Hi Steven,
The output of "genautomata -stats btver2.md" is given below:
Automaton `btver2'
4 NDFA states, 8 NDFA arcs
4 DFA states, 8 DFA arcs
3 minimal DFA states, 7 minimal DFA arcs
176 all insns 3 insn equivalence classes
1 locked states
6 transition comb vector els, 9 trans table els: use simple vect
9 min delay table els, compression factor 8
Automaton `btver2_int'
128 NDFA states, 492 NDFA arcs
128 DFA states, 492 DFA arcs
128 minimal DFA states, 492 minimal DFA arcs
176 all insns 15 insn equivalence classes
0 locked states
589 transition comb vector els, 1920 trans table els: use comb vect
1920 min delay table els, compression factor 1
Automaton `btver2_agu'
48 NDFA states, 184 NDFA arcs
48 DFA states, 184 DFA arcs
39 minimal DFA states, 154 minimal DFA arcs
176 all insns 10 insn equivalence classes
0 locked states
176 transition comb vector els, 390 trans table els: use simple vect
390 min delay table els, compression factor 2
Automaton `btver2_fp'
995 NDFA states, 5567 NDFA arcs
995 DFA states, 5567 DFA arcs
995 minimal DFA states, 5567 minimal DFA arcs
176 all insns 54 insn equivalence classes
0 locked states
6394 transition comb vector els, 53730 trans table els: use comb vect
53730 min delay table els, compression factor 1
1453 all allocated states, 6220 all allocated arcs
4080 all allocated alternative states
7165 all transition comb vector els, 56049 all trans table els
56049 all min delay table els
1 all locked states
Regards,
Venkat.
-----Original Message-----
From: Steven Bosscher [mailto:[email protected]]
Sent: Thursday, January 17, 2013 7:10 PM
To: Kumar, Venkataramanan
Cc: Jan Hubicka; Uros Bizjak; [email protected]
Subject: Re: [Patch i386]: btver2 pipeline descriptions.
On Thu, Jan 17, 2013 at 2:18 PM, Kumar, Venkataramanan wrote:
> Hi Honza,
>
> There is no change when I find top 10 symbols using nm -t d --size-sort ./cc1
> | c++filt | tail.
More interesting would be the output of genautomata with the "-stats" flag.
Ciao!
Steven
