Release/2.8.0

[jholloc]

Release targetting getting UDA to build on Windows CI runner

[SimonPinches]

Is the build / install successful on system based on OpenSSL/1.1 and OpenSSL/3, e.g. RHEL9 ?

[stephen-dixon]

Would anyone have an issue with reverting the API calls that were broken earlier in 2.7.0 when the new "client_flags" argument was introduced? I wanted to add a header of wrapper functions for other (accidental) breaking changes to the API (such as idamFree -> udaFree) but obviously function overloads with the same name will not work in the C layer, which applies to all the client_flags functions.

In my opinion the additional argument is redundant as the call to udaClientFlags() can be made within in each function instead of taking it as an input parameter.

This change will definitely take place in the uda/3.0.0 release, but I'd also like to include it in the 2.7.6 release and update 2.8.0 to follow this change.

For reference, most instances occur within the source/client/accAPI.h file

e.g.

LIBRARY_API void setIdamProperty(const char* property, CLIENT_FLAGS* client_flags);

The proposed new function signature would be e.g.

LIBRARY_API void setIdamProperty(const char* property);

[olivhoenen]

@stephen-dixon I think it is fine if this change is being added to 2.8.0.

[olivhoenen]

Can someone add https://github.com/deepakmaroo to this PR?

[stephen-dixon]

Can someone add https://github.com/deepakmaroo to this PR?

Do they need to be added as a collaborator to the uda project? I don't have the privileges to do this

[stephen-dixon]

@olivhoenen @DorisBDR Apologies for the confusion, but on reflection I realise we cannot remove the client flags from the API function signatures in the way I previously described, so I won't try to change them in 2.7.6.

I will still add a header to include mappings for other name changes (such as idamFree -> udaFree) and I'll commit those changes to this branch too.

[deepakmaroo]

Hello Team,
In a use case, I am using IMAS plugin. I have build UDA 2.8.0 and uda-plugins (IMAS) on sdcc, but when tried to access uda_cli -h localhost -p 56565 "IMAS::open(uri='imas:mdsplus?path=/work/imas/shared/imasdb/ITER/3/134174/117', mode='open')"
it ends with error:
getPluginAddress.cpp:40 >> Cannot open the target shared library libimas_plugin.so: libuda_cpp.so.0: cannot open shared object file: No such file or directory [1007 help() 0 -1 IMAS IMAS ] [handleRequest] [Protocol 10 Error (Receiving Client Block)]

It works when renamed ./lib/libuda_cpp.so to ./lib/libuda_cpp.so.0

UDA 2.7.5 and previous modules has libuda_cpp.so.0 library in ./lib directory

which is missing in 2.8.0.

In initPluginList, getPluginAddress is expecting libuda_cpp.so.0 but build and installed is libuda_cpp.so.2

[stephen-dixon]

@DorisBDR this should now be ready for testing

[DorisBDR]

i'm having two issues with 2.8.0

1. compilation error
   c/source/plugins/bytes/bytesPlugin.cpp:175:48: error: cannot bind non-const lvalue reference of type ‘char*&’ to an rvalue of type ‘char*’
   ERROR] /mnt/ITER/abadiel/udacert/uda280/mudaidamtrunk/target/main/c/source/plugins/bytes/bytesPlugin.cpp:175:48: error: cannot bind non-const lvalue reference of type ‘char*&’ to an rvalue of type ‘char*’
   [WARN] boost::split(allowed_paths, std::getenv("UDA_BYTES_PLUGIN_ALLOWED_PATHS"), boost::is_any_of(";"));
   [WARN] ~~~~~~~~~~~^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
   [WARN] In file included from /usr/include/boost/algorithm/string.hpp:23,
   [WARN] from /mnt/ITER/abadiel/udacert/uda280/mudaidamtrunk/target/main/c/source/plugins/bytes/bytesPlugin.cpp:21:
   [WARN] /usr/include/boost/algorithm/string/split.hpp:140:35: note: initializing argument 2 of ‘SequenceSequenceT& boost::algorithm::split(SequenceSequenceT&, RangeT&, PredicateT, boost::algorithm::token_compress_mode_type) [with SequenceSequenceT = std::vector<std::__cxx11::basic_string >; RangeT = char*; PredicateT = boost::algorithm::detail::is_any_ofF]’
   [WARN] inline SequenceSequenceT& split(
   [WARN] ^~~~~
   [WARN] At global scope:

