diff options
author | Joseph Myers <joseph@codesourcery.com> | 2017-12-06 13:42:58 +0000 |
---|---|---|
committer | Joseph Myers <joseph@codesourcery.com> | 2017-12-06 13:42:58 +0000 |
commit | 53994f12633b863eb833fc3b1f6971ae8de61ff8 (patch) | |
tree | 7a4dc01f70bf0c49d5c4daf83b24521f7fe1244c /sysdeps/ieee754/ldbl-128/e_j0l.c | |
parent | 62ea2193ee4b538b13da1c579113761e0b92376c (diff) | |
download | glibc-53994f12633b863eb833fc3b1f6971ae8de61ff8.tar.gz glibc-53994f12633b863eb833fc3b1f6971ae8de61ff8.tar.xz glibc-53994f12633b863eb833fc3b1f6971ae8de61ff8.zip |
Make some ldbl-128, ldbl-128ibm arrays const.
I noticed that an x86_64 build of libm unexpectedly contained more non-constant data than an older version (before _Float128 support) did. The problem is non-const arrays in the ldbl-128 j0l and j1l implementations; this patch makes those arrays, and the corresponding ldbl-128ibm ones, const. Tested for x86_64, and tested compilation for powerpc with build-many-glibcs.py. * sysdeps/ieee754/ldbl-128/e_j0l.c (Y0_2N): Make const. (Y0_2D): Likewise. * sysdeps/ieee754/ldbl-128/e_j1l.c (Y0_2N): Likewise. (Y0_2D): Likewise. * sysdeps/ieee754/ldbl-128ibm/e_j0l.c (Y0_2N): Likewise. (Y0_2D): Likewise. * sysdeps/ieee754/ldbl-128ibm/e_j1l.c (Y0_2N): Likewise. (Y0_2D): Likewise.
Diffstat (limited to 'sysdeps/ieee754/ldbl-128/e_j0l.c')
-rw-r--r-- | sysdeps/ieee754/ldbl-128/e_j0l.c | 4 |
1 files changed, 2 insertions, 2 deletions
diff --git a/sysdeps/ieee754/ldbl-128/e_j0l.c b/sysdeps/ieee754/ldbl-128/e_j0l.c index fb8d3518ce..9190e27c2a 100644 --- a/sysdeps/ieee754/ldbl-128/e_j0l.c +++ b/sysdeps/ieee754/ldbl-128/e_j0l.c @@ -796,7 +796,7 @@ strong_alias (__ieee754_j0l, __j0l_finite) Peak absolute error 1.7e-36 (relative where Y0 > 1) 0 <= x <= 2 */ #define NY0_2N 7 -static _Float128 Y0_2N[NY0_2N + 1] = { +static const _Float128 Y0_2N[NY0_2N + 1] = { L(-1.062023609591350692692296993537002558155E19), L(2.542000883190248639104127452714966858866E19), L(-1.984190771278515324281415820316054696545E18), @@ -807,7 +807,7 @@ static _Float128 Y0_2N[NY0_2N + 1] = { L(8.230845651379566339707130644134372793322E6), }; #define NY0_2D 7 -static _Float128 Y0_2D[NY0_2D + 1] = { +static const _Float128 Y0_2D[NY0_2D + 1] = { L(1.438972634353286978700329883122253752192E20), L(1.856409101981569254247700169486907405500E18), L(1.219693352678218589553725579802986255614E16), |