[PATCH v4] DRM: add DRM Driver for Samsung SoC EXYNOS4210.

[PATCH v4] DRM: add DRM Driver for Samsung SoC EXYNOS4210.

[Thomas Hellstrom]

On 09/09/2011 01:38 PM, Inki Dae wrote:
> This patch is a DRM Driver for Samsung SoC Exynos4210 and now enables
> only FIMD yet but we will add HDMI support also in the future.
>
> from now on, I will remove RFC prefix because I think we have got comments
> enough.
>
> this patch is based on git repository below:
> git://git.kernel.org/pub/scm/linux/kernel/git/airlied/drm-2.6.git,
> branch name: drm-next
> commit-id: bcc65fd8e929a9d9d34d814d6efc1d2793546922
>
> you can refer to our working repository below:
> http://git.infradead.org/users/kmpark/linux-2.6-samsung
> branch name: samsung-drm
>
> We tried to re-use lowlevel codes of the FIMD driver(s3c-fb.c
> based on Linux framebuffer) but couldn't so because lowlevel codes
> of s3c-fb.c are included internally and so FIMD module of this driver has
> its own lowlevel codes.
>
> We used GEM framework for buffer management and DMA APIs(dma_alloc_*)
> for buffer allocation. by using DMA API, we could use CMA later.
>
> Refer to this link for CMA(Continuous Memory Allocator):
> http://lkml.org/lkml/2011/7/20/45
>
> this driver supports only physically continuous memory(non-iommu).
>
> Links to previous versions of the patchset:
> v1:<  https://lwn.net/Articles/454380/>
> v2:<  http://www.spinics.net/lists/kernel/msg1224275.html>
> v3:<  http://www.gossamer-threads.com/lists/linux/kernel/1423684>
>
> Changelog v2:
> DRM: add DRM_IOCTL_SAMSUNG_GEM_MMAP ioctl command.
>
>      this feature maps user address space to physical memory region
>      once user application requests DRM_IOCTL_SAMSUNG_GEM_MMAP ioctl.
>
> DRM: code clean and add exception codes.
>
> Changelog v3:
> DRM: Support multiple irq.
>
>      FIMD and HDMI have their own irq handler but DRM Framework can regiter
>      only one irq handler this patch supports mutiple irq for Samsung SoC.
>
> DRM: Consider modularization.
>
>      each DRM, FIMD could be built as a module.
>
> DRM: Have indenpendent crtc object.
>
>      crtc isn't specific to SoC Platform so this patch gets a crtc
>      to be used as common object.
>      created crtc could be attached to any encoder object.
>
> DRM: code clean and add exception codes.
>
> Changelog v4:
> DRM: remove is_defult from samsung_fb.
>
>      is_default isn't used for default framebuffer.
>
> DRM: code refactoring to fimd module.
>      this patch is be considered with multiple display objects and
>      would use its own request_irq() to register a irq handler instead of
>      drm framework's one.
>
> DRM: remove find_samsung_drm_gem_object()
>
> DRM: move kernel private data structures and definitions to driver folder.
>
>      samsung_drm.h would contain only public information for userspace
>      ioctl interface.
>
> DRM: code refactoring to gem modules.
>      buffer module isn't dependent of gem module anymore.
>
> DRM: fixed security issue.
>
> DRM: remove encoder porinter from specific connector.
>
>      samsung connector doesn't need to have generic encoder.
>
> DRM: code clean and add exception codes.
>
> Signed-off-by: Inki Dae<inki.dae@samsung.com>
> Signed-off-by: Joonyoung Shim<jy0922.shim@samsung.com>
> Signed-off-by: SeungWoo Kim<sw0312.kim@samsung.com>
> Signed-off-by: kyungmin.park<kyungmin.park@samsung.com>
> ---
>
> +static struct drm_ioctl_desc samsung_ioctls[] = {
> +	DRM_IOCTL_DEF_DRV(SAMSUNG_GEM_CREATE, samsung_drm_gem_create_ioctl,
> +			DRM_UNLOCKED | DRM_AUTH),
>    

Hi!

With reference my previous security comment.

Let's say you have a compromised video player running as a DRM client, 
that tries to repeatedly allocate huge GEM buffers...

What will happen when all DMA memory is exhausted? Will this cause other 
device drivers to see an OOM, or only DRM?

The old DRI model basically allowed any authorized DRI client to exhaust 
video ram or AGP memory, but never system memory. Newer DRI drivers 
typically only allow DRI masters to do that.
as

I don't think an authorized DRI client should be able to easily exhaust 
resources (DMA memory) used by other device drivers causing them to fail.

/Thomas


> +	DRM_IOCTL_DEF_DRV(SAMSUNG_GEM_MAP_OFFSET,
> +			samsung_drm_gem_map_offset_ioctl, DRM_UNLOCKED |
> +				DRM_AUTH),
> +	DRM_IOCTL_DEF_DRV(SAMSUNG_GEM_MMAP,
> +			samsung_drm_gem_mmap_ioctl, DRM_UNLOCKED | DRM_AUTH),
> +};
>
>    

[PATCH v4] DRM: add DRM Driver for Samsung SoC EXYNOS4210.

[Rob Clark]

On Sat, Sep 10, 2011 at 9:04 AM, Thomas Hellstrom <thomas@shipmail.org> wrote:
> On 09/09/2011 01:38 PM, Inki Dae wrote:
>>
>> This patch is a DRM Driver for Samsung SoC Exynos4210 and now enables
>> only FIMD yet but we will add HDMI support also in the future.
>>
>> from now on, I will remove RFC prefix because I think we have got comments
>> enough.
>>
>> this patch is based on git repository below:
>> git://git.kernel.org/pub/scm/linux/kernel/git/airlied/drm-2.6.git,
>> branch name: drm-next
>> commit-id: bcc65fd8e929a9d9d34d814d6efc1d2793546922
>>
>> you can refer to our working repository below:
>> http://git.infradead.org/users/kmpark/linux-2.6-samsung
>> branch name: samsung-drm
>>
>> We tried to re-use lowlevel codes of the FIMD driver(s3c-fb.c
>> based on Linux framebuffer) but couldn't so because lowlevel codes
>> of s3c-fb.c are included internally and so FIMD module of this driver has
>> its own lowlevel codes.
>>
>> We used GEM framework for buffer management and DMA APIs(dma_alloc_*)
>> for buffer allocation. by using DMA API, we could use CMA later.
>>
>> Refer to this link for CMA(Continuous Memory Allocator):
>> http://lkml.org/lkml/2011/7/20/45
>>
>> this driver supports only physically continuous memory(non-iommu).
>>
>> Links to previous versions of the patchset:
>> v1:< ?https://lwn.net/Articles/454380/>
>> v2:< ?http://www.spinics.net/lists/kernel/msg1224275.html>
>> v3:< ?http://www.gossamer-threads.com/lists/linux/kernel/1423684>
>>
>> Changelog v2:
>> DRM: add DRM_IOCTL_SAMSUNG_GEM_MMAP ioctl command.
>>
>> ? ? this feature maps user address space to physical memory region
>> ? ? once user application requests DRM_IOCTL_SAMSUNG_GEM_MMAP ioctl.
>>
>> DRM: code clean and add exception codes.
>>
>> Changelog v3:
>> DRM: Support multiple irq.
>>
>> ? ? FIMD and HDMI have their own irq handler but DRM Framework can regiter
>> ? ? only one irq handler this patch supports mutiple irq for Samsung SoC.
>>
>> DRM: Consider modularization.
>>
>> ? ? each DRM, FIMD could be built as a module.
>>
>> DRM: Have indenpendent crtc object.
>>
>> ? ? crtc isn't specific to SoC Platform so this patch gets a crtc
>> ? ? to be used as common object.
>> ? ? created crtc could be attached to any encoder object.
>>
>> DRM: code clean and add exception codes.
>>
>> Changelog v4:
>> DRM: remove is_defult from samsung_fb.
>>
>> ? ? is_default isn't used for default framebuffer.
>>
>> DRM: code refactoring to fimd module.
>> ? ? this patch is be considered with multiple display objects and
>> ? ? would use its own request_irq() to register a irq handler instead of
>> ? ? drm framework's one.
>>
>> DRM: remove find_samsung_drm_gem_object()
>>
>> DRM: move kernel private data structures and definitions to driver folder.
>>
>> ? ? samsung_drm.h would contain only public information for userspace
>> ? ? ioctl interface.
>>
>> DRM: code refactoring to gem modules.
>> ? ? buffer module isn't dependent of gem module anymore.
>>
>> DRM: fixed security issue.
>>
>> DRM: remove encoder porinter from specific connector.
>>
>> ? ? samsung connector doesn't need to have generic encoder.
>>
>> DRM: code clean and add exception codes.
>>
>> Signed-off-by: Inki Dae<inki.dae@samsung.com>
>> Signed-off-by: Joonyoung Shim<jy0922.shim@samsung.com>
>> Signed-off-by: SeungWoo Kim<sw0312.kim@samsung.com>
>> Signed-off-by: kyungmin.park<kyungmin.park@samsung.com>
>> ---
>>
>> +static struct drm_ioctl_desc samsung_ioctls[] = {
>> + ? ? ? DRM_IOCTL_DEF_DRV(SAMSUNG_GEM_CREATE,
>> samsung_drm_gem_create_ioctl,
>> + ? ? ? ? ? ? ? ? ? ? ? DRM_UNLOCKED | DRM_AUTH),
>>
>
> Hi!
>
> With reference my previous security comment.
>
> Let's say you have a compromised video player running as a DRM client, that
> tries to repeatedly allocate huge GEM buffers...
>
> What will happen when all DMA memory is exhausted? Will this cause other
> device drivers to see an OOM, or only DRM?
>
> The old DRI model basically allowed any authorized DRI client to exhaust
> video ram or AGP memory, but never system memory. Newer DRI drivers
> typically only allow DRI masters to do that.
> as
>
> I don't think an authorized DRI client should be able to easily exhaust
> resources (DMA memory) used by other device drivers causing them to fail.

I'm not entirely sure what else can be done, other than have a
threshold on max MB allocatable of buffer memory..

In the samsung driver case, he is only allocating scanout memory from
CMA, so the limit will be the CMA region size.. beyond that you can't
get physically contiguous memory.  So I think this driver is safe.

In the omap driver case, I'm also allowing allocation of shmem backed
buffers.  So perhaps I need a threshold of how many buffers can have
pages attached.  Offhand, I'd have to think about how to handle that..
 have any of the other UMA gfx drivers dealt with this yet?

BR,
-R

> /Thomas
>
>
>> + ? ? ? DRM_IOCTL_DEF_DRV(SAMSUNG_GEM_MAP_OFFSET,
>> + ? ? ? ? ? ? ? ? ? ? ? samsung_drm_gem_map_offset_ioctl, DRM_UNLOCKED |
>> + ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? DRM_AUTH),
>> + ? ? ? DRM_IOCTL_DEF_DRV(SAMSUNG_GEM_MMAP,
>> + ? ? ? ? ? ? ? ? ? ? ? samsung_drm_gem_mmap_ioctl, DRM_UNLOCKED |
>> DRM_AUTH),
>> +};
>>
>>
>
>
>
> _______________________________________________
> dri-devel mailing list
> dri-devel at lists.freedesktop.org
> http://lists.freedesktop.org/mailman/listinfo/dri-devel
>

[PATCH v4] DRM: add DRM Driver for Samsung SoC EXYNOS4210.

[Thomas Hellstrom]

On 09/10/2011 07:31 PM, Rob Clark wrote:
> On Sat, Sep 10, 2011 at 9:04 AM, Thomas Hellstrom<thomas@shipmail.org>  wrote:
>    
>> On 09/09/2011 01:38 PM, Inki Dae wrote:
>>      
>>> This patch is a DRM Driver for Samsung SoC Exynos4210 and now enables
>>> only FIMD yet but we will add HDMI support also in the future.
>>>
>>> from now on, I will remove RFC prefix because I think we have got comments
>>> enough.
>>>
>>> this patch is based on git repository below:
>>> git://git.kernel.org/pub/scm/linux/kernel/git/airlied/drm-2.6.git,
>>> branch name: drm-next
>>> commit-id: bcc65fd8e929a9d9d34d814d6efc1d2793546922
>>>
>>> you can refer to our working repository below:
>>> http://git.infradead.org/users/kmpark/linux-2.6-samsung
>>> branch name: samsung-drm
>>>
>>> We tried to re-use lowlevel codes of the FIMD driver(s3c-fb.c
>>> based on Linux framebuffer) but couldn't so because lowlevel codes
>>> of s3c-fb.c are included internally and so FIMD module of this driver has
>>> its own lowlevel codes.
>>>
>>> We used GEM framework for buffer management and DMA APIs(dma_alloc_*)
>>> for buffer allocation. by using DMA API, we could use CMA later.
>>>
>>> Refer to this link for CMA(Continuous Memory Allocator):
>>> http://lkml.org/lkml/2011/7/20/45
>>>
>>> this driver supports only physically continuous memory(non-iommu).
>>>
>>> Links to previous versions of the patchset:
>>> v1:<    https://lwn.net/Articles/454380/>
>>> v2:<    http://www.spinics.net/lists/kernel/msg1224275.html>
>>> v3:<    http://www.gossamer-threads.com/lists/linux/kernel/1423684>
>>>
>>> Changelog v2:
>>> DRM: add DRM_IOCTL_SAMSUNG_GEM_MMAP ioctl command.
>>>
>>>      this feature maps user address space to physical memory region
>>>      once user application requests DRM_IOCTL_SAMSUNG_GEM_MMAP ioctl.
>>>
>>> DRM: code clean and add exception codes.
>>>
>>> Changelog v3:
>>> DRM: Support multiple irq.
>>>
>>>      FIMD and HDMI have their own irq handler but DRM Framework can regiter
>>>      only one irq handler this patch supports mutiple irq for Samsung SoC.
>>>
>>> DRM: Consider modularization.
>>>
>>>      each DRM, FIMD could be built as a module.
>>>
>>> DRM: Have indenpendent crtc object.
>>>
>>>      crtc isn't specific to SoC Platform so this patch gets a crtc
>>>      to be used as common object.
>>>      created crtc could be attached to any encoder object.
>>>
>>> DRM: code clean and add exception codes.
>>>
>>> Changelog v4:
>>> DRM: remove is_defult from samsung_fb.
>>>
>>>      is_default isn't used for default framebuffer.
>>>
>>> DRM: code refactoring to fimd module.
>>>      this patch is be considered with multiple display objects and
>>>      would use its own request_irq() to register a irq handler instead of
>>>      drm framework's one.
>>>
>>> DRM: remove find_samsung_drm_gem_object()
>>>
>>> DRM: move kernel private data structures and definitions to driver folder.
>>>
>>>      samsung_drm.h would contain only public information for userspace
>>>      ioctl interface.
>>>
>>> DRM: code refactoring to gem modules.
>>>      buffer module isn't dependent of gem module anymore.
>>>
>>> DRM: fixed security issue.
>>>
>>> DRM: remove encoder porinter from specific connector.
>>>
>>>      samsung connector doesn't need to have generic encoder.
>>>
>>> DRM: code clean and add exception codes.
>>>
>>> Signed-off-by: Inki Dae<inki.dae@samsung.com>
>>> Signed-off-by: Joonyoung Shim<jy0922.shim@samsung.com>
>>> Signed-off-by: SeungWoo Kim<sw0312.kim@samsung.com>
>>> Signed-off-by: kyungmin.park<kyungmin.park@samsung.com>
>>> ---
>>>
>>> +static struct drm_ioctl_desc samsung_ioctls[] = {
>>> +       DRM_IOCTL_DEF_DRV(SAMSUNG_GEM_CREATE,
>>> samsung_drm_gem_create_ioctl,
>>> +                       DRM_UNLOCKED | DRM_AUTH),
>>>
>>>        
>> Hi!
>>
>> With reference my previous security comment.
>>
>> Let's say you have a compromised video player running as a DRM client, that
>> tries to repeatedly allocate huge GEM buffers...
>>
>> What will happen when all DMA memory is exhausted? Will this cause other
>> device drivers to see an OOM, or only DRM?
>>
>> The old DRI model basically allowed any authorized DRI client to exhaust
>> video ram or AGP memory, but never system memory. Newer DRI drivers
>> typically only allow DRI masters to do that.
>> as
>>
>> I don't think an authorized DRI client should be able to easily exhaust
>> resources (DMA memory) used by other device drivers causing them to fail.
>>      
> I'm not entirely sure what else can be done, other than have a
> threshold on max MB allocatable of buffer memory..
>    

Yes, I think that's what needs to be done, and that threshold should be 
low enough to keep other device drivers running in the worst allocation 
case.

> In the samsung driver case, he is only allocating scanout memory from
> CMA, so the limit will be the CMA region size.. beyond that you can't
> get physically contiguous memory.  So I think this driver is safe.
>    

It's not really what well-behaved user-space drivers do that should be a 
concern, but what compromized application *might* do that is a concern.

> In the omap driver case, I'm also allowing allocation of shmem backed
> buffers.  So perhaps I need a threshold of how many buffers can have
> pages attached.  Offhand, I'd have to think about how to handle that..
>   have any of the other UMA gfx drivers dealt with this yet?
>    

I think shmem pages are more well behaved since they are allocated out 
of the page cache. When the page cache is exhausted, the kernel resorts 
to shrinkers to claim more memory, but the structures backing the shmem 
buffers use pinned memory and may exhaust available kmalloc memory, 
unless, as you say, their number is somehow limited. I'm not sure 
whether the shmem subsystem actually does this already...

I don't know how / if the Intel driver implements this. The ttm-aware 
drivers does some simple accounting of system memory used for graphics 
purposes, and limits it to a preset percentage of available system 
memory, and that percentage is adjustable using a sysfs entry. This 
requires, however, that the drivers register all long-lived data 
structures created as a result of a user-space call with the accounting 
system.

/Thomas


> BR,
> -R
>    





>    
>> /Thomas
>>
>>
>>      
>>> +       DRM_IOCTL_DEF_DRV(SAMSUNG_GEM_MAP_OFFSET,
>>> +                       samsung_drm_gem_map_offset_ioctl, DRM_UNLOCKED |
>>> +                               DRM_AUTH),
>>> +       DRM_IOCTL_DEF_DRV(SAMSUNG_GEM_MMAP,
>>> +                       samsung_drm_gem_mmap_ioctl, DRM_UNLOCKED |
>>> DRM_AUTH),
>>> +};
>>>
>>>
>>>        
>>
>>
>> _______________________________________________
>> dri-devel mailing list
>> dri-devel at lists.freedesktop.org
>> http://lists.freedesktop.org/mailman/listinfo/dri-devel
>>
>>      

[PATCH v4] DRM: add DRM Driver for Samsung SoC EXYNOS4210.

[Thomas Hellstrom]

On 09/11/2011 11:26 PM, Thomas Hellstrom wrote:
> On 09/10/2011 07:31 PM, Rob Clark wrote:
>> On Sat, Sep 10, 2011 at 9:04 AM, Thomas 
>> Hellstrom<thomas@shipmail.org>  wrote:
>>> On 09/09/2011 01:38 PM, Inki Dae wrote:
>>>> This patch is a DRM Driver for Samsung SoC Exynos4210 and now enables
>>>> only FIMD yet but we will add HDMI support also in the future.
>>>>
>>>> from now on, I will remove RFC prefix because I think we have got 
>>>> comments
>>>> enough.
>>>>
>>>> this patch is based on git repository below:
>>>> git://git.kernel.org/pub/scm/linux/kernel/git/airlied/drm-2.6.git,
>>>> branch name: drm-next
>>>> commit-id: bcc65fd8e929a9d9d34d814d6efc1d2793546922
>>>>
>>>> you can refer to our working repository below:
>>>> http://git.infradead.org/users/kmpark/linux-2.6-samsung
>>>> branch name: samsung-drm
>>>>
>>>> We tried to re-use lowlevel codes of the FIMD driver(s3c-fb.c
>>>> based on Linux framebuffer) but couldn't so because lowlevel codes
>>>> of s3c-fb.c are included internally and so FIMD module of this 
>>>> driver has
>>>> its own lowlevel codes.
>>>>
>>>> We used GEM framework for buffer management and DMA APIs(dma_alloc_*)
>>>> for buffer allocation. by using DMA API, we could use CMA later.
>>>>
>>>> Refer to this link for CMA(Continuous Memory Allocator):
>>>> http://lkml.org/lkml/2011/7/20/45
>>>>
>>>> this driver supports only physically continuous memory(non-iommu).
>>>>
>>>> Links to previous versions of the patchset:
>>>> v1:<    https://lwn.net/Articles/454380/>
>>>> v2:<    http://www.spinics.net/lists/kernel/msg1224275.html>
>>>> v3:<    http://www.gossamer-threads.com/lists/linux/kernel/1423684>
>>>>
>>>> Changelog v2:
>>>> DRM: add DRM_IOCTL_SAMSUNG_GEM_MMAP ioctl command.
>>>>
>>>>      this feature maps user address space to physical memory region
>>>>      once user application requests DRM_IOCTL_SAMSUNG_GEM_MMAP ioctl.
>>>>
>>>> DRM: code clean and add exception codes.
>>>>
>>>> Changelog v3:
>>>> DRM: Support multiple irq.
>>>>
>>>>      FIMD and HDMI have their own irq handler but DRM Framework can 
>>>> regiter
>>>>      only one irq handler this patch supports mutiple irq for 
>>>> Samsung SoC.
>>>>
>>>> DRM: Consider modularization.
>>>>
>>>>      each DRM, FIMD could be built as a module.
>>>>
>>>> DRM: Have indenpendent crtc object.
>>>>
>>>>      crtc isn't specific to SoC Platform so this patch gets a crtc
>>>>      to be used as common object.
>>>>      created crtc could be attached to any encoder object.
>>>>
>>>> DRM: code clean and add exception codes.
>>>>
>>>> Changelog v4:
>>>> DRM: remove is_defult from samsung_fb.
>>>>
>>>>      is_default isn't used for default framebuffer.
>>>>
>>>> DRM: code refactoring to fimd module.
>>>>      this patch is be considered with multiple display objects and
>>>>      would use its own request_irq() to register a irq handler 
>>>> instead of
>>>>      drm framework's one.
>>>>
>>>> DRM: remove find_samsung_drm_gem_object()
>>>>
>>>> DRM: move kernel private data structures and definitions to driver 
>>>> folder.
>>>>
>>>>      samsung_drm.h would contain only public information for userspace
>>>>      ioctl interface.
>>>>
>>>> DRM: code refactoring to gem modules.
>>>>      buffer module isn't dependent of gem module anymore.
>>>>
>>>> DRM: fixed security issue.
>>>>
>>>> DRM: remove encoder porinter from specific connector.
>>>>
>>>>      samsung connector doesn't need to have generic encoder.
>>>>
>>>> DRM: code clean and add exception codes.
>>>>
>>>> Signed-off-by: Inki Dae<inki.dae@samsung.com>
>>>> Signed-off-by: Joonyoung Shim<jy0922.shim@samsung.com>
>>>> Signed-off-by: SeungWoo Kim<sw0312.kim@samsung.com>
>>>> Signed-off-by: kyungmin.park<kyungmin.park@samsung.com>
>>>> ---
>>>>
>>>> +static struct drm_ioctl_desc samsung_ioctls[] = {
>>>> +       DRM_IOCTL_DEF_DRV(SAMSUNG_GEM_CREATE,
>>>> samsung_drm_gem_create_ioctl,
>>>> +                       DRM_UNLOCKED | DRM_AUTH),
>>>>
>>> Hi!
>>>
>>> With reference my previous security comment.
>>>
>>> Let's say you have a compromised video player running as a DRM 
>>> client, that
>>> tries to repeatedly allocate huge GEM buffers...
>>>
>>> What will happen when all DMA memory is exhausted? Will this cause 
>>> other
>>> device drivers to see an OOM, or only DRM?
>>>
>>> The old DRI model basically allowed any authorized DRI client to 
>>> exhaust
>>> video ram or AGP memory, but never system memory. Newer DRI drivers
>>> typically only allow DRI masters to do that.
>>> as
>>>
>>> I don't think an authorized DRI client should be able to easily exhaust
>>> resources (DMA memory) used by other device drivers causing them to 
>>> fail.
>> I'm not entirely sure what else can be done, other than have a
>> threshold on max MB allocatable of buffer memory..
>
> Yes, I think that's what needs to be done, and that threshold should 
> be low enough to keep other device drivers running in the worst 
> allocation case.
>
>> In the samsung driver case, he is only allocating scanout memory from
>> CMA, so the limit will be the CMA region size.. beyond that you can't
>> get physically contiguous memory.  So I think this driver is safe.
>
> It's not really what well-behaved user-space drivers do that should be 
> a concern, but what compromized application *might* do that is a concern.

Hmm. I might have missed your point here. If the buffer allocation ioctl 
only allows allocating CMA memory, then I agree the driver fits the old 
DRI security model, as long as no other devices on the platform will 
ever use CMA.

But in that case, there really should be a way for the driver to say 
"Hey, all CMA memory on this system is mine", in the same way 
traditional video drivers can claim the VRAM PCI resource.

This is to avoid the possibility that future drivers that need CMA will 
be vulnerable to DOS-attacks from ill-behaved DRI clients.

/Thomas

[PATCH v4] DRM: add DRM Driver for Samsung SoC EXYNOS4210.

[Inki Dae]

Hi Thomas.

> -----Original Message-----
> From: Thomas Hellstrom [mailto:thomas at shipmail.org]
> Sent: Monday, September 12, 2011 3:32 PM
> To: Rob Clark
> Cc: Inki Dae; kyungmin.park at samsung.com; sw0312.kim at samsung.com; linux-
> arm-kernel at lists.infradead.org; dri-devel at lists.freedesktop.org
> Subject: Re: [PATCH v4] DRM: add DRM Driver for Samsung SoC EXYNOS4210.
> 
> On 09/11/2011 11:26 PM, Thomas Hellstrom wrote:
> > On 09/10/2011 07:31 PM, Rob Clark wrote:
> >> On Sat, Sep 10, 2011 at 9:04 AM, Thomas
> >> Hellstrom<thomas@shipmail.org>  wrote:
> >>> On 09/09/2011 01:38 PM, Inki Dae wrote:
> >>>> This patch is a DRM Driver for Samsung SoC Exynos4210 and now enables
> >>>> only FIMD yet but we will add HDMI support also in the future.
> >>>>
> >>>> from now on, I will remove RFC prefix because I think we have got
> >>>> comments
> >>>> enough.
> >>>>
> >>>> this patch is based on git repository below:
> >>>> git://git.kernel.org/pub/scm/linux/kernel/git/airlied/drm-2.6.git,
> >>>> branch name: drm-next
> >>>> commit-id: bcc65fd8e929a9d9d34d814d6efc1d2793546922
> >>>>
> >>>> you can refer to our working repository below:
> >>>> http://git.infradead.org/users/kmpark/linux-2.6-samsung
> >>>> branch name: samsung-drm
> >>>>
> >>>> We tried to re-use lowlevel codes of the FIMD driver(s3c-fb.c
> >>>> based on Linux framebuffer) but couldn't so because lowlevel codes
> >>>> of s3c-fb.c are included internally and so FIMD module of this
> >>>> driver has
> >>>> its own lowlevel codes.
> >>>>
> >>>> We used GEM framework for buffer management and DMA APIs(dma_alloc_*)
> >>>> for buffer allocation. by using DMA API, we could use CMA later.
> >>>>
> >>>> Refer to this link for CMA(Continuous Memory Allocator):
> >>>> http://lkml.org/lkml/2011/7/20/45
> >>>>
> >>>> this driver supports only physically continuous memory(non-iommu).
> >>>>
> >>>> Links to previous versions of the patchset:
> >>>> v1:<    https://lwn.net/Articles/454380/>
> >>>> v2:<    http://www.spinics.net/lists/kernel/msg1224275.html>
> >>>> v3:<    http://www.gossamer-threads.com/lists/linux/kernel/1423684>
> >>>>
> >>>> Changelog v2:
> >>>> DRM: add DRM_IOCTL_SAMSUNG_GEM_MMAP ioctl command.
> >>>>
> >>>>      this feature maps user address space to physical memory region
> >>>>      once user application requests DRM_IOCTL_SAMSUNG_GEM_MMAP ioctl.
> >>>>
> >>>> DRM: code clean and add exception codes.
> >>>>
> >>>> Changelog v3:
> >>>> DRM: Support multiple irq.
> >>>>
> >>>>      FIMD and HDMI have their own irq handler but DRM Framework can
> >>>> regiter
> >>>>      only one irq handler this patch supports mutiple irq for
> >>>> Samsung SoC.
> >>>>
> >>>> DRM: Consider modularization.
> >>>>
> >>>>      each DRM, FIMD could be built as a module.
> >>>>
> >>>> DRM: Have indenpendent crtc object.
> >>>>
> >>>>      crtc isn't specific to SoC Platform so this patch gets a crtc
> >>>>      to be used as common object.
> >>>>      created crtc could be attached to any encoder object.
> >>>>
> >>>> DRM: code clean and add exception codes.
> >>>>
> >>>> Changelog v4:
> >>>> DRM: remove is_defult from samsung_fb.
> >>>>
> >>>>      is_default isn't used for default framebuffer.
> >>>>
> >>>> DRM: code refactoring to fimd module.
> >>>>      this patch is be considered with multiple display objects and
> >>>>      would use its own request_irq() to register a irq handler
> >>>> instead of
> >>>>      drm framework's one.
> >>>>
> >>>> DRM: remove find_samsung_drm_gem_object()
> >>>>
> >>>> DRM: move kernel private data structures and definitions to driver
> >>>> folder.
> >>>>
> >>>>      samsung_drm.h would contain only public information for
userspace
> >>>>      ioctl interface.
> >>>>
> >>>> DRM: code refactoring to gem modules.
> >>>>      buffer module isn't dependent of gem module anymore.
> >>>>
> >>>> DRM: fixed security issue.
> >>>>
> >>>> DRM: remove encoder porinter from specific connector.
> >>>>
> >>>>      samsung connector doesn't need to have generic encoder.
> >>>>
> >>>> DRM: code clean and add exception codes.
> >>>>
> >>>> Signed-off-by: Inki Dae<inki.dae@samsung.com>
> >>>> Signed-off-by: Joonyoung Shim<jy0922.shim@samsung.com>
> >>>> Signed-off-by: SeungWoo Kim<sw0312.kim@samsung.com>
> >>>> Signed-off-by: kyungmin.park<kyungmin.park@samsung.com>
> >>>> ---
> >>>>
> >>>> +static struct drm_ioctl_desc samsung_ioctls[] = {
> >>>> +       DRM_IOCTL_DEF_DRV(SAMSUNG_GEM_CREATE,
> >>>> samsung_drm_gem_create_ioctl,
> >>>> +                       DRM_UNLOCKED | DRM_AUTH),
> >>>>
> >>> Hi!
> >>>
> >>> With reference my previous security comment.
> >>>
> >>> Let's say you have a compromised video player running as a DRM
> >>> client, that
> >>> tries to repeatedly allocate huge GEM buffers...
> >>>
> >>> What will happen when all DMA memory is exhausted? Will this cause
> >>> other
> >>> device drivers to see an OOM, or only DRM?
> >>>
> >>> The old DRI model basically allowed any authorized DRI client to
> >>> exhaust
> >>> video ram or AGP memory, but never system memory. Newer DRI drivers
> >>> typically only allow DRI masters to do that.
> >>> as
> >>>
> >>> I don't think an authorized DRI client should be able to easily
> exhaust
> >>> resources (DMA memory) used by other device drivers causing them to
> >>> fail.
> >> I'm not entirely sure what else can be done, other than have a
> >> threshold on max MB allocatable of buffer memory..
> >
> > Yes, I think that's what needs to be done, and that threshold should
> > be low enough to keep other device drivers running in the worst
> > allocation case.
> >
> >> In the samsung driver case, he is only allocating scanout memory from
> >> CMA, so the limit will be the CMA region size.. beyond that you can't
> >> get physically contiguous memory.  So I think this driver is safe.
> >
> > It's not really what well-behaved user-space drivers do that should be
> > a concern, but what compromized application *might* do that is a
concern.
> 
> Hmm. I might have missed your point here. If the buffer allocation ioctl
> only allows allocating CMA memory, then I agree the driver fits the old
> DRI security model, as long as no other devices on the platform will
> ever use CMA.
> 
> But in that case, there really should be a way for the driver to say
> "Hey, all CMA memory on this system is mine", in the same way
> traditional video drivers can claim the VRAM PCI resource.
> 

CMA could reserve memory region for a specific driver so DRM Client could
request memory allocation from only the region.

