On Tue, 03 Jan 2012 16:45:49 -0800 Ian Romanick <[email protected]> wrote: > > Maybe stuff this into a separate function that's a no-op in the< 3 > > case? That would clean things up a little and save an #ifdef in the > > middle of a function (always a nice thing). > > > > Or just require updated DRI2 bits to build and avoid the ifdefs > > altogether, since they tend to cause trouble anyway. > > I had thought about both of these, and ultimately we just want at later. > However, that requires a new Mesa release. I thought it was better to > leave the "obvious" clutter of the #ifdef in the code for later clean-up > than an extra function. > > What do you think? >
I think a separate function is a bit nicer in any case. Having huge functions with or without ifdefs should probably be avoided regardless. That could be an incremental patch though; there are probably a couple of other version check branches that could be pulled out for clarity. Looking forward to the next release though! -- Jesse Barnes, Intel Open Source Technology Center
signature.asc
Description: PGP signature
_______________________________________________ [email protected]: X.Org development Archives: http://lists.x.org/archives/xorg-devel Info: http://lists.x.org/mailman/listinfo/xorg-devel
