Gating the code doesn't actually hide the perf problem. Because the perf problem occurs in getPresumedLoc which calculates line and column number. The debug code was calling getPresumedLoc twice once to only get the line and once to only get the column, but this results in line and column actually being calculated twice.
Perhaps the debug code should be refactored to only call getPresumedLoc once? On Thu, Oct 18, 2012 at 3:05 PM, Eric Christopher <[email protected]>wrote: > > > > Hrm, I skimmed over the "no known current users". I hate such > configuration flags, but this one makes sense. > > > > Agreed. I just thought it was a waste to remove the code completely > since someone may want the information. > > -eric > _______________________________________________ > cfe-commits mailing list > [email protected] > http://lists.cs.uiuc.edu/mailman/listinfo/cfe-commits > -- ~Craig
_______________________________________________ cfe-commits mailing list [email protected] http://lists.cs.uiuc.edu/mailman/listinfo/cfe-commits