> This is to avoid the possibility that future drivers that need CMA will
> be vulnerable to DOS-attacks from ill-behaved DRI clients.
> 

Thomas, if any application has root authority for ill-purpose then isn't it
possible to be vulnerable to DOS-attacks? I think DRM_AUTH means root
authority. I know DRM Framework gives any root application DRM_AUTH
authority for compatibility.

> /Thomas
> 

[PATCH v4] DRM: add DRM Driver for Samsung SoC EXYNOS4210.

[Rob Clark]

On Tue, Sep 13, 2011 at 9:03 PM, Inki Dae <inki.dae@samsung.com> wrote:
> Hi Thomas.
>
>> -----Original Message-----
>> From: Thomas Hellstrom [mailto:thomas at shipmail.org]
>> Sent: Monday, September 12, 2011 3:32 PM
>> To: Rob Clark
>> Cc: Inki Dae; kyungmin.park at samsung.com; sw0312.kim at samsung.com; linux-
>> arm-kernel at lists.infradead.org; dri-devel at lists.freedesktop.org
>> Subject: Re: [PATCH v4] DRM: add DRM Driver for Samsung SoC EXYNOS4210.
>>
>> On 09/11/2011 11:26 PM, Thomas Hellstrom wrote:
>> > On 09/10/2011 07:31 PM, Rob Clark wrote:
>> >> On Sat, Sep 10, 2011 at 9:04 AM, Thomas
>> >> Hellstrom<thomas@shipmail.org> ?wrote:
>> >>> On 09/09/2011 01:38 PM, Inki Dae wrote:
>> >>>> This patch is a DRM Driver for Samsung SoC Exynos4210 and now enables
>> >>>> only FIMD yet but we will add HDMI support also in the future.
>> >>>>
>> >>>> from now on, I will remove RFC prefix because I think we have got
>> >>>> comments
>> >>>> enough.
>> >>>>
>> >>>> this patch is based on git repository below:
>> >>>> git://git.kernel.org/pub/scm/linux/kernel/git/airlied/drm-2.6.git,
>> >>>> branch name: drm-next
>> >>>> commit-id: bcc65fd8e929a9d9d34d814d6efc1d2793546922
>> >>>>
>> >>>> you can refer to our working repository below:
>> >>>> http://git.infradead.org/users/kmpark/linux-2.6-samsung
>> >>>> branch name: samsung-drm
>> >>>>
>> >>>> We tried to re-use lowlevel codes of the FIMD driver(s3c-fb.c
>> >>>> based on Linux framebuffer) but couldn't so because lowlevel codes
>> >>>> of s3c-fb.c are included internally and so FIMD module of this
>> >>>> driver has
>> >>>> its own lowlevel codes.
>> >>>>
>> >>>> We used GEM framework for buffer management and DMA APIs(dma_alloc_*)
>> >>>> for buffer allocation. by using DMA API, we could use CMA later.
>> >>>>
>> >>>> Refer to this link for CMA(Continuous Memory Allocator):
>> >>>> http://lkml.org/lkml/2011/7/20/45
>> >>>>
>> >>>> this driver supports only physically continuous memory(non-iommu).
>> >>>>
>> >>>> Links to previous versions of the patchset:
>> >>>> v1:< ? ?https://lwn.net/Articles/454380/>
>> >>>> v2:< ? ?http://www.spinics.net/lists/kernel/msg1224275.html>
>> >>>> v3:< ? ?http://www.gossamer-threads.com/lists/linux/kernel/1423684>
>> >>>>
>> >>>> Changelog v2:
>> >>>> DRM: add DRM_IOCTL_SAMSUNG_GEM_MMAP ioctl command.
>> >>>>
>> >>>> ? ? ?this feature maps user address space to physical memory region
>> >>>> ? ? ?once user application requests DRM_IOCTL_SAMSUNG_GEM_MMAP ioctl.
>> >>>>
>> >>>> DRM: code clean and add exception codes.
>> >>>>
>> >>>> Changelog v3:
>> >>>> DRM: Support multiple irq.
>> >>>>
>> >>>> ? ? ?FIMD and HDMI have their own irq handler but DRM Framework can
>> >>>> regiter
>> >>>> ? ? ?only one irq handler this patch supports mutiple irq for
>> >>>> Samsung SoC.
>> >>>>
>> >>>> DRM: Consider modularization.
>> >>>>
>> >>>> ? ? ?each DRM, FIMD could be built as a module.
>> >>>>
>> >>>> DRM: Have indenpendent crtc object.
>> >>>>
>> >>>> ? ? ?crtc isn't specific to SoC Platform so this patch gets a crtc
>> >>>> ? ? ?to be used as common object.
>> >>>> ? ? ?created crtc could be attached to any encoder object.
>> >>>>
>> >>>> DRM: code clean and add exception codes.
>> >>>>
>> >>>> Changelog v4:
>> >>>> DRM: remove is_defult from samsung_fb.
>> >>>>
>> >>>> ? ? ?is_default isn't used for default framebuffer.
>> >>>>
>> >>>> DRM: code refactoring to fimd module.
>> >>>> ? ? ?this patch is be considered with multiple display objects and
>> >>>> ? ? ?would use its own request_irq() to register a irq handler
>> >>>> instead of
>> >>>> ? ? ?drm framework's one.
>> >>>>
>> >>>> DRM: remove find_samsung_drm_gem_object()
>> >>>>
>> >>>> DRM: move kernel private data structures and definitions to driver
>> >>>> folder.
>> >>>>
>> >>>> ? ? ?samsung_drm.h would contain only public information for
> userspace
>> >>>> ? ? ?ioctl interface.
>> >>>>
>> >>>> DRM: code refactoring to gem modules.
>> >>>> ? ? ?buffer module isn't dependent of gem module anymore.
>> >>>>
>> >>>> DRM: fixed security issue.
>> >>>>
>> >>>> DRM: remove encoder porinter from specific connector.
>> >>>>
>> >>>> ? ? ?samsung connector doesn't need to have generic encoder.
>> >>>>
>> >>>> DRM: code clean and add exception codes.
>> >>>>
>> >>>> Signed-off-by: Inki Dae<inki.dae@samsung.com>
>> >>>> Signed-off-by: Joonyoung Shim<jy0922.shim@samsung.com>
>> >>>> Signed-off-by: SeungWoo Kim<sw0312.kim@samsung.com>
>> >>>> Signed-off-by: kyungmin.park<kyungmin.park@samsung.com>
>> >>>> ---
>> >>>>
>> >>>> +static struct drm_ioctl_desc samsung_ioctls[] = {
>> >>>> + ? ? ? DRM_IOCTL_DEF_DRV(SAMSUNG_GEM_CREATE,
>> >>>> samsung_drm_gem_create_ioctl,
>> >>>> + ? ? ? ? ? ? ? ? ? ? ? DRM_UNLOCKED | DRM_AUTH),
>> >>>>
>> >>> Hi!
>> >>>
>> >>> With reference my previous security comment.
>> >>>
>> >>> Let's say you have a compromised video player running as a DRM
>> >>> client, that
>> >>> tries to repeatedly allocate huge GEM buffers...
>> >>>
>> >>> What will happen when all DMA memory is exhausted? Will this cause
>> >>> other
>> >>> device drivers to see an OOM, or only DRM?
>> >>>
>> >>> The old DRI model basically allowed any authorized DRI client to
>> >>> exhaust
>> >>> video ram or AGP memory, but never system memory. Newer DRI drivers
>> >>> typically only allow DRI masters to do that.
>> >>> as
>> >>>
>> >>> I don't think an authorized DRI client should be able to easily
>> exhaust
>> >>> resources (DMA memory) used by other device drivers causing them to
>> >>> fail.
>> >> I'm not entirely sure what else can be done, other than have a
>> >> threshold on max MB allocatable of buffer memory..
>> >
>> > Yes, I think that's what needs to be done, and that threshold should
>> > be low enough to keep other device drivers running in the worst
>> > allocation case.
>> >
>> >> In the samsung driver case, he is only allocating scanout memory from
>> >> CMA, so the limit will be the CMA region size.. beyond that you can't
>> >> get physically contiguous memory. ?So I think this driver is safe.
>> >
>> > It's not really what well-behaved user-space drivers do that should be
>> > a concern, but what compromized application *might* do that is a
> concern.
>>
>> Hmm. I might have missed your point here. If the buffer allocation ioctl
>> only allows allocating CMA memory, then I agree the driver fits the old
>> DRI security model, as long as no other devices on the platform will
>> ever use CMA.
>>
>> But in that case, there really should be a way for the driver to say
>> "Hey, all CMA memory on this system is mine", in the same way
>> traditional video drivers can claim the VRAM PCI resource.
>>
>
> CMA could reserve memory region for a specific driver so DRM Client could
> request memory allocation from only the region.
>
>> This is to avoid the possibility that future drivers that need CMA will
>> be vulnerable to DOS-attacks from ill-behaved DRI clients.
>>
>
> Thomas, if any application has root authority for ill-purpose then isn't it
> possible to be vulnerable to DOS-attacks? I think DRM_AUTH means root
> authority. I know DRM Framework gives any root application DRM_AUTH
> authority for compatibility.

DRM_AUTH just means that the client has authenticated w/ X11 (meaning
that it has permission to connect to x server)..

But I think that since memory allocation is limited to the size of the
CMA region, that this puts a reasonable cap of the memory that can be
allocated by the client.  If this is a problem, it certainly isn't the
worst problem.  You could still limit via file permissions the users
that can open the DRM device file, so it is really no worse than other
devices like v4l2..

BR,
-R


>> /Thomas
>>
>
>
>

[PATCH v4] DRM: add DRM Driver for Samsung SoC EXYNOS4210.

[Inki Dae]

> -----Original Message-----
> From: Rob Clark [mailto:robdclark at gmail.com]
> Sent: Wednesday, September 14, 2011 11:26 AM
> To: Inki Dae
> Cc: Thomas Hellstrom; kyungmin.park at samsung.com; sw0312.kim at samsung.com;
> linux-arm-kernel at lists.infradead.org; dri-devel at lists.freedesktop.org
> Subject: Re: [PATCH v4] DRM: add DRM Driver for Samsung SoC EXYNOS4210.
> 
> On Tue, Sep 13, 2011 at 9:03 PM, Inki Dae <inki.dae@samsung.com> wrote:
> > Hi Thomas.
> >
> >> -----Original Message-----
> >> From: Thomas Hellstrom [mailto:thomas at shipmail.org]
> >> Sent: Monday, September 12, 2011 3:32 PM
> >> To: Rob Clark
> >> Cc: Inki Dae; kyungmin.park at samsung.com; sw0312.kim at samsung.com; linux-
> >> arm-kernel at lists.infradead.org; dri-devel at lists.freedesktop.org
> >> Subject: Re: [PATCH v4] DRM: add DRM Driver for Samsung SoC EXYNOS4210.
> >>
> >> On 09/11/2011 11:26 PM, Thomas Hellstrom wrote:
> >> > On 09/10/2011 07:31 PM, Rob Clark wrote:
> >> >> On Sat, Sep 10, 2011 at 9:04 AM, Thomas
> >> >> Hellstrom<thomas@shipmail.org> ?wrote:
> >> >>> On 09/09/2011 01:38 PM, Inki Dae wrote:
> >> >>>> This patch is a DRM Driver for Samsung SoC Exynos4210 and now
> enables
> >> >>>> only FIMD yet but we will add HDMI support also in the future.
> >> >>>>
> >> >>>> from now on, I will remove RFC prefix because I think we have got
> >> >>>> comments
> >> >>>> enough.
> >> >>>>
> >> >>>> this patch is based on git repository below:
> >> >>>> git://git.kernel.org/pub/scm/linux/kernel/git/airlied/drm-2.6.git,
> >> >>>> branch name: drm-next
> >> >>>> commit-id: bcc65fd8e929a9d9d34d814d6efc1d2793546922
> >> >>>>
> >> >>>> you can refer to our working repository below:
> >> >>>> http://git.infradead.org/users/kmpark/linux-2.6-samsung
> >> >>>> branch name: samsung-drm
> >> >>>>
> >> >>>> We tried to re-use lowlevel codes of the FIMD driver(s3c-fb.c
> >> >>>> based on Linux framebuffer) but couldn't so because lowlevel codes
> >> >>>> of s3c-fb.c are included internally and so FIMD module of this
> >> >>>> driver has
> >> >>>> its own lowlevel codes.
> >> >>>>
> >> >>>> We used GEM framework for buffer management and DMA
> APIs(dma_alloc_*)
> >> >>>> for buffer allocation. by using DMA API, we could use CMA later.
> >> >>>>
> >> >>>> Refer to this link for CMA(Continuous Memory Allocator):
> >> >>>> http://lkml.org/lkml/2011/7/20/45
> >> >>>>
> >> >>>> this driver supports only physically continuous memory(non-iommu).
> >> >>>>
> >> >>>> Links to previous versions of the patchset:
> >> >>>> v1:< ? ?https://lwn.net/Articles/454380/>
> >> >>>> v2:< ? ?http://www.spinics.net/lists/kernel/msg1224275.html>
> >> >>>> v3:< ? ?http://www.gossamer-
> threads.com/lists/linux/kernel/1423684>
> >> >>>>
> >> >>>> Changelog v2:
> >> >>>> DRM: add DRM_IOCTL_SAMSUNG_GEM_MMAP ioctl command.
> >> >>>>
> >> >>>> ? ? ?this feature maps user address space to physical memory
> region
> >> >>>> ? ? ?once user application requests DRM_IOCTL_SAMSUNG_GEM_MMAP
> ioctl.
> >> >>>>
> >> >>>> DRM: code clean and add exception codes.
> >> >>>>
> >> >>>> Changelog v3:
> >> >>>> DRM: Support multiple irq.
> >> >>>>
> >> >>>> ? ? ?FIMD and HDMI have their own irq handler but DRM Framework
> can
> >> >>>> regiter
> >> >>>> ? ? ?only one irq handler this patch supports mutiple irq for
> >> >>>> Samsung SoC.
> >> >>>>
> >> >>>> DRM: Consider modularization.
> >> >>>>
> >> >>>> ? ? ?each DRM, FIMD could be built as a module.
> >> >>>>
> >> >>>> DRM: Have indenpendent crtc object.
> >> >>>>
> >> >>>> ? ? ?crtc isn't specific to SoC Platform so this patch gets a crtc
> >> >>>> ? ? ?to be used as common object.
> >> >>>> ? ? ?created crtc could be attached to any encoder object.
> >> >>>>
> >> >>>> DRM: code clean and add exception codes.
> >> >>>>
> >> >>>> Changelog v4:
> >> >>>> DRM: remove is_defult from samsung_fb.
> >> >>>>
> >> >>>> ? ? ?is_default isn't used for default framebuffer.
> >> >>>>
> >> >>>> DRM: code refactoring to fimd module.
> >> >>>> ? ? ?this patch is be considered with multiple display objects and
> >> >>>> ? ? ?would use its own request_irq() to register a irq handler
> >> >>>> instead of
> >> >>>> ? ? ?drm framework's one.
> >> >>>>
> >> >>>> DRM: remove find_samsung_drm_gem_object()
> >> >>>>
> >> >>>> DRM: move kernel private data structures and definitions to driver
> >> >>>> folder.
> >> >>>>
> >> >>>> ? ? ?samsung_drm.h would contain only public information for
> > userspace
> >> >>>> ? ? ?ioctl interface.
> >> >>>>
> >> >>>> DRM: code refactoring to gem modules.
> >> >>>> ? ? ?buffer module isn't dependent of gem module anymore.
> >> >>>>
> >> >>>> DRM: fixed security issue.
> >> >>>>
> >> >>>> DRM: remove encoder porinter from specific connector.
> >> >>>>
> >> >>>> ? ? ?samsung connector doesn't need to have generic encoder.
> >> >>>>
> >> >>>> DRM: code clean and add exception codes.
> >> >>>>
> >> >>>> Signed-off-by: Inki Dae<inki.dae@samsung.com>
> >> >>>> Signed-off-by: Joonyoung Shim<jy0922.shim@samsung.com>
> >> >>>> Signed-off-by: SeungWoo Kim<sw0312.kim@samsung.com>
> >> >>>> Signed-off-by: kyungmin.park<kyungmin.park@samsung.com>
> >> >>>> ---
> >> >>>>
> >> >>>> +static struct drm_ioctl_desc samsung_ioctls[] = {
> >> >>>> + ? ? ? DRM_IOCTL_DEF_DRV(SAMSUNG_GEM_CREATE,
> >> >>>> samsung_drm_gem_create_ioctl,
> >> >>>> + ? ? ? ? ? ? ? ? ? ? ? DRM_UNLOCKED | DRM_AUTH),
> >> >>>>
> >> >>> Hi!
> >> >>>
> >> >>> With reference my previous security comment.
> >> >>>
> >> >>> Let's say you have a compromised video player running as a DRM
> >> >>> client, that
> >> >>> tries to repeatedly allocate huge GEM buffers...
> >> >>>
> >> >>> What will happen when all DMA memory is exhausted? Will this cause
> >> >>> other
> >> >>> device drivers to see an OOM, or only DRM?
> >> >>>
> >> >>> The old DRI model basically allowed any authorized DRI client to
> >> >>> exhaust
> >> >>> video ram or AGP memory, but never system memory. Newer DRI drivers
> >> >>> typically only allow DRI masters to do that.
> >> >>> as
> >> >>>
> >> >>> I don't think an authorized DRI client should be able to easily
> >> exhaust
> >> >>> resources (DMA memory) used by other device drivers causing them to
> >> >>> fail.
> >> >> I'm not entirely sure what else can be done, other than have a
> >> >> threshold on max MB allocatable of buffer memory..
> >> >
> >> > Yes, I think that's what needs to be done, and that threshold should
> >> > be low enough to keep other device drivers running in the worst
> >> > allocation case.
> >> >
> >> >> In the samsung driver case, he is only allocating scanout memory
> from
> >> >> CMA, so the limit will be the CMA region size.. beyond that you
> can't
> >> >> get physically contiguous memory. ?So I think this driver is safe.
> >> >
> >> > It's not really what well-behaved user-space drivers do that should
> be
> >> > a concern, but what compromized application *might* do that is a
> > concern.
> >>
> >> Hmm. I might have missed your point here. If the buffer allocation
> ioctl
> >> only allows allocating CMA memory, then I agree the driver fits the old
> >> DRI security model, as long as no other devices on the platform will
> >> ever use CMA.
> >>
> >> But in that case, there really should be a way for the driver to say
> >> "Hey, all CMA memory on this system is mine", in the same way
> >> traditional video drivers can claim the VRAM PCI resource.
> >>
> >
> > CMA could reserve memory region for a specific driver so DRM Client
> could
> > request memory allocation from only the region.
> >
> >> This is to avoid the possibility that future drivers that need CMA will
> >> be vulnerable to DOS-attacks from ill-behaved DRI clients.
> >>
> >
> > Thomas, if any application has root authority for ill-purpose then isn't
> it
> > possible to be vulnerable to DOS-attacks? I think DRM_AUTH means root
> > authority. I know DRM Framework gives any root application DRM_AUTH
> > authority for compatibility.
> 
> DRM_AUTH just means that the client has authenticated w/ X11 (meaning
> that it has permission to connect to x server)..
> 

Yes, I understood so. but see drm_open_helper() of drm_fops.c file please.
in this function, you can see a line below.
/* for compatibility root is always authenticated */
priv->authenticated = capable(CAP_SYS_ADMIN)

I think the code above says that any application with root permission is
authenticated.


> But I think that since memory allocation is limited to the size of the
> CMA region, that this puts a reasonable cap of the memory that can be
> allocated by the client.  If this is a problem, it certainly isn't the
> worst problem.  You could still limit via file permissions the users
> that can open the DRM device file, so it is really no worse than other
> devices like v4l2..
> 
> BR,
> -R
> 
> 
> >> /Thomas
> >>
> >
> >
> >

[PATCH v4] DRM: add DRM Driver for Samsung SoC EXYNOS4210.

[Thomas Hellstrom]

On 09/14/2011 07:55 AM, Inki Dae wrote:
>
>    
>> -----Original Message-----
>> From: Rob Clark [mailto:robdclark at gmail.com]
>> Sent: Wednesday, September 14, 2011 11:26 AM
>> To: Inki Dae
>> Cc: Thomas Hellstrom; kyungmin.park at samsung.com; sw0312.kim at samsung.com;
>> linux-arm-kernel at lists.infradead.org; dri-devel at lists.freedesktop.org
>> Subject: Re: [PATCH v4] DRM: add DRM Driver for Samsung SoC EXYNOS4210.
>>
>> On Tue, Sep 13, 2011 at 9:03 PM, Inki Dae<inki.dae@samsung.com>  wrote:
>>      
>>> Hi Thomas.
>>>
>>>        
>>>> -----Original Message-----
>>>> From: Thomas Hellstrom [mailto:thomas at shipmail.org]
>>>> Sent: Monday, September 12, 2011 3:32 PM
>>>> To: Rob Clark
>>>> Cc: Inki Dae; kyungmin.park at samsung.com; sw0312.kim at samsung.com; linux-
>>>> arm-kernel at lists.infradead.org; dri-devel at lists.freedesktop.org
>>>> Subject: Re: [PATCH v4] DRM: add DRM Driver for Samsung SoC EXYNOS4210.
>>>>
>>>> On 09/11/2011 11:26 PM, Thomas Hellstrom wrote:
>>>>          
>>>>> On 09/10/2011 07:31 PM, Rob Clark wrote:
>>>>>            
>>>>>> On Sat, Sep 10, 2011 at 9:04 AM, Thomas
>>>>>> Hellstrom<thomas@shipmail.org>    wrote:
>>>>>>              
>>>>>>> On 09/09/2011 01:38 PM, Inki Dae wrote:
>>>>>>>                
>>>>>>>> This patch is a DRM Driver for Samsung SoC Exynos4210 and now
>>>>>>>>                  
>> enables
>>      
>>>>>>>> only FIMD yet but we will add HDMI support also in the future.
>>>>>>>>
>>>>>>>> from now on, I will remove RFC prefix because I think we have got
>>>>>>>> comments
>>>>>>>> enough.
>>>>>>>>
>>>>>>>> this patch is based on git repository below:
>>>>>>>> git://git.kernel.org/pub/scm/linux/kernel/git/airlied/drm-2.6.git,
>>>>>>>> branch name: drm-next
>>>>>>>> commit-id: bcc65fd8e929a9d9d34d814d6efc1d2793546922
>>>>>>>>
>>>>>>>> you can refer to our working repository below:
>>>>>>>> http://git.infradead.org/users/kmpark/linux-2.6-samsung
>>>>>>>> branch name: samsung-drm
>>>>>>>>
>>>>>>>> We tried to re-use lowlevel codes of the FIMD driver(s3c-fb.c
>>>>>>>> based on Linux framebuffer) but couldn't so because lowlevel codes
>>>>>>>> of s3c-fb.c are included internally and so FIMD module of this
>>>>>>>> driver has
>>>>>>>> its own lowlevel codes.
>>>>>>>>
>>>>>>>> We used GEM framework for buffer management and DMA
>>>>>>>>                  
>> APIs(dma_alloc_*)
>>      
>>>>>>>> for buffer allocation. by using DMA API, we could use CMA later.
>>>>>>>>
>>>>>>>> Refer to this link for CMA(Continuous Memory Allocator):
>>>>>>>> http://lkml.org/lkml/2011/7/20/45
>>>>>>>>
>>>>>>>> this driver supports only physically continuous memory(non-iommu).
>>>>>>>>
>>>>>>>> Links to previous versions of the patchset:
>>>>>>>> v1:<      https://lwn.net/Articles/454380/>
>>>>>>>> v2:<      http://www.spinics.net/lists/kernel/msg1224275.html>
>>>>>>>> v3:<      http://www.gossamer-
>>>>>>>>                  
>> threads.com/lists/linux/kernel/1423684>
>>      
>>>>>>>> Changelog v2:
>>>>>>>> DRM: add DRM_IOCTL_SAMSUNG_GEM_MMAP ioctl command.
>>>>>>>>
>>>>>>>>       this feature maps user address space to physical memory
>>>>>>>>                  
>> region
>>      
>>>>>>>>       once user application requests DRM_IOCTL_SAMSUNG_GEM_MMAP
>>>>>>>>                  
>> ioctl.
>>      
>>>>>>>> DRM: code clean and add exception codes.
>>>>>>>>
>>>>>>>> Changelog v3:
>>>>>>>> DRM: Support multiple irq.
>>>>>>>>
>>>>>>>>       FIMD and HDMI have their own irq handler but DRM Framework
>>>>>>>>                  
>> can
>>      
>>>>>>>> regiter
>>>>>>>>       only one irq handler this patch supports mutiple irq for
>>>>>>>> Samsung SoC.
>>>>>>>>
>>>>>>>> DRM: Consider modularization.
>>>>>>>>
>>>>>>>>       each DRM, FIMD could be built as a module.
>>>>>>>>
>>>>>>>> DRM: Have indenpendent crtc object.
>>>>>>>>
>>>>>>>>       crtc isn't specific to SoC Platform so this patch gets a crtc
>>>>>>>>       to be used as common object.
>>>>>>>>       created crtc could be attached to any encoder object.
>>>>>>>>
>>>>>>>> DRM: code clean and add exception codes.
>>>>>>>>
>>>>>>>> Changelog v4:
>>>>>>>> DRM: remove is_defult from samsung_fb.
>>>>>>>>
>>>>>>>>       is_default isn't used for default framebuffer.
>>>>>>>>
>>>>>>>> DRM: code refactoring to fimd module.
>>>>>>>>       this patch is be considered with multiple display objects and
>>>>>>>>       would use its own request_irq() to register a irq handler
>>>>>>>> instead of
>>>>>>>>       drm framework's one.
>>>>>>>>
>>>>>>>> DRM: remove find_samsung_drm_gem_object()
>>>>>>>>
>>>>>>>> DRM: move kernel private data structures and definitions to driver
>>>>>>>> folder.
>>>>>>>>
>>>>>>>>       samsung_drm.h would contain only public information for
>>>>>>>>                  
>>> userspace
>>>        
>>>>>>>>       ioctl interface.
>>>>>>>>
>>>>>>>> DRM: code refactoring to gem modules.
>>>>>>>>       buffer module isn't dependent of gem module anymore.
>>>>>>>>
>>>>>>>> DRM: fixed security issue.
>>>>>>>>
>>>>>>>> DRM: remove encoder porinter from specific connector.
>>>>>>>>
>>>>>>>>       samsung connector doesn't need to have generic encoder.
>>>>>>>>
>>>>>>>> DRM: code clean and add exception codes.
>>>>>>>>
>>>>>>>> Signed-off-by: Inki Dae<inki.dae@samsung.com>
>>>>>>>> Signed-off-by: Joonyoung Shim<jy0922.shim@samsung.com>
>>>>>>>> Signed-off-by: SeungWoo Kim<sw0312.kim@samsung.com>
>>>>>>>> Signed-off-by: kyungmin.park<kyungmin.park@samsung.com>
>>>>>>>> ---
>>>>>>>>
>>>>>>>> +static struct drm_ioctl_desc samsung_ioctls[] = {
>>>>>>>> +       DRM_IOCTL_DEF_DRV(SAMSUNG_GEM_CREATE,
>>>>>>>> samsung_drm_gem_create_ioctl,
>>>>>>>> +                       DRM_UNLOCKED | DRM_AUTH),
>>>>>>>>
>>>>>>>>                  
>>>>>>> Hi!
>>>>>>>
>>>>>>> With reference my previous security comment.
>>>>>>>
>>>>>>> Let's say you have a compromised video player running as a DRM
>>>>>>> client, that
>>>>>>> tries to repeatedly allocate huge GEM buffers...
>>>>>>>
>>>>>>> What will happen when all DMA memory is exhausted? Will this cause
>>>>>>> other
>>>>>>> device drivers to see an OOM, or only DRM?
>>>>>>>
>>>>>>> The old DRI model basically allowed any authorized DRI client to
>>>>>>> exhaust
>>>>>>> video ram or AGP memory, but never system memory. Newer DRI drivers
>>>>>>> typically only allow DRI masters to do that.
>>>>>>> as
>>>>>>>
>>>>>>> I don't think an authorized DRI client should be able to easily
>>>>>>>                
>>>> exhaust
>>>>          
>>>>>>> resources (DMA memory) used by other device drivers causing them to
>>>>>>> fail.
>>>>>>>                
>>>>>> I'm not entirely sure what else can be done, other than have a
>>>>>> threshold on max MB allocatable of buffer memory..
>>>>>>              
>>>>> Yes, I think that's what needs to be done, and that threshold should
>>>>> be low enough to keep other device drivers running in the worst
>>>>> allocation case.
>>>>>
>>>>>            
>>>>>> In the samsung driver case, he is only allocating scanout memory
>>>>>>              
>> from
>>      
>>>>>> CMA, so the limit will be the CMA region size.. beyond that you
>>>>>>              
>> can't
>>      
>>>>>> get physically contiguous memory.  So I think this driver is safe.
>>>>>>              
>>>>> It's not really what well-behaved user-space drivers do that should
>>>>>            
>> be
>>      
>>>>> a concern, but what compromized application *might* do that is a
>>>>>            
>>> concern.
>>>        
>>>> Hmm. I might have missed your point here. If the buffer allocation
>>>>          
>> ioctl
>>      
>>>> only allows allocating CMA memory, then I agree the driver fits the old
>>>> DRI security model, as long as no other devices on the platform will
>>>> ever use CMA.
>>>>
>>>> But in that case, there really should be a way for the driver to say
>>>> "Hey, all CMA memory on this system is mine", in the same way
>>>> traditional video drivers can claim the VRAM PCI resource.
>>>>
>>>>          
>>> CMA could reserve memory region for a specific driver so DRM Client
>>>        
>> could
>>      
>>> request memory allocation from only the region.
>>>
>>>        
>>>> This is to avoid the possibility that future drivers that need CMA will
>>>> be vulnerable to DOS-attacks from ill-behaved DRI clients.
>>>>
>>>>          
>>> Thomas, if any application has root authority for ill-purpose then isn't
>>>        
>> it
>>      
>>> possible to be vulnerable to DOS-attacks? I think DRM_AUTH means root
>>> authority. I know DRM Framework gives any root application DRM_AUTH
>>> authority for compatibility.
>>>        
>> DRM_AUTH just means that the client has authenticated w/ X11 (meaning
>> that it has permission to connect to x server)..
>>
>>      
> Yes, I understood so. but see drm_open_helper() of drm_fops.c file please.
> in this function, you can see a line below.
> /* for compatibility root is always authenticated */
> priv->authenticated = capable(CAP_SYS_ADMIN)
>
> I think the code above says that any application with root permission is
> authenticated.
>
>    

Yes, that is true. A root client may be assumed to have AUTH 
permissions, but the inverse does not hold, meaning that an AUTH client 
may *not* be assumed to have root permissions. I think there is a 
ROOT_ONLY ioctl flag for that.

