[Bug middle-end/78977] [7 Regression] g++7 snprintf() of double produces wrong code with -O3

[jakub at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=78977

--- Comment #7 from Jakub Jelinek <jakub at gcc dot gnu.org> ---
(In reply to Hannes Hauswedell from comment #6)
> (In reply to Martin Sebor from comment #5)
> > (In reply to Hannes Hauswedell from comment #3)
> > > 
> > > Great, let me know if it's merged, then I will try a newer snapshot!
> > 
> > The patch has been committed (r244037).  Please give it a try and let us
> > know how it goes.
> 
> Thanks! I have checked out master via git and verified that it includes the
> fix. I then tried to tie this into the FreeBSD-port (to make use of the
> local patches and build flags), but unfortunately it still doesn't build.
> I normally don't build gcc myself so it would likely take me considerable
> time to track these build problems and fix them. The next weekly snapshot
> should be built on 20170108 and then land in the FreeBSD-ports very soon
> after.
> Can we wait for this, or are we in a hurry -- for when is the GCC7 release
> scheduled?

Could you in the mean time attach the preprocessed source + options (and
confirm whether it is appendNumber in the preprocessed source or something
else)?
Then we can easily check if the patch actually changed anything or not.