2. [WARN] At global scope:
   [ERROR] cc1plus: error: unrecognized command line option ‘-Wno-use-after-free’ [-Werror]
   [ERROR] cc1plus: error: unrecognized command line option ‘-Wno-mismatched-new-delete’ [-Werror]
   [ERROR] cc1plus: error: unrecognized command line option ‘-Wno-use-after-free’ [-Werror]
   [ERROR] cc1plus: all warnings being treated as errors

FYI
[abadiel@ccs73n-1 mudaidamtrunk]$ gcc --version
gcc (GCC) 8.5.0 20210514 (Red Hat 8.5.0-4)
Copyright (C) 2018 Free Software Foundation, Inc.
This is free software; see the source for copying conditions. There is NO
warranty; not even for MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.

[abadiel@ccs73n-1 mudaidamtrunk]$

replacing the code with the following two lines solved the issue in my case
std::string envuda=std::getenv("UDA_BYTES_PLUGIN_ALLOWED_PATHS");
boost::split(allowed_paths, envuda, boost::is_any_of(";"));

[DorisBDR]

another issue is with the version
src/main/c/CMakeLists.txt -> git archive shows 2.7.6 i had to update to 2.8.0 (in my case i downloaded a zip of the uda)

[stephen-dixon]

another issue is with the version src/main/c/CMakeLists.txt -> git archive shows 2.7.6 i had to update to 2.8.0 (in my case i downloaded a zip of the uda)

I think the git archive will only show 2.8.0 as the version when we merge the PR and tag a new release. Was the version 2.7.6 exactly, or was it something like 2.7.6.XXXX?

[DorisBDR]

another issue is with the version src/main/c/CMakeLists.txt -> git archive shows 2.7.6 i had to update to 2.8.0 (in my case i downloaded a zip of the uda)

I think the git archive will only show 2.8.0 as the version when we merge the PR and tag a new release. Was the version 2.7.6 exactly, or was it something like 2.7.6.XXXX?

it was something like 2.7.6.XXXX

[stephen-dixon]

another issue is with the version src/main/c/CMakeLists.txt -> git archive shows 2.7.6 i had to update to 2.8.0 (in my case i downloaded a zip of the uda)

I think the git archive will only show 2.8.0 as the version when we merge the PR and tag a new release. Was the version 2.7.6 exactly, or was it something like 2.7.6.XXXX?

it was something like 2.7.6.XXXX

OK this is expected behaviour until we tag the next release

[stephen-dixon]

For the unrecognised command-line options I think it's a case of this: (ie warnings are only raised when there are other issues during compilation)

When an unrecognized warning option is requested (e.g., -Wunknown-warning),
GCC emits a diagnostic stating that the option is not recognized.

However, if the -Wno- form is used, the behavior is slightly different:
no diagnostic is produced for -Wno-unknown-warning unless other diagnostics
are being produced.

This allows the use of new -Wno- options with old compilers, but if something
goes wrong, the compiler warns that an unrecognized option is present.

[DorisBDR]

i have a doubt ...in this version shall we see systemd too? where and how it is managed
thanks

[DorisBDR]

ok i found these 2 files c/source/etc/uda@.service.in and c/source/etc/uda.socket.in i guess it is this one

[stephen-dixon]

ok i found these 2 files c/source/etc/uda@.service.in and c/source/etc/uda.socket.in i guess it is this one

Yes that's right. I think the syntax for deploying the service is something like

sudo cp <install_dir>/etc/uda.socket <install_dir>l/etc/uda@.service /etc/systemd/system &&
sudo systemctl start uda.socket &&
sudo systemctl enable uda.socket