On Thu, Dec 15, 2016 at 01:03:41PM +0100, Hendrik Leppkes wrote:
> On Thu, Dec 15, 2016 at 12:57 PM, Diego Biurrun <[email protected]> wrote:
> > On Thu, Dec 15, 2016 at 08:43:16AM +0100, Steve Lhomme wrote:
> >> On Wed, Dec 14, 2016 at 5:12 PM, Diego Biurrun <[email protected]> wrote:
> >> > On Wed, Dec 14, 2016 at 04:46:54PM +0100, Steve Lhomme wrote:
> >> >> From: Steve Lhomme <[email protected]>
> >> >>
> >> >> Without any optimization flags, MSVC does no dead code elimination 
> >> >> (DCE) at
> >> >> all, even for the most trivial cases. DCE is a prerequisite for 
> >> >> building libav
> >> >> correctly, otherwise there are undefined references to functions for 
> >> >> other
> >> >> architectures and disabled components.
> >> >>
> >> >> -Os -Og is the minimal optimization flags for MSVC that does include 
> >> >> DCE. It
> >> >
> >> > -Os -Og are .. that do ..
> >>
> >> I would say "the setting '-Os -Og' is". One could say "the flags '-Os -Og' 
> >> are".
> >
> > Now that you added the quotes it is indeed right ;-p
> >
> >> >> warns that -Og will be removed but it doesn't work without. -O1 
> >> >> includes these
> >> >> flags and some other ones not necessary to link properly.
> >> >
> >> > What does "warn" mean? And is this combination really preferable to 
> >> > plain -O1?
> >>
> >> Every C file compiled warns that Og is deprecated and may be removed
> >> soon. I prefer it to -O1 because it involves less optimizations, which
> >> is the goal of --disable-optimizations. O1 sets "-Og -Os -Oy -Ob2 -Gs
> >> -GF -Gy".
> >
> > Adding a warning for every file is not something we should do lightly.
> > Warnings are valuable and they become complete noise when the silly
> > warnings drown out the real ones.
> >
> > If Og is deprecated, what is its replacement?
> 
> Microsoft is discontinuing the fine-grained optimization options and
> recommends to just stick to O1 and O2.

Then I suggest we follow that and just use -O1 here, even though a smaller
set is normally the more elegant solution.

Diego
_______________________________________________
libav-devel mailing list
[email protected]
https://lists.libav.org/mailman/listinfo/libav-devel

Reply via email to