diff options
author | Paul E. Murphy <murphyp@linux.vnet.ibm.com> | 2016-08-25 11:25:33 -0500 |
---|---|---|
committer | Paul E. Murphy <murphyp@linux.vnet.ibm.com> | 2016-09-01 09:28:05 -0500 |
commit | 7b7c39450b3c4ab35b4960346e61d7b177ee728e (patch) | |
tree | 251612e6e9a4192b1e5ca8010589ec64a692b9fa /ChangeLog | |
parent | 4d728087ef8cc826b05bd21d0c74d4eca9b1a27d (diff) | |
download | glibc-7b7c39450b3c4ab35b4960346e61d7b177ee728e.tar.gz glibc-7b7c39450b3c4ab35b4960346e61d7b177ee728e.tar.xz glibc-7b7c39450b3c4ab35b4960346e61d7b177ee728e.zip |
Make common fdim implementation generic.
The only difference is the usage of math_narrow_eval when building s_fdiml.c. This should be harmless for long double, but I did observe some code generation changes on m68k, but lack the resources to test it. Likewise, to more easily support overriding symbol generation, the aliasing macros are always conditionally defined on their absence to reduce boilerplate. I also ran builds for i486, ppc64, sparcv9, aarch64, s390x and observed no changes to s_fdim* objects.
Diffstat (limited to 'ChangeLog')
-rw-r--r-- | ChangeLog | 34 |
1 files changed, 34 insertions, 0 deletions
diff --git a/ChangeLog b/ChangeLog index 7c9ecd505b..8366d8538c 100644 --- a/ChangeLog +++ b/ChangeLog @@ -1,3 +1,37 @@ +2016-09-01 Paul E. Murphy <murphyp@linux.vnet.ibm.com> + + * math/Makefile (gen-libm-calls): Add s_fdim. + (libm-calls): Move to above. + + * math/s_fdim.c: Renamed and refactored into to ... + * math/s_fdim_template.c: New file. + * math/s_fdiml.c: Removed. + * sysdeps/ieee754/ldbl-opt/s_fdim.c: Removed. + * sysdeps/ieee754/ldbl-opt/s_fdiml.c: Removed. + + * sysdeps/generic/math-type-macros-double.h: + (declare_mgen_alias): Don't declare if already declared. + * sysdeps/ieee754/ldbl-opt/math-type-macros-ldouble.h: + Likewise. + + * sysdeps/ieee754/ldbl-opt/math-type-macros-double.h: + (declare_mgen_alias): Don't declare if already declared. + (M_LIBM_NEED_COMPAT): Likewise. + (declare_mgen_libm_compat): Likewise. + (LDOUBLE_fdiml_libm_version): New macro. + + * sysdeps/sparc/sparc32/sparcv9/fpu/multiarch/s_fdim.c: + Include math/s_fdim.c now that ldbl-opt version is not needed. + (declare_mgen_alias): New macro to disable aliasing. + + * sysdeps/sparc/sparc32/sparcv9/fpu/multiarch/s_fdimf-vis3.c: + (declare_mgen_alias): Likewise. + * sysdeps/sparc/sparc32/sparcv9/fpu/multiarch/s_fdimf.c: + (declare_mgen_alias): Likewise. + * sysdeps/sparc/sparc32/sparcv9/fpu/multiarch/s_fdim-vis3.c: + (declare_mgen_alias): Likewise. + (M_LIBM_NEED_COMPAT): Likewise. + 2016-09-01 Florian Weimer <fweimer@redhat.com> [BZ #20525] |