Hi, I'd be willing to take a stab at this (but admittedly, If I have to do some major fiddling with the Stats infrastructure, this would be on the backburner for a week or so).
On Thu, Apr 7, 2011 at 12:20 PM, Beckmann, Brad <brad.beckm...@amd.com> wrote: > Currently Ruby does dynamically set the size of the bitsets and assigns > bitsets of different lengths to each > other. However, I don't think that is really necessary. Rather it is just > convenient because Ruby uses the machineCount function to set the bitset size > and that function is not available during construction. You can certainly > reorganize SLICC to set the max size at construction, but it will take a > little effort. I'm not sure you can, but if you can dynamically allocate a bitset then we can make this work. > If someone is willing to replace the current Ruby Set implementation with > stl::bitset, I'll volunteer to make the > necessary SLICC changes. Sounds good. I'll take a crack at changing this, post it to reviewboard, and then when you add your SLICC change we can do some initial testing. (again: I need to see what's going with the RubyStats though before spending too much time) -- - Korey _______________________________________________ m5-dev mailing list m5-dev@m5sim.org http://m5sim.org/mailman/listinfo/m5-dev