This adds support for the SME parts of arm_sme.h.
gcc/
* doc/invoke.texi: Document +sme-i16i64 and +sme-f64f64.
* config.gcc (aarch64*-*-*): Add arm_sme.h to the list of headers
to install and aarch64-sve-builtins-sme.o to the list of objects
to build.
* config/aarch64/aarch64-c.cc (aarch64_update_cpp_builtins): Define
or undefine TARGET_SME, TARGET_SME_I16I64 and TARGET_SME_F64F64.
(aarch64_pragma_aarch64): Handle arm_sme.h.
* config/aarch64/aarch64-option-extensions.def (sme-i16i64)
(sme-f64f64): New extensions.
* config/aarch64/aarch64-protos.h (aarch64_sme_vq_immediate)
(aarch64_addsvl_addspl_immediate_p, aarch64_output_addsvl_addspl)
(aarch64_output_sme_zero_za): Declare.
(aarch64_output_move_struct): Delete.
(aarch64_sme_ldr_vnum_offset): Declare.
(aarch64_sve::handle_arm_sme_h): Likewise.
* config/aarch64/aarch64.h (AARCH64_ISA_SM_ON): New macro.
(AARCH64_ISA_SME_I16I64, AARCH64_ISA_SME_F64F64): Likewise.
(TARGET_STREAMING, TARGET_STREAMING_SME): Likewise.
(TARGET_SME_I16I64, TARGET_SME_F64F64): Likewise.
* config/aarch64/aarch64.cc (aarch64_sve_rdvl_factor_p): Rename to...
(aarch64_sve_rdvl_addvl_factor_p): ...this.
(aarch64_sve_rdvl_immediate_p): Update accordingly.
(aarch64_rdsvl_immediate_p, aarch64_add_offset): Likewise.
(aarch64_sme_vq_immediate): Likewise. Make public.
(aarch64_sve_addpl_factor_p): New function.
(aarch64_sve_addvl_addpl_immediate_p): Use
aarch64_sve_rdvl_addvl_factor_p and aarch64_sve_addpl_factor_p.
(aarch64_addsvl_addspl_immediate_p): New function.
(aarch64_output_addsvl_addspl): Likewise.
(aarch64_cannot_force_const_mem): Return true for RDSVL immediates.
(aarch64_classify_index): Handle .Q scaling for VNx1TImode.
(aarch64_classify_address): Likewise for vnum offsets.
(aarch64_output_sme_zero_za): New function.
(aarch64_sme_ldr_vnum_offset_p): Likewise.
* config/aarch64/predicates.md (aarch64_addsvl_addspl_immediate):
New predicate.
(aarch64_pluslong_operand): Include it for SME.
* config/aarch64/constraints.md (Ucj, Uav): New constraints.
* config/aarch64/iterators.md (VNx1TI_ONLY): New mode iterator.
(SME_ZA_I, SME_ZA_SDI, SME_ZA_SDF_I, SME_MOP_BHI): Likewise.
(SME_MOP_HSDF): Likewise.
(UNSPEC_SME_ADDHA, UNSPEC_SME_ADDVA, UNSPEC_SME_FMOPA)
(UNSPEC_SME_FMOPS, UNSPEC_SME_LD1_HOR, UNSPEC_SME_LD1_VER)
(UNSPEC_SME_READ_HOR, UNSPEC_SME_READ_VER, UNSPEC_SME_SMOPA)
(UNSPEC_SME_SMOPS, UNSPEC_SME_ST1_HOR, UNSPEC_SME_ST1_VER)
(UNSPEC_SME_SUMOPA, UNSPEC_SME_SUMOPS, UNSPEC_SME_UMOPA)
(UNSPEC_SME_UMOPS, UNSPEC_SME_USMOPA, UNSPEC_SME_USMOPS)
(UNSPEC_SME_WRITE_HOR, UNSPEC_SME_WRITE_VER): New unspecs.
(elem_bits): Handle x2 and x4 structure modes, plus VNx1TI.
(Vetype, Vesize, VPRED): Handle VNx1TI.
(b): New mode attribute.
(SME_LD1, SME_READ, SME_ST1, SME_WRITE, SME_BINARY_SDI, SME_INT_MOP)
(SME_FP_MOP): New int iterators.
(optab): Handle SME unspecs.
(hv): New int attribute.
* config/aarch64/aarch64.md (*add<mode>3_aarch64): Handle ADDSVL
and ADDSPL.
* config/aarch64/aarch64-sme.md (UNSPEC_SME_LDR): New unspec.
(@aarch64_sme_<optab><mode>, @aarch64_sme_<optab><mode>_plus)
(aarch64_sme_ldr0, @aarch64_sme_ldrn<mode>): New patterns.
(UNSPEC_SME_STR): New unspec.
(@aarch64_sme_<optab><mode>, @aarch64_sme_<optab><mode>_plus)
(aarch64_sme_str0, @aarch64_sme_strn<mode>): New patterns.
(@aarch64_sme_<optab><v_int_container><mode>): Likewise.
(*aarch64_sme_<optab><v_int_container><mode>_plus): Likewise.
(@aarch64_sme_<optab><VNx1TI_ONLY:mode><SVE_FULL:mode>): Likewise.
(@aarch64_sme_<optab><v_int_container><mode>): Likewise.
(*aarch64_sme_<optab><v_int_container><mode>_plus): Likewise.
(@aarch64_sme_<optab><VNx1TI_ONLY:mode><SVE_FULL:mode>): Likewise.
(UNSPEC_SME_ZERO): New unspec.
(aarch64_sme_zero): New pattern.
(@aarch64_sme_<SME_BINARY_SDI:optab><mode>): Likewise.
(@aarch64_sme_<SME_INT_MOP:optab><mode>): Likewise.
(@aarch64_sme_<SME_FP_MOP:optab><mode>): Likewise.
* config/aarch64/aarch64-sve-builtins.def: Add ZA type suffixes.
Include aarch64-sve-builtins-sme.def.
(DEF_SME_ZA_FUNCTION): New macro.
* config/aarch64/aarch64-sve-builtins.h (CP_READ_ZA): New call
property.
(CP_WRITE_ZA): Likewise.
(PRED_za_m): New predication type.
(type_suffix_index): Handle DEF_SME_ZA_SUFFIX.
(type_suffix_info): Add vector_p and za_p fields.
(function_instance::num_za_tiles): New member function.
(function_builder::get_attributes): Add an aarch64_feature_flags
argument.
(function_expander::get_contiguous_base): Take a base argument
number, a vnum argument number, and an argument that indicates
whether the vnum parameter is a factor of the SME vector length
or the prevailing vector length.
(function_expander::add_integer_operand): Take a poly_int64.
(sve_switcher::sve_switcher): Take a base set of flags.
(sme_switcher): New class.
(scalar_types): Add a null entry for NUM_VECTOR_TYPES.
* config/aarch64/aarch64-sve-builtins.cc: Include
aarch64-sve-builtins-sme.h.
(pred_suffixes): Add an entry for PRED_za_m.
(type_suffixes): Initialize vector_p and za_p. Handle ZA suffixes.
(TYPES_all_za, TYPES_d_za, TYPES_za_bhsd_data, TYPES_za_all_data)
(TYPES_za_s_integer, TYPES_za_d_integer, TYPES_mop_base)
(TYPES_mop_base_signed, TYPES_mop_base_unsigned, TYPES_mop_i16i64)
(TYPES_mop_i16i64_signed, TYPES_mop_i16i64_unsigned, TYPES_za): New
type suffix macros.
(preds_m, preds_za_m): New predication lists.
(function_groups): Handle DEF_SME_ZA_FUNCTION.
(scalar_types): Add an entry for NUM_VECTOR_TYPES.
(find_type_suffix_for_scalar_type): Check positively for vectors
rather than negatively for predicates.
(check_required_extensions): Handle PSTATE.SM and PSTATE.ZA
requirements.
(report_out_of_range): Handle the case where the minimum and
maximum are the same.
(function_instance::reads_global_state_p): Return true for functions
that read ZA.
(function_instance::modifies_global_state_p): Return true for functions
that write to ZA.
(sve_switcher::sve_switcher): Add a base flags argument.
(function_builder::get_name): Handle "__arm_" prefixes.
(add_attribute): Add an overload that takes a namespaces.
(add_shared_state_attribute): New function.
(function_builder::get_attributes): Take the required feature flags
as argument. Add streaming and ZA attributes where appropriate.
(function_builder::add_unique_function): Update calls accordingly.
(function_resolver::check_gp_argument): Assert that the predication
isn't ZA _m predication.
(function_checker::function_checker): Don't bias the argument
number for ZA _m predication.
(function_expander::get_contiguous_base): Add arguments that
specify the base argument number, the vnum argument number,
and an argument that indicates whether the vnum parameter is
a factor of the SME vector length or the prevailing vector length.
Handle the SME case.
(function_expander::add_input_operand): Handle pmode_register_operand.
(function_expander::add_integer_operand): Take a poly_int64.
(init_builtins): Call handle_arm_sme_h for LTO.
(handle_arm_sve_h): Skip SME intrinsics.
(handle_arm_sme_h): New function.
* config/aarch64/aarch64-sve-builtins-functions.h
(read_write_za, write_za): New classes.
(unspec_based_sme_function, za_arith_function): New using aliases.
(quiet_za_arith_function): Likewise.
* config/aarch64/aarch64-sve-builtins-shapes.h
(binary_za_int_m, binary_za_m, binary_za_uint_m, bool_inherent)
(inherent_za, inherent_mask_za, ldr_za, load_za, read_za_m, store_za)
(str_za, unary_za_m, write_za_m): Declare.
* config/aarch64/aarch64-sve-builtins-shapes.cc (apply_predication):
Expect za_m functions to have an existing governing predicate.
(binary_za_m_base, binary_za_int_m_def, binary_za_m_def): New classes.
(binary_za_uint_m_def, bool_inherent_def, inherent_za_def): Likewise.
(inherent_mask_za_def, ldr_za_def, load_za_def, read_za_m_def)
(store_za_def, str_za_def, unary_za_m_def, write_za_m_def): Likewise.
* config/aarch64/arm_sme.h: New file.
* config/aarch64/aarch64-sve-builtins-sme.h: Likewise.
* config/aarch64/aarch64-sve-builtins-sme.cc: Likewise.
* config/aarch64/aarch64-sve-builtins-sme.def: Likewise.
* config/aarch64/t-aarch64 (aarch64-sve-builtins.o): Depend on
aarch64-sve-builtins-sme.def and aarch64-sve-builtins-sme.h.
(aarch64-sve-builtins-sme.o): New rule.
gcc/testsuite/
* lib/target-supports.exp: Add sme and sme-i16i64 features.
* gcc.target/aarch64/pragma_cpp_predefs_4.c: Test __ARM_FEATURE_SME*
macros.
* gcc.target/aarch64/sve/acle/asm/test_sve_acle.h: Allow functions
to be marked as __arm_streaming, __arm_streaming_compatible, and
__arm_inout("za").
* g++.target/aarch64/sve/acle/general-c++/func_redef_4.c: Mark the
function as __arm_streaming_compatible.
* g++.target/aarch64/sve/acle/general-c++/func_redef_5.c: Likewise.
* g++.target/aarch64/sve/acle/general-c++/func_redef_7.c: Likewise.
* gcc.target/aarch64/sve/acle/general-c/func_redef_4.c: Likewise.
* gcc.target/aarch64/sve/acle/general-c/func_redef_5.c: Likewise.
* g++.target/aarch64/sme/aarch64-sme-acle-asm.exp: New test harness.
* gcc.target/aarch64/sme/aarch64-sme-acle-asm.exp: Likewise.
* gcc.target/aarch64/sve/acle/general-c/binary_za_int_m_1.c: New test.
* gcc.target/aarch64/sve/acle/general-c/binary_za_m_1.c: Likewise.
* gcc.target/aarch64/sve/acle/general-c/binary_za_m_2.c: Likewise.
* gcc.target/aarch64/sve/acle/general-c/binary_za_uint_m_1.c: Likewise.
* gcc.target/aarch64/sve/acle/general-c/read_za_m_1.c: Likewise.
* gcc.target/aarch64/sve/acle/general-c/unary_za_m_1.c: Likewise.
* gcc.target/aarch64/sve/acle/general-c/write_za_m_1.c: Likewise.
;;
aarch64*-*-*)
cpu_type=aarch64
- extra_headers="arm_fp16.h arm_neon.h arm_bf16.h arm_acle.h arm_sve.h"
+ extra_headers="arm_fp16.h arm_neon.h arm_bf16.h arm_acle.h arm_sve.h arm_sme.h"
c_target_objs="aarch64-c.o"
cxx_target_objs="aarch64-c.o"
d_target_objs="aarch64-d.o"
- extra_objs="aarch64-builtins.o aarch-common.o aarch64-sve-builtins.o aarch64-sve-builtins-shapes.o aarch64-sve-builtins-base.o aarch64-sve-builtins-sve2.o cortex-a57-fma-steering.o aarch64-speculation.o falkor-tag-collision-avoidance.o aarch-bti-insert.o aarch64-cc-fusion.o"
+ extra_objs="aarch64-builtins.o aarch-common.o aarch64-sve-builtins.o aarch64-sve-builtins-shapes.o aarch64-sve-builtins-base.o aarch64-sve-builtins-sve2.o aarch64-sve-builtins-sme.o cortex-a57-fma-steering.o aarch64-speculation.o falkor-tag-collision-avoidance.o aarch-bti-insert.o aarch64-cc-fusion.o"
target_gtfiles="\$(srcdir)/config/aarch64/aarch64-builtins.cc \$(srcdir)/config/aarch64/aarch64-sve-builtins.h \$(srcdir)/config/aarch64/aarch64-sve-builtins.cc"
target_has_targetm_common=yes
;;
"__ARM_FEATURE_LS64", pfile);
aarch64_def_or_undef (AARCH64_ISA_RCPC, "__ARM_FEATURE_RCPC", pfile);
+ aarch64_def_or_undef (TARGET_SME, "__ARM_FEATURE_SME", pfile);
+ aarch64_def_or_undef (TARGET_SME_I16I64, "__ARM_FEATURE_SME_I16I64", pfile);
+ aarch64_def_or_undef (TARGET_SME_F64F64, "__ARM_FEATURE_SME_F64F64", pfile);
+
/* Not for ACLE, but required to keep "float.h" correct if we switch
target between implementations that do or do not support ARMv8.2-A
16-bit floating-point extensions. */
const char *name = TREE_STRING_POINTER (x);
if (strcmp (name, "arm_sve.h") == 0)
aarch64_sve::handle_arm_sve_h ();
+ else if (strcmp (name, "arm_sme.h") == 0)
+ aarch64_sve::handle_arm_sme_h ();
else if (strcmp (name, "arm_neon.h") == 0)
handle_arm_neon_h ();
else if (strcmp (name, "arm_acle.h") == 0)
AARCH64_OPT_EXTENSION("sme", SME, (BF16, SVE2), (), (), "sme")
+AARCH64_OPT_EXTENSION("sme-i16i64", SME_I16I64, (SME), (), (), "")
+
+AARCH64_OPT_EXTENSION("sme-f64f64", SME_F64F64, (SME), (), (), "")
+
#undef AARCH64_OPT_EXTENSION
int aarch64_add_offset_temporaries (rtx);
void aarch64_split_add_offset (scalar_int_mode, rtx, rtx, rtx, rtx, rtx);
bool aarch64_rdsvl_immediate_p (const_rtx);
+rtx aarch64_sme_vq_immediate (machine_mode mode, HOST_WIDE_INT,
+ aarch64_feature_flags);
char *aarch64_output_rdsvl (const_rtx);
+bool aarch64_addsvl_addspl_immediate_p (const_rtx);
+char *aarch64_output_addsvl_addspl (rtx);
bool aarch64_mov_operand_p (rtx, machine_mode);
rtx aarch64_reverse_mask (machine_mode, unsigned int);
bool aarch64_offset_7bit_signed_scaled_p (machine_mode, poly_int64);
bool aarch64_use_return_insn_p (void);
const char *aarch64_output_casesi (rtx *);
const char *aarch64_output_load_tp (rtx);
+const char *aarch64_output_sme_zero_za (rtx);
arm_pcs aarch64_tlsdesc_abi_id ();
enum aarch64_symbol_type aarch64_classify_symbol (rtx, HOST_WIDE_INT);
int aarch64_uxt_size (int, HOST_WIDE_INT);
int aarch64_vec_fpconst_pow_of_2 (rtx);
rtx aarch64_mask_from_zextract_ops (rtx, rtx);
-const char *aarch64_output_move_struct (rtx *operands);
rtx aarch64_return_addr_rtx (void);
rtx aarch64_return_addr (int, rtx);
rtx aarch64_simd_gen_const_vector_dup (machine_mode, HOST_WIDE_INT);
bool aarch64_sve_ldr_operand_p (rtx);
bool aarch64_sve_prefetch_operand_p (rtx, machine_mode);
bool aarch64_sve_struct_memory_operand_p (rtx);
+bool aarch64_sme_ldr_vnum_offset_p (rtx, rtx);
rtx aarch64_simd_vect_par_cnst_half (machine_mode, int, bool);
rtx aarch64_gen_stepped_int_parallel (unsigned int, int, int);
bool aarch64_stepped_int_parallel_p (rtx, int);
namespace aarch64_sve {
void init_builtins ();
void handle_arm_sve_h ();
+ void handle_arm_sme_h ();
tree builtin_decl (unsigned, bool);
bool builtin_type_p (const_tree);
bool builtin_type_p (const_tree, unsigned int *, unsigned int *);
;; ---- Test current state
;; ---- PSTATE.SM management
;; ---- PSTATE.ZA management
+;;
+;; == Loads, stores and moves
+;; ---- Single-vector loads
+;; ---- Single-vector stores
+;; ---- Single-vector moves
+;; ---- Zeroing
+;;
+;; == Binary arithmetic
+;; ---- Binary arithmetic on ZA tile
+;;
+;; == Ternary arithmetic
+;; ---- [INT] Sum of outer products
+;; ---- [FP] Sum of outer products
;; =========================================================================
;; == State management
DONE;
}
)
+
+;; =========================================================================
+;; == Loads, stores and moves
+;; =========================================================================
+
+;; -------------------------------------------------------------------------
+;; ---- Single-vector loads
+;; -------------------------------------------------------------------------
+;; Includes:
+;; - LD1
+;; - LDR
+;; -------------------------------------------------------------------------
+
+(define_c_enum "unspec" [
+ UNSPEC_SME_LDR
+])
+
+(define_insn "@aarch64_sme_<optab><mode>"
+ [(set (reg:SME_ZA_I ZA_REGNUM)
+ (unspec:SME_ZA_I
+ [(reg:SME_ZA_I ZA_REGNUM)
+ (reg:DI SME_STATE_REGNUM)
+ (match_operand:DI 0 "const_int_operand")
+ (match_operand:SI 1 "register_operand" "Ucj")
+ (match_operand:<VPRED> 2 "register_operand" "Upl")
+ (match_operand:SME_ZA_I 3 "aarch64_sve_ldff1_operand" "Utf")]
+ SME_LD1))]
+ "TARGET_STREAMING_SME"
+ "ld1<Vesize>\t{ za%0<hv>.<Vetype>[%w1, 0] }, %2/z, %3"
+)
+
+(define_insn "@aarch64_sme_<optab><mode>_plus"
+ [(set (reg:SME_ZA_I ZA_REGNUM)
+ (unspec:SME_ZA_I
+ [(reg:SME_ZA_I ZA_REGNUM)
+ (reg:DI SME_STATE_REGNUM)
+ (match_operand:DI 0 "const_int_operand")
+ (plus:SI (match_operand:SI 1 "register_operand" "Ucj")
+ (match_operand:SI 2 "const_int_operand"))
+ (match_operand:<VPRED> 3 "register_operand" "Upl")
+ (match_operand:SME_ZA_I 4 "aarch64_sve_ldff1_operand" "Utf")]
+ SME_LD1))]
+ "TARGET_STREAMING_SME
+ && UINTVAL (operands[2]) < 128 / <elem_bits>"
+ "ld1<Vesize>\t{ za%0<hv>.<Vetype>[%w1, %2] }, %3/z, %4"
+)
+
+(define_insn "aarch64_sme_ldr0"
+ [(set (reg:VNx16QI ZA_REGNUM)
+ (unspec:VNx16QI
+ [(reg:VNx16QI ZA_REGNUM)
+ (reg:DI SME_STATE_REGNUM)
+ (match_operand:SI 0 "register_operand" "Ucj")
+ (mem:VNx16QI (match_operand 1 "pmode_register_operand" "rk"))]
+ UNSPEC_SME_LDR))]
+ "TARGET_SME"
+ "ldr\tza[%w0, 0], [%1, #0, mul vl]"
+)
+
+(define_insn "@aarch64_sme_ldrn<mode>"
+ [(set (reg:VNx16QI ZA_REGNUM)
+ (unspec:VNx16QI
+ [(reg:VNx16QI ZA_REGNUM)
+ (reg:DI SME_STATE_REGNUM)
+ (plus:SI (match_operand:SI 0 "register_operand" "Ucj")
+ (match_operand:SI 1 "const_int_operand"))
+ (mem:VNx16QI
+ (plus:P (match_operand:P 2 "register_operand" "rk")
+ (match_operand:P 3 "aarch64_mov_operand")))]
+ UNSPEC_SME_LDR))]
+ "TARGET_SME
+ && aarch64_sme_ldr_vnum_offset_p (operands[1], operands[3])"
+ "ldr\tza[%w0, %1], [%2, #%1, mul vl]"
+)
+
+;; -------------------------------------------------------------------------
+;; ---- Single-vector stores
+;; -------------------------------------------------------------------------
+;; Includes:
+;; - ST1
+;; - STR
+;; -------------------------------------------------------------------------
+
+(define_c_enum "unspec" [
+ UNSPEC_SME_STR
+])
+
+(define_insn "@aarch64_sme_<optab><mode>"
+ [(set (match_operand:SME_ZA_I 0 "aarch64_sve_ldff1_operand" "+Utf")
+ (unspec:SME_ZA_I
+ [(reg:SME_ZA_I ZA_REGNUM)
+ (reg:DI SME_STATE_REGNUM)
+ (match_dup 0)
+ (match_operand:DI 1 "const_int_operand")
+ (match_operand:SI 2 "register_operand" "Ucj")
+ (match_operand:<VPRED> 3 "register_operand" "Upl")]
+ SME_ST1))]
+ "TARGET_STREAMING_SME"
+ "st1<Vesize>\t{ za%1<hv>.<Vetype>[%w2, 0] }, %3, %0"
+)
+
+(define_insn "@aarch64_sme_<optab><mode>_plus"
+ [(set (match_operand:SME_ZA_I 0 "aarch64_sve_ldff1_operand" "+Utf")
+ (unspec:SME_ZA_I
+ [(reg:SME_ZA_I ZA_REGNUM)
+ (reg:DI SME_STATE_REGNUM)
+ (match_dup 0)
+ (match_operand:DI 1 "const_int_operand")
+ (plus:SI (match_operand:SI 2 "register_operand" "Ucj")
+ (match_operand:SI 3 "const_int_operand"))
+ (match_operand:<VPRED> 4 "register_operand" "Upl")]
+ SME_ST1))]
+ "TARGET_STREAMING_SME
+ && UINTVAL (operands[3]) < 128 / <elem_bits>"
+ "st1<Vesize>\t{ za%1<hv>.<Vetype>[%w2, %3] }, %4, %0"
+)
+
+(define_insn "aarch64_sme_str0"
+ [(set (mem:VNx16QI (match_operand 1 "pmode_register_operand" "rk"))
+ (unspec:VNx16QI
+ [(reg:VNx16QI ZA_REGNUM)
+ (reg:DI SME_STATE_REGNUM)
+ (mem:VNx16QI (match_dup 1))
+ (match_operand:SI 0 "register_operand" "Ucj")]
+ UNSPEC_SME_STR))]
+ "TARGET_SME"
+ "str\tza[%w0, 0], [%1, #0, mul vl]"
+)
+
+(define_insn "@aarch64_sme_strn<mode>"
+ [(set (mem:VNx16QI
+ (plus:P (match_operand:P 2 "register_operand" "rk")
+ (match_operand:P 3 "aarch64_mov_operand")))
+ (unspec:VNx16QI
+ [(reg:VNx16QI ZA_REGNUM)
+ (reg:DI SME_STATE_REGNUM)
+ (mem:VNx16QI (plus:P (match_dup 2) (match_dup 3)))
+ (plus:SI (match_operand:SI 0 "register_operand" "Ucj")
+ (match_operand:SI 1 "const_int_operand"))]
+ UNSPEC_SME_STR))]
+ "TARGET_SME
+ && aarch64_sme_ldr_vnum_offset_p (operands[1], operands[3])"
+ "str\tza[%w0, %1], [%2, #%1, mul vl]"
+)
+
+;; -------------------------------------------------------------------------
+;; ---- Single-vector moves
+;; -------------------------------------------------------------------------
+;; Includes:
+;; - MOVA
+;; -------------------------------------------------------------------------
+
+(define_insn "@aarch64_sme_<optab><v_int_container><mode>"
+ [(set (match_operand:SVE_FULL 0 "register_operand" "=w")
+ (unspec:SVE_FULL
+ [(reg:<V_INT_CONTAINER> ZA_REGNUM)
+ (reg:DI SME_STATE_REGNUM)
+ (match_operand:SVE_FULL 1 "register_operand" "0")
+ (match_operand:<VPRED> 2 "register_operand" "Upl")
+ (match_operand:DI 3 "const_int_operand")
+ (match_operand:SI 4 "register_operand" "Ucj")]
+ SME_READ))]
+ "TARGET_STREAMING_SME"
+ "mova\t%0.<Vetype>, %2/m, za%3<hv>.<Vetype>[%w4, 0]"
+)
+
+(define_insn "*aarch64_sme_<optab><v_int_container><mode>_plus"
+ [(set (match_operand:SVE_FULL 0 "register_operand" "=w")
+ (unspec:SVE_FULL
+ [(reg:<V_INT_CONTAINER> ZA_REGNUM)
+ (reg:DI SME_STATE_REGNUM)
+ (match_operand:SVE_FULL 1 "register_operand" "0")
+ (match_operand:<VPRED> 2 "register_operand" "Upl")
+ (match_operand:DI 3 "const_int_operand")
+ (plus:SI (match_operand:SI 4 "register_operand" "Ucj")
+ (match_operand:SI 5 "const_int_operand"))]
+ SME_READ))]
+ "TARGET_STREAMING_SME
+ && UINTVAL (operands[5]) < 128 / <elem_bits>"
+ "mova\t%0.<Vetype>, %2/m, za%3<hv>.<Vetype>[%w4, %5]"
+)
+
+(define_insn "@aarch64_sme_<optab><VNx1TI_ONLY:mode><SVE_FULL:mode>"
+ [(set (match_operand:SVE_FULL 0 "register_operand" "=w")
+ (unspec:SVE_FULL
+ [(reg:VNx1TI_ONLY ZA_REGNUM)
+ (reg:DI SME_STATE_REGNUM)
+ (match_operand:SVE_FULL 1 "register_operand" "0")
+ (match_operand:VNx2BI 2 "register_operand" "Upl")
+ (match_operand:DI 3 "const_int_operand")
+ (match_operand:SI 4 "register_operand" "Ucj")]
+ SME_READ))]
+ "TARGET_STREAMING_SME"
+ "mova\t%0.q, %2/m, za%3<hv>.q[%w4, 0]"
+)
+
+(define_insn "@aarch64_sme_<optab><v_int_container><mode>"
+ [(set (reg:<V_INT_CONTAINER> ZA_REGNUM)
+ (unspec:<V_INT_CONTAINER>
+ [(reg:SVE_FULL ZA_REGNUM)
+ (reg:DI SME_STATE_REGNUM)
+ (match_operand:DI 0 "const_int_operand")
+ (match_operand:SI 1 "register_operand" "Ucj")
+ (match_operand:<VPRED> 2 "register_operand" "Upl")
+ (match_operand:SVE_FULL 3 "register_operand" "w")]
+ SME_WRITE))]
+ "TARGET_STREAMING_SME"
+ "mova\tza%0<hv>.<Vetype>[%w1, 0], %2/m, %3.<Vetype>"
+)
+
+(define_insn "*aarch64_sme_<optab><v_int_container><mode>_plus"
+ [(set (reg:<V_INT_CONTAINER> ZA_REGNUM)
+ (unspec:<V_INT_CONTAINER>
+ [(reg:SVE_FULL ZA_REGNUM)
+ (reg:DI SME_STATE_REGNUM)
+ (match_operand:DI 0 "const_int_operand")
+ (plus:SI (match_operand:SI 1 "register_operand" "Ucj")
+ (match_operand:SI 2 "const_int_operand"))
+ (match_operand:<VPRED> 3 "register_operand" "Upl")
+ (match_operand:SVE_FULL 4 "register_operand" "w")]
+ SME_WRITE))]
+ "TARGET_STREAMING_SME
+ && UINTVAL (operands[2]) < 128 / <elem_bits>"
+ "mova\tza%0<hv>.<Vetype>[%w1, %2], %3/m, %4.<Vetype>"
+)
+
+(define_insn "@aarch64_sme_<optab><VNx1TI_ONLY:mode><SVE_FULL:mode>"
+ [(set (reg:VNx1TI_ONLY ZA_REGNUM)
+ (unspec:VNx1TI_ONLY
+ [(reg:VNx1TI_ONLY ZA_REGNUM)
+ (reg:DI SME_STATE_REGNUM)
+ (match_operand:DI 0 "const_int_operand")
+ (match_operand:SI 1 "register_operand" "Ucj")
+ (match_operand:VNx2BI 2 "register_operand" "Upl")
+ (match_operand:SVE_FULL 3 "register_operand" "w")]
+ SME_WRITE))]
+ "TARGET_STREAMING_SME"
+ "mova\tza%0<hv>.q[%w1, 0], %2/m, %3.q"
+)
+
+;; -------------------------------------------------------------------------
+;; ---- Zeroing
+;; -------------------------------------------------------------------------
+;; Includes:
+;; - ZERO
+;; -------------------------------------------------------------------------
+
+(define_c_enum "unspec" [UNSPEC_SME_ZERO])
+
+(define_insn "aarch64_sme_zero_za"
+ [(set (reg:VNx16QI ZA_REGNUM)
+ (unspec:VNx16QI [(reg:VNx16QI ZA_REGNUM)
+ (reg:DI SME_STATE_REGNUM)
+ (match_operand:DI 0 "const_int_operand")]
+ UNSPEC_SME_ZERO))]
+ "TARGET_SME"
+ {
+ return aarch64_output_sme_zero_za (operands[0]);
+ }
+)
+
+;; =========================================================================
+;; == Binary arithmetic
+;; =========================================================================
+
+;; -------------------------------------------------------------------------
+;; ---- Binary arithmetic on ZA tile
+;; -------------------------------------------------------------------------
+;; Includes:
+;; - ADDHA
+;; - ADDVA
+;; -------------------------------------------------------------------------
+
+(define_insn "@aarch64_sme_<optab><mode>"
+ [(set (reg:SME_ZA_SDI ZA_REGNUM)
+ (unspec:SME_ZA_SDI
+ [(reg:SME_ZA_SDI ZA_REGNUM)
+ (reg:DI SME_STATE_REGNUM)
+ (match_operand:DI 0 "const_int_operand")
+ (match_operand:<VPRED> 1 "register_operand" "Upl")
+ (match_operand:<VPRED> 2 "register_operand" "Upl")
+ (match_operand:SME_ZA_SDI 3 "register_operand" "w")]
+ SME_BINARY_SDI))]
+ "TARGET_STREAMING_SME"
+ "<optab>\tza%0.<Vetype>, %1/m, %2/m, %3.<Vetype>"
+)
+
+;; =========================================================================
+;; == Ternary arithmetic
+;; =========================================================================
+
+;; -------------------------------------------------------------------------
+;; ---- [INT] Sum of outer products
+;; -------------------------------------------------------------------------
+;; Includes:
+;; - SMOPA
+;; - SMOPS
+;; - SUMOPA
+;; - SUMOPS
+;; - UMOPA
+;; - UMOPS
+;; - USMOPA
+;; - USMOPS
+;; -------------------------------------------------------------------------
+
+(define_insn "@aarch64_sme_<optab><VNx4SI_ONLY:mode><VNx16QI_ONLY:mode>"
+ [(set (reg:VNx4SI_ONLY ZA_REGNUM)
+ (unspec:VNx4SI_ONLY
+ [(reg:VNx4SI_ONLY ZA_REGNUM)
+ (reg:DI SME_STATE_REGNUM)
+ (match_operand:DI 0 "const_int_operand")
+ (match_operand:<VNx4SI_ONLY:VPRED> 1 "register_operand" "Upl")
+ (match_operand:<VNx4SI_ONLY:VPRED> 2 "register_operand" "Upl")
+ (match_operand:VNx16QI_ONLY 3 "register_operand" "w")
+ (match_operand:VNx16QI_ONLY 4 "register_operand" "w")]
+ SME_INT_MOP))]
+ "TARGET_STREAMING_SME"
+ "<optab>\tza%0.s, %1/m, %2/m, %3.b, %4.b"
+)
+
+(define_insn "@aarch64_sme_<optab><VNx2DI_ONLY:mode><VNx8HI_ONLY:mode>"
+ [(set (reg:VNx2DI_ONLY ZA_REGNUM)
+ (unspec:VNx2DI_ONLY
+ [(reg:VNx2DI_ONLY ZA_REGNUM)
+ (reg:DI SME_STATE_REGNUM)
+ (match_operand:DI 0 "const_int_operand")
+ (match_operand:<VNx2DI_ONLY:VPRED> 1 "register_operand" "Upl")
+ (match_operand:<VNx2DI_ONLY:VPRED> 2 "register_operand" "Upl")
+ (match_operand:VNx8HI_ONLY 3 "register_operand" "w")
+ (match_operand:VNx8HI_ONLY 4 "register_operand" "w")]
+ SME_INT_MOP))]
+ "TARGET_STREAMING_SME && TARGET_SME_I16I64"
+ "<optab>\tza%0.d, %1/m, %2/m, %3.h, %4.h"
+)
+
+;; -------------------------------------------------------------------------
+;; ---- [FP] Sum of outer products
+;; -------------------------------------------------------------------------
+;; Includes:
+;; - BFMOPA
+;; - BFMOPS
+;; - FMOPA
+;; - FMOPS
+;; -------------------------------------------------------------------------
+
+(define_insn "@aarch64_sme_<optab><SME_ZA_SDF_I:mode><SME_MOP_HSDF:mode>"
+ [(set (reg:SME_ZA_SDF_I ZA_REGNUM)
+ (unspec:SME_ZA_SDF_I
+ [(reg:SME_ZA_SDF_I ZA_REGNUM)
+ (reg:DI SME_STATE_REGNUM)
+ (match_operand:DI 0 "const_int_operand")
+ (match_operand:<SME_ZA_SDF_I:VPRED> 1 "register_operand" "Upl")
+ (match_operand:<SME_ZA_SDF_I:VPRED> 2 "register_operand" "Upl")
+ (match_operand:SME_MOP_HSDF 3 "register_operand" "w")
+ (match_operand:SME_MOP_HSDF 4 "register_operand" "w")]
+ SME_FP_MOP))]
+ "TARGET_STREAMING_SME
+ && (<SME_ZA_SDF_I:elem_bits> == 32) == (<SME_MOP_HSDF:elem_bits> <= 32)"
+ "<b><optab>\tza%0.<SME_ZA_SDF_I:Vetype>, %1/m, %2/m, %3.<SME_MOP_HSDF:Vetype>, %4.<SME_MOP_HSDF:Vetype>"
+)
}
};
+/* Wrap T, which is derived from function_base, and indicate that it
+ additionally has the call properties in PROPERTIES. */
+template<typename T, unsigned int PROPERTIES>
+class add_call_properties : public T
+{
+public:
+ using T::T;
+
+ unsigned int
+ call_properties (const function_instance &fi) const override
+ {
+ return T::call_properties (fi) | PROPERTIES;
+ }
+};
+
+template<typename T>
+using read_write_za = add_call_properties<T, CP_READ_ZA | CP_WRITE_ZA>;
+
+template<typename T>
+using write_za = add_call_properties<T, CP_WRITE_ZA>;
+
/* A function_base that sometimes or always operates on tuples of
vectors. */
class multi_vector_function : public function_base
typedef unspec_based_function_exact_insn<code_for_aarch64_sve_sub_lane>
unspec_based_sub_lane_function;
+/* General SME unspec-based functions, parameterized on the vector mode. */
+class sme_1mode_function : public read_write_za<unspec_based_function_base>
+{
+public:
+ using parent = read_write_za<unspec_based_function_base>;
+
+ CONSTEXPR sme_1mode_function (int unspec_for_sint, int unspec_for_uint,
+ int unspec_for_fp)
+ : parent (unspec_for_sint, unspec_for_uint, unspec_for_fp, 1)
+ {}
+
+ rtx
+ expand (function_expander &e) const override
+ {
+ auto icode = code_for_aarch64_sme (unspec_for (e), e.tuple_mode (1));
+ return e.use_exact_insn (icode);
+ }
+};
+
+/* General SME unspec-based functions, parameterized on both the ZA mode
+ and the vector mode. */
+template<insn_code (*CODE) (int, machine_mode, machine_mode)>
+class sme_2mode_function_t : public read_write_za<unspec_based_function_base>
+{
+public:
+ using parent = read_write_za<unspec_based_function_base>;
