diff options
| author | Joe Ramsay <Joe.Ramsay@arm.com> | 2024-11-01 15:48:54 +0000 |
|---|---|---|
| committer | Wilco Dijkstra <wilco.dijkstra@arm.com> | 2024-11-01 16:10:41 +0000 |
| commit | 2d82d781a539ce8e82178fc1fa2c99ae1884e7fe (patch) | |
| tree | f12a3d4fae269805d922457e80a46873cfd60ee2 /sysdeps/aarch64/fpu/erfcf_data.c | |
| parent | 6d477b8de8c3a84a061a78c05e5300143b297850 (diff) | |
| download | glibc-2d82d781a539ce8e82178fc1fa2c99ae1884e7fe.tar.xz glibc-2d82d781a539ce8e82178fc1fa2c99ae1884e7fe.zip | |
AArch64: Remove SVE erf and erfc tables
By using a combination of mask-and-add instead of the shift-based
index calculation the routines can share the same table as other
variants with no performance degradation.
The tables change name because of other changes in downstream AOR.
Reviewed-by: Wilco Dijkstra <Wilco.Dijkstra@arm.com>
Diffstat (limited to 'sysdeps/aarch64/fpu/erfcf_data.c')
| -rw-r--r-- | sysdeps/aarch64/fpu/erfcf_data.c | 8 |
1 files changed, 4 insertions, 4 deletions
diff --git a/sysdeps/aarch64/fpu/erfcf_data.c b/sysdeps/aarch64/fpu/erfcf_data.c index 77fb889a78..d45087bbb9 100644 --- a/sysdeps/aarch64/fpu/erfcf_data.c +++ b/sysdeps/aarch64/fpu/erfcf_data.c @@ -19,14 +19,14 @@ #include "vecmath_config.h" -/* Lookup table used in erfcf. +/* Lookup table used in vector erfcf. For each possible rounded input r (multiples of 1/64), between r = 0.0 and r = 10.0625 (645 values): - - the first entry __erfcf_data.tab.erfc contains the values of erfc(r), - - the second entry __erfcf_data.tab.scale contains the values of + - the first entry __v_erfcf_data.tab.erfc contains the values of erfc(r), + - the second entry __v_erfcf_data.tab.scale contains the values of 2/sqrt(pi)*exp(-r^2). Both values may go into subnormal range, therefore they are scaled by a large enough value 2^47 (fits in 8 bits). */ -const struct erfcf_data __erfcf_data = { +const struct v_erfcf_data __v_erfcf_data = { .tab = { { 0x1p47, 0x1.20dd76p47 }, { 0x1.f6f944p46, 0x1.20cb68p47 }, { 0x1.edf3aap46, 0x1.209546p47 }, |