[snip] > With a test program I've determined that GL_LINEAR_MIPMAP_NEAREST and > GL_NEAREST_MIPMAP_LINEAR seem to be transposed. I think the Radeon > docs may be in error. > > Try GL_NEAREST_MIPMAP_LINEAR in your program and see if that has the > desired effect. I'll simply swap the two cases in the switch-statement > in the Radeon driver then. confirmed... after exchanging theese two modes everything looks ok. > Tux's texture looks a bit funny (too white) using the DRI trunk but > looks better using the 3.5 branch. :-) there is still another problem though... I have reported this earlier, but it looks like it is still present: #0 0x469b7793 in triangle () from /usr/X11R6/lib/modules/dri/radeon_dri.so #1 0x46987344 in indexed_render_tris () from /usr/X11R6/lib/modules/dri/radeon_dri.so #2 0x469874c5 in gl_render_elts () from /usr/X11R6/lib/modules/dri/radeon_dri.so #3 0x46924961 in gl_run_pipeline () from /usr/X11R6/lib/modules/dri/radeon_dri.so #4 0x468d5303 in gl_cva_force_precalc () from /usr/X11R6/lib/modules/dri/radeon_dri.so #5 0x4697fa56 in _mesa_DrawElements () from /usr/X11R6/lib/modules/dri/radeon_dri.so #6 0x464ca215 in gl_device_t::hw_flush (this=0x83a9740, type=G_TRIANGLE_LIST, vertex_count=112) at gl_drv.cpp:2034 this happens when I start using CVA. I am pushing lots of rather small vertex arrays down the GL. Without the CVA, everything looks ok (no currputed geometry)... if I reenable the cva again in the run time, the crash comes several frames after... sometimes immediately 1 frame after the switch, sometimes it takes a second to crash... I think I will have to try with mesa 3.5 and see if it is still there. Regards, Petr _______________________________________________ Dri-devel mailing list [EMAIL PROTECTED] http://lists.sourceforge.net/lists/listinfo/dri-devel
