armnn: ArmNN 21.05 fails to build with gcc11

ArmNN 21.05 fails to build with gcc11 becauseof:

[   89s] [  2%] Building CXX object CMakeFiles/armnnUtils.dir/src/armnnUtils/FloatingPointConverter.cpp.o
[   89s] /usr/bin/c++ -DARMCOMPUTECL_ENABLED -DARMCOMPUTENEON_ENABLED -DARMNNREF_ENABLED -DARMNN_SERIALIZER -DARMNN_SERIALIZER_SCHEMA_PATH=\"/home/abuild/rpmbuild/BUILD/armnn-21.05/src/armnnSerializer/ArmnnSchema.fbs\" -DARMNN_TF_LITE_PARSER -DBOOST_ALL_DYN_LINK -DBOOST_ALL_NO_LIB -DDYNAMIC_BACKEND_BUILD_DIR=\"/home/abuild/rpmbuild/BUILD/armnn-21.05/build\" -I/home/abuild/rpmbuild/BUILD/armnn-21.05/include -I/home/abuild/rpmbuild/BUILD/armnn-21.05/profiling -I/home/abuild/rpmbuild/BUILD/armnn-21.05/src/backends -isystem /home/abuild/rpmbuild/BUILD/armnn-21.05/third-party -isystem /usr/include/half -mbranch-protection=standard -O2 -Wall -D_FORTIFY_SOURCE=2 -fstack-protector-strong -funwind-tables -fasynchronous-unwind-tables -fstack-clash-protection -Werror=return-type  -g -pthread  -Wno-error=deprecated-copy -Wno-error=deprecated-declarations -Wno-error=implicit-fallthrough -Wno-error=unused-parameter -std=c++14 -Wall -Wextra -Werror -Wold-style-cast -Wno-missing-braces -Wconversion -Wsign-conversion -O2 -g  -fPIC -MD -MT CMakeFiles/armnnUtils.dir/src/armnnUtils/FloatingPointConverter.cpp.o -MF CMakeFiles/armnnUtils.dir/src/armnnUtils/FloatingPointConverter.cpp.o.d -o CMakeFiles/armnnUtils.dir/src/armnnUtils/FloatingPointConverter.cpp.o -c /home/abuild/rpmbuild/BUILD/armnn-21.05/src/armnnUtils/FloatingPointConverter.cpp
[   89s] In file included from /home/abuild/rpmbuild/BUILD/armnn-21.05/src/armnnUtils/FloatingPointConverter.cpp:8:
[   89s] /home/abuild/rpmbuild/BUILD/armnn-21.05/src/armnnUtils/BFloat16.hpp: In member function 'float armnn::BFloat16::ToFloat32() const':
[   89s] /home/abuild/rpmbuild/BUILD/armnn-21.05/src/armnnUtils/BFloat16.hpp:89:17: error: 'u32' is used uninitialized [-Werror=uninitialized]
[   89s]    89 |         return *f32;
[   89s]       |                 ^~~
[   89s] /home/abuild/rpmbuild/BUILD/armnn-21.05/src/armnnUtils/BFloat16.hpp:87:24: note: 'u32' declared here
[   89s]    87 |         const uint32_t u32 = static_cast<uint32_t>(m_Value << 16u);
[   89s]       |                        ^~~
[   89s] cc1plus: all warnings being treated as errors
[   89s] make[2]: *** [CMakeFiles/armnnUtils.dir/build.make:177: CMakeFiles/armnnUtils.dir/src/armnnUtils/FloatingPointConverter.cpp.o] Error 1