[Bug math/18790] New: [ldbl-128ibm] tanhl inaccurate

[Bug math/18790] New: [ldbl-128ibm] tanhl inaccurate

[jsm28 at gcc dot gnu.org]

https://sourceware.org/bugzilla/show_bug.cgi?id=18790

            Bug ID: 18790
           Summary: [ldbl-128ibm] tanhl inaccurate
           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
  Target Milestone: ---

ldbl-128ibm tanhl uses a too-small threshold to decide when to return +/-1,
resulting in large errors, e.g.:

Failure: Test: tanh (0x1.6p+4)
Result:
 is:          1.00000000000000000000e+00   0x1.00000000000000000000p+0
 should be:   9.99999999999999999844e-01   0x1.fffffffffffffffa4230p-1
 difference:  1.55622644822670250709e-19   0x1.6f741de1748000000000p-63
 ulp       :  12625608903588.0000
 max.ulp   :  1.0000

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

[Bug math/18790] [ldbl-128ibm] tanhl inaccurate

[cvs-commit at gcc dot gnu.org]

https://sourceware.org/bugzilla/show_bug.cgi?id=18790

--- 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  7ee06ef15827f8f2582a94b1ceaad47eaf502744 (commit)
      from  eb32b0d40308166c4d8f6330cc2958cb1e545075 (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=7ee06ef15827f8f2582a94b1ceaad47eaf502744

commit 7ee06ef15827f8f2582a94b1ceaad47eaf502744
Author: Joseph Myers <joseph@codesourcery.com>
Date:   Mon Aug 10 20:35:30 2015 +0000

    Fix ldbl-128ibm tanhl inaccuracy (bug 18790).

    ldbl-128ibm tanhl uses a too-small threshold to decide when to return
    +/-1, resulting in large errors.  This patch changes it to a more
    appropriate threshold (the requirement is for 2*exp(-2|x|) to be small
    in terms of ulps of 1).

    Tested for x86_64, x86 and powerpc.

        [BZ #18790]
        * sysdeps/ieee754/ldbl-128ibm/s_tanhl.c (__tanhl): Increase
        threshold for returning +/- 1.
        * math/auto-libm-test-in: Add more tests of tanh.
        * math/auto-libm-test-out: Regenerated.
        * sysdeps/i386/fpu/libm-test-ulps: Update.

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

Summary of changes:
 ChangeLog                             |    9 +
 NEWS                                  |    2 +-
 math/auto-libm-test-in                |   36 ++
 math/auto-libm-test-out               |  900 +++++++++++++++++++++++++++++++++
 sysdeps/i386/fpu/libm-test-ulps       |    2 +-
 sysdeps/ieee754/ldbl-128ibm/s_tanhl.c |   10 +-
 6 files changed, 952 insertions(+), 7 deletions(-)

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

[Bug math/18790] [ldbl-128ibm] tanhl inaccurate

[jsm28 at gcc dot gnu.org]

https://sourceware.org/bugzilla/show_bug.cgi?id=18790

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

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |RESOLVED
         Resolution|---                         |FIXED
   Target Milestone|---                         |2.23

--- Comment #2 from Joseph Myers <jsm28 at gcc dot gnu.org> ---
Fixed for 2.23.

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