diff options
Diffstat (limited to 'gcc/config/i386')
| -rw-r--r-- | gcc/config/i386/i386-expand.cc | 9 | ||||
| -rw-r--r-- | gcc/config/i386/i386.cc | 15 | ||||
| -rw-r--r-- | gcc/config/i386/sse.md | 3 |
3 files changed, 23 insertions, 4 deletions
diff --git a/gcc/config/i386/i386-expand.cc b/gcc/config/i386/i386-expand.cc index fd9bcaa..438fa4e 100644 --- a/gcc/config/i386/i386-expand.cc +++ b/gcc/config/i386/i386-expand.cc @@ -4159,12 +4159,18 @@ static bool ix86_expand_sse_fp_minmax (rtx dest, enum rtx_code code, rtx cmp_op0, rtx cmp_op1, rtx if_true, rtx if_false) { - machine_mode mode; + machine_mode mode = GET_MODE (dest); bool is_min; rtx tmp; if (code == LT) ; + else if (code == LE && !HONOR_NANS (mode)) + { + /* We can swap LE to GE and then invert to LT. */ + std::swap (cmp_op0, cmp_op1); + std::swap (if_true, if_false); + } else if (code == UNGE) std::swap (if_true, if_false); else @@ -4177,7 +4183,6 @@ ix86_expand_sse_fp_minmax (rtx dest, enum rtx_code code, rtx cmp_op0, else return false; - mode = GET_MODE (dest); if (immediate_operand (if_false, mode)) if_false = force_reg (mode, if_false); if (immediate_operand (if_true, mode)) diff --git a/gcc/config/i386/i386.cc b/gcc/config/i386/i386.cc index db43045..75a9cb6 100644 --- a/gcc/config/i386/i386.cc +++ b/gcc/config/i386/i386.cc @@ -26397,7 +26397,20 @@ ix86_vector_costs::add_stmt_cost (int count, vect_cost_for_stmt kind, (TREE_OPERAND (gimple_assign_rhs1 (def), 0)))))) { if (fp) - m_num_sse_needed[where]++; + { + /* Scalar FP values residing in x87 registers need to be + spilled and reloaded. */ + auto mode2 = TYPE_MODE (TREE_TYPE (op)); + if (IS_STACK_MODE (mode2)) + { + int cost + = (ix86_cost->hard_register.fp_store[mode2 == SFmode + ? 0 : 1] + + ix86_cost->sse_load[sse_store_index (mode2)]); + stmt_cost += COSTS_N_INSNS (cost) / 2; + } + m_num_sse_needed[where]++; + } else { m_num_gpr_needed[where]++; diff --git a/gcc/config/i386/sse.md b/gcc/config/i386/sse.md index 0be898c..fb79b2e 100644 --- a/gcc/config/i386/sse.md +++ b/gcc/config/i386/sse.md @@ -4900,7 +4900,8 @@ (match_operand:SI 3 "<cmp_imm_predicate>")] UNSPEC_PCMP_ITER))] "TARGET_AVX512F && ix86_pre_reload_split () - && rtx_equal_p (operands[1], operands[2])" + && rtx_equal_p (operands[1], operands[2]) + && (!MEM_P (operands[1]) || !MEM_VOLATILE_P (operands[1]))" "#" "&& 1" [(set (match_dup 0) (match_dup 4))] |