[postgis-devel] [Internet] Re: Postgis 2.4.6 regession tests fails on Solaris

[Raúl Marín Rodríguez]

>   something else I didn't think of

There is the different floating point arithmetics that the compiler can make
use of. IIRC correctly, I've had issues with different compilers using
different bit sizes for floating point operations 64 or 80 bit, which
after rounding can lead to different values. This can even happen whithin the
same compiler: vectorized code uses 64b operators while "normal" uses whatever
is native (usually 80b in modern x86 cpus).

Debugging differences like this is hard even locally, so you can understand
that nobody is too exited to do it without a machine with the "failing"
architecture.

--
Raúl Marín Rodríguez
carto.com