+
+ CONSTEXPR sme_2mode_function_t (int unspec_for_sint, int unspec_for_uint,
+ int unspec_for_fp)
+ : parent (unspec_for_sint, unspec_for_uint, unspec_for_fp, 1)
+ {}
+
+ rtx
+ expand (function_expander &e) const override
+ {
+ insn_code icode = CODE (unspec_for (e), e.vector_mode (0),
+ e.tuple_mode (1));
+ return e.use_exact_insn (icode);
+ }
+};
+
+using sme_2mode_function = sme_2mode_function_t<code_for_aarch64_sme>;
+
/* A function that acts like unspec_based_function_exact_insn<INT_CODE>
when operating on integers, but that expands to an (fma ...)-style
aarch64_sve* operation when applied to floats. */
apply_predication (const function_instance &instance, tree return_type,
vec<tree> &argument_types)
{
- if (instance.pred != PRED_none)
+ /* There are currently no SME ZA instructions that have both merging and
+ unpredicated forms, so for simplicity, the predicates are always included
+ in the original format string. */
+ if (instance.pred != PRED_none && instance.pred != PRED_za_m)
{
argument_types.quick_insert (0, get_svbool_t ());
/* For unary merge operations, the first argument is a vector with
}
};
+/* Base class for binary_za_m and similar shapes. */
+template<type_class_index TCLASS = function_resolver::SAME_TYPE_CLASS,
+ unsigned int BITS = function_resolver::SAME_SIZE>
+struct binary_za_m_base : public overloaded_base<1>
+{
+ tree
+ resolve (function_resolver &r) const override
+ {
+ type_suffix_index type;
+ if (!r.check_num_arguments (5)
+ || !r.require_integer_immediate (0)
+ || !r.require_vector_type (1, VECTOR_TYPE_svbool_t)
+ || !r.require_vector_type (2, VECTOR_TYPE_svbool_t)
+ || (type = r.infer_vector_type (3)) == NUM_TYPE_SUFFIXES
+ || !r.require_derived_vector_type (4, 3, type, TCLASS, BITS))
+ return error_mark_node;
+
+ return r.resolve_to (r.mode_suffix_id, r.type_suffix_ids[0], type);
+ }
+
+ bool
+ check (function_checker &c) const override
+ {
+ return c.require_immediate_range (0, 0, c.num_za_tiles () - 1);
+ }
+};
+
/* Base class for inc_dec and inc_dec_pat. */
struct inc_dec_base : public overloaded_base<0>
{
};
SHAPE (binary_wide_opt_n)
+/* void svfoo_t0[_t1]_g(uint64_t, svbool_t, svbool_t, sv<t1>x<g>_t,
+ sv<t1:int>x<g>_t)
+
+ where the first argument is a ZA tile. */
+struct binary_za_int_m_def : public binary_za_m_base<TYPE_signed>
+{
+ void
+ build (function_builder &b, const function_group_info &group) const override
+ {
+ b.add_overloaded_functions (group, MODE_none);
+ build_all (b, "_,su64,vp,vp,t1,ts1", group, MODE_none);
+ }
+};
+SHAPE (binary_za_int_m)
+
+/* void svfoo_t0[_t1]_g(uint64_t, svbool_t, svbool_t, sv<t1>x<g>_t,
+ sv<t1>x<g>_t)
+
+ where the first argument is a ZA tile. */
+struct binary_za_m_def : public binary_za_m_base<>
+{
+ void
+ build (function_builder &b, const function_group_info &group) const override
+ {
+ b.add_overloaded_functions (group, MODE_none);
+ /* Allow the overloaded form to be specified seperately, with just
+ a single suffix. This is necessary for the 64-bit SME MOP intrinsics,
+ which have some forms dependent on FEAT_SME_I16I64 and some forms
+ dependent on FEAT_SME_F64F64. The resolver needs to be defined
+ for base SME. */
+ if (group.types[0][1] != NUM_TYPE_SUFFIXES)
+ build_all (b, "_,su64,vp,vp,t1,t1", group, MODE_none);
+ }
+};
+SHAPE (binary_za_m)
+
+/* void svfoo_t0[_t1]_g(uint64_t, svbool_t, svbool_t, sv<t1>x<g>_t,
+ sv<t1:uint>x<g>_t)
+
+ where the first argument is a ZA tile. */
+struct binary_za_uint_m_def : public binary_za_m_base<TYPE_unsigned>
+{
+ void
+ build (function_builder &b, const function_group_info &group) const override
+ {
+ b.add_overloaded_functions (group, MODE_none);
+ build_all (b, "_,su64,vp,vp,t1,tu1", group, MODE_none);
+ }
+};
+SHAPE (binary_za_uint_m)
+
+/* bool svfoo(). */
+struct bool_inherent_def : public nonoverloaded_base
+{
+ void
+ build (function_builder &b, const function_group_info &group) const override
+ {
+ build_all (b, "sp", group, MODE_none);
+ }
+};
+SHAPE (bool_inherent)
+
/* sv<t0>_t svfoo[_t0](sv<t0>_t, sv<t0>_t)
<t0>_t svfoo[_n_t0](<t0>_t, sv<t0>_t). */
struct clast_def : public overloaded_base<0>
};
SHAPE (inherent_b)
+/* void svfoo_t0(). */
+struct inherent_za_def : public nonoverloaded_base
+{
+ void
+ build (function_builder &b, const function_group_info &group) const override
+ {
+ build_all (b, "_", group, MODE_none);
+ }
+};
+SHAPE (inherent_za)
+
+/* void svfoo_t0(uint64_t)
+
+ where the argument is an integer constant that specifies an 8-bit mask. */
+struct inherent_mask_za_def : public nonoverloaded_base
+{
+ void
+ build (function_builder &b, const function_group_info &group) const override
+ {
+ build_all (b, "_,su64", group, MODE_none);
+ }
+
+ bool
+ check (function_checker &c) const override
+ {
+ return c.require_immediate_range (0, 0, 255);
+ }
+};
+SHAPE (inherent_mask_za)
+
+/* void svfoo_t0(uint32_t, const void *)
+ void svfoo_vnum_t0(uint32_t, const void *, int64_t)
+
+ where the first argument is a variable ZA slice. */
+struct ldr_za_def : public nonoverloaded_base
+{
+ void
+ build (function_builder &b, const function_group_info &group) const override
+ {
+ build_all (b, "_,su32,al", group, MODE_none);
+ build_all (b, "_,su32,al,ss64", group, MODE_vnum);
+ }
+};
+SHAPE (ldr_za)
+
/* sv<t0>[xN]_t svfoo[_t0](const <t0>_t *)
sv<t0>[xN]_t svfoo_vnum[_t0](const <t0>_t *, int64_t). */
struct load_def : public load_contiguous_base
};
SHAPE (load_replicate)
+/* void svfoo_t0(uint64_t, uint32_t, svbool_t, const void *)
+ void svfoo_vnum_t0(uint64_t, uint32_t, svbool_t, const void *, int64_t)
+
+ where the first two fields form a (ZA tile, slice) pair. */
+struct load_za_def : public nonoverloaded_base
+{
+ void
+ build (function_builder &b, const function_group_info &group) const override
+ {
+ build_all (b, "_,su64,su32,vp,al", group, MODE_none);
+ build_all (b, "_,su64,su32,vp,al,ss64", group, MODE_vnum);
+ }
+
+ bool
+ check (function_checker &c) const override
+ {
+ return c.require_immediate_range (0, 0, c.num_za_tiles () - 1);
+ }
+};
+SHAPE (load_za)
+
/* svbool_t svfoo(enum svpattern). */
struct pattern_pred_def : public nonoverloaded_base
{
};
SHAPE (rdffr)
+/* sv<t1>_t svfoo_t0[_t1](uint64_t, uint32_t)
+
+ where the first two fields form a (ZA tile, slice) pair. */
+struct read_za_m_def : public overloaded_base<1>
+{
+ bool
+ has_merge_argument_p (const function_instance &, unsigned int) const override
+ {
+ return true;
+ }
+
+ void
+ build (function_builder &b, const function_group_info &group) const override
+ {
+ b.add_overloaded_functions (group, MODE_none);
+ build_all (b, "t1,su64,su32", group, MODE_none);
+ }
+
+ tree
+ resolve (function_resolver &r) const override
+ {
+ gcc_assert (r.pred == PRED_m);
+ type_suffix_index type;
+ if (!r.check_num_arguments (4)
+ || (type = r.infer_vector_type (0)) == NUM_TYPE_SUFFIXES
+ || !r.require_vector_type (1, VECTOR_TYPE_svbool_t)
+ || !r.require_integer_immediate (2)
+ || !r.require_scalar_type (3, "uint32_t"))
+ return error_mark_node;
+
+ return r.resolve_to (r.mode_suffix_id, r.type_suffix_ids[0], type);
+ }
+
+ bool
+ check (function_checker &c) const override
+ {
+ gcc_assert (c.pred == PRED_m);
+ return c.require_immediate_range (1, 0, c.num_za_tiles () - 1);
+ }
+};
+SHAPE (read_za_m)
+
/* <t0>_t svfoo[_t0](sv<t0>_t). */
struct reduction_def : public overloaded_base<0>
{
};
SHAPE (store_scatter_offset_restricted)
+/* void svfoo_t0(uint64_t, uint32_t, svbool_t, void *)
+ void svfoo_vnum_t0(uint64_t, uint32_t, svbool_t, void *, int64_t)
+
+ where the first two fields form a (ZA tile, slice) pair. */
+struct store_za_def : public nonoverloaded_base
+{
+ void
+ build (function_builder &b, const function_group_info &group) const override
+ {
+ build_all (b, "_,su64,su32,vp,as", group, MODE_none);
+ build_all (b, "_,su64,su32,vp,as,ss64", group, MODE_vnum);
+ }
+
+ bool
+ check (function_checker &c) const override
+ {
+ return c.require_immediate_range (0, 0, c.num_za_tiles () - 1);
+ }
+};
+SHAPE (store_za)
+
+/* void svfoo_t0(uint32_t, void *)
+ void svfoo_vnum_t0(uint32_t, void *, int64_t)
+
+ where the first argument is a variable ZA slice. */
+struct str_za_def : public nonoverloaded_base
+{
+ void
+ build (function_builder &b, const function_group_info &group) const override
+ {
+ build_all (b, "_,su32,as", group, MODE_none);
+ build_all (b, "_,su32,as,ss64", group, MODE_vnum);
+ }
+};
+SHAPE (str_za)
+
/* sv<t0>_t svfoo[_t0](sv<t0>xN_t, sv<t0:uint>_t). */
struct tbl_tuple_def : public overloaded_base<0>
{
};
SHAPE (unary_widen)
+/* void svfoo_t0[_t1](uint64_t, svbool_t, svbool_t, sv<t1>_t)
+
+ where the first argument is a ZA tile. */
+struct unary_za_m_def : public overloaded_base<1>
+{
+ void
+ build (function_builder &b, const function_group_info &group) const override
+ {
+ b.add_overloaded_functions (group, MODE_none);
+ build_all (b, "_,su64,vp,vp,t1", group, MODE_none);
+ }
+
+ tree
+ resolve (function_resolver &r) const override
+ {
+ type_suffix_index type;
+ if (!r.check_num_arguments (4)
+ || !r.require_integer_immediate (0)
+ || !r.require_vector_type (1, VECTOR_TYPE_svbool_t)
+ || !r.require_vector_type (2, VECTOR_TYPE_svbool_t)
+ || (type = r.infer_vector_type (3)) == NUM_TYPE_SUFFIXES)
+ return error_mark_node;
+
+ return r.resolve_to (r.mode_suffix_id, r.type_suffix_ids[0], type);
+ }
+
+ bool
+ check (function_checker &c) const override
+ {
+ return c.require_immediate_range (0, 0, c.num_za_tiles () - 1);
+ }
+};
+SHAPE (unary_za_m)
+
+/* void svfoo_t0[_t1](uint64_t, uint32_t, svbool_t, sv<t1>_t)
+
+ where the first two fields form a (ZA tile, slice) pair. */
+struct write_za_m_def : public overloaded_base<1>
+{
+ void
+ build (function_builder &b, const function_group_info &group) const override
+ {
+ b.add_overloaded_functions (group, MODE_none);
+ build_all (b, "_,su64,su32,vp,t1", group, MODE_none);
+ }
+
+ tree
+ resolve (function_resolver &r) const override
+ {
+ type_suffix_index type;
+ if (!r.check_num_arguments (4)
+ || !r.require_integer_immediate (0)
+ || !r.require_scalar_type (1, "uint32_t")
+ || !r.require_vector_type (2, VECTOR_TYPE_svbool_t)
+ || (type = r.infer_vector_type (3)) == NUM_TYPE_SUFFIXES)
+ return error_mark_node;
+
+ return r.resolve_to (r.mode_suffix_id, r.type_suffix_ids[0], type);
+ }
+
+ bool
+ check (function_checker &c) const override
+ {
+ return c.require_immediate_range (0, 0, c.num_za_tiles () - 1);
+ }
+};
+SHAPE (write_za_m)
+
}
extern const function_shape *const binary_uint64_opt_n;
extern const function_shape *const binary_wide;
extern const function_shape *const binary_wide_opt_n;
+ extern const function_shape *const binary_za_int_m;
+ extern const function_shape *const binary_za_m;
+ extern const function_shape *const binary_za_uint_m;
+ extern const function_shape *const bool_inherent;
extern const function_shape *const clast;
extern const function_shape *const compare;
extern const function_shape *const compare_opt_n;
extern const function_shape *const inc_dec_pred_scalar;
extern const function_shape *const inherent;
extern const function_shape *const inherent_b;
+ extern const function_shape *const inherent_za;
+ extern const function_shape *const inherent_mask_za;
+ extern const function_shape *const ldr_za;
extern const function_shape *const load;
extern const function_shape *const load_ext;
extern const function_shape *const load_ext_gather_index;
extern const function_shape *const load_gather_sv_restricted;
extern const function_shape *const load_gather_vs;
extern const function_shape *const load_replicate;
+ extern const function_shape *const load_za;
extern const function_shape *const mmla;
extern const function_shape *const pattern_pred;
extern const function_shape *const prefetch;
extern const function_shape *const prefetch_gather_offset;
extern const function_shape *const ptest;
extern const function_shape *const rdffr;
+ extern const function_shape *const read_za_m;
extern const function_shape *const reduction;
extern const function_shape *const reduction_wide;
extern const function_shape *const reinterpret;
extern const function_shape *const store_scatter_index_restricted;
extern const function_shape *const store_scatter_offset;
extern const function_shape *const store_scatter_offset_restricted;
+ extern const function_shape *const store_za;
+ extern const function_shape *const str_za;
extern const function_shape *const tbl_tuple;
extern const function_shape *const ternary_bfloat;
extern const function_shape *const ternary_bfloat_lane;
extern const function_shape *const unary_to_uint;
extern const function_shape *const unary_uint;
extern const function_shape *const unary_widen;
+ extern const function_shape *const unary_za_m;
+ extern const function_shape *const write_za_m;
}
}
--- /dev/null
+/* ACLE support for AArch64 SME.
+ Copyright (C) 2023 Free Software Foundation, Inc.
+
+ This file is part of GCC.
+
+ GCC is free software; you can redistribute it and/or modify it
+ under the terms of the GNU General Public License as published by
+ the Free Software Foundation; either version 3, or (at your option)
+ any later version.
+
+ GCC is distributed in the hope that it will be useful, but
+ WITHOUT ANY WARRANTY; without even the implied warranty of
+ MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
+ General Public License for more details.
+
+ You should have received a copy of the GNU General Public License
+ along with GCC; see the file COPYING3. If not see
+ <http://www.gnu.org/licenses/>. */
+
+#include "config.h"
+#include "system.h"
+#include "coretypes.h"
+#include "tm.h"
+#include "tree.h"
+#include "rtl.h"
+#include "tm_p.h"
+#include "memmodel.h"
+#include "insn-codes.h"
+#include "optabs.h"
+#include "recog.h"
+#include "expr.h"
+#include "basic-block.h"
+#include "function.h"
+#include "fold-const.h"
+#include "gimple.h"
+#include "gimple-iterator.h"
+#include "gimplify.h"
+#include "explow.h"
+#include "emit-rtl.h"
+#include "aarch64-sve-builtins.h"
+#include "aarch64-sve-builtins-shapes.h"
+#include "aarch64-sve-builtins-base.h"
+#include "aarch64-sve-builtins-sme.h"
+#include "aarch64-sve-builtins-functions.h"
+
+using namespace aarch64_sve;
+
+namespace {
+
+class load_store_za_base : public function_base
+{
+public:
+ tree
+ memory_scalar_type (const function_instance &) const override
+ {
+ return void_type_node;
+ }
+};
+
+class read_write_za_base : public function_base
+{
+public:
+ constexpr read_write_za_base (int unspec) : m_unspec (unspec) {}
+
+ rtx
+ expand (function_expander &e) const override
+ {
+ auto za_mode = e.vector_mode (0);
+ auto z_mode = e.vector_mode (1);
+ auto icode = (za_mode == VNx1TImode
+ ? code_for_aarch64_sme (m_unspec, za_mode, z_mode)
+ : code_for_aarch64_sme (m_unspec, z_mode, z_mode));
+ return e.use_exact_insn (icode);
+ }
+
+ int m_unspec;
+};
+
+using load_za_base = add_call_properties<load_store_za_base,
+ CP_READ_MEMORY | CP_READ_ZA
+ | CP_WRITE_ZA>;
+
+using store_za_base = add_call_properties<load_store_za_base,
+ CP_WRITE_MEMORY | CP_READ_ZA>;
+
+/* E is a load or store intrinsic that accesses a ZA slice of mode MEM_MODE.
+ The intrinsic has a vnum parameter at index ARGNO. Return true if the
+ vnum argument is a constant that is a valid ZA offset for the underlying
+ instruction. */
+
+static bool
+has_in_range_vnum_arg (function_expander &e, machine_mode mem_mode,
+ unsigned int argno)
+{
+ return (e.mode_suffix_id == MODE_vnum
+ && CONST_INT_P (e.args[argno])
+ && UINTVAL (e.args[argno]) < 16 / GET_MODE_UNIT_SIZE (mem_mode));
+}
+
+/* E is a ZA load or store intrinsic that uses instruction ICODE. Add a
+ 32-bit operand that gives the total ZA slice. (The instruction hard-codes
+ the constant offset to 0, so there is no operand for that.)
+
+ Argument ARGNO is the intrinsic's slice argument. If the intrinsic is
+ a _vnum intrinsic, argument VNUM_ARGNO is the intrinsic's vnum operand,
+ which must be added to the slice argument. */
+
+static void
+add_load_store_slice_operand (function_expander &e, insn_code icode,
+ unsigned int argno, unsigned int vnum_argno)
+{
+ rtx base = e.args[argno];
+ if (e.mode_suffix_id == MODE_vnum)
+ {
+ rtx vnum = lowpart_subreg (SImode, e.args[vnum_argno], DImode);
+ base = simplify_gen_binary (PLUS, SImode, base, vnum);
+ }
+ e.add_input_operand (icode, base);
+}
+
+/* Add a memory operand for ZA LD1 or ST1 intrinsic E. BASE_ARGNO is
+ the index of the base argument. */
+
+static void
+add_load_store_operand (function_expander &e, unsigned int base_argno)
+{
+ auto mode = e.vector_mode (0);
+ rtx base = e.get_contiguous_base (mode, base_argno, base_argno + 1,
+ AARCH64_FL_SM_ON);
+ auto mem = gen_rtx_MEM (mode, force_reg (Pmode, base));
+ set_mem_align (mem, BITS_PER_UNIT);
+ e.add_fixed_operand (mem);
+}
+
+/* Expand ZA LDR or STR intrinsic E. There are two underlying instructions:
+
+ - BASE_CODE has a zero ZA slice offset
+ - VNUM_CODE has a constant operand for the ZA slice offset. */
+
+static rtx
+expand_ldr_str_za (function_expander &e, insn_code base_code,
+ insn_code vnum_code)
+{
+ if (has_in_range_vnum_arg (e, VNx16QImode, 2))
+ {
+ rtx mem_offset = aarch64_sme_vq_immediate (Pmode,
+ UINTVAL (e.args[2]) * 16,
+ AARCH64_ISA_MODE);
+ e.add_input_operand (vnum_code, e.args[0]);
+ e.add_input_operand (vnum_code, e.args[2]);
+ e.add_input_operand (vnum_code, e.args[1]);
+ e.add_input_operand (vnum_code, mem_offset);
+ return e.generate_insn (vnum_code);
+ }
+ else
+ {
+ rtx base = e.get_contiguous_base (VNx16QImode, 1, 2, AARCH64_FL_SM_ON);
+ add_load_store_slice_operand (e, base_code, 0, 2);
+ e.add_input_operand (base_code, base);
+ return e.generate_insn (base_code);
+ }
+}
+
+/* Expand ZA LD1 or ST1 intrinsic E. UNSPEC is the load or store unspec.
+ IS_LOAD is true if E is a load, false if it is a store. */
+
+static rtx
+expand_ld1_st1 (function_expander &e, int unspec, bool is_load)
+{
+ bool is_vnum = has_in_range_vnum_arg (e, e.vector_mode (0), 4);
+ auto icode = (is_vnum
+ ? code_for_aarch64_sme_plus (unspec, e.vector_mode (0))
+ : code_for_aarch64_sme (unspec, e.vector_mode (0)));
+ if (!is_load)
+ add_load_store_operand (e, 3);
+ e.add_input_operand (icode, e.args[0]);
+ if (is_vnum)
+ {
+ e.add_input_operand (icode, e.args[1]);
+ e.add_input_operand (icode, e.args[4]);
+ }
+ else
+ add_load_store_slice_operand (e, icode, 1, 4);
+ e.add_input_operand (icode, e.args[2]);
+ if (is_load)
+ add_load_store_operand (e, 3);
+ return e.generate_insn (icode);
+}
+
+class arm_has_sme_impl : public function_base
+{
+ gimple *
+ fold (gimple_folder &f) const override
+ {
+ if (TARGET_SME)
+ return f.fold_to_cstu (1);
+ return nullptr;
+ }
+
+ rtx
+ expand (function_expander &e) const override
+ {
+ if (TARGET_SME)
+ return const1_rtx;
+ emit_insn (gen_aarch64_get_sme_state ());
+ return expand_simple_binop (DImode, LSHIFTRT,
+ gen_rtx_REG (DImode, R0_REGNUM),
+ gen_int_mode (63, QImode),
+ e.possible_target, true, OPTAB_LIB_WIDEN);
+ }
+};
+
+class arm_in_streaming_mode_impl : public function_base
+{
+ gimple *
+ fold (gimple_folder &f) const override
+ {
+ if (TARGET_STREAMING)
+ return f.fold_to_cstu (1);
+ if (TARGET_NON_STREAMING)
+ return f.fold_to_cstu (0);
+ return nullptr;
+ }
+
+ rtx
+ expand (function_expander &e) const override
+ {
+ if (TARGET_STREAMING)
+ return const1_rtx;
+
+ if (TARGET_NON_STREAMING)
+ return const0_rtx;
+
+ rtx reg;
+ if (TARGET_SME)
+ {
+ reg = gen_reg_rtx (DImode);
+ emit_insn (gen_aarch64_read_svcr (reg));
+ }
+ else
+ {
+ emit_insn (gen_aarch64_get_sme_state ());
+ reg = gen_rtx_REG (DImode, R0_REGNUM);
+ }
+ return expand_simple_binop (DImode, AND, reg, gen_int_mode (1, DImode),
+ e.possible_target, true, OPTAB_LIB_WIDEN);
+ }
+};
+
+/* Implements svcnts[bhwd]. */
+class svcnts_bhwd_impl : public function_base
+{
+public:
+ constexpr svcnts_bhwd_impl (machine_mode ref_mode) : m_ref_mode (ref_mode) {}
+
+ unsigned int
+ get_shift () const
+ {
+ return exact_log2 (GET_MODE_UNIT_SIZE (m_ref_mode));
+ }
+
+ gimple *
+ fold (gimple_folder &f) const override
+ {
+ if (TARGET_STREAMING)
+ return f.fold_to_cstu (GET_MODE_NUNITS (m_ref_mode));
+ return nullptr;
+ }
+
+ rtx
+ expand (function_expander &e) const override
+ {
+ rtx cntsb = aarch64_sme_vq_immediate (DImode, 16, AARCH64_ISA_MODE);
+ auto shift = get_shift ();
+ if (!shift)
+ return cntsb;
+
+ return expand_simple_binop (DImode, LSHIFTRT, cntsb,
+ gen_int_mode (shift, QImode),
+ e.possible_target, true, OPTAB_LIB_WIDEN);
+ }
+
+ /* The mode of the vector associated with the [bhwd] suffix. */
+ machine_mode m_ref_mode;
+};
+
+class svld1_za_impl : public load_za_base
+{
+public:
+ constexpr svld1_za_impl (int unspec) : m_unspec (unspec) {}
+
+ rtx
+ expand (function_expander &e) const override
+ {
+ return expand_ld1_st1 (e, m_unspec, true);
+ }
+
+ int m_unspec;
+};
+
+class svldr_za_impl : public load_za_base
+{
+public:
+ rtx
+ expand (function_expander &e) const override
+ {
+ return expand_ldr_str_za (e, CODE_FOR_aarch64_sme_ldr0,
+ code_for_aarch64_sme_ldrn (Pmode));
+ }
+};
+
+using svread_za_tile_impl = add_call_properties<read_write_za_base,
+ CP_READ_ZA>;
+
+class svst1_za_impl : public store_za_base
+{
+public:
+ constexpr svst1_za_impl (int unspec) : m_unspec (unspec) {}
+
+ rtx
+ expand (function_expander &e) const override
+ {
+ return expand_ld1_st1 (e, m_unspec, false);
+ }
+
+ int m_unspec;
+};
+
+class svstr_za_impl : public store_za_base
+{
+public:
+ rtx
+ expand (function_expander &e) const override
+ {
+ return expand_ldr_str_za (e, CODE_FOR_aarch64_sme_str0,
+ code_for_aarch64_sme_strn (Pmode));
+ }
+};
+
+class svundef_za_impl : public write_za<function_base>
+{
+public:
+ rtx
+ expand (function_expander &) const override
+ {
+ rtx target = gen_rtx_REG (VNx16QImode, ZA_REGNUM);
+ emit_clobber (copy_rtx (target));
+ return const0_rtx;
+ }
+};
+
+using svwrite_za_tile_impl = add_call_properties<read_write_za_base,
+ CP_READ_ZA | CP_WRITE_ZA>;
+
+class svzero_mask_za_impl : public write_za<function_base>
+{
+public:
+ rtx
+ expand (function_expander &e) const override
+ {
+ return e.use_exact_insn (CODE_FOR_aarch64_sme_zero_za);
+ }
+};
+
+class svzero_za_impl : public write_za<function_base>
+{
+public:
+ rtx
+ expand (function_expander &) const override
+ {
+ emit_insn (gen_aarch64_sme_zero_za (gen_int_mode (0xff, SImode)));
+ return const0_rtx;
+ }
+};
+
+} /* end anonymous namespace */
+
+namespace aarch64_sve {
+
+FUNCTION (arm_has_sme, arm_has_sme_impl, )
+FUNCTION (arm_in_streaming_mode, arm_in_streaming_mode_impl, )
+FUNCTION (svaddha_za, sme_1mode_function, (UNSPEC_SME_ADDHA,
+ UNSPEC_SME_ADDHA, -1))
+FUNCTION (svaddva_za, sme_1mode_function, (UNSPEC_SME_ADDVA,
+ UNSPEC_SME_ADDVA, -1))
+FUNCTION (svcntsb, svcnts_bhwd_impl, (VNx16QImode))
+FUNCTION (svcntsd, svcnts_bhwd_impl, (VNx2DImode))
+FUNCTION (svcntsh, svcnts_bhwd_impl, (VNx8HImode))
+FUNCTION (svcntsw, svcnts_bhwd_impl, (VNx4SImode))
+FUNCTION (svld1_hor_za, svld1_za_impl, (UNSPEC_SME_LD1_HOR))
+FUNCTION (svld1_ver_za, svld1_za_impl, (UNSPEC_SME_LD1_VER))
+FUNCTION (svldr_za, svldr_za_impl, )
+FUNCTION (svmopa_za, sme_2mode_function, (UNSPEC_SME_SMOPA, UNSPEC_SME_UMOPA,
+ UNSPEC_SME_FMOPA))
+FUNCTION (svmops_za, sme_2mode_function, (UNSPEC_SME_SMOPS, UNSPEC_SME_UMOPS,
+ UNSPEC_SME_FMOPS))
+FUNCTION (svread_hor_za, svread_za_tile_impl, (UNSPEC_SME_READ_HOR))
+FUNCTION (svread_ver_za, svread_za_tile_impl, (UNSPEC_SME_READ_VER))
+FUNCTION (svst1_hor_za, svst1_za_impl, (UNSPEC_SME_ST1_HOR))
+FUNCTION (svst1_ver_za, svst1_za_impl, (UNSPEC_SME_ST1_VER))
+FUNCTION (svstr_za, svstr_za_impl, )
+FUNCTION (svsumopa_za, sme_2mode_function, (UNSPEC_SME_SUMOPA, -1, -1))
+FUNCTION (svsumops_za, sme_2mode_function, (UNSPEC_SME_SUMOPS, -1, -1))
+FUNCTION (svundef_za, svundef_za_impl, )
+FUNCTION (svusmopa_za, sme_2mode_function, (-1, UNSPEC_SME_USMOPA, -1))
+FUNCTION (svusmops_za, sme_2mode_function, (-1, UNSPEC_SME_USMOPS, -1))
+FUNCTION (svwrite_hor_za, svwrite_za_tile_impl, (UNSPEC_SME_WRITE_HOR))
+FUNCTION (svwrite_ver_za, svwrite_za_tile_impl, (UNSPEC_SME_WRITE_VER))
+FUNCTION (svzero_mask_za, svzero_mask_za_impl, )
+FUNCTION (svzero_za, svzero_za_impl, )
+
+} /* end namespace aarch64_sve */
--- /dev/null
+/* ACLE support for AArch64 SME.
+ Copyright (C) 2023 Free Software Foundation, Inc.
+
+ This file is part of GCC.
+
+ GCC is free software; you can redistribute it and/or modify it
+ under the terms of the GNU General Public License as published by
+ the Free Software Foundation; either version 3, or (at your option)
+ any later version.
+
+ GCC is distributed in the hope that it will be useful, but
+ WITHOUT ANY WARRANTY; without even the implied warranty of
+ MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
+ General Public License for more details.
+
+ You should have received a copy of the GNU General Public License
+ along with GCC; see the file COPYING3. If not see
+ <http://www.gnu.org/licenses/>. */
+
+#define REQUIRED_EXTENSIONS 0
+DEF_SVE_FUNCTION (arm_has_sme, bool_inherent, none, none)
+DEF_SVE_FUNCTION (arm_in_streaming_mode, bool_inherent, none, none)
+#undef REQUIRED_EXTENSIONS
+
+#define REQUIRED_EXTENSIONS AARCH64_FL_SME
+DEF_SVE_FUNCTION (svcntsb, count_inherent, none, none)
+DEF_SVE_FUNCTION (svcntsd, count_inherent, none, none)
+DEF_SVE_FUNCTION (svcntsh, count_inherent, none, none)
+DEF_SVE_FUNCTION (svcntsw, count_inherent, none, none)
+DEF_SME_ZA_FUNCTION (svldr, ldr_za, za, none)
+DEF_SME_ZA_FUNCTION (svstr, str_za, za, none)
+DEF_SME_ZA_FUNCTION (svundef, inherent_za, za, none)
+DEF_SME_ZA_FUNCTION (svzero, inherent_za, za, none)
+DEF_SME_ZA_FUNCTION (svzero_mask, inherent_mask_za, za, none)
+#undef REQUIRED_EXTENSIONS
+
+#define REQUIRED_EXTENSIONS AARCH64_FL_SME | AARCH64_FL_SM_ON
+DEF_SME_ZA_FUNCTION (svaddha, unary_za_m, za_s_integer, za_m)
+DEF_SME_ZA_FUNCTION (svaddva, unary_za_m, za_s_integer, za_m)
+DEF_SME_ZA_FUNCTION (svld1_hor, load_za, all_za, none)
+DEF_SME_ZA_FUNCTION (svld1_ver, load_za, all_za, none)
+DEF_SME_ZA_FUNCTION (svmopa, binary_za_m, mop_base, za_m)
+DEF_SME_ZA_FUNCTION (svmopa, binary_za_m, d_za, za_m)
+DEF_SME_ZA_FUNCTION (svmops, binary_za_m, mop_base, za_m)
+DEF_SME_ZA_FUNCTION (svmops, binary_za_m, d_za, za_m)
+DEF_SME_ZA_FUNCTION (svread_hor, read_za_m, za_all_data, m)
+DEF_SME_ZA_FUNCTION (svread_ver, read_za_m, za_all_data, m)
+DEF_SME_ZA_FUNCTION (svst1_hor, store_za, all_za, none)
+DEF_SME_ZA_FUNCTION (svst1_ver, store_za, all_za, none)
+DEF_SME_ZA_FUNCTION (svsumopa, binary_za_uint_m, mop_base_signed, za_m)
+DEF_SME_ZA_FUNCTION (svsumops, binary_za_uint_m, mop_base_signed, za_m)
+DEF_SME_ZA_FUNCTION (svusmopa, binary_za_int_m, mop_base_unsigned, za_m)
+DEF_SME_ZA_FUNCTION (svusmops, binary_za_int_m, mop_base_unsigned, za_m)
+DEF_SME_ZA_FUNCTION (svwrite_hor, write_za_m, za_all_data, za_m)
+DEF_SME_ZA_FUNCTION (svwrite_ver, write_za_m, za_all_data, za_m)
+#undef REQUIRED_EXTENSIONS
+
+#define REQUIRED_EXTENSIONS (AARCH64_FL_SME \
+ | AARCH64_FL_SME_I16I64 \
+ | AARCH64_FL_SM_ON)
+DEF_SME_ZA_FUNCTION (svaddha, unary_za_m, za_d_integer, za_m)
+DEF_SME_ZA_FUNCTION (svaddva, unary_za_m, za_d_integer, za_m)
+DEF_SME_ZA_FUNCTION (svmopa, binary_za_m, mop_i16i64, za_m)
+DEF_SME_ZA_FUNCTION (svmops, binary_za_m, mop_i16i64, za_m)
+DEF_SME_ZA_FUNCTION (svsumopa, binary_za_uint_m, mop_i16i64_signed, za_m)
+DEF_SME_ZA_FUNCTION (svsumops, binary_za_uint_m, mop_i16i64_signed, za_m)
+DEF_SME_ZA_FUNCTION (svusmopa, binary_za_int_m, mop_i16i64_unsigned, za_m)
+DEF_SME_ZA_FUNCTION (svusmops, binary_za_int_m, mop_i16i64_unsigned, za_m)
+#undef REQUIRED_EXTENSIONS
+
+#define REQUIRED_EXTENSIONS (AARCH64_FL_SME \
+ | AARCH64_FL_SME_F64F64 \
+ | AARCH64_FL_SM_ON)
+DEF_SME_ZA_FUNCTION (svmopa, binary_za_m, za_d_float, za_m)
+DEF_SME_ZA_FUNCTION (svmops, binary_za_m, za_d_float, za_m)
+#undef REQUIRED_EXTENSIONS
--- /dev/null
+/* ACLE support for AArch64 SME.
