO_NOEXEC | PANFROST_BO_HEAP)))
> @@ -92,21 +93,29 @@ static int panfrost_ioctl_create_bo(struct drm_device *dev, void *data,
> !(args->flags & PANFROST_BO_NOEXEC))
> return -EINVAL;
>
> - bo = panfrost_gem_create_with_handle(file, dev, args->size, args->flags,
> - &args->handle);
> + bo = panfrost_gem_create(dev, args->size, args->flags);
> if (IS_ERR(bo))
> return PTR_ERR(bo);
>
> + ret = drm_gem_handle_create(file, &bo->base.base, &args->handle);
> + if (ret)
> + goto out;
> +
> mapping = panfrost_gem_mapping_get(bo, priv);
> - if (!mapping) {
> - drm_gem_object_put(&bo->base.base);
> - return -EINVAL;
> + if (mapping) {
> + args->offset = mapping->mmnode.start << PAGE_SHIFT;
> + panfrost_gem_mapping_put(mapping);
> + } else {
> + /* This can only happen if the handle from
> + * drm_gem_handle_create() has already been guessed and freed
> + * by user space
> + */
> + ret = -EINVAL;
> }
>
> - args->offset = mapping->mmnode.start << PAGE_SHIFT;
> - panfrost_gem_mapping_put(mapping);
> -
> - return 0;
> +out:
> + drm_gem_object_put(&bo->base.base);
> + return ret;
> }
>
> /**
> diff --git a/drivers/gpu/drm/panfrost/panfrost_gem.c b/drivers/gpu/drm/panfrost/panfrost_gem.c
> index 293e799e2fe8..3c812fbd126f 100644
> --- a/drivers/gpu/drm/panfrost/panfrost_gem.c
> +++ b/drivers/gpu/drm/panfrost/panfrost_gem.c
> @@ -235,12 +235,8 @@ struct drm_gem_object *panfrost_gem_create_object(struct drm_device *dev, size_t
> }
>
> struct panfrost_gem_object *
> -panfrost_gem_create_with_handle(struct drm_file *file_priv,
> - struct drm_device *dev, size_t size,
> - u32 flags,
> - uint32_t *handle)
> +panfrost_gem_create(struct drm_device *dev, size_t size, u32 flags)
> {
> - int ret;
> struct drm_gem_shmem_object *shmem;
> struct panfrost_gem_object *bo;
>
> @@ -256,16 +252,6 @@ panfrost_gem_create_with_handle(struct drm_file *file_priv,
> bo->noexec = !!(flags & PANFROST_BO_NOEXEC);
> bo->is_heap = !!(flags & PANFROST_BO_HEAP);
>
> - /*
> - * Allocate an id of idr table where the obj is registered
> - * and handle has the id what user can see.
> - */
> - ret = drm_gem_handle_create(file_priv, &shmem->base, handle);
> - /* drop reference from allocate - handle holds it now. */
> - drm_gem_object_put(&shmem->base);
> - if (ret)
> - return ERR_PTR(ret);
> -
> return bo;
> }
>
> diff --git a/drivers/gpu/drm/panfrost/panfrost_gem.h b/drivers/gpu/drm/panfrost/panfrost_gem.h
> index 8088d5fd8480..ad2877eeeccd 100644
> --- a/drivers/gpu/drm/panfrost/panfrost_gem.h
> +++ b/drivers/gpu/drm/panfrost/panfrost_gem.h
> @@ -69,10 +69,7 @@ panfrost_gem_prime_import_sg_table(struct drm_device *dev,
> struct sg_table *sgt);
>
> struct panfrost_gem_object *
> -panfrost_gem_create_with_handle(struct drm_file *file_priv,
> - struct drm_device *dev, size_t size,
> - u32 flags,
> - uint32_t *handle);
> +panfrost_gem_create(struct drm_device *dev, size_t size, u32 flags);
>
> int panfrost_gem_open(struct drm_gem_object *obj, struct drm_file *file_priv);
> void panfrost_gem_close(struct drm_gem_object *obj,
> --
> 2.34.1
>
Powered by blists - more mailing lists