On Thu, Dec 12, 2013 at 05:23:08PM +0000, Chris Wilson wrote: > On Thu, Dec 12, 2013 at 05:05:18PM +0000, Damien Lespiau wrote: > > On Thu, Dec 12, 2013 at 04:58:21PM +0000, Chris Wilson wrote: > > > On Thu, Dec 12, 2013 at 02:36:37PM +0000, Damien Lespiau wrote: > > > > If we make sure that all the dev_priv->info usages are wrapped by > > > > INTEL_INFO(), we can easily modify the ->info field to be structure and > > > > not a pointer while keeping the const protection in the INTEL_INFO() > > > > macro. > > > > > > Yuck. > > > > Would Jani's suggestion to transmogrify INTEL_INFO() into a function > > make you happier? Or is it the back and forth from *dev_priv to *dev > > to dev_priv->info that is of utmost disgust? > > Ultimately, I'd like to see the pointer dance die. But in the meantime, > I'd like to see the macro die, dev_priv->info is far less scary than > INTEL_INFO(dev_priv->dev)->info, even to_i915(dev)->info is more > pleasant to read. I guess another option would be to make keep the dev_priv.info as a const struct, and just explicitly cast the const away when modifying the struct. -- Ville Syrjälä Intel OTC _______________________________________________ Intel-gfx mailing list Intel-gfx@xxxxxxxxxxxxxxxxxxxxx http://lists.freedesktop.org/mailman/listinfo/intel-gfx