+ Copyright (C) 2023 Free Software Foundation, Inc.
+
+ This file is part of GCC.
+
+ GCC is free software; you can redistribute it and/or modify it
+ under the terms of the GNU General Public License as published by
+ the Free Software Foundation; either version 3, or (at your option)
+ any later version.
+
+ GCC is distributed in the hope that it will be useful, but
+ WITHOUT ANY WARRANTY; without even the implied warranty of
+ MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
+ General Public License for more details.
+
+ You should have received a copy of the GNU General Public License
+ along with GCC; see the file COPYING3. If not see
+ <http://www.gnu.org/licenses/>. */
+
+#ifndef GCC_AARCH64_SVE_BUILTINS_SME_H
+#define GCC_AARCH64_SVE_BUILTINS_SME_H
+
+namespace aarch64_sve
+{
+ namespace functions
+ {
+ extern const function_base *const arm_has_sme;
+ extern const function_base *const arm_in_streaming_mode;
+ extern const function_base *const svaddha_za;
+ extern const function_base *const svaddva_za;
+ extern const function_base *const svcntsb;
+ extern const function_base *const svcntsd;
+ extern const function_base *const svcntsh;
+ extern const function_base *const svcntsw;
+ extern const function_base *const svld1_hor_za;
+ extern const function_base *const svld1_ver_za;
+ extern const function_base *const svldr_za;
+ extern const function_base *const svmopa_za;
+ extern const function_base *const svmops_za;
+ extern const function_base *const svread_hor_za;
+ extern const function_base *const svread_ver_za;
+ extern const function_base *const svst1_hor_za;
+ extern const function_base *const svst1_ver_za;
+ extern const function_base *const svstr_za;
+ extern const function_base *const svsumopa_za;
+ extern const function_base *const svsumops_za;
+ extern const function_base *const svusmopa_za;
+ extern const function_base *const svusmops_za;
+ extern const function_base *const svwrite_hor_za;
+ extern const function_base *const svwrite_ver_za;
+ extern const function_base *const svundef_za;
+ extern const function_base *const svzero_za;
+ extern const function_base *const svzero_mask_za;
+ }
+}
+
+#endif
#include "aarch64-sve-builtins.h"
#include "aarch64-sve-builtins-base.h"
#include "aarch64-sve-builtins-sve2.h"
+#include "aarch64-sve-builtins-sme.h"
#include "aarch64-sve-builtins-shapes.h"
namespace aarch64_sve {
"_m",
"_x",
"_z",
+ "_m",
""
};
TYPE_##CLASS == TYPE_signed || TYPE_##CLASS == TYPE_unsigned, \
TYPE_##CLASS == TYPE_unsigned, \
TYPE_##CLASS == TYPE_float, \
+ TYPE_##CLASS != TYPE_bool, \
TYPE_##CLASS == TYPE_bool, \
+ false, \
+ 0, \
+ MODE },
+#define DEF_SME_ZA_SUFFIX(NAME, BITS, MODE) \
+ { "_" #NAME, \
+ NUM_VECTOR_TYPES, \
+ NUM_TYPE_CLASSES, \
+ BITS, \
+ BITS / BITS_PER_UNIT, \
+ false, \
+ false, \
+ false, \
+ false, \
+ false, \
+ true, \
0, \
MODE },
#include "aarch64-sve-builtins.def"
{ "", NUM_VECTOR_TYPES, TYPE_bool, 0, 0, false, false, false, false,
- 0, VOIDmode }
+ false, false, 0, VOIDmode }
};
CONSTEXPR const group_suffix_info group_suffixes[] = {
TYPES_while1 (D, b32), \
TYPES_while1 (D, b64)
+/* _za8 _za16 _za32 _za64 _za128. */
+#define TYPES_all_za(S, D) \
+ S (za8), S (za16), S (za32), S (za64), S (za128)
+
+/* _za64. */
+#define TYPES_d_za(S, D) \
+ S (za64)
+
+/* { _za8 } x { _s8 _u8 }
+
+ { _za16 } x { _bf16 _f16 _s16 _u16 }
+
+ { _za32 } x { _f32 _s32 _u32 }
+
+ { _za64 } x { _f64 _s64 _u64 }. */
+#define TYPES_za_bhsd_data(S, D) \
+ D (za8, s8), D (za8, u8), \
+ D (za16, bf16), D (za16, f16), D (za16, s16), D (za16, u16), \
+ D (za32, f32), D (za32, s32), D (za32, u32), \
+ D (za64, f64), D (za64, s64), D (za64, u64)
+
+/* Likewise, plus:
+
+ { _za128 } x { _bf16 }
+ { _f16 _f32 _f64 }
+ { _s8 _s16 _s32 _s64 }
+ { _u8 _u16 _u32 _u64 }. */
+
+#define TYPES_za_all_data(S, D) \
+ TYPES_za_bhsd_data (S, D), \
+ TYPES_reinterpret1 (D, za128)
+
+/* _za32 x { _s32 _u32 }. */
+#define TYPES_za_s_integer(S, D) \
+ D (za32, s32), D (za32, u32)
+
+
+/* _za64_f64. */
+#define TYPES_za_d_float(S, D) \
+ D (za64, f64)
+
+/* _za64 x { _s64 _u64 }. */
+#define TYPES_za_d_integer(S, D) \
+ D (za64, s64), D (za64, u64)
+
+/* _za32 x { _s8 _u8 _bf16 _f16 _f32 }. */
+#define TYPES_mop_base(S, D) \
+ D (za32, s8), D (za32, u8), D (za32, bf16), D (za32, f16), D (za32, f32)
+
+/* _za32_s8. */
+#define TYPES_mop_base_signed(S, D) \
+ D (za32, s8)
+
+/* _za32_u8. */
+#define TYPES_mop_base_unsigned(S, D) \
+ D (za32, u8)
+
+/* _za64 x { _s16 _u16 }. */
+#define TYPES_mop_i16i64(S, D) \
+ D (za64, s16), D (za64, u16)
+
+/* _za64_s16. */
+#define TYPES_mop_i16i64_signed(S, D) \
+ D (za64, s16)
+
+/* _za64_u16. */
+#define TYPES_mop_i16i64_unsigned(S, D) \
+ D (za64, u16)
+
+/* _za. */
+#define TYPES_za(S, D) \
+ S (za)
+
/* Describe a pair of type suffixes in which only the first is used. */
#define DEF_VECTOR_TYPE(X) { TYPE_SUFFIX_ ## X, NUM_TYPE_SUFFIXES }
DEF_SVE_TYPES_ARRAY (inc_dec_n);
DEF_SVE_TYPES_ARRAY (reinterpret);
DEF_SVE_TYPES_ARRAY (while);
+DEF_SVE_TYPES_ARRAY (all_za);
+DEF_SVE_TYPES_ARRAY (d_za);
+DEF_SVE_TYPES_ARRAY (za_all_data);
+DEF_SVE_TYPES_ARRAY (za_s_integer);
+DEF_SVE_TYPES_ARRAY (za_d_float);
+DEF_SVE_TYPES_ARRAY (za_d_integer);
+DEF_SVE_TYPES_ARRAY (mop_base);
+DEF_SVE_TYPES_ARRAY (mop_base_signed);
+DEF_SVE_TYPES_ARRAY (mop_base_unsigned);
+DEF_SVE_TYPES_ARRAY (mop_i16i64);
+DEF_SVE_TYPES_ARRAY (mop_i16i64_signed);
+DEF_SVE_TYPES_ARRAY (mop_i16i64_unsigned);
+DEF_SVE_TYPES_ARRAY (za);
static const group_suffix_index groups_none[] = {
GROUP_none, NUM_GROUP_SUFFIXES
explicit suffix. */
static const predication_index preds_implicit[] = { PRED_implicit, NUM_PREDS };
+/* Used by functions that only support "_m" predication. */
+static const predication_index preds_m[] = { PRED_m, NUM_PREDS };
+
/* Used by functions that allow merging and "don't care" predication,
but are not suitable for predicated MOVPRFX. */
static const predication_index preds_mx[] = {
/* Used by (mostly predicate) functions that only support "_z" predication. */
static const predication_index preds_z[] = { PRED_z, NUM_PREDS };
+/* Used by SME instructions that always merge into ZA. */
+static const predication_index preds_za_m[] = { PRED_za_m, NUM_PREDS };
+
/* A list of all SVE ACLE functions. */
static CONSTEXPR const function_group_info function_groups[] = {
#define DEF_SVE_FUNCTION_GS(NAME, SHAPE, TYPES, GROUPS, PREDS) \
{ #NAME, &functions::NAME, &shapes::SHAPE, types_##TYPES, groups_##GROUPS, \
preds_##PREDS, REQUIRED_EXTENSIONS },
+#define DEF_SME_ZA_FUNCTION_GS(NAME, SHAPE, TYPES, GROUPS, PREDS) \
+ { #NAME, &functions::NAME##_za, &shapes::SHAPE, types_##TYPES, \
+ groups_##GROUPS, preds_##PREDS, (REQUIRED_EXTENSIONS | AARCH64_FL_ZA_ON) },
#include "aarch64-sve-builtins.def"
};
/* The scalar type associated with each vector type. */
-extern GTY(()) tree scalar_types[NUM_VECTOR_TYPES];
-tree scalar_types[NUM_VECTOR_TYPES];
+extern GTY(()) tree scalar_types[NUM_VECTOR_TYPES + 1];
+tree scalar_types[NUM_VECTOR_TYPES + 1];
/* The single-predicate and single-vector types, with their built-in
"__SV..._t" name. Allow an index of NUM_VECTOR_TYPES, which always
/* A linear search should be OK here, since the code isn't hot and
the number of types is only small. */
for (unsigned int suffix_i = 0; suffix_i < NUM_TYPE_SUFFIXES; ++suffix_i)
- if (!type_suffixes[suffix_i].bool_p)
+ if (type_suffixes[suffix_i].vector_p)
{
vector_type_index vector_i = type_suffixes[suffix_i].vector_type;
if (matches_type_p (scalar_types[vector_i], type))
return false;
}
+ if (missing_extensions & AARCH64_FL_SM_ON)
+ {
+ error_at (location, "ACLE function %qD can only be called when"
+ " SME streaming mode is enabled", fndecl);
+ return false;
+ }
+
+ if (missing_extensions & AARCH64_FL_ZA_ON)
+ {
+ error_at (location, "ACLE function %qD can only be called from"
+ " a function that has %qs state", fndecl, "za");
+ return false;
+ }
+
static const struct {
aarch64_feature_flags flag;
const char *name;
HOST_WIDE_INT actual, HOST_WIDE_INT min,
HOST_WIDE_INT max)
{
- error_at (location, "passing %wd to argument %d of %qE, which expects"
- " a value in the range [%wd, %wd]", actual, argno + 1, fndecl,
- min, max);
+ if (min == max)
+ error_at (location, "passing %wd to argument %d of %qE, which expects"
+ " the value %wd", actual, argno + 1, fndecl, min);
+ else
+ error_at (location, "passing %wd to argument %d of %qE, which expects"
+ " a value in the range [%wd, %wd]", actual, argno + 1, fndecl,
+ min, max);
}
/* Report that LOCATION has a call to FNDECL in which argument ARGNO has
return true;
/* Handle direct reads of global state. */
- return flags & (CP_READ_MEMORY | CP_READ_FFR);
+ return flags & (CP_READ_MEMORY | CP_READ_FFR | CP_READ_ZA);
}
/* Return true if calls to the function could modify some form of
return true;
/* Handle direct modifications of global state. */
- return flags & (CP_WRITE_MEMORY | CP_WRITE_FFR);
+ return flags & (CP_WRITE_MEMORY | CP_WRITE_FFR | CP_WRITE_ZA);
}
/* Return true if calls to the function could raise a signal. */
return value->instance == key;
}
-sve_switcher::sve_switcher ()
- : aarch64_simd_switcher (AARCH64_FL_F16 | AARCH64_FL_SVE)
+sve_switcher::sve_switcher (aarch64_feature_flags flags)
+ : aarch64_simd_switcher (AARCH64_FL_F16 | AARCH64_FL_SVE | flags)
{
/* Changing the ISA flags and have_regs_of_mode should be enough here.
We shouldn't need to pay the compile-time cost of a full target
function_builder::get_name (const function_instance &instance,
bool overloaded_p)
{
+ /* __arm_* functions are listed as arm_*, so that the associated GCC
+ code is not in the implementation namespace. */
+ if (strncmp (instance.base_name, "arm_", 4) == 0)
+ append_name ("__");
append_name (instance.base_name);
if (overloaded_p)
switch (instance.displacement_units ())
return tree_cons (get_identifier (name), NULL_TREE, attrs);
}
-/* Return the appropriate function attributes for INSTANCE. */
+/* Add attribute NS::NAME to ATTRS. */
+static tree
+add_attribute (const char *ns, const char *name, tree value, tree attrs)
+{
+ return tree_cons (build_tree_list (get_identifier (ns),
+ get_identifier (name)),
+ value, attrs);
+}
+
+/* Attribute arm::NAME describes shared state that is an input if IS_IN
+ and an output if IS_OUT. Check whether a call with call properties
+ CALL_FLAGS needs such an attribute. Add it to in-progress attribute
+ list ATTRS if so. Return the new attribute list. */
+static tree
+add_shared_state_attribute (const char *name, bool is_in, bool is_out,
+ unsigned int call_flags, tree attrs)
+{
+ struct state_flag_info
+ {
+ const char *name;
+ unsigned int read_flag;
+ unsigned int write_flag;
+ };
+ static state_flag_info state_flags[] =
+ {
+ { "za", CP_READ_ZA, CP_WRITE_ZA }
+ };
+
+ tree args = NULL_TREE;
+ for (const auto &state_flag : state_flags)
+ {
+ auto all_flags = state_flag.read_flag | state_flag.write_flag;
+ auto these_flags = ((is_in ? state_flag.read_flag : 0)
+ | (is_out ? state_flag.write_flag : 0));
+ if ((call_flags & all_flags) == these_flags)
+ {
+ tree value = build_string (strlen (state_flag.name) + 1,
+ state_flag.name);
+ args = tree_cons (NULL_TREE, value, args);
+ }
+ }
+ if (args)
+ attrs = add_attribute ("arm", name, args, attrs);
+ return attrs;
+}
+
+/* Return the appropriate function attributes for INSTANCE, which requires
+ the feature flags in REQUIRED_EXTENSIONS. */
tree
-function_builder::get_attributes (const function_instance &instance)
+function_builder::get_attributes (const function_instance &instance,
+ aarch64_feature_flags required_extensions)
{
tree attrs = NULL_TREE;
+ if (required_extensions & AARCH64_FL_SM_ON)
+ attrs = add_attribute ("arm", "streaming", NULL_TREE, attrs);
+ else if (!(required_extensions & AARCH64_FL_SM_OFF))
+ attrs = add_attribute ("arm", "streaming_compatible", NULL_TREE, attrs);
+
+ attrs = add_shared_state_attribute ("in", true, false,
+ instance.call_properties (), attrs);
+ attrs = add_shared_state_attribute ("out", false, true,
+ instance.call_properties (), attrs);
+ attrs = add_shared_state_attribute ("inout", true, true,
+ instance.call_properties (), attrs);
+
if (!instance.modifies_global_state_p ())
{
if (instance.reads_global_state_p ())
tree fntype = build_function_type_array (return_type,
argument_types.length (),
argument_types.address ());
- tree attrs = get_attributes (instance);
+ tree attrs = get_attributes (instance, required_extensions);
registered_function &rfn = add_function (instance, name, fntype, attrs,
required_extensions, false, false);
if (strcmp (name, overload_name) != 0)
{
/* Attribute lists shouldn't be shared. */
- tree attrs = get_attributes (instance);
+ tree attrs = get_attributes (instance, required_extensions);
bool placeholder_p = !(m_direct_overloads || force_direct_overloads);
add_function (instance, overload_name, fntype, attrs,
required_extensions, false, placeholder_p);
function_resolver::check_gp_argument (unsigned int nops,
unsigned int &i, unsigned int &nargs)
{
+ gcc_assert (pred != PRED_za_m);
i = 0;
if (pred != PRED_none)
{
unsigned int nargs, tree *args)
: function_call_info (location, instance, fndecl),
m_fntype (fntype), m_nargs (nargs), m_args (args),
- /* We don't have to worry about unary _m operations here, since they
- never have arguments that need checking. */
- m_base_arg (pred != PRED_none ? 1 : 0)
+ m_base_arg (pred != PRED_none && pred != PRED_za_m ? 1 : 0)
{
}
}
/* Return the base address for a contiguous load or store function.
- MEM_MODE is the mode of the addressed memory. */
+ MEM_MODE is the mode of the addressed memory, BASE_ARGNO is
+ the index of the base argument, and VNUM_ARGNO is the index of
+ the vnum offset argument (if any). VL_ISA_MODE is AARCH64_FL_SM_ON
+ if the vnum argument is a factor of the SME vector length, 0 if it
+ is a factor of the current prevailing vector length. */
rtx
-function_expander::get_contiguous_base (machine_mode mem_mode)
+function_expander::get_contiguous_base (machine_mode mem_mode,
+ unsigned int base_argno,
+ unsigned int vnum_argno,
+ aarch64_feature_flags vl_isa_mode)
{
- rtx base = convert_to_pmode (args[1]);
+ rtx base = convert_to_pmode (args[base_argno]);
if (mode_suffix_id == MODE_vnum)
{
- /* Use the size of the memory mode for extending loads and truncating
- stores. Use the size of a full vector for non-extending loads
- and non-truncating stores (including svld[234] and svst[234]). */
- poly_int64 size = ordered_min (GET_MODE_SIZE (mem_mode),
- BYTES_PER_SVE_VECTOR);
- rtx offset = gen_int_mode (size, Pmode);
- offset = simplify_gen_binary (MULT, Pmode, args[2], offset);
- base = simplify_gen_binary (PLUS, Pmode, base, offset);
+ rtx vnum = args[vnum_argno];
+ if (vnum != const0_rtx)
+ {
+ /* Use the size of the memory mode for extending loads and truncating
+ stores. Use the size of a full vector for non-extending loads
+ and non-truncating stores (including svld[234] and svst[234]). */
+ poly_int64 size = ordered_min (GET_MODE_SIZE (mem_mode),
+ BYTES_PER_SVE_VECTOR);
+ rtx offset;
+ if ((vl_isa_mode & AARCH64_FL_SM_ON)
+ && !TARGET_STREAMING
+ && !size.is_constant ())
+ {
+ gcc_assert (known_eq (size, BYTES_PER_SVE_VECTOR));
+ if (CONST_INT_P (vnum) && IN_RANGE (INTVAL (vnum), -32, 31))
+ offset = aarch64_sme_vq_immediate (Pmode, INTVAL (vnum) * 16,
+ AARCH64_ISA_MODE);
+ else
+ {
+ offset = aarch64_sme_vq_immediate (Pmode, 16,
+ AARCH64_ISA_MODE);
+ offset = simplify_gen_binary (MULT, Pmode, vnum, offset);
+ }
+ }
+ else
+ {
+ offset = gen_int_mode (size, Pmode);
+ offset = simplify_gen_binary (MULT, Pmode, vnum, offset);
+ }
+ base = simplify_gen_binary (PLUS, Pmode, base, offset);
+ }
}
return base;
}
machine_mode mode = operand.mode;
if (mode == VOIDmode)
{
- /* The only allowable use of VOIDmode is the wildcard
- aarch64_any_register_operand, which is used to avoid
- combinatorial explosion in the reinterpret patterns. */
- gcc_assert (operand.predicate == aarch64_any_register_operand);
- mode = GET_MODE (x);
+ /* The only allowable uses of VOIDmode are:
+
+ - the wildcard aarch64_any_register_operand, which is used
+ to avoid combinatorial explosion in the reinterpret patterns
+
+ - pmode_register_operand, which always has mode Pmode. */
+ if (operand.predicate == aarch64_any_register_operand)
+ mode = GET_MODE (x);
+ else if (operand.predicate == pmode_register_operand)
+ mode = Pmode;
+ else
+ gcc_unreachable ();
}
else if (!VECTOR_MODE_P (GET_MODE (x)) && VECTOR_MODE_P (mode))
x = expand_vector_broadcast (mode, x);
/* Add an integer operand with value X to the instruction. */
void
-function_expander::add_integer_operand (HOST_WIDE_INT x)
+function_expander::add_integer_operand (poly_int64 x)
{
m_ops.safe_grow (m_ops.length () + 1, true);
create_integer_operand (&m_ops.last (), x);
sve_switcher sve;
register_builtin_types ();
if (in_lto_p)
- handle_arm_sve_h ();
+ {
+ handle_arm_sve_h ();
+ handle_arm_sme_h ();
+ }
}
/* Register vector type TYPE under its arm_sve.h name. */
function_table = new hash_table<registered_function_hasher> (1023);
function_builder builder;
for (unsigned int i = 0; i < ARRAY_SIZE (function_groups); ++i)
- builder.register_function_group (function_groups[i]);
+ if (!(function_groups[i].required_extensions & AARCH64_FL_SME))
+ builder.register_function_group (function_groups[i]);
}
/* Return the function decl with SVE function subcode CODE, or error_mark_node
return (*registered_functions)[code]->decl;
}
+/* Implement #pragma GCC aarch64 "arm_sme.h". */
+void
+handle_arm_sme_h ()
+{
+ if (!function_table)
+ {
+ error ("%qs defined without first defining %qs",
+ "arm_sme.h", "arm_sve.h");
+ return;
+ }
+
+ static bool initialized_p;
+ if (initialized_p)
+ {
+ error ("duplicate definition of %qs", "arm_sme.h");
+ return;
+ }
+ initialized_p = true;
+
+ sme_switcher sme;
+
+ function_builder builder;
+ for (unsigned int i = 0; i < ARRAY_SIZE (function_groups); ++i)
+ if (function_groups[i].required_extensions & AARCH64_FL_SME)
+ builder.register_function_group (function_groups[i]);
+}
+
/* If we're implementing manual overloading, check whether the SVE
function with subcode CODE is overloaded, and if so attempt to
determine the corresponding non-overloaded function. The call
#define DEF_SVE_TYPE_SUFFIX(A, B, C, D, E)
#endif
+#ifndef DEF_SME_ZA_SUFFIX
+#define DEF_SME_ZA_SUFFIX(A, B, C)
+#endif
+
#ifndef DEF_SVE_GROUP_SUFFIX
#define DEF_SVE_GROUP_SUFFIX(A, B, C)
#endif
DEF_SVE_FUNCTION_GS (NAME, SHAPE, TYPES, none, PREDS)
#endif
+#ifndef DEF_SME_ZA_FUNCTION_GS
+#define DEF_SME_ZA_FUNCTION_GS(NAME, SHAPE, TYPES, GROUP, PREDS) \
+ DEF_SVE_FUNCTION_GS(NAME, SHAPE, TYPES, GROUP, PREDS)
+#endif
+
+#ifndef DEF_SME_ZA_FUNCTION
+#define DEF_SME_ZA_FUNCTION(NAME, SHAPE, TYPES, PREDS) \
+ DEF_SME_ZA_FUNCTION_GS (NAME, SHAPE, TYPES, none, PREDS)
+#endif
+
DEF_SVE_MODE (n, none, none, none)
DEF_SVE_MODE (index, none, none, elements)
DEF_SVE_MODE (offset, none, none, bytes)
DEF_SVE_TYPE_SUFFIX (u32, svuint32_t, unsigned, 32, VNx4SImode)
DEF_SVE_TYPE_SUFFIX (u64, svuint64_t, unsigned, 64, VNx2DImode)
+/* Associate _za with bytes. This is needed for svldr_vnum_za and
+ svstr_vnum_za, whose ZA offset can be in the range [0, 15], as for za8. */
+DEF_SME_ZA_SUFFIX (za, 8, VNx16QImode)
+
+DEF_SME_ZA_SUFFIX (za8, 8, VNx16QImode)
+DEF_SME_ZA_SUFFIX (za16, 16, VNx8HImode)
+DEF_SME_ZA_SUFFIX (za32, 32, VNx4SImode)
+DEF_SME_ZA_SUFFIX (za64, 64, VNx2DImode)
+DEF_SME_ZA_SUFFIX (za128, 128, VNx1TImode)
+
DEF_SVE_GROUP_SUFFIX (x2, 0, 2)
DEF_SVE_GROUP_SUFFIX (x3, 0, 3)
DEF_SVE_GROUP_SUFFIX (x4, 0, 4)
#include "aarch64-sve-builtins-base.def"
#include "aarch64-sve-builtins-sve2.def"
+#include "aarch64-sve-builtins-sme.def"
+#undef DEF_SME_ZA_FUNCTION
#undef DEF_SVE_FUNCTION
+#undef DEF_SME_ZA_FUNCTION_GS
#undef DEF_SVE_FUNCTION_GS
#undef DEF_SVE_GROUP_SUFFIX
+#undef DEF_SME_ZA_SUFFIX
#undef DEF_SVE_TYPE_SUFFIX
#undef DEF_SVE_TYPE
#undef DEF_SVE_MODE
const unsigned int CP_WRITE_MEMORY = 1U << 4;
const unsigned int CP_READ_FFR = 1U << 5;
const unsigned int CP_WRITE_FFR = 1U << 6;
+const unsigned int CP_READ_ZA = 1U << 7;
+const unsigned int CP_WRITE_ZA = 1U << 8;
/* Enumerates the SVE predicate and (data) vector types, together called
"vector types" for brevity. */
/* Zero predication: set inactive lanes of the vector result to zero. */
PRED_z,
+ /* Merging predication for SME's ZA: merge into slices of the array
+ instead of overwriting the whole slices. */
+ PRED_za_m,
+
NUM_PREDS
};
{
#define DEF_SVE_TYPE_SUFFIX(NAME, ACLE_TYPE, CLASS, BITS, MODE) \
TYPE_SUFFIX_ ## NAME,
+#define DEF_SME_ZA_SUFFIX(NAME, BITS, MODE) \
+ TYPE_SUFFIX_ ## NAME,
#include "aarch64-sve-builtins.def"
NUM_TYPE_SUFFIXES
};
unsigned int unsigned_p : 1;
/* True if the suffix is for a floating-point type. */
unsigned int float_p : 1;
+ /* True if the suffix is for a vector type (integer or float). */
+ unsigned int vector_p : 1;
/* True if the suffix is for a boolean type. */
unsigned int bool_p : 1;
- unsigned int spare : 12;
+ /* True if the suffix is for SME's ZA. */
+ unsigned int za_p : 1;
+ unsigned int spare : 10;
/* The associated vector or predicate mode. */
machine_mode vector_mode : 16;
tree displacement_vector_type () const;
units_index displacement_units () const;
+ unsigned int num_za_tiles () const;
+
const type_suffix_info &type_suffix (unsigned int) const;
const group_suffix_info &group_suffix () const;
tree scalar_type (unsigned int) const;
tree vector_type (unsigned int) const;
tree tuple_type (unsigned int) const;
- unsigned int elements_per_vq (unsigned int i) const;
+ unsigned int elements_per_vq (unsigned int) const;
machine_mode vector_mode (unsigned int) const;
machine_mode tuple_mode (unsigned int) const;
machine_mode gp_mode (unsigned int) const;
char *get_name (const function_instance &, bool);
- tree get_attributes (const function_instance &);
+ tree get_attributes (const function_instance &, aarch64_feature_flags);
registered_function &add_function (const function_instance &,
const char *, tree, tree,
bool overlaps_input_p (rtx);
rtx convert_to_pmode (rtx);
- rtx get_contiguous_base (machine_mode);
+ rtx get_contiguous_base (machine_mode, unsigned int = 1, unsigned int = 2,
+ aarch64_feature_flags = 0);
rtx get_fallback_value (machine_mode, unsigned int,
unsigned int, unsigned int &);
rtx get_reg_target ();
void add_output_operand (insn_code);
void add_input_operand (insn_code, rtx);
- void add_integer_operand (HOST_WIDE_INT);
+ void add_integer_operand (poly_int64);
void add_mem_operand (machine_mode, rtx);
void add_address_operand (rtx);
void add_fixed_operand (rtx);
class sve_switcher : public aarch64_simd_switcher
{
public:
- sve_switcher ();
+ sve_switcher (aarch64_feature_flags = 0);
~sve_switcher ();
private:
bool m_old_have_regs_of_mode[MAX_MACHINE_MODE];
};
+/* Extends sve_switch enough for defining arm_sme.h. */
+class sme_switcher : public sve_switcher
+{
+public:
+ sme_switcher () : sve_switcher (AARCH64_FL_SME) {}
+};
+
extern const type_suffix_info type_suffixes[NUM_TYPE_SUFFIXES + 1];
extern const mode_suffix_info mode_suffixes[MODE_none + 1];
extern const group_suffix_info group_suffixes[NUM_GROUP_SUFFIXES];
-extern tree scalar_types[NUM_VECTOR_TYPES];
+extern tree scalar_types[NUM_VECTOR_TYPES + 1];
extern tree acle_vector_types[MAX_TUPLE_SIZE][NUM_VECTOR_TYPES + 1];
extern tree acle_svpattern;
extern tree acle_svprfop;
return acle_vector_types[0][mode_suffix ().displacement_vector_type];
}
+/* Return the number of ZA tiles associated with the _za<N> suffix
+ (which is always the first type suffix). */
+inline unsigned int
+function_instance::num_za_tiles () const
+{
+ auto &suffix = type_suffix (0);
+ gcc_checking_assert (suffix.za_p);
+ return suffix.element_bytes;
+}
+
/* If the function takes a vector or scalar displacement, return the units
in which the displacement is measured, otherwise return UNITS_none. */
inline units_index
}
/* Return true if a single RDVL instruction can multiply FACTOR by the
- number of 128-bit quadwords in an SVE vector. */
+ number of 128-bit quadwords in an SVE vector. This is also the
+ range of ADDVL. */
static bool
-aarch64_sve_rdvl_factor_p (HOST_WIDE_INT factor)
+aarch64_sve_rdvl_addvl_factor_p (HOST_WIDE_INT factor)
{
return (multiple_p (factor, 16)
&& IN_RANGE (factor, -32 * 16, 31 * 16));
}
+/* Return true if ADDPL can be used to add FACTOR multiplied by the number
+ of quadwords in an SVE vector. */
+
+static bool
+aarch64_sve_addpl_factor_p (HOST_WIDE_INT factor)
+{
+ return (multiple_p (factor, 2)
+ && IN_RANGE (factor, -32 * 2, 31 * 2));
+}
+
/* Return true if we can move VALUE into a register using a single
RDVL instruction. */
aarch64_sve_rdvl_immediate_p (poly_int64 value)
{
HOST_WIDE_INT factor = value.coeffs[0];
- return value.coeffs[1] == factor && aarch64_sve_rdvl_factor_p (factor);
+ return value.coeffs[1] == factor && aarch64_sve_rdvl_addvl_factor_p (factor);
}
/* Likewise for rtx X. */
HOST_WIDE_INT factor = value.coeffs[0];
if (factor == 0 || value.coeffs[1] != factor)
return false;
- /* FACTOR counts VG / 2, so a value of 2 is one predicate width
- and a value of 16 is one vector width. */
- return (((factor & 15) == 0 && IN_RANGE (factor, -32 * 16, 31 * 16))
- || ((factor & 1) == 0 && IN_RANGE (factor, -32 * 2, 31 * 2)));
+ return (aarch64_sve_rdvl_addvl_factor_p (factor)
+ || aarch64_sve_addpl_factor_p (factor));
}
/* Likewise for rtx X. */
number of 128-bit quadwords in an SME vector. ISA_MODE is the
ISA mode in which the calculation is being performed. */
-static rtx
+rtx
aarch64_sme_vq_immediate (machine_mode mode, HOST_WIDE_INT factor,
aarch64_feature_flags isa_mode)
{
- gcc_assert (aarch64_sve_rdvl_factor_p (factor));
+ gcc_assert (aarch64_sve_rdvl_addvl_factor_p (factor));
if (isa_mode & AARCH64_FL_SM_ON)
/* We're in streaming mode, so we can use normal poly-int values. */
return gen_int_mode ({ factor, factor }, mode);
{
HOST_WIDE_INT factor;
return (aarch64_sme_vq_unspec_p (x, &factor)
- && aarch64_sve_rdvl_factor_p (factor));
+ && aarch64_sve_rdvl_addvl_factor_p (factor));
}
/* Return the asm string for an RDSVL instruction that calculates X,
return buffer;
}
+/* Return true if X is a constant that can be added using ADDSVL or ADDSPL. */
+
+bool
+aarch64_addsvl_addspl_immediate_p (const_rtx x)
+{
+ HOST_WIDE_INT factor;
+ return (aarch64_sme_vq_unspec_p (x, &factor)
+ && (aarch64_sve_rdvl_addvl_factor_p (factor)
+ || aarch64_sve_addpl_factor_p (factor)));
+}
+
+/* X is a constant that satisfies aarch64_addsvl_addspl_immediate_p.
