linux-kernel - Re: [PATCH] drm/cma: Fix recent regression of mmap() in the MMU case.

Open Source and information security mailing list archives

Message-ID: <20170418065234.6y543izrwwus4rq3@phenom.ffwll.local>
Date:   Tue, 18 Apr 2017 08:52:34 +0200
From:   Daniel Vetter <daniel@...ll.ch>
To:     Eric Anholt <eric@...olt.net>
Cc:     dri-devel@...ts.freedesktop.org,
        Yannick Fertre <yannick.fertre@...com>,
        linux-kernel@...r.kernel.org
Subject: Re: [PATCH] drm/cma: Fix recent regression of mmap() in the MMU case.

On Mon, Apr 17, 2017 at 04:31:24PM -0700, Eric Anholt wrote:
> The stub get_unmapped_area() function was actually getting called, so
> all of our mmap()s failed.
> 
> Cc: Yannick Fertre <yannick.fertre@...com>
> Fixes: 97bf3a9aa60f ("drm/cma: Update DEFINE_DRM_GEM_CMA_FOPS to add get_unmapped_area")
> Signed-off-by: Eric Anholt <eric@...olt.net>

Oh drat, I assumed the core would only call this when needed, but things
don't work like that.

Reviewed-by: Daniel Vetter <daniel.vetter@...ll.ch>

The strange thing is that in some cases ->mm->get_unmapped_area overwrites
fops->get_unmapped_area, but not for do_mmap in the mmu case. Seems not
entirely thought out to me ...
-Daniel

> ---
>  include/drm/drm_gem_cma_helper.h | 18 ++++++++----------
>  1 file changed, 8 insertions(+), 10 deletions(-)
> 
> diff --git a/include/drm/drm_gem_cma_helper.h b/include/drm/drm_gem_cma_helper.h
> index 7320b140545a..b42529e0fae0 100644
> --- a/include/drm/drm_gem_cma_helper.h
> +++ b/include/drm/drm_gem_cma_helper.h
> @@ -26,6 +26,13 @@ to_drm_gem_cma_obj(struct drm_gem_object *gem_obj)
>  	return container_of(gem_obj, struct drm_gem_cma_object, base);
>  }
>  
> +#ifndef CONFIG_MMU
> +#define DRM_GEM_CMA_UNMAPPED_AREA_FOPS \
> +	.get_unmapped_area	= drm_gem_cma_get_unmapped_area,
> +#else
> +#define DRM_GEM_CMA_UNMAPPED_AREA_FOPS
> +#endif
> +
>  /**
>   * DEFINE_DRM_GEM_CMA_FOPS() - macro to generate file operations for CMA drivers
>   * @name: name for the generated structure
> @@ -50,7 +57,7 @@ to_drm_gem_cma_obj(struct drm_gem_object *gem_obj)
>  		.read		= drm_read,\
>  		.llseek		= noop_llseek,\
>  		.mmap		= drm_gem_cma_mmap,\
> -		.get_unmapped_area	= drm_gem_cma_get_unmapped_area,\
> +		DRM_GEM_CMA_UNMAPPED_AREA_FOPS \
>  	}
>  
>  /* free GEM object */
> @@ -86,15 +93,6 @@ unsigned long drm_gem_cma_get_unmapped_area(struct file *filp,
>  					    unsigned long len,
>  					    unsigned long pgoff,
>  					    unsigned long flags);
> -#else
> -static inline unsigned long drm_gem_cma_get_unmapped_area(struct file *filp,
> -							  unsigned long addr,
> -							  unsigned long len,
> -							  unsigned long pgoff,
> -							  unsigned long flags)
> -{
> -	return -EINVAL;
> -}
>  #endif
>  
>  #ifdef CONFIG_DEBUG_FS
> -- 
> 2.11.0
> 
> _______________________________________________
> dri-devel mailing list
> dri-devel@...ts.freedesktop.org
> https://lists.freedesktop.org/mailman/listinfo/dri-devel

-- 
Daniel Vetter
Software Engineer, Intel Corporation
http://blog.ffwll.ch

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives