[U-Boot] [PATCH 3/5] fdt: add fdt_add_display_timings(..)

[Christian Gmeiner]

2014-01-16 12:44 GMT+01:00 Christian Gmeiner <christian.gmeiner at gmail.com>:
> Hi all
>
> 2014/1/14 Christian Gmeiner <christian.gmeiner at gmail.com>:
>> 2014/1/12 Anatolij Gustschin <agust at denx.de>:
>>> Hi Stefano,
>>>
>>> On Wed, 08 Jan 2014 11:53:39 +0100
>>> Stefano Babic <sbabic at denx.de> wrote:
>>> ...
>>>> Agree that we have to sync u-boot and kernel, and this can be a way in
>>>> the short term.
>>>>
>>>> I am asking if this is in the long term the best way to do it. You are
>>>> converting EDID values to fb_videomode *mode, and then again to the
>>>> device node as required by DT.
>>>> We have already had some talks about moving U-Boot configuration to DT,
>>>> that is U-Boot can be also configured by a DT file (see for example
>>>> support for Nvidia processors, they already support DT in U-Boot).
>>>>
>>>> Anatolji, what do you think as best solution we have to follow for
>>>> display setting ?
>>>
>>> many drivers use struct fb_videomode internally and this display-timings
>>> binding already exists in linux, so I think a function for converting
>>> from fb_videomode to DT is useful. However we should probably extend
>>> this current implementation of the function, e.g. rename it to
>>> fdt_update_display_timings() and pass more arguments: node compatible
>>> and the name of the parent node containing the display-timings node.
>>> The code for searching the display-timings node is also needed for
>>> other boards, so if it is in the function itself, it will simplify
>>> the usage.
>>>
>>> This function could look for display-timings node and create it if
>>> it doesn't exist. Or update the existing node with new info.
>>>
>>
>> Thanks for your comments... will come up with something in the next version
>> of the patch series.
>>
>
> I have some time to work on this patch.
>
> Anatolij are you happy witht the following functions?
>
> int fdt_find_display_timings(void *fdt, const char *compat, const char *parent)
> {
>     int coff = fdt_node_offset_by_compatible(fdt, -1, compat);
>     int poff = fdt_subnode_offset(fdt, coff, parent);
>     int timings = fdt_subnode_offset(fdt, poff, "display-timings");
>
>     return timings;
> }
>
> int fdt_update_display_timings(void *fdt, const char *compat, const
> char *parent, struct fb_videomode *mode)
> {
>     int timings = fdt_find_display_timings(fdt, compat, parent);
>
>     /* check if display-timings subnode does exist */
>     if (timings == -FDT_ERR_NOTFOUND) {
>         return timings;
>     }
>
>     /* set all needed properties */
>     if (timings != -FDT_ERR_NOTFOUND) {
>         fdt_setprop_u32(fdt, noff, "clock-frequency", mode->pixclock);
>         fdt_setprop_u32(fdt, noff, "hactive", mode->xres);
>         fdt_setprop_u32(fdt, noff, "vactive", mode->yres);
>         fdt_setprop_u32(fdt, noff, "hback-porch", mode->left_margin);
>         fdt_setprop_u32(fdt, noff, "hfront-porch", mode->right_margin);
>         fdt_setprop_u32(fdt, noff, "vback-porch", mode->upper_margin);
>         fdt_setprop_u32(fdt, noff, "vfront-porch", mode->lower_margin);
>         fdt_setprop_u32(fdt, noff, "hsync-len", mode->hsync_len);
>         fdt_setprop_u32(fdt, noff, "vsync-len", mode->vsync_len);
>     }
>
>     return 0;
> }
>
> This would allow us to update an existing display-timings node via
>
> fdt_update_display_timings(blob, "fsl,imx6q-ldb", "lvds-channel", mode);
>
>
> Note: not even compile tested :)
>

I got no feedback over some months - bad. But okay I really want to get all
the bits of our imx6 based board upstream as soon as possible. So has something
changed in that area?

In the end I want to read EDID stored in an eeprom and put it into the
device tree (patching or creating new node).

greets
--
Christian Gmeiner, MSc

https://soundcloud.com/christian-gmeiner