ATF Test failure

[Paul Goyette <paul%whooppee.com@localhost>]

After a rather lengthy period of build breakage, the amd64 port is now 
building successfully.  However, some tests are apparently taking a much 
longer time to execute, causing the overall execution time for the tests 
to exceed the allotted 3 hours.  Prior to the breakage, a normal test 
pass would require approximately 2.4 hours, while we now exceed 3.0 
hours to exdecute just the first 365 tests case (out of 567 total 
cases).

I have not compared every test involved to identify which ones are 
taking so long to complete.  However, one specific test has been 
identified.  The test case grow_16M_v0_8192 from test program 
sbin/resize_ffs/t_grow used to run in approximately 40 seconds, but now 
fails to complete in more than 10 minutes - 15 times as long as before.

There were quite a few commits during the build breakage (beginning at 
2013-05-02 03:50:08, and ending at 2013-05-04 01:20:00), but most of the 
liekly suspects are rump-related.  Commits for disklabel, vndcompress, 
installboot, and various arm-specific commits are unlikely to be the 
culprits here.


-------------------------------------------------------------------------
| Paul Goyette     | PGP Key fingerprint:     | E-mail addresses:       |
| Customer Service | FA29 0E3B 35AF E8AE 6651 | paul at whooppee.com    |
| Network Engineer | 0786 F758 55DE 53BA 7731 | pgoyette at juniper.net |
| Kernel Developer |                          | pgoyette at netbsd.org  |
-------------------------------------------------------------------------