[
https://issues.apache.org/jira/browse/STDCXX-898?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12601289#action_12601289
]
Travis Vitek commented on STDCXX-898:
-------------------------------------
This test fails 13 assertions on all platforms I checked with the exception of
AIX. I did a quick test and on Linux and SunOS it is the same exact set of
failures. So, one option would be to hardcode the test to expect traps on all
platforms except AIX. Another option would be to remove the traps portion of
this test as it is thoroughly tested in
[18.limits.traps.cpp|http://svn.apache.org/viewvc/stdcxx/branches/4.2.x/tests/support/18.limits.traps.cpp?view=markup].
It seems to me that the best thing to do here is to remove the failing
conditions as they are being tested.
> [MSVC] 18.numeric.special.int test fails
> ----------------------------------------
>
> Key: STDCXX-898
> URL: https://issues.apache.org/jira/browse/STDCXX-898
> Project: C++ Standard Library
> Issue Type: Bug
> Components: Tests
> Affects Versions: 4.2.1
> Environment: MSVC
> Reporter: Farid Zaripov
> Assignee: Travis Vitek
> Priority: Trivial
> Fix For: 4.2.2
>
> Original Estimate: 1h
> Remaining Estimate: 1h
>
> The 18.numeric.special.int test fails with the following assertions:
> {noformat}
> # ASSERTION (S7) (4 lines):
> # TEXT: std::numeric_limits<bool>::traps == 0, got true
> # CLAUSE: numeric.special
> # LINE: 587
> # ASSERTION (S7) (4 lines):
> # TEXT: std::numeric_limits<char>::traps == 0, got true
> # CLAUSE: numeric.special
> # LINE: 587
> # ASSERTION (S7) (4 lines):
> # TEXT: std::numeric_limits<signed char>::traps == 0, got true
> # CLAUSE: numeric.special
> # LINE: 587
> # ASSERTION (S7) (4 lines):
> # TEXT: std::numeric_limits<unsigned char>::traps == 0, got true
> # CLAUSE: numeric.special
> # LINE: 587
> # ASSERTION (S7) (4 lines):
> # TEXT: std::numeric_limits<short>::traps == 0, got true
> # CLAUSE: numeric.special
> # LINE: 587
> # ASSERTION (S7) (4 lines):
> # TEXT: std::numeric_limits<unsigned short>::traps == 0, got true
> # CLAUSE: numeric.special
> # LINE: 587
> # ASSERTION (S7) (4 lines):
> # TEXT: std::numeric_limits<int>::traps == 0, got true
> # CLAUSE: numeric.special
> # LINE: 587
> # ASSERTION (S7) (4 lines):
> # TEXT: std::numeric_limits<unsigned int>::traps == 0, got true
> # CLAUSE: numeric.special
> # LINE: 587
> # ASSERTION (S7) (4 lines):
> # TEXT: std::numeric_limits<long>::traps == 0, got true
> # CLAUSE: numeric.special
> # LINE: 587
> # ASSERTION (S7) (4 lines):
> # TEXT: std::numeric_limits<unsigned long>::traps == 0, got true
> # CLAUSE: numeric.special
> # LINE: 587
> # ASSERTION (S7) (4 lines):
> # TEXT: std::numeric_limits<_RWSTD_LONG_LONG>::traps == 0, got true
> # CLAUSE: numeric.special
> # LINE: 587
> # ASSERTION (S7) (4 lines):
> # TEXT: std::numeric_limits<unsigned _RWSTD_LONG_LONG>::traps == 0, got true
> # CLAUSE: numeric.special
> # LINE: 587
> # ASSERTION (S7) (4 lines):
> # TEXT: std::numeric_limits<wchar_t>::traps == 0, got true
> # CLAUSE: numeric.special
> # LINE: 587
> {noformat}
> Problem is in that the test unconditionally expects that
> numeric_limits<>::traps is false for all integral types. Perhaps we need to
> hardcode the expected results depending on the platform (as done in
> 18.numeric.special.float test).
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
