UHD build apparently failed

As I am installing GNU Radio from this link:
http://blogs.bu.edu/mhirsch/2012/07/installing-gnu-radio-in-ubuntu-12-04-x64/comment-page-1/
(which installs GNU Radio & UHD )
I got the following error , what can i do ? Thanks :slight_smile:

[ 26%] Building CXX object
lib/CMakeFiles/uhd.dir/transport/nirio/rpc/rpc_client.cpp.o
/home/dell/GNUradio/uhd/host/lib/transport/nirio/rpc/rpc_client.cpp: In
constructor ‘uhd::usrprio_rpc::rpc_client::rpc_client(const string&,
const
string&, uint32_t, uint32_t)’:
/home/dell/GNUradio/uhd/host/lib/transport/nirio/rpc/rpc_client.cpp:51:9:
error: ‘connect’ is not a member of ‘boost::asio’
/home/dell/GNUradio/uhd/host/lib/transport/nirio/rpc/rpc_client.cpp:51:9:
note: suggested alternatives:
/usr/include/x86_64-linux-gnu/sys/socket.h:129:12: note: ‘connect’
/usr/include/boost/asio/detail/impl/socket_ops.ipp:383:5: note:
‘boost::asio::detail::socket_ops::connect’
make[2]: ***
[lib/CMakeFiles/uhd.dir/transport/nirio/rpc/rpc_client.cpp.o]
Error 1
make[1]: *** [lib/CMakeFiles/uhd.dir/all] Error 2
make: *** [all] Error 2
UHD build apparently failed
Exiting UHD build
Send success/fail info to sbrac.org?

On 03/21/2014 09:40 PM, Sara C. wrote:

/home/dell/GNUradio/uhd/host/lib/transport/nirio/rpc/rpc_client.cpp:51:9:
UHD build apparently failed
Exiting UHD build
Send success/fail info to sbrac.org http://sbrac.org?


Discuss-gnuradio mailing list
[email protected]
Discuss-gnuradio Info Page
I see the same problem on Fedora 14, and even when I specify:

-DENABLE_X300=OFF

In the cmake, I still see this build failure due to nirio requiring a
newer boost.

Ben/Michael have you a comment?

Hi Marcus, Sara

I think I fixed this somewhen back in december, when RHEL6 support
broke.
Which branch (tag) are you building off? UHD 3.7.0 (master) should have
those changes.

Cheers,

Moritz

On 03/23/2014 10:23 AM, Moritz Fischer wrote:

Hi Marcus, Sara

I think I fixed this somewhen back in december, when RHEL6 support broke.
Which branch (tag) are you building off? UHD 3.7.0 (master) should
have those changes.

Cheers,

Moritz

Here’s a bit of git log of what I’m building:

commit a8caec5f93976c081d488118f53a72dca49efdf6
Author: Ashish C. [email protected]
Date: Wed Feb 19 19:22:52 2014 -0800

 b200: Fixed bug in rx_dsp_core_3000 that would assume 3 halfbands

and X300 s

commit 7fef199d194c9a63b3312845979fa353f90f4d23
Author: Nicholas C. [email protected]
Date: Fri Feb 14 06:49:10 2014 -0800

 Updated images URL to UHD 3.7.0 images

commit ebf1f1780d692c25e445b4df1c19d01ba43d6a94
Author: Ben H. [email protected]
Date: Thu Feb 13 17:30:24 2014 -0800

 Putting the ICMP destination unreachable fix in X300 firmware.

commit ff1546f8137f7f92bb250f685561b0c34cc0e053
Author: Ben H. [email protected]
Date: Fri Feb 14 12:05:07 2014 -0800

 Pushing the bulk of UHD-3.7.0 code.

