Hi everyone,
I just noticed that the percentages for some of the distribution stats for the
O3CPU (such as cpu.commit.COM:committed_per_cycle, cpu0.fetch.rateDist, but
not cpu.iq.ISSUE:FU_type_0) are off by a factor of 100, e.g.:
system.cpu1.commit.COM:committed_per_cycle.start_dist #
Number of insts commited each cycle
system.cpu1.commit.COM:committed_per_cycle.samples 699261
system.cpu1.commit.COM:committed_per_cycle.min_value 0
0 346161 4950.38%
1 105400 1507.31%
2 81999 1172.65%
3 72064 1030.57%
4 70346 1006.00%
5 5584 79.86%
6 4121 58.93%
7 943 13.49%
8 12643 180.81%
system.cpu1.commit.COM:committed_per_cycle.max_value 8
system.cpu1.commit.COM:committed_per_cycle.end_dist
Here's a patch that fixes it, although I dunno if this was the right place to
fix it:
--- ../m5-2.0b3-orig/src/base/stats/text.cc 2007-05-17 04:26:07.000000000
+0200
+++ src/base/stats/text.cc 2007-09-04 16:14:15.798458307 +0200
@@ -479,7 +479,7 @@
continue;
_min = i * bucket_size + min;
- _pdf = vec[i] / total * 100.0;
+ _pdf = vec[i] / total ; // do not multiply by 100 here... This
seems to be already done elsewhere.
_cdf += _pdf;
Best regards
Jonas.
--
Dipl.-Ing. Jonas Diemer
Institut für Datentechnik und Kommunikationsnetze
(Institute of Computer and Communication Network Engineering)
Hans-Sommer-Str. 66
D-38106 Braunschweig
Germany
Telefon: +49 531 391 3752
Telefax: +49 531 391 4587
E-Mail: [EMAIL PROTECTED]
Web: http://www.ida.ing.tu-bs.de/
_______________________________________________
m5-users mailing list
[email protected]
http://m5sim.org/cgi-bin/mailman/listinfo/m5-users