The problem I'm seeing compared to other drivers is the following:

Imagine for example that you have a disc driver that allocates temporary 
memory out of the same DMA pool as the DRM driver.

Now you have a video player that is a DRM client. It contains a security 
flaw and is compromized by somebody trying to play a specially crafted 
video stream. The video player starts to allocate gem buffers until it 
receives an -ENOMEM. Then it stops allocating and does nothing.

Now the system tries an important disc access (paging for example). This 
fails, because the video player has exhausted all DMA memory and the 
disc driver fails to allocate.

The system is dead.

The point is:

If there is a chance that other drivers will use the same DMA/CMA pool 
as the DRM driver, DRM must leave enough DMA/CMA memory for those 
drivers to work.

The difference compared to other drm drivers:

There are other drm drivers that work the same way, with a static 
allocator. For example "via" and "sis". But those drivers completely 
claim the resources they are using. Nobody else is expected to use VRAM 
/ AGP.

In the Samsung case, it's not clear to me whether the DMA/CMA pool *can* 
be shared with other devices.
If it is, IMHO you must implement an allocation limit in DRM, if not, 
the driver should probably be safe.

Thanks,
Thomas

















>    
>> But I think that since memory allocation is limited to the size of the
>> CMA region, that this puts a reasonable cap of the memory that can be
>> allocated by the client.  If this is a problem, it certainly isn't the
>> worst problem.  You could still limit via file permissions the users
>> that can open the DRM device file, so it is really no worse than other
>> devices like v4l2..
>>
>> BR,
>> -R
>>
>>
>>      
>>>> /Thomas
>>>>
>>>>          
>>>
>>>
>>>        
>    

[PATCH v4] DRM: add DRM Driver for Samsung SoC EXYNOS4210.

[Rob Clark]

On Wed, Sep 14, 2011 at 2:57 AM, Thomas Hellstrom <thomas@shipmail.org> wrote:
>>>>>>>>>
>>>>>>>>> +static struct drm_ioctl_desc samsung_ioctls[] = {
>>>>>>>>> + ? ? ? DRM_IOCTL_DEF_DRV(SAMSUNG_GEM_CREATE,
>>>>>>>>> samsung_drm_gem_create_ioctl,
>>>>>>>>> + ? ? ? ? ? ? ? ? ? ? ? DRM_UNLOCKED | DRM_AUTH),
>>>>>>>>>
>>>>>>>>>
>>>>>>>>
>>>>>>>> Hi!
>>>>>>>>
>>>>>>>> With reference my previous security comment.
>>>>>>>>
>>>>>>>> Let's say you have a compromised video player running as a DRM
>>>>>>>> client, that
>>>>>>>> tries to repeatedly allocate huge GEM buffers...
>>>>>>>>
>>>>>>>> What will happen when all DMA memory is exhausted? Will this cause
>>>>>>>> other
>>>>>>>> device drivers to see an OOM, or only DRM?
>>>>>>>>
>>>>>>>> The old DRI model basically allowed any authorized DRI client to
>>>>>>>> exhaust
>>>>>>>> video ram or AGP memory, but never system memory. Newer DRI drivers
>>>>>>>> typically only allow DRI masters to do that.
>>>>>>>> as
>>>>>>>>
>>>>>>>> I don't think an authorized DRI client should be able to easily
>>>>>>>>
>>>>>
>>>>> exhaust
>>>>>
>>>>>>>>
>>>>>>>> resources (DMA memory) used by other device drivers causing them to
>>>>>>>> fail.
>>>>>>>>
>>>>>>>
>>>>>>> I'm not entirely sure what else can be done, other than have a
>>>>>>> threshold on max MB allocatable of buffer memory..
>>>>>>>
>>>>>>
>>>>>> Yes, I think that's what needs to be done, and that threshold should
>>>>>> be low enough to keep other device drivers running in the worst
>>>>>> allocation case.
>>>>>>
>>>>>>
>>>>>>>
>>>>>>> In the samsung driver case, he is only allocating scanout memory
>>>>>>>
>>>
>>> from
>>>
>>>>>>>
>>>>>>> CMA, so the limit will be the CMA region size.. beyond that you
>>>>>>>
>>>
>>> can't
>>>
>>>>>>>
>>>>>>> get physically contiguous memory. ?So I think this driver is safe.
>>>>>>>
>>>>>>
>>>>>> It's not really what well-behaved user-space drivers do that should
>>>>>>
>>>
>>> be
>>>
>>>>>>
>>>>>> a concern, but what compromized application *might* do that is a
>>>>>>
>>>>
>>>> concern.
>>>>
>>>>>
>>>>> Hmm. I might have missed your point here. If the buffer allocation
>>>>>
>>>
>>> ioctl
>>>
>>>>>
>>>>> only allows allocating CMA memory, then I agree the driver fits the old
>>>>> DRI security model, as long as no other devices on the platform will
>>>>> ever use CMA.
>>>>>
>>>>> But in that case, there really should be a way for the driver to say
>>>>> "Hey, all CMA memory on this system is mine", in the same way
>>>>> traditional video drivers can claim the VRAM PCI resource.
>>>>>
>>>>>
>>>>
>>>> CMA could reserve memory region for a specific driver so DRM Client
>>>>
>>>
>>> could
>>>
>>>>
>>>> request memory allocation from only the region.
>>>>
>>>>
>>>>>
>>>>> This is to avoid the possibility that future drivers that need CMA will
>>>>> be vulnerable to DOS-attacks from ill-behaved DRI clients.
>>>>>
>>>>>
>>>>
>>>> Thomas, if any application has root authority for ill-purpose then isn't
>>>>
>>>
>>> it
>>>
>>>>
>>>> possible to be vulnerable to DOS-attacks? I think DRM_AUTH means root
>>>> authority. I know DRM Framework gives any root application DRM_AUTH
>>>> authority for compatibility.
>>>>
>>>
>>> DRM_AUTH just means that the client has authenticated w/ X11 (meaning
>>> that it has permission to connect to x server)..
>>>
>>>
>>
>> Yes, I understood so. but see drm_open_helper() of drm_fops.c file please.
>> in this function, you can see a line below.
>> /* for compatibility root is always authenticated */
>> priv->authenticated = capable(CAP_SYS_ADMIN)
>>
>> I think the code above says that any application with root permission is
>> authenticated.
>>
>>
>
> Yes, that is true. A root client may be assumed to have AUTH permissions,
> but the inverse does not hold, meaning that an AUTH client may *not* be
> assumed to have root permissions. I think there is a ROOT_ONLY ioctl flag
> for that.
>
> The problem I'm seeing compared to other drivers is the following:
>
> Imagine for example that you have a disc driver that allocates temporary
> memory out of the same DMA pool as the DRM driver.
>
> Now you have a video player that is a DRM client. It contains a security
> flaw and is compromized by somebody trying to play a specially crafted video
> stream. The video player starts to allocate gem buffers until it receives an
> -ENOMEM. Then it stops allocating and does nothing.
>
> Now the system tries an important disc access (paging for example). This
> fails, because the video player has exhausted all DMA memory and the disc
> driver fails to allocate.
>
> The system is dead.
>
> The point is:
>
> If there is a chance that other drivers will use the same DMA/CMA pool as
> the DRM driver, DRM must leave enough DMA/CMA memory for those drivers to
> work.

ah, ok, I get your point


> The difference compared to other drm drivers:
>
> There are other drm drivers that work the same way, with a static allocator.
> For example "via" and "sis". But those drivers completely claim the
> resources they are using. Nobody else is expected to use VRAM / AGP.
>
> In the Samsung case, it's not clear to me whether the DMA/CMA pool *can* be
> shared with other devices.
> If it is, IMHO you must implement an allocation limit in DRM, if not, the
> driver should probably be safe.

It is possible to create a device private CMA pool..  although OTOH it
might be desirable to let some other drivers (like v4l2) use buffers
from the same pool..

I'm not entirely sure what will happen w/ dma_alloc_coherent, etc, if
the global CMA pool is exhausted.

Marek?  I guess you know what would happen?

BR,
-R

> Thanks,
> Thomas

[Linaro-mm-sig] [PATCH v4] DRM: add DRM Driver for Samsung SoC EXYNOS4210.

[Kyungmin Park]

On Thu, Sep 15, 2011 at 6:53 AM, Rob Clark <robdclark@gmail.com> wrote:
> On Wed, Sep 14, 2011 at 2:57 AM, Thomas Hellstrom <thomas@shipmail.org> wrote:
>>>>>>>>>>
>>>>>>>>>> +static struct drm_ioctl_desc samsung_ioctls[] = {
>>>>>>>>>> + ? ? ? DRM_IOCTL_DEF_DRV(SAMSUNG_GEM_CREATE,
>>>>>>>>>> samsung_drm_gem_create_ioctl,
>>>>>>>>>> + ? ? ? ? ? ? ? ? ? ? ? DRM_UNLOCKED | DRM_AUTH),
>>>>>>>>>>
>>>>>>>>>>
>>>>>>>>>
>>>>>>>>> Hi!
>>>>>>>>>
>>>>>>>>> With reference my previous security comment.
>>>>>>>>>
>>>>>>>>> Let's say you have a compromised video player running as a DRM
>>>>>>>>> client, that
>>>>>>>>> tries to repeatedly allocate huge GEM buffers...
>>>>>>>>>
>>>>>>>>> What will happen when all DMA memory is exhausted? Will this cause
>>>>>>>>> other
>>>>>>>>> device drivers to see an OOM, or only DRM?
>>>>>>>>>
>>>>>>>>> The old DRI model basically allowed any authorized DRI client to
>>>>>>>>> exhaust
>>>>>>>>> video ram or AGP memory, but never system memory. Newer DRI drivers
>>>>>>>>> typically only allow DRI masters to do that.
>>>>>>>>> as
>>>>>>>>>
>>>>>>>>> I don't think an authorized DRI client should be able to easily
>>>>>>>>>
>>>>>>
>>>>>> exhaust
>>>>>>
>>>>>>>>>
>>>>>>>>> resources (DMA memory) used by other device drivers causing them to
>>>>>>>>> fail.
>>>>>>>>>
>>>>>>>>
>>>>>>>> I'm not entirely sure what else can be done, other than have a
>>>>>>>> threshold on max MB allocatable of buffer memory..
>>>>>>>>
>>>>>>>
>>>>>>> Yes, I think that's what needs to be done, and that threshold should
>>>>>>> be low enough to keep other device drivers running in the worst
>>>>>>> allocation case.
>>>>>>>
>>>>>>>
>>>>>>>>
>>>>>>>> In the samsung driver case, he is only allocating scanout memory
>>>>>>>>
>>>>
>>>> from
>>>>
>>>>>>>>
>>>>>>>> CMA, so the limit will be the CMA region size.. beyond that you
>>>>>>>>
>>>>
>>>> can't
>>>>
>>>>>>>>
>>>>>>>> get physically contiguous memory. ?So I think this driver is safe.
>>>>>>>>
>>>>>>>
>>>>>>> It's not really what well-behaved user-space drivers do that should
>>>>>>>
>>>>
>>>> be
>>>>
>>>>>>>
>>>>>>> a concern, but what compromized application *might* do that is a
>>>>>>>
>>>>>
>>>>> concern.
>>>>>
>>>>>>
>>>>>> Hmm. I might have missed your point here. If the buffer allocation
>>>>>>
>>>>
>>>> ioctl
>>>>
>>>>>>
>>>>>> only allows allocating CMA memory, then I agree the driver fits the old
>>>>>> DRI security model, as long as no other devices on the platform will
>>>>>> ever use CMA.
>>>>>>
>>>>>> But in that case, there really should be a way for the driver to say
>>>>>> "Hey, all CMA memory on this system is mine", in the same way
>>>>>> traditional video drivers can claim the VRAM PCI resource.
>>>>>>
>>>>>>
>>>>>
>>>>> CMA could reserve memory region for a specific driver so DRM Client
>>>>>
>>>>
>>>> could
>>>>
>>>>>
>>>>> request memory allocation from only the region.
>>>>>
>>>>>
>>>>>>
>>>>>> This is to avoid the possibility that future drivers that need CMA will
>>>>>> be vulnerable to DOS-attacks from ill-behaved DRI clients.
>>>>>>
>>>>>>
>>>>>
>>>>> Thomas, if any application has root authority for ill-purpose then isn't
>>>>>
>>>>
>>>> it
>>>>
>>>>>
>>>>> possible to be vulnerable to DOS-attacks? I think DRM_AUTH means root
>>>>> authority. I know DRM Framework gives any root application DRM_AUTH
>>>>> authority for compatibility.
>>>>>
>>>>
>>>> DRM_AUTH just means that the client has authenticated w/ X11 (meaning
>>>> that it has permission to connect to x server)..
>>>>
>>>>
>>>
>>> Yes, I understood so. but see drm_open_helper() of drm_fops.c file please.
>>> in this function, you can see a line below.
>>> /* for compatibility root is always authenticated */
>>> priv->authenticated = capable(CAP_SYS_ADMIN)
>>>
>>> I think the code above says that any application with root permission is
>>> authenticated.
>>>
>>>
>>
>> Yes, that is true. A root client may be assumed to have AUTH permissions,
>> but the inverse does not hold, meaning that an AUTH client may *not* be
>> assumed to have root permissions. I think there is a ROOT_ONLY ioctl flag
>> for that.
>>
>> The problem I'm seeing compared to other drivers is the following:
>>
>> Imagine for example that you have a disc driver that allocates temporary
>> memory out of the same DMA pool as the DRM driver.
>>
>> Now you have a video player that is a DRM client. It contains a security
>> flaw and is compromized by somebody trying to play a specially crafted video
>> stream. The video player starts to allocate gem buffers until it receives an
>> -ENOMEM. Then it stops allocating and does nothing.
>>
>> Now the system tries an important disc access (paging for example). This
>> fails, because the video player has exhausted all DMA memory and the disc
>> driver fails to allocate.
>>
>> The system is dead.
>>
>> The point is:
>>
>> If there is a chance that other drivers will use the same DMA/CMA pool as
>> the DRM driver, DRM must leave enough DMA/CMA memory for those drivers to
>> work.
>
> ah, ok, I get your point
>
>
>> The difference compared to other drm drivers:
>>
>> There are other drm drivers that work the same way, with a static allocator.
>> For example "via" and "sis". But those drivers completely claim the
>> resources they are using. Nobody else is expected to use VRAM / AGP.
>>
>> In the Samsung case, it's not clear to me whether the DMA/CMA pool *can* be
>> shared with other devices.
>> If it is, IMHO you must implement an allocation limit in DRM, if not, the
>> driver should probably be safe.
>
> It is possible to create a device private CMA pool.. ?although OTOH it
> might be desirable to let some other drivers (like v4l2) use buffers
> from the same pool..
That's the final goal. memory sharing among multimedia devices (v4l2),
display(fimd, hdmi), and 3D (mali).
currently mali drivers is tightly coupled with UMP and it's hard to
use the CMA. it's need to solve this issue also.

In case of multimedia drivers, it has fixed memory area as scenario
and display also has similar restriction. I don't expect DRI request
the more memory than CMA pool has.
>
> I'm not entirely sure what will happen w/ dma_alloc_coherent, etc, if
> the global CMA pool is exhausted.
The primary goal of CMA is guarantee the memory for multimedia. So if
other device use the multimedia CMA pool, it should release the used
memory. and it's hard to release the memory. it should has display CMA
pool and don't share the other multimedia devices CMA pool.

CMA can share the same CMA pool and specify the device usage which CMA
pool used.
>
> Marek? ?I guess you know what would happen?

He will comeback when 21 Sep.

Thank you,
Kyungmin Park

[PATCH v4] DRM: add DRM Driver for Samsung SoC EXYNOS4210.

[Marek Szyprowski]

Hello,

On Wednesday, September 14, 2011 11:53 PM Rob Clark wrote:

> On Wed, Sep 14, 2011 at 2:57 AM, Thomas Hellstrom <thomas@shipmail.org> wrote:

(snipped)

> > Yes, that is true. A root client may be assumed to have AUTH permissions,
> > but the inverse does not hold, meaning that an AUTH client may *not* be
> > assumed to have root permissions. I think there is a ROOT_ONLY ioctl flag
> > for that.
> >
> > The problem I'm seeing compared to other drivers is the following:
> >
> > Imagine for example that you have a disc driver that allocates temporary
> > memory out of the same DMA pool as the DRM driver.
> >
> > Now you have a video player that is a DRM client. It contains a security
> > flaw and is compromized by somebody trying to play a specially crafted video
> > stream. The video player starts to allocate gem buffers until it receives an
> > -ENOMEM. Then it stops allocating and does nothing.
> >
> > Now the system tries an important disc access (paging for example). This
> > fails, because the video player has exhausted all DMA memory and the disc
> > driver fails to allocate.
> >
> > The system is dead.
> >
> > The point is:
> >
> > If there is a chance that other drivers will use the same DMA/CMA pool as
> > the DRM driver, DRM must leave enough DMA/CMA memory for those drivers to
> > work.
> 
> ah, ok, I get your point
> 
> > The difference compared to other drm drivers:
> >
> > There are other drm drivers that work the same way, with a static allocator.
> > For example "via" and "sis". But those drivers completely claim the
> > resources they are using. Nobody else is expected to use VRAM / AGP.
> >
> > In the Samsung case, it's not clear to me whether the DMA/CMA pool *can* be
> > shared with other devices.
> > If it is, IMHO you must implement an allocation limit in DRM, if not, the
> > driver should probably be safe.
> 
> It is possible to create a device private CMA pool..  although OTOH it
> might be desirable to let some other drivers (like v4l2) use buffers
> from the same pool..

Creating a device private CMA pool doesn't prevent other drivers to access
memory from it if there is a way to pass a buffer to them (i.e. using 
dma_buf method). However the driver must be able to address that memory.
CMA private pools were designed mainly to resolve the problem that some
weird hardware can access certain types of buffers only at particular
memory address ranges which matches particular memory bank.
 
> I'm not entirely sure what will happen w/ dma_alloc_coherent, etc, if
> the global CMA pool is exhausted.
> 
> Marek?  I guess you know what would happen?

The allocation will simply fail and dma_alloc_coherent will return NULL.

Best regards
-- 
Marek Szyprowski
Samsung Poland R&D Center

[PATCH v4] DRM: add DRM Driver for Samsung SoC EXYNOS4210.

[Rob Clark]

On Wed, Sep 21, 2011 at 7:41 AM, Marek Szyprowski
<m.szyprowski@samsung.com> wrote:
>> I'm not entirely sure what will happen w/ dma_alloc_coherent, etc, if
>> the global CMA pool is exhausted.
>>
>> Marek? ?I guess you know what would happen?
>
> The allocation will simply fail and dma_alloc_coherent will return NULL.
>

Ok, so we should probably setup device private pools so that userspace
cannot allocate so much buffer memory that dma_alloc_coherent fails
for other devices..

This is what I've done so far for omapdrm driver, although it is part
of the other patch that adds the platform-device..

Thx

BR,
-R

[PATCH v4] DRM: add DRM Driver for Samsung SoC EXYNOS4210.

[Inki Dae]

Hi, Thomas.

> -----Original Message-----
> From: Thomas Hellstrom [mailto:thomas at shipmail.org]
> Sent: Wednesday, September 14, 2011 4:57 PM
> To: Inki Dae
> Cc: 'Rob Clark'; kyungmin.park at samsung.com; sw0312.kim at samsung.com; linux-
> arm-kernel at lists.infradead.org; dri-devel at lists.freedesktop.org
> Subject: Re: [PATCH v4] DRM: add DRM Driver for Samsung SoC EXYNOS4210.
> 
> On 09/14/2011 07:55 AM, Inki Dae wrote:
> >
> >
> >> -----Original Message-----
> >> From: Rob Clark [mailto:robdclark at gmail.com]
> >> Sent: Wednesday, September 14, 2011 11:26 AM
> >> To: Inki Dae
> >> Cc: Thomas Hellstrom; kyungmin.park at samsung.com;
sw0312.kim at samsung.com;
> >> linux-arm-kernel at lists.infradead.org; dri-devel at lists.freedesktop.org
> >> Subject: Re: [PATCH v4] DRM: add DRM Driver for Samsung SoC EXYNOS4210.
> >>
> >> On Tue, Sep 13, 2011 at 9:03 PM, Inki Dae<inki.dae@samsung.com>  wrote:
> >>
> >>> Hi Thomas.
> >>>
> >>>
> >>>> -----Original Message-----
> >>>> From: Thomas Hellstrom [mailto:thomas at shipmail.org]
> >>>> Sent: Monday, September 12, 2011 3:32 PM
> >>>> To: Rob Clark
> >>>> Cc: Inki Dae; kyungmin.park at samsung.com; sw0312.kim at samsung.com;
> linux-
> >>>> arm-kernel at lists.infradead.org; dri-devel at lists.freedesktop.org
> >>>> Subject: Re: [PATCH v4] DRM: add DRM Driver for Samsung SoC
> EXYNOS4210.
> >>>>
> >>>> On 09/11/2011 11:26 PM, Thomas Hellstrom wrote:
> >>>>
> >>>>> On 09/10/2011 07:31 PM, Rob Clark wrote:
> >>>>>
> >>>>>> On Sat, Sep 10, 2011 at 9:04 AM, Thomas
> >>>>>> Hellstrom<thomas@shipmail.org>    wrote:
> >>>>>>
> >>>>>>> On 09/09/2011 01:38 PM, Inki Dae wrote:
> >>>>>>>
> >>>>>>>> This patch is a DRM Driver for Samsung SoC Exynos4210 and now
> >>>>>>>>
> >> enables
> >>
> >>>>>>>> only FIMD yet but we will add HDMI support also in the future.
> >>>>>>>>
> >>>>>>>> from now on, I will remove RFC prefix because I think we have got
> >>>>>>>> comments
> >>>>>>>> enough.
> >>>>>>>>
> >>>>>>>> this patch is based on git repository below:
> >>>>>>>>
git://git.kernel.org/pub/scm/linux/kernel/git/airlied/drm-2.6.git,
> >>>>>>>> branch name: drm-next
> >>>>>>>> commit-id: bcc65fd8e929a9d9d34d814d6efc1d2793546922
> >>>>>>>>
> >>>>>>>> you can refer to our working repository below:
> >>>>>>>> http://git.infradead.org/users/kmpark/linux-2.6-samsung
> >>>>>>>> branch name: samsung-drm
> >>>>>>>>
> >>>>>>>> We tried to re-use lowlevel codes of the FIMD driver(s3c-fb.c
> >>>>>>>> based on Linux framebuffer) but couldn't so because lowlevel
> codes
> >>>>>>>> of s3c-fb.c are included internally and so FIMD module of this
> >>>>>>>> driver has
> >>>>>>>> its own lowlevel codes.
> >>>>>>>>
> >>>>>>>> We used GEM framework for buffer management and DMA
> >>>>>>>>
> >> APIs(dma_alloc_*)
> >>
> >>>>>>>> for buffer allocation. by using DMA API, we could use CMA later.
> >>>>>>>>
> >>>>>>>> Refer to this link for CMA(Continuous Memory Allocator):
> >>>>>>>> http://lkml.org/lkml/2011/7/20/45
> >>>>>>>>
> >>>>>>>> this driver supports only physically continuous
memory(non-iommu).
> >>>>>>>>
> >>>>>>>> Links to previous versions of the patchset:
> >>>>>>>> v1:<      https://lwn.net/Articles/454380/>
> >>>>>>>> v2:<      http://www.spinics.net/lists/kernel/msg1224275.html>
> >>>>>>>> v3:<      http://www.gossamer-
> >>>>>>>>
> >> threads.com/lists/linux/kernel/1423684>
> >>
> >>>>>>>> Changelog v2:
> >>>>>>>> DRM: add DRM_IOCTL_SAMSUNG_GEM_MMAP ioctl command.
> >>>>>>>>
> >>>>>>>>       this feature maps user address space to physical memory
> >>>>>>>>
> >> region
> >>
> >>>>>>>>       once user application requests DRM_IOCTL_SAMSUNG_GEM_MMAP
> >>>>>>>>
> >> ioctl.
> >>
> >>>>>>>> DRM: code clean and add exception codes.
> >>>>>>>>
> >>>>>>>> Changelog v3:
> >>>>>>>> DRM: Support multiple irq.
> >>>>>>>>
> >>>>>>>>       FIMD and HDMI have their own irq handler but DRM Framework
> >>>>>>>>
> >> can
> >>
> >>>>>>>> regiter
> >>>>>>>>       only one irq handler this patch supports mutiple irq for
> >>>>>>>> Samsung SoC.
> >>>>>>>>
> >>>>>>>> DRM: Consider modularization.
> >>>>>>>>
> >>>>>>>>       each DRM, FIMD could be built as a module.
> >>>>>>>>
> >>>>>>>> DRM: Have indenpendent crtc object.
> >>>>>>>>
> >>>>>>>>       crtc isn't specific to SoC Platform so this patch gets a
> crtc
> >>>>>>>>       to be used as common object.
> >>>>>>>>       created crtc could be attached to any encoder object.
> >>>>>>>>
> >>>>>>>> DRM: code clean and add exception codes.
> >>>>>>>>
> >>>>>>>> Changelog v4:
> >>>>>>>> DRM: remove is_defult from samsung_fb.
> >>>>>>>>
> >>>>>>>>       is_default isn't used for default framebuffer.
> >>>>>>>>
> >>>>>>>> DRM: code refactoring to fimd module.
> >>>>>>>>       this patch is be considered with multiple display objects
> and
> >>>>>>>>       would use its own request_irq() to register a irq handler
> >>>>>>>> instead of
> >>>>>>>>       drm framework's one.
> >>>>>>>>
> >>>>>>>> DRM: remove find_samsung_drm_gem_object()
> >>>>>>>>
> >>>>>>>> DRM: move kernel private data structures and definitions to
> driver
> >>>>>>>> folder.
> >>>>>>>>
> >>>>>>>>       samsung_drm.h would contain only public information for
> >>>>>>>>
> >>> userspace
> >>>
> >>>>>>>>       ioctl interface.
> >>>>>>>>
> >>>>>>>> DRM: code refactoring to gem modules.
> >>>>>>>>       buffer module isn't dependent of gem module anymore.
> >>>>>>>>
> >>>>>>>> DRM: fixed security issue.
> >>>>>>>>
> >>>>>>>> DRM: remove encoder porinter from specific connector.
> >>>>>>>>
> >>>>>>>>       samsung connector doesn't need to have generic encoder.
> >>>>>>>>
> >>>>>>>> DRM: code clean and add exception codes.
> >>>>>>>>
> >>>>>>>> Signed-off-by: Inki Dae<inki.dae@samsung.com>
> >>>>>>>> Signed-off-by: Joonyoung Shim<jy0922.shim@samsung.com>
> >>>>>>>> Signed-off-by: SeungWoo Kim<sw0312.kim@samsung.com>
> >>>>>>>> Signed-off-by: kyungmin.park<kyungmin.park@samsung.com>
> >>>>>>>> ---
> >>>>>>>>
> >>>>>>>> +static struct drm_ioctl_desc samsung_ioctls[] = {
> >>>>>>>> +       DRM_IOCTL_DEF_DRV(SAMSUNG_GEM_CREATE,
> >>>>>>>> samsung_drm_gem_create_ioctl,
> >>>>>>>> +                       DRM_UNLOCKED | DRM_AUTH),
> >>>>>>>>
> >>>>>>>>
> >>>>>>> Hi!
> >>>>>>>
> >>>>>>> With reference my previous security comment.
> >>>>>>>
> >>>>>>> Let's say you have a compromised video player running as a DRM
> >>>>>>> client, that
> >>>>>>> tries to repeatedly allocate huge GEM buffers...
> >>>>>>>
> >>>>>>> What will happen when all DMA memory is exhausted? Will this cause
> >>>>>>> other
> >>>>>>> device drivers to see an OOM, or only DRM?
> >>>>>>>
> >>>>>>> The old DRI model basically allowed any authorized DRI client to
> >>>>>>> exhaust
> >>>>>>> video ram or AGP memory, but never system memory. Newer DRI
> drivers
> >>>>>>> typically only allow DRI masters to do that.
> >>>>>>> as
> >>>>>>>
> >>>>>>> I don't think an authorized DRI client should be able to easily
> >>>>>>>
> >>>> exhaust
> >>>>
> >>>>>>> resources (DMA memory) used by other device drivers causing them
> to
> >>>>>>> fail.
> >>>>>>>
> >>>>>> I'm not entirely sure what else can be done, other than have a
> >>>>>> threshold on max MB allocatable of buffer memory..
> >>>>>>
> >>>>> Yes, I think that's what needs to be done, and that threshold should
> >>>>> be low enough to keep other device drivers running in the worst
> >>>>> allocation case.
> >>>>>
> >>>>>
> >>>>>> In the samsung driver case, he is only allocating scanout memory
> >>>>>>
> >> from
> >>
> >>>>>> CMA, so the limit will be the CMA region size.. beyond that you
> >>>>>>
> >> can't
> >>
> >>>>>> get physically contiguous memory.  So I think this driver is safe.
> >>>>>>
> >>>>> It's not really what well-behaved user-space drivers do that should
> >>>>>
> >> be
> >>
> >>>>> a concern, but what compromized application *might* do that is a
> >>>>>
> >>> concern.
> >>>
> >>>> Hmm. I might have missed your point here. If the buffer allocation
> >>>>
> >> ioctl
> >>
> >>>> only allows allocating CMA memory, then I agree the driver fits the
> old
> >>>> DRI security model, as long as no other devices on the platform will
> >>>> ever use CMA.
> >>>>
> >>>> But in that case, there really should be a way for the driver to say
> >>>> "Hey, all CMA memory on this system is mine", in the same way
> >>>> traditional video drivers can claim the VRAM PCI resource.
> >>>>
> >>>>
> >>> CMA could reserve memory region for a specific driver so DRM Client
> >>>
> >> could
> >>
> >>> request memory allocation from only the region.
> >>>
> >>>
> >>>> This is to avoid the possibility that future drivers that need CMA
> will
> >>>> be vulnerable to DOS-attacks from ill-behaved DRI clients.
> >>>>
> >>>>
> >>> Thomas, if any application has root authority for ill-purpose then
> isn't
> >>>
> >> it
> >>
> >>> possible to be vulnerable to DOS-attacks? I think DRM_AUTH means root
> >>> authority. I know DRM Framework gives any root application DRM_AUTH
> >>> authority for compatibility.
> >>>
> >> DRM_AUTH just means that the client has authenticated w/ X11 (meaning
> >> that it has permission to connect to x server)..
> >>
> >>
> > Yes, I understood so. but see drm_open_helper() of drm_fops.c file
> please.
> > in this function, you can see a line below.
> > /* for compatibility root is always authenticated */
> > priv->authenticated = capable(CAP_SYS_ADMIN)
> >
> > I think the code above says that any application with root permission is
> > authenticated.
> >
> >
> 
> Yes, that is true. A root client may be assumed to have AUTH
> permissions, but the inverse does not hold, meaning that an AUTH client
> may *not* be assumed to have root permissions. I think there is a
> ROOT_ONLY ioctl flag for that.
> 
> The problem I'm seeing compared to other drivers is the following:
> 
> Imagine for example that you have a disc driver that allocates temporary
> memory out of the same DMA pool as the DRM driver.
> 
> Now you have a video player that is a DRM client. It contains a security
> flaw and is compromized by somebody trying to play a specially crafted
> video stream. The video player starts to allocate gem buffers until it
> receives an -ENOMEM. Then it stops allocating and does nothing.
> 
> Now the system tries an important disc access (paging for example). This
> fails, because the video player has exhausted all DMA memory and the
> disc driver fails to allocate.
> 
> The system is dead.
> 

Ok, I understood. but in case of using CMA, DRM driver would have private
memory pool which is reserved only for itself. so although the video player
requested gem buffer allocation until it receives an -ENOMEM and then the
system is try to access the disc, it would work fine. because the region
requested by system and the region requested by DRM driver could entirely be
separated. DRM driver wouldn't have implications for the system region.

