[llvm-commits] [llvm-gcc-4.2] r51608 - /llvm-gcc-4.2/trunk/gcc/llvm-backend.cpp

[Duncan Sands]

Hi Evan,

> We are running the inliner at -O2, no? We are no longer checking the  
> optimizer level. The issue here is llvm-gcc still run the inliner even  
> with -fno-inline-function.

I get it now - sorry for the confusion.  Your patch seems fine :)

> Can we rely on the gcc inliner to handle functions that are marked  
> inline until we add the inline hint to llvm? Could that be the reason   
> why WebKit regressed when the gcc inliner is disabled completely?

Rather than adding an inline hint, there could be a per-function
threshold.  I'm not saying that I think this is a good idea, I'm
just tossing the thought out for discussion.

If we turn down the gcc inliner to the minimum (as I did before)
then it won't inline functions marked inline, only those marked
always_inline.

Ciao,

Duncan.