Hi,
On 10/1/25 12:26 PM, Guillermo Rodriguez Garcia wrote:
Hi Marius,
Yes, indeed, setting transform=normal fixes the issue.
Is this the expected behaviour?
Yes. Re-reading that MR actually explains the fact that it allows
users to override that. My initial implementation wasn't giving you that
option at all.
Speculate this is the normal behaviour as *not* all drivers perform the
rotation on the their own. Happy to take documentation patches that
would further clarify this.
In my experience, a rotation=... in the device tree is a configuration
parameter for the display controller (LCDC/DRM driver) that tells the
kernel driver that the panel is mounted physically rotated, and the
driver should compensate by rotating the framebuffer output before
sending it to the panel. I would not expect userspace to apply an
additional rotation.
Weston seems to be interpreting this as "the panel is physically
rotated, and it is Weston's job to compensate for that rotation,
unless a different transform is explicitly requested". I find this a
bit confusing.
Best regards,
Guillermo
El mar, 30 sept 2025 a las 23:21, Marius Vlad
(<[email protected]>) escribió:
Hi,
Could you try overriding the transform to 'normal' for that particular
output and see if that makes a difference?
If the LCD driver does the flip on its own, you should be able
to tell Weston to avoid doing the same.
On 9/30/25 8:06 PM, Guillermo Rodriguez Garcia wrote:
Hi Daniel,
Thanks for your answer. This is Weston 10. If I am correct this
includes the code from
https://gitlab.freedesktop.org/wayland/weston/-/merge_requests/315
which added support for the panel rotation property.
The thing is, I believe Weston *is* actually parsing the property. But
I also suspect that it is applying the same rotation again (which in
effect "unrotates" the display). In other words:
- When the system boots, the LCD controller driver parses the device
tree, sees rotation = <180>, and rotates the display
- At this point everything is correct. modetest reports that the
display is upside down and if I draw directly to the display, the
image is actually rotated, as it should
- Then I start Weston
- And the image looks "unrotated" again. But I suspect that this is
because Weston applied another 180 rotation to a display that was
already rotated by the LCD controller driver.
Does this make sense?
Guillermo
El mar, 30 sept 2025 a las 14:16, Daniel Stone
(<[email protected]>) escribió:
Hi,
On Tue, 30 Sept 2025 at 12:45, Guillermo Rodriguez Garcia
<[email protected]> wrote:
My problem is that when I start Weston, the image is NOT rotated. Or rather it is
probably rotated twice (modetest still reports "upside down"). I am not
specifying any transform in weston.ini. Is this the expected behaviour?
Which version of Weston are you using? We have supported the 'panel
rotation' property for a long time, but maybe you are on a very old
version which doesn't have this, or a vendor derivative which doesn't
work with it.
Cheers,
Daniel