Hi Marcus,

On Sun, Mar 23, 2014 at 3:24 PM, Marcus D. Leech [email protected]
wrote:

Here’s a bit of git log of what I’m building:

commit a8caec5f93976c081d488118f53a72dca49efdf6
Author: Ashish C. [email protected] [email protected]
Date: Wed Feb 19 19:22:52 2014 -0800

b200: Fixed bug in rx_dsp_core_3000 that would assume 3 halfbands and

X300 s

Crap, that’s 7 commits down from the fix in our internal master. I’ll
see
if I can get stuff pushed tomorrow.

In the meantime you can try the attached patch.

Happy hacking,

Moritz

boost::container is quite new (introduced in 1.48 afaik); I’m pulling
master right now, but boost::container::vector should be STL-compatible.
replace the incluse by #include and try :wink:

On 03/23/2014 12:08 PM, Moritz Fischer wrote:

Moritz
It got further, the nirio stuff made it past, but, then:

/home/mleech/uhd/host/include/uhd/transport/nirio/nirio_fifo.ipp:377:24:
warning: expected [error|warning|ignored] after ‘#pragma GCC diagnostic’
[ 44%] Building CXX object lib/CMakeFiles/uhd.dir/usrp/dboard_base.cpp.o
[ 44%] Building CXX object
lib/CMakeFiles/uhd.dir/usrp/dboard_eeprom.cpp.o
[ 44%] Building CXX object lib/CMakeFiles/uhd.dir/usrp/dboard_id.cpp.o
[ 45%] Building CXX object
lib/CMakeFiles/uhd.dir/usrp/dboard_iface.cpp.o
[ 45%] Building CXX object
lib/CMakeFiles/uhd.dir/usrp/dboard_manager.cpp.o
[ 46%] Building CXX object lib/CMakeFiles/uhd.dir/usrp/gps_ctrl.cpp.o
/home/mleech/uhd/host/lib/usrp/gps_ctrl.cpp:33:38: fatal error:
boost/container/vector.hpp: No such file or directory
compilation terminated.
make[2]: *** [lib/CMakeFiles/uhd.dir/usrp/gps_ctrl.cpp.o] Error 1
make[1]: *** [lib/CMakeFiles/uhd.dir/all] Error 2
make: *** [all] Error 2

Hi Marcuses,

on my machine it builds without the include. Can you confirm that?

Cheers,

Moritz

On 03/23/2014 02:42 PM, Marcus M. wrote:

boost::container is quite new (introduced in 1.48 afaik); I’m pulling
master right now, but boost::container::vector should be STL-compatible.
replace the incluse by #include and try :wink:
That got me past gps_ctrl.cpp, but now:

/home/mleech/uhd/host/lib/usrp/common/adf435x_common.cpp: In function
‘adf435x_tuning_settings tune_adf435x_synth(double, double, const
adf435x_tuning_constraints&, double&)’:
/home/mleech/uhd/host/lib/usrp/common/adf435x_common.cpp:81:29: error:
‘floor’ is not a member of ‘std’
/home/mleech/uhd/host/lib/usrp/common/adf435x_common.cpp:128:66: error:
‘ceil’ is not a member of ‘std’
make[2]: *** [lib/CMakeFiles/uhd.dir/usrp/common/adf435x_common.cpp.o]
Error 1
make[1]: *** [lib/CMakeFiles/uhd.dir/all] Error 2
make: *** [all] Error 2

That stuff used to work just fine with my older (F14) system here.
Frustrating.

Confirmed, fc19, boost 1.53.0

On 03/23/2014 07:08 PM, Marcus M. wrote:

This calls for #include ; I wonder how that is working on my f19
Yup, that took care of it.

Then there’s the b200_iface.cpp problem with libusb on my system not
having libusb_error_name, so I had to copy the macro from
transport/zero_copy that’s conditional on HAVE_LIBUSB_ERROR_NAME,
and it all compiled just fine.

This calls for #include ; I wonder how that is working on my f19

Hi all,

thanks for testing. Parts of that is already internally fixed on our
branches for the rest I created patches.
I expect this to be pushed to the public master in the next days.

Thanks for hanging in there,

Moritz