A agree with Slava here. Removing 'inline' like this is not the right thing. We should find places that matter (reduces binary size and has either no negative or positive impact on performance). Also, we need to check compilers on other platforms than linux as well.
Thanks, -- Mads On Thu, Jul 22, 2010 at 1:54 PM, <[email protected]> wrote: > I am against such removal. > > Inline is both hint for a compiler and a programmer. > > If compiler is currently clever enough to inline function without our hint > that > does not matter it will continue to do so. > > When programmer included inline here (or in some other place) he probably > wanted > to say "hey! I do not want call overhead here. It's a small frequently > called > function". We should preserve such hints because they make code reading > better. > > Of course removing inline from large functions might be a good idea (it > might > accidentaly save us some code space/stack space) but I think you should not > just > remove inline everywhere (especially not from small functions). > > If you are in doubt you can also do a svn blame and chat with the person > who > wrote this function or who is touching this component now. > > > http://codereview.chromium.org/3012018/show > > -- > v8-dev mailing list > [email protected] > http://groups.google.com/group/v8-dev > -- v8-dev mailing list [email protected] http://groups.google.com/group/v8-dev