> The point is:
> 
> If there is a chance that other drivers will use the same DMA/CMA pool
> as the DRM driver, DRM must leave enough DMA/CMA memory for those
> drivers to work.
> 
> The difference compared to other drm drivers:
> 
> There are other drm drivers that work the same way, with a static
> allocator. For example "via" and "sis". But those drivers completely
> claim the resources they are using. Nobody else is expected to use VRAM
> / AGP.
> 

I think if we use private cma region for DRM driver then it is similar to
via and sis way.

> In the Samsung case, it's not clear to me whether the DMA/CMA pool *can*
> be shared with other devices.
> If it is, IMHO you must implement an allocation limit in DRM, if not,
> the driver should probably be safe.
> 

As your saying, we would implement an allocation limit in DRM; DRM driver
have private region that any other drivers can't access to this one. in
fact, such limitation should be set at machine code. if this way has any
problems for security also, we CAN CHANGE DRM permission anytime. :) thank
you for your detailed account and that was very helpful to me.

> Thanks,
> Thomas
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> >
> >> But I think that since memory allocation is limited to the size of the
> >> CMA region, that this puts a reasonable cap of the memory that can be
> >> allocated by the client.  If this is a problem, it certainly isn't the
> >> worst problem.  You could still limit via file permissions the users
> >> that can open the DRM device file, so it is really no worse than other
> >> devices like v4l2..
> >>
> >> BR,
> >> -R
> >>
> >>
> >>
> >>>> /Thomas
> >>>>
> >>>>
> >>>
> >>>
> >>>
> >
> 

[PATCH v4] DRM: add DRM Driver for Samsung SoC EXYNOS4210.

[Thomas Hellstrom]

On 09/15/2011 03:20 AM, Inki Dae wrote:
> Hi, Thomas.
>
>    
>> -----Original Message-----
>> From: Thomas Hellstrom [mailto:thomas at shipmail.org]
>> Sent: Wednesday, September 14, 2011 4:57 PM
>> To: Inki Dae
>> Cc: 'Rob Clark'; kyungmin.park at samsung.com; sw0312.kim at samsung.com; linux-
>> arm-kernel at lists.infradead.org; dri-devel at lists.freedesktop.org
>> Subject: Re: [PATCH v4] DRM: add DRM Driver for Samsung SoC EXYNOS4210.
>>
>> On 09/14/2011 07:55 AM, Inki Dae wrote:
>>      
>>>
>>>        
>>>> -----Original Message-----
>>>> From: Rob Clark [mailto:robdclark at gmail.com]
>>>> Sent: Wednesday, September 14, 2011 11:26 AM
>>>> To: Inki Dae
>>>> Cc: Thomas Hellstrom; kyungmin.park at samsung.com;
>>>>          
> sw0312.kim at samsung.com;
>    
>>>> linux-arm-kernel at lists.infradead.org; dri-devel at lists.freedesktop.org
>>>> Subject: Re: [PATCH v4] DRM: add DRM Driver for Samsung SoC EXYNOS4210.
>>>>
>>>> On Tue, Sep 13, 2011 at 9:03 PM, Inki Dae<inki.dae@samsung.com>   wrote:
>>>>
>>>>          
>>>>> Hi Thomas.
>>>>>
>>>>>
>>>>>            
>>>>>> -----Original Message-----
>>>>>> From: Thomas Hellstrom [mailto:thomas at shipmail.org]
>>>>>> Sent: Monday, September 12, 2011 3:32 PM
>>>>>> To: Rob Clark
>>>>>> Cc: Inki Dae; kyungmin.park at samsung.com; sw0312.kim at samsung.com;
>>>>>>              
>> linux-
>>      
>>>>>> arm-kernel at lists.infradead.org; dri-devel at lists.freedesktop.org
>>>>>> Subject: Re: [PATCH v4] DRM: add DRM Driver for Samsung SoC
>>>>>>              
>> EXYNOS4210.
>>      
>>>>>> On 09/11/2011 11:26 PM, Thomas Hellstrom wrote:
>>>>>>
>>>>>>              
>>>>>>> On 09/10/2011 07:31 PM, Rob Clark wrote:
>>>>>>>
>>>>>>>                
>>>>>>>> On Sat, Sep 10, 2011 at 9:04 AM, Thomas
>>>>>>>> Hellstrom<thomas@shipmail.org>     wrote:
>>>>>>>>
>>>>>>>>                  
>>>>>>>>> On 09/09/2011 01:38 PM, Inki Dae wrote:
>>>>>>>>>
>>>>>>>>>                    
>>>>>>>>>> This patch is a DRM Driver for Samsung SoC Exynos4210 and now
>>>>>>>>>>
>>>>>>>>>>                      
>>>> enables
>>>>
>>>>          
>>>>>>>>>> only FIMD yet but we will add HDMI support also in the future.
>>>>>>>>>>
>>>>>>>>>> from now on, I will remove RFC prefix because I think we have got
>>>>>>>>>> comments
>>>>>>>>>> enough.
>>>>>>>>>>
>>>>>>>>>> this patch is based on git repository below:
>>>>>>>>>>
>>>>>>>>>>                      
> git://git.kernel.org/pub/scm/linux/kernel/git/airlied/drm-2.6.git,
>    
>>>>>>>>>> branch name: drm-next
>>>>>>>>>> commit-id: bcc65fd8e929a9d9d34d814d6efc1d2793546922
>>>>>>>>>>
>>>>>>>>>> you can refer to our working repository below:
>>>>>>>>>> http://git.infradead.org/users/kmpark/linux-2.6-samsung
>>>>>>>>>> branch name: samsung-drm
>>>>>>>>>>
>>>>>>>>>> We tried to re-use lowlevel codes of the FIMD driver(s3c-fb.c
>>>>>>>>>> based on Linux framebuffer) but couldn't so because lowlevel
>>>>>>>>>>                      
>> codes
>>      
>>>>>>>>>> of s3c-fb.c are included internally and so FIMD module of this
>>>>>>>>>> driver has
>>>>>>>>>> its own lowlevel codes.
>>>>>>>>>>
>>>>>>>>>> We used GEM framework for buffer management and DMA
>>>>>>>>>>
>>>>>>>>>>                      
>>>> APIs(dma_alloc_*)
>>>>
>>>>          
>>>>>>>>>> for buffer allocation. by using DMA API, we could use CMA later.
>>>>>>>>>>
>>>>>>>>>> Refer to this link for CMA(Continuous Memory Allocator):
>>>>>>>>>> http://lkml.org/lkml/2011/7/20/45
>>>>>>>>>>
>>>>>>>>>> this driver supports only physically continuous
>>>>>>>>>>                      
> memory(non-iommu).
>    
>>>>>>>>>> Links to previous versions of the patchset:
>>>>>>>>>> v1:<       https://lwn.net/Articles/454380/>
>>>>>>>>>> v2:<       http://www.spinics.net/lists/kernel/msg1224275.html>
>>>>>>>>>> v3:<       http://www.gossamer-
>>>>>>>>>>
>>>>>>>>>>                      
>>>> threads.com/lists/linux/kernel/1423684>
>>>>
>>>>          
>>>>>>>>>> Changelog v2:
>>>>>>>>>> DRM: add DRM_IOCTL_SAMSUNG_GEM_MMAP ioctl command.
>>>>>>>>>>
>>>>>>>>>>        this feature maps user address space to physical memory
>>>>>>>>>>
>>>>>>>>>>                      
>>>> region
>>>>
>>>>          
>>>>>>>>>>        once user application requests DRM_IOCTL_SAMSUNG_GEM_MMAP
>>>>>>>>>>
>>>>>>>>>>                      
>>>> ioctl.
>>>>
>>>>          
>>>>>>>>>> DRM: code clean and add exception codes.
>>>>>>>>>>
>>>>>>>>>> Changelog v3:
>>>>>>>>>> DRM: Support multiple irq.
>>>>>>>>>>
>>>>>>>>>>        FIMD and HDMI have their own irq handler but DRM Framework
>>>>>>>>>>
>>>>>>>>>>                      
>>>> can
>>>>
>>>>          
>>>>>>>>>> regiter
>>>>>>>>>>        only one irq handler this patch supports mutiple irq for
>>>>>>>>>> Samsung SoC.
>>>>>>>>>>
>>>>>>>>>> DRM: Consider modularization.
>>>>>>>>>>
>>>>>>>>>>        each DRM, FIMD could be built as a module.
>>>>>>>>>>
>>>>>>>>>> DRM: Have indenpendent crtc object.
>>>>>>>>>>
>>>>>>>>>>        crtc isn't specific to SoC Platform so this patch gets a
>>>>>>>>>>                      
>> crtc
>>      
>>>>>>>>>>        to be used as common object.
>>>>>>>>>>        created crtc could be attached to any encoder object.
>>>>>>>>>>
>>>>>>>>>> DRM: code clean and add exception codes.
>>>>>>>>>>
>>>>>>>>>> Changelog v4:
>>>>>>>>>> DRM: remove is_defult from samsung_fb.
>>>>>>>>>>
>>>>>>>>>>        is_default isn't used for default framebuffer.
>>>>>>>>>>
>>>>>>>>>> DRM: code refactoring to fimd module.
>>>>>>>>>>        this patch is be considered with multiple display objects
>>>>>>>>>>                      
>> and
>>      
>>>>>>>>>>        would use its own request_irq() to register a irq handler
>>>>>>>>>> instead of
>>>>>>>>>>        drm framework's one.
>>>>>>>>>>
>>>>>>>>>> DRM: remove find_samsung_drm_gem_object()
>>>>>>>>>>
>>>>>>>>>> DRM: move kernel private data structures and definitions to
>>>>>>>>>>                      
>> driver
>>      
>>>>>>>>>> folder.
>>>>>>>>>>
>>>>>>>>>>        samsung_drm.h would contain only public information for
>>>>>>>>>>
>>>>>>>>>>                      
>>>>> userspace
>>>>>
>>>>>            
>>>>>>>>>>        ioctl interface.
>>>>>>>>>>
>>>>>>>>>> DRM: code refactoring to gem modules.
>>>>>>>>>>        buffer module isn't dependent of gem module anymore.
>>>>>>>>>>
>>>>>>>>>> DRM: fixed security issue.
>>>>>>>>>>
>>>>>>>>>> DRM: remove encoder porinter from specific connector.
>>>>>>>>>>
>>>>>>>>>>        samsung connector doesn't need to have generic encoder.
>>>>>>>>>>
>>>>>>>>>> DRM: code clean and add exception codes.
>>>>>>>>>>
>>>>>>>>>> Signed-off-by: Inki Dae<inki.dae@samsung.com>
>>>>>>>>>> Signed-off-by: Joonyoung Shim<jy0922.shim@samsung.com>
>>>>>>>>>> Signed-off-by: SeungWoo Kim<sw0312.kim@samsung.com>
>>>>>>>>>> Signed-off-by: kyungmin.park<kyungmin.park@samsung.com>
>>>>>>>>>> ---
>>>>>>>>>>
>>>>>>>>>> +static struct drm_ioctl_desc samsung_ioctls[] = {
>>>>>>>>>> +       DRM_IOCTL_DEF_DRV(SAMSUNG_GEM_CREATE,
>>>>>>>>>> samsung_drm_gem_create_ioctl,
>>>>>>>>>> +                       DRM_UNLOCKED | DRM_AUTH),
>>>>>>>>>>
>>>>>>>>>>
>>>>>>>>>>                      
>>>>>>>>> Hi!
>>>>>>>>>
>>>>>>>>> With reference my previous security comment.
>>>>>>>>>
>>>>>>>>> Let's say you have a compromised video player running as a DRM
>>>>>>>>> client, that
>>>>>>>>> tries to repeatedly allocate huge GEM buffers...
>>>>>>>>>
>>>>>>>>> What will happen when all DMA memory is exhausted? Will this cause
>>>>>>>>> other
>>>>>>>>> device drivers to see an OOM, or only DRM?
>>>>>>>>>
>>>>>>>>> The old DRI model basically allowed any authorized DRI client to
>>>>>>>>> exhaust
>>>>>>>>> video ram or AGP memory, but never system memory. Newer DRI
>>>>>>>>>                    
>> drivers
>>      
>>>>>>>>> typically only allow DRI masters to do that.
>>>>>>>>> as
>>>>>>>>>
>>>>>>>>> I don't think an authorized DRI client should be able to easily
>>>>>>>>>
>>>>>>>>>                    
>>>>>> exhaust
>>>>>>
>>>>>>              
>>>>>>>>> resources (DMA memory) used by other device drivers causing them
>>>>>>>>>                    
>> to
>>      
>>>>>>>>> fail.
>>>>>>>>>
>>>>>>>>>                    
>>>>>>>> I'm not entirely sure what else can be done, other than have a
>>>>>>>> threshold on max MB allocatable of buffer memory..
>>>>>>>>
>>>>>>>>                  
>>>>>>> Yes, I think that's what needs to be done, and that threshold should
>>>>>>> be low enough to keep other device drivers running in the worst
>>>>>>> allocation case.
>>>>>>>
>>>>>>>
>>>>>>>                
>>>>>>>> In the samsung driver case, he is only allocating scanout memory
>>>>>>>>
>>>>>>>>                  
>>>> from
>>>>
>>>>          
>>>>>>>> CMA, so the limit will be the CMA region size.. beyond that you
>>>>>>>>
>>>>>>>>                  
>>>> can't
>>>>
>>>>          
>>>>>>>> get physically contiguous memory.  So I think this driver is safe.
>>>>>>>>
>>>>>>>>                  
>>>>>>> It's not really what well-behaved user-space drivers do that should
>>>>>>>
>>>>>>>                
>>>> be
>>>>
>>>>          
>>>>>>> a concern, but what compromized application *might* do that is a
>>>>>>>
>>>>>>>                
>>>>> concern.
>>>>>
>>>>>            
>>>>>> Hmm. I might have missed your point here. If the buffer allocation
>>>>>>
>>>>>>              
>>>> ioctl
>>>>
>>>>          
>>>>>> only allows allocating CMA memory, then I agree the driver fits the
>>>>>>              
>> old
>>      
>>>>>> DRI security model, as long as no other devices on the platform will
>>>>>> ever use CMA.
>>>>>>
>>>>>> But in that case, there really should be a way for the driver to say
>>>>>> "Hey, all CMA memory on this system is mine", in the same way
>>>>>> traditional video drivers can claim the VRAM PCI resource.
>>>>>>
>>>>>>
>>>>>>              
>>>>> CMA could reserve memory region for a specific driver so DRM Client
>>>>>
>>>>>            
>>>> could
>>>>
>>>>          
>>>>> request memory allocation from only the region.
>>>>>
>>>>>
>>>>>            
>>>>>> This is to avoid the possibility that future drivers that need CMA
>>>>>>              
>> will
>>      
>>>>>> be vulnerable to DOS-attacks from ill-behaved DRI clients.
>>>>>>
>>>>>>
>>>>>>              
>>>>> Thomas, if any application has root authority for ill-purpose then
>>>>>            
>> isn't
>>      
>>>>>            
>>>> it
>>>>
>>>>          
>>>>> possible to be vulnerable to DOS-attacks? I think DRM_AUTH means root
>>>>> authority. I know DRM Framework gives any root application DRM_AUTH
>>>>> authority for compatibility.
>>>>>
>>>>>            
>>>> DRM_AUTH just means that the client has authenticated w/ X11 (meaning
>>>> that it has permission to connect to x server)..
>>>>
>>>>
>>>>          
>>> Yes, I understood so. but see drm_open_helper() of drm_fops.c file
>>>        
>> please.
>>      
>>> in this function, you can see a line below.
>>> /* for compatibility root is always authenticated */
>>> priv->authenticated = capable(CAP_SYS_ADMIN)
>>>
>>> I think the code above says that any application with root permission is
>>> authenticated.
>>>
>>>
>>>        
>> Yes, that is true. A root client may be assumed to have AUTH
>> permissions, but the inverse does not hold, meaning that an AUTH client
>> may *not* be assumed to have root permissions. I think there is a
>> ROOT_ONLY ioctl flag for that.
>>
>> The problem I'm seeing compared to other drivers is the following:
>>
>> Imagine for example that you have a disc driver that allocates temporary
>> memory out of the same DMA pool as the DRM driver.
>>
>> Now you have a video player that is a DRM client. It contains a security
>> flaw and is compromized by somebody trying to play a specially crafted
>> video stream. The video player starts to allocate gem buffers until it
>> receives an -ENOMEM. Then it stops allocating and does nothing.
>>
>> Now the system tries an important disc access (paging for example). This
>> fails, because the video player has exhausted all DMA memory and the
>> disc driver fails to allocate.
>>
>> The system is dead.
>>
>>      
> Ok, I understood. but in case of using CMA, DRM driver would have private
> memory pool which is reserved only for itself. so although the video player
> requested gem buffer allocation until it receives an -ENOMEM and then the
> system is try to access the disc, it would work fine. because the region
> requested by system and the region requested by DRM driver could entirely be
> separated. DRM driver wouldn't have implications for the system region.
>
>    
>> The point is:
>>
>> If there is a chance that other drivers will use the same DMA/CMA pool
>> as the DRM driver, DRM must leave enough DMA/CMA memory for those
>> drivers to work.
>>
>> The difference compared to other drm drivers:
>>
>> There are other drm drivers that work the same way, with a static
>> allocator. For example "via" and "sis". But those drivers completely
>> claim the resources they are using. Nobody else is expected to use VRAM
>> / AGP.
>>
>>      
> I think if we use private cma region for DRM driver then it is similar to
> via and sis way.
>
>    

In that case, I agree the driver should be OK.

...

Thanks,
Thomas

[PATCH v4] DRM: add DRM Driver for Samsung SoC EXYNOS4210.

[Inki Dae]

Hi Thomas.
Sorry for being late.

> -----Original Message-----
> From: Thomas Hellstrom [mailto:thomas at shipmail.org]
> Sent: Saturday, September 10, 2011 11:04 PM
> To: Inki Dae
> Cc: airlied at linux.ie; dri-devel at lists.freedesktop.org;
> sw0312.kim at samsung.com; kyungmin.park at samsung.com; linux-arm-
> kernel at lists.infradead.org
> Subject: Re: [PATCH v4] DRM: add DRM Driver for Samsung SoC EXYNOS4210.
> 
> On 09/09/2011 01:38 PM, Inki Dae wrote:
> > This patch is a DRM Driver for Samsung SoC Exynos4210 and now enables
> > only FIMD yet but we will add HDMI support also in the future.
> >
> > from now on, I will remove RFC prefix because I think we have got
> comments
> > enough.
> >
> > this patch is based on git repository below:
> > git://git.kernel.org/pub/scm/linux/kernel/git/airlied/drm-2.6.git,
> > branch name: drm-next
> > commit-id: bcc65fd8e929a9d9d34d814d6efc1d2793546922
> >
> > you can refer to our working repository below:
> > http://git.infradead.org/users/kmpark/linux-2.6-samsung
> > branch name: samsung-drm
> >
> > We tried to re-use lowlevel codes of the FIMD driver(s3c-fb.c
> > based on Linux framebuffer) but couldn't so because lowlevel codes
> > of s3c-fb.c are included internally and so FIMD module of this driver
> has
> > its own lowlevel codes.
> >
> > We used GEM framework for buffer management and DMA APIs(dma_alloc_*)
> > for buffer allocation. by using DMA API, we could use CMA later.
> >
> > Refer to this link for CMA(Continuous Memory Allocator):
> > http://lkml.org/lkml/2011/7/20/45
> >
> > this driver supports only physically continuous memory(non-iommu).
> >
> > Links to previous versions of the patchset:
> > v1:<  https://lwn.net/Articles/454380/>
> > v2:<  http://www.spinics.net/lists/kernel/msg1224275.html>
> > v3:<  http://www.gossamer-threads.com/lists/linux/kernel/1423684>
> >
> > Changelog v2:
> > DRM: add DRM_IOCTL_SAMSUNG_GEM_MMAP ioctl command.
> >
> >      this feature maps user address space to physical memory region
> >      once user application requests DRM_IOCTL_SAMSUNG_GEM_MMAP ioctl.
> >
> > DRM: code clean and add exception codes.
> >
> > Changelog v3:
> > DRM: Support multiple irq.
> >
> >      FIMD and HDMI have their own irq handler but DRM Framework can
> regiter
> >      only one irq handler this patch supports mutiple irq for Samsung
SoC.
> >
> > DRM: Consider modularization.
> >
> >      each DRM, FIMD could be built as a module.
> >
> > DRM: Have indenpendent crtc object.
> >
> >      crtc isn't specific to SoC Platform so this patch gets a crtc
> >      to be used as common object.
> >      created crtc could be attached to any encoder object.
> >
> > DRM: code clean and add exception codes.
> >
> > Changelog v4:
> > DRM: remove is_defult from samsung_fb.
> >
> >      is_default isn't used for default framebuffer.
> >
> > DRM: code refactoring to fimd module.
> >      this patch is be considered with multiple display objects and
> >      would use its own request_irq() to register a irq handler instead
of
> >      drm framework's one.
> >
> > DRM: remove find_samsung_drm_gem_object()
> >
> > DRM: move kernel private data structures and definitions to driver
> folder.
> >
> >      samsung_drm.h would contain only public information for userspace
> >      ioctl interface.
> >
> > DRM: code refactoring to gem modules.
> >      buffer module isn't dependent of gem module anymore.
> >
> > DRM: fixed security issue.
> >
> > DRM: remove encoder porinter from specific connector.
> >
> >      samsung connector doesn't need to have generic encoder.
> >
> > DRM: code clean and add exception codes.
> >
> > Signed-off-by: Inki Dae<inki.dae@samsung.com>
> > Signed-off-by: Joonyoung Shim<jy0922.shim@samsung.com>
> > Signed-off-by: SeungWoo Kim<sw0312.kim@samsung.com>
> > Signed-off-by: kyungmin.park<kyungmin.park@samsung.com>
> > ---
> >
> > +static struct drm_ioctl_desc samsung_ioctls[] = {
> > +	DRM_IOCTL_DEF_DRV(SAMSUNG_GEM_CREATE, samsung_drm_gem_create_ioctl,
> > +			DRM_UNLOCKED | DRM_AUTH),
> >
> 
> Hi!
> 
> With reference my previous security comment.
> 
> Let's say you have a compromised video player running as a DRM client,
> that tries to repeatedly allocate huge GEM buffers...
> 
> What will happen when all DMA memory is exhausted? Will this cause other
> device drivers to see an OOM, or only DRM?
> 

Out driver allocates physically continuous memory from the region reserved
by CMA or DMA Region in case of not using CMA. if all memory is exhausted at
allocation moment, it just returns "Allocation fail" and then the
application would be terminated.

> The old DRI model basically allowed any authorized DRI client to exhaust
> video ram or AGP memory, but never system memory. Newer DRI drivers
> typically only allow DRI masters to do that.
> as
> 

I thought it's enough for an authorized DRI client to request memory
allocation. because an authorized DRI client has root authority. I know, in
case of v4l2 based device drivers, the applications can request buffer
allocation through v4l2 interface, REQBUFS. and also if memory allocation is
repeated, we can see any error in the near future. and I wonder why only DRM
master could do that. it might be my missing points so I am happy to you
give me your words?

> I don't think an authorized DRI client should be able to easily exhaust
> resources (DMA memory) used by other device drivers causing them to fail.
> 
> /Thomas
> 
> 
> > +	DRM_IOCTL_DEF_DRV(SAMSUNG_GEM_MAP_OFFSET,
> > +			samsung_drm_gem_map_offset_ioctl, DRM_UNLOCKED |
> > +				DRM_AUTH),
> > +	DRM_IOCTL_DEF_DRV(SAMSUNG_GEM_MMAP,
> > +			samsung_drm_gem_mmap_ioctl, DRM_UNLOCKED |
DRM_AUTH),
> > +};
> >
> >
> 

[PATCH v4] DRM: add DRM Driver for Samsung SoC EXYNOS4210.

[Konrad Rzeszutek Wilk]

On Thu, Sep 15, 2011 at 06:37:39PM +0900, Inki Dae wrote:
> Hello, Konrad Rzeszutek Wilk.
> 
> Your review and comments are so very detail. it was very helpful. thank you
> again.

.. snip ..
> > > +		/* FIXME: error check */
> > 
> > You might as well do it now - before you do the next posting.
> > 
> 
> Ok, get it. you mean, send this patch modified now, not next posting.?

When you are ready to send V5 (hopefully the last version) of these patches.

.. snip..
> > > + * @ops: pointer to samsung_drm_overlay_ops.
> > > + *
> > > + * this structure is common to Samsung SoC and would be copied
> > > + * to hardware specific overlay info.
> > 
> > Uh?
> 
> This means that contents of samsung_drm_overlay object will copied to fimd's
> one, struct fimd_win_data.

