Experimenting with tree inliner parameter settings

[Richard Guenther]

On Sun, 4 May 2003, John David Anglin wrote:

> > So the question is: What do these numbers do for your code,
> > both in terms of compile time and in runtime performance?
> > Richard, Dave, Gerald, can you please test the settings
> > below on your favorite code and on the test case for
> > PR 10160 and PR 8361 (and maybe PR 10316)?
> >
> > max-inline-insns-single=230
> > max-inline-insns-auto=230
> > max-inline-insns=500
> > min-inline-insns=80
> > max-inline-slope=20
>
> Here is some testing on the PA under hpux.

Here is the above on my favorite testcase (PR already closed):

g++-3.3 (GCC) 3.3 20030428 (prerelease), -fno-exceptions -O2

default params:

Execution times (seconds)
 garbage collection    :   0.32 ( 3%) usr   0.00 ( 0%) sys   0.00 ( 0%)
 parser                :   2.33 (20%) usr   0.37 (32%) sys   3.50 (27%)
 name lookup           :   1.14 (10%) usr   0.33 (28%) sys   1.00 ( 8%)
 expand                :   0.92 ( 8%) usr   0.12 (10%) sys   1.50 (12%)
 CSE                   :   1.41 (12%) usr   0.00 ( 0%) sys   1.00 ( 8%)
 global CSE            :   0.38 ( 3%) usr   0.03 ( 3%) sys   0.50 ( 4%)
 loop analysis         :   0.27 ( 2%) usr   0.00 ( 0%) sys   0.00 ( 0%)
 CSE 2                 :   0.56 ( 5%) usr   0.00 ( 0%) sys   0.50 ( 4%)
 local alloc           :   0.21 ( 2%) usr   0.00 ( 0%) sys   0.00 ( 0%)
 global alloc          :   0.38 ( 3%) usr   0.01 ( 1%) sys   0.00 ( 0%)
 scheduling 2          :   0.33 ( 3%) usr   0.01 ( 1%) sys   0.00 ( 0%)
 rest of compilation   :   0.20 ( 2%) usr   0.00 ( 0%) sys   0.00 ( 0%)
 TOTAL                 :  11.80             1.16            13.00

with params from above:

Execution times (seconds)
 garbage collection    :   0.31 ( 4%) usr   0.00 ( 0%) sys   0.00 ( 0%)
 parser                :   2.63 (30%) usr   0.33 (43%) sys   1.50 (16%)
 name lookup           :   1.05 (12%) usr   0.20 (26%) sys   1.00 (11%)
 expand                :   0.44 ( 5%) usr   0.01 ( 1%) sys   2.50 (26%)
 CSE                   :   0.74 ( 8%) usr   0.00 ( 0%) sys   0.00 ( 0%)
 global CSE            :   0.21 ( 2%) usr   0.01 ( 1%) sys   0.50 ( 5%)
 loop analysis         :   0.09 ( 1%) usr   0.00 ( 0%) sys   0.00 ( 0%)
 CSE 2                 :   0.29 ( 3%) usr   0.00 ( 0%) sys   1.00 (11%)
 local alloc           :   0.10 ( 1%) usr   0.00 ( 0%) sys   0.00 ( 0%)
 global alloc          :   0.25 ( 3%) usr   0.00 ( 0%) sys   0.50 ( 5%)
 scheduling 2          :   0.18 ( 2%) usr   0.02 ( 3%) sys   0.00 ( 0%)
 rest of compilation   :   0.12 ( 1%) usr   0.00 ( 0%) sys   0.00 ( 0%)
 TOTAL                 :   8.79             0.76             9.50

Performance for those params is bad, as expected, but it was the
same before.

Richard.

--
Richard Guenther <richard dot guenther at uni-tuebingen dot de>
WWW: http://www.tat.physik.uni-tuebingen.de/~rguenth/