That won't quite work (size would go to 0 when compacting). However, maybe rereading size will do. Let me commit it as is now to commit closures CL. And then I'll experiment with it more.
Thanks a lot for review in such inconvenient time! yours, anton. On Sat, Jan 15, 2011 at 12:22 AM, <[email protected]> wrote: > Can we avoid breaking the (old nice) invariant by setting the cache size > only > after calling the factory function? (The generated code in the classic > backend > seems to do exactly this so it seems like the only place to fix is > Runtime_GetFromCache.) Otherwise, LGTM. > > > http://codereview.chromium.org/6309004/diff/1/src/objects-debug.cc > File src/objects-debug.cc (right): > > http://codereview.chromium.org/6309004/diff/1/src/objects-debug.cc#newcode666 > src/objects-debug.cc:666: int size = > Smi::cast(get(kCacheSizeIndex))->value(); > Probably needs USE(size) now. > > http://codereview.chromium.org/6309004/ > -- v8-dev mailing list [email protected] http://groups.google.com/group/v8-dev