Ok, lets use that comment instead then.
> 
> > > + */
> > > +struct samsung_drm_overlay {
> > > +	unsigned int offset_x;
> > > +	unsigned int offset_y;
> > > +	unsigned int width;
> > > +	unsigned int height;
> > > +	unsigned int paddr;
> > 
> > You don't want to use 'dma_addr_t' ?
> > 
> 
> You are right. I will correct it. thank you.
> 
> > > +	void __iomem *vaddr;
> > > +	unsigned int buf_off;
> > > +	unsigned int end_buf_off;
> > > +	unsigned int buf_offsize;
> > > +	unsigned int line_size;
> > > +
> > > +	bool default_win;
> > > +	bool color_key;
> > > +	unsigned int index_color;
> > > +	bool local_path;
> > > +	bool transparency;
> > > +	bool activated;
> > > +};
> > > +
> > > +/**
> > > + * Samsung DRM Display Structure.
> > > + *	- this structure is common to analog tv, digital tv and lcd panel.
> > > + *
> > > + * @dev: pointer to specific device object.
> > 
> > ??
> 
> Ah, this should be removed. Thank you.
> 
> > > + * @is_connected: check for that display is connected or not.
> > > + * @get_edid: get edid modes from display driver.
> > > + * @get_timing: get timing object from display driver.
> > > + * @check_timing: check if timing is valid or not.
> > > + * @power_on: display device on or off.
> > > + */
> > > +struct samsung_drm_display {
> > > +	unsigned int type;
> > > +	bool (*is_connected)(struct device *dev);
> > > +	int (*get_edid)(struct device *dev, struct drm_connector *connector,
> > > +				u8 *edid, int len);
> > > +	void *(*get_timing)(struct device *dev);
> > > +	int (*check_timing)(struct device *dev, void *timing);
> > > +	int (*power_on)(struct device *dev, int mode);
> > > +};
> > > +
> > > +/**
> > > + * Samsung drm manager ops
> > > + *
> > > + * @mode_set: convert drm_display_mode to hw specific display mode and
> > > + *	      would be called by encoder->mode_set().
> > > + * @commit: set current hw specific display mode to hw.
> > > + * @enable_vblank: specific driver callback for enabling vblank
> > interrupt.
> > > + * @disable_vblank: specific driver callback for disabling vblank
> > interrupt.
> > > + */
> > > +struct samsung_drm_manager_ops {
> > > +	void (*mode_set)(struct device *subdrv_dev, void *mode);
> > > +	void (*commit)(struct device *subdrv_dev);
> > > +	int (*enable_vblank)(struct device *subdrv_dev);
> > > +	void (*disable_vblank)(struct device *subdrv_dev);
> > > +};
> > > +
> > > +/**
> > > + * Samsung drm common manager structure.
> > > + *
> > > + * @dev: pointer to device object for subdrv device driver.
> > > + * @ops: ops pointer to samsung drm common framebuffer.
> > > + *	 ops of fimd or hdmi driver should be set to this ones.
> > > + */
> > > +struct samsung_drm_manager {
> > > +	struct device *dev;
> > > +	int pipe;
> > 
> > No comment for that?
> 
> Ok, get it. I will add some comment for that. thank you.
> 
> > > +	struct samsung_drm_manager_ops *ops;
> > > +	struct samsung_drm_overlay_ops *overlay_ops;
> > > +	struct samsung_drm_display *display;
> > 
> > And you are missing the comments for these two.
> 
> Also.
> 
> > > +};
> > > +
> > > +/**
> > 
> > I just noticed it, but the '**' is not the kerneldoc
> > style comment. You might want to remove them from all the files.
> > 
> 
> Ok, get it. thank you.
> 
> > > + * Samsung drm private structure.
> > 
> > Ok, you are defining it in a public header so you should at
> > least document what the fields mean.
> > 
> 
> Ok, get it. I will add comments to all fields and thank you for your
> pointing.
> 
> > If you don't want the public to use it - make another header
> > file, helpfully called 'xxx_private.h'
> > 
> > 
> > > + */
> > > +struct samsung_drm_private {
> > > +	struct drm_fb_helper *fb_helper;
> > > +
> > > +	/* for pageflip */
> > > +	struct list_head pageflip_event_list;
> > > +	bool pageflip_event;
> > > +
> > > +	struct drm_crtc *crtc[MAX_CRTC];
> > > +
> > > +	/* add some structures. */
> > 
> > Umm, which ones?
> > 
> 
> This comment will be removed. Thank you.
> 
> > > +};
> > > +
> > > +struct samsung_drm_subdrv {
> > > +	struct list_head list;
> > > +	struct drm_device *drm_dev;
> > > +
> > > +	/* driver ops */
> > > +	int (*probe)(struct drm_device *dev);
> > > +	void (*remove)(struct drm_device *dev);
> > > +
> > > +	struct samsung_drm_manager manager;
> > > +	struct drm_encoder *encoder;
> > > +	struct drm_connector *connector;
> > > +};
> > > +
> > > +int samsung_drm_device_register(struct drm_device *dev);
> > > +void samsung_drm_device_unregister(struct drm_device *dev);
> > > +int samsung_drm_subdrv_register(struct samsung_drm_subdrv *drm_subdrv);
> > > +void samsung_drm_subdrv_unregister(struct samsung_drm_subdrv
> > *drm_subdrv);
> > > +
> > > +#endif
> > > diff --git a/drivers/gpu/drm/samsung/samsung_drm_encoder.c
> > b/drivers/gpu/drm/samsung/samsung_drm_encoder.c
> > > new file mode 100644
> > > index 0000000..c875968
> > > --- /dev/null
> > > +++ b/drivers/gpu/drm/samsung/samsung_drm_encoder.c
> > > @@ -0,0 +1,261 @@
> > > +/*
> > > + * Copyright (c) 2011 Samsung Electronics Co., Ltd.
> > > + * Authors:
> > > + *	Inki Dae <inki.dae@samsung.com>
> > > + *	Joonyoung Shim <jy0922.shim@samsung.com>
> > > + *	SeungWoo Kim <sw0312.kim@samsung.com>
> > > + *
> > > + * Permission is hereby granted, free of charge, to any person
> > obtaining a
> > > + * copy of this software and associated documentation files (the
> > "Software"),
> > > + * to deal in the Software without restriction, including without
> > limitation
> > > + * the rights to use, copy, modify, merge, publish, distribute,
> > sublicense,
> > > + * and/or sell copies of the Software, and to permit persons to whom
> > the
> > > + * Software is furnished to do so, subject to the following conditions:
> > > + *
> > > + * The above copyright notice and this permission notice (including the
> > next
> > > + * paragraph) shall be included in all copies or substantial portions
> > of the
> > > + * Software.
> > > + *
> > > + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
> > EXPRESS OR
> > > + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
> > MERCHANTABILITY,
> > > + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT
> > SHALL
> > > + * VA LINUX SYSTEMS AND/OR ITS SUPPLIERS BE LIABLE FOR ANY CLAIM,
> > DAMAGES OR
> > > + * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR
> OTHERWISE,
> > > + * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE
> > OR
> > > + * OTHER DEALINGS IN THE SOFTWARE.
> > > + */
> > > +
> > > +#include "drmP.h"
> > > +#include "drm_crtc_helper.h"
> > > +
> > > +#include "samsung_drm_drv.h"
> > > +#include "samsung_drm_crtc.h"
> > > +#include "samsung_drm_encoder.h"
> > > +
> > > +#define to_samsung_encoder(x)	container_of(x, struct
> > samsung_drm_encoder,\
> > > +				drm_encoder)
> > > +
> > > +struct samsung_drm_encoder {
> > > +	struct drm_encoder		drm_encoder;
> > > +	struct samsung_drm_manager	*manager;
> > > +};
> > > +
> > > +static void samsung_drm_encoder_dpms(struct drm_encoder *encoder, int
> > mode)
> > > +{
> > > +	struct drm_device *dev = encoder->dev;
> > > +	struct drm_connector *connector;
> > > +	struct samsung_drm_manager *manager =
> > samsung_drm_get_manager(encoder);
> > > +
> > > +	DRM_DEBUG_KMS("%s, encoder dpms: %d\n", __FILE__, mode);
> > > +
> > > +	list_for_each_entry(connector, &dev->mode_config.connector_list,
> > head) {
> > > +		if (connector->encoder == encoder) {
> > > +			struct samsung_drm_display *display =
> manager->display;
> > > +
> > > +			if (display && display->power_on)
> > > +				display->power_on(manager->dev, mode);
> > > +		}
> > > +	}
> > > +}
> > > +
> > > +static bool
> > > +samsung_drm_encoder_mode_fixup(struct drm_encoder *encoder,
> > > +			       struct drm_display_mode *mode,
> > > +			       struct drm_display_mode *adjusted_mode)
> > > +{
> > > +	DRM_DEBUG_KMS("%s\n", __FILE__);
> > > +
> > > +	return true;
> > > +}
> > > +
> > > +static void samsung_drm_encoder_mode_set(struct drm_encoder *encoder,
> > > +					 struct drm_display_mode *mode,
> > > +					 struct drm_display_mode
> *adjusted_mode)
> > > +{
> > > +	struct drm_device *dev = encoder->dev;
> > > +	struct drm_connector *connector;
> > > +	struct samsung_drm_manager *manager =
> > samsung_drm_get_manager(encoder);
> > > +	struct samsung_drm_manager_ops *manager_ops = manager->ops;
> > > +	struct samsung_drm_overlay_ops *overlay_ops = manager->overlay_ops;
> > > +	struct samsung_drm_overlay *overlay = get_samsung_drm_overlay(dev,
> > > +						encoder->crtc);
> > > +
> > > +	DRM_DEBUG_KMS("%s\n", __FILE__);
> > > +
> > > +	mode = adjusted_mode;
> > > +
> > > +	list_for_each_entry(connector, &dev->mode_config.connector_list,
> > head) {
> > > +		if (connector->encoder == encoder) {
> > > +			if (manager_ops && manager_ops->mode_set)
> > > +				manager_ops->mode_set(manager->dev, mode);
> > > +
> > > +			if (overlay_ops && overlay_ops->mode_set)
> > > +				overlay_ops->mode_set(manager->dev,
> overlay);
> > > +		}
> > > +	}
> > > +}
> > > +
> > > +static void samsung_drm_encoder_prepare(struct drm_encoder *encoder)
> > > +{
> > > +	DRM_DEBUG_KMS("%s\n", __FILE__);
> > > +}
> > > +
> > > +static void samsung_drm_encoder_commit(struct drm_encoder *encoder)
> > > +{
> > > +	struct samsung_drm_manager *manager =
> > samsung_drm_get_manager(encoder);
> > > +	struct samsung_drm_manager_ops *manager_ops = manager->ops;
> > > +	struct samsung_drm_overlay_ops *overlay_ops = manager->overlay_ops;
> > > +
> > > +	DRM_DEBUG_KMS("%s\n", __FILE__);
> > > +
> > > +	if (manager_ops && manager_ops->commit)
> > > +		manager_ops->commit(manager->dev);
> > > +
> > > +	if (overlay_ops && overlay_ops->commit)
> > > +		overlay_ops->commit(manager->dev);
> > > +}
> > > +
> > > +static struct drm_crtc *
> > > +samsung_drm_encoder_get_crtc(struct drm_encoder *encoder)
> > > +{
> > > +	return encoder->crtc;
> > > +}
> > > +
> > > +static struct drm_encoder_helper_funcs samsung_encoder_helper_funcs = {
> > > +	.dpms		= samsung_drm_encoder_dpms,
> > > +	.mode_fixup	= samsung_drm_encoder_mode_fixup,
> > > +	.mode_set	= samsung_drm_encoder_mode_set,
> > > +	.prepare	= samsung_drm_encoder_prepare,
> > > +	.commit		= samsung_drm_encoder_commit,
> > > +	.get_crtc	= samsung_drm_encoder_get_crtc,
> > > +};
> > > +
> > > +static void samsung_drm_encoder_destroy(struct drm_encoder *encoder)
> > > +{
> > > +	struct samsung_drm_encoder *samsung_encoder =
> > > +		to_samsung_encoder(encoder);
> > > +
> > > +	DRM_DEBUG_KMS("%s\n", __FILE__);
> > > +
> > > +	samsung_encoder->manager->pipe = -1;
> > > +
> > > +	drm_encoder_cleanup(encoder);
> > > +	encoder->dev->mode_config.num_encoder--;
> > > +	kfree(samsung_encoder);
> > > +}
> > > +
> > > +static struct drm_encoder_funcs samsung_encoder_funcs = {
> > > +	.destroy = samsung_drm_encoder_destroy,
> > > +};
> > > +
> > > +struct drm_encoder *
> > > +samsung_drm_encoder_create(struct drm_device *dev,
> > > +			   struct samsung_drm_manager *manager,
> > > +			   unsigned int possible_crtcs)
> > > +{
> > > +	struct drm_encoder *encoder;
> > > +	struct samsung_drm_encoder *samsung_encoder;
> > > +
> > > +	DRM_DEBUG_KMS("%s\n", __FILE__);
> > > +
> > > +	if (!manager || !possible_crtcs)
> > > +		return NULL;
> > > +
> > > +	if (!manager->dev)
> > > +		return NULL;
> > > +
> > > +	samsung_encoder = kzalloc(sizeof(*samsung_encoder), GFP_KERNEL);
> > > +	if (!samsung_encoder) {
> > > +		DRM_ERROR("failed to allocate encoder\n");
> > > +		return NULL;
> > > +	}
> > > +
> > > +	samsung_encoder->manager = manager;
> > > +	encoder = &samsung_encoder->drm_encoder;
> > > +	encoder->possible_crtcs = possible_crtcs;
> > > +
> > > +	DRM_DEBUG_KMS("possible_crtcs = 0x%x\n", encoder->possible_crtcs);
> > > +
> > > +	drm_encoder_init(dev, encoder, &samsung_encoder_funcs,
> > > +			DRM_MODE_ENCODER_TMDS);
> > > +
> > > +	drm_encoder_helper_add(encoder, &samsung_encoder_helper_funcs);
> > > +
> > > +	DRM_DEBUG_KMS("encoder has been created\n");
> > > +
> > > +	return encoder;
> > > +}
> > > +
> > > +struct samsung_drm_manager *samsung_drm_get_manager(struct drm_encoder
> > *encoder)
> > > +{
> > > +	return to_samsung_encoder(encoder)->manager;
> > > +}
> > > +
> > > +void samsung_drm_fn_encoder(struct drm_crtc *crtc, void *data,
> > > +			    void (*fn)(struct drm_encoder *, void *))
> > > +{
> > > +	struct drm_device *dev = crtc->dev;
> > > +	struct drm_encoder *encoder;
> > > +
> > > +	list_for_each_entry(encoder, &dev->mode_config.encoder_list, head)
> > {
> > > +		if (encoder->crtc != crtc)
> > > +			continue;
> > > +
> > > +		fn(encoder, data);
> > > +	}
> > > +}
> > > +
> > > +void samsung_drm_enable_vblank(struct drm_encoder *encoder, void *data)
> > > +{
> > > +	struct samsung_drm_manager *manager =
> > > +		to_samsung_encoder(encoder)->manager;
> > > +	struct samsung_drm_manager_ops *manager_ops = manager->ops;
> > > +	int crtc = *(int *)data;
> > > +
> > > +	if (manager->pipe == -1)
> > > +		manager->pipe = crtc;
> > > +
> > > +	/* old_crtc checking needs? FIXME!!! */
> > 
> > Is this still pertient?
> > 
> 
> I will remove this comment after looking over more. Thank you.
> 
> > > +
> > > +	if (manager_ops->enable_vblank)
> > > +		manager_ops->enable_vblank(manager->dev);
> > > +}
> > > +
> > > +void samsung_drm_disable_vblank(struct drm_encoder *encoder, void
> *data)
> > > +{
> > > +	struct samsung_drm_manager *manager =
> > > +		to_samsung_encoder(encoder)->manager;
> > > +	struct samsung_drm_manager_ops *manager_ops = manager->ops;
> > > +	int crtc = *(int *)data;
> > 
> > You don't want to check whether data is NULL before you
> > derefernce it?
> > 
> 
> Ok, get it. thank you.
> 
> > > +
> > > +	if (manager->pipe == -1)
> > > +		manager->pipe = crtc;
> > > +
> > > +	/* old_crtc checking needs? FIXME!!! */
> > 
> > I am not really sure what it means.. It probably means something to you -
> > but
> > perhaps it might make sense to expand it in case somebody else wants to
> > implement this?
> > 
> 
> As I mentioned above, I will remove this comment after looking over more.
> Thank you.
> 
> > > +
> > > +	if (manager_ops->disable_vblank)
> > > +		manager_ops->disable_vblank(manager->dev);
> > > +}
> > > +
> > > +void samsung_drm_encoder_crtc_commit(struct drm_encoder *encoder, void
> > *data)
> > > +{
> > > +	struct samsung_drm_manager *manager =
> > > +		to_samsung_encoder(encoder)->manager;
> > > +	struct samsung_drm_overlay_ops *overlay_ops = manager->overlay_ops;
> > > +
> > > +	overlay_ops->commit(manager->dev);
> > > +}
> > > +
> > > +void samsung_drm_encoder_crtc_mode_set(struct drm_encoder *encoder,
> > void *data)
> > > +{
> > > +	struct samsung_drm_manager *manager =
> > > +		to_samsung_encoder(encoder)->manager;
> > > +	struct samsung_drm_overlay_ops *overlay_ops = manager->overlay_ops;
> > > +	struct samsung_drm_overlay *overlay = data;
> > > +
> > > +	overlay_ops->mode_set(manager->dev, overlay);
> > > +}
> > > +
> > > +MODULE_AUTHOR("Inki Dae <inki.dae@samsung.com>");
> > 
> > You don't want to include the rest of the authors?
> 
> Definitely no. I will add other authors. Thank you.
> 
> > > +MODULE_DESCRIPTION("Samsung SoC DRM Encoder Driver");
> > > +MODULE_LICENSE("GPL");
> > > diff --git a/drivers/gpu/drm/samsung/samsung_drm_encoder.h
> > b/drivers/gpu/drm/samsung/samsung_drm_encoder.h
> > > new file mode 100644
> > > index 0000000..99040b2
> > > --- /dev/null
> > > +++ b/drivers/gpu/drm/samsung/samsung_drm_encoder.h
> > > @@ -0,0 +1,45 @@
> > > +/*
> > > + * Copyright (c) 2011 Samsung Electronics Co., Ltd.
> > > + * Authors:
> > > + *	Inki Dae <inki.dae@samsung.com>
> > > + *	Joonyoung Shim <jy0922.shim@samsung.com>
> > > + *	SeungWoo Kim <sw0312.kim@samsung.com>
> > > + *
> > > + * Permission is hereby granted, free of charge, to any person
> > obtaining a
> > > + * copy of this software and associated documentation files (the
> > "Software"),
> > > + * to deal in the Software without restriction, including without
> > limitation
> > > + * the rights to use, copy, modify, merge, publish, distribute,
> > sublicense,
> > > + * and/or sell copies of the Software, and to permit persons to whom
> > the
> > > + * Software is furnished to do so, subject to the following conditions:
> > > + *
> > > + * The above copyright notice and this permission notice (including the
> > next
> > > + * paragraph) shall be included in all copies or substantial portions
> > of the
> > > + * Software.
> > > + *
> > > + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
> > EXPRESS OR
> > > + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
> > MERCHANTABILITY,
> > > + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT
> > SHALL
> > > + * VA LINUX SYSTEMS AND/OR ITS SUPPLIERS BE LIABLE FOR ANY CLAIM,
> > DAMAGES OR
> > > + * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR
> OTHERWISE,
> > > + * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE
> > OR
> > > + * OTHER DEALINGS IN THE SOFTWARE.
> > > + */
> > > +
> > > +#ifndef _SAMSUNG_DRM_ENCODER_H_
> > > +#define _SAMSUNG_DRM_ENCODER_H_
> > > +
> > > +struct samsung_drm_manager;
> > > +
> > > +struct drm_encoder *samsung_drm_encoder_create(struct drm_device *dev,
> > > +					       struct samsung_drm_manager
> *mgr,
> > > +					       unsigned int possible_crtcs);
> > > +struct samsung_drm_manager *
> > > +samsung_drm_get_manager(struct drm_encoder *encoder);
> > > +void samsung_drm_fn_encoder(struct drm_crtc *crtc, void *data,
> > > +			    void (*fn)(struct drm_encoder *, void *));
> > > +void samsung_drm_enable_vblank(struct drm_encoder *encoder, void
> *data);
> > > +void samsung_drm_disable_vblank(struct drm_encoder *encoder, void
> > *data);
> > > +void samsung_drm_encoder_crtc_commit(struct drm_encoder *encoder, void
> > *data);
> > > +void samsung_drm_encoder_crtc_mode_set(struct drm_encoder *encoder,
> > void *data);
> > > +
> > > +#endif
> > > diff --git a/drivers/gpu/drm/samsung/samsung_drm_fb.c
> > b/drivers/gpu/drm/samsung/samsung_drm_fb.c
> > > new file mode 100644
> > > index 0000000..f087ecf
> > > --- /dev/null
> > > +++ b/drivers/gpu/drm/samsung/samsung_drm_fb.c
> > > @@ -0,0 +1,262 @@
> > > +/*
> > > + * Copyright (c) 2011 Samsung Electronics Co., Ltd.
> > > + * Authors:
> > > + *	Inki Dae <inki.dae@samsung.com>
> > > + *	Joonyoung Shim <jy0922.shim@samsung.com>
> > > + *	SeungWoo Kim <sw0312.kim@samsung.com>
> > > + *
> > > + * Permission is hereby granted, free of charge, to any person
> > obtaining a
> > > + * copy of this software and associated documentation files (the
> > "Software"),
> > > + * to deal in the Software without restriction, including without
> > limitation
> > > + * the rights to use, copy, modify, merge, publish, distribute,
> > sublicense,
> > > + * and/or sell copies of the Software, and to permit persons to whom
> > the
> > > + * Software is furnished to do so, subject to the following conditions:
> > > + *
> > > + * The above copyright notice and this permission notice (including the
> > next
> > > + * paragraph) shall be included in all copies or substantial portions
> > of the
> > > + * Software.
> > > + *
> > > + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
> > EXPRESS OR
> > > + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
> > MERCHANTABILITY,
> > > + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT
> > SHALL
> > > + * VA LINUX SYSTEMS AND/OR ITS SUPPLIERS BE LIABLE FOR ANY CLAIM,
> > DAMAGES OR
> > > + * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR
> OTHERWISE,
> > > + * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE
> > OR
> > > + * OTHER DEALINGS IN THE SOFTWARE.
> > > + */
> > > +
> > > +#include "drmP.h"
> > > +#include "drm_crtc.h"
> > > +#include "drm_crtc_helper.h"
> > > +
> > > +#include "samsung_drm_fb.h"
> > > +#include "samsung_drm_buf.h"
> > > +#include "samsung_drm_gem.h"
> > > +
> > > +#define to_samsung_fb(x)	container_of(x, struct samsung_drm_fb,
> > fb)
> > > +
> > > +struct samsung_drm_fb {
> > > +	struct drm_framebuffer		fb;
> > > +	struct samsung_drm_gem_obj	*samsung_gem_obj;
> > > +
> > > +	unsigned int			fb_size;
> > > +	dma_addr_t			paddr;
> > > +	void __iomem			*vaddr;
> > > +};
> > > +
> > > +static void samsung_drm_fb_destroy(struct drm_framebuffer *fb)
> > > +{
> > > +	struct samsung_drm_fb *samsung_fb = to_samsung_fb(fb);
> > > +
> > > +	DRM_DEBUG_KMS("%s\n", __FILE__);
> > > +
> > > +	drm_framebuffer_cleanup(fb);
> > > +
> > > +	/* default framebuffer has no gem object so it releases buffer. */
> > 
> > What is 'it' ?
> > 
> 
> Please, ignore 'it'. and  I will correct this comment. Thank you.
> 
> > > +	if (!samsung_fb->samsung_gem_obj)
> > > +		samsung_drm_buf_destroy(fb->dev,
> > > +				samsung_fb->samsung_gem_obj->entry);
> > > +
> > > +	kfree(samsung_fb);
> > 
> > samsung_fb = NULL;
> > 
> 
> Thank you.
> 
> > > +}
> > > +
> > > +static int samsung_drm_fb_create_handle(struct drm_framebuffer *fb,
> > > +					struct drm_file *file_priv,
> > > +					unsigned int *handle)
> > > +{
> > > +	struct samsung_drm_fb *samsung_fb = to_samsung_fb(fb);
> > > +
> > > +	DRM_DEBUG_KMS("%s\n", __FILE__);
> > > +
> > > +	return drm_gem_handle_create(file_priv,
> > > +					&samsung_fb->samsung_gem_obj->base,
> > > +					handle);
> > > +}
> > > +
> > > +static int samsung_drm_fb_dirty(struct drm_framebuffer *fb,
> > > +				struct drm_file *file_priv, unsigned flags,
> > > +				unsigned color, struct drm_clip_rect *clips,
> > > +				unsigned num_clips)
> > > +{
> > > +	DRM_DEBUG_KMS("%s\n", __FILE__);
> > > +
> > > +	/*
> > > +	 * update framebuffer and its hardware.
> > > +	 * - this callback would be called by user application
> > > +	 *	with DRM_IOCTL_MODE_DIRTYFB command.
> > > +	 *
> > > +	 * ps. Userspace can notify the driver via this callback
> > > +	 * that an area of the framebuffer has been changed then should
> > > +	 * be flushed to the display hardware.
> > > +	 */
> > > +
> > > +	return 0;
> > > +}
> > > +
> > > +static struct drm_framebuffer_funcs samsung_drm_fb_funcs = {
> > > +	.destroy	= samsung_drm_fb_destroy,
> > > +	.create_handle	= samsung_drm_fb_create_handle,
> > > +	.dirty		= samsung_drm_fb_dirty,
> > > +};
> > > +
> > > +static struct drm_framebuffer *
> > > +samsung_drm_fb_init(struct drm_file *file_priv, struct drm_device *dev,
> > > +		    struct drm_mode_fb_cmd *mode_cmd)
> > > +{
> > > +	struct samsung_drm_fb *samsung_fb;
> > > +	struct drm_framebuffer *fb;
> > > +	struct samsung_drm_gem_obj *samsung_gem_obj = NULL;
> > > +	struct drm_gem_object *obj;
> > > +	unsigned int size;
> > > +	int ret;
> > > +
> > > +	DRM_DEBUG_KMS("%s\n", __FILE__);
> > > +
> > > +	mode_cmd->pitch = max(mode_cmd->pitch,
> > > +			mode_cmd->width * (mode_cmd->bpp >> 3));
> > > +
> > > +	DRM_LOG_KMS("drm fb create(%dx%d)\n",
> > > +			mode_cmd->width, mode_cmd->height);
> > > +
> > > +	samsung_fb = kzalloc(sizeof(*samsung_fb), GFP_KERNEL);
> > > +	if (!samsung_fb) {
> > > +		DRM_ERROR("failed to allocate samsung drm framebuffer.\n");
> > > +		return ERR_PTR(-ENOMEM);
> > > +	}
> > > +
> > > +	fb = &samsung_fb->fb;
> > > +	ret = drm_framebuffer_init(dev, fb, &samsung_drm_fb_funcs);
> > > +	if (ret) {
> > > +		DRM_ERROR("failed to initialize framebuffer.\n");
> > > +		goto err_init;
> > > +	}
> > > +
> > > +	DRM_LOG_KMS("create: fb id: %d\n", fb->base.id);
> > > +
> > > +	size = mode_cmd->pitch * mode_cmd->height;
> > > +
> > > +	/*
> > > +	 * mode_cmd->handle could be pointer to a buffer allocated by user
> > > +	 * application using KMS library.
> > > +	 */
> > > +	if (!mode_cmd->handle) {
> > > +		/*
> > > +		 * in case that file_priv is NULL, it allocates only buffer
> > and
> > > +		 * this buffer would be used for default framebuffer.
> > > +		 */
> > > +		if (!file_priv) {
> > > +			struct samsung_drm_buf_entry *entry;
> > > +
> > > +			entry = samsung_drm_buf_create(dev, size);
> > > +			if (IS_ERR(entry)) {
> > > +				ret = PTR_ERR(entry);
> > > +				goto err_buf_create;
> > > +			}
> > > +
> > > +			samsung_fb->vaddr = entry->vaddr;
> > > +			samsung_fb->paddr = entry->paddr;
> > > +
> > > +			DRM_LOG_KMS("default fb: paddr = 0x%x, size =
> 0x%x\n",
> > > +					entry->paddr, size);
> > > +
> > > +			goto out;
> > > +		} else {
> > > +			samsung_gem_obj = samsung_drm_gem_create(file_priv,
> > dev,
> > > +							size,
> > > +							&mode_cmd->handle);
> > > +			if (IS_ERR(samsung_gem_obj)) {
> > > +				ret = PTR_ERR(samsung_gem_obj);
> > > +				goto err_gem_create;
> > > +			}
> > > +		}
> > > +	} else {
> > > +		obj = drm_gem_object_lookup(dev, file_priv, mode_cmd-
> > >handle);
> > > +		if (!obj) {
> > > +			DRM_ERROR("failed to lookup gem object.\n");
> > > +			goto err_lookup;
> > > +		}
> > > +
> > > +		samsung_gem_obj = to_samsung_gem_obj(obj);
> > > +
> > > +		drm_gem_object_unreference_unlocked(obj);
> > > +	}
> > > +
> > > +	/*
> > > +	 * in case of getting samsung_gem_obj from either handle or
> > > +	 * new creation.
> > > +	 */
> > > +	samsung_fb->vaddr = samsung_gem_obj->entry->vaddr;
> > > +	samsung_fb->paddr = samsung_gem_obj->entry->paddr;
> > > +
> > > +	DRM_LOG_KMS("paddr = 0x%x, size = 0x%x, gem object = 0x%x\n",
> > > +			samsung_fb->paddr, size,
> (u32)&samsung_gem_obj->base);
> > 
> > Why truncating it to be 4GB? Is it potentially possible that this
> > could run a machine with 5GB?
> > 
> 
> You are right. I will correct it. thank you.
> 
> > > +
> > > +out:
> > > +	samsung_fb->samsung_gem_obj = samsung_gem_obj;
> > > +	samsung_fb->fb_size = size;
> > > +
> > > +	drm_helper_mode_fill_fb_struct(fb, mode_cmd);
> > > +
> > > +	return fb;
> > > +
> > > +err_lookup:
> > > +err_gem_create:
> > > +err_buf_create:
> > 
> > Why don't you just coalesce them all together and call it:
> > err:
> > 
> 
> Ok, get it. your saying is more clean. Thank you.
> 
> 
> > > +	drm_framebuffer_cleanup(fb);
> > > +
> > > +err_init:
> > > +	kfree(samsung_fb);
> > > +
> > > +	return ERR_PTR(ret);
> > > +}
> > > +
> > > +struct drm_framebuffer *samsung_drm_fb_create(struct drm_device *dev,
> > > +					      struct drm_file *file_priv,
> > > +					      struct drm_mode_fb_cmd
> *mode_cmd)
> > > +{
> > > +	DRM_DEBUG_KMS("%s\n", __FILE__);
> > > +
> > > +	return samsung_drm_fb_init(file_priv, dev, mode_cmd);
> > > +}
> > > +
> > > +void samsung_drm_fb_update_buf_off(struct drm_framebuffer *fb,
> > > +				   unsigned int x, unsigned int y,
> > > +				   struct samsung_drm_buffer_info *info)
> > > +{
> > > +	struct samsung_drm_fb *samsung_fb = to_samsung_fb(fb);
> > > +	unsigned long offset;
> > > +
> > > +	DRM_DEBUG_KMS("%s\n", __FILE__);
> > > +
> > > +	offset = x * (fb->bits_per_pixel >> 3);
> > > +	offset += y * fb->pitch;
> > > +
> > > +	info->base_addr = samsung_fb->paddr;
> > > +	info->vaddr = samsung_fb->vaddr + offset;
> > > +	info->paddr = samsung_fb->paddr + offset;
> > > +
> > > +	DRM_DEBUG_KMS("updated vaddr = 0x%x, paddr = 0x%x, offset = 0x%x\n",
> > > +			(unsigned int)info->vaddr,
> > > +			(unsigned int)info->paddr,
> > > +			(unsigned int)offset);
> > > +}
> > > +
> > > +static struct drm_mode_config_funcs samsung_drm_mode_config_funcs = {
> > > +	.fb_create = samsung_drm_fb_create,
> > > +};
> > > +
> > > +void samsung_drm_mode_config_init(struct drm_device *dev)
> > > +{
> > > +	dev->mode_config.min_width = 0;
> > > +	dev->mode_config.min_height = 0;
> > > +
> > > +	/*
> > > +	 * It sets max width and height as default value(4096x4096).
> > > +	 * this value would be used to check for framebuffer size
> > limitation
> > > +	 * at drm_mode_addfb().
> > > +	 */
> > > +	dev->mode_config.max_width = 4096;
> > > +	dev->mode_config.max_height = 4096;
> > > +
> > > +	dev->mode_config.funcs = &samsung_drm_mode_config_funcs;
> > > +}
> > > diff --git a/drivers/gpu/drm/samsung/samsung_drm_fb.h
> > b/drivers/gpu/drm/samsung/samsung_drm_fb.h
> > > new file mode 100644
> > > index 0000000..6256089
> > > --- /dev/null
> > > +++ b/drivers/gpu/drm/samsung/samsung_drm_fb.h
> > > @@ -0,0 +1,47 @@
> > > +/*
> > > + * Copyright (c) 2011 Samsung Electronics Co., Ltd.
> > > + * Authors:
> > > + *	Inki Dae <inki.dae@samsung.com>
> > > + *	Joonyoung Shim <jy0922.shim@samsung.com>
> > > + *	SeungWoo Kim <sw0312.kim@samsung.com>
> > > + *
> > > + * Permission is hereby granted, free of charge, to any person
> > obtaining a
> > > + * copy of this software and associated documentation files (the
> > "Software"),
> > > + * to deal in the Software without restriction, including without
> > limitation
> > > + * the rights to use, copy, modify, merge, publish, distribute,
> > sublicense,
> > > + * and/or sell copies of the Software, and to permit persons to whom
> > the
> > > + * Software is furnished to do so, subject to the following conditions:
> > > + *
> > > + * The above copyright notice and this permission notice (including the
> > next
> > > + * paragraph) shall be included in all copies or substantial portions
> > of the
> > > + * Software.
> > > + *
> > > + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
> > EXPRESS OR
> > > + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
> > MERCHANTABILITY,
> > > + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT
> > SHALL
> > > + * VA LINUX SYSTEMS AND/OR ITS SUPPLIERS BE LIABLE FOR ANY CLAIM,
> > DAMAGES OR
> > > + * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR
> OTHERWISE,
> > > + * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE
> > OR
> > > + * OTHER DEALINGS IN THE SOFTWARE.
> > > + */
> > > +
> > > +#ifndef _SAMSUNG_DRM_FB_H_
> > > +#define _SAMSUNG_DRM_FB_H
> > > +
> > > +struct samsung_drm_buffer_info {
> > > +	unsigned long base_addr;
> > > +	dma_addr_t paddr;
> > > +	void __iomem *vaddr;
> > > +};
> > > +
> > > +void samsung_drm_fb_update_buf_off(struct drm_framebuffer *fb,
> > > +				   unsigned int x, unsigned int y,
> > > +				   struct samsung_drm_buffer_info *info);
> > > +
> > > +struct drm_framebuffer *samsung_drm_fb_create(struct drm_device *dev,
> > > +					      struct drm_file *filp,
> > > +					      struct drm_mode_fb_cmd
> *mode_cmd);
> > > +
> > > +void samsung_drm_mode_config_init(struct drm_device *dev);
> > > +
> > > +#endif
> > > diff --git a/drivers/gpu/drm/samsung/samsung_drm_fbdev.c
> > b/drivers/gpu/drm/samsung/samsung_drm_fbdev.c
> > > new file mode 100644
> > > index 0000000..1c46bc6
> > > --- /dev/null
> > > +++ b/drivers/gpu/drm/samsung/samsung_drm_fbdev.c
> > > @@ -0,0 +1,409 @@
> > > +/*
> > > + * Copyright (c) 2011 Samsung Electronics Co., Ltd.
> > > + * Authors:
> > > + *	Inki Dae <inki.dae@samsung.com>
> > > + *	Joonyoung Shim <jy0922.shim@samsung.com>
> > > + *	SeungWoo Kim <sw0312.kim@samsung.com>
> > > + *
> > > + * Permission is hereby granted, free of charge, to any person
> > obtaining a
> > > + * copy of this software and associated documentation files (the
> > "Software"),
> > > + * to deal in the Software without restriction, including without
> > limitation
> > > + * the rights to use, copy, modify, merge, publish, distribute,
> > sublicense,
> > > + * and/or sell copies of the Software, and to permit persons to whom
> > the
> > > + * Software is furnished to do so, subject to the following conditions:
> > > + *
> > > + * The above copyright notice and this permission notice (including the
> > next
> > > + * paragraph) shall be included in all copies or substantial portions
> > of the
> > > + * Software.
> > > + *
> > > + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
> > EXPRESS OR
> > > + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
> > MERCHANTABILITY,
> > > + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT
> > SHALL
> > > + * VA LINUX SYSTEMS AND/OR ITS SUPPLIERS BE LIABLE FOR ANY CLAIM,
> > DAMAGES OR
> > > + * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR
> OTHERWISE,
> > > + * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE
> > OR
> > > + * OTHER DEALINGS IN THE SOFTWARE.
> > > + */
> > > +
> > > +#include "drmP.h"
> > > +#include "drm_crtc.h"
> > > +#include "drm_fb_helper.h"
> > > +#include "drm_crtc_helper.h"
> > > +
> > > +#include "samsung_drm_drv.h"
> > > +#include "samsung_drm_fb.h"
> > > +
> > > +#define to_samsung_fbdev(x)	container_of(x, struct
> > samsung_drm_fbdev,\
> > > +				drm_fb_helper)
> > > +
> > > +struct samsung_drm_fbdev {
> > > +	struct drm_fb_helper	drm_fb_helper;
> > > +	struct drm_framebuffer	*fb;
> > > +};
> > > +
> > > +static inline unsigned int chan_to_field(unsigned int chan,
> > > +					 struct fb_bitfield *bf)
> > > +{
> > > +	chan &= 0xffff;
> > > +	chan >>= 16 - bf->length;
> > 
> > Any chance that bf->length can be bigger than 16? And cause this
> > to return um wrong vlaues?
> > 
> > perhaps chan >>= (16 - max(16,bf->length));
> > 
> 
> Ok, get it. thank you.
> 
> > ?
> > > +
> > > +	return chan << bf->offset;
> > > +}
> > > +
> > > +static int samsung_drm_fbdev_setcolreg(unsigned regno, unsigned red,
> > > +				       unsigned green, unsigned blue,
> > > +				       unsigned transp, struct fb_info
> *info)
> > > +{
> > > +	unsigned int val;
> > > +
> > > +	DRM_DEBUG_KMS("%s\n", __FILE__);
> > > +
> > > +	switch (info->fix.visual) {
> > > +	case FB_VISUAL_TRUECOLOR:
> > > +		if (regno < 16) {
> > 
> > Not <=?
> 
> Ah, you are right. 16bit is high color. thank you.
> 
> > > +			u32 *pal = info->pseudo_palette;
> > > +
> > > +			val = chan_to_field(red, &info->var.red);
> > > +			val |= chan_to_field(green, &info->var.green);
> > > +			val |= chan_to_field(blue, &info->var.blue);
> > > +
> > > +			pal[regno] = val;
> > > +		}
> > 
> > so if regno > 16 then we still return 0. Should we return -EINVAL instead.
> > 
> 
> Ok, get it. thank you.
> 
> > > +		break;
> > > +	default:
> > > +		return 1;
> > 
> > -EINVAL?
> > 
> 
> Thank you.
> 
> > > +	}
> > > +
> > > +	return 0;
> > > +}
> > > +
> > > +static struct fb_ops samsung_drm_fb_ops = {
> > > +	.owner		= THIS_MODULE,
> > > +	.fb_fillrect	= cfb_fillrect,
> > > +	.fb_copyarea	= cfb_copyarea,
> > > +	.fb_imageblit	= cfb_imageblit,
> > > +	.fb_check_var	= drm_fb_helper_check_var,
> > > +	.fb_set_par	= drm_fb_helper_set_par,
> > > +	.fb_setcolreg	= samsung_drm_fbdev_setcolreg,
> > > +	.fb_blank	= drm_fb_helper_blank,
> > > +	.fb_pan_display	= drm_fb_helper_pan_display,
> > > +	.fb_setcmap	= drm_fb_helper_setcmap,
> > > +};
> > > +
> > > +static void samsung_drm_fbdev_update(struct drm_fb_helper *helper,
> > > +				     struct drm_framebuffer *fb,
> > > +				     unsigned int fb_width,
> > > +				     unsigned int fb_height)
> > > +{
> > > +	struct fb_info *fbi = helper->fbdev;
> > > +	struct drm_device *dev = helper->dev;
> > > +	struct samsung_drm_fbdev *samsung_fb = to_samsung_fbdev(helper);
> > > +	struct samsung_drm_buffer_info buffer_info;
> > > +	unsigned int size = fb_width * fb_height * (fb->bits_per_pixel >>
> > 3);
> > > +
> > > +	DRM_DEBUG_KMS("%s\n", __FILE__);
> > > +
> > > +	samsung_fb->fb = fb;
> > > +
> > > +	drm_fb_helper_fill_fix(fbi, fb->pitch, fb->depth);
> > > +	drm_fb_helper_fill_var(fbi, helper, fb_width, fb_height);
> > > +
> > > +	samsung_drm_fb_update_buf_off(fb, fbi->var.xoffset, fbi-
> > >var.yoffset,
> > > +			&buffer_info);
> > > +
> > > +	dev->mode_config.fb_base = buffer_info.base_addr;
> > > +
> > > +	fbi->screen_base = buffer_info.vaddr;
> > > +	fbi->screen_size = size;
> > > +	fbi->fix.smem_start = buffer_info.paddr;
> > > +	fbi->fix.smem_len = size;
> > > +}
> > > +
> > > +static int samsung_drm_fbdev_create(struct drm_fb_helper *helper,
> > > +				    struct drm_fb_helper_surface_size
> *sizes)
> > > +{
> > > +	struct samsung_drm_fbdev *samsung_fbdev = to_samsung_fbdev(helper);
> > > +	struct drm_device *dev = helper->dev;
> > > +	struct fb_info *fbi;
> > > +	struct drm_mode_fb_cmd mode_cmd = { 0 };
> > > +	struct platform_device *pdev = dev->platformdev;
> > > +	int ret;
> > > +
> > > +	DRM_DEBUG_KMS("%s\n", __FILE__);
> > > +
> > > +	DRM_DEBUG_KMS("surface width(%d), height(%d) and bpp(%d\n",
> > > +			sizes->surface_width, sizes->surface_height,
> > > +			sizes->surface_bpp);
> > > +
> > > +	mode_cmd.width = sizes->surface_width;
> > > +	mode_cmd.height = sizes->surface_height;
> > > +	mode_cmd.bpp = sizes->surface_bpp;
> > > +	mode_cmd.depth = sizes->surface_depth;
> > > +
> > > +	mutex_lock(&dev->struct_mutex);
> > > +
> > > +	fbi = framebuffer_alloc(0, &pdev->dev);
> > > +	if (!fbi) {
> > > +		DRM_ERROR("failed to allocate fb info.\n");
> > > +		ret = -ENOMEM;
> > > +		goto out;
> > > +	}
> > > +
> > > +	samsung_fbdev->fb = samsung_drm_fb_create(dev, NULL, &mode_cmd);
> > > +	if (IS_ERR(samsung_fbdev->fb)) {
> > 
> > You probably want to do IS_ERR_OR_NULL?
> > 
> 
> Ok, that's more safe. thank you.
> 
> > > +		DRM_ERROR("failed to create drm dramebuffer.\n");
> > > +		ret = PTR_ERR(samsung_fbdev->fb);
> > 
> > framebuffer_release ?
> > > +		goto out;
> > > +	}
> > > +
> 
> I missed it. thank you.
> 
> > > +	helper->fb = samsung_fbdev->fb;
> > > +	helper->fbdev = fbi;
> > > +
> > > +	fbi->par = helper;
> > > +	fbi->flags = FBINFO_FLAG_DEFAULT;
> > > +	fbi->fbops = &samsung_drm_fb_ops;
> > > +
> > > +	ret = fb_alloc_cmap(&fbi->cmap, 256, 0);
> > > +	if (ret) {
> > 
> > framebuffer_release ?
> > 
> 
> Also. Thank you.
> 
> > Or just add a new label that will make that call.
> > 
> > > +		DRM_ERROR("failed to allocate cmap.\n");
> > > +		goto out;
> > > +	}
> > > +
> > > +	samsung_drm_fbdev_update(helper, helper->fb, sizes->fb_width,
> > > +			sizes->fb_height);
> > > +
> > > +out:
> > > +	mutex_unlock(&dev->struct_mutex);
> > > +	return ret;
> > > +}
> > > +
> > > +static bool
> > > +samsung_drm_fbdev_is_samefb(struct drm_framebuffer *fb,
> > > +			    struct drm_fb_helper_surface_size *sizes)
> > > +{
> > > +	if (fb->width != sizes->surface_width)
> > > +		return false;
> > > +	if (fb->height != sizes->surface_height)
> > > +		return false;
> > > +	if (fb->bits_per_pixel != sizes->surface_bpp)
> > > +		return false;
> > > +	if (fb->depth != sizes->surface_depth)
> > > +		return false;
> > > +
> > > +	return true;
> > > +}
> > > +
> > > +static int samsung_drm_fbdev_recreate(struct drm_fb_helper *helper,
> > > +				      struct drm_fb_helper_surface_size
> *sizes)
> > > +{
> > > +	struct drm_device *dev = helper->dev;
> > > +	struct samsung_drm_fbdev *samsung_fbdev = to_samsung_fbdev(helper);
> > > +	struct drm_framebuffer *fb = samsung_fbdev->fb;
> > > +	struct drm_mode_fb_cmd mode_cmd = { 0 };
> > > +
> > > +	if (helper->fb != fb) {
> > > +		DRM_ERROR("drm framebuffer is different\n");
> > > +		return -EINVAL;
> > > +	}
> > > +
> > > +	if (samsung_drm_fbdev_is_samefb(fb, sizes))
> > > +		return 0;
> > > +
> > > +	mode_cmd.width = sizes->surface_width;
> > > +	mode_cmd.height = sizes->surface_height;
> > > +	mode_cmd.bpp = sizes->surface_bpp;
> > > +	mode_cmd.depth = sizes->surface_depth;
> > > +
> > > +	if (fb->funcs->destroy)
> > > +		fb->funcs->destroy(fb);
> > > +
> > > +	samsung_fbdev->fb = samsung_drm_fb_create(dev, NULL, &mode_cmd);
> > > +	if (IS_ERR(samsung_fbdev->fb)) {
> > > +		DRM_ERROR("failed to allocate fb.\n");
> > > +		return PTR_ERR(samsung_fbdev->fb);
> > > +	}
> > > +
> > > +	helper->fb = samsung_fbdev->fb;
> > > +	samsung_drm_fbdev_update(helper, helper->fb, sizes->fb_width,
> > > +			sizes->fb_height);
> > > +
> > > +	return 0;
> > > +}
> > > +
> > > +static int samsung_drm_fbdev_probe(struct drm_fb_helper *helper,
> > > +				   struct drm_fb_helper_surface_size *sizes)
> > > +{
> > > +	int ret = 0;
> > > +
> > > +	DRM_DEBUG_KMS("%s\n", __FILE__);
> > > +
> > > +	if (!helper->fb) {
> > > +		ret = samsung_drm_fbdev_create(helper, sizes);
> > > +		if (ret < 0) {
> > > +			DRM_ERROR("failed to create fbdev.\n");
> > > +			return ret;
> > > +		}
> > > +
> > > +		/*
> > > +		 * fb_helper expects a value more than 1 if succeed
> > > +		 * because register_framebuffer() should be called.
> > > +		 */
> > > +		ret = 1;
> > > +	} else {
> > > +		ret = samsung_drm_fbdev_recreate(helper, sizes);
> > > +		if (ret < 0) {
> > > +			DRM_ERROR("failed to reconfigure fbdev\n");
> > > +			return ret;
> > > +		}
> > 
> > No need to do the same thing you did before?
> > 
> 
> Ah, I will correct it. thank you.
> 
> >  ret = 1?
> > 
> 
> This code must be strange. please, see new_fb =
> (*fb_helper->funcs->fb_prob)(fb_helper, &sizes) of
> drm_fb_helper_single_fb_probe(). drm_fb_helper.c requires three return
> values.

