[Bug math/18217] New: erfcf spurious underflows

[Bug math/18217] New: erfcf spurious underflows

[jsm28 at gcc dot gnu.org]

https://sourceware.org/bugzilla/show_bug.cgi?id=18217

            Bug ID: 18217
           Summary: erfcf spurious underflows
           Product: glibc
           Version: 2.21
            Status: NEW
          Severity: normal
          Priority: P2
         Component: math
          Assignee: unassigned at sourceware dot org
          Reporter: jsm28 at gcc dot gnu.org

The flt-32 implementation of erfcf produces spurious underflow exceptions.  For
example, on x86_64, erfcf (0x1p-55f) produces an underflow exception (while the
return value is correctly very close to 1).

-- 
You are receiving this mail because:
You are on the CC list for the bug.

[Bug math/18217] erfcf spurious underflows

[cvs-commit at gcc dot gnu.org]

https://sourceware.org/bugzilla/show_bug.cgi?id=18217

--- Comment #1 from cvs-commit at gcc dot gnu.org <cvs-commit at gcc dot gnu.org> ---
This is an automated email from the git hooks/post-receive script. It was
generated because a ref change was pushed to the repository containing
the project "GNU C Library master sources".

The branch, master has been updated
       via  fbc68f03b0e56db6a07d789fd403eb79d726fe96 (commit)
      from  9a71f1fcf53615c00b5f9e5da4bba92bccb0efb4 (commit)

Those revisions listed above that are new to this repository have
not appeared on any other notification email; so we list those
revisions in full, below.

- Log -----------------------------------------------------------------
https://sourceware.org/git/gitweb.cgi?p=glibc.git;h=fbc68f03b0e56db6a07d789fd403eb79d726fe96

commit fbc68f03b0e56db6a07d789fd403eb79d726fe96
Author: Joseph Myers <joseph@codesourcery.com>
Date:   Fri May 15 00:16:10 2015 +0000

    Fix erfcf spurious underflows (bug 18217).

    The flt-32 implementation of erfcf produces spurious underflow
    exceptions for some arguments close to 0, because of calculations
    squaring the argument and then multiplying by small constants.  This
    patch fixes this by adjusting the threshold for arguments for which
    the result is so close to 1 that 1 - x will give the right result from
    2**-56 to 2**-26.  (If 1 - x * 2/sqrt(pi) were used, the errors would be
    on the order of x^3 and a much larger threshold could be used.)

    Tested for x86_64 and x86.

        [BZ #18217]
        * sysdeps/ieee754/flt-32/s_erff.c (__erfcf): Use 2**-26 not 2**-56
        as threshold for returning 1 - x.
        * math/auto-libm-test-in: Add more tests of erfc.
        * math/auto-libm-test-out: Regenerated.

-----------------------------------------------------------------------

Summary of changes:
 ChangeLog                       |    8 ++++++
 NEWS                            |    4 +-
 math/auto-libm-test-in          |    2 +
 math/auto-libm-test-out         |   50 +++++++++++++++++++++++++++++++++++++++
 sysdeps/ieee754/flt-32/s_erff.c |    2 +-
 5 files changed, 63 insertions(+), 3 deletions(-)

-- 
You are receiving this mail because:
You are on the CC list for the bug.

[Bug math/18217] erfcf spurious underflows

[jsm28 at gcc dot gnu.org]

https://sourceware.org/bugzilla/show_bug.cgi?id=18217

Joseph Myers <jsm28 at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |RESOLVED
         Resolution|---                         |FIXED

--- Comment #2 from Joseph Myers <jsm28 at gcc dot gnu.org> ---
Fixed for 2.22.

-- 
You are receiving this mail because:
You are on the CC list for the bug.