SYS-OSS/FRET-qemu
4
0
Fork 0
Code Issues Pull Requests Projects Releases Wiki Activity

target/arm: Use float_round_to_odd in helper_fcvtx_f64_to_f32

Browse Source 
Softfloat has native support for round-to-odd.  Use it.

Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
Message-id: 20241206031428.78634-1-richard.henderson@linaro.org
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
This commit is contained in:
Richard Henderson 2024-12-05 21:14:28 -06:00 committed by Peter Maydell
parent 53ac794af6
commit c66be53ce7
1 changed files with 4 additions and 14 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

18
target/arm/tcg/helper-a64.c
View File

@ -408,23 +408,13 @@ float64 HELPER(frecpx_f64)(float64 a, void *fpstp)
float32 HELPER(fcvtx_f64_to_f32)(float64 a, CPUARMState *env) float32 HELPER(fcvtx_f64_to_f32)(float64 a, CPUARMState *env)
{ {
/* Von Neumann rounding is implemented by using round-to-zero
* and then setting the LSB of the result if Inexact was raised.
*/
float32 r; float32 r;
float_status *fpst = &env->vfp.fp_status; float_status *fpst = &env->vfp.fp_status;
float_status tstat = *fpst; int old = get_float_rounding_mode(fpst);
int exflags;
set_float_rounding_mode(float_round_to_zero, &tstat); set_float_rounding_mode(float_round_to_odd, fpst);
set_float_exception_flags(0, &tstat); r = float64_to_float32(a, fpst);
r = float64_to_float32(a, &tstat); set_float_rounding_mode(old, fpst);
exflags = get_float_exception_flags(&tstat);
if (exflags & float_flag_inexact) {
r = make_float32(float32_val(r) | 1);
}
exflags |= get_float_exception_flags(fpst);
set_float_exception_flags(exflags, fpst);
return r; return r;
} }