Hi Basile,
thank you for the answer.
No, I did not observe the error you mentioned.
Intel MKL is a free library but it is not open source so I cannot compile
it directly but I can install it. I don't think the accuracy is a problem
here since I am just calling the external library so it should use the
options intel chose to compile it. I was wondering about a way to
understand the differences in the linking when using
-funsafe-math-optimizations.
Best regards,
Nicola
Il giorno gio 5 giu 2025 alle ore 14:36 Basile Starynkevitch <
basile@xxxxxxxxxxxxxxxxx> ha scritto:
> On Thu, 2025-06-05 at 14:23 +0200, Nicola Giuliani via Gcc-help wrote:
> > Hi everybody,
> >
> > My name is Nicola Giuliani and I am a researcher in numerical analysis. I
> > have encountered a strange behavior when compiling a very simple piece of
> > code using g++. In this piece of code I link the InterlOneAPI MKL PARDISO
> > solver and ask it to solve a linear system. I don't do any mathematical
> > operation, I simply call the MKL libraries that I have dynamically
> linked.
>
>
> I suppose MKL PARDISO is a set of open source libraries, whose source code
> you
> could compile or statically analyze.
> >
> > This is the original compilation line I use
> >
> > g++ kernel_only_pardiso.cpp -m64 -L${MKLROOT}/lib -Wl,--no-as-needed
> > -lmkl_intel_lp64 -lmkl_gnu_thread -lmkl_core -lgomp -lpthread -lm -ldl
> > -I"${MKLROOT}/include" -o gnu_gnu.out
> >
>
> [...]
> >
> https://community.intel.com/t5/Intel-oneAPI-Math-Kernel-Library/PARDISO-performance-fluctuation-on-INTEL-processor/m-p/1693601/emcs_t/S2h8ZW1haWx8bWVudGlvbl9zdWJzY3JpcHRpb258TUJCNzJMQ0ZJRUFWQ0J8MTY5MzYwMXxBVF9NRU5USU9OU3xoSw
>
>
> Possible open source tools related to estimate accuracy of C or C++ code
> might
> be:
> https://frama-c.com/
>
> (I have been employed and part of the team working on it; see
> also https://arxiv.org/abs/1109.0779 and
> https://github.com/bstarynk/bismon that
> I don't maintain anymore)
>
> https://www-pequan.lip6.fr/cadna/
>
> Perhaps they could help you to understand the issues.
>
> https://www.absint.com/products.htm is selling related tools.
>
> BTW did you observe (or can you reproduce) the following bug on GCC trunk
> (on
> x86-64/Linux/Debian/Trixie)?
>
> the https://gcc.gnu.org/bugzilla/show_bug.cgi?id=120556
> https://gcc.gnu.org/pipermail/gcc-bugs/2025-June/916978.html
>
>
> Regards.
> --
> Basile STARYNKEVITCH <basile@xxxxxxxxxxxxxxxxx>
> 8 rue de la Faïencerie
> http://starynkevitch.net/Basile/
> 92340 Bourg-la-Reine https://github.com/bstarynk
> France
> https://github.com/RefPerSys/RefPerSys
>
