Re: Cost model for indirect call speculation

2013-08-12 Thread David Malcolm
On Mon, 2013-08-12 at 01:11 +0200, Jan Hubicka wrote: > Hi, > this patch adds simple cost model into indirect call speculation. [...] > Index: cgraph.c [...] > *** cgraph_resolve_speculation (struct cgrap > ! fprintf (dump_file, "Speculative indirect call %s/%i => %s/%i has

Re: Cost model for indirect call speculation

2013-08-12 Thread Andi Kleen
Xinliang David Li writes: > I like the approach in general -- in the past, indirect call promotion > and function inlining heuristics are disconnected -- which can lead to > either missing promotions or useless ones. This approach solves the > problem. > > On Sun, Aug 11, 2013 at 4:11 PM, Jan Hu

Re: Cost model for indirect call speculation

2013-08-11 Thread Xinliang David Li
I like the approach in general -- in the past, indirect call promotion and function inlining heuristics are disconnected -- which can lead to either missing promotions or useless ones. This approach solves the problem. On Sun, Aug 11, 2013 at 4:11 PM, Jan Hubicka wrote: > Hi, > this patch adds s