Skip to content
Browse files

WinRT: workaround for a possible bug in the Win10 Store's Certificati…

…on Kit

DXGI fails to report any displays in at least one of the
"Windows App Certification Kit 10.0"'s tests for Store Apps.  This was
causing SDL's video initialization code to fail, when the suspect test
("Direct3D Feature Test") was run, as DXGI was unable to report a
display-output at adapter-index 0, output-index 0.

The workaround that is applied here attempts to detect this case, then
use a hopefully-reasonable alternative means to calculate at least one
display output.
  • Loading branch information
DavidLudwig committed Dec 6, 2015
1 parent 57f2f3b commit 781455fd8bcd110290e367ee3a8fee044fbba831
Showing with 45 additions and 0 deletions.
  1. +45 −0 src/video/winrt/SDL_winrtvideo.cpp
@@ -315,6 +315,51 @@ WINRT_AddDisplaysForAdapter (_THIS, IDXGIFactory2 * dxgiFactory2, int adapterInd

for (int outputIndex = 0; ; ++outputIndex) {
if (WINRT_AddDisplaysForOutput(_this, dxgiAdapter1, outputIndex) < 0) {
/* HACK: The Windows App Certification Kit 10.0 can fail, when
running the Store Apps' test, "Direct3D Feature Test". The
certification kit's error is:
"Application App was not running at the end of the test. It likely crashed or was terminated for having become unresponsive."
This was caused by SDL/WinRT's DXGI failing to report any
outputs. Attempts to get the 1st display-output from the
1st display-adapter can fail, with IDXGIAdapter::EnumOutputs
returning DXGI_ERROR_NOT_FOUND. This could be a bug in Windows,
the Windows App Certification Kit, or possibly in SDL/WinRT's
display detection code. Either way, try to detect when this
happens, and use a hackish means to create a reasonable-as-possible
'display mode'. -- DavidL
if (adapterIndex == 0 && outputIndex == 0) {
SDL_VideoDisplay display;
SDL_DisplayMode mode;
ApplicationView ^ appView = ApplicationView::GetForCurrentView();
SDL_zero(mode); = "DXGI Display-detection Workaround";

/* HACK: ApplicationView's VisibleBounds property, appeared, via testing, to
give a better approximation of display-size, than did CoreWindow's
Bounds property, insofar that ApplicationView::VisibleBounds seems like
it will, at least some of the time, give the full display size (during the
failing test), whereas CoreWindow will not. -- DavidL
mode.w = WINRT_DIPS_TO_PHYSICAL_PIXELS(appView->VisibleBounds.Width);
mode.h = WINRT_DIPS_TO_PHYSICAL_PIXELS(appView->VisibleBounds.Height);

mode.format = DXGI_FORMAT_B8G8R8A8_UNORM;
mode.refresh_rate = 0; /* Display mode is unknown, so just fill in zero, as specified by SDL's header files */
display.desktop_mode = mode;
display.current_mode = mode;
if ((SDL_AddDisplayMode(&display, &mode) < 0) ||
(SDL_AddVideoDisplay(&display) < 0))
return SDL_SetError("Failed to apply DXGI Display-detection workaround");


0 comments on commit 781455f

Please sign in to comment.