Is your workaround for building inside scratchbox to set the CROSS_COMPILE
variable to empty, i.e. pass -device-option CROSS_COMPILE= ?
Unfortunately that had still caused the error when I tried, so I used
CROSS_COMPILE=/usr/bin/. Very nasty, I agree. :-)
It also shows that we would need a
On Tue, Oct 2, 2012 at 9:04 AM, Laszlo Papp lp...@kde.org wrote:
Is your workaround for building inside scratchbox to set the CROSS_COMPILE
variable to empty, i.e. pass -device-option CROSS_COMPILE= ?
Unfortunately that had still caused the error when I tried, so I used
On Thursday, September 27, 2012 07:46:57 PM Laszlo Papp wrote:
Can you do what it says?
I have spent a couple of hours with figuring this problem out, and it
seems the following change in qmake has made the build policy about
errors stricter:
https://codereview.qt-project.org/#change,35183
git bisect --run should help. It works unattended if you have a test case
that
you check with a shell script, and spits out the commit that introduced the
regression.
Yes, I know bisect. It is unfortunately still too much work for my current
schedule, but thank you anyway.
I will have to
On quinta-feira, 27 de setembro de 2012 11.51.16, Laszlo Papp wrote:
Hi,
Two weeks old tarball finally worked from master, but it is now broken
again when processing the arm.pro (?). See the details here:
https://build.pub.meego.com/package/live_build_log?arch=armv7elpackage=qt5-
Can you do what it says?
I have spent a couple of hours with figuring this problem out, and it
seems the following change in qmake has made the build policy about
errors stricter:
https://codereview.qt-project.org/#change,35183
I think this change is actually a good idea, so I do not blame
