search for: triconnected

...hopefully faster. At the moment run time is comparable to dominance tree >> calculation. > > Both algorithms are linear so there is really no big difference in time imo. Sure. However in terms of maintainability it is nice to be able to reuse existing analysis instead of write another triconnected component analysis upfront. > I believe the biggest difference that you mention is that you can capture more > complicated regions without having to modify the CFG with the current > algorithm. Yes. >> If you want, have a look into some results I got with a pass extracting >&gt...

...is comparable to dominance tree >>>> calculation. >>> >>> Both algorithms are linear so there is really no big difference in time imo. >> Sure. However in terms of maintainability it is nice to be able to reuse >> existing analysis instead of write another triconnected component >> analysis upfront. >> >>> I believe the biggest difference that you mention is that you can capture more >>> complicated regions without having to modify the CFG with the current >>> algorithm. >> Yes. >> >>>> If you want, h...