+ Return the asm string for the associated instruction. */
+
+char *
+aarch64_output_addsvl_addspl (rtx x)
+{
+ static char buffer[sizeof ("addspl\t%x0, %x1, #-") + 3 * sizeof (int)];
+ HOST_WIDE_INT factor;
+ if (!aarch64_sme_vq_unspec_p (x, &factor))
+ gcc_unreachable ();
+ if (aarch64_sve_rdvl_addvl_factor_p (factor))
+ snprintf (buffer, sizeof (buffer), "addsvl\t%%x0, %%x1, #%d",
+ (int) factor / 16);
+ else if (aarch64_sve_addpl_factor_p (factor))
+ snprintf (buffer, sizeof (buffer), "addspl\t%%x0, %%x1, #%d",
+ (int) factor / 2);
+ else
+ gcc_unreachable ();
+ return buffer;
+}
+
/* Multipliers for repeating bitmasks of width 32, 16, 8, 4, and 2. */
static const unsigned HOST_WIDE_INT bitmask_imm_mul[] =
{
/* Try to use an unshifted CNT[BHWD] or RDVL. */
if (aarch64_sve_cnt_factor_p (factor)
- || aarch64_sve_rdvl_factor_p (factor))
+ || aarch64_sve_rdvl_addvl_factor_p (factor))
{
val = gen_int_mode (poly_int64 (factor, factor), mode);
shift = 0;
&& contains_reg_of_mode[GENERAL_REGS][GET_MODE (SUBREG_REG (index))])
index = SUBREG_REG (index);
- if (aarch64_sve_data_mode_p (mode))
+ if (aarch64_sve_data_mode_p (mode) || mode == VNx1TImode)
{
if (type != ADDRESS_REG_REG
|| (1 << shift) != GET_MODE_UNIT_SIZE (mode))
&& ((vec_flags == 0
&& known_lt (GET_MODE_SIZE (mode), 16))
|| vec_flags == VEC_ADVSIMD
- || vec_flags & VEC_SVE_DATA));
+ || vec_flags & VEC_SVE_DATA
+ || mode == VNx1TImode));
/* For SVE, only accept [Rn], [Rn, #offset, MUL VL] and [Rn, Rm, LSL #shift].
The latter is not valid for SVE predicates, and that's rejected through
/* Make "m" use the LD1 offset range for SVE data modes, so
that pre-RTL optimizers like ivopts will work to that
instead of the wider LDR/STR range. */
- if (vec_flags == VEC_SVE_DATA)
+ if (vec_flags == VEC_SVE_DATA || mode == VNx1TImode)
return (type == ADDR_QUERY_M
? offset_4bit_signed_scaled_p (mode, offset)
: offset_9bit_signed_scaled_p (mode, offset));
return "";
}
+/* Return the asm string for an SME ZERO instruction whose 8-bit mask
+ operand is MASK. */
+const char *
+aarch64_output_sme_zero_za (rtx mask)
+{
+ auto mask_val = UINTVAL (mask);
+ if (mask_val == 0)
+ return "zero\t{}";
+
+ if (mask_val == 0xff)
+ return "zero\t{ za }";
+
+ static constexpr std::pair<unsigned int, char> tiles[] = {
+ { 0xff, 'b' },
+ { 0x55, 'h' },
+ { 0x11, 's' },
+ { 0x01, 'd' }
+ };
+ /* The last entry in the list has the form "za7.d }", but that's the
+ same length as "za7.d, ". */
+ static char buffer[sizeof("zero\t{ ") + sizeof ("za7.d, ") * 8 + 1];
+ unsigned int i = 0;
+ i += snprintf (buffer + i, sizeof (buffer) - i, "zero\t");
+ const char *prefix = "{ ";
+ for (auto &tile : tiles)
+ {
+ auto tile_mask = tile.first;
+ unsigned int tile_index = 0;
+ while (tile_mask < 0x100)
+ {
+ if ((mask_val & tile_mask) == tile_mask)
+ {
+ i += snprintf (buffer + i, sizeof (buffer) - i, "%sza%d.%c",
+ prefix, tile_index, tile.second);
+ prefix = ", ";
+ mask_val &= ~tile_mask;
+ }
+ tile_mask <<= 1;
+ tile_index += 1;
+ }
+ }
+ gcc_assert (mask_val == 0 && i + 3 <= sizeof (buffer));
+ snprintf (buffer + i, sizeof (buffer) - i, " }");
+ return buffer;
+}
/* Return size in bits of an arithmetic operand which is shifted/scaled and
masked such that it is suitable for a UXTB, UXTH, or UXTW extend
&& offset_4bit_signed_scaled_p (SVE_BYTE_MODE, last));
}
+/* Return true if OFFSET is a constant integer and if VNUM is
+ OFFSET * the number of bytes in an SVE vector. This is the requirement
+ that exists in SME LDR and STR instructions, where the VL offset must
+ equal the ZA slice offset. */
+bool
+aarch64_sme_ldr_vnum_offset_p (rtx offset, rtx vnum)
+{
+ if (!CONST_INT_P (offset) || !IN_RANGE (INTVAL (offset), 0, 15))
+ return false;
+
+ if (TARGET_STREAMING)
+ {
+ poly_int64 const_vnum;
+ return (poly_int_rtx_p (vnum, &const_vnum)
+ && known_eq (const_vnum,
+ INTVAL (offset) * BYTES_PER_SVE_VECTOR));
+ }
+ else
+ {
+ HOST_WIDE_INT factor;
+ return (aarch64_sme_vq_unspec_p (vnum, &factor)
+ && factor == INTVAL (offset) * 16);
+ }
+}
+
/* Emit a register copy from operand to operand, taking care not to
early-clobber source registers in the process.
/* Macros to test ISA flags. */
#define AARCH64_ISA_SM_OFF (aarch64_isa_flags & AARCH64_FL_SM_OFF)
+#define AARCH64_ISA_SM_ON (aarch64_isa_flags & AARCH64_FL_SM_ON)
#define AARCH64_ISA_ZA_ON (aarch64_isa_flags & AARCH64_FL_ZA_ON)
#define AARCH64_ISA_MODE (aarch64_isa_flags & AARCH64_FL_ISA_MODES)
#define AARCH64_ISA_CRC (aarch64_isa_flags & AARCH64_FL_CRC)
#define AARCH64_ISA_SVE2_SHA3 (aarch64_isa_flags & AARCH64_FL_SVE2_SHA3)
#define AARCH64_ISA_SVE2_SM4 (aarch64_isa_flags & AARCH64_FL_SVE2_SM4)
#define AARCH64_ISA_SME (aarch64_isa_flags & AARCH64_FL_SME)
+#define AARCH64_ISA_SME_I16I64 (aarch64_isa_flags & AARCH64_FL_SME_I16I64)
+#define AARCH64_ISA_SME_F64F64 (aarch64_isa_flags & AARCH64_FL_SME_F64F64)
#define AARCH64_ISA_V8_3A (aarch64_isa_flags & AARCH64_FL_V8_3A)
#define AARCH64_ISA_DOTPROD (aarch64_isa_flags & AARCH64_FL_DOTPROD)
#define AARCH64_ISA_AES (aarch64_isa_flags & AARCH64_FL_AES)
/* The current function is a normal non-streaming function. */
#define TARGET_NON_STREAMING (AARCH64_ISA_SM_OFF)
+/* The current function has a streaming body. */
+#define TARGET_STREAMING (AARCH64_ISA_SM_ON)
+
/* The current function has a streaming-compatible body. */
#define TARGET_STREAMING_COMPATIBLE \
((aarch64_isa_flags & AARCH64_FL_SM_STATE) == 0)
imply anything about the state of PSTATE.SM. */
#define TARGET_SME (AARCH64_ISA_SME)
+/* Streaming-mode SME instructions. */
+#define TARGET_STREAMING_SME (TARGET_STREAMING && TARGET_SME)
+
+/* The FEAT_SME_I16I64 extension to SME, enabled through +sme-i16i64. */
+#define TARGET_SME_I16I64 (AARCH64_ISA_SME_I16I64)
+
+/* The FEAT_SME_F64F64 extension to SME, enabled through +sme-f64f64. */
+#define TARGET_SME_F64F64 (AARCH64_ISA_SME_F64F64)
+
/* ARMv8.3-A features. */
#define TARGET_ARMV8_3 (AARCH64_ISA_V8_3A)
[ r , rk , Uaa ; multiple , * ] #
[ r , 0 , Uai ; alu_imm , sve ] << aarch64_output_sve_scalar_inc_dec (operands[2]);
[ rk , rk , Uav ; alu_imm , sve ] << aarch64_output_sve_addvl_addpl (operands[2]);
+ [ rk , rk , UaV ; alu_imm , sme ] << aarch64_output_addsvl_addspl (operands[2]);
}
;; The "alu_imm" types for INC/DEC and ADDVL/ADDPL are just placeholders.
)
--- /dev/null
+/* AArch64 SME intrinsics include file.
+ Copyright (C) 2023 Free Software Foundation, Inc.
+
+ This file is part of GCC.
+
+ GCC is free software; you can redistribute it and/or modify it
+ under the terms of the GNU General Public License as published
+ by the Free Software Foundation; either version 3, or (at your
+ option) any later version.
+
+ GCC is distributed in the hope that it will be useful, but WITHOUT
+ ANY WARRANTY; without even the implied warranty of MERCHANTABILITY
+ or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public
+ License for more details.
+
+ Under Section 7 of GPL version 3, you are granted additional
+ permissions described in the GCC Runtime Library Exception, version
+ 3.1, as published by the Free Software Foundation.
+
+ You should have received a copy of the GNU General Public License and
+ a copy of the GCC Runtime Library Exception along with this program;
+ see the files COPYING3 and COPYING.RUNTIME respectively. If not, see
+ <http://www.gnu.org/licenses/>. */
+
+#ifndef _ARM_SME_H_
+#define _ARM_SME_H_
+
+#include <arm_sve.h>
+#pragma GCC aarch64 "arm_sme.h"
+
+void __arm_za_disable(void) __arm_streaming_compatible;
+
+void *__arm_sc_memcpy(void *, const void *, __SIZE_TYPE__)
+ __arm_streaming_compatible;
+
+void *__arm_sc_memmove(void *, const void *, __SIZE_TYPE__)
+ __arm_streaming_compatible;
+
+void *__arm_sc_memset(void *, int, __SIZE_TYPE__)
+ __arm_streaming_compatible;
+
+void *__arm_sc_memchr(void *, int, __SIZE_TYPE__)
+ __arm_streaming_compatible;
+
+#endif
(define_register_constraint "k" "STACK_REG"
"@internal The stack register.")
+(define_register_constraint "Ucj" "W12_W15_REGS"
+ "@internal r12-r15, which can be used to index ZA.")
+
(define_register_constraint "Ucs" "TAILCALL_ADDR_REGS"
"@internal Registers suitable for an indirect tail call")
a single ADDVL or ADDPL."
(match_operand 0 "aarch64_sve_addvl_addpl_immediate"))
+(define_constraint "UaV"
+ "@internal
+ A constraint that matches a VG-based constant that can be added by
+ a single ADDSVL or ADDSPL."
+ (match_operand 0 "aarch64_addsvl_addspl_immediate"))
+
(define_constraint "Uat"
"@internal
A constraint that matches a VG-based constant that can be added by
(define_mode_iterator VNx4SF_ONLY [VNx4SF])
(define_mode_iterator VNx2DI_ONLY [VNx2DI])
(define_mode_iterator VNx2DF_ONLY [VNx2DF])
+(define_mode_iterator VNx1TI_ONLY [VNx1TI])
;; All fully-packed SVE vector modes.
(define_mode_iterator SVE_FULL [VNx16QI VNx8HI VNx4SI VNx2DI
;; Bfloat16 modes to which V4SF can be converted
(define_mode_iterator V4SF_TO_BF [V4BF V8BF])
+;; The modes used to represent different ZA access sizes.
+(define_mode_iterator SME_ZA_I [VNx16QI VNx8HI VNx4SI VNx2DI VNx1TI])
+(define_mode_iterator SME_ZA_SDI [VNx4SI (VNx2DI "TARGET_SME_I16I64")])
+
+(define_mode_iterator SME_ZA_SDF_I [VNx4SI (VNx2DI "TARGET_SME_F64F64")])
+
+;; The modes for which outer product instructions are supported.
+(define_mode_iterator SME_MOP_BHI [VNx16QI (VNx8HI "TARGET_SME_I16I64")])
+(define_mode_iterator SME_MOP_HSDF [VNx8BF VNx8HF VNx4SF
+ (VNx2DF "TARGET_SME_F64F64")])
+
;; ------------------------------------------------------------------
;; Unspec enumerations for Advance SIMD. These could well go into
;; aarch64.md but for their use in int_iterators here.
UNSPEC_BFCVTN2 ; Used in aarch64-simd.md.
UNSPEC_BFCVT ; Used in aarch64-simd.md.
UNSPEC_FCVTXN ; Used in aarch64-simd.md.
+
+ ;; All used in aarch64-sme.md
+ UNSPEC_SME_ADDHA
+ UNSPEC_SME_ADDVA
+ UNSPEC_SME_FMOPA
+ UNSPEC_SME_FMOPS
+ UNSPEC_SME_LD1_HOR
+ UNSPEC_SME_LD1_VER
+ UNSPEC_SME_READ_HOR
+ UNSPEC_SME_READ_VER
+ UNSPEC_SME_SMOPA
+ UNSPEC_SME_SMOPS
+ UNSPEC_SME_ST1_HOR
+ UNSPEC_SME_ST1_VER
+ UNSPEC_SME_SUMOPA
+ UNSPEC_SME_SUMOPS
+ UNSPEC_SME_UMOPA
+ UNSPEC_SME_UMOPS
+ UNSPEC_SME_USMOPA
+ UNSPEC_SME_USMOPS
+ UNSPEC_SME_WRITE_HOR
+ UNSPEC_SME_WRITE_VER
])
;; ------------------------------------------------------------------
;; element.
(define_mode_attr elem_bits [(VNx16BI "8") (VNx8BI "16")
(VNx4BI "32") (VNx2BI "64")
- (VNx16QI "8") (VNx8HI "16")
- (VNx4SI "32") (VNx2DI "64")
- (VNx8HF "16") (VNx4SF "32") (VNx2DF "64")])
+ (VNx16QI "8") (VNx32QI "8") (VNx64QI "8")
+ (VNx8HI "16") (VNx16HI "16") (VNx32HI "16")
+ (VNx8HF "16") (VNx16HF "16") (VNx32HF "16")
+ (VNx8BF "16") (VNx16BF "16") (VNx32BF "16")
+ (VNx4SI "32") (VNx8SI "32") (VNx16SI "32")
+ (VNx4SF "32") (VNx8SF "32") (VNx16SF "32")
+ (VNx2DI "64") (VNx4DI "64") (VNx8DI "64")
+ (VNx2DF "64") (VNx4DF "64") (VNx8DF "64")
+ (VNx1TI "128")])
;; The number of bits in a vector container.
(define_mode_attr container_bits [(VNx16QI "8")
(VNx4SF "s") (VNx2SF "s")
(VNx2DI "d")
(VNx2DF "d")
+ (VNx1TI "q")
(BF "h") (V4BF "h") (V8BF "h")
(HF "h")
(SF "s") (DF "d")
(VNx4SF "w") (VNx2SF "w")
(VNx2DI "d")
(VNx2DF "d")
+ (VNx1TI "q")
(VNx32QI "b") (VNx48QI "b") (VNx64QI "b")
(VNx16HI "h") (VNx24HI "h") (VNx32HI "h")
(VNx16HF "h") (VNx24HF "h") (VNx32HF "h")
(VNx4SF "VNx4BI") (VNx2SF "VNx2BI")
(VNx2DI "VNx2BI")
(VNx2DF "VNx2BI")
+ (VNx1TI "VNx2BI")
(VNx32QI "VNx16BI")
(VNx16HI "VNx8BI") (VNx16HF "VNx8BI")
(VNx16BF "VNx8BI")
(V8HI "vec") (V2SI "vec") (V4SI "vec")
(V2DI "vec") (DI "offset")])
+(define_mode_attr b [(VNx8BF "b") (VNx8HF "") (VNx4SF "") (VNx2DF "")])
+
;; -------------------------------------------------------------------
;; Code Iterators
;; -------------------------------------------------------------------
(define_int_iterator FCMUL_OP [UNSPEC_FCMUL
UNSPEC_FCMUL_CONJ])
+(define_int_iterator SME_LD1 [UNSPEC_SME_LD1_HOR UNSPEC_SME_LD1_VER])
+(define_int_iterator SME_READ [UNSPEC_SME_READ_HOR UNSPEC_SME_READ_VER])
+(define_int_iterator SME_ST1 [UNSPEC_SME_ST1_HOR UNSPEC_SME_ST1_VER])
+(define_int_iterator SME_WRITE [UNSPEC_SME_WRITE_HOR UNSPEC_SME_WRITE_VER])
+
+(define_int_iterator SME_BINARY_SDI [UNSPEC_SME_ADDHA UNSPEC_SME_ADDVA])
+
+(define_int_iterator SME_INT_MOP [UNSPEC_SME_SMOPA UNSPEC_SME_SMOPS
+ UNSPEC_SME_SUMOPA UNSPEC_SME_SUMOPS
+ UNSPEC_SME_UMOPA UNSPEC_SME_UMOPS
+ UNSPEC_SME_USMOPA UNSPEC_SME_USMOPS])
+
+(define_int_iterator SME_FP_MOP [UNSPEC_SME_FMOPA UNSPEC_SME_FMOPS])
+
;; Iterators for atomic operations.
(define_int_iterator ATOMIC_LDOP
(UNSPEC_PMULLT "pmullt")
(UNSPEC_PMULLT_PAIR "pmullt_pair")
(UNSPEC_SMATMUL "smatmul")
+ (UNSPEC_SME_ADDHA "addha")
+ (UNSPEC_SME_ADDVA "addva")
+ (UNSPEC_SME_FMOPA "fmopa")
+ (UNSPEC_SME_FMOPS "fmops")
+ (UNSPEC_SME_LD1_HOR "ld1_hor")
+ (UNSPEC_SME_LD1_VER "ld1_ver")
+ (UNSPEC_SME_READ_HOR "read_hor")
+ (UNSPEC_SME_READ_VER "read_ver")
+ (UNSPEC_SME_SMOPA "smopa")
+ (UNSPEC_SME_SMOPS "smops")
+ (UNSPEC_SME_ST1_HOR "st1_hor")
+ (UNSPEC_SME_ST1_VER "st1_ver")
+ (UNSPEC_SME_SUMOPA "sumopa")
+ (UNSPEC_SME_SUMOPS "sumops")
+ (UNSPEC_SME_UMOPA "umopa")
+ (UNSPEC_SME_UMOPS "umops")
+ (UNSPEC_SME_USMOPA "usmopa")
+ (UNSPEC_SME_USMOPS "usmops")
+ (UNSPEC_SME_WRITE_HOR "write_hor")
+ (UNSPEC_SME_WRITE_VER "write_ver")
(UNSPEC_SQCADD90 "sqcadd90")
(UNSPEC_SQCADD270 "sqcadd270")
(UNSPEC_SQRDCMLAH "sqrdcmlah")
(define_int_attr unspec [(UNSPEC_WHILERW "UNSPEC_WHILERW")
(UNSPEC_WHILEWR "UNSPEC_WHILEWR")])
+(define_int_attr hv [(UNSPEC_SME_LD1_HOR "h")
+ (UNSPEC_SME_LD1_VER "v")
+ (UNSPEC_SME_READ_HOR "h")
+ (UNSPEC_SME_READ_VER "v")
+ (UNSPEC_SME_ST1_HOR "h")
+ (UNSPEC_SME_ST1_VER "v")
+ (UNSPEC_SME_WRITE_HOR "h")
+ (UNSPEC_SME_WRITE_VER "v")])
+
;; Iterators and attributes for fpcr fpsr getter setters
(define_int_iterator GET_FPSCR
(and (match_code "const_poly_int")
(match_test "aarch64_add_offset_temporaries (op) == 1")))
+(define_predicate "aarch64_addsvl_addspl_immediate"
+ (and (match_code "const")
+ (match_test "aarch64_addsvl_addspl_immediate_p (op)")))
+
(define_predicate "aarch64_pluslong_operand"
(ior (match_operand 0 "register_operand")
(match_operand 0 "aarch64_pluslong_immediate")
(and (match_test "TARGET_SVE")
- (match_operand 0 "aarch64_sve_plus_immediate"))))
+ (match_operand 0 "aarch64_sve_plus_immediate"))
+ (and (match_test "TARGET_SME")
+ (match_operand 0 "aarch64_addsvl_addspl_immediate"))))
(define_predicate "aarch64_pluslong_or_poly_operand"
(ior (match_operand 0 "aarch64_pluslong_operand")
$(srcdir)/config/aarch64/aarch64-sve-builtins.def \
$(srcdir)/config/aarch64/aarch64-sve-builtins-base.def \
$(srcdir)/config/aarch64/aarch64-sve-builtins-sve2.def \
+ $(srcdir)/config/aarch64/aarch64-sve-builtins-sme.def \
$(CONFIG_H) $(SYSTEM_H) coretypes.h $(TM_H) $(TREE_H) $(RTL_H) \
$(TM_P_H) memmodel.h insn-codes.h $(OPTABS_H) $(RECOG_H) $(DIAGNOSTIC_H) \
$(EXPR_H) $(BASIC_BLOCK_H) $(FUNCTION_H) fold-const.h $(GIMPLE_H) \
$(srcdir)/config/aarch64/aarch64-sve-builtins.h \
$(srcdir)/config/aarch64/aarch64-sve-builtins-shapes.h \
$(srcdir)/config/aarch64/aarch64-sve-builtins-base.h \
- $(srcdir)/config/aarch64/aarch64-sve-builtins-sve2.h
+ $(srcdir)/config/aarch64/aarch64-sve-builtins-sve2.h \
+ $(srcdir)/config/aarch64/aarch64-sve-builtins-sme.h
$(COMPILER) -c $(ALL_COMPILERFLAGS) $(ALL_CPPFLAGS) $(INCLUDES) \
$(srcdir)/config/aarch64/aarch64-sve-builtins.cc
$(COMPILER) -c $(ALL_COMPILERFLAGS) $(ALL_CPPFLAGS) $(INCLUDES) \
$(srcdir)/config/aarch64/aarch64-sve-builtins-sve2.cc
+aarch64-sve-builtins-sme.o: \
+ $(srcdir)/config/aarch64/aarch64-sve-builtins-sme.cc \
+ $(CONFIG_H) $(SYSTEM_H) coretypes.h $(TM_H) $(TREE_H) $(RTL_H) \
+ $(TM_P_H) memmodel.h insn-codes.h $(OPTABS_H) $(RECOG_H) \
+ $(EXPR_H) $(BASIC_BLOCK_H) $(FUNCTION_H) fold-const.h $(GIMPLE_H) \
+ gimple-iterator.h gimplify.h explow.h $(EMIT_RTL_H) \
+ $(srcdir)/config/aarch64/aarch64-sve-builtins.h \
+ $(srcdir)/config/aarch64/aarch64-sve-builtins-shapes.h \
+ $(srcdir)/config/aarch64/aarch64-sve-builtins-sme.h \
+ $(srcdir)/config/aarch64/aarch64-sve-builtins-functions.h
+ $(COMPILER) -c $(ALL_COMPILERFLAGS) $(ALL_CPPFLAGS) $(INCLUDES) \
+ $(srcdir)/config/aarch64/aarch64-sve-builtins-sme.cc
+
aarch64-builtin-iterators.h: $(srcdir)/config/aarch64/geniterators.sh \
$(srcdir)/config/aarch64/iterators.md
$(SHELL) $(srcdir)/config/aarch64/geniterators.sh \
Enable the Common Short Sequence Compression instructions.
@item sme
Enable the Scalable Matrix Extension.
+@item sme-i16i64
+Enable the FEAT_SME_I16I64 extension to SME.
+@item sme-f64f64
+Enable the FEAT_SME_F64F64 extension to SME.
@end table
--- /dev/null
+# Assembly-based regression-test driver for the SME ACLE.
+# Copyright (C) 2009-2023 Free Software Foundation, Inc.
+#
+# This file is part of GCC.
+#
+# GCC is free software; you can redistribute it and/or modify it
+# under the terms of the GNU General Public License as published by
+# the Free Software Foundation; either version 3, or (at your option)
+# any later version.
+#
+# GCC is distributed in the hope that it will be useful, but
+# WITHOUT ANY WARRANTY; without even the implied warranty of
+# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
+# General Public License for more details.
+#
+# You should have received a copy of the GNU General Public License
+# along with GCC; see the file COPYING3. If not see
+# <http://www.gnu.org/licenses/>. */
+
+# GCC testsuite that uses the `dg.exp' driver.
+
+# Exit immediately if this isn't an AArch64 target.
+if {![istarget aarch64*-*-*] } {
+ return
+}
+
+# Load support procs.
+load_lib g++-dg.exp
+
+# Initialize `dg'.
+dg-init
+
+# Force SME if we're not testing it already.
+if { [check_effective_target_aarch64_sme] } {
+ set sme_flags ""
+} else {
+ set sme_flags "-march=armv9-a+sme"
+}
+
+# Turn off any codegen tweaks by default that may affect expected assembly.
+# Tests relying on those should turn them on explicitly.
+set sme_flags "$sme_flags -mtune=generic -moverride=tune=none"
+
+global gcc_runtest_parallelize_limit_minor
+if { [info exists gcc_runtest_parallelize_limit_minor] } {
+ set old_limit_minor $gcc_runtest_parallelize_limit_minor
+ set gcc_runtest_parallelize_limit_minor 1
+}
+
+torture-init
+set-torture-options {
+ "-std=c++11 -O0 -g"
+ "-std=c++14 -O1 -g"
+ "-std=c++17 -Og -g"
+ "-std=c++23 -Os -g"
+ "-std=gnu++11 -O2 -fno-schedule-insns -fno-schedule-insns2 -DCHECK_ASM --save-temps"
+ "-std=gnu++23 -Ofast -g"
+} {
+ "-DTEST_FULL"
+ "-DTEST_OVERLOADS"
+}
+
+# Main loop.
+set gcc_subdir [string replace $subdir 0 2 gcc]
+set files [glob -nocomplain $srcdir/$gcc_subdir/acle-asm/*.c]
+set save-dg-do-what-default ${dg-do-what-default}
+if { [check_effective_target_aarch64_asm_sme-i16i64_ok] } {
+ set dg-do-what-default assemble
+} else {
+ set dg-do-what-default compile
+}
+gcc-dg-runtest [lsort $files] "" "$sme_flags -fno-ipa-icf"
+set dg-do-what-default ${save-dg-do-what-default}
+
+torture-finish
+
+if { [info exists gcc_runtest_parallelize_limit_minor] } {
+ set gcc_runtest_parallelize_limit_minor $old_limit_minor
+}
+
+# All done.
+dg-finish
to be diagnosed. Any attempt to call the function before including
arm_sve.h will lead to a link failure. (Same for taking its address,
etc.) */
-extern __SVUint8_t svadd_u8_x (__SVBool_t, __SVUint8_t, __SVUint8_t);
+extern __SVUint8_t svadd_u8_x (__SVBool_t, __SVUint8_t, __SVUint8_t)
+ __arm_streaming_compatible;
#pragma GCC aarch64 "arm_sve.h"
__SVUint8_t
svadd_u8_x (__SVBool_t pg, __SVUint8_t x, __SVUint8_t y)
+ __arm_streaming_compatible
{
return x;
}
__SVUint8_t
svadd_x (__SVBool_t pg, __SVUint8_t x, __SVUint8_t y)
+ __arm_streaming_compatible
{
return x;
}
#ifdef __ARM_FEATURE_SVE2_SHA3
#error Foo
#endif
+
+#pragma GCC target "+sme"
+#ifndef __ARM_FEATURE_SME
+#error Foo
+#endif
+
+#pragma GCC target "+sme+nofp"
+#ifdef __ARM_FEATURE_SME
+#error Foo
+#endif
+
+#pragma GCC target "+sme+nosimd"
+#ifdef __ARM_FEATURE_SME
+#error Foo
+#endif
+
+#pragma GCC target "+sme+nobf16"
+#ifdef __ARM_FEATURE_SME
+#error Foo
+#endif
+
+#pragma GCC target "+nothing+sme"
+#ifdef __ARM_FEATURE_SME_I16I64
+#error Foo
+#endif
+#ifdef __ARM_FEATURE_SME_F64F64
+#error Foo
+#endif
+
+#pragma GCC target "+sme-i16i64"
+#ifndef __ARM_FEATURE_SME_I16I64
+#error Foo
+#endif
+
+#pragma GCC target "+sme-f64f64"
+#ifndef __ARM_FEATURE_SME_F64F64
+#error Foo
+#endif
--- /dev/null
+# Assembly-based regression-test driver for the SME ACLE.
+# Copyright (C) 2009-2023 Free Software Foundation, Inc.
+#
+# This file is part of GCC.
+#
+# GCC is free software; you can redistribute it and/or modify it
+# under the terms of the GNU General Public License as published by
+# the Free Software Foundation; either version 3, or (at your option)
+# any later version.
+#
+# GCC is distributed in the hope that it will be useful, but
+# WITHOUT ANY WARRANTY; without even the implied warranty of
+# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
+# General Public License for more details.
+#
+# You should have received a copy of the GNU General Public License
+# along with GCC; see the file COPYING3. If not see
+# <http://www.gnu.org/licenses/>. */
+
+# GCC testsuite that uses the `dg.exp' driver.
+
+# Exit immediately if this isn't an AArch64 target.
+if {![istarget aarch64*-*-*] } {
+ return
+}
+
+# Load support procs.
+load_lib gcc-dg.exp
+
+# Initialize `dg'.
+dg-init
+
+# Force SME if we're not testing it already.
+if { [check_effective_target_aarch64_sme] } {
+ set sme_flags ""
+} else {
+ set sme_flags "-march=armv9-a+sme"
+}
+
+# Turn off any codegen tweaks by default that may affect expected assembly.
+# Tests relying on those should turn them on explicitly.
+set sme_flags "$sme_flags -mtune=generic -moverride=tune=none"
+
+global gcc_runtest_parallelize_limit_minor
+if { [info exists gcc_runtest_parallelize_limit_minor] } {
+ set old_limit_minor $gcc_runtest_parallelize_limit_minor
+ set gcc_runtest_parallelize_limit_minor 1
+}
+
+torture-init
+set-torture-options {
+ "-std=c90 -O0 -g"
+ "-std=c99 -Og -g"
+ "-std=c11 -Os -g"
+ "-std=c23 -O2 -fno-schedule-insns -fno-schedule-insns2 -DCHECK_ASM --save-temps"
+ "-std=gnu90 -O3 -g"
+ "-std=gnu23 -Ofast -g"
+} {
+ "-DTEST_FULL"
+ "-DTEST_OVERLOADS"
+}
+
+# Main loop.
+set files [glob -nocomplain $srcdir/$subdir/acle-asm/*.c]
+set save-dg-do-what-default ${dg-do-what-default}
+if { [check_effective_target_aarch64_asm_sme-i16i64_ok] } {
+ set dg-do-what-default assemble
+} else {
+ set dg-do-what-default compile
+}
+gcc-dg-runtest [lsort $files] "" "$sme_flags -fno-ipa-icf"
+set dg-do-what-default ${save-dg-do-what-default}
+
+torture-finish
+
+if { [info exists gcc_runtest_parallelize_limit_minor] } {
+ set gcc_runtest_parallelize_limit_minor $old_limit_minor
+}
+
+# All done.
+dg-finish
--- /dev/null
+/* { dg-final { check-function-bodies "**" "" "-DCHECK_ASM" } } */
+
+#include "test_sme_acle.h"
+
+/*
+** addha_za32_s32_0_p0_p1_z0:
+** addha za0\.s, p0/m, p1/m, z0\.s
+** ret
+*/
+TEST_UNIFORM_ZA (addha_za32_s32_0_p0_p1_z0, svint32_t,
+ svaddha_za32_s32_m (0, p0, p1, z0),
+ svaddha_za32_m (0, p0, p1, z0))
+
+/*
+** addha_za32_s32_0_p1_p0_z1:
+** addha za0\.s, p1/m, p0/m, z1\.s
+** ret
+*/
+TEST_UNIFORM_ZA (addha_za32_s32_0_p1_p0_z1, svint32_t,
+ svaddha_za32_s32_m (0, p1, p0, z1),
+ svaddha_za32_m (0, p1, p0, z1))
+
+/*
+** addha_za32_s32_1_p0_p1_z0:
+** addha za1\.s, p0/m, p1/m, z0\.s
+** ret
+*/
+TEST_UNIFORM_ZA (addha_za32_s32_1_p0_p1_z0, svint32_t,
+ svaddha_za32_s32_m (1, p0, p1, z0),
+ svaddha_za32_m (1, p0, p1, z0))
+
+/*
+** addha_za32_s32_3_p0_p1_z0:
+** addha za3\.s, p0/m, p1/m, z0\.s
+** ret
+*/
+TEST_UNIFORM_ZA (addha_za32_s32_3_p0_p1_z0, svint32_t,
+ svaddha_za32_s32_m (3, p0, p1, z0),
+ svaddha_za32_m (3, p0, p1, z0))
+
+/*
+** addha_za32_u32_0_p0_p1_z0:
+** addha za0\.s, p0/m, p1/m, z0\.s
+** ret
+*/
+TEST_UNIFORM_ZA (addha_za32_u32_0_p0_p1_z0, svuint32_t,
+ svaddha_za32_u32_m (0, p0, p1, z0),
+ svaddha_za32_m (0, p0, p1, z0))
--- /dev/null
+/* { dg-final { check-function-bodies "**" "" "-DCHECK_ASM" } } */
+
+#include "test_sme_acle.h"
+
+#pragma GCC target "+sme-i16i64"
+
+/*
+** addha_za64_s64_0_p0_p1_z0:
+** addha za0\.d, p0/m, p1/m, z0\.d
+** ret
+*/
+TEST_UNIFORM_ZA (addha_za64_s64_0_p0_p1_z0, svint64_t,
+ svaddha_za64_s64_m (0, p0, p1, z0),
+ svaddha_za64_m (0, p0, p1, z0))
+
+/*
+** addha_za64_s64_0_p1_p0_z1:
+** addha za0\.d, p1/m, p0/m, z1\.d
+** ret
+*/
+TEST_UNIFORM_ZA (addha_za64_s64_0_p1_p0_z1, svint64_t,
+ svaddha_za64_s64_m (0, p1, p0, z1),
+ svaddha_za64_m (0, p1, p0, z1))
+
+/*
+** addha_za64_s64_1_p0_p1_z0:
+** addha za1\.d, p0/m, p1/m, z0\.d
+** ret
+*/
+TEST_UNIFORM_ZA (addha_za64_s64_1_p0_p1_z0, svint64_t,
+ svaddha_za64_s64_m (1, p0, p1, z0),
+ svaddha_za64_m (1, p0, p1, z0))
+
+/*
+** addha_za64_s64_7_p0_p1_z0:
+** addha za7\.d, p0/m, p1/m, z0\.d
+** ret
+*/
+TEST_UNIFORM_ZA (addha_za64_s64_7_p0_p1_z0, svint64_t,
+ svaddha_za64_s64_m (7, p0, p1, z0),
+ svaddha_za64_m (7, p0, p1, z0))
+
+/*
+** addha_za64_u64_0_p0_p1_z0:
+** addha za0\.d, p0/m, p1/m, z0\.d
+** ret
+*/
+TEST_UNIFORM_ZA (addha_za64_u64_0_p0_p1_z0, svuint64_t,
+ svaddha_za64_u64_m (0, p0, p1, z0),
+ svaddha_za64_m (0, p0, p1, z0))
--- /dev/null
+/* { dg-final { check-function-bodies "**" "" "-DCHECK_ASM" } } */
+
+#include "test_sme_acle.h"
+
+/*
+** addva_za32_s32_0_p0_p1_z0:
+** addva za0\.s, p0/m, p1/m, z0\.s
+** ret
+*/
+TEST_UNIFORM_ZA (addva_za32_s32_0_p0_p1_z0, svint32_t,
+ svaddva_za32_s32_m (0, p0, p1, z0),
+ svaddva_za32_m (0, p0, p1, z0))
+
+/*
+** addva_za32_s32_0_p1_p0_z1:
+** addva za0\.s, p1/m, p0/m, z1\.s
+** ret
+*/
+TEST_UNIFORM_ZA (addva_za32_s32_0_p1_p0_z1, svint32_t,
+ svaddva_za32_s32_m (0, p1, p0, z1),
+ svaddva_za32_m (0, p1, p0, z1))
+
+/*
+** addva_za32_s32_1_p0_p1_z0:
+** addva za1\.s, p0/m, p1/m, z0\.s
+** ret
+*/
+TEST_UNIFORM_ZA (addva_za32_s32_1_p0_p1_z0, svint32_t,
+ svaddva_za32_s32_m (1, p0, p1, z0),
+ svaddva_za32_m (1, p0, p1, z0))
+
+/*
+** addva_za32_s32_3_p0_p1_z0:
+** addva za3\.s, p0/m, p1/m, z0\.s
+** ret
+*/
+TEST_UNIFORM_ZA (addva_za32_s32_3_p0_p1_z0, svint32_t,
+ svaddva_za32_s32_m (3, p0, p1, z0),
+ svaddva_za32_m (3, p0, p1, z0))
+
+/*
+** addva_za32_u32_0_p0_p1_z0:
+** addva za0\.s, p0/m, p1/m, z0\.s
+** ret
+*/
+TEST_UNIFORM_ZA (addva_za32_u32_0_p0_p1_z0, svuint32_t,
+ svaddva_za32_u32_m (0, p0, p1, z0),
+ svaddva_za32_m (0, p0, p1, z0))
--- /dev/null
+/* { dg-final { check-function-bodies "**" "" "-DCHECK_ASM" } } */
+
+#include "test_sme_acle.h"
+
+#pragma GCC target "+sme-i16i64"
+
+/*
+** addva_za64_s64_0_p0_p1_z0:
+** addva za0\.d, p0/m, p1/m, z0\.d
+** ret
+*/
+TEST_UNIFORM_ZA (addva_za64_s64_0_p0_p1_z0, svint64_t,
+ svaddva_za64_s64_m (0, p0, p1, z0),
+ svaddva_za64_m (0, p0, p1, z0))
+
+/*
+** addva_za64_s64_0_p1_p0_z1:
+** addva za0\.d, p1/m, p0/m, z1\.d
+** ret
+*/
+TEST_UNIFORM_ZA (addva_za64_s64_0_p1_p0_z1, svint64_t,
+ svaddva_za64_s64_m (0, p1, p0, z1),
+ svaddva_za64_m (0, p1, p0, z1))
+
+/*
+** addva_za64_s64_1_p0_p1_z0:
+** addva za1\.d, p0/m, p1/m, z0\.d
+** ret
+*/
+TEST_UNIFORM_ZA (addva_za64_s64_1_p0_p1_z0, svint64_t,
+ svaddva_za64_s64_m (1, p0, p1, z0),
+ svaddva_za64_m (1, p0, p1, z0))
+
+/*
+** addva_za64_s64_7_p0_p1_z0:
+** addva za7\.d, p0/m, p1/m, z0\.d
+** ret
+*/
+TEST_UNIFORM_ZA (addva_za64_s64_7_p0_p1_z0, svint64_t,
+ svaddva_za64_s64_m (7, p0, p1, z0),
+ svaddva_za64_m (7, p0, p1, z0))
+
+/*
+** addva_za64_u64_0_p0_p1_z0:
+** addva za0\.d, p0/m, p1/m, z0\.d
+** ret
+*/
+TEST_UNIFORM_ZA (addva_za64_u64_0_p0_p1_z0, svuint64_t,
+ svaddva_za64_u64_m (0, p0, p1, z0),
+ svaddva_za64_m (0, p0, p1, z0))
--- /dev/null
+/* { dg-final { check-function-bodies "**" "" "-DCHECK_ASM" } } */
+
+#define STREAMING_COMPATIBLE
+#define NO_SHARED_ZA
+#include "test_sme_acle.h"
+
+#pragma GCC target "+nosme"
+
+/*
+** test_nosme:
+** ...
