[mythtv] Ticket #13060: Duplicate definition error in recorders/v4lchannel.cpp causes compile fail
R. G. Newbury
newbury at mandamus.org
Sat Jul 1 18:08:17 UTC 2017
On 27/06/17 05:09 PM, Gary Buhrmaster wrote:
> On Tue, Jun 27, 2017 at 8:20 PM, R. G. Newbury <newbury at mandamus.org> wrote:
>> I saw the ifdef in v4lchannel.h.
>> I presume that v4l is defined, so that the ifdef then includes videodev2.h
>> But the result is a duplicate inclusion.
>> AT LEAST FOR ME.... I have been building mythtv with the same scripts for
>> over 10 years now
> Well, that seems highly unlikely they are the same
> since qt5 did not exist 10 years ago (so while they
> may be similar to 10 years ago, they have clearly
> changed, and while you did not say exactly the same
> scripts, you did imply you have made no changes.)
I have had the same axe for 20 years, too. 5 new handles, and 2 new
heads, but it's the same axe.
Of course, certain things change, but not the overall structure. Main
changes have been in the inclusion/exclusion of compile switches as
>> so I don't think that things have changed.
>> The first few lines of the call to configure are (for the laptop, Fedora 23)
>> ./configure \
>> --prefix=/usr/local \
>> --compile-type=release \
>> --qmake=/usr/bin/qmake-qt5 \
>> --dvb-path=/usr/src/kernels/`uname -r`/include/uapi/ \
>> --arch=x86_64 \
>> etc. etc.
>> So the header path *should be* correct....
> I would look at the dvb-path include files carefully.
> And possibly remove --dvb-path entirely from the script
> if configure properly auto-detects the paths (which it
> generally should these days).
Deleted and did a git-clone for a brand new setup.
Took out the --dvb-path line in configure.
And it compiled cleanly.
So it was either a brain-fart from configure due to the inclusion of the
Or the change to sacrificing a Rhode Island Red *cockerel* that did the
More information about the mythtv-dev