So confusing...
> 
> > > +	}
> > > +
> > > +	return ret;
> > > +}
> > > +
> > > +static struct drm_fb_helper_funcs samsung_drm_fb_helper_funcs = {
> > > +	.fb_probe =	samsung_drm_fbdev_probe,
> > > +};
> > > +
> > > +int samsung_drm_fbdev_init(struct drm_device *dev)
> > > +{
> > > +	struct samsung_drm_fbdev *fbdev;
> > > +	struct samsung_drm_private *private = dev->dev_private;
> > > +	struct drm_fb_helper *helper;
> > > +	unsigned int num_crtc;
> > > +	int ret;
> > > +
> > > +	DRM_DEBUG_KMS("%s\n", __FILE__);
> > > +
> > > +	if (!dev->mode_config.num_crtc || !dev->mode_config.num_connector)
> > > +		return 0;
> > > +
> > > +	fbdev = kzalloc(sizeof(*fbdev), GFP_KERNEL);
> > > +	if (!fbdev) {
> > > +		DRM_ERROR("failed to allocate drm fbdev.\n");
> > > +		return -ENOMEM;
> > > +	}
> > > +
> > > +	private->fb_helper = helper = &fbdev->drm_fb_helper;
> > > +	helper->funcs = &samsung_drm_fb_helper_funcs;
> > > +
> > > +	num_crtc = dev->mode_config.num_crtc;
> > > +
> > > +	ret = drm_fb_helper_init(dev, helper, num_crtc, 4);
> > 
> > Is that value '4'
> 
> Yes, this driver is going to provide maximum four connectors.
> 
> > > +	if (ret < 0) {
> > > +		DRM_ERROR("failed to initialize drm fb helper.\n");
> > > +		goto fail;
> > > +	}
> > > +
> > > +	ret = drm_fb_helper_single_add_all_connectors(helper);
> > > +	if (ret < 0) {
> > > +		DRM_ERROR("failed to register drm_fb_helper_connector.\n");
> > > +		goto fail;
> > > +
> > > +	}
> > > +
> > > +	ret = drm_fb_helper_initial_config(helper, 32);
> > 
> > and '32' should perhaps be in #defines?
> 
> Ok, get it. thank you.
> 
> > > +	if (ret < 0) {
> > > +		DRM_ERROR("failed to set up hw configuration.\n");
> > 
> > Which I think leaks memory. The drm_fb_helper_single_add_all_connectors
> > does a bunch of kzallocs. Should you free them here?
> > 
> 
> Ah, you are right. thank you.
> 
> > > +		goto fail;
> > > +	}
> > > +
> > > +	return 0;
> > > +
> > > +fail:
> > > +	private->fb_helper = NULL;
> > > +	kfree(fbdev);
> > > +
> > > +	return ret;
> > > +}
> > > +
> > > +static void samsung_drm_fbdev_destroy(struct drm_device *dev,
> > > +				      struct drm_fb_helper *fb_helper)
> > > +{
> > > +	struct drm_framebuffer *fb;
> > > +	struct fb_info *info;
> > > +
> > > +	/* release drm framebuffer and real buffer */
> > > +	if (fb_helper->fb && fb_helper->fb->funcs) {
> > > +		fb = fb_helper->fb;
> > > +		if (fb->funcs->destroy)
> > > +			fb->funcs->destroy(fb);
> > > +	}
> > > +
> > > +	/* release linux framebuffer */
> > > +	if (fb_helper->fbdev) {
> > 
> > You can declare the 'info' here.
> 
> Ok, get it. thank you.
> 
> > > +		info = fb_helper->fbdev;
> > > +		unregister_framebuffer(info);
> > > +		if (info->cmap.len)
> > > +			fb_dealloc_cmap(&info->cmap);
> > > +		framebuffer_release(info);
> > > +	}
> > > +
> > > +	drm_fb_helper_fini(fb_helper);
> > > +}
> > > +
> > > +void samsung_drm_fbdev_fini(struct drm_device *dev)
> > > +{
> > > +	struct samsung_drm_private *private = dev->dev_private;
> > > +	struct samsung_drm_fbdev *fbdev;
> > > +
> > > +	if (!private || !private->fb_helper)
> > > +		return;
> > > +
> > > +	fbdev = to_samsung_fbdev(private->fb_helper);
> > > +
> > > +	samsung_drm_fbdev_destroy(dev, private->fb_helper);
> > > +	kfree(fbdev);
> > > +	private->fb_helper = NULL;
> > > +}
> > > +
> > > +void samsung_drm_fbdev_restore_mode(struct drm_device *dev)
> > > +{
> > > +	struct samsung_drm_private *private = dev->dev_private;
> > > +
> > > +	if (!private || !private->fb_helper)
> > > +		return;
> > > +
> > > +	drm_fb_helper_restore_fbdev_mode(private->fb_helper);
> > > +}
> > > +
> > > +int samsung_drm_fbdev_reinit(struct drm_device *dev)
> > > +{
> > > +	struct samsung_drm_private *private = dev->dev_private;
> > > +	int ret;
> > > +
> > > +	if (!private)
> > > +		return -EINVAL;
> > > +
> > > +	if (!dev->mode_config.num_connector) {
> > > +		samsung_drm_fbdev_fini(dev);
> > > +		return 0;
> > > +	}
> > > +
> > > +	if (private->fb_helper) {
> > > +		struct drm_fb_helper *fb_helper = private->fb_helper;
> > > +
> > > +		drm_fb_helper_fini(fb_helper);
> > > +
> > > +		ret = drm_fb_helper_init(dev, fb_helper,
> > > +				dev->mode_config.num_crtc, 4);
> > > +		if (ret < 0) {
> > > +			DRM_ERROR("failed to initialize drm fb helper\n");
> > > +			return ret;
> > > +		}
> > > +
> > > +		ret = drm_fb_helper_single_add_all_connectors(fb_helper);
> > > +		if (ret < 0) {
> > > +			DRM_ERROR("failed to add fb helper to
> connectors\n");
> > 
> > You should free what drm_fb_helper_init allocated
> 
> Ok, get it. thank you.
> 
> > > +			return ret;
> > > +		}
> > > +
> > > +		ret = drm_fb_helper_initial_config(fb_helper, 32);
> > > +		if (ret < 0) {
> > 
> > .. and free what drm_fb_helper_single_add_all_connectors allocated too.
> > 
> 
> Ok, thank you.
> 
> > > +			DRM_ERROR("failed to set up hw configuration.\n");
> > > +			return ret;
> > > +		}
> > > +	} else
> > > +		ret = samsung_drm_fbdev_init(dev);
> > > +
> > > +	return ret;
> > > +}
> > > diff --git a/drivers/gpu/drm/samsung/samsung_drm_fbdev.h
> > b/drivers/gpu/drm/samsung/samsung_drm_fbdev.h
> > > new file mode 100644
> > > index 0000000..3ef4e0d
> > > --- /dev/null
> > > +++ b/drivers/gpu/drm/samsung/samsung_drm_fbdev.h
> > > @@ -0,0 +1,37 @@
> > > +/*
> > > + * Copyright (c) 2011 Samsung Electronics Co., Ltd.
> > > + *
> > > + * Authors:
> > > + *	Inki Dae <inki.dae@samsung.com>
> > > + *	Joonyoung Shim <jy0922.shim@samsung.com>
> > > + *	SeungWoo Kim <sw0312.kim@samsung.com>
> > > + *
> > > + * Permission is hereby granted, free of charge, to any person
> > obtaining a
> > > + * copy of this software and associated documentation files (the
> > "Software"),
> > > + * to deal in the Software without restriction, including without
> > limitation
> > > + * the rights to use, copy, modify, merge, publish, distribute,
> > sublicense,
> > > + * and/or sell copies of the Software, and to permit persons to whom
> > the
> > > + * Software is furnished to do so, subject to the following conditions:
> > > + *
> > > + * The above copyright notice and this permission notice (including the
> > next
> > > + * paragraph) shall be included in all copies or substantial portions
> > of the
> > > + * Software.
> > > + *
> > > + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
> > EXPRESS OR
> > > + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
> > MERCHANTABILITY,
> > > + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT
> > SHALL
> > > + * VA LINUX SYSTEMS AND/OR ITS SUPPLIERS BE LIABLE FOR ANY CLAIM,
> > DAMAGES OR
> > > + * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR
> OTHERWISE,
> > > + * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE
> > OR
> > > + * OTHER DEALINGS IN THE SOFTWARE.
> > > + */
> > > +
> > > +#ifndef _SAMSUNG_DRM_FBDEV_H_
> > > +#define _SAMSUNG_DRM_FBDEV_H_
> > > +
> > > +int samsung_drm_fbdev_init(struct drm_device *dev);
> > > +int samsung_drm_fbdev_reinit(struct drm_device *dev);
> > > +void samsung_drm_fbdev_fini(struct drm_device *dev);
> > > +void samsung_drm_fbdev_restore_mode(struct drm_device *dev);
> > > +
> > > +#endif
> > > diff --git a/drivers/gpu/drm/samsung/samsung_drm_fimd.c
> > b/drivers/gpu/drm/samsung/samsung_drm_fimd.c
> > > new file mode 100644
> > > index 0000000..d823e19
> > > --- /dev/null
> > > +++ b/drivers/gpu/drm/samsung/samsung_drm_fimd.c
> > > @@ -0,0 +1,643 @@
> > > +/*
> > > + * Copyright (C) 2011 Samsung Electronics Co.Ltd
> > > + * Authors:
> > > + *	Joonyoung Shim <jy0922.shim@samsung.com>
> > > + *	Inki Dae <inki.dae@samsung.com>
> > > + *
> > > + * This program is free software; you can redistribute  it and/or
> > modify it
> > > + * under  the terms of  the GNU General  Public License as published by
> > the
> > > + * Free Software Foundation;  either version 2 of the  License, or (at
> > your
> > > + * option) any later version.
> > > + *
> > > + */
> > 
> > What is 'FIMD'? Is there an explanation of what that stands for?
> > 
> 
> That's Fully Interactive Mobile Display and this is display controller. 