+** bl __arm_sme_state
+** lsr x0, x0, #?63
+** ...
+*/
+PROTO (test_nosme, int, ()) { return __arm_has_sme (); }
+
+#pragma GCC target "+sme"
+
+/*
+** test_sme:
+** mov w0, #?1
+** ret
+*/
+PROTO (test_sme, int, ()) { return __arm_has_sme (); }
--- /dev/null
+/* { dg-final { check-function-bodies "**" "" "-DCHECK_ASM" } } */
+
+#define NON_STREAMING
+#include "test_sme_acle.h"
+
+/*
+** test_sme:
+** mov w0, #?0
+** ret
+*/
+PROTO (test_sme, int, ()) { return __arm_in_streaming_mode (); }
--- /dev/null
+/* { dg-final { check-function-bodies "**" "" "-DCHECK_ASM" } } */
+
+#define NO_SHARED_ZA
+#include "test_sme_acle.h"
+
+/*
+** test_sme:
+** mov w0, #?1
+** ret
+*/
+PROTO (test_sme, int, ()) { return __arm_in_streaming_mode (); }
--- /dev/null
+/* { dg-final { check-function-bodies "**" "" "-DCHECK_ASM" } } */
+
+#define STREAMING_COMPATIBLE
+#define NO_SHARED_ZA
+#include "test_sme_acle.h"
+
+#pragma GCC target "+nosme"
+
+/*
+** test_nosme:
+** ...
+** bl __arm_sme_state
+** and w0, w0, #?1
+** ...
+*/
+PROTO (test_nosme, int, ()) { return __arm_in_streaming_mode (); }
+
+#pragma GCC target "+sme"
+
+/*
+** test_sme:
+** mrs x([0-9]+), svcr
+** and w0, w\1, #?1
+** ret
+*/
+PROTO (test_sme, int, ()) { return __arm_in_streaming_mode (); }
--- /dev/null
+/* { dg-final { check-function-bodies "**" "" "-DCHECK_ASM" } } */
+
+#define NO_SHARED_ZA
+#include "test_sme_acle.h"
+
+/*
+** cntb_1:
+** cntb x0
+** ret
+*/
+PROTO (cntb_1, uint64_t, ()) { return svcntsb (); }
+
+/*
+** cntb_2:
+** cntb x0, all, mul #2
+** ret
+*/
+PROTO (cntb_2, uint64_t, ()) { return svcntsb () * 2; }
+
+/*
+** cntb_3:
+** cntb x0, all, mul #3
+** ret
+*/
+PROTO (cntb_3, uint64_t, ()) { return svcntsb () * 3; }
+
+/*
+** cntb_4:
+** cntb x0, all, mul #4
+** ret
+*/
+PROTO (cntb_4, uint64_t, ()) { return svcntsb () * 4; }
+
+/*
+** cntb_8:
+** cntb x0, all, mul #8
+** ret
+*/
+PROTO (cntb_8, uint64_t, ()) { return svcntsb () * 8; }
+
+/*
+** cntb_15:
+** cntb x0, all, mul #15
+** ret
+*/
+PROTO (cntb_15, uint64_t, ()) { return svcntsb () * 15; }
+
+/*
+** cntb_16:
+** cntb x0, all, mul #16
+** ret
+*/
+PROTO (cntb_16, uint64_t, ()) { return svcntsb () * 16; }
+
+/*
+** cntb_17:
+** rdvl x0, #17
+** ret
+*/
+PROTO (cntb_17, uint64_t, ()) { return svcntsb () * 17; }
+
+/*
+** cntb_31:
+** rdvl x0, #31
+** ret
+*/
+PROTO (cntb_31, uint64_t, ()) { return svcntsb () * 31; }
+
+/*
+** cntb_32:
+** cntb (x[0-9]+)
+** lsl x0, \1, 5
+** ret
+*/
+PROTO (cntb_32, uint64_t, ()) { return svcntsb () * 32; }
+
+/* Other sequences would be OK. */
+/*
+** cntb_33:
+** cntb (x[0-9]+)
+** lsl x0, \1, 5
+** incb x0
+** ret
+*/
+PROTO (cntb_33, uint64_t, ()) { return svcntsb () * 33; }
+
+/*
+** cntb_64:
+** cntb (x[0-9]+)
+** lsl x0, \1, 6
+** ret
+*/
+PROTO (cntb_64, uint64_t, ()) { return svcntsb () * 64; }
+
+/*
+** cntb_128:
+** cntb (x[0-9]+)
+** lsl x0, \1, 7
+** ret
+*/
+PROTO (cntb_128, uint64_t, ()) { return svcntsb () * 128; }
+
+/* Other sequences would be OK. */
+/*
+** cntb_129:
+** cntb (x[0-9]+)
+** lsl x0, \1, 7
+** incb x0
+** ret
+*/
+PROTO (cntb_129, uint64_t, ()) { return svcntsb () * 129; }
+
+/*
+** cntb_m1:
+** rdvl x0, #-1
+** ret
+*/
+PROTO (cntb_m1, uint64_t, ()) { return -svcntsb (); }
+
+/*
+** cntb_m13:
+** rdvl x0, #-13
+** ret
+*/
+PROTO (cntb_m13, uint64_t, ()) { return -svcntsb () * 13; }
+
+/*
+** cntb_m15:
+** rdvl x0, #-15
+** ret
+*/
+PROTO (cntb_m15, uint64_t, ()) { return -svcntsb () * 15; }
+
+/*
+** cntb_m16:
+** rdvl x0, #-16
+** ret
+*/
+PROTO (cntb_m16, uint64_t, ()) { return -svcntsb () * 16; }
+
+/*
+** cntb_m17:
+** rdvl x0, #-17
+** ret
+*/
+PROTO (cntb_m17, uint64_t, ()) { return -svcntsb () * 17; }
+
+/*
+** cntb_m32:
+** rdvl x0, #-32
+** ret
+*/
+PROTO (cntb_m32, uint64_t, ()) { return -svcntsb () * 32; }
+
+/*
+** cntb_m33:
+** rdvl x0, #-32
+** decb x0
+** ret
+*/
+PROTO (cntb_m33, uint64_t, ()) { return -svcntsb () * 33; }
+
+/*
+** cntb_m34:
+** rdvl (x[0-9]+), #-17
+** lsl x0, \1, #?1
+** ret
+*/
+PROTO (cntb_m34, uint64_t, ()) { return -svcntsb () * 34; }
+
+/*
+** cntb_m64:
+** rdvl (x[0-9]+), #-1
+** lsl x0, \1, #?6
+** ret
+*/
+PROTO (cntb_m64, uint64_t, ()) { return -svcntsb () * 64; }
+
+/*
+** incb_1:
+** incb x0
+** ret
+*/
+PROTO (incb_1, uint64_t, (uint64_t x0)) { return x0 + svcntsb (); }
+
+/*
+** incb_2:
+** incb x0, all, mul #2
+** ret
+*/
+PROTO (incb_2, uint64_t, (uint64_t x0)) { return x0 + svcntsb () * 2; }
+
+/*
+** incb_3:
+** incb x0, all, mul #3
+** ret
+*/
+PROTO (incb_3, uint64_t, (uint64_t x0)) { return x0 + svcntsb () * 3; }
+
+/*
+** incb_4:
+** incb x0, all, mul #4
+** ret
+*/
+PROTO (incb_4, uint64_t, (uint64_t x0)) { return x0 + svcntsb () * 4; }
+
+/*
+** incb_8:
+** incb x0, all, mul #8
+** ret
+*/
+PROTO (incb_8, uint64_t, (uint64_t x0)) { return x0 + svcntsb () * 8; }
+
+/*
+** incb_15:
+** incb x0, all, mul #15
+** ret
+*/
+PROTO (incb_15, uint64_t, (uint64_t x0)) { return x0 + svcntsb () * 15; }
+
+/*
+** incb_16:
+** incb x0, all, mul #16
+** ret
+*/
+PROTO (incb_16, uint64_t, (uint64_t x0)) { return x0 + svcntsb () * 16; }
+
+/*
+** incb_17:
+** addvl x0, x0, #17
+** ret
+*/
+PROTO (incb_17, uint64_t, (uint64_t x0)) { return x0 + svcntsb () * 17; }
+
+/*
+** incb_31:
+** addvl x0, x0, #31
+** ret
+*/
+PROTO (incb_31, uint64_t, (uint64_t x0)) { return x0 + svcntsb () * 31; }
+
+/*
+** decb_1:
+** decb x0
+** ret
+*/
+PROTO (decb_1, uint64_t, (uint64_t x0)) { return x0 - svcntsb (); }
+
+/*
+** decb_2:
+** decb x0, all, mul #2
+** ret
+*/
+PROTO (decb_2, uint64_t, (uint64_t x0)) { return x0 - svcntsb () * 2; }
+
+/*
+** decb_3:
+** decb x0, all, mul #3
+** ret
+*/
+PROTO (decb_3, uint64_t, (uint64_t x0)) { return x0 - svcntsb () * 3; }
+
+/*
+** decb_4:
+** decb x0, all, mul #4
+** ret
+*/
+PROTO (decb_4, uint64_t, (uint64_t x0)) { return x0 - svcntsb () * 4; }
+
+/*
+** decb_8:
+** decb x0, all, mul #8
+** ret
+*/
+PROTO (decb_8, uint64_t, (uint64_t x0)) { return x0 - svcntsb () * 8; }
+
+/*
+** decb_15:
+** decb x0, all, mul #15
+** ret
+*/
+PROTO (decb_15, uint64_t, (uint64_t x0)) { return x0 - svcntsb () * 15; }
+
+/*
+** decb_16:
+** decb x0, all, mul #16
+** ret
+*/
+PROTO (decb_16, uint64_t, (uint64_t x0)) { return x0 - svcntsb () * 16; }
+
+/*
+** decb_17:
+** addvl x0, x0, #-17
+** ret
+*/
+PROTO (decb_17, uint64_t, (uint64_t x0)) { return x0 - svcntsb () * 17; }
+
+/*
+** decb_31:
+** addvl x0, x0, #-31
+** ret
+*/
+PROTO (decb_31, uint64_t, (uint64_t x0)) { return x0 - svcntsb () * 31; }
+
+/*
+** decb_32:
+** addvl x0, x0, #-32
+** ret
+*/
+PROTO (decb_32, uint64_t, (uint64_t x0)) { return x0 - svcntsb () * 32; }
--- /dev/null
+/* { dg-final { check-function-bodies "**" "" "-DCHECK_ASM" } } */
+
+#define STREAMING_COMPATIBLE
+#define NO_SHARED_ZA
+#include "test_sme_acle.h"
+
+/*
+** cntsb:
+** rdsvl x0, #1
+** ret
+*/
+PROTO (cntsb, uint64_t, ()) { return svcntsb (); }
--- /dev/null
+/* { dg-final { check-function-bodies "**" "" "-DCHECK_ASM" } } */
+
+#define NO_SHARED_ZA
+#include "test_sme_acle.h"
+
+/*
+** cntd_1:
+** cntd x0
+** ret
+*/
+PROTO (cntd_1, uint64_t, ()) { return svcntsd (); }
+
+/*
+** cntd_2:
+** cntw x0
+** ret
+*/
+PROTO (cntd_2, uint64_t, ()) { return svcntsd () * 2; }
+
+/*
+** cntd_3:
+** cntd x0, all, mul #3
+** ret
+*/
+PROTO (cntd_3, uint64_t, ()) { return svcntsd () * 3; }
+
+/*
+** cntd_4:
+** cnth x0
+** ret
+*/
+PROTO (cntd_4, uint64_t, ()) { return svcntsd () * 4; }
+
+/*
+** cntd_8:
+** cntb x0
+** ret
+*/
+PROTO (cntd_8, uint64_t, ()) { return svcntsd () * 8; }
+
+/*
+** cntd_15:
+** cntd x0, all, mul #15
+** ret
+*/
+PROTO (cntd_15, uint64_t, ()) { return svcntsd () * 15; }
+
+/*
+** cntd_16:
+** cntb x0, all, mul #2
+** ret
+*/
+PROTO (cntd_16, uint64_t, ()) { return svcntsd () * 16; }
+
+/* Other sequences would be OK. */
+/*
+** cntd_17:
+** rdvl (x[0-9]+), #17
+** asr x0, \1, 3
+** ret
+*/
+PROTO (cntd_17, uint64_t, ()) { return svcntsd () * 17; }
+
+/*
+** cntd_32:
+** cntb x0, all, mul #4
+** ret
+*/
+PROTO (cntd_32, uint64_t, ()) { return svcntsd () * 32; }
+
+/*
+** cntd_64:
+** cntb x0, all, mul #8
+** ret
+*/
+PROTO (cntd_64, uint64_t, ()) { return svcntsd () * 64; }
+
+/*
+** cntd_128:
+** cntb x0, all, mul #16
+** ret
+*/
+PROTO (cntd_128, uint64_t, ()) { return svcntsd () * 128; }
+
+/*
+** cntd_m1:
+** cntd (x[0-9]+)
+** neg x0, \1
+** ret
+*/
+PROTO (cntd_m1, uint64_t, ()) { return -svcntsd (); }
+
+/*
+** cntd_m13:
+** cntd (x[0-9]+), all, mul #13
+** neg x0, \1
+** ret
+*/
+PROTO (cntd_m13, uint64_t, ()) { return -svcntsd () * 13; }
+
+/*
+** cntd_m15:
+** cntd (x[0-9]+), all, mul #15
+** neg x0, \1
+** ret
+*/
+PROTO (cntd_m15, uint64_t, ()) { return -svcntsd () * 15; }
+
+/*
+** cntd_m16:
+** rdvl x0, #-2
+** ret
+*/
+PROTO (cntd_m16, uint64_t, ()) { return -svcntsd () * 16; }
+
+/* Other sequences would be OK. */
+/*
+** cntd_m17:
+** rdvl (x[0-9]+), #-17
+** asr x0, \1, 3
+** ret
+*/
+PROTO (cntd_m17, uint64_t, ()) { return -svcntsd () * 17; }
+
+/*
+** incd_1:
+** incd x0
+** ret
+*/
+PROTO (incd_1, uint64_t, (uint64_t x0)) { return x0 + svcntsd (); }
+
+/*
+** incd_2:
+** incw x0
+** ret
+*/
+PROTO (incd_2, uint64_t, (uint64_t x0)) { return x0 + svcntsd () * 2; }
+
+/*
+** incd_3:
+** incd x0, all, mul #3
+** ret
+*/
+PROTO (incd_3, uint64_t, (uint64_t x0)) { return x0 + svcntsd () * 3; }
+
+/*
+** incd_4:
+** inch x0
+** ret
+*/
+PROTO (incd_4, uint64_t, (uint64_t x0)) { return x0 + svcntsd () * 4; }
+
+/*
+** incd_7:
+** incd x0, all, mul #7
+** ret
+*/
+PROTO (incd_7, uint64_t, (uint64_t x0)) { return x0 + svcntsd () * 7; }
+
+/*
+** incd_8:
+** incb x0
+** ret
+*/
+PROTO (incd_8, uint64_t, (uint64_t x0)) { return x0 + svcntsd () * 8; }
+
+/*
+** incd_9:
+** incd x0, all, mul #9
+** ret
+*/
+PROTO (incd_9, uint64_t, (uint64_t x0)) { return x0 + svcntsd () * 9; }
+
+/*
+** incd_15:
+** incd x0, all, mul #15
+** ret
+*/
+PROTO (incd_15, uint64_t, (uint64_t x0)) { return x0 + svcntsd () * 15; }
+
+/*
+** incd_16:
+** incb x0, all, mul #2
+** ret
+*/
+PROTO (incd_16, uint64_t, (uint64_t x0)) { return x0 + svcntsd () * 16; }
+
+/*
+** incd_18:
+** incw x0, all, mul #9
+** ret
+*/
+PROTO (incd_18, uint64_t, (uint64_t x0)) { return x0 + svcntsd () * 18; }
+
+/*
+** incd_30:
+** incw x0, all, mul #15
+** ret
+*/
+PROTO (incd_30, uint64_t, (uint64_t x0)) { return x0 + svcntsd () * 30; }
+
+/*
+** decd_1:
+** decd x0
+** ret
+*/
+PROTO (decd_1, uint64_t, (uint64_t x0)) { return x0 - svcntsd (); }
+
+/*
+** decd_2:
+** decw x0
+** ret
+*/
+PROTO (decd_2, uint64_t, (uint64_t x0)) { return x0 - svcntsd () * 2; }
+
+/*
+** decd_3:
+** decd x0, all, mul #3
+** ret
+*/
+PROTO (decd_3, uint64_t, (uint64_t x0)) { return x0 - svcntsd () * 3; }
+
+/*
+** decd_4:
+** dech x0
+** ret
+*/
+PROTO (decd_4, uint64_t, (uint64_t x0)) { return x0 - svcntsd () * 4; }
+
+/*
+** decd_7:
+** decd x0, all, mul #7
+** ret
+*/
+PROTO (decd_7, uint64_t, (uint64_t x0)) { return x0 - svcntsd () * 7; }
+
+/*
+** decd_8:
+** decb x0
+** ret
+*/
+PROTO (decd_8, uint64_t, (uint64_t x0)) { return x0 - svcntsd () * 8; }
+
+/*
+** decd_9:
+** decd x0, all, mul #9
+** ret
+*/
+PROTO (decd_9, uint64_t, (uint64_t x0)) { return x0 - svcntsd () * 9; }
+
+/*
+** decd_15:
+** decd x0, all, mul #15
+** ret
+*/
+PROTO (decd_15, uint64_t, (uint64_t x0)) { return x0 - svcntsd () * 15; }
+
+/*
+** decd_16:
+** decb x0, all, mul #2
+** ret
+*/
+PROTO (decd_16, uint64_t, (uint64_t x0)) { return x0 - svcntsd () * 16; }
+
+/*
+** decd_18:
+** decw x0, all, mul #9
+** ret
+*/
+PROTO (decd_18, uint64_t, (uint64_t x0)) { return x0 - svcntsd () * 18; }
+
+/*
+** decd_30:
+** decw x0, all, mul #15
+** ret
+*/
+PROTO (decd_30, uint64_t, (uint64_t x0)) { return x0 - svcntsd () * 30; }
--- /dev/null
+/* { dg-final { check-function-bodies "**" "" "-DCHECK_ASM" } } */
+
+#define STREAMING_COMPATIBLE
+#define NO_SHARED_ZA
+#include "test_sme_acle.h"
+
+/*
+** cntsd:
+** rdsvl (x[0-9])+, #1
+** lsr x0, \1, #?3
+** ret
+*/
+PROTO (cntsd, uint64_t, ()) { return svcntsd (); }
--- /dev/null
+/* { dg-final { check-function-bodies "**" "" "-DCHECK_ASM" } } */
+
+#define NO_SHARED_ZA
+#include "test_sme_acle.h"
+
+/*
+** cnth_1:
+** cnth x0
+** ret
+*/
+PROTO (cnth_1, uint64_t, ()) { return svcntsh (); }
+
+/*
+** cnth_2:
+** cntb x0
+** ret
+*/
+PROTO (cnth_2, uint64_t, ()) { return svcntsh () * 2; }
+
+/*
+** cnth_3:
+** cnth x0, all, mul #3
+** ret
+*/
+PROTO (cnth_3, uint64_t, ()) { return svcntsh () * 3; }
+
+/*
+** cnth_4:
+** cntb x0, all, mul #2
+** ret
+*/
+PROTO (cnth_4, uint64_t, ()) { return svcntsh () * 4; }
+
+/*
+** cnth_8:
+** cntb x0, all, mul #4
+** ret
+*/
+PROTO (cnth_8, uint64_t, ()) { return svcntsh () * 8; }
+
+/*
+** cnth_15:
+** cnth x0, all, mul #15
+** ret
+*/
+PROTO (cnth_15, uint64_t, ()) { return svcntsh () * 15; }
+
+/*
+** cnth_16:
+** cntb x0, all, mul #8
+** ret
+*/
+PROTO (cnth_16, uint64_t, ()) { return svcntsh () * 16; }
+
+/* Other sequences would be OK. */
+/*
+** cnth_17:
+** rdvl (x[0-9]+), #17
+** asr x0, \1, 1
+** ret
+*/
+PROTO (cnth_17, uint64_t, ()) { return svcntsh () * 17; }
+
+/*
+** cnth_32:
+** cntb x0, all, mul #16
+** ret
+*/
+PROTO (cnth_32, uint64_t, ()) { return svcntsh () * 32; }
+
+/*
+** cnth_64:
+** cntb (x[0-9]+)
+** lsl x0, \1, 5
+** ret
+*/
+PROTO (cnth_64, uint64_t, ()) { return svcntsh () * 64; }
+
+/*
+** cnth_128:
+** cntb (x[0-9]+)
+** lsl x0, \1, 6
+** ret
+*/
+PROTO (cnth_128, uint64_t, ()) { return svcntsh () * 128; }
+
+/*
+** cnth_m1:
+** cnth (x[0-9]+)
+** neg x0, \1
+** ret
+*/
+PROTO (cnth_m1, uint64_t, ()) { return -svcntsh (); }
+
+/*
+** cnth_m13:
+** cnth (x[0-9]+), all, mul #13
+** neg x0, \1
+** ret
+*/
+PROTO (cnth_m13, uint64_t, ()) { return -svcntsh () * 13; }
+
+/*
+** cnth_m15:
+** cnth (x[0-9]+), all, mul #15
+** neg x0, \1
+** ret
+*/
+PROTO (cnth_m15, uint64_t, ()) { return -svcntsh () * 15; }
+
+/*
+** cnth_m16:
+** rdvl x0, #-8
+** ret
+*/
+PROTO (cnth_m16, uint64_t, ()) { return -svcntsh () * 16; }
+
+/* Other sequences would be OK. */
+/*
+** cnth_m17:
+** rdvl (x[0-9]+), #-17
+** asr x0, \1, 1
+** ret
+*/
+PROTO (cnth_m17, uint64_t, ()) { return -svcntsh () * 17; }
+
+/*
+** inch_1:
+** inch x0
+** ret
+*/
+PROTO (inch_1, uint64_t, (uint64_t x0)) { return x0 + svcntsh (); }
+
+/*
+** inch_2:
+** incb x0
+** ret
+*/
+PROTO (inch_2, uint64_t, (uint64_t x0)) { return x0 + svcntsh () * 2; }
+
+/*
+** inch_3:
+** inch x0, all, mul #3
+** ret
+*/
+PROTO (inch_3, uint64_t, (uint64_t x0)) { return x0 + svcntsh () * 3; }
+
+/*
+** inch_4:
+** incb x0, all, mul #2
+** ret
+*/
+PROTO (inch_4, uint64_t, (uint64_t x0)) { return x0 + svcntsh () * 4; }
+
+/*
+** inch_7:
+** inch x0, all, mul #7
+** ret
+*/
+PROTO (inch_7, uint64_t, (uint64_t x0)) { return x0 + svcntsh () * 7; }
+
+/*
+** inch_8:
+** incb x0, all, mul #4
+** ret
+*/
+PROTO (inch_8, uint64_t, (uint64_t x0)) { return x0 + svcntsh () * 8; }
+
+/*
+** inch_9:
+** inch x0, all, mul #9
+** ret
+*/
+PROTO (inch_9, uint64_t, (uint64_t x0)) { return x0 + svcntsh () * 9; }
+
+/*
+** inch_15:
+** inch x0, all, mul #15
+** ret
+*/
+PROTO (inch_15, uint64_t, (uint64_t x0)) { return x0 + svcntsh () * 15; }
+
+/*
+** inch_16:
+** incb x0, all, mul #8
+** ret
+*/
+PROTO (inch_16, uint64_t, (uint64_t x0)) { return x0 + svcntsh () * 16; }
+
+/*
+** inch_18:
+** incb x0, all, mul #9
+** ret
+*/
+PROTO (inch_18, uint64_t, (uint64_t x0)) { return x0 + svcntsh () * 18; }
+
+/*
+** inch_30:
+** incb x0, all, mul #15
+** ret
+*/
+PROTO (inch_30, uint64_t, (uint64_t x0)) { return x0 + svcntsh () * 30; }
+
+/*
+** dech_1:
+** dech x0
+** ret
+*/
+PROTO (dech_1, uint64_t, (uint64_t x0)) { return x0 - svcntsh (); }
+
+/*
+** dech_2:
+** decb x0
+** ret
+*/
+PROTO (dech_2, uint64_t, (uint64_t x0)) { return x0 - svcntsh () * 2; }
+
+/*
+** dech_3:
+** dech x0, all, mul #3
+** ret
+*/
+PROTO (dech_3, uint64_t, (uint64_t x0)) { return x0 - svcntsh () * 3; }
+
+/*
+** dech_4:
+** decb x0, all, mul #2
+** ret
+*/
+PROTO (dech_4, uint64_t, (uint64_t x0)) { return x0 - svcntsh () * 4; }
+
+/*
+** dech_7:
+** dech x0, all, mul #7
+** ret
+*/
+PROTO (dech_7, uint64_t, (uint64_t x0)) { return x0 - svcntsh () * 7; }
+
+/*
+** dech_8:
+** decb x0, all, mul #4
+** ret
+*/
+PROTO (dech_8, uint64_t, (uint64_t x0)) { return x0 - svcntsh () * 8; }
+
+/*
+** dech_9:
+** dech x0, all, mul #9
+** ret
+*/
+PROTO (dech_9, uint64_t, (uint64_t x0)) { return x0 - svcntsh () * 9; }
+
+/*
+** dech_15:
+** dech x0, all, mul #15
+** ret
+*/
+PROTO (dech_15, uint64_t, (uint64_t x0)) { return x0 - svcntsh () * 15; }
+
+/*
+** dech_16:
+** decb x0, all, mul #8
+** ret
+*/
+PROTO (dech_16, uint64_t, (uint64_t x0)) { return x0 - svcntsh () * 16; }
+
+/*
+** dech_18:
+** decb x0, all, mul #9
+** ret
+*/
+PROTO (dech_18, uint64_t, (uint64_t x0)) { return x0 - svcntsh () * 18; }
+
+/*
+** dech_30:
+** decb x0, all, mul #15
+** ret
+*/
+PROTO (dech_30, uint64_t, (uint64_t x0)) { return x0 - svcntsh () * 30; }
--- /dev/null
+/* { dg-final { check-function-bodies "**" "" "-DCHECK_ASM" } } */
+
+#define STREAMING_COMPATIBLE
+#define NO_SHARED_ZA
+#include "test_sme_acle.h"
+
+/*
+** cntsh:
+** rdsvl (x[0-9])+, #1
+** lsr x0, \1, #?1
+** ret
+*/
+PROTO (cntsh, uint64_t, ()) { return svcntsh (); }
--- /dev/null
+/* { dg-final { check-function-bodies "**" "" "-DCHECK_ASM" } } */
+
+#define NO_SHARED_ZA
+#include "test_sme_acle.h"
+
+/*
+** cntw_1:
+** cntw x0
+** ret
+*/
+PROTO (cntw_1, uint64_t, ()) { return svcntsw (); }
+
+/*
+** cntw_2:
+** cnth x0
+** ret
+*/
+PROTO (cntw_2, uint64_t, ()) { return svcntsw () * 2; }
+
+/*
+** cntw_3:
+** cntw x0, all, mul #3
+** ret
+*/
+PROTO (cntw_3, uint64_t, ()) { return svcntsw () * 3; }
+
+/*
+** cntw_4:
+** cntb x0
+** ret
+*/
+PROTO (cntw_4, uint64_t, ()) { return svcntsw () * 4; }
+
+/*
+** cntw_8:
+** cntb x0, all, mul #2
+** ret
+*/
+PROTO (cntw_8, uint64_t, ()) { return svcntsw () * 8; }
+
+/*
+** cntw_15:
+** cntw x0, all, mul #15
+** ret
+*/
+PROTO (cntw_15, uint64_t, ()) { return svcntsw () * 15; }
+
+/*
+** cntw_16:
+** cntb x0, all, mul #4
+** ret
+*/
+PROTO (cntw_16, uint64_t, ()) { return svcntsw () * 16; }
+
+/* Other sequences would be OK. */
+/*
+** cntw_17:
+** rdvl (x[0-9]+), #17
+** asr x0, \1, 2
+** ret
+*/
+PROTO (cntw_17, uint64_t, ()) { return svcntsw () * 17; }
+
+/*
+** cntw_32:
+** cntb x0, all, mul #8
+** ret
+*/
+PROTO (cntw_32, uint64_t, ()) { return svcntsw () * 32; }
+
+/*
+** cntw_64:
+** cntb x0, all, mul #16
+** ret
+*/
+PROTO (cntw_64, uint64_t, ()) { return svcntsw () * 64; }
+
+/*
+** cntw_128:
+** cntb (x[0-9]+)
+** lsl x0, \1, 5
+** ret
+*/
+PROTO (cntw_128, uint64_t, ()) { return svcntsw () * 128; }
+
+/*
+** cntw_m1:
+** cntw (x[0-9]+)
+** neg x0, \1
+** ret
+*/
+PROTO (cntw_m1, uint64_t, ()) { return -svcntsw (); }
+
+/*
+** cntw_m13:
+** cntw (x[0-9]+), all, mul #13
+** neg x0, \1
+** ret
+*/
+PROTO (cntw_m13, uint64_t, ()) { return -svcntsw () * 13; }
+
+/*
+** cntw_m15:
+** cntw (x[0-9]+), all, mul #15
+** neg x0, \1
+** ret
+*/
+PROTO (cntw_m15, uint64_t, ()) { return -svcntsw () * 15; }
+
+/*
+** cntw_m16:
+** rdvl (x[0-9]+), #-4
+** ret
+*/
+PROTO (cntw_m16, uint64_t, ()) { return -svcntsw () * 16; }
+
+/* Other sequences would be OK. */
+/*
+** cntw_m17:
+** rdvl (x[0-9]+), #-17
+** asr x0, \1, 2
+** ret
+*/
+PROTO (cntw_m17, uint64_t, ()) { return -svcntsw () * 17; }
+
+/*
+** incw_1:
+** incw x0
+** ret
+*/
+PROTO (incw_1, uint64_t, (uint64_t x0)) { return x0 + svcntsw (); }
+
+/*
+** incw_2:
+** inch x0
+** ret
+*/
+PROTO (incw_2, uint64_t, (uint64_t x0)) { return x0 + svcntsw () * 2; }
+
+/*
+** incw_3:
+** incw x0, all, mul #3
+** ret
+*/
+PROTO (incw_3, uint64_t, (uint64_t x0)) { return x0 + svcntsw () * 3; }
+
+/*
+** incw_4:
+** incb x0
+** ret
+*/
+PROTO (incw_4, uint64_t, (uint64_t x0)) { return x0 + svcntsw () * 4; }
+
+/*
+** incw_7:
+** incw x0, all, mul #7
+** ret
+*/
+PROTO (incw_7, uint64_t, (uint64_t x0)) { return x0 + svcntsw () * 7; }
+
+/*
+** incw_8:
+** incb x0, all, mul #2
+** ret
+*/
+PROTO (incw_8, uint64_t, (uint64_t x0)) { return x0 + svcntsw () * 8; }
+
+/*
+** incw_9:
+** incw x0, all, mul #9
+** ret
+*/
+PROTO (incw_9, uint64_t, (uint64_t x0)) { return x0 + svcntsw () * 9; }
+
+/*
+** incw_15:
+** incw x0, all, mul #15
+** ret
+*/
+PROTO (incw_15, uint64_t, (uint64_t x0)) { return x0 + svcntsw () * 15; }
+
+/*
+** incw_16:
+** incb x0, all, mul #4
+** ret
+*/
+PROTO (incw_16, uint64_t, (uint64_t x0)) { return x0 + svcntsw () * 16; }
+
+/*
+** incw_18:
+** inch x0, all, mul #9
+** ret
+*/
+PROTO (incw_18, uint64_t, (uint64_t x0)) { return x0 + svcntsw () * 18; }
+
+/*
+** incw_30:
+** inch x0, all, mul #15
+** ret
+*/
+PROTO (incw_30, uint64_t, (uint64_t x0)) { return x0 + svcntsw () * 30; }
+
+/*
+** decw_1:
+** decw x0
+** ret
+*/
+PROTO (decw_1, uint64_t, (uint64_t x0)) { return x0 - svcntsw (); }
+
+/*
+** decw_2:
+** dech x0
+** ret
+*/
+PROTO (decw_2, uint64_t, (uint64_t x0)) { return x0 - svcntsw () * 2; }
+
+/*
+** decw_3:
+** decw x0, all, mul #3
+** ret
+*/
+PROTO (decw_3, uint64_t, (uint64_t x0)) { return x0 - svcntsw () * 3; }
+
+/*
+** decw_4:
+** decb x0
+** ret
+*/
+PROTO (decw_4, uint64_t, (uint64_t x0)) { return x0 - svcntsw () * 4; }
+
+/*
+** decw_7:
+** decw x0, all, mul #7
+** ret
+*/
+PROTO (decw_7, uint64_t, (uint64_t x0)) { return x0 - svcntsw () * 7; }
+
+/*
+** decw_8:
+** decb x0, all, mul #2
+** ret
+*/
+PROTO (decw_8, uint64_t, (uint64_t x0)) { return x0 - svcntsw () * 8; }
+
+/*
+** decw_9:
+** decw x0, all, mul #9
+** ret
+*/
+PROTO (decw_9, uint64_t, (uint64_t x0)) { return x0 - svcntsw () * 9; }
+
+/*
+** decw_15:
+** decw x0, all, mul #15
+** ret
+*/
+PROTO (decw_15, uint64_t, (uint64_t x0)) { return x0 - svcntsw () * 15; }
+
+/*
+** decw_16:
+** decb x0, all, mul #4
+** ret
+*/
+PROTO (decw_16, uint64_t, (uint64_t x0)) { return x0 - svcntsw () * 16; }
+
+/*
+** decw_18:
+** dech x0, all, mul #9
+** ret
+*/
+PROTO (decw_18, uint64_t, (uint64_t x0)) { return x0 - svcntsw () * 18; }
+
+/*
+** decw_30:
+** dech x0, all, mul #15
+** ret
+*/
+PROTO (decw_30, uint64_t, (uint64_t x0)) { return x0 - svcntsw () * 30; }
--- /dev/null
+/* { dg-final { check-function-bodies "**" "" "-DCHECK_ASM" } } */
+
+#define STREAMING_COMPATIBLE
+#define NO_SHARED_ZA
+#include "test_sme_acle.h"
+
+/*
+** cntsw:
+** rdsvl (x[0-9])+, #1
