Errors in compiling darktable subpackages

[Germano Massullo]

I am working on adding more features to darktable (basecurve tool and
noise tool subpackages).

In log [1] from line

BUILDSTDERR: /usr/bin/ld:
CMakeFiles/dt-curve-tool.dir/dt-curve-tool.c.o: relocation R_X86_64_32S
against symbol `spline_set' can not be used when making a shared object;
recompile with -fPIC

there are many errors, and I don't understand if they are upstream
programming errors or more likely errors in my spec file.

Build URL is
https://copr.fedorainfracloud.org/coprs/germano/darktable_2_4/build/779148/

Thank you for your time

[1]:
https://copr-be.cloud.fedoraproject.org/results/germano/darktable_2_4/fed...

[Fabio Valentini]

On Thu, Jul 19, 2018 at 9:28 PM Germano Massullo
<germano.massullo(a)gmail.com&gt; wrote:
>  I am working on adding more features to darktable (basecurve tool and noise tool
> subpackages).
>
>  In log [1] from line
>
>  BUILDSTDERR: /usr/bin/ld: CMakeFiles/dt-curve-tool.dir/dt-curve-tool.c.o: relocation
> R_X86_64_32S against symbol `spline_set' can not be used when making a shared object;
> recompile with -fPIC 
I used to get these errors too for some of my packages. It turned out
that upstream made some mistakes in their CMakeLists.txt files:
The errors were usually caused by some CMakeLists.txt file overriding
the system-provided CFLAGS with something else, instead of only
appending necessary flags.

I hope that helps.

Fabio

> there are many errors, and I don't understand if they are upstream programming
errors or more likely errors in my spec file.
>
> Build URL is
https://copr.fedorainfracloud.org/coprs/germano/darktable_2_4/build/779148/
>
> Thank you for your time
>
> [1]:
https://copr-be.cloud.fedoraproject.org/results/germano/darktable_2_4/fed...
>
> _______________________________________________
> devel mailing list -- devel(a)lists.fedoraproject.org
> To unsubscribe send an email to devel-leave(a)lists.fedoraproject.org
> Fedora Code of Conduct: https://getfedora.org/code-of-conduct.html
> List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines
> List Archives:
https://lists.fedoraproject.org/archives/list/[e-mail].o...

[Germano Massullo]

On 7/19/18 10:15 PM, Fabio Valentini wrote:
>  I used to get these errors too for some of my packages. It turned
> out
>  that upstream made some mistakes in their CMakeLists.txt files:
>  The errors were usually caused by some CMakeLists.txt file overriding
>  the system-provided CFLAGS with something else, instead of only
>  appending necessary flags. 
Thank you for your hint.
I am also looking at OpenSUSE version of darktable spec file [1] but I
find it very difficult to read with all of those macros etc.

[1]:
https://build.opensuse.org/package/view_file/home:darix:darktable:master/...

[Germano Massullo]

A user in darktable mailing list [1] said

"If you are using -specs=/usr/lib/rpm/redhat/redhat-hardened-ld at link
time, you also need to use
-specs=/usr/lib/rpm/redhat/redhat-hardened-cc1 at compile time, and as
you are compiling and linking at the same time, you need either both, or
drop the -specs=/usr/lib/rpm/redhat/redhat-hardened-ld."

so I am starting investigating about redhat-hardened usage

[1]:
https://www.mail-archive.com/[e-mail]/msg03229.html

[Germano Massullo]

A little update: upon FAS misc suggestion I started trying to comment
from line 5 to 15 of file
https://github.com/darktable-org/darktable/blob/darktable-2.4.x/tools/bas...
and then the previous error seems to be disappeared.
https://copr.fedorainfracloud.org/coprs/germano/darktable_2_4/build/779751/

Now I am investigating about how to fix the other CMakeLists errors