Hi everyone, welcome to the third blog of my GSoC'25 journey. A major update of my work from the third week is successful compilation of Test Open Maps of package hashmaps. Now, Test Open Maps is compiling to 100% without any Semantic Error, ASR Verify Pass Error, Internal Compilation Error or CodeGeneration Error. But still the complete test file is not passing the test due to some segmentation fault encountered during the runtime and some error stop failures on conditional checks.
I started the week by fixing Issue #7438 as the error was related to basic Assignment operation which is used multiple times in the package hashmaps. The Link to the Merged Pull Request is here. After this, I improved the workaround of deserialization error and created the workarounds for various other issues for whom fixing the MREs was a bit tricky task.
Appropriate Workarounds created for various issues are :-
- Improved Workaround for Deserialization Error
- Workaround for Move Alloc Error
- Workaround for Polymorphic Argument Error
- Workaround for Linker Error
With these workarounds and some other fixes, Test Open Maps is now working compiling to 100% using LFortran. Also, with these workaround, Test Open Map works well and passes all the tests using GFortran. This ensures that all the workarounds are correct and working well which was not the case before as some workarounds were causing test to fail via GFortran.
Following is the list of Merged Pull Requests for this week that helped in compilation of Test Open Maps :-
After the successful compilation of Test Open Maps, the next step is to ensure that the test works well and executes correctly and passes the test without any error stop failure or segmentation fault failures. I encountered some segmentation fault and created MREs and workarounds for them.
Following is the list of MREs regarding segmentation fault created this week :-
I succeeded in creating workarounds for them. Following is the list of workarounds related to segmentation fault :-
Still, I haven't got rid of segmentation fault completely and for my next week, my aim would be to identify all the segmentation fault failures and other error stop failures, create workarounds for them or fix them and make test open maps execute correctly. After doing this, I will be creating a clean branch to collect all the required workaround and add these workarounds to our base branch to setup Test Open Maps at CI. I am quite optimistic that I will set up CI for test open maps in the following week and if time permits, I will also start my work on other Test files and Example files of package hashmaps.
Overall, I worked for 28 hours this week and enjoyed the work that I did in the third week and would like to thank Ondrej Certik, Harshita Kalani, Pranav Goswami and all the other LFortran members for their reviews and suggestions which helped me a lot to tackle new difficulties. I am looking forward to continue my journey in the next week with the same excitement and enthusiasm and plan to complete my proposed tasks as quickly as I can.