[Bug rtl-optimization/119010] [15 Regression] 444.namd shows a huge compile-time regression with -mtune=znver5

[rguenth at gcc dot gnu.org via Gcc-bugs]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=119010

--- Comment #2 from Richard Biener <rguenth at gcc dot gnu.org> ---
Hmm, the automaton is the same as the znver4 one it seems.  Here's it's
statistics, znver4_fpu is by far the largest:

Automaton `znver4'
        5 NDFA states,             15 NDFA arcs 
        5 DFA states,              15 DFA arcs
        5 minimal DFA states,      15 minimal DFA arcs
     1818 all insns          4 insn equivalence classes
    0 locked states
   18 transition comb vector els,    20 trans table els: use simple vect
   20 min delay table els, compression factor 8

Automaton `znver4_ieu'
      419 NDFA states,           3563 NDFA arcs 
      419 DFA states,            3563 DFA arcs
      419 minimal DFA states,    3563 minimal DFA arcs
     1818 all insns         20 insn equivalence classes
    0 locked states
 5006 transition comb vector els,  8380 trans table els: use simple vect
 8380 min delay table els, compression factor 2

Automaton `znver4_idiv'
       15 NDFA states,             39 NDFA arcs 
       15 DFA states,              39 DFA arcs
       12 minimal DFA states,      33 minimal DFA arcs
     1818 all insns          8 insn equivalence classes
    0 locked states
   33 transition comb vector els,    96 trans table els: use comb vect
   96 min delay table els, compression factor 2

Automaton `znver4_fdiv'
       15 NDFA states,             43 NDFA arcs 
       15 DFA states,              43 DFA arcs
       11 minimal DFA states,      35 minimal DFA arcs
     1818 all insns         11 insn equivalence classes
    0 locked states
   41 transition comb vector els,   121 trans table els: use comb vect
  121 min delay table els, compression factor 2

Automaton `znver4_agu'
       38 NDFA states,            211 NDFA arcs
       38 DFA states,             211 DFA arcs
       38 minimal DFA states,     211 minimal DFA arcs
     1818 all insns         18 insn equivalence classes
    0 locked states
  257 transition comb vector els,   684 trans table els: use comb vect
  684 min delay table els, compression factor 2

Automaton `znver4_fpu'
     2456 NDFA states,          37781 NDFA arcs
     2456 DFA states,           37781 DFA arcs
     2456 minimal DFA states,   37781 minimal DFA arcs
     1818 all insns         37 insn equivalence classes
    0 locked states
57619 transition comb vector els, 90872 trans table els: use simple vect
90872 min delay table els, compression factor 2

Automaton `znver4_fp_store'
      480 NDFA states,           3344 NDFA arcs
      480 DFA states,            3344 DFA arcs
      480 minimal DFA states,    3344 minimal DFA arcs
     1818 all insns         16 insn equivalence classes
    0 locked states
 3969 transition comb vector els,  7680 trans table els: use simple vect
 7680 min delay table els, compression factor 2