patch ping: Dump the "degree of overlap" to compare static profile with instrumentation profile

2011-10-07 Thread Dehao Chen
http://gcc.gnu.org/ml/gcc-patches/2011-09/msg01293.html

patch ping: Dump the "degree of overlap" to compare static profile with instrumentation profile

2011-10-18 Thread Dehao Chen
http://gcc.gnu.org/ml/gcc-patches/2011-09/msg01293.html Thanks! Dehao

Re: patch ping: Dump the "degree of overlap" to compare static profile with instrumentation profile

2011-10-07 Thread Xinliang David Li
Yes, this is useful for static prediction tuning. David On Fri, Oct 7, 2011 at 6:40 PM, Dehao Chen wrote: > > http://gcc.gnu.org/ml/gcc-patches/2011-09/msg01293.html

Re: patch ping: Dump the "degree of overlap" to compare static profile with instrumentation profile

2011-10-19 Thread Jan Hubicka
> http://gcc.gnu.org/ml/gcc-patches/2011-09/msg01293.html Hi, this seems fine for mainline. I am not 100% sure it is best choice of metric (i.e. mostly one is interested if hot blocks identified by static profile cover hot blocks of real profile), but we could get some experience with it. Thanks