Re: fatal: internal error Ver 5.2.0 readall()

[arnold]

Eli Zaretskii <eliz@gnu.org> wrote:

> > From: arnold@skeeve.com
> > Date: Sun, 11 Sep 2022 01:07:52 -0600
> > Cc: jnaman2@gmail.com, bug-gawk@gnu.org
> > 
> > Eli Zaretskii <eliz@gnu.org> wrote:
> > 
> > > > From: arnold@skeeve.com
> > > > Date: Sun, 11 Sep 2022 00:45:48 -0600
> > > > Cc: eliz@gnu.org
> > > > 
> > > > Hi.
> > > > 
> > > > This does not reproduce on x86_64 under Linux or MacOS. It also
> > > > doesn't reproduce on Linux if I compile gawk in 32 bit mode.
> > > > 
> > > > Eli - Can you take a look please?
> > >
> > > I get the same error as the OP.
> > 
> > So, can you please point a debugger at it to try to figure out what's
> > happening?
>
> Doesn't the error message tell what is happening?
>
> If not, where to look for more details?
>
> Btw, if I compile Gawk without optimizations, the readall program
> succeeds(?) in the unoptimized version, and produces the following
> output:
>
>   arr1: len=0
>   0 = readall("foo.binawk");
>   arr1: len=0
>
> Does this help?

I will take it from here.  Thanks

Arnold