[Bug middle-end/55219] [4.7/4.8 regression] attempting to compile a pre-processed unit eats up memory until OOM kills the cc1 process

[jakub at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=55219



--- Comment #6 from Jakub Jelinek <jakub at gcc dot gnu.org> 2012-11-07 
14:18:36 UTC ---

Slightly deobfuscated testcase:

int x, c, d, e, f, g, h, i;

double j;

const int k;

const enum { B } a;

void

fn1 (void)

{

  h = (g ? c : g ? f : g ? e : g ? i : g ? f : g ? e : g ? d : x)

      + (a ? : a ? : a ? : a ? : a ? : a ? : a ? : a ? : a ? : a ? : a

         ? j : a ? : 0 ? : a ? : a ? : a ? : a ? : a ? : a ? k : a ? : x);

}



Worst case we could introduce a static counter that could be increased at the

beginning of the fold_binary_op_with_conditional_arg and cleared at the end of

that function if it was originally 0, and just give up if there are more than

say a few hundred calls.