2008/10/18 Vitaliy Margolen wine-patches@kievinfo.com:
diff --git a/include/wine/wined3d_gl.h b/include/wine/wined3d_gl.h index 5e55a0d..b26ab78 100644 --- a/include/wine/wined3d_gl.h +++ b/include/wine/wined3d_gl.h @@ -1819,6 +1819,7 @@ typedef void (WINE_GLAPI * PGLFNGLVERTEXBLENDARB) (GLint count); #define GL_FRAMEBUFFER_INCOMPLETE_FORMATS_EXT 0x8CDA #define GL_FRAMEBUFFER_INCOMPLETE_DRAW_BUFFER_EXT 0x8CDB #define GL_FRAMEBUFFER_INCOMPLETE_READ_BUFFER_EXT 0x8CDC +#define GL_FRAMEBUFFER_INCOMPLETE_MULTISAMPLE_EXT 0x8D56 #define GL_FRAMEBUFFER_UNSUPPORTED_EXT 0x8CDD #define GL_FRAMEBUFFER_BINDING_EXT 0x8CA6 #define GL_RENDERBUFFER_BINDING_EXT 0x8CA7
That constant is part of EXT_framebuffer_multisample, not EXT_framebuffer_object.
Henri Verbeet wrote:
2008/10/18 Vitaliy Margolen wine-patches@kievinfo.com:
diff --git a/include/wine/wined3d_gl.h b/include/wine/wined3d_gl.h index 5e55a0d..b26ab78 100644 --- a/include/wine/wined3d_gl.h +++ b/include/wine/wined3d_gl.h @@ -1819,6 +1819,7 @@ typedef void (WINE_GLAPI * PGLFNGLVERTEXBLENDARB) (GLint count); #define GL_FRAMEBUFFER_INCOMPLETE_FORMATS_EXT 0x8CDA #define GL_FRAMEBUFFER_INCOMPLETE_DRAW_BUFFER_EXT 0x8CDB #define GL_FRAMEBUFFER_INCOMPLETE_READ_BUFFER_EXT 0x8CDC +#define GL_FRAMEBUFFER_INCOMPLETE_MULTISAMPLE_EXT 0x8D56 #define GL_FRAMEBUFFER_UNSUPPORTED_EXT 0x8CDD #define GL_FRAMEBUFFER_BINDING_EXT 0x8CA6 #define GL_RENDERBUFFER_BINDING_EXT 0x8CA7
That constant is part of EXT_framebuffer_multisample, not EXT_framebuffer_object.
Meaning? You don't like the place where I put it into wined3d_gl.h? I can change that to numerical order:
#define GL_RENDERBUFFER_STENCIL_SIZE_EXT 0x8D55 +#define GL_FRAMEBUFFER_INCOMPLETE_MULTISAMPLE_EXT 0x8D56 #define GL_FRAMEBUFFER_ATTACHMENT_OBJECT_TYPE_EXT 0x8CD0
Vitaliy
2008/10/18 Vitaliy Margolen wine-devel@kievinfo.com:
Meaning? You don't like the place where I put it into wined3d_gl.h? I can change that to numerical order:
You should add support for GL_EXT_framebuffer_multisample if you want to use that constant. It shouldn't be much work, the extension only introduces 3 new constants and a function.
Henri Verbeet wrote:
2008/10/18 Vitaliy Margolen wine-devel@kievinfo.com:
Meaning? You don't like the place where I put it into wined3d_gl.h? I can change that to numerical order:
You should add support for GL_EXT_framebuffer_multisample if you want to use that constant. It shouldn't be much work, the extension only introduces 3 new constants and a function.
Without that I'm getting messages like this:
fixme:d3d:context_check_fbo_status FBO status Unrecognied FBO status 0x00008d56 fixme:d3d:context_check_fbo_status Color attachment 0: (0x5e44590) WINED3DFMT_X8R8G8B8 1024x768 fixme:d3d:context_check_fbo_status Depth attachment: (0x34efb20) WINED3DFMT_D24S8 1024x768
Vitaliy.
2008/10/18 Vitaliy Margolen wine-devel@kievinfo.com:
Without that I'm getting messages like this:
fixme:d3d:context_check_fbo_status FBO status Unrecognied FBO status 0x00008d56 fixme:d3d:context_check_fbo_status Color attachment 0: (0x5e44590) WINED3DFMT_X8R8G8B8 1024x768 fixme:d3d:context_check_fbo_status Depth attachment: (0x34efb20) WINED3DFMT_D24S8 1024x768
Vitaliy.
Actually, the patch already got committed. I'll just send a patch to fix the header.
Henri Verbeet wrote:
2008/10/18 Vitaliy Margolen wine-devel@kievinfo.com:
Without that I'm getting messages like this:
fixme:d3d:context_check_fbo_status FBO status Unrecognied FBO status 0x00008d56 fixme:d3d:context_check_fbo_status Color attachment 0: (0x5e44590) WINED3DFMT_X8R8G8B8 1024x768 fixme:d3d:context_check_fbo_status Depth attachment: (0x34efb20) WINED3DFMT_D24S8 1024x768
Vitaliy.
Actually, the patch already got committed. I'll just send a patch to fix the header.
Oh, Alexandre working overtime? <g>
These messages are from STALKER Clear Sky. But I suspect there is more to it. Like more driver bugs. I'm seeing major graphics corruptions, black screen and eventual system crash.
Vitaliy.
2008/10/18 Vitaliy Margolen wine-devel@kievinfo.com:
Oh, Alexandre working overtime? <g>
These messages are from STALKER Clear Sky. But I suspect there is more to it. Like more driver bugs. I'm seeing major graphics corruptions, black screen and eventual system crash.
It probably tries to use multisampling somewhere. Unfortunately that's pretty broken at the moment. I think the main problem is that we don't pay attention to the MultiSampleType field for render targets at all, but I haven't looked into it much yet.