On Tue, 27 Mar 2018 18:52:56 +0200 Mathieu Malaterre <malat@xxxxxxxxxx> wrote: MM> On Tue, Mar 27, 2018 at 6:35 PM, Vadim Zeitlin <vz-gcc@xxxxxxxxxxxx> wrote: MM> > On Tue, 27 Mar 2018 18:31:26 +0200 Mathieu Malaterre <malat@xxxxxxxxxx> wrote: MM> > MM> > MM> Well the issue is totally reproducible here. MM> > MM> > Thanks again but, just to be sure: have you retested in your (main) system MM> > after upgrading binutils or did you create a chroot for the second test? MM> > Because I see that the binutils version has changed, but I'm not sure why. MM> MM> I created a chroot exactly as you mentioned. I've updated your MM> original bug report with my (possible) findings. Thanks! In the meanwhile, I've added Buster sources to my Stretch system and installed g++-mingw-w64-i686 7.2.0-20+20.2 from Buster outside of chroot. The puzzling thing is that I still can reproduce the problem, even though my binutils-mingw-w64-i686 is still at 2.28-5+7.4+b4. But at least I know that using a chroot is _not_ a problem now. It could still be a Debian-specific problem, I guess, so it might be worth rebuilding gcc from the upstream sources and I might do it later, but for now I'll stop flailing blindly in the dark and wait if somebody can manage to actually debug this. Thanks again to everybody for your help, even if the bug is not fixed yet, it's still a relief to have a confirmation that I haven't gone crazy! VZ
Attachment:
pgpnIBu41tVpS.pgp
Description: PGP signature
