No, I think the issue was that (a) it is a check on every marked object in release builds, and (b) it was not correct because there is more than one call site where we mark objects and only one of them was testing for global objects.
On Thu, Oct 30, 2008 at 5:36 PM, <[EMAIL PROTECTED]> wrote: > > Global object counter is pretty useful for debugging memory leaks. > Does it cause trouble for inlining? > > On 2008/10/30 16:04:25, Patrick Johnson wrote: > > LGTM > > > > I only reviewed the change to remove the global objects counter. > We're not > > using it either as far as I know. > > > > http://codereview.chromium.org/8910 > > > > --~--~---------~--~----~------------~-------~--~----~ v8-dev mailing list [email protected] http://groups.google.com/group/v8-dev -~----------~----~----~----~------~----~------~--~---
