Reference to a global expression defined in a .global module causes a segfault

[tueda]

The following program doesn't make any sense in practice (.global just saves declarations), but it causes a segmentation fault:

Global F = 1;
.global
Local G = F;
.end

[vermaseren]

Hi Takahiro,

What happens, I guess, is that the .global does not work out the first
expression. After the .global that one is kind of 'hanging' in the sense that
the subexpression subterm that refers to the 1 now refers to the subexpression
that contains the F and hence we get a circular reference.

I guess the proper thing would be that the .global does not clear the
compiler buffer. The alternative would be to also let it work as a .sort

What do you think better?

Cheers

Jos

Quoting Takahiro Ueda notifications@github.com:

The following program doesn't make any sense in practice (.global
just saves declarations), but it causes a segmentation fault:

Global F = 1;
.global
Local G = F;
.end

---

Reply to this email directly or view it on GitHub:
#4

[tueda]

Hi Jos,

I guess the proper thing would be that the .global does not clear the compiler buffer. The alternative would be to also let it work as a .sort
What do you think better?

I think the former is fine if there is no side effect. Or maybe just printing an error or a warning would also be OK. It's a little thing and I guess no one will get in trouble about it except those who mistakenly use .global instead of .store like me :)