Some 12-core machines run into lots of timeouts when running 18 WebKitTestRunners (24 virtual cores * 0.75, which is a quotient for WebKit2). I'm not really sure why it's a problem with GuardMalloc while it's not a problem with regular builds. We definitely have enough RAM for all. But it seems like bringing the number further down to 12 doesn't reduce performance anyway. rdar://problem/24711525
Created attachment 276452 [details] proposed fix
Comment on attachment 276452 [details] proposed fix r=me
Comment on attachment 276452 [details] proposed fix Clearing flags on attachment: 276452 Committed r199573: <http://trac.webkit.org/changeset/199573>
All reviewed patches have been landed. Closing bug.