+** lsr x0, \1, #?2
+** ret
+*/
+PROTO (cntsw, uint64_t, ()) { return svcntsw (); }
--- /dev/null
+/* { dg-final { check-function-bodies "**" "" "-DCHECK_ASM" } } */
+
+#include "test_sme_acle.h"
+
+/*
+** ld1_vnum_za128_0_0_0:
+** mov (w1[2-5]), (?:wzr|#?0)
+** ld1q { za0h\.q\[\1, 0\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_vnum_za128_0_0_0,
+ svld1_hor_vnum_za128 (0, 0, p0, x1, 0),
+ svld1_hor_vnum_za128 (0, 0, p0, x1, 0))
+
+/*
+** ld1_vnum_za128_7_1_0:
+** mov (w1[2-5]), #?1
+** ld1q { za7h\.q\[\1, 0\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_vnum_za128_7_1_0,
+ svld1_hor_vnum_za128 (7, 1, p0, x1, 0),
+ svld1_hor_vnum_za128 (7, 1, p0, x1, 0))
+
+/*
+** ld1_vnum_za128_11_1_5:
+** incb x1, all, mul #5
+** mov (w1[2-5]), #?6
+** ld1q { za11h\.q\[\1, 0\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_vnum_za128_11_1_5,
+ svld1_hor_vnum_za128 (11, 1, p0, x1, 5),
+ svld1_hor_vnum_za128 (11, 1, p0, x1, 5))
+
+/*
+** ld1_vnum_za128_3_w0_0:
+** mov (w1[2-5]), w0
+** ld1q { za3h\.q\[\1, 0\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_vnum_za128_3_w0_0,
+ svld1_hor_vnum_za128 (3, w0, p0, x1, 0),
+ svld1_hor_vnum_za128 (3, w0, p0, x1, 0))
+
+/*
+** ld1_vnum_za128_5_w0_0:
+** incb x1, all, mul #13
+** add (w1[2-5]), w0, #?13
+** ld1q { za5h\.q\[\1, 0\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_vnum_za128_5_w0_0,
+ svld1_hor_vnum_za128 (5, w0, p0, x1, 13),
+ svld1_hor_vnum_za128 (5, w0, p0, x1, 13))
+
+/*
+** ld1_vnum_za128_11_w0_0:
+** cntb (x[0-9]+)
+** madd (x[0-9]+), (?:\1, x2|x2, \1), x1
+** add (w1[2-5]), (?:w0, w2|w2, w0)
+** ld1q { za11h\.q\[\3, 0\] }, p0/z, \[\2\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_vnum_za128_11_w0_0,
+ svld1_hor_vnum_za128 (11, w0, p0, x1, x2),
+ svld1_hor_vnum_za128 (11, w0, p0, x1, x2))
+
+/*
+** ld1_vnum_za128_15_w0p1_0:
+** add (w1[2-5]), w0, #?1
+** ld1q { za15h\.q\[\1, 0\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_vnum_za128_15_w0p1_0,
+ svld1_hor_vnum_za128 (15, w0 + 1, p0, x1, 0),
+ svld1_hor_vnum_za128 (15, w0 + 1, p0, x1, 0))
--- /dev/null
+/* { dg-final { check-function-bodies "**" "" "-DCHECK_ASM" } } */
+
+#include "test_sme_acle.h"
+
+/*
+** ld1_vnum_za16_1_0_1:
+** incb x1
+** mov (w1[2-5]), (?:wzr|#?0)
+** ld1h { za1h\.h\[\1, 1\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_vnum_za16_1_0_1,
+ svld1_hor_vnum_za16 (1, 0, p0, x1, 1),
+ svld1_hor_vnum_za16 (1, 0, p0, x1, 1))
+
+/*
+** ld1_vnum_za16_1_1_1:
+** incb x1
+** mov (w1[2-5]), #?1
+** ld1h { za1h\.h\[\1, 1\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_vnum_za16_1_1_1,
+ svld1_hor_vnum_za16 (1, 1, p0, x1, 1),
+ svld1_hor_vnum_za16 (1, 1, p0, x1, 1))
+
+/*
+** ld1_vnum_za16_0_0_8:
+** incb x1, all, mul #8
+** mov (w1[2-5]), #?8
+** ld1h { za0h\.h\[\1, 0\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_vnum_za16_0_0_8,
+ svld1_hor_vnum_za16 (0, 0, p0, x1, 8),
+ svld1_hor_vnum_za16 (0, 0, p0, x1, 8))
+
+/*
+** ld1_vnum_za16_0_1_8:
+** incb x1, all, mul #8
+** mov (w1[2-5]), #?9
+** ld1h { za0h\.h\[\1, 0\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_vnum_za16_0_1_8,
+ svld1_hor_vnum_za16 (0, 1, p0, x1, 8),
+ svld1_hor_vnum_za16 (0, 1, p0, x1, 8))
+
+/*
+** ld1_vnum_za16_0_w0_0:
+** mov (w1[2-5]), w0
+** ld1h { za0h\.h\[\1, 0\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_vnum_za16_0_w0_0,
+ svld1_hor_vnum_za16 (0, w0, p0, x1, 0),
+ svld1_hor_vnum_za16 (0, w0, p0, x1, 0))
+
+/*
+** ld1_vnum_za16_0_w0_1:
+** incb x1
+** mov (w1[2-5]), w0
+** ld1h { za0h\.h\[\1, 1\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_vnum_za16_0_w0_1,
+ svld1_hor_vnum_za16 (0, w0, p0, x1, 1),
+ svld1_hor_vnum_za16 (0, w0, p0, x1, 1))
+
+/*
+** ld1_vnum_za16_0_w0_7:
+** incb x1, all, mul #7
+** mov (w1[2-5]), w0
+** ld1h { za0h\.h\[\1, 7\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_vnum_za16_0_w0_7,
+ svld1_hor_vnum_za16 (0, w0, p0, x1, 7),
+ svld1_hor_vnum_za16 (0, w0, p0, x1, 7))
+
+/*
+** ld1_vnum_za16_1_w0_8:
+** incb x1, all, mul #8
+** add (w1[2-5]), w0, #?8
+** ld1h { za1h\.h\[\1, 0\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_vnum_za16_1_w0_8,
+ svld1_hor_vnum_za16 (1, w0, p0, x1, 8),
+ svld1_hor_vnum_za16 (1, w0, p0, x1, 8))
+
+/*
+** ld1_vnum_za16_1_w0_13:
+** incb x1, all, mul #13
+** add (w1[2-5]), w0, #?13
+** ld1h { za1h\.h\[\1, 0\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_vnum_za16_1_w0_13,
+ svld1_hor_vnum_za16 (1, w0, p0, x1, 13),
+ svld1_hor_vnum_za16 (1, w0, p0, x1, 13))
+
+/*
+** ld1_vnum_za16_0_w0_x2:
+** cntb (x[0-9]+)
+** madd (x[0-9]+), (?:\1, x2|x2, \1), x1
+** add (w1[2-5]), (?:w0, w2|w2, w0)
+** ld1h { za0h\.h\[\3, 0\] }, p0/z, \[\2\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_vnum_za16_0_w0_x2,
+ svld1_hor_vnum_za16 (0, w0, p0, x1, x2),
+ svld1_hor_vnum_za16 (0, w0, p0, x1, x2))
+
+/*
+** ld1_vnum_za16_1_w0p1_0:
+** mov (w1[2-5]), w0
+** ld1h { za1h\.h\[\1, 1\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_vnum_za16_1_w0p1_0,
+ svld1_hor_vnum_za16 (1, w0 + 1, p0, x1, 0),
+ svld1_hor_vnum_za16 (1, w0 + 1, p0, x1, 0))
--- /dev/null
+/* { dg-final { check-function-bodies "**" "" "-DCHECK_ASM" } } */
+
+#include "test_sme_acle.h"
+
+/*
+** ld1_vnum_za32_3_0_1:
+** incb x1
+** mov (w1[2-5]), (?:wzr|#?0)
+** ld1w { za3h\.s\[\1, 1\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_vnum_za32_3_0_1,
+ svld1_hor_vnum_za32 (3, 0, p0, x1, 1),
+ svld1_hor_vnum_za32 (3, 0, p0, x1, 1))
+
+/*
+** ld1_vnum_za32_2_1_1:
+** incb x1
+** mov (w1[2-5]), #?1
+** ld1w { za2h\.s\[\1, 1\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_vnum_za32_2_1_1,
+ svld1_hor_vnum_za32 (2, 1, p0, x1, 1),
+ svld1_hor_vnum_za32 (2, 1, p0, x1, 1))
+
+/*
+** ld1_vnum_za32_0_0_4:
+** incb x1, all, mul #4
+** mov (w1[2-5]), #?4
+** ld1w { za0h\.s\[\1, 0\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_vnum_za32_0_0_4,
+ svld1_hor_vnum_za32 (0, 0, p0, x1, 4),
+ svld1_hor_vnum_za32 (0, 0, p0, x1, 4))
+
+/*
+** ld1_vnum_za32_2_1_4:
+** incb x1, all, mul #4
+** mov (w1[2-5]), #?5
+** ld1w { za2h\.s\[\1, 0\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_vnum_za32_2_1_4,
+ svld1_hor_vnum_za32 (2, 1, p0, x1, 4),
+ svld1_hor_vnum_za32 (2, 1, p0, x1, 4))
+
+/*
+** ld1_vnum_za32_0_w0_0:
+** mov (w1[2-5]), w0
+** ld1w { za0h\.s\[\1, 0\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_vnum_za32_0_w0_0,
+ svld1_hor_vnum_za32 (0, w0, p0, x1, 0),
+ svld1_hor_vnum_za32 (0, w0, p0, x1, 0))
+
+/*
+** ld1_vnum_za32_0_w0_1:
+** incb x1
+** mov (w1[2-5]), w0
+** ld1w { za0h\.s\[\1, 1\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_vnum_za32_0_w0_1,
+ svld1_hor_vnum_za32 (0, w0, p0, x1, 1),
+ svld1_hor_vnum_za32 (0, w0, p0, x1, 1))
+
+/*
+** ld1_vnum_za32_0_w0_3:
+** incb x1, all, mul #3
+** mov (w1[2-5]), w0
+** ld1w { za0h\.s\[\1, 3\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_vnum_za32_0_w0_3,
+ svld1_hor_vnum_za32 (0, w0, p0, x1, 3),
+ svld1_hor_vnum_za32 (0, w0, p0, x1, 3))
+
+/*
+** ld1_vnum_za32_1_w0_4:
+** incb x1, all, mul #4
+** add (w1[2-5]), w0, #?4
+** ld1w { za1h\.s\[\1, 0\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_vnum_za32_1_w0_4,
+ svld1_hor_vnum_za32 (1, w0, p0, x1, 4),
+ svld1_hor_vnum_za32 (1, w0, p0, x1, 4))
+
+/*
+** ld1_vnum_za32_3_w0_13:
+** incb x1, all, mul #13
+** add (w1[2-5]), w0, #?13
+** ld1w { za3h\.s\[\1, 0\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_vnum_za32_3_w0_13,
+ svld1_hor_vnum_za32 (3, w0, p0, x1, 13),
+ svld1_hor_vnum_za32 (3, w0, p0, x1, 13))
+
+/*
+** ld1_vnum_za32_0_w0_x2:
+** cntb (x[0-9]+)
+** madd (x[0-9]+), (?:\1, x2|x2, \1), x1
+** add (w1[2-5]), (?:w0, w2|w2, w0)
+** ld1w { za0h\.s\[\3, 0\] }, p0/z, \[\2\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_vnum_za32_0_w0_x2,
+ svld1_hor_vnum_za32 (0, w0, p0, x1, x2),
+ svld1_hor_vnum_za32 (0, w0, p0, x1, x2))
+
+/*
+** ld1_vnum_za32_1_w0p1_0:
+** mov (w1[2-5]), w0
+** ld1w { za1h\.s\[\1, 1\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_vnum_za32_1_w0p1_0,
+ svld1_hor_vnum_za32 (1, w0 + 1, p0, x1, 0),
+ svld1_hor_vnum_za32 (1, w0 + 1, p0, x1, 0))
--- /dev/null
+/* { dg-final { check-function-bodies "**" "" "-DCHECK_ASM" } } */
+
+#include "test_sme_acle.h"
+
+/*
+** ld1_vnum_za64_3_0_1:
+** incb x1
+** mov (w1[2-5]), (?:wzr|#?0)
+** ld1d { za3h\.d\[\1, 1\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_vnum_za64_3_0_1,
+ svld1_hor_vnum_za64 (3, 0, p0, x1, 1),
+ svld1_hor_vnum_za64 (3, 0, p0, x1, 1))
+
+/*
+** ld1_vnum_za64_7_1_1:
+** incb x1
+** mov (w1[2-5]), #?1
+** ld1d { za7h\.d\[\1, 1\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_vnum_za64_7_1_1,
+ svld1_hor_vnum_za64 (7, 1, p0, x1, 1),
+ svld1_hor_vnum_za64 (7, 1, p0, x1, 1))
+
+/*
+** ld1_vnum_za64_0_0_2:
+** incb x1, all, mul #2
+** mov (w1[2-5]), #?2
+** ld1d { za0h\.d\[\1, 0\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_vnum_za64_0_0_2,
+ svld1_hor_vnum_za64 (0, 0, p0, x1, 2),
+ svld1_hor_vnum_za64 (0, 0, p0, x1, 2))
+
+/*
+** ld1_vnum_za64_5_1_2:
+** incb x1, all, mul #2
+** mov (w1[2-5]), #?3
+** ld1d { za5h\.d\[\1, 0\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_vnum_za64_5_1_2,
+ svld1_hor_vnum_za64 (5, 1, p0, x1, 2),
+ svld1_hor_vnum_za64 (5, 1, p0, x1, 2))
+
+/*
+** ld1_vnum_za64_0_w0_0:
+** mov (w1[2-5]), w0
+** ld1d { za0h\.d\[\1, 0\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_vnum_za64_0_w0_0,
+ svld1_hor_vnum_za64 (0, w0, p0, x1, 0),
+ svld1_hor_vnum_za64 (0, w0, p0, x1, 0))
+
+/*
+** ld1_vnum_za64_0_w0_1:
+** incb x1
+** mov (w1[2-5]), w0
+** ld1d { za0h\.d\[\1, 1\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_vnum_za64_0_w0_1,
+ svld1_hor_vnum_za64 (0, w0, p0, x1, 1),
+ svld1_hor_vnum_za64 (0, w0, p0, x1, 1))
+
+/*
+** ld1_vnum_za64_6_w0_2:
+** incb x1, all, mul #2
+** add (w1[2-5]), w0, #?2
+** ld1d { za6h\.d\[\1, 0\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_vnum_za64_6_w0_2,
+ svld1_hor_vnum_za64 (6, w0, p0, x1, 2),
+ svld1_hor_vnum_za64 (6, w0, p0, x1, 2))
+
+/*
+** ld1_vnum_za64_2_w0_13:
+** incb x1, all, mul #13
+** add (w1[2-5]), w0, #?13
+** ld1d { za2h\.d\[\1, 0\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_vnum_za64_2_w0_13,
+ svld1_hor_vnum_za64 (2, w0, p0, x1, 13),
+ svld1_hor_vnum_za64 (2, w0, p0, x1, 13))
+
+/*
+** ld1_vnum_za64_4_w0_x2:
+** cntb (x[0-9]+)
+** madd (x[0-9]+), (?:\1, x2|x2, \1), x1
+** add (w1[2-5]), (?:w0, w2|w2, w0)
+** ld1d { za4h\.d\[\3, 0\] }, p0/z, \[\2\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_vnum_za64_4_w0_x2,
+ svld1_hor_vnum_za64 (4, w0, p0, x1, x2),
+ svld1_hor_vnum_za64 (4, w0, p0, x1, x2))
+
+/*
+** ld1_vnum_za64_1_w0p1_0:
+** mov (w1[2-5]), w0
+** ld1d { za1h\.d\[\1, 1\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_vnum_za64_1_w0p1_0,
+ svld1_hor_vnum_za64 (1, w0 + 1, p0, x1, 0),
+ svld1_hor_vnum_za64 (1, w0 + 1, p0, x1, 0))
--- /dev/null
+/* { dg-final { check-function-bodies "**" "" "-DCHECK_ASM" } } */
+
+#include "test_sme_acle.h"
+
+/*
+** ld1_vnum_za8_0_0_1:
+** incb x1
+** mov (w1[2-5]), (?:wzr|#?0)
+** ld1b { za0h\.b\[\1, 1\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_vnum_za8_0_0_1,
+ svld1_hor_vnum_za8 (0, 0, p0, x1, 1),
+ svld1_hor_vnum_za8 (0, 0, p0, x1, 1))
+
+/*
+** ld1_vnum_za8_0_1_1:
+** incb x1
+** mov (w1[2-5]), #?1
+** ld1b { za0h\.b\[\1, 1\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_vnum_za8_0_1_1,
+ svld1_hor_vnum_za8 (0, 1, p0, x1, 1),
+ svld1_hor_vnum_za8 (0, 1, p0, x1, 1))
+
+/*
+** ld1_vnum_za8_0_0_16:
+** incb x1, all, mul #16
+** mov (w1[2-5]), #?16
+** ld1b { za0h\.b\[\1, 0\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_vnum_za8_0_0_16,
+ svld1_hor_vnum_za8 (0, 0, p0, x1, 16),
+ svld1_hor_vnum_za8 (0, 0, p0, x1, 16))
+
+/*
+** ld1_vnum_za8_0_1_16:
+** incb x1, all, mul #16
+** mov (w1[2-5]), #?17
+** ld1b { za0h\.b\[\1, 0\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_vnum_za8_0_1_16,
+ svld1_hor_vnum_za8 (0, 1, p0, x1, 16),
+ svld1_hor_vnum_za8 (0, 1, p0, x1, 16))
+
+/*
+** ld1_vnum_za8_0_w0_0:
+** mov (w1[2-5]), w0
+** ld1b { za0h\.b\[\1, 0\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_vnum_za8_0_w0_0,
+ svld1_hor_vnum_za8 (0, w0, p0, x1, 0),
+ svld1_hor_vnum_za8 (0, w0, p0, x1, 0))
+
+/*
+** ld1_vnum_za8_0_w0_1:
+** incb x1
+** mov (w1[2-5]), w0
+** ld1b { za0h\.b\[\1, 1\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_vnum_za8_0_w0_1,
+ svld1_hor_vnum_za8 (0, w0, p0, x1, 1),
+ svld1_hor_vnum_za8 (0, w0, p0, x1, 1))
+
+/*
+** ld1_vnum_za8_0_w0_15:
+** incb x1, all, mul #15
+** mov (w1[2-5]), w0
+** ld1b { za0h\.b\[\1, 15\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_vnum_za8_0_w0_15,
+ svld1_hor_vnum_za8 (0, w0, p0, x1, 15),
+ svld1_hor_vnum_za8 (0, w0, p0, x1, 15))
+
+/*
+** ld1_vnum_za8_0_w0_16:
+** incb x1, all, mul #16
+** add (w1[2-5]), w0, #?16
+** ld1b { za0h\.b\[\1, 0\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_vnum_za8_0_w0_16,
+ svld1_hor_vnum_za8 (0, w0, p0, x1, 16),
+ svld1_hor_vnum_za8 (0, w0, p0, x1, 16))
+
+/*
+** ld1_vnum_za8_0_w0_x2:
+** cntb (x[0-9]+)
+** mul (x[0-9]+), (?:\1, x2|x2, \1)
+** add (w1[2-5]), (?:w0, w2|w2, w0)
+** ld1b { za0h\.b\[\3, 0\] }, p0/z, \[x1, \2\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_vnum_za8_0_w0_x2,
+ svld1_hor_vnum_za8 (0, w0, p0, x1, x2),
+ svld1_hor_vnum_za8 (0, w0, p0, x1, x2))
+
+/*
+** ld1_vnum_za8_0_w0p1_0:
+** mov (w1[2-5]), w0
+** ld1b { za0h\.b\[\1, 1\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_vnum_za8_0_w0p1_0,
+ svld1_hor_vnum_za8 (0, w0 + 1, p0, x1, 0),
+ svld1_hor_vnum_za8 (0, w0 + 1, p0, x1, 0))
--- /dev/null
+/* { dg-final { check-function-bodies "**" "" "-DCHECK_ASM" } } */
+
+#include "test_sme_acle.h"
+
+/*
+** ld1_za128_0_0:
+** mov (w1[2-5]), (?:wzr|#?0)
+** ld1q { za0h\.q\[\1, 0\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_za128_0_0,
+ svld1_hor_za128 (0, 0, p0, x1),
+ svld1_hor_za128 (0, 0, p0, x1))
+
+/*
+** ld1_za128_0_1:
+** mov (w1[2-5]), #?1
+** ld1q { za0h\.q\[\1, 0\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_za128_0_1,
+ svld1_hor_za128 (0, 1, p0, x1),
+ svld1_hor_za128 (0, 1, p0, x1))
+
+/*
+** ld1_za128_0_w0:
+** mov (w1[2-5]), w0
+** ld1q { za0h\.q\[\1, 0\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_za128_0_w0,
+ svld1_hor_za128 (0, w0, p0, x1),
+ svld1_hor_za128 (0, w0, p0, x1))
+
+/*
+** ld1_za128_0_w0_p1:
+** add (w1[2-5]), w0, #?1
+** ld1q { za0h\.q\[\1, 0\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_za128_0_w0_p1,
+ svld1_hor_za128 (0, w0 + 1, p0, x1),
+ svld1_hor_za128 (0, w0 + 1, p0, x1))
+
+/*
+** ld1_za128_7_w0:
+** mov (w1[2-5]), w0
+** ld1q { za7h\.q\[\1, 0\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_za128_7_w0,
+ svld1_hor_za128 (7, w0, p0, x1),
+ svld1_hor_za128 (7, w0, p0, x1))
+
+/*
+** ld1_za128_13_w0:
+** mov (w1[2-5]), w0
+** ld1q { za13h\.q\[\1, 0\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_za128_13_w0,
+ svld1_hor_za128 (13, w0, p0, x1),
+ svld1_hor_za128 (13, w0, p0, x1))
+
+/*
+** ld1_za128_15_w0:
+** mov (w1[2-5]), w0
+** ld1q { za15h\.q\[\1, 0\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_za128_15_w0,
+ svld1_hor_za128 (15, w0, p0, x1),
+ svld1_hor_za128 (15, w0, p0, x1))
+
+/*
+** ld1_za128_9_w0_index:
+** mov (w1[2-5]), w0
+** ld1q { za9h\.q\[\1, 0\] }, p0/z, \[x1, x2, lsl #?4\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_za128_9_w0_index,
+ svld1_hor_za128 (9, w0, p0, x1 + x2 * 16),
+ svld1_hor_za128 (9, w0, p0, x1 + x2 * 16))
--- /dev/null
+/* { dg-final { check-function-bodies "**" "" "-DCHECK_ASM" } } */
+
+#include "test_sme_acle.h"
+
+/*
+** ld1_za16_0_0:
+** mov (w1[2-5]), (?:wzr|#?0)
+** ld1h { za0h\.h\[\1, 0\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_za16_0_0,
+ svld1_hor_za16 (0, 0, p0, x1),
+ svld1_hor_za16 (0, 0, p0, x1))
+
+/* It would also be OK (and perhaps better) to move 0 into a register
+ and use an offset of 7. */
+/*
+** ld1_za16_0_7:
+** mov (w1[2-5]), #?7
+** ld1h { za0h\.h\[\1, 0\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_za16_0_7,
+ svld1_hor_za16 (0, 7, p0, x1),
+ svld1_hor_za16 (0, 7, p0, x1))
+
+/*
+** ld1_za16_0_8:
+** mov (w1[2-5]), #?8
+** ld1h { za0h\.h\[\1, 0\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_za16_0_8,
+ svld1_hor_za16 (0, 8, p0, x1),
+ svld1_hor_za16 (0, 8, p0, x1))
+
+/*
+** ld1_za16_0_w0:
+** mov (w1[2-5]), w0
+** ld1h { za0h\.h\[\1, 0\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_za16_0_w0,
+ svld1_hor_za16 (0, w0, p0, x1),
+ svld1_hor_za16 (0, w0, p0, x1))
+
+/*
+** ld1_za16_0_w0_p1:
+** mov (w1[2-5]), w0
+** ld1h { za0h\.h\[\1, 1\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_za16_0_w0_p1,
+ svld1_hor_za16 (0, w0 + 1, p0, x1),
+ svld1_hor_za16 (0, w0 + 1, p0, x1))
+
+/*
+** ld1_za16_0_w0_p7:
+** mov (w1[2-5]), w0
+** ld1h { za0h\.h\[\1, 7\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_za16_0_w0_p7,
+ svld1_hor_za16 (0, w0 + 7, p0, x1),
+ svld1_hor_za16 (0, w0 + 7, p0, x1))
+
+/*
+** ld1_za16_1_w0:
+** mov (w1[2-5]), w0
+** ld1h { za1h\.h\[\1, 0\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_za16_1_w0,
+ svld1_hor_za16 (1, w0, p0, x1),
+ svld1_hor_za16 (1, w0, p0, x1))
+
+
+/*
+** ld1_za16_1_w0_p1:
+** mov (w1[2-5]), w0
+** ld1h { za1h\.h\[\1, 1\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_za16_1_w0_p1,
+ svld1_hor_za16 (1, w0 + 1, p0, x1),
+ svld1_hor_za16 (1, w0 + 1, p0, x1))
+
+/*
+** ld1_za16_1_w0_p7:
+** mov (w1[2-5]), w0
+** ld1h { za1h\.h\[\1, 7\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_za16_1_w0_p7,
+ svld1_hor_za16 (1, w0 + 7, p0, x1),
+ svld1_hor_za16 (1, w0 + 7, p0, x1))
+
+/*
+** ld1_za16_1_w0_p5_index:
+** mov (w1[2-5]), w0
+** ld1h { za1h\.h\[\1, 5\] }, p0/z, \[x1, x2, lsl #?1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_za16_1_w0_p5_index,
+ svld1_hor_za16 (1, w0 + 5, p0, x1 + x2 * 2),
+ svld1_hor_za16 (1, w0 + 5, p0, x1 + x2 * 2))
+
+/*
+** ld1_za16_0_w0_p8:
+** add (w1[2-5]), w0, #?8
+** ld1h { za0h\.h\[\1, 0\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_za16_0_w0_p8,
+ svld1_hor_za16 (0, w0 + 8, p0, x1),
+ svld1_hor_za16 (0, w0 + 8, p0, x1))
+
+/*
+** ld1_za16_0_w0_m1:
+** sub (w1[2-5]), w0, #?1
+** ld1h { za0h\.h\[\1, 0\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_za16_0_w0_m1,
+ svld1_hor_za16 (0, w0 - 1, p0, x1),
+ svld1_hor_za16 (0, w0 - 1, p0, x1))
--- /dev/null
+/* { dg-final { check-function-bodies "**" "" "-DCHECK_ASM" } } */
+
+#include "test_sme_acle.h"
+
+/*
+** ld1_za32_0_0:
+** mov (w1[2-5]), (?:w0|#?0)
+** ld1w { za0h\.s\[\1, 0\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_za32_0_0,
+ svld1_hor_za32 (0, 0, p0, x1),
+ svld1_hor_za32 (0, 0, p0, x1))
+
+/* It would also be OK (and perhaps better) to move 0 into a register
+ and use an offset of 3. */
+/*
+** ld1_za32_0_3:
+** mov (w1[2-5]), #?3
+** ld1w { za0h\.s\[\1, 0\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_za32_0_3,
+ svld1_hor_za32 (0, 3, p0, x1),
+ svld1_hor_za32 (0, 3, p0, x1))
+
+/*
+** ld1_za32_0_4:
+** mov (w1[2-5]), #?4
+** ld1w { za0h\.s\[\1, 0\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_za32_0_4,
+ svld1_hor_za32 (0, 4, p0, x1),
+ svld1_hor_za32 (0, 4, p0, x1))
+
+/*
+** ld1_za32_0_w0:
+** mov (w1[2-5]), w0
+** ld1w { za0h\.s\[\1, 0\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_za32_0_w0,
+ svld1_hor_za32 (0, w0, p0, x1),
+ svld1_hor_za32 (0, w0, p0, x1))
+
+/*
+** ld1_za32_0_w0_p1:
+** mov (w1[2-5]), w0
+** ld1w { za0h\.s\[\1, 1\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_za32_0_w0_p1,
+ svld1_hor_za32 (0, w0 + 1, p0, x1),
+ svld1_hor_za32 (0, w0 + 1, p0, x1))
+
+/*
+** ld1_za32_0_w0_p3:
+** mov (w1[2-5]), w0
+** ld1w { za0h\.s\[\1, 3\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_za32_0_w0_p3,
+ svld1_hor_za32 (0, w0 + 3, p0, x1),
+ svld1_hor_za32 (0, w0 + 3, p0, x1))
+
+/*
+** ld1_za32_3_w0:
+** mov (w1[2-5]), w0
+** ld1w { za3h\.s\[\1, 0\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_za32_3_w0,
+ svld1_hor_za32 (3, w0, p0, x1),
+ svld1_hor_za32 (3, w0, p0, x1))
+
+/*
+** ld1_za32_3_w0_p1:
+** mov (w1[2-5]), w0
+** ld1w { za3h\.s\[\1, 1\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_za32_3_w0_p1,
+ svld1_hor_za32 (3, w0 + 1, p0, x1),
+ svld1_hor_za32 (3, w0 + 1, p0, x1))
+
+/*
+** ld1_za32_3_w0_p3:
+** mov (w1[2-5]), w0
+** ld1w { za3h\.s\[\1, 3\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_za32_3_w0_p3,
+ svld1_hor_za32 (3, w0 + 3, p0, x1),
+ svld1_hor_za32 (3, w0 + 3, p0, x1))
+
+/*
+** ld1_za32_1_w0_p2_index:
+** mov (w1[2-5]), w0
+** ld1w { za1h\.s\[\1, 2\] }, p0/z, \[x1, x2, lsl #?2\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_za32_1_w0_p2_index,
+ svld1_hor_za32 (1, w0 + 2, p0, x1 + x2 * 4),
+ svld1_hor_za32 (1, w0 + 2, p0, x1 + x2 * 4))
+
+/*
+** ld1_za32_0_w0_p4:
+** add (w1[2-5]), w0, #?4
+** ld1w { za0h\.s\[\1, 0\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_za32_0_w0_p4,
+ svld1_hor_za32 (0, w0 + 4, p0, x1),
+ svld1_hor_za32 (0, w0 + 4, p0, x1))
+
+/*
+** ld1_za32_0_w0_m1:
+** sub (w1[2-5]), w0, #?1
+** ld1w { za0h\.s\[\1, 0\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_za32_0_w0_m1,
+ svld1_hor_za32 (0, w0 - 1, p0, x1),
+ svld1_hor_za32 (0, w0 - 1, p0, x1))
--- /dev/null
+/* { dg-final { check-function-bodies "**" "" "-DCHECK_ASM" } } */
+
+#include "test_sme_acle.h"
+
+/*
+** ld1_za64_0_0:
+** mov (w1[2-5]), (?:wzr|#?0)
+** ld1d { za0h\.d\[\1, 0\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_za64_0_0,
+ svld1_hor_za64 (0, 0, p0, x1),
+ svld1_hor_za64 (0, 0, p0, x1))
+
+/* It would also be OK (and perhaps better) to move 0 into a register
+ and use an offset of 1. */
+/*
+** ld1_za64_0_1:
+** mov (w1[2-5]), #?1
+** ld1d { za0h\.d\[\1, 0\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_za64_0_1,
+ svld1_hor_za64 (0, 1, p0, x1),
+ svld1_hor_za64 (0, 1, p0, x1))
+
+/*
+** ld1_za64_0_2:
+** mov (w1[2-5]), #?2
+** ld1d { za0h\.d\[\1, 0\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_za64_0_2,
+ svld1_hor_za64 (0, 2, p0, x1),
+ svld1_hor_za64 (0, 2, p0, x1))
+
+/*
+** ld1_za64_0_w0:
+** mov (w1[2-5]), w0
+** ld1d { za0h\.d\[\1, 0\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_za64_0_w0,
+ svld1_hor_za64 (0, w0, p0, x1),
+ svld1_hor_za64 (0, w0, p0, x1))
+
+/*
+** ld1_za64_0_w0_p1:
+** mov (w1[2-5]), w0
+** ld1d { za0h\.d\[\1, 1\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_za64_0_w0_p1,
