> Sent: Monday, January 14, 2019 at 12:56 PM > From: rol...@logikalsolutions.com > To: "Jason H" <jh...@gmx.com> > Cc: "Elvis Stansvik" <elvst...@gmail.com>, "interest@qt-project.org Interest" > <interest@qt-project.org> > Subject: Re: [Interest] 5.11.3 make install failure > > > Quoting Jason H <jh...@gmx.com>: > > > Yes, I've seen that too, even on my 8 core/16 thread build server > > which 64GB, _never_ touches swap. > > It took repeated `make`s to finally build. Never failed repeatedly > > in the same place. I think this is a GCC issue, or some combination > > of linux+gcc. Decreasing the J made it happen less, but maybe that's > > just because it's compiling less? Without a reliable failure case I > > could not identify it further. > > > > Before you leave the client for the day, I'd probably put `make` in > > a loop until the exit code is 0, then issue a `make install` > > > > Well, 5.12.0 repeatedly fails in the exact same spot. > > I can't waste more time on this. 5.9.7 successfully builds first pass > every time no matter the -j option. Everything after that seems to wig > out and die no matter what configure command I use.
I agree with you, that it should only take one invocation. In my experience, it's not Qt, it's the code. Though if what Thiago suggested, make -kjN || make -j1 is what they are using on the COIN servers, it might be why they have not seen it yet. Meanwhile, report that spot (if you haven't already). I've not tried compiling 5.12 yet, just 5.11.3. What about your experience with 5.11.3? _______________________________________________ Interest mailing list Interest@qt-project.org https://lists.qt-project.org/listinfo/interest
