bin/53678: Various t_asan_poison test cases failing on i386

[gson%gson.org@localhost (Andreas Gustafsson)]

>Number:         53678
>Category:       bin
>Synopsis:       Various t_asan_poison test cases failing on i386
>Confidential:   no
>Severity:       non-critical
>Priority:       medium
>Responsible:    bin-bug-people
>State:          open
>Class:          sw-bug
>Submitter-Id:   net
>Arrival-Date:   Sun Oct 21 10:05:00 +0000 2018
>Originator:     Andreas Gustafsson
>Release:        NetBSD-current, source date >= c. 2018.04.11.06.41.23
>Organization:
>Environment:
System: NetBSD
Architecture: i386
Machine: i386
>Description:

The following test cases have been failing on the qemu-based TNF i386
testbed ever since they were created in April 2018:

  usr.bin/cc/t_asan_poison/poison
  usr.bin/cc/t_asan_poison/poison_pic
  usr.bin/c++/t_asan_poison/poison
  usr.bin/c++/t_asan_poison/poison_pic

The following test cases are failing on my own i386 testbed, running
on real hardware:

  usr.bin/cc/t_asan_poison/poison_pie
  usr.bin/c++/t_asan_poison/poison_pie

Each fails only on the specific testbed reported, and passes on the
other.  They do not fail on amd64.

Log output from recent failures:

  http://releng.netbsd.org/b5reports/i386/2018/2018.10.20.08.47.03/test.html#usr.bin_cc_t_asan_poison_poison

  http://www.gson.org/netbsd/bugs/build/i386-baremetal/2018/2018.10.18.19.02.46/test.html#usr.bin_cc_t_asan_poison_poison_pie

>How-To-Repeat:

Run the ATF tests on i386, either under qemu or on real hardware
depending on which test of failures you want to reproduce.

>Fix: