On Mon, 6 Sep 2004 16:55:10 +0200, Felix Kühling <[EMAIL PROTECTED]> wrote: > after upgrading the DRM (it has been a while since the last cvs update) > the Savage driver stopped working. I tracked it down to the DRM refusing > to create multiple framebuffer-type mappings. If it finds an existing mapping > it returns it instead of creating a new one. However, the Savage driver > needs multiple framebuffer-type mappings. The real frame buffer is > supplemented by a so-called aperture where front/back/depth buffers can > be accessed at fixed addresses (independent of their actual locations in > the frame buffer) linearly even if the framebuffer is physically tiled. > This behaviour is controlled by a set of tiled surface registers (IIRC). > The HwMC code makes yet another framebuffer-type mapping. :-/
Does the Savage DRM driver know enough to make these extra mappings? Or is computation in user space required? Is it possible for the Savage DRMdriver to create the necessary maps using initmap like I did in the Radeon example? I was unaware of this feature of the Savage chipset when I made the changes and I don't own one to test with. Is there a representative PCI card available for this chipset? -- Jon Smirl [EMAIL PROTECTED] ------------------------------------------------------- This SF.Net email is sponsored by BEA Weblogic Workshop FREE Java Enterprise J2EE developer tools! Get your free copy of BEA WebLogic Workshop 8.1 today. http://ads.osdn.com/?ad_idP47&alloc_id808&op=click -- _______________________________________________ Dri-devel mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/dri-devel
