[RFC v3 1/3] fbmem: add default get_fb_unmapped_area function

[Laurent Pinchart]

Hi Benjamin,

On Thursday 01 Dec 2016 08:24:50 Benjamin Gaignard wrote:
> 2016-12-01 2:13 GMT+01:00 Laurent Pinchart:
> > On Wednesday 30 Nov 2016 20:46:23 Benjamin Gaignard wrote:
> >> 2016-11-30 20:39 GMT+01:00 Laurent Pinchart:
> >>> On Wednesday 30 Nov 2016 20:34:17 Benjamin Gaignard wrote:
> >>>> If HAVE_ARCH_FB_UNMAPPED_AREA is set and get_fb_unmapped_area
> >>>> function not defined in platform architecture let use a default
> >>>> function.
> >>>> 
> >>>> Signed-off-by: Benjamin Gaignard <benjamin.gaignard at linaro.org>
> >>>> ---
> >>>> 
> >>>>  drivers/video/fbdev/core/fbmem.c | 15 +++++++++++++++
> >>>>  1 file changed, 15 insertions(+)
> >>>> 
> >>>> diff --git a/drivers/video/fbdev/core/fbmem.c
> >>>> b/drivers/video/fbdev/core/fbmem.c index 76c1ad9..54488ee 100644
> >>>> --- a/drivers/video/fbdev/core/fbmem.c
> >>>> +++ b/drivers/video/fbdev/core/fbmem.c
> >>>> @@ -1492,6 +1492,21 @@ static long fb_compat_ioctl(struct file *file,
> >>>> unsigned int cmd,
> >>>>    return 0;
> >>>>  }
> >>>> 
> >>>> +#if defined(HAVE_ARCH_FB_UNMAPPED_AREA) &&
> >>>> !defined(get_fb_unmapped_area)
> >>> 
> >>> I think I've asked this twice already, how is that possible ?
> >> 
> >> As you said before only sparc and blackfin architectures have define
> >> HAVE_ARCH_FB_UNMAPPED_AREA and get_fb_unmapped_area().
> >> For stm32f4 (ARMv7m) I plan to only add HAVE_ARCH_FB_UNMAPPED_AREA and
> >> use this default function
> > 
> > Ah, so it's for support of a future patch. You should explain that in the
> > commit message.
> 
> OK
> 
> > The other question I asked was whether it would be possible to come up
> > with a generic implementation of this function that would work on all
> > architectures. At first sight the blackfin implementation is similar, but
> > the sparc one is quite different. Why does it have to be arch-specific ?
> 
> I don't know but I guess it is link to how memory is mmapped by the
> architecture

Well, how about trying to find out then ? :-)

> >> By testing those two flags I avoid to define twince
> >> get_fb_unmapped_area() if the architecture already
> >> provide it.
> >> 
> >>>> +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;
> >>>> +}
> >>>> +#endif
> >>>> +
> >>>>  static const struct file_operations fb_fops = {
> >>>>       .owner =        THIS_MODULE,
> >>>>       .read =         fb_read,

-- 
Regards,

Laurent Pinchart