diff options
author | Richard Henderson <richard.henderson@linaro.org> | 2025-02-01 16:39:54 +0000 |
---|---|---|
committer | Peter Maydell <peter.maydell@linaro.org> | 2025-02-11 16:22:07 +0000 |
commit | a66c4585fff70ffc4a61e0f5f5528320a55cd9cd (patch) | |
tree | 8c170182fb82cf4d6236055d5e71ee1e4e69eeba | |
parent | f67a16e7d754cde9a55dd1e26ea4db48701b6fb9 (diff) | |
download | qemu-a66c4585fff70ffc4a61e0f5f5528320a55cd9cd.zip qemu-a66c4585fff70ffc4a61e0f5f5528320a55cd9cd.tar.gz qemu-a66c4585fff70ffc4a61e0f5f5528320a55cd9cd.tar.bz2 |
target/arm: Handle FPCR.AH in SVE FMLSLB, FMLSLT (vectors)
Handle FPCR.AH's requirement to not negate the sign of a NaN in SVE
FMLSL (indexed), using the usual trick of negating by XOR when AH=0
and by muladd flags when AH=1.
Since we have the CPUARMState* in the helper anyway, we can
look directly at env->vfp.fpcr and don't need toa pass in the
FPCR.AH value via the SIMD data word.
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
Message-id: 20250129013857.135256-33-richard.henderson@linaro.org
[PMM: tweaked commit message]
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
-rw-r--r-- | target/arm/tcg/vec_helper.c | 15 |
1 files changed, 12 insertions, 3 deletions
diff --git a/target/arm/tcg/vec_helper.c b/target/arm/tcg/vec_helper.c index 42bb43ac..aefcd07 100644 --- a/target/arm/tcg/vec_helper.c +++ b/target/arm/tcg/vec_helper.c @@ -2178,19 +2178,28 @@ void HELPER(sve2_fmlal_zzzw_s)(void *vd, void *vn, void *vm, void *va, CPUARMState *env, uint32_t desc) { intptr_t i, oprsz = simd_oprsz(desc); - uint16_t negn = extract32(desc, SIMD_DATA_SHIFT, 1) << 15; + bool is_s = extract32(desc, SIMD_DATA_SHIFT, 1); intptr_t sel = extract32(desc, SIMD_DATA_SHIFT + 1, 1) * sizeof(float16); float_status *status = &env->vfp.fp_status_a64; bool fz16 = get_flush_inputs_to_zero(&env->vfp.fp_status_f16_a64); + int negx = 0, negf = 0; + + if (is_s) { + if (env->vfp.fpcr & FPCR_AH) { + negf = float_muladd_negate_product; + } else { + negx = 0x8000; + } + } for (i = 0; i < oprsz; i += sizeof(float32)) { - float16 nn_16 = *(float16 *)(vn + H1_2(i + sel)) ^ negn; + float16 nn_16 = *(float16 *)(vn + H1_2(i + sel)) ^ negx; float16 mm_16 = *(float16 *)(vm + H1_2(i + sel)); float32 nn = float16_to_float32_by_bits(nn_16, fz16); float32 mm = float16_to_float32_by_bits(mm_16, fz16); float32 aa = *(float32 *)(va + H1_4(i)); - *(float32 *)(vd + H1_4(i)) = float32_muladd(nn, mm, aa, 0, status); + *(float32 *)(vd + H1_4(i)) = float32_muladd(nn, mm, aa, negf, status); } } |