Import Thomas Helstrom's SSE memcpy code from the via X.org driver.

Add a TextureMemCpy callback, called from texstore.c when copying
texture data via the memcpy_texture() path.
Enable this code in the via driver - 100% speedup in texdown.c results.

src/mesa/drivers/common/driverfuncs.c

@@ -107,6 +107,7 @@ _mesa_init_driver_functions(struct dd_function_table *driver)
    driver->DeleteTexture = _mesa_delete_texture_object;
    driver->NewTextureImage = _mesa_new_texture_image;
    driver->FreeTexImageData = _mesa_free_texture_image_data; 
+   driver->TextureMemCpy = _mesa_memcpy; 
    driver->IsTextureResident = NULL;
    driver->PrioritizeTexture = NULL;
    driver->ActiveTexture = NULL;