On Mon, Feb 13, 2023 at 08:50:53PM +0100, Ralf Hemmecke wrote:
> I relized that the fricas-aldor interface is currently broken.
> (configure --enable-aldor)
> To be more exact, the problematic commit is
>
> 991f83471e23485f1cb68a78e6b2f02e8d137fba
>
> That innocent looking commit
Actually, this is change that exposed latent problems in our
code. One problem was in interpreter and was worked around
in previous commit. But there were also several fixes to
Spad compiler in the period between time when I first tried
this change and time when I could commit it without causing
regression. There were several other algebra changes that
required fixes to Spad compiler. This change could work
as a trigger for problem fixed in Spad, but not fixed in
Aldor.
> hangs the compilation of the interface when
> doing it with make -j8. With -j1 it gives a segfault when compiling
> ap/HOMOL.ap, see below.
>
> Interestingly, comparing the output of "cd BUILDDIR/src/aldor; make -j1"
> with what happens just one commit before, yields the exact same compilation
> commands up to HOMOL.ap.
>
> I'm currently a bit helpless of what exactly causes the problem that lets
> aldor jump into a segfault on compiling HOMOL.ap.
AFAICS this is Aldor crashing on content of HOMOL.ap. So one should
look at changes in HOMOL.ap and previous .ap files (a lot of type info
is defined in earlier files). Based on that one should be able to
say if this is interface probles, that is .ap files are wrong,
or Aldor problem (crahs on valid code).
--
Waldek Hebisch
--
You received this message because you are subscribed to the Google Groups
"FriCAS - computer algebra system" group.
To unsubscribe from this group and stop receiving emails from it, send an email
to fricas-devel+unsubscr...@googlegroups.com.
To view this discussion on the web visit
https://groups.google.com/d/msgid/fricas-devel/20230213221434.u22dab6et5eitdw4%40fricas.math.uni.wroc.pl.
