Hi, Denis Oliver Kropp wrote: > On 12/13/2010 05:43 PM, André Draszik wrote: >> Not sure how to best deal with the calling of dfb_gfxcard_blit( srect, >> drect->x, drect->y, state ); if the rectangle sizes match. >> Should dfb_state_check()/_acquire() maybe do DFXL_BLIT | >> DFXL_STRETCHBLIT in the first place, instead of releasing and releasing >> and reacquiring the STRETCHBLIT state? > > SetState() is specified to receive exactly one flag, so better not use the > BLIT | STRETCHBLIT approach, but always call StretchBlit() and let the driver > do optimizations if at all possible. > > But I'm open for discussion...
OK, in that case, once the stretch blit state has been acquired, releasing it and acquiring the blit state (as in the patch) should not have any (large) overhead, as none of the flags will have changed. As Timothy said in his mail, I think that stretch blit should behave as before, optimising to do a simple blit if possible, otherwise all drivers will have to be modified - lots of code duplication... Cheers, Andre' _______________________________________________ directfb-dev mailing list directfb-dev@directfb.org http://mail.directfb.org/cgi-bin/mailman/listinfo/directfb-dev
