This probably explains your problem:

:51:46 dpkg-shlibdeps: warning: symbol 
_vlib_init_function_map_stat_segment_init used by 
debian/vpp-lib/usr/lib/x86_64-linux-gnu/libvlib.so.0.0.0 found in none of the 
libraries 13:51:46 dpkg-shlibdeps: warning: symbol 
_vlib_init_function_vlibmemory_init used by 
debian/vpp-lib/usr/lib/x86_64-linux-gnu/libvlib.so.0.0.0 found in none of the 
libraries 13:51:46 dpkg-shlibdeps: warning: symbol 
_vlib_init_function_map_api_segment_init used by 
debian/vpp-lib/usr/lib/x86_64-linux-gnu/libvlib.so.0.0.0 found in none of the 
libraries 13:51:46 dpkg-shlibdeps: warning: symbol 
_vlib_init_function_vpe_api_init used by 
debian/vpp-lib/usr/lib/x86_64-linux-gnu/libvlib.so.0.0.0 found in none of the 
libraries

— 
Damjan

> On 13 Jun 2018, at 22:59, Matthew Smith <mgsm...@netgate.com> wrote:
> 
> 
> Hi Dave,
> 
> VPP itself builds fine from the top of tree. We build several C applications 
> that act as clients to the shared memory API. Linking one of those 
> applications against libvlib is when the problems occur. This worked prior to 
> merging recent upstream changes into our local repo this morning.
> 
> When the vpp executable is built/linked, the object files containing the 
> definitions of the functions map_stat_segment_init() and vpe_api_init() are 
> available and the symbols can be resolved. Those symbols cannot be resolved 
> while building our application.
> 
> -Matt
> 
> 
>> On Jun 13, 2018, at 1:21 PM, Dave Barach via Lists.Fd.Io 
>> <dbarach=cisco....@lists.fd.io> wrote:
>> 
>> +1, or “make distclean” in build-root...
>>
>> From: vpp-dev@lists.fd.io <vpp-dev@lists.fd.io> On Behalf Of Andrew 
>> Yourtchenko
>> Sent: Wednesday, June 13, 2018 2:16 PM
>> To: Jon Loeliger <j...@netgate.com>
>> Cc: vpp-dev <vpp-dev@lists.fd.io>
>> Subject: Re: [vpp-dev] Missing functions?
>>
>> Jon,
>>
>> My bet something in your tree was compiled previously, but is not rebuilt 
>> properly when you did git pull and recompile.
>>
>> (The function names seem to belong to the new stats infra that I have heard 
>> about, but didn’t get to learn about yet).
>>
>> If this theory holds water, “git clean -fdx” followed by make build should 
>> get you back in action.
>>
>> Bonus points if you can tell how to repro this from a clean checkout, to be 
>> able to understand if this is something that can be taken care of.
>>
>> I sometimes get symptoms like this when I jump between my local branches 
>> that are too far apart (say, 17.xx and 18.xx), but that happened 
>> sufficiently rarely (maybe two-three times this year) to just kick it using 
>> the method above and move on with the life. 
>>
>> --a
>> 
>> On 13 Jun 2018, at 19:23, Jon Loeliger <j...@netgate.com> wrote:
>> 
>>
>> Hi!
>>
>> With the current VPP top-of-tree here:
>>
>> commit c7d50970d4ed8a4889b4374e6a1559aef7d3dcc0
>> Author: Andrew Yourtchenko <ayour...@gmail.com>
>> Date:   Tue Jun 12 15:15:49 2018 +0200
>>
>>     acl-plugin: change the src/dst L3 info in 5tuple struct to be always 
>> contiguous with L4 data
>>
>> We seem to have some new linking issues like this:
>>
>> ..../build-root/install-vpp-native/vpp/lib64/libvlib.so: undefined reference 
>> to `_vlib_init_function_map_stat_segment_init'
>> ..../build-root/install-vpp-native/vpp/lib64/libvlib.so: undefined reference 
>> to `_vlib_init_function_vpe_api_init'
>>
>> Has anyone else seen issues like that?
>>
>> Thanks,
>> jdl
>>
> 
> 

-=-=-=-=-=-=-=-=-=-=-=-
Links: You receive all messages sent to this group.

View/Reply Online (#9603): https://lists.fd.io/g/vpp-dev/message/9603
Mute This Topic: https://lists.fd.io/mt/22083250/21656
Group Owner: vpp-dev+ow...@lists.fd.io
Unsubscribe: https://lists.fd.io/g/vpp-dev/unsub  [arch...@mail-archive.com]
-=-=-=-=-=-=-=-=-=-=-=-

Reply via email to