Assem Medhat

use `git clean -dfx` then build again. but make sure to stage your changes, if you have untracked files they we'll get removed after running `git clean -dfx`.

I don't know how bison typically work, but it's a declarative parser; You just tell it how the language expression and primitives are combined to build the language. so using...

Is this LLVM 15 or 16?

I get some similar issue here in this PR https://github.com/lfortran/lfortran/pull/5039 in here https://github.com/lfortran/lfortran/actions/runs/11317506647/job/31471034781?pr=5039. after investigating the issue. I see that ASR is correct but the llvm output neglects some pieces...

I'll try to see what's the issue. mine works also when I use `print`.

I have Ubuntu. and yeah it takes so long, I though it was normal for stack trace unwinding.

Yeah We have the backend ready for tackling such errors in the runtime, for now it either segfaults or print some random data. I think if you run the same...

It's undefined behaviour, but for this case it works with random numbers id you fliped the situation, this should segfault at some point ``` program badint print "(A)",1111 end program...

Performance of PRIMA with `lf-cleaned-branch-11` branch | | LFortran (`message.f90` wokraround) ---------------------- _`--fast`_ | lfortran (`message.f90`wokraround)(NOFAST) | gfortran (`message.f90`wokraround)_`-O3 -march=native -ffast-math`_ | gfortran(`message.f90`wokraround)(NOFAST) | |----------|:----------------------------------------------------------:|:----------------------------------------------:|:-------------------------------------------------------------------------------:|:---------------------------------------------:| | Bobyqa_1 | real 0m0.006suser...

I'll try handling this one.