[Bug tree-optimization/108500] [11/12 Regression] -O -finline-small-functions results in "internal compiler error: Segmentation fault" on a very large program (700k function calls)

["rguenth at gcc dot gnu.org" <gcc-bugzilla@gcc.gnu.org>]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=108500

--- Comment #16 from Richard Biener <rguenth at gcc dot gnu.org> ---
(In reply to Richard Biener from comment #14)
> Martin, can you look at the SRA issue?  Do you want me to create a separate
> bugreport for this?  The IL into SRA looks like
> 
>   <bb 2> :
>   s2D.2755 = {};
>   s1D.2756 = {};
>   _unusedD.2002766 = s1D.2756;
>   sD.2002767 = s2D.2755;
>   s2D.2755 = sD.2002767;
>   _unusedD.2002766 ={v} {CLOBBER(eol)};
>   sD.2002767 ={v} {CLOBBER(eol)};
>   _unusedD.2002764 = s1D.2756;
>   sD.2002765 = s2D.2755;
>   s2D.2755 = sD.2002765;
>   _unusedD.2002764 ={v} {CLOBBER(eol)};
>   sD.2002765 ={v} {CLOBBER(eol)};
>   _unusedD.2002762 = s1D.2756;
>   sD.2002763 = s2D.2755;
>   s2D.2755 = sD.2002763;
>   _unusedD.2002762 ={v} {CLOBBER(eol)};
>   sD.2002763 ={v} {CLOBBER(eol)};
>   _unusedD.2002760 = s1D.2756;
>   sD.2002761 = s2D.2755;
>   s2D.2755 = sD.2002761;
>   _unusedD.2002760 ={v} {CLOBBER(eol)};
>   sD.2002761 ={v} {CLOBBER(eol)};
> ...

struct s1 {
   char a[20] ;
   char b[20] ;
};
struct s2 {
   int id;
   char a[20];
};
static inline  __attribute__((always_inline)) struct s2
f(struct s1 _unused, struct s2 s)
{ 
  return s;
} 

volatile struct s2 x;
int main(void)
{
  struct s2 s2 = {0};
  struct s1 s1 = {0};
#define TEN \
  s2 = f(s1,s2); \
  s2 = f(s1,s2); \
  s2 = f(s1,s2); \
  s2 = f(s1,s2); \
  s2 = f(s1,s2); \
  s2 = f(s1,s2); \
  s2 = f(s1,s2); \
  s2 = f(s1,s2); \
  s2 = f(s1,s2); \
  s2 = f(s1,s2);
  TEN
  TEN
  x = s2;
  return 0;
}

shows this.  With

diff --git a/gcc/tree-sra.cc b/gcc/tree-sra.cc
index ad0c738645d..ab3f47badcb 100644
--- a/gcc/tree-sra.cc
+++ b/gcc/tree-sra.cc
@@ -2980,6 +2980,7 @@ propagate_subaccesses_from_lhs (struct access *lacc,
struct access *racc)
 static void
 propagate_all_subaccesses (void)
 {
+  unsigned cnt = 0;
   propagation_budget = new hash_map<tree, unsigned>;
   while (rhs_work_queue_head)
     {
@@ -2994,6 +2995,7 @@ propagate_all_subaccesses (void)
        {
          struct access *lacc = link->lacc;

+         cnt++;
          if (!bitmap_bit_p (candidate_bitmap, DECL_UID (lacc->base)))
            continue;
          lacc = lacc->group_representative;
@@ -3019,6 +3021,7 @@ propagate_all_subaccesses (void)
            while (lacc);
        }
     }
+  fprintf (stderr, "%d\n", cnt);

   while (lhs_work_queue_head)
     {

we have with one TEN
> ./cc1 -quiet t.c -O
0
220
120

and with two TEN
> ./cc1 -quiet t.c -O
0
840
440

and with four TEN
> ./cc1 -quiet t.c -O
0
3280
1680

that's quadratic and has a quite high linear factor as well.