[RFC 1/1] drm: allow to use mmuless SoC

Benjamin Gaignard benjamin.gaignard at linaro.org
Wed Nov 30 12:35:00 UTC 2016


2016-11-30 12:35 GMT+01:00 Laurent Pinchart <laurent.pinchart at ideasonboard.com>:
> Hi Benjamin,
>
> Thank you for the patch.
>
> On Wednesday 30 Nov 2016 12:21:24 Benjamin Gaignard wrote:
>> Some platforms without MMU have display drivers where a drm/kms driver
>> could be implemented.
>> Before doing such kind of thing drm/kms must allow to use mmuless
>> devices.
>> This patch proposes to remove MMU configuration flag and add some
>> cma helpers functions to help implementing mmuless display driver
>>
>> Signed-off-by: Benjamin Gaignard <benjamin.gaignard at linaro.org>
>> ---
>>  drivers/gpu/drm/Kconfig              |  4 +--
>>  drivers/gpu/drm/drm_fb_cma_helper.c  | 20 ++++++++++++
>>  drivers/gpu/drm/drm_gem_cma_helper.c | 62 +++++++++++++++++++++++++++++++++
>>  drivers/gpu/drm/drm_vm.c             |  4 +++
>
> I'll let someone more knowledgeable than me comment on these two files.
>
>>  drivers/video/fbdev/core/fbmem.c     |  8 +++++
>>  include/drm/drm_gem_cma_helper.h     |  6 ++++
>>  6 files changed, 102 insertions(+), 2 deletions(-)
>
> [snip]
>
>> diff --git a/drivers/gpu/drm/drm_fb_cma_helper.c
>> b/drivers/gpu/drm/drm_fb_cma_helper.c index 81b3558..a3cef79 100644
>> --- a/drivers/gpu/drm/drm_fb_cma_helper.c
>> +++ b/drivers/gpu/drm/drm_fb_cma_helper.c
>> @@ -343,10 +343,30 @@ int drm_fb_cma_debugfs_show(struct seq_file *m, void
>> *arg)
>>
>>  static int drm_fb_cma_mmap(struct fb_info *info, struct vm_area_struct
>> *vma) {
>> +#ifdef CONFIG_MMU
>>       return dma_mmap_writecombine(info->device, vma, info->screen_base,
>>                                    info->fix.smem_start, info-
>>fix.smem_len);
>> +#else
>> +     return vm_iomap_memory(vma, vma->vm_start, info->fix.smem_len);
>> +#endif
>>  }
>>
>> +#ifdef HAVE_ARCH_FB_UNMAPPED_AREA
>
> Shouldn't this be #ifndef ? When the symbol is defined the implementation is
> provided by the architecture.
>
> By the way, given that you provide a generic implementation here, what's the
> reason some architecture need to implement the function in a special way ?
> Couldn't we come up with a generic implementation across all architectures ?
>
>> +unsigned long get_fb_unmapped_area(struct file *filp,
>> +                                unsigned long addr, unsigned long len,
>> +                                unsigned long pgoff, unsigned long flags)
>> +{
>> +     struct fb_info * const info = filp->private_data;
>> +     unsigned long fb_size = PAGE_ALIGN(info->fix.smem_len);
>> +
>> +     if (pgoff > fb_size || len > fb_size - pgoff)
>> +             return -EINVAL;
>> +
>> +     return (unsigned long)info->screen_base + pgoff;
>> +}
>> +EXPORT_SYMBOL_GPL(get_fb_unmapped_area);
>
> Shouldn't this be defined somewhere in drivers/video/ instead, as the code is
> fbdev-specific and only used in drivers/video/fbdev/core/fbmem.c ?
>

maybe I could migrate this code in drivers/video/fbdev/core/fbmem.c and play
with HAVE_ARCH_FB_UNMAPPED_AREA and get_fb_unmapped_area to know
if the symbol already exist or not.
I think that would also fix your previous and next remark because I
this case I will not
have declare the prototype.

>> +#endif
>> +
>>  static struct fb_ops drm_fbdev_cma_ops = {
>>       .owner          = THIS_MODULE,
>>       DRM_FB_HELPER_DEFAULT_OPS,
>
> [snip]
>
>> diff --git a/drivers/video/fbdev/core/fbmem.c
>> b/drivers/video/fbdev/core/fbmem.c index 76c1ad9..d6a83bd 100644
>> --- a/drivers/video/fbdev/core/fbmem.c
>> +++ b/drivers/video/fbdev/core/fbmem.c
>> @@ -1492,6 +1492,14 @@ static long fb_compat_ioctl(struct file *file,
>> unsigned int cmd, return 0;
>>  }
>>
>> +#ifdef HAVE_ARCH_FB_UNMAPPED_AREA
>> +extern unsigned long get_fb_unmapped_area(struct file *filp,
>> +                                       unsigned long addr,
>> +                                       unsigned long len,
>> +                                       unsigned long pgoff,
>> +                                       unsigned long flags);
>> +#endif
>
> This should come from a header file.
>
>>  static const struct file_operations fb_fops = {
>>       .owner =        THIS_MODULE,
>>       .read =         fb_read,
>> diff --git a/include/drm/drm_gem_cma_helper.h
>> b/include/drm/drm_gem_cma_helper.h index acd6af8..79e9dbd 100644
>> --- a/include/drm/drm_gem_cma_helper.h
>> +++ b/include/drm/drm_gem_cma_helper.h
>> @@ -51,6 +51,12 @@ int drm_gem_cma_dumb_map_offset(struct drm_file
>> *file_priv, struct drm_gem_cma_object *drm_gem_cma_create(struct drm_device
>> *drm, size_t size);
>>
>> +unsigned long drm_gem_cma_get_unmapped_area(struct file *filp,
>> +                                         unsigned long addr,
>> +                                         unsigned long len,
>> +                                         unsigned long pgoff,
>> +                                         unsigned long flags);
>> +
>>  extern const struct vm_operations_struct drm_gem_cma_vm_ops;
>>
>>  #ifdef CONFIG_DEBUG_FS
>
> --
> Regards,
>
> Laurent Pinchart
>



-- 
Benjamin Gaignard

Graphic Study Group

Linaro.org │ Open source software for ARM SoCs

Follow Linaro: Facebook | Twitter | Blog


More information about the dri-devel mailing list