[Buildroot] [PATCH 1/2] cc-tool: drop BR2_PACKAGE_BOOST_SIGNALS select

Fabrice Fontaine fontaine.fabrice at gmail.com
Sun Jan 13 13:59:36 UTC 2019


Dear Thomas,

Le dim. 13 janv. 2019 à 14:27, Thomas Petazzoni
<thomas.petazzoni at bootlin.com> a écrit :
>
> Hello,
>
> On Sat, 12 Jan 2019 21:50:25 +0100, Fabrice Fontaine wrote:
> > cc-tool depends on signals2 not signals
> >
> > Signed-off-by: Fabrice Fontaine <fontaine.fabrice at gmail.com>
>
> This commit message is a bit terse:
>
>  - Where do you see that cc-tool depends on signals2 not signals?
>
>  - Why it doesn't need to select some BOOST_SIGNALS2 option ?
>
> in cc-tool/boost.m4, I see:
>
> # Look for Boost.Signals.  For the documentation of PREFERRED-RT-OPT, see the
> # documentation of BOOST_FIND_LIB above.
> BOOST_DEFUN([Signals],
> [BOOST_FIND_LIB([signals], [$1],
>                 [boost/signal.hpp],
>                 [boost::signal<void ()> s;])
> ])# BOOST_SIGNALS
>
> # BOOST_SIGNALS2()
> # -----------------
> # Look for Boost.Signals2
> BOOST_DEFUN([Signals2],
> [BOOST_FIND_HEADER([boost/signals2.hpp])])
>
> So it's not that clear cut that it uses *only* signals2. A bit of
> clarification would be nice. Though indeed, grepping for "signals"
> through the code only shows:
>
> src/data/progress_watcher.h:#include <boost/signals2.hpp>
> src/data/progress_watcher.h:    typedef boost::signals2::signal<void (uint_t done_size, uint_t full_size)> OnProgress;
>
> So is the check for "Signals" in m4/boost.m4 some dead code ?
Yes, this is dead code and we don't need to select signals2 as this is
a header-only boost library. I'll send a v2 with those clarifications.
>
> Thanks,
>
> Thomas
> --
> Thomas Petazzoni, CTO, Bootlin
> Embedded Linux and Kernel engineering
> https://bootlin.com
Best Regards,

Fabrice


More information about the buildroot mailing list