mesos-reviews mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Benjamin Bannier" <benjamin.bann...@mesosphere.io>
Subject Review Request 41872: Used only exactly representable floats in test calculation.
Date Mon, 04 Jan 2016 09:23:05 GMT

-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/41872/
-----------------------------------------------------------

Review request for mesos, Alexander Rojas and Michael Park.


Bugs: MESOS-4272
    https://issues.apache.org/jira/browse/MESOS-4272


Repository: mesos


Description
-------

yIn the IEEE floating point standard representation usually only values
where the fractional part of the normalized form is a multiple of 2^n
can be represented exactly; a value of 0.33 cannot be represented
exactly.  This can lead to an incorrectly failing tests where the
calculated value does not match the expectation.

Instead switch to a value of 0.25 which can be represented exactly.


Diffs
-----

  3rdparty/libprocess/3rdparty/stout/tests/duration_tests.cpp 0aafc8643cf6afcb58befa6c1e7d427ac529be03


Diff: https://reviews.apache.org/r/41872/diff/


Testing
-------


Thanks,

Benjamin Bannier


Mime
  • Unnamed multipart/alternative (inline, None, 0 bytes)
View raw message