[xserver,2/2,v4] xwayland: Apply output rotation for screen size

Submitted by Olivier Fourdan on Feb. 8, 2017, 8:23 a.m.

Details

Message ID 20170208082320.7414-1-ofourdan@redhat.com
State Accepted
Commit 058809c43ec578a407cf40d4c3e54a42503e3562
Headers show
Series "xwayland: Pretend we support all transformations" ( rev: 4 ) in X.org (DEPRECATED - USE GITLAB)

Not browsing as part of any series.

Commit Message

Olivier Fourdan Feb. 8, 2017, 8:23 a.m.
Previously, we would swap the width/height of the Xwayland output based
on the output rotation, so that the overall screen size would match the
actual rotation of each output.

Problem is the RandR's ConstrainCursorHarder() handler will also apply
the output rotation, meaning that when the output is rotated, the
pointer will be constrained within the wrong dimension.

Moreover, XRandR assumes the original output width/height are unchanged
when the output is rotated, so by changing the Xwayland output width and
height based on rotation, Xwayland causes XRandr to report the wrong
output sizes (an output of size 1024x768 rotated left or right should
remain 1024x768, not 768x1024).

So to avoid this issue and keep things consistent between Wayland and
Xwayland outputs, leave the actual width/height unchanged but apply the
rotation when computing the screen size. This fixes both the output size
being wrong in "xrandr -q" and the pointer being constrained in the
wrong dimension with rotated with weston.

Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=99663
Signed-off-by: Olivier Fourdan <ofourdan@redhat.com>
---
 v3: Split the patch in two as there two issues. The second issue being
     the pointer events not being reported when a rotation is applied
     this is because of the RR ConstrainCursorHarder() handler that we
     should not need in Xwayland.
 v4: Do not disable the ConstrainCursorHarder() handler set by
     RRScreenInit(), simply leave the oupout size unchanged and apply 
     the rotation only when needed to compute the overall screen size...

 hw/xwayland/xwayland-output.c | 28 ++++++++++++++++------------
 1 file changed, 16 insertions(+), 12 deletions(-)

Patch hide | download patch | download mbox

diff --git a/hw/xwayland/xwayland-output.c b/hw/xwayland/xwayland-output.c
index bdf270a..7d6c7b0 100644
--- a/hw/xwayland/xwayland-output.c
+++ b/hw/xwayland/xwayland-output.c
@@ -108,14 +108,8 @@  output_handle_mode(void *data, struct wl_output *wl_output, uint32_t flags,
     if (!(flags & WL_OUTPUT_MODE_CURRENT))
         return;
 
-    if (xwl_output->rotation & (RR_Rotate_0 | RR_Rotate_180)) {
-        xwl_output->width = width;
-        xwl_output->height = height;
-    } else {
-        xwl_output->width = height;
-        xwl_output->height = width;
-    }
-
+    xwl_output->width = width;
+    xwl_output->height = height;
     xwl_output->refresh = refresh;
 }
 
@@ -123,11 +117,21 @@  static inline void
 output_get_new_size(struct xwl_output *xwl_output,
                     int *height, int *width)
 {
-    if (*width < xwl_output->x + xwl_output->width)
-        *width = xwl_output->x + xwl_output->width;
+    int output_width, output_height;
+
+    if (xwl_output->rotation & (RR_Rotate_0 | RR_Rotate_180)) {
+        output_width = xwl_output->width;
+        output_height = xwl_output->height;
+    } else {
+        output_width = xwl_output->height;
+        output_height = xwl_output->width;
+    }
+
+    if (*width < xwl_output->x + output_width)
+        *width = xwl_output->x + output_width;
 
-    if (*height < xwl_output->y + xwl_output->height)
-        *height = xwl_output->y + xwl_output->height;
+    if (*height < xwl_output->y + output_height)
+        *height = xwl_output->y + output_height;
 }
 
 /* Approximate some kind of mmpd (m.m. per dot) of the screen given the outputs

Comments

On Wed, 2017-02-08 at 09:23 +0100, Olivier Fourdan wrote:
> Previously, we would swap the width/height of the Xwayland output based
> on the output rotation, so that the overall screen size would match the
> actual rotation of each output.

This series makes sense, but I'm a little lost on which versions of
each patch are intended at this point. I assume it's this and v3 of
1/2?

(In general for short serieses I'd prefer a complete resend than trying
to sort out revisions within the thread.)

- ajax
Hey Adam,

----- Original Message -----
> On Wed, 2017-02-08 at 09:23 +0100, Olivier Fourdan wrote:
> > Previously, we would swap the width/height of the Xwayland output based
> > on the output rotation, so that the overall screen size would match the
> > actual rotation of each output.
> 
> This series makes sense, but I'm a little lost on which versions of
> each patch are intended at this point. I assume it's this and v3 of
> 1/2?
> 
> (In general for short serieses I'd prefer a complete resend than trying
> to sort out revisions within the thread.)

Yeah... It's all confusing! Sorry!

Those are the two patches:

https://patchwork.freedesktop.org/patch/137446/
https://patchwork.freedesktop.org/patch/137635/

Cheers,
Olivier
On Wed, 2017-02-08 at 13:20 -0500, Olivier Fourdan wrote:

> Yeah... It's all confusing! Sorry!
> 
> Those are the two patches:
> 
> https://patchwork.freedesktop.org/patch/137446/
> https://patchwork.freedesktop.org/patch/137635/

Merged:

remote: I: patch #137446 updated using rev afeace27d3818274b75d59375771dc964d2f56bb.
remote: I: patch #137635 updated using rev 058809c43ec578a407cf40d4c3e54a42503e3562.
remote: I: 2 patch(es) updated to state Accepted.
To ssh://git.freedesktop.org/git/xorg/xserver
   38696ea..058809c  master -> master

- ajax