OK, can you include that comment and perhaps a link to the specification?
(if such thing exists).
> 
> 
> > > +
> > > +/*
> > > + * TODO list
> > > + *  - Code cleanup(FIXME and TODO parts)
> > > + *  - Clock gating and Power management
> > > + *  - Writeback feature
> > > + */
> > > +
> > > +#include "drmP.h"
> > > +
> > > +#include <linux/kernel.h>
> > > +#include <linux/module.h>
> > > +#include <linux/platform_device.h>
> > > +#include <linux/clk.h>
> > > +
> > > +#include <drm/samsung_drm.h>
> > > +#include <plat/regs-fb-v4.h>
> > > +
> > > +#include "samsung_drm_drv.h"
> > > +#include "samsung_drm_fbdev.h"
> > > +#include "samsung_drm_crtc.h"
> > > +
> > > +/* irq_flags bits */
> > > +#define FIMD_VSYNC_IRQ_EN	0
> > 
> > Which is just one value right? 0?
> > 
> 
> This definition isn't needed. so I will remove it and directly use 0
> instead.
> 
> > 
> > > +
> > > +#define VIDOSD_A(win)	(VIDOSD_BASE + 0x00 + (win) * 16)
> > > +#define VIDOSD_B(win)	(VIDOSD_BASE + 0x04 + (win) * 16)
> > > +#define VIDOSD_C(win)	(VIDOSD_BASE + 0x08 + (win) * 16)
> > > +#define VIDOSD_D(win)	(VIDOSD_BASE + 0x0C + (win) * 16)
> > 
> > What about a comment for these very important low-level values?
> > > +
> 
> I think it's good to leave some comments to them also. I will add some
> comments. thank you.
> 
> 
> > > +#define VIDWx_BUF_START(win, buf)	(VIDW_BUF_START(buf) + (win) * 8)
> > > +#define VIDWx_BUF_END(win, buf)		(VIDW_BUF_END(buf) + (win)
> > * 8)
> > > +#define VIDWx_BUF_SIZE(win, buf)	(VIDW_BUF_SIZE(buf) + (win) * 4)
> > > +
> > > +#define WINDOWS_NR	5
> > > +
> > > +#define get_fimd_context(dev)
> > 	platform_get_drvdata(to_platform_device(dev))
> > > +
> > > +struct fimd_win_data {
> > > +	unsigned int		offset_x;
> > > +	unsigned int		offset_y;
> > > +	unsigned int		width;
> > > +	unsigned int		height;
> > > +	unsigned int		paddr;
> > 
> > dma_addr_t ?
> > 
> 
> Get it. thank you.
> 
> > > +	void __iomem		*vaddr;
> > > +	unsigned int		end_buf_off;
> > 
> > buf_off ?
> 
> I have a feeling that this variable could be removed. thank you.
> 
> > > +	unsigned int		buf_offsize;
> > > +	unsigned int		line_size;	/* bytes */
> > > +};
> > > +
> > > +struct fimd_context {
> > > +	struct samsung_drm_subdrv	subdrv;
> > > +	int				irq_no;
> > 
> > Just 'irq'
> 
> Ok, get it.
> 
> > 
> > > +	struct drm_crtc			*crtc;
> > > +	struct clk			*bus_clk;
> > > +	struct clk			*lcd_clk;
> > > +	struct resource			*regs_res;
> > > +	void __iomem			*regs;
> > > +	struct fimd_win_data		win_data[WINDOWS_NR];
> > > +	unsigned int			clkdiv;
> > > +	unsigned int			default_win;
> > > +	unsigned int			bpp;
> > > +	unsigned long			irq_flags;
> > > +	u32				vidcon0;
> > > +	u32				vidcon1;
> > > +
> > > +	struct fb_videomode		*timing;
> > > +};
> > > +
> > > +static bool fimd_display_is_connected(struct device *dev)
> > > +{
> > > +	struct fimd_context *ctx = get_fimd_context(dev);
> > > +
> > > +	/* TODO. */
> > > +
> > > +	return true;
> > > +}
> > > +
> > > +static void *fimd_get_timing(struct device *dev)
> > > +{
> > > +	struct fimd_context *ctx = get_fimd_context(dev);
> > > +
> > > +	return ctx->timing;
> > > +}
> > > +
> > > +static int fimd_check_timing(struct device *dev, void *timing)
> > > +{
> > > +	struct fimd_context *ctx = get_fimd_context(dev);
> > > +
> > > +	/* TODO. */
> > > +
> > > +	return 0;
> > > +}
> > > +
> > > +static int fimd_display_power_on(struct device *dev, int mode)
> > > +{
> > > +	struct fimd_context *ctx = get_fimd_context(dev);
> > > +
> > > +	/* TODO. */
> > > +
> > > +	return 0;
> > > +}
> > > +
> > > +static struct samsung_drm_display fimd_display = {
> > > +	.type = SAMSUNG_DISPLAY_TYPE_LCD,
> > > +	.is_connected = fimd_display_is_connected,
> > > +	.get_timing = fimd_get_timing,
> > > +	.check_timing = fimd_check_timing,
> > > +	.power_on = fimd_display_power_on,
> > > +};
> > > +
> > > +static void fimd_commit(struct device *dev)
> > > +{
> > > +	struct fimd_context *ctx = get_fimd_context(dev);
> > > +	void __iomem *regs = ctx->regs;
> > > +	struct fb_videomode *timing = ctx->timing;
> > > +	u32 val;
> > > +
> > > +	/* vidcon0 */
> > > +	val = ctx->vidcon0;
> > > +	val &= ~(VIDCON0_CLKVAL_F_MASK | VIDCON0_CLKDIR);
> > > +
> > > +	if (ctx->clkdiv > 1)
> > > +		val |= VIDCON0_CLKVAL_F(ctx->clkdiv - 1) | VIDCON0_CLKDIR;
> > > +	else
> > > +		val &= ~VIDCON0_CLKDIR;	/* 1:1 clock */
> > > +
> > > +	val |= VIDCON0_ENVID | VIDCON0_ENVID_F;
> > > +	writel(val, regs + VIDCON0);
> > > +
> > > +	/* vidcon1 */
> > 
> > That comment is bit useless.. It is clear what you are
> > doing.
> 
> Ok, get it. thank you.
> 
> > > +	writel(ctx->vidcon1, regs + VIDCON1);
> > > +
> > > +	/* vidtcon0 */
> > 
> > Ah, so the order is important. Perhaps you should
> > just have a comment before you do any writes that
> > enumarates in which order it should be done?
> > 
> 
> The order hasn't implications for working. no problem if this setting is
> done before dma of fimd is enabled. Ah.. such comments should be added.
> thank you.
> 
> > And why the order is important (in case the future
> > chipsets have it fixed/changed).
> > 
> 
> I understood what you mean. Thank you for your pointing.
> 
> > > +	val = VIDTCON0_VBPD(timing->upper_margin - 1) |
> > > +	       VIDTCON0_VFPD(timing->lower_margin - 1) |
> > > +	       VIDTCON0_VSPW(timing->vsync_len - 1);
> > > +	writel(val, regs + VIDTCON0);
> > > +
> > > +	/* vidtcon1 */
> > > +	val = VIDTCON1_HBPD(timing->left_margin - 1) |
> > > +	       VIDTCON1_HFPD(timing->right_margin - 1) |
> > > +	       VIDTCON1_HSPW(timing->hsync_len - 1);
> > > +	writel(val, regs + VIDTCON1);
> > > +
> > > +	/* vidtcon2 */
> > > +	val = VIDTCON2_LINEVAL(timing->yres - 1) |
> > > +	       VIDTCON2_HOZVAL(timing->xres - 1);
> > > +	writel(val, regs + VIDTCON2);
> > > +}
> > > +
> > > +static int fimd_enable_vblank(struct device *dev)
> > > +{
> > > +	struct fimd_context *ctx = get_fimd_context(dev);
> > > +	void __iomem *regs = ctx->regs;
> > > +	u32 val;
> > > +
> > > +	if (!test_and_set_bit(FIMD_VSYNC_IRQ_EN, &ctx->irq_flags)) {
> > > +		val = readl(regs + VIDINTCON0);
> > > +
> > > +		val |= VIDINTCON0_INT_ENABLE;
> > > +		val |= VIDINTCON0_INT_FRAME;
> > > +
> > > +		val &= ~VIDINTCON0_FRAMESEL0_MASK;
> > > +		val |= VIDINTCON0_FRAMESEL0_VSYNC;
> > > +		val &= ~VIDINTCON0_FRAMESEL1_MASK;
> > > +		val |= VIDINTCON0_FRAMESEL1_NONE;
> > > +
> > > +		writel(val, regs + VIDINTCON0);
> > > +	}
> > > +
> > > +	return 0;
> > > +}
> > > +
> > > +static void fimd_disable_vblank(struct device *dev)
> > > +{
> > > +	struct fimd_context *ctx = get_fimd_context(dev);
> > > +	void __iomem *regs = ctx->regs;
> > > +	u32 val;
> > > +
> > > +	if (test_and_clear_bit(FIMD_VSYNC_IRQ_EN, &ctx->irq_flags)) {
> > > +		val = readl(regs + VIDINTCON0);
> > > +
> > > +		val &= ~VIDINTCON0_INT_FRAME;
> > > +		val &= ~VIDINTCON0_INT_ENABLE;
> > > +
> > > +		writel(val, regs + VIDINTCON0);
> > > +	}
> > > +}
> > > +
> > > +static struct samsung_drm_manager_ops fimd_manager_ops = {
> > > +	.commit = fimd_commit,
> > > +	.enable_vblank = fimd_enable_vblank,
> > > +	.disable_vblank = fimd_disable_vblank,
> > > +};
> > > +
> > > +static void fimd_win_mode_set(struct device *dev,
> > > +			      struct samsung_drm_overlay *overlay)
> > > +{
> > > +	struct fimd_context *ctx = get_fimd_context(dev);
> > > +	struct fimd_win_data *win_data;
> > > +
> > > +	if (!overlay) {
> > > +		dev_err(dev, "overlay is NULL\n");
> > > +		return;
> > > +	}
> > > +
> > > +	win_data = &ctx->win_data[ctx->default_win];
> > > +
> > > +	win_data->offset_x = overlay->offset_x;
> > > +	win_data->offset_y = overlay->offset_y;
> > > +	win_data->width = overlay->width;
> > > +	win_data->height = overlay->height;
> > > +	win_data->paddr = overlay->paddr;
> > > +	win_data->vaddr = overlay->vaddr;
> > > +	win_data->end_buf_off = overlay->end_buf_off * (ctx->bpp >> 3);
> > > +	win_data->buf_offsize = overlay->buf_offsize * (ctx->bpp >> 3);
> > > +	win_data->line_size = overlay->line_size * (ctx->bpp >> 3);
> > > +}
> > > +
> > > +static void fimd_win_commit(struct device *dev)
> > > +{
> > > +	struct fimd_context *ctx = get_fimd_context(dev);
> > > +	void __iomem *regs = ctx->regs;
> > > +	struct fimd_win_data *win_data;
> > > +	int win = ctx->default_win;
> > > +	u32 val;
> > > +
> > > +	if (win < 0 || win > WINDOWS_NR)
> > > +		return;
> > > +
> > > +	win_data = &ctx->win_data[win];
> > > +
> > > +	/* protect windows */
> > > +	val = readl(regs + SHADOWCON);
> > > +	val |= SHADOWCON_WINx_PROTECT(win);
> > > +	writel(val, regs + SHADOWCON);
> > > +
> > > +	/* buffer start address */
> > > +	val = win_data->paddr;
> > > +	writel(val, regs + VIDWx_BUF_START(win, 0));
> > > +
> > > +	/* buffer end address */
> > > +	val = win_data->paddr + win_data->end_buf_off;
> > > +	writel(val, regs + VIDWx_BUF_END(win, 0));
> > > +
> > > +	/* buffer size */
> > > +	val = VIDW_BUF_SIZE_OFFSET(win_data->buf_offsize) |
> > > +		VIDW_BUF_SIZE_PAGEWIDTH(win_data->line_size);
> > > +	writel(val, regs + VIDWx_BUF_SIZE(win, 0));
> > > +
> > > +	/* OSD position */
> > > +	val = VIDOSDxA_TOPLEFT_X(win_data->offset_x) |
> > > +		VIDOSDxA_TOPLEFT_Y(win_data->offset_y);
> > > +	writel(val, regs + VIDOSD_A(win));
> > > +
> > > +	val = VIDOSDxB_BOTRIGHT_X(win_data->offset_x + win_data->width - 1)
> > |
> > > +		VIDOSDxB_BOTRIGHT_Y(win_data->offset_y + win_data->height -
> > 1);
> > > +	writel(val, regs + VIDOSD_B(win));
> > > +
> > > +	/* TODO: OSD alpha */
> > > +
> > > +	/* OSD size */
> > > +	if (win != 3 && win != 4) {
> > > +		u32 offset = VIDOSD_D(win);
> > > +		if (win == 0)
> > > +			offset = VIDOSD_C(win);
> > > +		val = win_data->width * win_data->height;
> > > +		writel(val, regs + offset);
> > > +	}
> > > +
> > > +	/* FIXME: remove fixed values */
> > 
> > <nods> Yes.
> > 
> 
> That FIXME would be updated soon. :)
> 
> > > +	val = WINCONx_ENWIN;
> > > +	val |= WINCON0_BPPMODE_24BPP_888;
> > > +	val |= WINCONx_WSWP;
> > > +	val |= WINCONx_BURSTLEN_16WORD;
> > > +	writel(val, regs + WINCON(win));
> > > +
> > > +	/* TODO: colour key */
> > > +
> > > +	/* Enable DMA channel and unprotect windows */
> > > +	val = readl(regs + SHADOWCON);
> > > +	val |= SHADOWCON_CHx_ENABLE(win);
> > > +	val &= ~SHADOWCON_WINx_PROTECT(win);
> > > +	writel(val, regs + SHADOWCON);
> > > +}
> > > +
> > > +static void fimd_win_disable(struct device *dev)
> > > +{
> > > +	struct fimd_context *ctx = get_fimd_context(dev);
> > > +	void __iomem *regs = ctx->regs;
> > > +	struct fimd_win_data *win_data;
> > > +	int win = ctx->default_win;
> > > +	u32 val;
> > > +
> > > +	if (win < 0 || win > WINDOWS_NR)
> > > +		return;
> > > +
> > > +	win_data = &ctx->win_data[win];
> > > +
> > > +	/* protect windows */
> > > +	val = readl(regs + SHADOWCON);
> > > +	val |= SHADOWCON_WINx_PROTECT(win);
> > > +	writel(val, regs + SHADOWCON);
> > > +
> > > +	/* wincon */
> > > +	val = readl(regs + WINCON(win));
> > > +	val &= ~WINCONx_ENWIN;
> > > +	writel(val, regs + WINCON(win));
> > > +
> > > +	/* unprotect windows */
> > > +	val = readl(regs + SHADOWCON);
> > > +	val &= ~SHADOWCON_CHx_ENABLE(win);
> > > +	val &= ~SHADOWCON_WINx_PROTECT(win);
> > > +	writel(val, regs + SHADOWCON);
> > > +}
> > > +
> > > +static struct samsung_drm_overlay_ops fimd_overlay_ops = {
> > > +	.mode_set = fimd_win_mode_set,
> > > +	.commit = fimd_win_commit,
> > > +	.disable = fimd_win_disable,
> > > +};
> > > +
> > > +/* for pageflip event */
> > > +static void fimd_finish_pageflip(struct drm_device *drm_dev, int crtc)
> > > +{
> > > +	struct samsung_drm_private *dev_priv = drm_dev->dev_private;
> > > +	struct drm_pending_vblank_event *e, *t;
> > > +	struct timeval now;
> > > +	unsigned long flags;
> > > +
> > > +	if (!dev_priv->pageflip_event)
> > > +		return;
> > > +
> > > +	spin_lock_irqsave(&drm_dev->event_lock, flags);
> > > +
> > > +	list_for_each_entry_safe(e, t, &dev_priv->pageflip_event_list,
> > > +			base.link) {
> > > +		do_gettimeofday(&now);
> > > +		e->event.sequence = 0;
> > > +		e->event.tv_sec = now.tv_sec;
> > > +		e->event.tv_usec = now.tv_usec;
> > > +
> > > +		list_move_tail(&e->base.link, &e->base.file_priv-
> > >event_list);
> > > +		wake_up_interruptible(&e->base.file_priv->event_wait);
> > > +	}
> > > +
> > > +	drm_vblank_put(drm_dev, crtc);
> > 
> > You can make this call outside the spinlock. But it looks like pretty
> > much everybody is doing it this way?
> > 
> 
> No problem to move it outside(to fimd_irq_handler) and spinlock is used only
> here. thank you.
> 
> > > +	dev_priv->pageflip_event = false;
> > > +
> > > +	spin_unlock_irqrestore(&drm_dev->event_lock, flags);
> > > +}
> > > +
> > > +static irqreturn_t fimd_irq_handler(int irq, void *dev_id)
> > > +{
> > > +	struct fimd_context *ctx = (struct fimd_context *)dev_id;
> > > +	struct samsung_drm_subdrv *subdrv = &ctx->subdrv;
> > > +	struct drm_device *drm_dev = subdrv->drm_dev;
> > > +	struct device *dev = subdrv->manager.dev;
> > > +	struct samsung_drm_manager *manager = &subdrv->manager;
> > > +	void __iomem *regs = ctx->regs;
> > > +	u32 val;
> > > +
> > > +	val = readl(regs + VIDINTCON1);
> > > +
> > > +	if (val & VIDINTCON1_INT_FRAME)
> > > +		/* VSYNC interrupt */
> > > +		writel(VIDINTCON1_INT_FRAME, regs + VIDINTCON1);
> > > +
> > > +	drm_handle_vblank(drm_dev, manager->pipe);
> > > +	fimd_finish_pageflip(drm_dev, manager->pipe);
> > > +
> > > +	return IRQ_HANDLED;
> > > +}
> > > +
> > > +static int fimd_subdrv_probe(struct drm_device *drm_dev)
> > > +{
> > > +	struct drm_driver *drm_driver = drm_dev->driver;
> > > +
> > > +	drm_dev->irq_enabled = 1;
> > 
> > Hehe.. Just like that ,eh?
> > 
> 
> I will add some comments. :)
> 
> > > +
> > > +	/* TODO. */
> > > +
> > > +	/* FIXME */
> > > +	drm_dev->vblank_disable_allowed = 1;
> > > +
> 
> Ditto.
> 
> > > +	return 0;
> > > +}
> > > +
> > > +static void fimd_subdrv_remove(struct drm_device *drm_dev)
> > > +{
> > > +	struct drm_driver *drm_driver = drm_dev->driver;
> > > +
> > > +	/* TODO. */
> > > +}
> > > +
> > > +static int fimd_calc_clkdiv(struct fimd_context *ctx,
> > > +			    struct fb_videomode *timing)
> > > +{
> > > +	unsigned long clk = clk_get_rate(ctx->lcd_clk);
> > > +	u32 retrace;
> > > +	u32 clkdiv;
> > > +	u32 best_framerate = 0;
> > > +	u32 framerate;
> > > +
> > > +	retrace = timing->left_margin + timing->hsync_len +
> > > +				timing->right_margin + timing->xres;
> > > +	retrace *= timing->upper_margin + timing->vsync_len +
> > > +				timing->lower_margin + timing->yres;
> > > +
> > > +	/* default framerate is 60Hz */
> > > +	if (!timing->refresh)
> > > +		timing->refresh = 60;
> > > +
> > > +	clk /= retrace;
> > > +
> > > +	for (clkdiv = 1; clkdiv < 0x100; clkdiv++) {
> > > +		int tmp;
> > > +
> > > +		/* get best framerate */
> > > +		framerate = clk / clkdiv;
> > > +		tmp = timing->refresh - framerate;
> > > +		if (tmp < 0) {
> > > +			best_framerate = framerate;
> > > +			continue;
> > > +		} else {
> > > +			if (!best_framerate)
> > > +				best_framerate = framerate;
> > > +			else if (tmp < (best_framerate - framerate))
> > > +				best_framerate = framerate ;
> > 
> > The ';' at the end should be moved.
> 
> Ah, get it. thank you.
> 
> > > +			break;
> > > +		}
> > > +	}
> > > +
> > > +	return clkdiv;
> > > +}
> > > +
> > > +static void fimd_clear_win(struct fimd_context *ctx, int win)
> > > +{
> > > +	void __iomem *regs = ctx->regs;
> > > +	u32 val;
> > > +
> > > +	writel(0, regs + WINCON(win));
> > > +	writel(0, regs + VIDOSD_A(win));
> > > +	writel(0, regs + VIDOSD_B(win));
> > > +	writel(0, regs + VIDOSD_C(win));
> > > +
> > > +	if (win == 1 || win == 2)
> > > +		writel(0, regs + VIDOSD_D(win));
> > > +
> > > +	val = readl(regs + SHADOWCON);
> > > +	val &= ~SHADOWCON_WINx_PROTECT(win);
> > > +	writel(val, regs + SHADOWCON);
> > > +}
> > > +
> > > +static int __devinit fimd_probe(struct platform_device *pdev)
> > > +{
> > > +	struct device *dev = &pdev->dev;
> > > +	struct fimd_context *ctx;
> > > +	struct samsung_drm_subdrv *subdrv;
> > > +	struct samsung_drm_fimd_pdata *pdata;
> > > +	struct fb_videomode *timing;
> > > +	struct resource *res;
> > > +	int win;
> > > +	int ret = -EINVAL;
> > > +
> > > +	printk(KERN_DEBUG "[%d] %s\n", __LINE__, __func__);
> > 
> > Hm printk? Use pr_dbg instead.
> > 
> 
> Ok, get it. thank you.
> 
> > > +
> > > +	pdata = pdev->dev.platform_data;
> > > +	if (!pdata) {
> > > +		dev_err(dev, "no platform data specified\n");
> > > +		return -EINVAL;
> > > +	}
> > > +
> > > +	timing = &pdata->timing;
> > > +	if (!timing) {
> > > +		dev_err(dev, "timing is null.\n");
> > > +		return -EINVAL;
> > > +	}
> > > +
> > > +	ctx = kzalloc(sizeof(*ctx), GFP_KERNEL);
> > > +	if (!ctx)
> > > +		return -ENOMEM;
> > > +
> > > +	ctx->bus_clk = clk_get(dev, "fimd");
> > > +	if (IS_ERR(ctx->bus_clk)) {
> > > +		dev_err(dev, "failed to get bus clock\n");
> > > +		ret = PTR_ERR(ctx->bus_clk);
> > > +		goto err_clk_get;
> > > +	}
> > > +
> > > +	clk_enable(ctx->bus_clk);
> > > +
> > > +	ctx->lcd_clk = clk_get(dev, "sclk_fimd");
> > > +	if (IS_ERR(ctx->lcd_clk)) {
> > > +		dev_err(dev, "failed to get lcd clock\n");
> > > +		ret = PTR_ERR(ctx->lcd_clk);
> > > +		goto err_bus_clk;
> > > +	}
> > > +
> > > +	clk_enable(ctx->lcd_clk);
> > > +
> > > +	res = platform_get_resource(pdev, IORESOURCE_MEM, 0);
> > > +	if (!res) {
> > > +		dev_err(dev, "failed to find registers\n");
> > > +		ret = -ENOENT;
> > > +		goto err_clk;
> > > +	}
> > > +
> > > +	ctx->regs_res = request_mem_region(res->start, resource_size(res),
> > > +					   dev_name(dev));
> > > +	if (!ctx->regs_res) {
> > > +		dev_err(dev, "failed to claim register region\n");
> > > +		ret = -ENOENT;
> > > +		goto err_clk;
> > > +	}
> > > +
> > > +	ctx->regs = ioremap(res->start, resource_size(res));
> > > +	if (!ctx->regs) {
> > > +		dev_err(dev, "failed to map registers\n");
> > > +		ret = -ENXIO;
> > > +		goto err_req_region_io;
> > > +	}
> > > +
> > > +	res = platform_get_resource(pdev, IORESOURCE_IRQ, 0);
> > > +	if (!res) {
> > > +		dev_err(dev, "irq request failed.\n");
> > > +		goto err_req_region_irq;
> > > +	}
> > > +
> > > +	ctx->irq_no = res->start;
> > > +
> > > +	for (win = 0; win < WINDOWS_NR; win++)
> > > +		fimd_clear_win(ctx, win);
> > > +
> > > +	ret = request_irq(ctx->irq_no, fimd_irq_handler, 0, "drm_fimd",
> > ctx);
> > > +	if (ret < 0) {
> > > +		dev_err(dev, "irq request failed.\n");
> > > +		goto err_req_irq;
> > > +	}
> > > +
> > > +	ctx->clkdiv = fimd_calc_clkdiv(ctx, timing);
> > > +	ctx->vidcon0 = pdata->vidcon0;
> > > +	ctx->vidcon1 = pdata->vidcon1;
> > > +	ctx->default_win = pdata->default_win;
> > > +	ctx->bpp = pdata->bpp;
> > > +	ctx->timing = timing;
> > > +
> > > +	timing->pixclock = clk_get_rate(ctx->lcd_clk) / ctx->clkdiv;
> > > +
> > > +	subdrv = &ctx->subdrv;
> > > +
> > > +	subdrv->probe = fimd_subdrv_probe;
> > > +	subdrv->remove = fimd_subdrv_remove;
> > > +	subdrv->manager.pipe = -1;
> > > +	subdrv->manager.ops = &fimd_manager_ops;
> > > +	subdrv->manager.overlay_ops = &fimd_overlay_ops;
> > > +	subdrv->manager.display = &fimd_display;
> > > +	subdrv->manager.dev = dev;
> > > +
> > > +	platform_set_drvdata(pdev, ctx);
> > > +	samsung_drm_subdrv_register(subdrv);
> > > +
> > > +	return 0;
> > > +
> > > +err_req_irq:
> > > +err_req_region_irq:
> > > +	iounmap(ctx->regs);
> > > +
> > > +err_req_region_io:
> > > +	release_resource(ctx->regs_res);
> > > +	kfree(ctx->regs_res);
> > > +
> > > +err_clk:
> > > +	clk_disable(ctx->lcd_clk);
> > > +	clk_put(ctx->lcd_clk);
> > > +
> > > +err_bus_clk:
> > > +	clk_disable(ctx->bus_clk);
> > > +	clk_put(ctx->bus_clk);
> > > +
> > > +err_clk_get:
> > > +	kfree(ctx);
> > > +	return ret;
> > > +}
> > > +
> > > +static int __devexit fimd_remove(struct platform_device *pdev)
> > > +{
> > > +	struct fimd_context *ctx = platform_get_drvdata(pdev);
> > > +
> > > +	samsung_drm_subdrv_unregister(&ctx->subdrv);
> > > +
> > > +	clk_disable(ctx->lcd_clk);
> > > +	clk_disable(ctx->bus_clk);
> > > +	clk_put(ctx->lcd_clk);
> > > +	clk_put(ctx->bus_clk);
> > > +
> > 
> > free_irq ?
> 
> Ah , I missed it. thank you.
> 
> > > +	iounmap(ctx->regs);
> > > +	release_resource(ctx->regs_res);
> > > +	kfree(ctx->regs_res);
> > > +
> > > +	kfree(ctx);
> > > +
> > > +	return 0;
> > > +}
> > > +
> > > +static struct platform_driver fimd_driver = {
> > > +	.probe		= fimd_probe,
> > > +	.remove		= __devexit_p(fimd_remove),
> > > +	.driver		= {
> > > +		.name	= "exynos4-fb",
> > > +		.owner	= THIS_MODULE,
> > > +	},
> > > +};
> > > +
> > > +static int __init fimd_init(void)
> > > +{
> > > +	return platform_driver_register(&fimd_driver);
> > > +}
> > > +
> > > +static void __exit fimd_exit(void)
> > > +{
> > > +	platform_driver_unregister(&fimd_driver);
> > > +}
> > > +
> > > +module_init(fimd_init);
> > > +module_exit(fimd_exit);
> > > +
> > > +MODULE_AUTHOR("Joonyoung Shim <jy0922.shim@samsung.com>");
> > > +MODULE_DESCRIPTION("Samsung DRM FIMD Driver");
> > > +MODULE_LICENSE("GPL");
> > > diff --git a/drivers/gpu/drm/samsung/samsung_drm_gem.c
> > b/drivers/gpu/drm/samsung/samsung_drm_gem.c
> > > new file mode 100644
> > > index 0000000..8f71ef0
> > > --- /dev/null
> > > +++ b/drivers/gpu/drm/samsung/samsung_drm_gem.c
> > > @@ -0,0 +1,492 @@
> > > +/* samsung_drm_gem.c
> > > + *
> > > + * Copyright (c) 2011 Samsung Electronics Co., Ltd.
> > > + * Author: Inki Dae <inki.dae@samsung.com>
> > > + *
> > > + * Permission is hereby granted, free of charge, to any person
> > obtaining a
> > > + * copy of this software and associated documentation files (the
> > "Software"),
> > > + * to deal in the Software without restriction, including without
> > limitation
> > > + * the rights to use, copy, modify, merge, publish, distribute,
> > sublicense,
> > > + * and/or sell copies of the Software, and to permit persons to whom
> > the
> > > + * Software is furnished to do so, subject to the following conditions:
> > > + *
> > > + * The above copyright notice and this permission notice (including the
> > next
> > > + * paragraph) shall be included in all copies or substantial portions
> > of the
> > > + * Software.
> > > + *
> > > + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
> > EXPRESS OR
> > > + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
> > MERCHANTABILITY,
> > > + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT
> > SHALL
> > > + * VA LINUX SYSTEMS AND/OR ITS SUPPLIERS BE LIABLE FOR ANY CLAIM,
> > DAMAGES OR
> > > + * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR
> OTHERWISE,
> > > + * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE
> > OR
> > > + * OTHER DEALINGS IN THE SOFTWARE.
> > > + */
> > > +
> > > +#include "drmP.h"
> > > +#include "drm.h"
> > > +
> > > +#include <drm/samsung_drm.h>
> > > +
> > > +#include "samsung_drm_drv.h"
> > > +#include "samsung_drm_gem.h"
> > > +#include "samsung_drm_buf.h"
> > > +
> > > +static unsigned int convert_to_vm_err_msg(int msg)
> > > +{
> > > +	unsigned int out_msg;
> > > +
> > > +	switch (msg) {
> > > +	case 0:
> > > +	case -ERESTARTSYS:
> > > +	case -EINTR:
> > > +		out_msg = VM_FAULT_NOPAGE;
> > > +		break;
> > > +
> > > +	case -ENOMEM:
> > > +		out_msg = VM_FAULT_OOM;
> > > +		break;
> > > +
> > > +	default:
> > > +		out_msg = VM_FAULT_SIGBUS;
> > > +		break;
> > > +	}
> > > +
> > > +	return out_msg;
> > > +}
> > > +
> > > +static unsigned int get_gem_mmap_offset(struct drm_gem_object *obj)
> > > +{
> > > +	DRM_DEBUG_KMS("%s\n", __FILE__);
> > > +
> > > +	return (unsigned int)obj->map_list.hash.key << PAGE_SHIFT;
> > > +}
> > > +
> > > +/**
> > > + * samsung_drm_gem_create_mmap_offset - create a fake mmap offset for
> > an object
> > > + * @obj: obj in question
> > > + *
> > > + * GEM memory mapping works by handing back to userspace a fake mmap
> > offset
> > > + * it can use in a subsequent mmap(2) call.  The DRM core code then
> > looks
> > > + * up the object based on the offset and sets up the various memory
> > mapping
> > > + * structures.
> > > + *
> > > + * This routine allocates and attaches a fake offset for @obj.
> > > + */
> > > +static int
> > > +samsung_drm_gem_create_mmap_offset(struct drm_gem_object *obj)
> > > +{
> > > +	struct drm_device *dev = obj->dev;
> > > +	struct drm_gem_mm *mm = dev->mm_private;
> > > +	struct drm_map_list *list;
> > > +	struct drm_local_map *map;
> > > +	int ret = 0;
> > > +
> > > +	/* Set the object up for mmap'ing */
> > > +	list = &obj->map_list;
> > > +	list->map = kzalloc(sizeof(struct drm_map_list), GFP_KERNEL);
> > > +	if (!list->map)
> > > +		return -ENOMEM;
> > > +
> > > +	map = list->map;
> > > +	map->type = _DRM_GEM;
> > > +	map->size = obj->size;
> > > +	map->handle = obj;
> > > +
> > > +	/* Get a DRM GEM mmap offset allocated... */
> > > +	list->file_offset_node = drm_mm_search_free(&mm->offset_manager,
> > > +						    obj->size / PAGE_SIZE,
> > > +						    0, 0);
> > > +	if (!list->file_offset_node) {
> > > +		DRM_ERROR("failed to allocate offset for bo %d\n",
> > > +			  obj->name);
> > > +		ret = -ENOSPC;
> > > +		goto out_free_list;
> > > +	}
> > > +
> > > +	list->file_offset_node = drm_mm_get_block(list->file_offset_node,
> > > +						  obj->size / PAGE_SIZE,
> > > +						  0);
> > > +	if (!list->file_offset_node) {
> > > +		ret = -ENOMEM;
> > > +		goto out_free_list;
> > > +	}
> > > +
> > > +	list->hash.key = list->file_offset_node->start;
> > > +	ret = drm_ht_insert_item(&mm->offset_hash, &list->hash);
> > > +	if (ret) {
> > > +		DRM_ERROR("failed to add to map hash\n");
> > > +		goto out_free_mm;
> > > +	}
> > > +
> > > +	return 0;
> > > +
> > > +out_free_mm:
> > > +	drm_mm_put_block(list->file_offset_node);
> > > +out_free_list:
> > > +	kfree(list->map);
> > > +	list->map = NULL;
> > > +
> > > +	return ret;
> > > +}
> > > +
> > > +static void
> > > +samsung_drm_gem_free_mmap_offset(struct drm_gem_object *obj)
> > > +{
> > > +	struct drm_device *dev = obj->dev;
> > > +	struct drm_gem_mm *mm = dev->mm_private;
> > > +	struct drm_map_list *list = &obj->map_list;
> > > +
> > > +	drm_ht_remove_item(&mm->offset_hash, &list->hash);
> > > +	drm_mm_put_block(list->file_offset_node);
> > > +	kfree(list->map);
> > > +	list->map = NULL;
> > > +}
> > > +
> > > +struct samsung_drm_gem_obj *samsung_drm_gem_create(struct drm_file
> > *file_priv,
> > > +		struct drm_device *dev, unsigned int size,
> > > +		unsigned int *handle)
> > > +{
> > > +	struct samsung_drm_gem_obj *samsung_gem_obj;
> > > +	struct samsung_drm_buf_entry *entry;
> > > +	struct drm_gem_object *obj;
> > > +	int ret;
> > > +
> > > +	DRM_DEBUG_KMS("%s\n", __FILE__);
> > > +
> > > +	size = roundup(size, PAGE_SIZE);
> > > +
> > > +	samsung_gem_obj = kzalloc(sizeof(*samsung_gem_obj), GFP_KERNEL);
> > > +	if (!samsung_gem_obj) {
> > > +		DRM_ERROR("failed to allocate samsung gem object.\n");
> > > +		return ERR_PTR(-ENOMEM);
> > > +	}
> > > +
> > > +	/* allocate the new buffer object and memory region. */
> > > +	entry = samsung_drm_buf_create(dev, size);
> > > +	if (!entry) {
> > > +		kfree(samsung_gem_obj);
> > > +		return ERR_PTR(-ENOMEM);
> > > +	}
> > > +
> > > +	samsung_gem_obj->entry = entry;
> > > +
> > > +	obj = &samsung_gem_obj->base;
> > > +
> > > +	ret = drm_gem_object_init(dev, obj, size);
> > > +	if (ret < 0) {
> > > +		DRM_ERROR("failed to initailize gem object.\n");
> > > +		goto err_obj_init;
> > > +	}
> > > +
> > > +	DRM_DEBUG_KMS("created file object = 0x%x\n", (unsigned int)obj-
> > >filp);
> > > +
> > > +	ret = samsung_drm_gem_create_mmap_offset(obj);
> > > +	if (ret < 0) {
> > > +		DRM_ERROR("failed to allocate mmap offset.\n");
> > > +		goto err_create_mmap_offset;
> > > +	}
> > > +
> > > +	/**
> > > +	 * allocate a 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, obj, handle);
> > > +	if (ret)
> > > +		goto err_handle_create;
> > > +
> > > +	DRM_DEBUG_KMS("gem handle = 0x%x\n", *handle);
> > > +
> > > +	/* drop reference from allocate - handle holds it now. */
> > > +	drm_gem_object_unreference_unlocked(obj);
> > > +
> > > +	return samsung_gem_obj;
> > > +
> > > +err_handle_create:
> > > +	samsung_drm_gem_free_mmap_offset(obj);
> > > +	drm_gem_object_release(obj);
> > > +
> > > +err_create_mmap_offset:
> > > +	/* add exception. FIXME. */
> > > +
> > > +err_obj_init:
> > > +	samsung_drm_buf_destroy(dev, samsung_gem_obj->entry);
> > > +
> > > +	kfree(samsung_gem_obj);
> > > +
> > > +	return ERR_PTR(ret);
> > > +}
> > > +
> > > +int samsung_drm_gem_create_ioctl(struct drm_device *dev, void *data,
> > > +		struct drm_file *file_priv)
> > > +{
> > > +	struct drm_samsung_gem_create *args = data;
> > > +	struct samsung_drm_gem_obj *samsung_gem_obj;
> > > +
> > > +	DRM_DEBUG_KMS("%s : size = 0x%x\n", __FILE__, args->size);
> > > +
> > > +	samsung_gem_obj = samsung_drm_gem_create(file_priv, dev, args->size,
> > > +			&args->handle);
> > > +	if (IS_ERR(samsung_gem_obj))
> > > +		return PTR_ERR(samsung_gem_obj);
> > > +
> > > +	return 0;
> > > +}
> > > +
> > > +int samsung_drm_gem_map_offset_ioctl(struct drm_device *dev, void
> *data,
> > > +		struct drm_file *file_priv)
> > > +{
> > > +	struct drm_samsung_gem_map_off *args = data;
> > > +
> > > +	DRM_DEBUG_KMS("%s\n", __FILE__);
> > > +
> > > +	DRM_DEBUG_KMS("handle = 0x%x, offset = 0x%x\n",
> > > +			args->handle, (u32)args->offset);
> > > +
> > > +	if (!(dev->driver->driver_features & DRIVER_GEM)) {
> > > +		DRM_ERROR("not support GEM.\n");
> > 
> > Um, You can make that error read better. Perhaps
> > 
> > drv_err(dev->dev,"Does not support GEM.\n");
> > 
> 
> Ok, get it. thank you.
> 
> > > +		return -ENODEV;
> > > +	}
> > > +
> > > +	return samsung_drm_gem_dumb_map_offset(file_priv, dev, args->handle,
> > > +			&args->offset);
> > > +}
> > > +
> > > +static int samsung_drm_gem_mmap_buffer(struct file *filp,
> > > +		struct vm_area_struct *vma)
> > > +{
> > > +	struct drm_gem_object *obj = filp->private_data;
> > > +	struct samsung_drm_gem_obj *samsung_gem_obj =
> > to_samsung_gem_obj(obj);
> > > +	struct samsung_drm_buf_entry *entry;
> > > +	unsigned long pfn, size;
> > > +
> > > +	DRM_DEBUG_KMS("%s\n", __FILE__);
> > > +
> > > +	vma->vm_flags |= (VM_IO | VM_RESERVED);
> > > +	vma->vm_page_prot = pgprot_noncached(vma->vm_page_prot);
> > > +	vma->vm_file = filp;
> > > +
> > > +	size = vma->vm_end - vma->vm_start;
> > > +	entry = samsung_gem_obj->entry;
> > > +
> > > +	/* check if the region to be mapped is valid or not. */
> > > +	if ((entry->paddr + vma->vm_pgoff + size) >
> > > +			(entry->paddr + entry->size)) {
> > > +		drm_gem_object_unreference_unlocked(obj);
> > > +		DRM_ERROR("desired size is bigger then real size.\n");
> > 
> > So .. you can't continue by just using the real size instead?
> > 
> 
> I am afraid I don't understand what you mean but I think that condition is
> fine. size is a vm area to user-desired size and you could request mapping
> as specific size. so it just check user-requested virtual space region it
> bigger than allocated physical memory region to be mapped. if there is my
> missing points, I would be happy to you give me your comments. thank you.

