User-Agent: Mozilla/5.0 (Macintosh; U; Intel Mac OS X 10.5; en-US; rv:1.9b4pre) Gecko/2008020904 Minefield/3.0b4pre Build Identifier: There is a section of code nsColumnSetFrame::Reflow that reads: // If balancing, then we allow the last column to grow to unbounded // height during the first reflow. This gives us a way to estimate // what the average column height should be, because we can measure // the heights of all the columns and sum them up. But don't do this // if we have a next in flow because we don't want to suck all its // content back here and then have to push it out again! nsIFrame* nextInFlow = GetNextInFlow(); PRBool unboundedLastColumn = isBalancing && nextInFlow; Based on the comment above it, I believe the last line should be: PRBool unboundedLastColumn = isBalancing && !nextInFlow; I don't know what symptoms this would exhibit during rendering. Reproducible: Didn't try Steps to Reproduce: 1. 2. 3.
Comment on attachment 302374 [details] [diff] [review] Patch to reverse the check This simple think-o should speed up columns rendering significantly.
Checking in layout/generic/nsColumnSetFrame.cpp; /cvsroot/mozilla/layout/generic/nsColumnSetFrame.cpp,v <-- nsColumnSetFrame.cpp new revision: 3.47; previous revision: 3.46 done