On 06/08/2017 06:01 PM, Emil Velikov wrote:
On 8 June 2017 at 11:10, Tapani Pälli <[email protected]> wrote:
Specification states that in case of error, value should not be
written, patch changes buffer age queries to return -1 in case of
error so that we can skip changing the value.
In addition, small change to droid_query_buffer_age to return 0
in case buffer does not have a back buffer available.
Fixes:
dEQP-EGL.functional.negative_partial_update.not_postable_surface
Great find Tapani. Please don't forget the stable tag.
Cc: [email protected]
Signed-off-by: Tapani Pälli <[email protected]>
---
src/egl/drivers/dri2/platform_android.c | 4 ++--
src/egl/drivers/dri2/platform_drm.c | 2 +-
src/egl/main/eglsurface.c | 6 +++++-
3 files changed, 8 insertions(+), 4 deletions(-)
diff --git a/src/egl/drivers/dri2/platform_android.c
b/src/egl/drivers/dri2/platform_android.c
index 48ecb9f..4c97935 100644
--- a/src/egl/drivers/dri2/platform_android.c
+++ b/src/egl/drivers/dri2/platform_android.c
@@ -614,10 +614,10 @@ droid_query_buffer_age(_EGLDriver *drv,
if (update_buffers(dri2_surf) < 0) {
Completely unrelated:
This combined with the following discrepancy might be the reason why
Android is special wrt get_back_bo()
On wayland/drm we have
- getBuffers{WithFormat,} (dri2/image respectively) -> misc -> get_back_bo
- swap_buffers -> age buffers -> get_back_bo -> pick unlocked (and
oldest for gbm) buffer -> create image -> init. age
- query_buffer_age -> get_back_bo -> return age
while on Android
- getBuffers{WithFormat,} (dri2/image respectively) -> misc ->
_missing_ get_back_bo from update_buffers()
- swap_buffers -> age buffers -> if we have back bo -> init. age
- query_buffer_age -> get_back_bo -> return age
Hence, things won't fly on Android if a query_buffer_age is not called
before swapbuffers.
_eglError(EGL_BAD_ALLOC, "droid_query_buffer_age");
- return 0;
+ return -1;
}
- return dri2_surf->back->age;
+ return dri2_surf->back ? dri2_surf->back->age : 0;
}
static EGLBoolean
diff --git a/src/egl/drivers/dri2/platform_drm.c
b/src/egl/drivers/dri2/platform_drm.c
index 2f04589..36c89fc 100644
--- a/src/egl/drivers/dri2/platform_drm.c
+++ b/src/egl/drivers/dri2/platform_drm.c
@@ -464,7 +464,7 @@ dri2_drm_query_buffer_age(_EGLDriver *drv,
if (get_back_bo(dri2_surf) < 0) {
_eglError(EGL_BAD_ALLOC, "dri2_query_buffer_age");
- return 0;
+ return -1;
dri2_wl_query_buffer_age() could use an identical fix.
Please either squash a wayland fix in (adding a note that x11 needs
similar fix) or split them up.
Oopsie yes, I'll add dri2_wl_query_buffer_age in the mix too, x11 will
be a separate thing, I took a look at it and it is .. different.
In either case
Reviewed-by: Emil Velikov <[email protected]>
Warning slight diversion incoming :-\
On platform_x11.c side similar fix is needed, but let's keep that separate.
Mostly since we need to update loader_dri3_query_buffer_age() as well
as libGLX :-\
... Not to mention there's some fun points around glXQueryDrawable.
Agreed, will take a look separately.
Spec states that the function returns "void", same as the XML file and
headers (Mesa, Khronos and Nvidia shipped ones). Yet nearly all
documentation pages [1] [2] [3] claim that "int" is returned.
An older man page [4] seems to be more accurate, yet the only correct
instance that I can find.
Might be worth checking with other vendors/Khronos and following from there.
Checking python-opengl - seems to be doing the right thing. So we
should be fine - barring that the manuals that need fixing ;-)
-Emil
[1]
https://www.khronos.org/registry/OpenGL-Refpages/gl2.1/xhtml/glXQueryDrawable.xml
[2]
http://pyopengl.sourceforge.net/documentation/manual-3.0/glXQueryDrawable.html
[3] http://code.nabla.net/doc/OpenGL/api/OpenGL/man/glXQueryDrawable.html
[4] http://nixdoc.net/man-pages/IRIX/man3/glxquerydrawable.3.html
_______________________________________________
mesa-dev mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/mesa-dev
