https://gcc.gnu.org/bugzilla/show_bug.cgi?id=125359
Iain Sandoe <iains at gcc dot gnu.org> changed:
What |Removed |Added
----------------------------------------------------------------------------
Assignee|iains at gcc dot gnu.org |unassigned at gcc dot
gnu.org
Status|WAITING |NEW
--- Comment #13 from Iain Sandoe <iains at gcc dot gnu.org> ---
(In reply to Zhongteng Gui from comment #12)
> (In reply to Iain Sandoe from comment #11)
> > Created attachment 65942 [details]
> > Patch to test
> >
> > This works of Darwin - and, I think, in principle should for other platforms
> > - but I am not that familiar with mingw.
> >
> > Please test...
> >
> > ... note that the classification of this bug should probably be either c++
> > or ipa
>
> Unfortunately, this doesn't fix this problem for MinGW. Tested at
> https://github.com/dragon-archer/gcc-for-windows/pull/14, and I can still
> reproduce the problem with the artifacts.
I guess that -fno-declone-ctor-dtor does work around it tho?
If that is the case, then you probably need to debug what combination of
COMDAT, Weak and Hidden properties is present at the point the decision is
made, and make sure to the set the no_external_definition flag accordingly.
Unfortunately, I do not have resources to work on the mingw problem right now.