On Wed, Dec 18, 2013 at 9:45 AM, Alexander Monakov <amona...@ispras.ru>wrote:
> > > On Tue, 17 Dec 2013, José Fonseca wrote: > > > The headers can be fixed. > > > > But I doubt that's all it takes. Neither GL or EGL fully abstract the > > windowing system API such as XLIB. If XLIB is not used for creating > > windows, something else (GBM?) must be used, and apitrace would need glue > > code (retrace/glws_egl_gbm*) to use it. > > > > Another alternative would be to use waffle. > > True; but (to clarify for Chris) that would only be needed to build retrace > (replay) code. This is not a hard requriement. On Linux, when libX11.so > is > not available, apitrace happily builds the cli and trace tools without > retracers. > > I see. In that case, could someone provide a diff of the relevent header as it is included in this system vs khronos headers? If the differences relatively to khronos headers are small, we may consider patching the khronos headers whenever they are refreshed (as done with other headers). If that proves unpractical, then we need to add more brains to the CMake files to prefer system headers when available/told... Jose
_______________________________________________ apitrace mailing list apitrace@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/apitrace
