On Mon, 2013-11-18 at 22:13 -0700, Jeff Law wrote: > On 11/18/13 13:25, David Malcolm wrote: > > > > I regenerated it and am bootstrapping now. I glanced over it and > > nothing major seems to have changed; just changes due to the movement of > > code between files. Am attaching the changed patch. > Yea, there's a lot of churn right now, so this is kind-of expected. > > Thanks for re-posting and verifying it's good to go. Never hurts to be > careful. > > > > > I think so, yes, though you'll have to cast it to the appropriate > > subclass by hand; rather than the status quo of getting multiple > > screenfuls of text, you'll just get the gimple_statement_base fields: > Right. It's parsing the pages of text that I find to be a totally > pointless waste of time. It's gotten worse, mostly because I was away > for too long and I don't have the accessors in muscle-memory. > > WRT elimination of the GIMPLE_CHECKs, understood on the state of things > and why the script didn't eliminate them in the various places I pointed > out. > > This is fine. Please install.
Thanks; I've committed the combined patch series to trunk as r205034.