+ svld1_hor_za64 (0, w0 + 1, p0, x1),
+ svld1_hor_za64 (0, w0 + 1, p0, x1))
+
+/*
+** ld1_za64_7_w0:
+** mov (w1[2-5]), w0
+** ld1d { za7h\.d\[\1, 0\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_za64_7_w0,
+ svld1_hor_za64 (7, w0, p0, x1),
+ svld1_hor_za64 (7, w0, p0, x1))
+
+/*
+** ld1_za64_7_w0_p1:
+** mov (w1[2-5]), w0
+** ld1d { za7h\.d\[\1, 1\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_za64_7_w0_p1,
+ svld1_hor_za64 (7, w0 + 1, p0, x1),
+ svld1_hor_za64 (7, w0 + 1, p0, x1))
+
+/*
+** ld1_za64_5_w0_p1_index:
+** mov (w1[2-5]), w0
+** ld1d { za5h\.d\[\1, 1\] }, p0/z, \[x1, x2, lsl #?3\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_za64_5_w0_p1_index,
+ svld1_hor_za64 (5, w0 + 1, p0, x1 + x2 * 8),
+ svld1_hor_za64 (5, w0 + 1, p0, x1 + x2 * 8))
+
+/*
+** ld1_za64_0_w0_p2:
+** add (w1[2-5]), w0, #?2
+** ld1d { za0h\.d\[\1, 0\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_za64_0_w0_p2,
+ svld1_hor_za64 (0, w0 + 2, p0, x1),
+ svld1_hor_za64 (0, w0 + 2, p0, x1))
+
+/*
+** ld1_za64_0_w0_m1:
+** sub (w1[2-5]), w0, #?1
+** ld1d { za0h\.d\[\1, 0\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_za64_0_w0_m1,
+ svld1_hor_za64 (0, w0 - 1, p0, x1),
+ svld1_hor_za64 (0, w0 - 1, p0, x1))
--- /dev/null
+/* { dg-final { check-function-bodies "**" "" "-DCHECK_ASM" } } */
+
+#include "test_sme_acle.h"
+
+/*
+** ld1_za8_0_0:
+** mov (w1[2-5]), (?:wzr|#?0)
+** ld1b { za0h\.b\[\1, 0\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_za8_0_0,
+ svld1_hor_za8 (0, 0, p0, x1),
+ svld1_hor_za8 (0, 0, p0, x1))
+
+/* It would also be OK (and perhaps better) to move 0 into a register
+ and use an offset of 15. */
+/*
+** ld1_za8_0_15:
+** mov (w1[2-5]), #?15
+** ld1b { za0h\.b\[\1, 0\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_za8_0_15,
+ svld1_hor_za8 (0, 15, p0, x1),
+ svld1_hor_za8 (0, 15, p0, x1))
+
+/*
+** ld1_za8_0_16:
+** mov (w1[2-5]), #?16
+** ld1b { za0h\.b\[\1, 0\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_za8_0_16,
+ svld1_hor_za8 (0, 16, p0, x1),
+ svld1_hor_za8 (0, 16, p0, x1))
+
+/*
+** ld1_za8_0_w0:
+** mov (w1[2-5]), w0
+** ld1b { za0h\.b\[\1, 0\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_za8_0_w0,
+ svld1_hor_za8 (0, w0, p0, x1),
+ svld1_hor_za8 (0, w0, p0, x1))
+
+/*
+** ld1_za8_0_w0_p1:
+** mov (w1[2-5]), w0
+** ld1b { za0h\.b\[\1, 1\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_za8_0_w0_p1,
+ svld1_hor_za8 (0, w0 + 1, p0, x1),
+ svld1_hor_za8 (0, w0 + 1, p0, x1))
+
+/*
+** ld1_za8_0_w0_p15:
+** mov (w1[2-5]), w0
+** ld1b { za0h\.b\[\1, 15\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_za8_0_w0_p15,
+ svld1_hor_za8 (0, w0 + 15, p0, x1),
+ svld1_hor_za8 (0, w0 + 15, p0, x1))
+
+/*
+** ld1_za8_0_w0_p13_index:
+** mov (w1[2-5]), w0
+** ld1b { za0h\.b\[\1, 15\] }, p0/z, \[x1, x2\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_za8_0_w0_p13_index,
+ svld1_hor_za8 (0, w0 + 15, p0, x1 + x2),
+ svld1_hor_za8 (0, w0 + 15, p0, x1 + x2))
+
+/*
+** ld1_za8_0_w0_p16:
+** add (w1[2-5]), w0, #?16
+** ld1b { za0h\.b\[\1, 0\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_za8_0_w0_p16,
+ svld1_hor_za8 (0, w0 + 16, p0, x1),
+ svld1_hor_za8 (0, w0 + 16, p0, x1))
+
+/*
+** ld1_za8_0_w0_m1:
+** sub (w1[2-5]), w0, #?1
+** ld1b { za0h\.b\[\1, 0\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_za8_0_w0_m1,
+ svld1_hor_za8 (0, w0 - 1, p0, x1),
+ svld1_hor_za8 (0, w0 - 1, p0, x1))
--- /dev/null
+/* { dg-final { check-function-bodies "**" "" "-DCHECK_ASM" } } */
+
+#include "test_sme_acle.h"
+
+/*
+** ld1_vnum_za128_0_0_0:
+** mov (w1[2-5]), (?:wzr|#?0)
+** ld1q { za0v\.q\[\1, 0\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_vnum_za128_0_0_0,
+ svld1_ver_vnum_za128 (0, 0, p0, x1, 0),
+ svld1_ver_vnum_za128 (0, 0, p0, x1, 0))
+
+/*
+** ld1_vnum_za128_7_1_0:
+** mov (w1[2-5]), #?1
+** ld1q { za7v\.q\[\1, 0\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_vnum_za128_7_1_0,
+ svld1_ver_vnum_za128 (7, 1, p0, x1, 0),
+ svld1_ver_vnum_za128 (7, 1, p0, x1, 0))
+
+/*
+** ld1_vnum_za128_11_1_5:
+** incb x1, all, mul #5
+** mov (w1[2-5]), #?6
+** ld1q { za11v\.q\[\1, 0\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_vnum_za128_11_1_5,
+ svld1_ver_vnum_za128 (11, 1, p0, x1, 5),
+ svld1_ver_vnum_za128 (11, 1, p0, x1, 5))
+
+/*
+** ld1_vnum_za128_3_w0_0:
+** mov (w1[2-5]), w0
+** ld1q { za3v\.q\[\1, 0\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_vnum_za128_3_w0_0,
+ svld1_ver_vnum_za128 (3, w0, p0, x1, 0),
+ svld1_ver_vnum_za128 (3, w0, p0, x1, 0))
+
+/*
+** ld1_vnum_za128_5_w0_0:
+** incb x1, all, mul #13
+** add (w1[2-5]), w0, #?13
+** ld1q { za5v\.q\[\1, 0\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_vnum_za128_5_w0_0,
+ svld1_ver_vnum_za128 (5, w0, p0, x1, 13),
+ svld1_ver_vnum_za128 (5, w0, p0, x1, 13))
+
+/*
+** ld1_vnum_za128_11_w0_0:
+** cntb (x[0-9]+)
+** madd (x[0-9]+), (?:\1, x2|x2, \1), x1
+** add (w1[2-5]), (?:w0, w2|w2, w0)
+** ld1q { za11v\.q\[\3, 0\] }, p0/z, \[\2\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_vnum_za128_11_w0_0,
+ svld1_ver_vnum_za128 (11, w0, p0, x1, x2),
+ svld1_ver_vnum_za128 (11, w0, p0, x1, x2))
+
+/*
+** ld1_vnum_za128_15_w0p1_0:
+** add (w1[2-5]), w0, #?1
+** ld1q { za15v\.q\[\1, 0\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_vnum_za128_15_w0p1_0,
+ svld1_ver_vnum_za128 (15, w0 + 1, p0, x1, 0),
+ svld1_ver_vnum_za128 (15, w0 + 1, p0, x1, 0))
--- /dev/null
+/* { dg-final { check-function-bodies "**" "" "-DCHECK_ASM" } } */
+
+#include "test_sme_acle.h"
+
+/*
+** ld1_vnum_za16_1_0_1:
+** incb x1
+** mov (w1[2-5]), (?:wzr|#?0)
+** ld1h { za1v\.h\[\1, 1\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_vnum_za16_1_0_1,
+ svld1_ver_vnum_za16 (1, 0, p0, x1, 1),
+ svld1_ver_vnum_za16 (1, 0, p0, x1, 1))
+
+/*
+** ld1_vnum_za16_1_1_1:
+** incb x1
+** mov (w1[2-5]), #?1
+** ld1h { za1v\.h\[\1, 1\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_vnum_za16_1_1_1,
+ svld1_ver_vnum_za16 (1, 1, p0, x1, 1),
+ svld1_ver_vnum_za16 (1, 1, p0, x1, 1))
+
+/*
+** ld1_vnum_za16_0_0_8:
+** incb x1, all, mul #8
+** mov (w1[2-5]), #?8
+** ld1h { za0v\.h\[\1, 0\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_vnum_za16_0_0_8,
+ svld1_ver_vnum_za16 (0, 0, p0, x1, 8),
+ svld1_ver_vnum_za16 (0, 0, p0, x1, 8))
+
+/*
+** ld1_vnum_za16_0_1_8:
+** incb x1, all, mul #8
+** mov (w1[2-5]), #?9
+** ld1h { za0v\.h\[\1, 0\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_vnum_za16_0_1_8,
+ svld1_ver_vnum_za16 (0, 1, p0, x1, 8),
+ svld1_ver_vnum_za16 (0, 1, p0, x1, 8))
+
+/*
+** ld1_vnum_za16_0_w0_0:
+** mov (w1[2-5]), w0
+** ld1h { za0v\.h\[\1, 0\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_vnum_za16_0_w0_0,
+ svld1_ver_vnum_za16 (0, w0, p0, x1, 0),
+ svld1_ver_vnum_za16 (0, w0, p0, x1, 0))
+
+/*
+** ld1_vnum_za16_0_w0_1:
+** incb x1
+** mov (w1[2-5]), w0
+** ld1h { za0v\.h\[\1, 1\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_vnum_za16_0_w0_1,
+ svld1_ver_vnum_za16 (0, w0, p0, x1, 1),
+ svld1_ver_vnum_za16 (0, w0, p0, x1, 1))
+
+/*
+** ld1_vnum_za16_0_w0_7:
+** incb x1, all, mul #7
+** mov (w1[2-5]), w0
+** ld1h { za0v\.h\[\1, 7\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_vnum_za16_0_w0_7,
+ svld1_ver_vnum_za16 (0, w0, p0, x1, 7),
+ svld1_ver_vnum_za16 (0, w0, p0, x1, 7))
+
+/*
+** ld1_vnum_za16_1_w0_8:
+** incb x1, all, mul #8
+** add (w1[2-5]), w0, #?8
+** ld1h { za1v\.h\[\1, 0\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_vnum_za16_1_w0_8,
+ svld1_ver_vnum_za16 (1, w0, p0, x1, 8),
+ svld1_ver_vnum_za16 (1, w0, p0, x1, 8))
+
+/*
+** ld1_vnum_za16_1_w0_13:
+** incb x1, all, mul #13
+** add (w1[2-5]), w0, #?13
+** ld1h { za1v\.h\[\1, 0\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_vnum_za16_1_w0_13,
+ svld1_ver_vnum_za16 (1, w0, p0, x1, 13),
+ svld1_ver_vnum_za16 (1, w0, p0, x1, 13))
+
+/*
+** ld1_vnum_za16_0_w0_x2:
+** cntb (x[0-9]+)
+** madd (x[0-9]+), (?:\1, x2|x2, \1), x1
+** add (w1[2-5]), (?:w0, w2|w2, w0)
+** ld1h { za0v\.h\[\3, 0\] }, p0/z, \[\2\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_vnum_za16_0_w0_x2,
+ svld1_ver_vnum_za16 (0, w0, p0, x1, x2),
+ svld1_ver_vnum_za16 (0, w0, p0, x1, x2))
+
+/*
+** ld1_vnum_za16_1_w0p1_0:
+** mov (w1[2-5]), w0
+** ld1h { za1v\.h\[\1, 1\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_vnum_za16_1_w0p1_0,
+ svld1_ver_vnum_za16 (1, w0 + 1, p0, x1, 0),
+ svld1_ver_vnum_za16 (1, w0 + 1, p0, x1, 0))
--- /dev/null
+/* { dg-final { check-function-bodies "**" "" "-DCHECK_ASM" } } */
+
+#include "test_sme_acle.h"
+
+/*
+** ld1_vnum_za32_3_0_1:
+** incb x1
+** mov (w1[2-5]), (?:wzr|#?0)
+** ld1w { za3v\.s\[\1, 1\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_vnum_za32_3_0_1,
+ svld1_ver_vnum_za32 (3, 0, p0, x1, 1),
+ svld1_ver_vnum_za32 (3, 0, p0, x1, 1))
+
+/*
+** ld1_vnum_za32_2_1_1:
+** incb x1
+** mov (w1[2-5]), #?1
+** ld1w { za2v\.s\[\1, 1\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_vnum_za32_2_1_1,
+ svld1_ver_vnum_za32 (2, 1, p0, x1, 1),
+ svld1_ver_vnum_za32 (2, 1, p0, x1, 1))
+
+/*
+** ld1_vnum_za32_0_0_4:
+** incb x1, all, mul #4
+** mov (w1[2-5]), #?4
+** ld1w { za0v\.s\[\1, 0\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_vnum_za32_0_0_4,
+ svld1_ver_vnum_za32 (0, 0, p0, x1, 4),
+ svld1_ver_vnum_za32 (0, 0, p0, x1, 4))
+
+/*
+** ld1_vnum_za32_2_1_4:
+** incb x1, all, mul #4
+** mov (w1[2-5]), #?5
+** ld1w { za2v\.s\[\1, 0\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_vnum_za32_2_1_4,
+ svld1_ver_vnum_za32 (2, 1, p0, x1, 4),
+ svld1_ver_vnum_za32 (2, 1, p0, x1, 4))
+
+/*
+** ld1_vnum_za32_0_w0_0:
+** mov (w1[2-5]), w0
+** ld1w { za0v\.s\[\1, 0\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_vnum_za32_0_w0_0,
+ svld1_ver_vnum_za32 (0, w0, p0, x1, 0),
+ svld1_ver_vnum_za32 (0, w0, p0, x1, 0))
+
+/*
+** ld1_vnum_za32_0_w0_1:
+** incb x1
+** mov (w1[2-5]), w0
+** ld1w { za0v\.s\[\1, 1\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_vnum_za32_0_w0_1,
+ svld1_ver_vnum_za32 (0, w0, p0, x1, 1),
+ svld1_ver_vnum_za32 (0, w0, p0, x1, 1))
+
+/*
+** ld1_vnum_za32_0_w0_3:
+** incb x1, all, mul #3
+** mov (w1[2-5]), w0
+** ld1w { za0v\.s\[\1, 3\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_vnum_za32_0_w0_3,
+ svld1_ver_vnum_za32 (0, w0, p0, x1, 3),
+ svld1_ver_vnum_za32 (0, w0, p0, x1, 3))
+
+/*
+** ld1_vnum_za32_1_w0_4:
+** incb x1, all, mul #4
+** add (w1[2-5]), w0, #?4
+** ld1w { za1v\.s\[\1, 0\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_vnum_za32_1_w0_4,
+ svld1_ver_vnum_za32 (1, w0, p0, x1, 4),
+ svld1_ver_vnum_za32 (1, w0, p0, x1, 4))
+
+/*
+** ld1_vnum_za32_3_w0_13:
+** incb x1, all, mul #13
+** add (w1[2-5]), w0, #?13
+** ld1w { za3v\.s\[\1, 0\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_vnum_za32_3_w0_13,
+ svld1_ver_vnum_za32 (3, w0, p0, x1, 13),
+ svld1_ver_vnum_za32 (3, w0, p0, x1, 13))
+
+/*
+** ld1_vnum_za32_0_w0_x2:
+** cntb (x[0-9]+)
+** madd (x[0-9]+), (?:\1, x2|x2, \1), x1
+** add (w1[2-5]), (?:w0, w2|w2, w0)
+** ld1w { za0v\.s\[\3, 0\] }, p0/z, \[\2\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_vnum_za32_0_w0_x2,
+ svld1_ver_vnum_za32 (0, w0, p0, x1, x2),
+ svld1_ver_vnum_za32 (0, w0, p0, x1, x2))
+
+/*
+** ld1_vnum_za32_1_w0p1_0:
+** mov (w1[2-5]), w0
+** ld1w { za1v\.s\[\1, 1\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_vnum_za32_1_w0p1_0,
+ svld1_ver_vnum_za32 (1, w0 + 1, p0, x1, 0),
+ svld1_ver_vnum_za32 (1, w0 + 1, p0, x1, 0))
--- /dev/null
+/* { dg-final { check-function-bodies "**" "" "-DCHECK_ASM" } } */
+
+#include "test_sme_acle.h"
+
+/*
+** ld1_vnum_za64_3_0_1:
+** incb x1
+** mov (w1[2-5]), (?:wzr|#?0)
+** ld1d { za3v\.d\[\1, 1\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_vnum_za64_3_0_1,
+ svld1_ver_vnum_za64 (3, 0, p0, x1, 1),
+ svld1_ver_vnum_za64 (3, 0, p0, x1, 1))
+
+/*
+** ld1_vnum_za64_7_1_1:
+** incb x1
+** mov (w1[2-5]), #?1
+** ld1d { za7v\.d\[\1, 1\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_vnum_za64_7_1_1,
+ svld1_ver_vnum_za64 (7, 1, p0, x1, 1),
+ svld1_ver_vnum_za64 (7, 1, p0, x1, 1))
+
+/*
+** ld1_vnum_za64_0_0_2:
+** incb x1, all, mul #2
+** mov (w1[2-5]), #?2
+** ld1d { za0v\.d\[\1, 0\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_vnum_za64_0_0_2,
+ svld1_ver_vnum_za64 (0, 0, p0, x1, 2),
+ svld1_ver_vnum_za64 (0, 0, p0, x1, 2))
+
+/*
+** ld1_vnum_za64_5_1_2:
+** incb x1, all, mul #2
+** mov (w1[2-5]), #?3
+** ld1d { za5v\.d\[\1, 0\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_vnum_za64_5_1_2,
+ svld1_ver_vnum_za64 (5, 1, p0, x1, 2),
+ svld1_ver_vnum_za64 (5, 1, p0, x1, 2))
+
+/*
+** ld1_vnum_za64_0_w0_0:
+** mov (w1[2-5]), w0
+** ld1d { za0v\.d\[\1, 0\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_vnum_za64_0_w0_0,
+ svld1_ver_vnum_za64 (0, w0, p0, x1, 0),
+ svld1_ver_vnum_za64 (0, w0, p0, x1, 0))
+
+/*
+** ld1_vnum_za64_0_w0_1:
+** incb x1
+** mov (w1[2-5]), w0
+** ld1d { za0v\.d\[\1, 1\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_vnum_za64_0_w0_1,
+ svld1_ver_vnum_za64 (0, w0, p0, x1, 1),
+ svld1_ver_vnum_za64 (0, w0, p0, x1, 1))
+
+/*
+** ld1_vnum_za64_6_w0_2:
+** incb x1, all, mul #2
+** add (w1[2-5]), w0, #?2
+** ld1d { za6v\.d\[\1, 0\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_vnum_za64_6_w0_2,
+ svld1_ver_vnum_za64 (6, w0, p0, x1, 2),
+ svld1_ver_vnum_za64 (6, w0, p0, x1, 2))
+
+/*
+** ld1_vnum_za64_2_w0_13:
+** incb x1, all, mul #13
+** add (w1[2-5]), w0, #?13
+** ld1d { za2v\.d\[\1, 0\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_vnum_za64_2_w0_13,
+ svld1_ver_vnum_za64 (2, w0, p0, x1, 13),
+ svld1_ver_vnum_za64 (2, w0, p0, x1, 13))
+
+/*
+** ld1_vnum_za64_4_w0_x2:
+** cntb (x[0-9]+)
+** madd (x[0-9]+), (?:\1, x2|x2, \1), x1
+** add (w1[2-5]), (?:w0, w2|w2, w0)
+** ld1d { za4v\.d\[\3, 0\] }, p0/z, \[\2\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_vnum_za64_4_w0_x2,
+ svld1_ver_vnum_za64 (4, w0, p0, x1, x2),
+ svld1_ver_vnum_za64 (4, w0, p0, x1, x2))
+
+/*
+** ld1_vnum_za64_1_w0p1_0:
+** mov (w1[2-5]), w0
+** ld1d { za1v\.d\[\1, 1\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_vnum_za64_1_w0p1_0,
+ svld1_ver_vnum_za64 (1, w0 + 1, p0, x1, 0),
+ svld1_ver_vnum_za64 (1, w0 + 1, p0, x1, 0))
--- /dev/null
+/* { dg-final { check-function-bodies "**" "" "-DCHECK_ASM" } } */
+
+#include "test_sme_acle.h"
+
+/*
+** ld1_vnum_za8_0_0_1:
+** incb x1
+** mov (w1[2-5]), (?:wzr|#?0)
+** ld1b { za0v\.b\[\1, 1\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_vnum_za8_0_0_1,
+ svld1_ver_vnum_za8 (0, 0, p0, x1, 1),
+ svld1_ver_vnum_za8 (0, 0, p0, x1, 1))
+
+/*
+** ld1_vnum_za8_0_1_1:
+** incb x1
+** mov (w1[2-5]), #?1
+** ld1b { za0v\.b\[\1, 1\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_vnum_za8_0_1_1,
+ svld1_ver_vnum_za8 (0, 1, p0, x1, 1),
+ svld1_ver_vnum_za8 (0, 1, p0, x1, 1))
+
+/*
+** ld1_vnum_za8_0_0_16:
+** incb x1, all, mul #16
+** mov (w1[2-5]), #?16
+** ld1b { za0v\.b\[\1, 0\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_vnum_za8_0_0_16,
+ svld1_ver_vnum_za8 (0, 0, p0, x1, 16),
+ svld1_ver_vnum_za8 (0, 0, p0, x1, 16))
+
+/*
+** ld1_vnum_za8_0_1_16:
+** incb x1, all, mul #16
+** mov (w1[2-5]), #?17
+** ld1b { za0v\.b\[\1, 0\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_vnum_za8_0_1_16,
+ svld1_ver_vnum_za8 (0, 1, p0, x1, 16),
+ svld1_ver_vnum_za8 (0, 1, p0, x1, 16))
+
+/*
+** ld1_vnum_za8_0_w0_0:
+** mov (w1[2-5]), w0
+** ld1b { za0v\.b\[\1, 0\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_vnum_za8_0_w0_0,
+ svld1_ver_vnum_za8 (0, w0, p0, x1, 0),
+ svld1_ver_vnum_za8 (0, w0, p0, x1, 0))
+
+/*
+** ld1_vnum_za8_0_w0_1:
+** incb x1
+** mov (w1[2-5]), w0
+** ld1b { za0v\.b\[\1, 1\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_vnum_za8_0_w0_1,
+ svld1_ver_vnum_za8 (0, w0, p0, x1, 1),
+ svld1_ver_vnum_za8 (0, w0, p0, x1, 1))
+
+/*
+** ld1_vnum_za8_0_w0_15:
+** incb x1, all, mul #15
+** mov (w1[2-5]), w0
+** ld1b { za0v\.b\[\1, 15\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_vnum_za8_0_w0_15,
+ svld1_ver_vnum_za8 (0, w0, p0, x1, 15),
+ svld1_ver_vnum_za8 (0, w0, p0, x1, 15))
+
+/*
+** ld1_vnum_za8_0_w0_16:
+** incb x1, all, mul #16
+** add (w1[2-5]), w0, #?16
+** ld1b { za0v\.b\[\1, 0\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_vnum_za8_0_w0_16,
+ svld1_ver_vnum_za8 (0, w0, p0, x1, 16),
+ svld1_ver_vnum_za8 (0, w0, p0, x1, 16))
+
+/*
+** ld1_vnum_za8_0_w0_x2:
+** cntb (x[0-9]+)
+** mul (x[0-9]+), (?:\1, x2|x2, \1)
+** add (w1[2-5]), (?:w0, w2|w2, w0)
+** ld1b { za0v\.b\[\3, 0\] }, p0/z, \[x1, \2\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_vnum_za8_0_w0_x2,
+ svld1_ver_vnum_za8 (0, w0, p0, x1, x2),
+ svld1_ver_vnum_za8 (0, w0, p0, x1, x2))
+
+/*
+** ld1_vnum_za8_0_w0p1_0:
+** mov (w1[2-5]), w0
+** ld1b { za0v\.b\[\1, 1\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_vnum_za8_0_w0p1_0,
+ svld1_ver_vnum_za8 (0, w0 + 1, p0, x1, 0),
+ svld1_ver_vnum_za8 (0, w0 + 1, p0, x1, 0))
--- /dev/null
+/* { dg-final { check-function-bodies "**" "" "-DCHECK_ASM" } } */
+
+#include "test_sme_acle.h"
+
+/*
+** ld1_za128_0_0:
+** mov (w1[2-5]), (?:wzr|#?0)
+** ld1q { za0v\.q\[\1, 0\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_za128_0_0,
+ svld1_ver_za128 (0, 0, p0, x1),
+ svld1_ver_za128 (0, 0, p0, x1))
+
+/*
+** ld1_za128_0_1:
+** mov (w1[2-5]), #?1
+** ld1q { za0v\.q\[\1, 0\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_za128_0_1,
+ svld1_ver_za128 (0, 1, p0, x1),
+ svld1_ver_za128 (0, 1, p0, x1))
+
+/*
+** ld1_za128_0_w0:
+** mov (w1[2-5]), w0
+** ld1q { za0v\.q\[\1, 0\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_za128_0_w0,
+ svld1_ver_za128 (0, w0, p0, x1),
+ svld1_ver_za128 (0, w0, p0, x1))
+
+/*
+** ld1_za128_0_w0_p1:
+** add (w1[2-5]), w0, #?1
+** ld1q { za0v\.q\[\1, 0\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_za128_0_w0_p1,
+ svld1_ver_za128 (0, w0 + 1, p0, x1),
+ svld1_ver_za128 (0, w0 + 1, p0, x1))
+
+/*
+** ld1_za128_7_w0:
+** mov (w1[2-5]), w0
+** ld1q { za7v\.q\[\1, 0\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_za128_7_w0,
+ svld1_ver_za128 (7, w0, p0, x1),
+ svld1_ver_za128 (7, w0, p0, x1))
+
+/*
+** ld1_za128_13_w0:
+** mov (w1[2-5]), w0
+** ld1q { za13v\.q\[\1, 0\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_za128_13_w0,
+ svld1_ver_za128 (13, w0, p0, x1),
+ svld1_ver_za128 (13, w0, p0, x1))
+
+/*
+** ld1_za128_15_w0:
+** mov (w1[2-5]), w0
+** ld1q { za15v\.q\[\1, 0\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_za128_15_w0,
+ svld1_ver_za128 (15, w0, p0, x1),
+ svld1_ver_za128 (15, w0, p0, x1))
+
+/*
+** ld1_za128_9_w0_index:
+** mov (w1[2-5]), w0
+** ld1q { za9v\.q\[\1, 0\] }, p0/z, \[x1, x2, lsl #?4\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_za128_9_w0_index,
+ svld1_ver_za128 (9, w0, p0, x1 + x2 * 16),
+ svld1_ver_za128 (9, w0, p0, x1 + x2 * 16))
--- /dev/null
+/* { dg-final { check-function-bodies "**" "" "-DCHECK_ASM" } } */
+
+#include "test_sme_acle.h"
+
+/*
+** ld1_za16_0_0:
+** mov (w1[2-5]), (?:wzr|#?0)
+** ld1h { za0v\.h\[\1, 0\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_za16_0_0,
+ svld1_ver_za16 (0, 0, p0, x1),
+ svld1_ver_za16 (0, 0, p0, x1))
+
+/* It would also be OK (and perhaps better) to move 0 into a register
+ and use an offset of 7. */
+/*
+** ld1_za16_0_7:
+** mov (w1[2-5]), #?7
+** ld1h { za0v\.h\[\1, 0\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_za16_0_7,
+ svld1_ver_za16 (0, 7, p0, x1),
+ svld1_ver_za16 (0, 7, p0, x1))
+
+/*
+** ld1_za16_0_8:
+** mov (w1[2-5]), #?8
+** ld1h { za0v\.h\[\1, 0\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_za16_0_8,
+ svld1_ver_za16 (0, 8, p0, x1),
+ svld1_ver_za16 (0, 8, p0, x1))
+
+/*
+** ld1_za16_0_w0:
+** mov (w1[2-5]), w0
+** ld1h { za0v\.h\[\1, 0\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_za16_0_w0,
+ svld1_ver_za16 (0, w0, p0, x1),
+ svld1_ver_za16 (0, w0, p0, x1))
+
+/*
+** ld1_za16_0_w0_p1:
+** mov (w1[2-5]), w0
+** ld1h { za0v\.h\[\1, 1\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_za16_0_w0_p1,
+ svld1_ver_za16 (0, w0 + 1, p0, x1),
+ svld1_ver_za16 (0, w0 + 1, p0, x1))
+
+/*
+** ld1_za16_0_w0_p7:
+** mov (w1[2-5]), w0
+** ld1h { za0v\.h\[\1, 7\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_za16_0_w0_p7,
+ svld1_ver_za16 (0, w0 + 7, p0, x1),
+ svld1_ver_za16 (0, w0 + 7, p0, x1))
+
+/*
+** ld1_za16_1_w0:
+** mov (w1[2-5]), w0
+** ld1h { za1v\.h\[\1, 0\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_za16_1_w0,
+ svld1_ver_za16 (1, w0, p0, x1),
+ svld1_ver_za16 (1, w0, p0, x1))
+
+
+/*
+** ld1_za16_1_w0_p1:
+** mov (w1[2-5]), w0
+** ld1h { za1v\.h\[\1, 1\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_za16_1_w0_p1,
+ svld1_ver_za16 (1, w0 + 1, p0, x1),
+ svld1_ver_za16 (1, w0 + 1, p0, x1))
+
+/*
+** ld1_za16_1_w0_p7:
+** mov (w1[2-5]), w0
+** ld1h { za1v\.h\[\1, 7\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_za16_1_w0_p7,
+ svld1_ver_za16 (1, w0 + 7, p0, x1),
+ svld1_ver_za16 (1, w0 + 7, p0, x1))
+
+/*
+** ld1_za16_1_w0_p5_index:
+** mov (w1[2-5]), w0
+** ld1h { za1v\.h\[\1, 5\] }, p0/z, \[x1, x2, lsl #?1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_za16_1_w0_p5_index,
+ svld1_ver_za16 (1, w0 + 5, p0, x1 + x2 * 2),
+ svld1_ver_za16 (1, w0 + 5, p0, x1 + x2 * 2))
+
+/*
+** ld1_za16_0_w0_p8:
+** add (w1[2-5]), w0, #?8
+** ld1h { za0v\.h\[\1, 0\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_za16_0_w0_p8,
+ svld1_ver_za16 (0, w0 + 8, p0, x1),
+ svld1_ver_za16 (0, w0 + 8, p0, x1))
+
+/*
+** ld1_za16_0_w0_m1:
+** sub (w1[2-5]), w0, #?1
+** ld1h { za0v\.h\[\1, 0\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_za16_0_w0_m1,
+ svld1_ver_za16 (0, w0 - 1, p0, x1),
+ svld1_ver_za16 (0, w0 - 1, p0, x1))
--- /dev/null
+/* { dg-final { check-function-bodies "**" "" "-DCHECK_ASM" } } */
+
+#include "test_sme_acle.h"
+
+/*
+** ld1_za32_0_0:
+** mov (w1[2-5]), (?:w0|#?0)
+** ld1w { za0v\.s\[\1, 0\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_za32_0_0,
+ svld1_ver_za32 (0, 0, p0, x1),
+ svld1_ver_za32 (0, 0, p0, x1))
+
+/* It would also be OK (and perhaps better) to move 0 into a register
+ and use an offset of 3. */
+/*
+** ld1_za32_0_3:
+** mov (w1[2-5]), #?3
+** ld1w { za0v\.s\[\1, 0\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_za32_0_3,
+ svld1_ver_za32 (0, 3, p0, x1),
+ svld1_ver_za32 (0, 3, p0, x1))
+
+/*
+** ld1_za32_0_4:
+** mov (w1[2-5]), #?4
