[Buildroot] [PATCH v2 0/2] sigrok fixes and pulseview package
bgolaszewski at baylibre.com
Mon Feb 23 11:32:47 UTC 2015
Thomas Petazzoni <thomas.petazzoni at free-electrons.com>:
> I am not really happy with this, because there are potentially plenty
> of external toolchains that support C++11. So I believe we really need
> to go ahead and add a:
> And make sure it gets properly selected depending on whether the
> toolchain has C++11 support or not. You can look at how
> BR2_TOOLCHAIN_HAS_SSP or BR2_TOOLCHAIN_HAS_THREADS are implemented for
> Moreover, you make this apparently depend on the selected C library.
> But isn't C++11 support purely a gcc problem ?
Frankly I no longer know what the problem was. I'd swear libsigrok configure
was failing after not passing some C++11 support tests until I switched to
glibc, so I thought uClibc might not implement some features C++11 requires.
But today I ran a clean build against uClibc and it compiled fine - both
libsigrokcxx and pulseview.
So as much as you're right about BR2_TOOLCHAIN_HAS_CXX11 I would like to
propose merging these two patches in which I removed (E)GLIBC dependencies.
Both patches have been rebased against buildroot-next. Tested on a BeagleBone
Black with enlightenment window manager via vncviewer.
Bartosz Golaszewski (2):
libsigrok: enable building C++ bindings
pulseview: new package
package/Config.in | 1 +
package/libsigrok/Config.in | 12 ++++++++++++
package/libsigrok/libsigrok.mk | 9 ++++++++-
package/pulseview/Config.in | 28 ++++++++++++++++++++++++++++
package/pulseview/pulseview.mk | 24 ++++++++++++++++++++++++
5 files changed, 73 insertions(+), 1 deletion(-)
create mode 100644 package/pulseview/Config.in
create mode 100644 package/pulseview/pulseview.mk
More information about the buildroot