Repository : ssh://darcs.haskell.org//srv/darcs/testsuite

On branch  : master

http://hackage.haskell.org/trac/ghc/changeset/e2ff6c505b178671f51f5d27cf57d4bb5f11ba16

>---------------------------------------------------------------

commit e2ff6c505b178671f51f5d27cf57d4bb5f11ba16
Author: Simon Marlow <[email protected]>
Date:   Fri Apr 1 12:19:35 2011 +0100

    avoid undefined behaviour in arith005
    
    Sometime earlier this month we started producing different results for
    some of the tests in arith005 on x86.  Closer inspection revealed that
    the differences where in operations like
    
      ceiling (1.18088e+11) :: Int
    
    which is not representable with a 32-bit Int.  In the case of truncate
    we decided that the result should be undefined (see #1254) and avoided
    testing the undefined results, but we never did the same for ceiling,
    floor, round, or properFraction.  This patch fixes that.

 tests/ghc-regress/numeric/should_run/arith005.hs   |   20 ++++++++++----------
 .../ghc-regress/numeric/should_run/arith005.stdout |   20 ++++++++++----------
 .../numeric/should_run/arith005.stdout-ws-64       |   20 ++++++++++----------
 3 files changed, 30 insertions(+), 30 deletions(-)


Diff suppressed because of size. To see it, use:

    git show e2ff6c505b178671f51f5d27cf57d4bb5f11ba16

_______________________________________________
Cvs-ghc mailing list
[email protected]
http://www.haskell.org/mailman/listinfo/cvs-ghc

Reply via email to