omi script provider does not compile on ubuntu 16 and probably others

[yakman2020]

A build break. g++ -g -std=gnu++98 -fPIC -Wall -fvisibility=hidden -fno-strict-aliasing -fstack-protector-all -D_GNU_SOURCE -D_XOPEN_SOURCE=600 -DENCRYPT_DECRPYT -DGSS_USE_IOV=1 -D_BSD_SOURCE -fpermissive -DPRINT_BOOKENDS=0 -I/home/v-brucc/omi/Unix/output/include -I//home/v-brucc/omi/Unix/output/include -c -MD mi_instance.cpp -o /home/v-brucc/scriptprovider/output/obj/mi_instance.o In file included from shared_protocol.hpp:9:0, from mi_value.hpp:13, from mi_instance.hpp:8, from mi_instance.cpp:3: unique_ptr.hpp: In instantiation of ‘struct util::unique_ptr_move<char [], util::default_delete<char []> >’: shared_protocol.hpp:226:46: required from ‘int protocol::recv(std::cxx11::basic_string<char_t, traits>, socket_wrapper&) [with char_t = char; traits = std::char_traits]’ mi_value.hpp:876:35: required from ‘static int scx::MI_Array::recv(scx::MI_Array::Ptr, socket_wrapper&) [with unsigned int TYPE_ID = 29u; scx::MI_Array::Ptr = util::internal_counted_ptr<scx::MI_Array<29u>, util::default_delete<scx::MI_Array<29u> > >]’ mi_instance.cpp:552:50: required from here unique_ptr.hpp:141:45: error: creating array of ‘char []’ typename unique_ptr<T, D>::deleter_type m_deleter; ^ GNUmakefile:67: recipe for target '/home/v-brucc/scriptprovider/output/obj/mi_instance.o' failed make[1]: [/home/v-brucc/scriptprovider/output/obj/mi_instance.o] Error 1 make[1]: Leaving directory '/home/v-brucc/scriptprovider/provider' GNUmakefile:11: recipe for target 'SCRIPTPROVIDER' failed make: [SCRIPTPROVIDER] Error 2 make -C /home/v-brucc/scriptprovider/provider install make[1]: Entering directory '/home/v-brucc/scriptprovider/provider' ...compiling: mi_instance.o g++ -g -std=gnu++98 -fPIC -Wall -fvisibility=hidden -fno-strict-aliasing -fstack-protector-all -D_GNU_SOURCE -D_XOPEN_SOURCE=600 -DENCRYPT_DECRPYT -DGSS_USE_IOV=1 -D_BSD_SOURCE -DPRINT_BOOKENDS=0 -I/home/v-brucc/omi/Unix/output/include -I//home/v-brucc/omi/Unix/output/include -c -MD mi_instance.cpp -o /home/v-brucc/scriptprovider/output/obj/mi_instance.o In file included from shared_protocol.hpp:9:0, from mi_value.hpp:13, from mi_instance.hpp:8, from mi_instance.cpp:3: unique_ptr.hpp: In instantiation of ‘struct util::unique_ptr_move<char [], util::default_delete<char []> >’: shared_protocol.hpp:226:46: required from ‘int protocol::recv(std::cxx11::basic_string<char_t, traits>, socket_wrapper&) [with char_t = char; traits = std::char_traits]’ mi_value.hpp:876:35: required from ‘static int scx::MI_Array::recv(scx::MI_Array::Ptr, socket_wrapper&) [with unsigned int TYPE_ID = 29u; scx::MI_Array::Ptr = util::internal_counted_ptr<scx::MI_Array<29u>, util::default_delete<scx::MI_Array<29u> > >]’ mi_instance.cpp:552:50: required from here unique_ptr.hpp:141:45: error: creating array of ‘char []’ typename unique_ptr<T, D>::deleter_type m_deleter; ^ GNUmakefile:67: recipe for target '/home/v-brucc/scriptprovider/output/obj/mi_instance.o' failed make[1]: *** [/home/v-brucc/scriptprovider/output/obj/mi_instance.o] Error 1 make[1]: Leaving directory '/home/v-brucc/scriptprovider/provider'

[yakman2020]

@alinbalutoiu @c64cosmin @Microsoft/omi-devs

[c64cosmin]

@yakman2020 Confirmed on ubuntu 16, looking for solutions.

[alinbalutoiu]

I just checked, it's because Ubuntu 16 comes with gcc version 5 by default. You can install gcc 4 and everything would work fine on Ubuntu 16 as well. We should look anyway into fixing those issues that happen with newer compiler versions.

[JumpingYang001]

Since we provide installer packages to customer, customer needn't to build the source code themselves, so it is just affect developer, so I mark this issue as P2 which means we can fix this build issue in next release.

[JumpingYang001]

If we want to fix all 15 platforms compiling issue, that might need much more time. Our current highest priority for issues is these 2 issues:

• the create operation with remoting issue
• the python process memory leak issue.

[JumpingYang001]

Since the first version of python script provider are released, I mark this issue from P2 to P1, thanks.