1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
|
# This test isn't very useful until we can test subroutine timings
# perhaps by adding an option to nytprofcsv to include them
# and adjusting test.pl to test for them (including the ~N fudge factor).
# Meanwhile the test is useful for sanity checking the subroutine timing
# code using a command like
# make && NYTPROF_TEST=trace=3 perl -Mblib test.pl -leave=1 -use_db_sub=0 t/test70-subexcl.*
my $T = $ENV{NYTPROF_TEST_PAUSE_TIME} || 0.2;
sub A { # inclusive ~= $T, exclusive ~= $T
select undef, undef, undef, $T;
}
sub B { # inclusive ~= $T*2, exclusive ~= $T
A();
select undef, undef, undef, $T;
}
sub C { # inclusive ~= $T*2, exclusive ~= 0.0
B();
}
sub D { # inclusive ~= $T*4, exclusive ~= 0.0
C();
C(); # cumulative_subr_secs non-zero on sub entry
}
D();
|
