| Message ID | 1296497206-15643-5-git-send-email-christophe.lyon@st.com |
|---|---|
| State | New |
| Headers | show |
On 31 January 2011 18:06, <christophe.lyon@st.com> wrote: > From: Christophe Lyon <christophe.lyon@st.com> > > Tweak decoding of the shift-by-imm and narrow 64 bit insns > (VSHRN, VRSHRN, VQSHRN, VQSHRUN, VQRSHRN, VQRSHRUN). > > Signed-off-by: Christophe Lyon <christophe.lyon@st.com> > --- > target-arm/translate.c | 28 ++++++++++++++++++---------- > 1 files changed, 18 insertions(+), 10 deletions(-) > > diff --git a/target-arm/translate.c b/target-arm/translate.c > index 9ca5b82..a614e34 100644 > --- a/target-arm/translate.c > +++ b/target-arm/translate.c > @@ -4831,21 +4831,29 @@ static int disas_neon_data_insn(CPUState * env, DisasContext *s, uint32_t insn) > if (size == 3) { > neon_load_reg64(cpu_V0, rm + pass); > if (q) { > - if (u) > - gen_helper_neon_rshl_u64(cpu_V0, cpu_V0, tmp64); > - else > - gen_helper_neon_rshl_s64(cpu_V0, cpu_V0, tmp64); > + if ((op == 8 && !u) || (op == 9 && u)) { This patch would improved if you started with: int input_unsigned = (op == 8) ? !u : u; and then used that here and: > + if ((op == 8 && !u) || (op == 9 && u)) { here... > + gen_neon_shift_narrow(size, tmp, tmp2, q, > + (op == 8) ? !u : u); here... > + gen_neon_shift_narrow(size, tmp3, tmp2, q, > + (op == 8) ? !u : u); and here. (What all these things are trying to check is whether the input elements for the operations are signed or unsigned; this isn't the same as the u bit, which is whether the output is unsigned.) -- PMM
diff --git a/target-arm/translate.c b/target-arm/translate.c index 9ca5b82..a614e34 100644 --- a/target-arm/translate.c +++ b/target-arm/translate.c @@ -4831,21 +4831,29 @@ static int disas_neon_data_insn(CPUState * env, DisasContext *s, uint32_t insn) if (size == 3) { neon_load_reg64(cpu_V0, rm + pass); if (q) { - if (u) - gen_helper_neon_rshl_u64(cpu_V0, cpu_V0, tmp64); - else - gen_helper_neon_rshl_s64(cpu_V0, cpu_V0, tmp64); + if ((op == 8 && !u) || (op == 9 && u)) { + gen_helper_neon_rshl_u64(cpu_V0, cpu_V0, + tmp64); + } else { + gen_helper_neon_rshl_s64(cpu_V0, cpu_V0, + tmp64); + } } else { - if (u) - gen_helper_neon_shl_u64(cpu_V0, cpu_V0, tmp64); - else - gen_helper_neon_shl_s64(cpu_V0, cpu_V0, tmp64); + if ((op == 8 && !u) || (op == 9 && u)) { + gen_helper_neon_shl_u64(cpu_V0, cpu_V0, + tmp64); + } else { + gen_helper_neon_shl_s64(cpu_V0, cpu_V0, + tmp64); + } } } else { tmp = neon_load_reg(rm + pass, 0); - gen_neon_shift_narrow(size, tmp, tmp2, q, u); + gen_neon_shift_narrow(size, tmp, tmp2, q, + (op == 8) ? !u : u); tmp3 = neon_load_reg(rm + pass, 1); - gen_neon_shift_narrow(size, tmp3, tmp2, q, u); + gen_neon_shift_narrow(size, tmp3, tmp2, q, + (op == 8) ? !u : u); tcg_gen_concat_i32_i64(cpu_V0, tmp, tmp3); dead_tmp(tmp); dead_tmp(tmp3);