Re: [lmi] 'wine test_coding_rules.exe' segfault

[Vadim Zeitlin]

On Fri, 15 May 2020 21:34:34 +0000 Greg Chicares <address@hidden> wrote:

GC> On 2020-05-15 17:57, Vadim Zeitlin wrote:
GC> > On Fri, 15 May 2020 17:18:07 +0000 Greg Chicares <address@hidden> wrote:
GC> > 
GC> > GC> To reproduce: build lmi with gcc_i686-w64-mingw32 version 9.3, then:
GC> [...]
GC> > GC> Segmentation fault
GC> > GC> make: [GNUmakefile:467: check_concinnity] Error 139 (ignored)
GC> > GC> 
GC> > GC> OTOH, it works (no segfault) with gcc-8.3 .
[...]
GC> >  Should I try to debug this
GC> 
GC> Sure, if you really want,

 I'll check if I can reproduce the problem with gcc 10.1, when it's
available and maybe spend more time on it then, but I won't do anything
right now if the immediate problem is fixed.

GC> but I think it's a matter of exceeding the command-line limit (of
GC> 'wine', probably), and the segfault or signal 11 I'm seeing indicates
GC> nothing more than that.

 Well, I don't see crashing with segfault as the best form of reporting an
overlong command line, but at least it's good to know that we have a work
around for now.

 Of course, knowing how you strive for minimalism, perhaps this could even
be seen as a feature: now we know that adding too many files to lmi will
fail the coding style test (if only by crashing the tester), so we'll
strive to be even more concise than before.

 Best regards,
VZ

pgpIjyHChPbAo.pgp
Description: PGP signature