kargl
kargl
The backtrace suggests llvm 15. Address not mapped to object. llvm::Value::getType (this=0x0) at /usr/local/llvm15/include/llvm/IR/Value.h:255 255 Type *getType() const { return VTy; } I tried building on a different system than...
I got sidetrack. Someone has submitted a port of lfortran 0.34.0 to FreeBSD. It appears as [/usr/ports/lang/lfortran](https://cgit.freebsd.org/ports/tree/lang/lfortran) under FreeBSD's ports collections. Spent all morning building that due to dependencies. Still...
Any progress on this bug? It prevents me from compiling the first Fortran code needed to data mine `ISO_FORTRAN_ENV` for supported real kinds. The upshot is that I can compile...
For the march towards beta, it may be profitable to simply not worry about stack tracing on non-linux targets. Perhaps, a judicious use of `#if defined(__linux__)`. I'll note that this...
This bug report is quite confusing. There is no "default initialization" occurring anywhere within this program. In fact, it is impossible for "default initialization" to occur. In addition, the program...
> > There is no "default initialization" occurring anywhere within this program. > > I'm sorry, I think what I wanted to write is "explicit initialisation", Yes. F2023, 3.45 default...
There is a bug, but it is not in gfortran. The bug is in the Fortran code. 10.1.5.2.4 Evaluation of numeric intrinsic operations The execution of any numeric operation whose...
It seems that you're not looking at F2023 or F2018 documentation. gfortran's behavior is correct. See F2023. 16.9.88 FINDLOC (ARRAY, VALUE, DIM [, MASK, KIND, BACK]) or FINDLOC (ARRAY, VALUE...
Thank you for the fix. While my use case for matio is restricted to a rather simple IO, I have found matio a big help in providing portable output to...