I meant that you return -EINVAL. But I am wondering if it would be possible
to just continue on, but ignore what the user specified.

I think the issue here is that you are outputing the DRM_ERROR and
I am not sure if it is that neccessary. Perhaps DRM_DEBUG, but DRM_ERROR
just seems a bit.. heavy handed.

> 
> > > +		return -EINVAL;
> > > +	}
> > > +
> > > +	pfn = (samsung_gem_obj->entry->paddr + vma->vm_pgoff) >> PAGE_SHIFT;
> > > +
> > > +	DRM_DEBUG_KMS("offset = 0x%x, pfn to be mapped = 0x%x\n",
> > > +			(u32)vma->vm_pgoff, (u32)pfn);
> > 
> > You can drop those u32 and use %lx
> 
> Thank you.
> 
> > > +
> > > +	if (remap_pfn_range(vma, vma->vm_start, pfn,
> > > +				vma->vm_end - vma->vm_start,
> > > +				vma->vm_page_prot)) {
> > > +		DRM_ERROR("failed to remap pfn range.\n");
> > 
> > Perhaps also include more details. Say the PFN in question, the offset,
> > etc.
> > Something that will help in the field?
> > 
> 
> Ok, get it. thank you.
> 
> > > +		return -EAGAIN;
> > > +	}
> > > +
> > > +	return 0;
> > > +}
> > > +
> > > +static const struct file_operations samsung_drm_gem_fops = {
> > > +	.mmap = samsung_drm_gem_mmap_buffer,
> > > +};
> > > +
> > > +int samsung_drm_gem_mmap_ioctl(struct drm_device *dev, void *data,
> > > +		struct drm_file *file_priv)
> > > +{
> > > +	struct drm_samsung_gem_mmap *args = data;
> > > +	struct drm_gem_object *obj;
> > > +	unsigned int addr;
> > > +
> > > +	DRM_DEBUG_KMS("%s\n", __FILE__);
> > > +
> > > +	if (!(dev->driver->driver_features & DRIVER_GEM)) {
> > > +		DRM_ERROR("not support GEM.\n");
> > 
> > Ditto.
> 
> Thank you.
> 
> > > +		return -ENODEV;
> > > +	}
> > > +
> > > +	obj = drm_gem_object_lookup(dev, file_priv, args->handle);
> > > +	if (!obj) {
> > > +		DRM_ERROR("failed to lookup gem object.\n");
> > > +		return -EINVAL;
> > > +	}
> > > +
> > > +	obj->filp->f_op = &samsung_drm_gem_fops;
> > > +	obj->filp->private_data = obj;
> > > +
> > > +	down_write(&current->mm->mmap_sem);
> > > +	addr = do_mmap(obj->filp, 0, args->size,
> > > +			PROT_READ | PROT_WRITE, MAP_SHARED, args->offset);
> > > +	up_write(&current->mm->mmap_sem);
> > > +
> > > +	drm_gem_object_unreference_unlocked(obj);
> > > +
> > > +	if (IS_ERR((void *)addr))
> > > +		return PTR_ERR((void *)addr);
> > > +
> > > +	args->mapped = addr;
> > > +
> > > +	DRM_DEBUG_KMS("mapped = 0x%x\n", (u32)args->mapped);
> > 
> > Is all this casting to (u32) that neccessary?
> 
> Definitely no, this casting should be changed to 64bit type. thank you.
> 
> > > +
> > > +	return 0;
> > > +}
> > > +
> > > +int samsung_drm_gem_init_object(struct drm_gem_object *obj)
> > > +{
> > > +	DRM_DEBUG_KMS("%s\n", __FILE__);
> > > +
> > > +	return 0;
> > > +}
> > > +
> > > +void samsung_drm_gem_free_object(struct drm_gem_object *gem_obj)
> > > +{
> > > +	struct samsung_drm_gem_obj *samsung_gem_obj;
> > > +
> > > +	DRM_DEBUG_KMS("%s\n", __FILE__);
> > > +
> > > +	DRM_DEBUG_KMS("handle count = %d\n",
> > > +			atomic_read(&gem_obj->handle_count));
> > > +
> > > +	if (gem_obj->map_list.map)
> > > +		samsung_drm_gem_free_mmap_offset(gem_obj);
> > > +
> > > +	/* release file pointer to gem object. */
> > > +	drm_gem_object_release(gem_obj);
> > > +
> > > +	samsung_gem_obj = to_samsung_gem_obj(gem_obj);
> > > +
> > > +	samsung_drm_buf_destroy(gem_obj->dev, samsung_gem_obj->entry);
> > > +}
> > > +
> > > +int samsung_drm_gem_dumb_create(struct drm_file *file_priv,
> > > +		struct drm_device *dev, struct drm_mode_create_dumb *args)
> > > +{
> > > +	struct samsung_drm_gem_obj *samsung_gem_obj;
> > > +
> > > +	DRM_DEBUG_KMS("%s\n", __FILE__);
> > > +
> > > +	/**
> > > +	 * alocate memory to be used for framebuffer.
> > > +	 * - this callback would be called by user application
> > > +	 *	with DRM_IOCTL_MODE_CREATE_DUMB command.
> > > +	 */
> > > +
> > > +	args->pitch = args->width * args->bpp >> 3;
> > > +	args->size = args->pitch * args->height;
> > > +
> > > +	samsung_gem_obj = samsung_drm_gem_create(file_priv, dev, args->size,
> > > +							&args->handle);
> > > +	if (IS_ERR(samsung_gem_obj))
> > > +		return PTR_ERR(samsung_gem_obj);
> > > +
> > > +	return 0;
> > > +}
> > > +
> > > +int samsung_drm_gem_dumb_map_offset(struct drm_file *file_priv,
> > > +		struct drm_device *dev, uint32_t handle, uint64_t *offset)
> > > +{
> > > +	struct samsung_drm_gem_obj *samsung_gem_obj;
> > > +	struct drm_gem_object *obj;
> > > +
> > > +	DRM_DEBUG_KMS("%s\n", __FILE__);
> > > +
> > > +	mutex_lock(&dev->struct_mutex);
> > > +
> > > +	/**
> > > +	 * get offset of memory allocated for drm framebuffer.
> > > +	 * - this callback would be called by user application
> > > +	 *	with DRM_IOCTL_MODE_MAP_DUMB command.
> > > +	 */
> > > +
> > > +	obj = drm_gem_object_lookup(dev, file_priv, handle);
> > > +	if (!obj) {
> > > +		DRM_ERROR("failed to lookup gem object.\n");
> > > +		mutex_unlock(&dev->struct_mutex);
> > > +		return -EINVAL;
> > > +	}
> > > +
> > > +	samsung_gem_obj = to_samsung_gem_obj(obj);
> > > +
> > > +	*offset = get_gem_mmap_offset(&samsung_gem_obj->base);
> > > +
> > > +	drm_gem_object_unreference(obj);
> > > +
> > > +	DRM_DEBUG_KMS("offset = 0x%x\n", (unsigned int)*offset);
> > > +
> > > +	mutex_unlock(&dev->struct_mutex);
> > > +
> > > +	return 0;
> > > +}
> > > +
> > > +int samsung_drm_gem_fault(struct vm_area_struct *vma, struct vm_fault
> > *vmf)
> > > +{
> > > +	struct drm_gem_object *obj = vma->vm_private_data;
> > > +	struct samsung_drm_gem_obj *samsung_gem_obj =
> > to_samsung_gem_obj(obj);
> > > +	struct drm_device *dev = obj->dev;
> > > +	unsigned long pfn;
> > > +	pgoff_t page_offset;
> > > +	int ret;
> > > +
> > > +	page_offset = ((unsigned long)vmf->virtual_address -
> > > +			vma->vm_start) >> PAGE_SHIFT;
> > > +
> > > +	mutex_lock(&dev->struct_mutex);
> > > +
> > > +	pfn = (samsung_gem_obj->entry->paddr >> PAGE_SHIFT) + page_offset;
> > > +
> > > +	ret = vm_insert_mixed(vma, (unsigned long)vmf->virtual_address,
> > pfn);
> > > +
> > > +	mutex_unlock(&dev->struct_mutex);
> > > +
> > > +	return convert_to_vm_err_msg(ret);
> > > +}
> > > +
> > > +int samsung_drm_gem_mmap(struct file *filp, struct vm_area_struct *vma)
> > > +{
> > > +	int ret;
> > > +
> > > +	DRM_DEBUG_KMS("%s\n", __FILE__);
> > > +
> > > +	/* set vm_area_struct. */
> > > +	ret = drm_gem_mmap(filp, vma);
> > > +	if (ret < 0) {
> > > +		DRM_ERROR("failed to mmap.\n");
> > > +		return ret;
> > > +	}
> > > +
> > > +	vma->vm_flags &= ~VM_PFNMAP;
> > > +	vma->vm_flags |= VM_MIXEDMAP;
> > > +
> > > +	return ret;
> > > +}
> > > +
> > > +
> > > +int samsung_drm_gem_dumb_destroy(struct drm_file *file_priv,
> > > +		struct drm_device *dev, unsigned int handle)
> > > +{
> > > +	int ret;
> > > +
> > > +	DRM_DEBUG_KMS("%s\n", __FILE__);
> > > +
> > > +	/**
> > > +	 * obj->refcount and obj->handle_count are decreased and
> > > +	 * if both them are 0 then samsung_drm_gem_free_object()
> > > +	 * would be called by callback to release resources.
> > > +	 */
> > > +	ret = drm_gem_handle_delete(file_priv, handle);
> > > +	if (ret < 0) {
> > > +		DRM_ERROR("failed to delete drm_gem_handle.\n");
> > > +		return ret;
> > > +	}
> > > +
> > > +	return 0;
> > > +}
> > > +
> > > +MODULE_AUTHOR("Inki Dae <inki.dae@samsung.com>");
> > > +MODULE_DESCRIPTION("Samsung SoC DRM GEM Module");
> > > +MODULE_LICENSE("GPL");
> > > diff --git a/drivers/gpu/drm/samsung/samsung_drm_gem.h
> > b/drivers/gpu/drm/samsung/samsung_drm_gem.h
> > > new file mode 100644
> > > index 0000000..c6cd5e5
> > > --- /dev/null
> > > +++ b/drivers/gpu/drm/samsung/samsung_drm_gem.h
> > > @@ -0,0 +1,98 @@
> > > +/* samsung_drm_gem.h
> > > + *
> > > + * Copyright (c) 2011 Samsung Electronics Co., Ltd.
> > > + * Authoer: Inki Dae <inki.dae@samsung.com>
> > > + *
> > > + * Permission is hereby granted, free of charge, to any person
> > obtaining a
> > > + * copy of this software and associated documentation files (the
> > "Software"),
> > > + * to deal in the Software without restriction, including without
> > limitation
> > > + * the rights to use, copy, modify, merge, publish, distribute,
> > sublicense,
> > > + * and/or sell copies of the Software, and to permit persons to whom
> > the
> > > + * Software is furnished to do so, subject to the following conditions:
> > > + *
> > > + * The above copyright notice and this permission notice (including the
> > next
> > > + * paragraph) shall be included in all copies or substantial portions
> > of the
> > > + * Software.
> > > + *
> > > + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
> > EXPRESS OR
> > > + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
> > MERCHANTABILITY,
> > > + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT
> > SHALL
> > > + * VA LINUX SYSTEMS AND/OR ITS SUPPLIERS BE LIABLE FOR ANY CLAIM,
> > DAMAGES OR
> > > + * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR
> OTHERWISE,
> > > + * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE
> > OR
> > > + * OTHER DEALINGS IN THE SOFTWARE.
> > > + */
> > > +
> > > +#ifndef _SAMSUNG_DRM_GEM_H_
> > > +#define _SAMSUNG_DRM_GEM_H_
> > > +
> > > +#define to_samsung_gem_obj(x)	container_of(x,\
> > > +			struct samsung_drm_gem_obj, base)
> > > +
> > > +/**
> > > + * samsung drm buffer structure.
> > > + *
> > > + * @entry: pointer to samsung drm buffer entry object.
> > > + * @flags: it means memory type to be alloated or cache attributes.
> > 
> > Hm, that sounds like an enum? Or a candidate for enums?
> 
> 
> Yes, it's better to use enums. Thank you.
> 
> 
> > > + * @handle: gem handle.
> > 
> > You are missing @base
> > 
> 
> Yes, I missed it. thank you.
> 
> > > + *
> > > + * ps. this object would be transfered to user as kms_bo.handle so
> > 
> > It is 'P.S.'
> 
> Thank you.
> 
> > > + *	user can access to memory through kms_bo.handle.
> > > + */
> > > +struct samsung_drm_gem_obj {
> > > +	struct drm_gem_object base;
> > > +	struct samsung_drm_buf_entry *entry;
> > > +	unsigned int flags;
> > 
> > Actually, I don't think you are using this? Could it be chopped off?
> 
> Could you give me your more comments. I am afraid what is 'it'. I assume
> maybe it's flags. Yes, flags isn't used yet. I will remove it until this
> feature would be added. thank you.


Yup. That is the one I was thinking off.
> 
> > > +};
> > > +
> > > +
> > > +/* create a new buffer and get a new gem handle. */
> > > +struct samsung_drm_gem_obj *samsung_drm_gem_create(struct drm_file
> > *file_priv,
> > > +		struct drm_device *dev, unsigned int size,
> > > +		unsigned int *handle);
> > > +
> > > +/*
> > > + * request gem object creation and buffer allocation as the size
> > > + * that it is calculated with framebuffer information such as width,
> > > + * height and bpp.
> > > + */
> > > +int samsung_drm_gem_create_ioctl(struct drm_device *dev, void *data,
> > > +		struct drm_file *file_priv);
> > > +
> > > +/* get buffer offset to map to user space. */
> > > +int samsung_drm_gem_map_offset_ioctl(struct drm_device *dev, void
> *data,
> > > +		struct drm_file *file_priv);
> > > +
> > > +/* unmap a buffer from user space. */
> > > +int samsung_drm_gem_munmap_ioctl(struct drm_device *dev, void *data,
> > > +		struct drm_file *file_priv);
> > > +
> > > +/* initialize gem object. */
> > > +int samsung_drm_gem_init_object(struct drm_gem_object *obj);
> > > +
> > > +/* free gem object. */
> > > +void samsung_drm_gem_free_object(struct drm_gem_object *gem_obj);
> > > +
> > > +/* create memory region for drm framebuffer. */
> > > +int samsung_drm_gem_dumb_create(struct drm_file *file_priv,
> > > +		struct drm_device *dev, struct drm_mode_create_dumb *args);
> > > +
> > > +/* map memory region for drm framebuffer to user space. */
> > > +int samsung_drm_gem_dumb_map_offset(struct drm_file *file_priv,
> > > +		struct drm_device *dev, uint32_t handle, uint64_t *offset);
> > > +
> > > +/* page fault handler and mmap fault address(virtual) to physical
> > memory. */
> > > +int samsung_drm_gem_fault(struct vm_area_struct *vma, struct vm_fault
> > *vmf);
> > > +
> > > +/* mmap gem object. */
> > > +int samsung_drm_gem_mmap_ioctl(struct drm_device *dev, void *data,
> > > +		struct drm_file *file_priv);
> > > +
> > > +/* set vm_flags and we can change vm attribute to other here. */
> > > +int samsung_drm_gem_mmap(struct file *filp, struct vm_area_struct
> *vma);
> > > +
> > > +/* destroy memory region allocated. */
> > > +int samsung_drm_gem_dumb_destroy(struct drm_file *file_priv,
> > > +		struct drm_device *dev, unsigned int handle);
> > > +
> > > +#endif
> > > diff --git a/include/drm/samsung_drm.h b/include/drm/samsung_drm.h
> > > new file mode 100644
> > > index 0000000..5f89cf1
> > > --- /dev/null
> > > +++ b/include/drm/samsung_drm.h
> > > @@ -0,0 +1,103 @@
> > > +/* samsung_drm.h
> > > + *
> > > + * Copyright (c) 2011 Samsung Electronics Co., Ltd.
> > > + * Authors:
> > > + *	Inki Dae <inki.dae@samsung.com>
> > > + *	Joonyoung Shim <jy0922.shim@samsung.com>
> > > + *
> > > + * Permission is hereby granted, free of charge, to any person
> > obtaining a
> > > + * copy of this software and associated documentation files (the
> > "Software"),
> > > + * to deal in the Software without restriction, including without
> > limitation
> > > + * the rights to use, copy, modify, merge, publish, distribute,
> > sublicense,
> > > + * and/or sell copies of the Software, and to permit persons to whom
> > the
> > > + * Software is furnished to do so, subject to the following conditions:
> > > + *
> > > + * The above copyright notice and this permission notice (including the
> > next
> > > + * paragraph) shall be included in all copies or substantial portions
> > of the
> > > + * Software.
> > > + *
> > > + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
> > EXPRESS OR
> > > + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
> > MERCHANTABILITY,
> > > + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT
> > SHALL
> > > + * VA LINUX SYSTEMS AND/OR ITS SUPPLIERS BE LIABLE FOR ANY CLAIM,
> > DAMAGES OR
> > > + * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR
> OTHERWISE,
> > > + * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE
> > OR
> > > + * OTHER DEALINGS IN THE SOFTWARE.
> > > + */
> > > +
> > > +#ifndef _SAMSUNG_DRM_H_
> > > +#define _SAMSUNG_DRM_H_
> > > +
> > > +/**
> > > + * User-desired buffer creation information structure.
> > > + *
> > > + * @size: requested size for the object.
> > > + *	- this size value would be page-aligned internally.
> > > + * @flags: user request for setting memory type or cache attributes.
> > > + * @handle: returned handle for the object.
> > > + */
> > > +struct drm_samsung_gem_create {
> > > +	unsigned int size;
> > > +	unsigned int flags;
> > > +
> > 
> > You can deleta that space.
> 
> Ok, get it. thank you.
> 
> > > +	unsigned int handle;
> > > +};
> > > +
> > > +/**
> > > + * A structure for getting buffer offset.
> > > + *
> > > + * @handle: a pointer to gem object created.
> > > + * @offset: relatived offset value of the memory region allocated.
> > > + *	- this value should be set by user.
> > > + */
> > > +struct drm_samsung_gem_map_off {
> > > +	unsigned int handle;
> > > +	uint64_t offset;
> > > +};
> > > +
> > > +/**
> > > + * A structure for mapping buffer.
> > > + *
> > > + * @handle: a pointer to gem object created.
> > > + * @offset: relatived offset value of the memory region allocated.
> > > + *	- this value should be set by user.
> > > + * @size: memory size to be mapped.
> > > + * @mapped: user virtual address to be mapped.
> > > + */
> > > +struct drm_samsung_gem_mmap {
> > > +	unsigned int handle;
> > > +	uint64_t offset;
> > > +	unsigned int size;
> > > +
> > 
> > Ditto
> 
> Thank you.
> 
> > > +	uint64_t mapped;
> > > +};
> > > +
> > > +#define DRM_SAMSUNG_GEM_CREATE		0x00
> > > +#define DRM_SAMSUNG_GEM_MAP_OFFSET	0x01
> > > +#define DRM_SAMSUNG_GEM_MMAP		0x02
> > > +
> > > +#define DRM_IOCTL_SAMSUNG_GEM_CREATE
> DRM_IOWR(DRM_COMMAND_BASE +
> > \
> > > +		DRM_SAMSUNG_GEM_CREATE, struct drm_samsung_gem_create)
> > > +
> > > +#define DRM_IOCTL_SAMSUNG_GEM_MAP_OFFSET	DRM_IOWR(DRM_COMMAND_BASE +
> > \
> > > +		DRM_SAMSUNG_GEM_MAP_OFFSET, struct drm_samsung_gem_map_off)
> > > +
> > > +#define DRM_IOCTL_SAMSUNG_GEM_MMAP	DRM_IOWR(DRM_COMMAND_BASE + \
> > > +		DRM_SAMSUNG_GEM_MMAP, struct drm_samsung_gem_mmap)
> > > +
> > > +/**
> > > + * Platform Specific Structure for DRM based FIMD.
> > > + *
> > > + * @timing: default video mode for initializing
> > > + * @default_win: default window layer number to be used for UI.
> > > + * @bpp: default bit per pixel.
> > > + */
> > > +struct samsung_drm_fimd_pdata {
> > > +	struct fb_videomode		timing;
> > > +	u32				vidcon0;
> > > +	u32				vidcon1;
> > > +	unsigned int			default_win;
> > > +	unsigned int			bpp;
> > > +};
> > > +
> > > +#endif
> > > --
> > > 1.7.0.4


You might also include in V5 the comments about CMA/DMA and about
exhaustion so that if somebody looks at the code in a year they
won't wonder about it anymore.

[PATCH v4] DRM: add DRM Driver for Samsung SoC EXYNOS4210.

[Konrad Rzeszutek Wilk]

> > > > > +		DRM_ERROR("desired size is bigger then real
> size.\n");
> > > >
> > > > So .. you can't continue by just using the real size instead?
> > > >
> > >
> > > I am afraid I don't understand what you mean but I think that condition
> > is
> > > fine. size is a vm area to user-desired size and you could request
> > mapping
> > > as specific size. so it just check user-requested virtual space region
> > it
> > > bigger than allocated physical memory region to be mapped. if there is
> > my
> > > missing points, I would be happy to you give me your comments. thank
> you.
> > 
> > I meant that you return -EINVAL. But I am wondering if it would be
> > possible
> > to just continue on, but ignore what the user specified.
> > 
> > I think the issue here is that you are outputing the DRM_ERROR and
> > I am not sure if it is that neccessary. Perhaps DRM_DEBUG, but DRM_ERROR
> > just seems a bit.. heavy handed.
> > 
> 
> I thought this condition would be critical issue as user application should
> be terminated. is your wondering if user application could go ahead after
> mmap failed? this is just my view so I would be happy you to give me your
> advice. Thank you.

I think terminating the appliaction is the right thing. But the DRM_ERROR
is not neccessary.

[PATCH v4] DRM: add DRM Driver for Samsung SoC EXYNOS4210.

[Inki Dae]

Hello, Konrad Rzeszutek Wilk.

> -----Original Message-----
> From: Konrad Rzeszutek Wilk [mailto:konrad.wilk at oracle.com]
> Sent: Thursday, September 22, 2011 3:53 AM
> To: Inki Dae
> Cc: airlied at linux.ie; dri-devel at lists.freedesktop.org;
> sw0312.kim at samsung.com; kyungmin.park at samsung.com; linux-arm-
> kernel at lists.infradead.org
> Subject: Re: [PATCH v4] DRM: add DRM Driver for Samsung SoC EXYNOS4210.
> 
> > > > > > +		DRM_ERROR("desired size is bigger then real
> > size.\n");
> > > > >
> > > > > So .. you can't continue by just using the real size instead?
> > > > >
> > > >
> > > > I am afraid I don't understand what you mean but I think that
> condition
> > > is
> > > > fine. size is a vm area to user-desired size and you could request
> > > mapping
> > > > as specific size. so it just check user-requested virtual space
> region
> > > it
> > > > bigger than allocated physical memory region to be mapped. if there
> is
> > > my
> > > > missing points, I would be happy to you give me your comments. thank
> > you.
> > >
> > > I meant that you return -EINVAL. But I am wondering if it would be
> > > possible
> > > to just continue on, but ignore what the user specified.
> > >
> > > I think the issue here is that you are outputing the DRM_ERROR and
> > > I am not sure if it is that neccessary. Perhaps DRM_DEBUG, but
> DRM_ERROR
> > > just seems a bit.. heavy handed.
> > >
> >
> > I thought this condition would be critical issue as user application
> should
> > be terminated. is your wondering if user application could go ahead
> after
> > mmap failed? this is just my view so I would be happy you to give me
> your
> > advice. Thank you.
> 
> I think terminating the appliaction is the right thing. But the DRM_ERROR
> is not neccessary.

I understood. I will remove DRM_ERROR or use DRM_DEBUG instead. thank you.