I stumbled upon this testing problem, where the tested colour differs between wine and windows.
Tested on ubuntu 17.10 and all windows versions on testbot.
Signed-off-by: Pablo Martin pmart-wine@riseup.net --- dlls/d3d11/tests/d3d11.c | 1 + 1 file changed, 1 insertion(+)
diff --git a/dlls/d3d11/tests/d3d11.c b/dlls/d3d11/tests/d3d11.c index 63f964a461..307a966ac3 100644 --- a/dlls/d3d11/tests/d3d11.c +++ b/dlls/d3d11/tests/d3d11.c @@ -11581,6 +11581,7 @@ static void test_copy_subresource_region(void) ID3D11DeviceContext_PSSetShader(context, ps, NULL, 0);
ID3D11DeviceContext_ClearRenderTargetView(context, test_context.backbuffer_rtv, red); + todo_wine check_texture_color(test_context.backbuffer, 0x800000ff, 0);
set_box(&box, 0, 0, 0, 2, 2, 1); ID3D11DeviceContext_CopySubresourceRegion(context, (ID3D11Resource *)dst_texture, 0,
On Tue, May 15, 2018 at 7:09 PM, Pablo Martin pmart-wine@riseup.net wrote:
I stumbled upon this testing problem, where the tested colour differs between wine and windows.
Tested on ubuntu 17.10 and all windows versions on testbot.
Signed-off-by: Pablo Martin pmart-wine@riseup.net
dlls/d3d11/tests/d3d11.c | 1 + 1 file changed, 1 insertion(+)
diff --git a/dlls/d3d11/tests/d3d11.c b/dlls/d3d11/tests/d3d11.c index 63f964a461..307a966ac3 100644 --- a/dlls/d3d11/tests/d3d11.c +++ b/dlls/d3d11/tests/d3d11.c @@ -11581,6 +11581,7 @@ static void test_copy_subresource_region(void) ID3D11DeviceContext_PSSetShader(context, ps, NULL, 0);
ID3D11DeviceContext_ClearRenderTargetView(context, test_context.backbuffer_rtv, red);
- todo_wine check_texture_color(test_context.backbuffer, 0x800000ff, 0);
It doesn't seem good idea to require the exact result. 0x7f is close enough to 0x80.
set_box(&box, 0, 0, 0, 2, 2, 1); ID3D11DeviceContext_CopySubresourceRegion(context, (ID3D11Resource *)dst_texture, 0,-- 2.14.1
On 2018-05-15 23:50, Józef Kucia wrote:
On Tue, May 15, 2018 at 7:09 PM, Pablo Martin pmart-wine@riseup.net wrote:
I stumbled upon this testing problem, where the tested colour differs between wine and windows.
- todo_wine check_texture_color(test_context.backbuffer, 0x800000ff, 0);
It doesn't seem good idea to require the exact result. 0x7f is close enough to 0x80.
I see, ok! I think the patch can safely be ignored then since there are similar tests in the file.
Thank you, and sorry for the noob mistake. I'll take that into account for other patches.
On Wed, May 16, 2018 at 12:37 AM, Pablo Martin pmart-wine@riseup.net wrote:
On 2018-05-15 23:50, Józef Kucia wrote:
On Tue, May 15, 2018 at 7:09 PM, Pablo Martin pmart-wine@riseup.net wrote:
I stumbled upon this testing problem, where the tested colour differs between wine and windows.
- todo_wine check_texture_color(test_context.backbuffer, 0x800000ff, 0);
It doesn't seem good idea to require the exact result. 0x7f is close enough to 0x80.
I see, ok! I think the patch can safely be ignored then since there are similar tests in the file.
Thank you, and sorry for the noob mistake. I'll take that into account for other patches.
Generally, if a test doesn't return the exact value on all platforms you want to use tolerance greater than 0, e.g.
check_texture_color(test_context.backbuffer, 0x800000ff, 1);
or
check_texture_color(test_context.backbuffer, 0x800000ff, 2);
-
Józef Kucia -
Pablo Martin