+** ld1w { za0v\.s\[\1, 0\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_za32_0_4,
+ svld1_ver_za32 (0, 4, p0, x1),
+ svld1_ver_za32 (0, 4, p0, x1))
+
+/*
+** ld1_za32_0_w0:
+** mov (w1[2-5]), w0
+** ld1w { za0v\.s\[\1, 0\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_za32_0_w0,
+ svld1_ver_za32 (0, w0, p0, x1),
+ svld1_ver_za32 (0, w0, p0, x1))
+
+/*
+** ld1_za32_0_w0_p1:
+** mov (w1[2-5]), w0
+** ld1w { za0v\.s\[\1, 1\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_za32_0_w0_p1,
+ svld1_ver_za32 (0, w0 + 1, p0, x1),
+ svld1_ver_za32 (0, w0 + 1, p0, x1))
+
+/*
+** ld1_za32_0_w0_p3:
+** mov (w1[2-5]), w0
+** ld1w { za0v\.s\[\1, 3\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_za32_0_w0_p3,
+ svld1_ver_za32 (0, w0 + 3, p0, x1),
+ svld1_ver_za32 (0, w0 + 3, p0, x1))
+
+/*
+** ld1_za32_3_w0:
+** mov (w1[2-5]), w0
+** ld1w { za3v\.s\[\1, 0\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_za32_3_w0,
+ svld1_ver_za32 (3, w0, p0, x1),
+ svld1_ver_za32 (3, w0, p0, x1))
+
+/*
+** ld1_za32_3_w0_p1:
+** mov (w1[2-5]), w0
+** ld1w { za3v\.s\[\1, 1\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_za32_3_w0_p1,
+ svld1_ver_za32 (3, w0 + 1, p0, x1),
+ svld1_ver_za32 (3, w0 + 1, p0, x1))
+
+/*
+** ld1_za32_3_w0_p3:
+** mov (w1[2-5]), w0
+** ld1w { za3v\.s\[\1, 3\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_za32_3_w0_p3,
+ svld1_ver_za32 (3, w0 + 3, p0, x1),
+ svld1_ver_za32 (3, w0 + 3, p0, x1))
+
+/*
+** ld1_za32_1_w0_p2_index:
+** mov (w1[2-5]), w0
+** ld1w { za1v\.s\[\1, 2\] }, p0/z, \[x1, x2, lsl #?2\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_za32_1_w0_p2_index,
+ svld1_ver_za32 (1, w0 + 2, p0, x1 + x2 * 4),
+ svld1_ver_za32 (1, w0 + 2, p0, x1 + x2 * 4))
+
+/*
+** ld1_za32_0_w0_p4:
+** add (w1[2-5]), w0, #?4
+** ld1w { za0v\.s\[\1, 0\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_za32_0_w0_p4,
+ svld1_ver_za32 (0, w0 + 4, p0, x1),
+ svld1_ver_za32 (0, w0 + 4, p0, x1))
+
+/*
+** ld1_za32_0_w0_m1:
+** sub (w1[2-5]), w0, #?1
+** ld1w { za0v\.s\[\1, 0\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_za32_0_w0_m1,
+ svld1_ver_za32 (0, w0 - 1, p0, x1),
+ svld1_ver_za32 (0, w0 - 1, p0, x1))
--- /dev/null
+/* { dg-final { check-function-bodies "**" "" "-DCHECK_ASM" } } */
+
+#include "test_sme_acle.h"
+
+/*
+** ld1_za64_0_0:
+** mov (w1[2-5]), (?:wzr|#?0)
+** ld1d { za0v\.d\[\1, 0\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_za64_0_0,
+ svld1_ver_za64 (0, 0, p0, x1),
+ svld1_ver_za64 (0, 0, p0, x1))
+
+/* It would also be OK (and perhaps better) to move 0 into a register
+ and use an offset of 1. */
+/*
+** ld1_za64_0_1:
+** mov (w1[2-5]), #?1
+** ld1d { za0v\.d\[\1, 0\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_za64_0_1,
+ svld1_ver_za64 (0, 1, p0, x1),
+ svld1_ver_za64 (0, 1, p0, x1))
+
+/*
+** ld1_za64_0_2:
+** mov (w1[2-5]), #?2
+** ld1d { za0v\.d\[\1, 0\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_za64_0_2,
+ svld1_ver_za64 (0, 2, p0, x1),
+ svld1_ver_za64 (0, 2, p0, x1))
+
+/*
+** ld1_za64_0_w0:
+** mov (w1[2-5]), w0
+** ld1d { za0v\.d\[\1, 0\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_za64_0_w0,
+ svld1_ver_za64 (0, w0, p0, x1),
+ svld1_ver_za64 (0, w0, p0, x1))
+
+/*
+** ld1_za64_0_w0_p1:
+** mov (w1[2-5]), w0
+** ld1d { za0v\.d\[\1, 1\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_za64_0_w0_p1,
+ svld1_ver_za64 (0, w0 + 1, p0, x1),
+ svld1_ver_za64 (0, w0 + 1, p0, x1))
+
+/*
+** ld1_za64_7_w0:
+** mov (w1[2-5]), w0
+** ld1d { za7v\.d\[\1, 0\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_za64_7_w0,
+ svld1_ver_za64 (7, w0, p0, x1),
+ svld1_ver_za64 (7, w0, p0, x1))
+
+/*
+** ld1_za64_7_w0_p1:
+** mov (w1[2-5]), w0
+** ld1d { za7v\.d\[\1, 1\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_za64_7_w0_p1,
+ svld1_ver_za64 (7, w0 + 1, p0, x1),
+ svld1_ver_za64 (7, w0 + 1, p0, x1))
+
+/*
+** ld1_za64_5_w0_p1_index:
+** mov (w1[2-5]), w0
+** ld1d { za5v\.d\[\1, 1\] }, p0/z, \[x1, x2, lsl #?3\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_za64_5_w0_p1_index,
+ svld1_ver_za64 (5, w0 + 1, p0, x1 + x2 * 8),
+ svld1_ver_za64 (5, w0 + 1, p0, x1 + x2 * 8))
+
+/*
+** ld1_za64_0_w0_p2:
+** add (w1[2-5]), w0, #?2
+** ld1d { za0v\.d\[\1, 0\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_za64_0_w0_p2,
+ svld1_ver_za64 (0, w0 + 2, p0, x1),
+ svld1_ver_za64 (0, w0 + 2, p0, x1))
+
+/*
+** ld1_za64_0_w0_m1:
+** sub (w1[2-5]), w0, #?1
+** ld1d { za0v\.d\[\1, 0\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_za64_0_w0_m1,
+ svld1_ver_za64 (0, w0 - 1, p0, x1),
+ svld1_ver_za64 (0, w0 - 1, p0, x1))
--- /dev/null
+/* { dg-final { check-function-bodies "**" "" "-DCHECK_ASM" } } */
+
+#include "test_sme_acle.h"
+
+/*
+** ld1_za8_0_0:
+** mov (w1[2-5]), (?:wzr|#?0)
+** ld1b { za0v\.b\[\1, 0\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_za8_0_0,
+ svld1_ver_za8 (0, 0, p0, x1),
+ svld1_ver_za8 (0, 0, p0, x1))
+
+/* It would also be OK (and perhaps better) to move 0 into a register
+ and use an offset of 15. */
+/*
+** ld1_za8_0_15:
+** mov (w1[2-5]), #?15
+** ld1b { za0v\.b\[\1, 0\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_za8_0_15,
+ svld1_ver_za8 (0, 15, p0, x1),
+ svld1_ver_za8 (0, 15, p0, x1))
+
+/*
+** ld1_za8_0_16:
+** mov (w1[2-5]), #?16
+** ld1b { za0v\.b\[\1, 0\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_za8_0_16,
+ svld1_ver_za8 (0, 16, p0, x1),
+ svld1_ver_za8 (0, 16, p0, x1))
+
+/*
+** ld1_za8_0_w0:
+** mov (w1[2-5]), w0
+** ld1b { za0v\.b\[\1, 0\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_za8_0_w0,
+ svld1_ver_za8 (0, w0, p0, x1),
+ svld1_ver_za8 (0, w0, p0, x1))
+
+/*
+** ld1_za8_0_w0_p1:
+** mov (w1[2-5]), w0
+** ld1b { za0v\.b\[\1, 1\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_za8_0_w0_p1,
+ svld1_ver_za8 (0, w0 + 1, p0, x1),
+ svld1_ver_za8 (0, w0 + 1, p0, x1))
+
+/*
+** ld1_za8_0_w0_p15:
+** mov (w1[2-5]), w0
+** ld1b { za0v\.b\[\1, 15\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_za8_0_w0_p15,
+ svld1_ver_za8 (0, w0 + 15, p0, x1),
+ svld1_ver_za8 (0, w0 + 15, p0, x1))
+
+/*
+** ld1_za8_0_w0_p13_index:
+** mov (w1[2-5]), w0
+** ld1b { za0v\.b\[\1, 15\] }, p0/z, \[x1, x2\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_za8_0_w0_p13_index,
+ svld1_ver_za8 (0, w0 + 15, p0, x1 + x2),
+ svld1_ver_za8 (0, w0 + 15, p0, x1 + x2))
+
+/*
+** ld1_za8_0_w0_p16:
+** add (w1[2-5]), w0, #?16
+** ld1b { za0v\.b\[\1, 0\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_za8_0_w0_p16,
+ svld1_ver_za8 (0, w0 + 16, p0, x1),
+ svld1_ver_za8 (0, w0 + 16, p0, x1))
+
+/*
+** ld1_za8_0_w0_m1:
+** sub (w1[2-5]), w0, #?1
+** ld1b { za0v\.b\[\1, 0\] }, p0/z, \[x1\]
+** ret
+*/
+TEST_LOAD_ZA (ld1_za8_0_w0_m1,
+ svld1_ver_za8 (0, w0 - 1, p0, x1),
+ svld1_ver_za8 (0, w0 - 1, p0, x1))
--- /dev/null
+/* { dg-final { check-function-bodies "**" "" "-DCHECK_ASM" } } */
+
+#include "test_sme_acle.h"
+
+/*
+** ldr_vnum_za_0_0:
+** mov (w1[2-5]), (?:wzr|#?0)
+** ldr za\[\1, 0\], \[x1(?:, #0, mul vl)?\]
+** ret
+*/
+TEST_LOAD_ZA (ldr_vnum_za_0_0,
+ svldr_vnum_za (0, x1, 0),
+ svldr_vnum_za (0, x1, 0))
+
+/*
+** ldr_vnum_za_0_1:
+** mov (w1[2-5]), (?:wzr|#?0)
+** ldr za\[\1, 1\], \[x1(?:, #1, mul vl)?\]
+** ret
+*/
+TEST_LOAD_ZA (ldr_vnum_za_0_1,
+ svldr_vnum_za (0, x1, 1),
+ svldr_vnum_za (0, x1, 1))
+
+/*
+** ldr_vnum_za_1_0:
+** mov (w1[2-5]), #?1
+** ldr za\[\1, 0\], \[x1(?:, #0, mul vl)?\]
+** ret
+*/
+TEST_LOAD_ZA (ldr_vnum_za_1_0,
+ svldr_vnum_za (1, x1, 0),
+ svldr_vnum_za (1, x1, 0))
+
+/*
+** ldr_vnum_za_1_2:
+** mov (w1[2-5]), #?1
+** ldr za\[\1, 2\], \[x1(?:, #2, mul vl)?\]
+** ret
+*/
+TEST_LOAD_ZA (ldr_vnum_za_1_2,
+ svldr_vnum_za (1, x1, 2),
+ svldr_vnum_za (1, x1, 2))
+
+/*
+** ldr_vnum_za_w0_0:
+** mov (w1[2-5]), w0
+** ldr za\[\1, 0\], \[x1(?:, #0, mul vl)?\]
+** ret
+*/
+TEST_LOAD_ZA (ldr_vnum_za_w0_0,
+ svldr_vnum_za (w0, x1, 0),
+ svldr_vnum_za (w0, x1, 0))
+
+/*
+** ldr_vnum_za_w0_1:
+** mov (w1[2-5]), w0
+** ldr za\[\1, 1\], \[x1, #1, mul vl\]
+** ret
+*/
+TEST_LOAD_ZA (ldr_vnum_za_w0_1,
+ svldr_vnum_za (w0, x1, 1),
+ svldr_vnum_za (w0, x1, 1))
+
+/*
+** ldr_vnum_za_w0_13:
+** mov (w1[2-5]), w0
+** ldr za\[\1, 13\], \[x1, #13, mul vl\]
+** ret
+*/
+TEST_LOAD_ZA (ldr_vnum_za_w0_13,
+ svldr_vnum_za (w0, x1, 13),
+ svldr_vnum_za (w0, x1, 13))
+
+/*
+** ldr_vnum_za_w0_15:
+** mov (w1[2-5]), w0
+** ldr za\[\1, 15\], \[x1, #15, mul vl\]
+** ret
+*/
+TEST_LOAD_ZA (ldr_vnum_za_w0_15,
+ svldr_vnum_za (w0, x1, 15),
+ svldr_vnum_za (w0, x1, 15))
+
+/*
+** ldr_vnum_za_w0_16:
+** (
+** add (w1[2-5]), w0, #?16
+** incb x1, all, mul #16
+** ldr za\[\1, 0\], \[x1(?:, #0, mul vl)?\]
+** |
+** incb x1, all, mul #16
+** add (w1[2-5]), w0, #?16
+** ldr za\[\2, 0\], \[x1(?:, #0, mul vl)?\]
+** )
+** ret
+*/
+TEST_LOAD_ZA (ldr_vnum_za_w0_16,
+ svldr_vnum_za (w0, x1, 16),
+ svldr_vnum_za (w0, x1, 16))
+
+/*
+** ldr_vnum_za_w0_m1:
+** (
+** sub (w1[2-5]), w0, #?1
+** decb x1
+** ldr za\[\1, 0\], \[x1(?:, #0, mul vl)?\]
+** |
+** decb x1
+** sub (w1[2-5]), w0, #?1
+** ldr za\[\2, 0\], \[x1(?:, #0, mul vl)?\]
+** )
+** ret
+*/
+TEST_LOAD_ZA (ldr_vnum_za_w0_m1,
+ svldr_vnum_za (w0, x1, -1),
+ svldr_vnum_za (w0, x1, -1))
+
+/*
+** ldr_vnum_za_w0p1_0:
+** add (w1[2-5]), w0, #?1
+** ldr za\[\1, 0\], \[x1(?:, #0, mul vl)?\]
+** ret
+*/
+TEST_LOAD_ZA (ldr_vnum_za_w0p1_0,
+ svldr_vnum_za (w0 + 1, x1, 0),
+ svldr_vnum_za (w0 + 1, x1, 0))
+
+/*
+** ldr_vnum_za_w0m1_1:
+** sub (w1[2-5]), w0, #?1
+** ldr za\[\1, 1\], \[x1(?:, #1, mul vl)?\]
+** ret
+*/
+TEST_LOAD_ZA (ldr_vnum_za_w0m1_1,
+ svldr_vnum_za (w0 - 1, x1, 1),
+ svldr_vnum_za (w0 - 1, x1, 1))
+
+/*
+** ldr_vnum_za_w0p2_3:
+** add (w1[2-5]), w0, #?2
+** ldr za\[\1, 3\], \[x1(?:, #3, mul vl)?\]
+** ret
+*/
+TEST_LOAD_ZA (ldr_vnum_za_w0p2_3,
+ svldr_vnum_za (w0 + 2, x1, 3),
+ svldr_vnum_za (w0 + 2, x1, 3))
--- /dev/null
+/* { dg-final { check-function-bodies "**" "" "-DCHECK_ASM" } } */
+
+#define STREAMING_COMPATIBLE
+#include "test_sme_acle.h"
+
+/*
+** ldr_vnum_za_0_0:
+** mov (w1[2-5]), (?:wzr|#?0)
+** ldr za\[\1, 0\], \[x1(?:, #0, mul vl)?\]
+** ret
+*/
+TEST_LOAD_ZA (ldr_vnum_za_0_0,
+ svldr_vnum_za (0, x1, 0),
+ svldr_vnum_za (0, x1, 0))
+
+/*
+** ldr_vnum_za_0_1:
+** mov (w1[2-5]), (?:wzr|#?0)
+** ldr za\[\1, 1\], \[x1(?:, #1, mul vl)?\]
+** ret
+*/
+TEST_LOAD_ZA (ldr_vnum_za_0_1,
+ svldr_vnum_za (0, x1, 1),
+ svldr_vnum_za (0, x1, 1))
+
+/*
+** ldr_vnum_za_1_0:
+** mov (w1[2-5]), #?1
+** ldr za\[\1, 0\], \[x1(?:, #0, mul vl)?\]
+** ret
+*/
+TEST_LOAD_ZA (ldr_vnum_za_1_0,
+ svldr_vnum_za (1, x1, 0),
+ svldr_vnum_za (1, x1, 0))
+
+/*
+** ldr_vnum_za_1_2:
+** mov (w1[2-5]), #?1
+** ldr za\[\1, 2\], \[x1(?:, #2, mul vl)?\]
+** ret
+*/
+TEST_LOAD_ZA (ldr_vnum_za_1_2,
+ svldr_vnum_za (1, x1, 2),
+ svldr_vnum_za (1, x1, 2))
+
+/*
+** ldr_vnum_za_w0_0:
+** mov (w1[2-5]), w0
+** ldr za\[\1, 0\], \[x1(?:, #0, mul vl)?\]
+** ret
+*/
+TEST_LOAD_ZA (ldr_vnum_za_w0_0,
+ svldr_vnum_za (w0, x1, 0),
+ svldr_vnum_za (w0, x1, 0))
+
+/*
+** ldr_vnum_za_w0_1:
+** mov (w1[2-5]), w0
+** ldr za\[\1, 1\], \[x1, #1, mul vl\]
+** ret
+*/
+TEST_LOAD_ZA (ldr_vnum_za_w0_1,
+ svldr_vnum_za (w0, x1, 1),
+ svldr_vnum_za (w0, x1, 1))
+
+/*
+** ldr_vnum_za_w0_13:
+** mov (w1[2-5]), w0
+** ldr za\[\1, 13\], \[x1, #13, mul vl\]
+** ret
+*/
+TEST_LOAD_ZA (ldr_vnum_za_w0_13,
+ svldr_vnum_za (w0, x1, 13),
+ svldr_vnum_za (w0, x1, 13))
+
+/*
+** ldr_vnum_za_w0_15:
+** mov (w1[2-5]), w0
+** ldr za\[\1, 15\], \[x1, #15, mul vl\]
+** ret
+*/
+TEST_LOAD_ZA (ldr_vnum_za_w0_15,
+ svldr_vnum_za (w0, x1, 15),
+ svldr_vnum_za (w0, x1, 15))
+
+/*
+** ldr_vnum_za_w0_16:
+** (
+** add (w1[2-5]), w0, #?16
+** addsvl (x[0-9]+), x1, #16
+** ldr za\[\1, 0\], \[\2(?:, #0, mul vl)?\]
+** |
+** addsvl (x[0-9]+), x1, #16
+** add (w1[2-5]), w0, #?16
+** ldr za\[\4, 0\], \[\3(?:, #0, mul vl)?\]
+** )
+** ret
+*/
+TEST_LOAD_ZA (ldr_vnum_za_w0_16,
+ svldr_vnum_za (w0, x1, 16),
+ svldr_vnum_za (w0, x1, 16))
+
+/*
+** ldr_vnum_za_w0_m1:
+** (
+** sub (w1[2-5]), w0, #?1
+** addsvl (x[0-9]+), x1, #-1
+** ldr za\[\1, 0\], \[\2(?:, #0, mul vl)?\]
+** |
+** addsvl (x[0-9]+), x1, #-1
+** sub (w1[2-5]), w0, #?1
+** ldr za\[\4, 0\], \[\3(?:, #0, mul vl)?\]
+** )
+** ret
+*/
+TEST_LOAD_ZA (ldr_vnum_za_w0_m1,
+ svldr_vnum_za (w0, x1, -1),
+ svldr_vnum_za (w0, x1, -1))
+
+/*
+** ldr_vnum_za_w0p1_0:
+** add (w1[2-5]), w0, #?1
+** ldr za\[\1, 0\], \[x1(?:, #0, mul vl)?\]
+** ret
+*/
+TEST_LOAD_ZA (ldr_vnum_za_w0p1_0,
+ svldr_vnum_za (w0 + 1, x1, 0),
+ svldr_vnum_za (w0 + 1, x1, 0))
+
+/*
+** ldr_vnum_za_w0m1_1:
+** sub (w1[2-5]), w0, #?1
+** ldr za\[\1, 1\], \[x1(?:, #1, mul vl)?\]
+** ret
+*/
+TEST_LOAD_ZA (ldr_vnum_za_w0m1_1,
+ svldr_vnum_za (w0 - 1, x1, 1),
+ svldr_vnum_za (w0 - 1, x1, 1))
+
+/*
+** ldr_vnum_za_w0p2_3:
+** add (w1[2-5]), w0, #?2
+** ldr za\[\1, 3\], \[x1(?:, #3, mul vl)?\]
+** ret
+*/
+TEST_LOAD_ZA (ldr_vnum_za_w0p2_3,
+ svldr_vnum_za (w0 + 2, x1, 3),
+ svldr_vnum_za (w0 + 2, x1, 3))
--- /dev/null
+/* { dg-final { check-function-bodies "**" "" "-DCHECK_ASM" } } */
+
+#include "test_sme_acle.h"
+
+/*
+** ldr_za_0:
+** mov (w1[2-5]), (?:wzr|#?0)
+** ldr za\[\1, 0\], \[x1(?:, #0, mul vl)?\]
+** ret
+*/
+TEST_LOAD_ZA (ldr_za_0,
+ svldr_za (0, x1),
+ svldr_za (0, x1))
+
+/*
+** ldr_za_1:
+** mov (w1[2-5]), #?1
+** ldr za\[\1, 0\], \[x1(?:, #0, mul vl)?\]
+** ret
+*/
+TEST_LOAD_ZA (ldr_za_1,
+ svldr_za (1, x1),
+ svldr_za (1, x1))
+
+/*
+** ldr_za_w0:
+** mov (w1[2-5]), w0
+** ldr za\[\1, 0\], \[x1(?:, #0, mul vl)?\]
+** ret
+*/
+TEST_LOAD_ZA (ldr_za_w0,
+ svldr_za (w0, x1),
+ svldr_za (w0, x1))
+
+/*
+** ldr_za_w0_1_vnum:
+** mov (w1[2-5]), w0
+** ldr za\[\1, 1\], \[x1, #1, mul vl\]
+** ret
+*/
+TEST_LOAD_ZA (ldr_za_w0_1_vnum,
+ svldr_za (w0 + 1, x1 + svcntsb ()),
+ svldr_za (w0 + 1, x1 + svcntsb ()))
+
+/*
+** ldr_za_w0_13_vnum:
+** mov (w1[2-5]), w0
+** ldr za\[\1, 13\], \[x1, #13, mul vl\]
+** ret
+*/
+TEST_LOAD_ZA (ldr_za_w0_13_vnum,
+ svldr_za (w0 + 13, x1 + svcntsb () * 13),
+ svldr_za (w0 + 13, x1 + svcntsb () * 13))
+
+/*
+** ldr_za_w0_15_vnum:
+** mov (w1[2-5]), w0
+** ldr za\[\1, 15\], \[x1, #15, mul vl\]
+** ret
+*/
+TEST_LOAD_ZA (ldr_za_w0_15_vnum,
+ svldr_za (w0 + 15, x1 + svcntsb () * 15),
+ svldr_za (w0 + 15, x1 + svcntsb () * 15))
+
+/*
+** ldr_za_w0_16_vnum:
+** (
+** add (w1[2-5]), w0, #?16
+** incb x1, all, mul #16
+** ldr za\[\1, 0\], \[x1(?:, #0, mul vl)?\]
+** |
+** incb x1, all, mul #16
+** add (w1[2-5]), w0, #?16
+** ldr za\[\2, 0\], \[x1(?:, #0, mul vl)?\]
+** )
+** ret
+*/
+TEST_LOAD_ZA (ldr_za_w0_16_vnum,
+ svldr_za (w0 + 16, x1 + svcntsb () * 16),
+ svldr_za (w0 + 16, x1 + svcntsb () * 16))
+
+/*
+** ldr_za_w0_m1_vnum:
+** (
+** sub (w1[2-5]), w0, #?1
+** decb x1
+** ldr za\[\1, 0\], \[x1(?:, #0, mul vl)?\]
+** |
+** decb x1
+** sub (w1[2-5]), w0, #?1
+** ldr za\[\2, 0\], \[x1(?:, #0, mul vl)?\]
+** )
+** ret
+*/
+TEST_LOAD_ZA (ldr_za_w0_m1_vnum,
+ svldr_za (w0 - 1, x1 - svcntsb ()),
+ svldr_za (w0 - 1, x1 - svcntsb ()))
+
+/*
+** ldr_za_w0p2:
+** add (w1[2-5]), w0, #?2
+** ldr za\[\1, 0\], \[x1(?:, #0, mul vl)?\]
+** ret
+*/
+TEST_LOAD_ZA (ldr_za_w0p2,
+ svldr_za (w0 + 2, x1),
+ svldr_za (w0 + 2, x1))
+
+/*
+** ldr_za_offset:
+** (
+** mov (w1[2-5]), w0
+** add (x[0-9]+), x1, #?1
+** ldr za\[\1, 0\], \[\2(?:, #0, mul vl)?\]
+** |
+** add (x[0-9]+), x1, #?1
+** mov (w1[2-5]), w0
+** ldr za\[\4, 0\], \[\3(?:, #0, mul vl)?\]
+** )
+** ret
+*/
+TEST_LOAD_ZA (ldr_za_offset,
+ svldr_za (w0, x1 + 1),
+ svldr_za (w0, x1 + 1))
--- /dev/null
+/* { dg-final { check-function-bodies "**" "" "-DCHECK_ASM" } } */
+
+#define STREAMING_COMPATIBLE
+#include "test_sme_acle.h"
+
+/*
+** ldr_za_0:
+** mov (w1[2-5]), (?:wzr|#?0)
+** ldr za\[\1, 0\], \[x1(?:, #0, mul vl)?\]
+** ret
+*/
+TEST_LOAD_ZA (ldr_za_0,
+ svldr_za (0, x1),
+ svldr_za (0, x1))
+
+/*
+** ldr_za_1:
+** mov (w1[2-5]), #?1
+** ldr za\[\1, 0\], \[x1(?:, #0, mul vl)?\]
+** ret
+*/
+TEST_LOAD_ZA (ldr_za_1,
+ svldr_za (1, x1),
+ svldr_za (1, x1))
+
+/*
+** ldr_za_w0:
+** mov (w1[2-5]), w0
+** ldr za\[\1, 0\], \[x1(?:, #0, mul vl)?\]
+** ret
+*/
+TEST_LOAD_ZA (ldr_za_w0,
+ svldr_za (w0, x1),
+ svldr_za (w0, x1))
+
+/*
+** ldr_za_w0_1_vnum:
+** mov (w1[2-5]), w0
+** ldr za\[\1, 1\], \[x1, #1, mul vl\]
+** ret
+*/
+TEST_LOAD_ZA (ldr_za_w0_1_vnum,
+ svldr_za (w0 + 1, x1 + svcntsb ()),
+ svldr_za (w0 + 1, x1 + svcntsb ()))
+
+/*
+** ldr_za_w0p2:
+** add (w1[2-5]), w0, #?2
+** ldr za\[\1, 0\], \[x1(?:, #0, mul vl)?\]
+** ret
+*/
+TEST_LOAD_ZA (ldr_za_w0p2,
+ svldr_za (w0 + 2, x1),
+ svldr_za (w0 + 2, x1))
+
+/*
+** ldr_za_offset:
+** (
+** mov (w1[2-5]), w0
+** add (x[0-9]+), x1, #?1
+** ldr za\[\1, 0\], \[\2(?:, #0, mul vl)?\]
+** |
+** add (x[0-9]+), x1, #?1
+** mov (w1[2-5]), w0
+** ldr za\[\4, 0\], \[\3(?:, #0, mul vl)?\]
+** )
+** ret
+*/
+TEST_LOAD_ZA (ldr_za_offset,
+ svldr_za (w0, x1 + 1),
+ svldr_za (w0, x1 + 1))
--- /dev/null
+/* { dg-final { check-function-bodies "**" "" "-DCHECK_ASM" } } */
+
+#include "test_sme_acle.h"
+
+/*
+** mopa_za32_s8_0_p0_p1_z0_z1:
+** smopa za0\.s, p0/m, p1/m, z0\.b, z1\.b
+** ret
+*/
+TEST_UNIFORM_ZA (mopa_za32_s8_0_p0_p1_z0_z1, svint8_t,
+ svmopa_za32_s8_m (0, p0, p1, z0, z1),
+ svmopa_za32_m (0, p0, p1, z0, z1))
+
+/*
+** mopa_za32_s8_0_p1_p0_z1_z0:
+** smopa za0\.s, p1/m, p0/m, z1\.b, z0\.b
+** ret
+*/
+TEST_UNIFORM_ZA (mopa_za32_s8_0_p1_p0_z1_z0, svint8_t,
+ svmopa_za32_s8_m (0, p1, p0, z1, z0),
+ svmopa_za32_m (0, p1, p0, z1, z0))
+
+/*
+** mopa_za32_s8_3_p0_p1_z0_z1:
+** smopa za3\.s, p0/m, p1/m, z0\.b, z1\.b
+** ret
+*/
+TEST_UNIFORM_ZA (mopa_za32_s8_3_p0_p1_z0_z1, svint8_t,
+ svmopa_za32_s8_m (3, p0, p1, z0, z1),
+ svmopa_za32_m (3, p0, p1, z0, z1))
+
+/*
+** mopa_za32_u8_0_p0_p1_z0_z1:
+** umopa za0\.s, p0/m, p1/m, z0\.b, z1\.b
+** ret
+*/
+TEST_UNIFORM_ZA (mopa_za32_u8_0_p0_p1_z0_z1, svuint8_t,
+ svmopa_za32_u8_m (0, p0, p1, z0, z1),
+ svmopa_za32_m (0, p0, p1, z0, z1))
+
+/*
+** mopa_za32_u8_3_p0_p1_z0_z1:
+** umopa za3\.s, p0/m, p1/m, z0\.b, z1\.b
+** ret
+*/
+TEST_UNIFORM_ZA (mopa_za32_u8_3_p0_p1_z0_z1, svuint8_t,
+ svmopa_za32_u8_m (3, p0, p1, z0, z1),
+ svmopa_za32_m (3, p0, p1, z0, z1))
+
+/*
+** mopa_za32_bf16_0_p0_p1_z0_z1:
+** bfmopa za0\.s, p0/m, p1/m, z0\.h, z1\.h
+** ret
+*/
+TEST_UNIFORM_ZA (mopa_za32_bf16_0_p0_p1_z0_z1, svbfloat16_t,
+ svmopa_za32_bf16_m (0, p0, p1, z0, z1),
+ svmopa_za32_m (0, p0, p1, z0, z1))
+
+/*
+** mopa_za32_bf16_3_p0_p1_z0_z1:
+** bfmopa za3\.s, p0/m, p1/m, z0\.h, z1\.h
+** ret
+*/
+TEST_UNIFORM_ZA (mopa_za32_bf16_3_p0_p1_z0_z1, svbfloat16_t,
+ svmopa_za32_bf16_m (3, p0, p1, z0, z1),
+ svmopa_za32_m (3, p0, p1, z0, z1))
+
+/*
+** mopa_za32_f16_0_p0_p1_z0_z1:
+** fmopa za0\.s, p0/m, p1/m, z0\.h, z1\.h
+** ret
+*/
+TEST_UNIFORM_ZA (mopa_za32_f16_0_p0_p1_z0_z1, svfloat16_t,
+ svmopa_za32_f16_m (0, p0, p1, z0, z1),
+ svmopa_za32_m (0, p0, p1, z0, z1))
+
+/*
+** mopa_za32_f16_3_p0_p1_z0_z1:
+** fmopa za3\.s, p0/m, p1/m, z0\.h, z1\.h
+** ret
+*/
+TEST_UNIFORM_ZA (mopa_za32_f16_3_p0_p1_z0_z1, svfloat16_t,
+ svmopa_za32_f16_m (3, p0, p1, z0, z1),
+ svmopa_za32_m (3, p0, p1, z0, z1))
+
+/*
+** mopa_za32_f32_0_p0_p1_z0_z1:
+** fmopa za0\.s, p0/m, p1/m, z0\.s, z1\.s
+** ret
+*/
+TEST_UNIFORM_ZA (mopa_za32_f32_0_p0_p1_z0_z1, svfloat32_t,
+ svmopa_za32_f32_m (0, p0, p1, z0, z1),
+ svmopa_za32_m (0, p0, p1, z0, z1))
+
+/*
+** mopa_za32_f32_3_p0_p1_z0_z1:
+** fmopa za3\.s, p0/m, p1/m, z0\.s, z1\.s
+** ret
+*/
+TEST_UNIFORM_ZA (mopa_za32_f32_3_p0_p1_z0_z1, svfloat32_t,
+ svmopa_za32_f32_m (3, p0, p1, z0, z1),
+ svmopa_za32_m (3, p0, p1, z0, z1))
--- /dev/null
+/* { dg-final { check-function-bodies "**" "" "-DCHECK_ASM" } } */
+
+#include "test_sme_acle.h"
+
+#pragma GCC target "+sme-i16i64"
+
+/*
+** mopa_za64_s16_0_p0_p1_z0_z1:
+** smopa za0\.d, p0/m, p1/m, z0\.h, z1\.h
+** ret
+*/
+TEST_UNIFORM_ZA (mopa_za64_s16_0_p0_p1_z0_z1, svint16_t,
+ svmopa_za64_s16_m (0, p0, p1, z0, z1),
+ svmopa_za64_m (0, p0, p1, z0, z1))
+
+/*
+** mopa_za64_s16_0_p1_p0_z1_z0:
+** smopa za0\.d, p1/m, p0/m, z1\.h, z0\.h
+** ret
+*/
+TEST_UNIFORM_ZA (mopa_za64_s16_0_p1_p0_z1_z0, svint16_t,
+ svmopa_za64_s16_m (0, p1, p0, z1, z0),
+ svmopa_za64_m (0, p1, p0, z1, z0))
+
+/*
+** mopa_za64_s16_7_p0_p1_z0_z1:
+** smopa za7\.d, p0/m, p1/m, z0\.h, z1\.h
+** ret
+*/
+TEST_UNIFORM_ZA (mopa_za64_s16_7_p0_p1_z0_z1, svint16_t,
+ svmopa_za64_s16_m (7, p0, p1, z0, z1),
+ svmopa_za64_m (7, p0, p1, z0, z1))
+
+/*
+** mopa_za64_u16_0_p0_p1_z0_z1:
+** umopa za0\.d, p0/m, p1/m, z0\.h, z1\.h
+** ret
+*/
+TEST_UNIFORM_ZA (mopa_za64_u16_0_p0_p1_z0_z1, svuint16_t,
+ svmopa_za64_u16_m (0, p0, p1, z0, z1),
+ svmopa_za64_m (0, p0, p1, z0, z1))
+
+/*
+** mopa_za64_u16_7_p0_p1_z0_z1:
+** umopa za7\.d, p0/m, p1/m, z0\.h, z1\.h
+** ret
+*/
+TEST_UNIFORM_ZA (mopa_za64_u16_7_p0_p1_z0_z1, svuint16_t,
+ svmopa_za64_u16_m (7, p0, p1, z0, z1),
+ svmopa_za64_m (7, p0, p1, z0, z1))
+
+#pragma GCC target "+nosme-i16i64+sme-f64f64"
+
+/*
+** mopa_za64_f64_0_p0_p1_z0_z1:
+** fmopa za0\.d, p0/m, p1/m, z0\.d, z1\.d
+** ret
+*/
+TEST_UNIFORM_ZA (mopa_za64_f64_0_p0_p1_z0_z1, svfloat64_t,
+ svmopa_za64_f64_m (0, p0, p1, z0, z1),
+ svmopa_za64_m (0, p0, p1, z0, z1))
+
+/*
+** mopa_za64_f64_7_p0_p1_z0_z1:
+** fmopa za7\.d, p0/m, p1/m, z0\.d, z1\.d
+** ret
+*/
+TEST_UNIFORM_ZA (mopa_za64_f64_7_p0_p1_z0_z1, svfloat64_t,
+ svmopa_za64_f64_m (7, p0, p1, z0, z1),
+ svmopa_za64_m (7, p0, p1, z0, z1))
--- /dev/null
+/* { dg-final { check-function-bodies "**" "" "-DCHECK_ASM" } } */
+
+#include "test_sme_acle.h"
+
+/*
+** mops_za32_s8_0_p0_p1_z0_z1:
+** smops za0\.s, p0/m, p1/m, z0\.b, z1\.b
+** ret
+*/
+TEST_UNIFORM_ZA (mops_za32_s8_0_p0_p1_z0_z1, svint8_t,
+ svmops_za32_s8_m (0, p0, p1, z0, z1),
+ svmops_za32_m (0, p0, p1, z0, z1))
+
+/*
+** mops_za32_s8_0_p1_p0_z1_z0:
+** smops za0\.s, p1/m, p0/m, z1\.b, z0\.b
+** ret
+*/
+TEST_UNIFORM_ZA (mops_za32_s8_0_p1_p0_z1_z0, svint8_t,
+ svmops_za32_s8_m (0, p1, p0, z1, z0),
+ svmops_za32_m (0, p1, p0, z1, z0))
+
+/*
+** mops_za32_s8_3_p0_p1_z0_z1:
+** smops za3\.s, p0/m, p1/m, z0\.b, z1\.b
+** ret
+*/
+TEST_UNIFORM_ZA (mops_za32_s8_3_p0_p1_z0_z1, svint8_t,
+ svmops_za32_s8_m (3, p0, p1, z0, z1),
+ svmops_za32_m (3, p0, p1, z0, z1))
+
+/*
+** mops_za32_u8_0_p0_p1_z0_z1:
+** umops za0\.s, p0/m, p1/m, z0\.b, z1\.b
+** ret
+*/
+TEST_UNIFORM_ZA (mops_za32_u8_0_p0_p1_z0_z1, svuint8_t,
+ svmops_za32_u8_m (0, p0, p1, z0, z1),
+ svmops_za32_m (0, p0, p1, z0, z1))
+
+/*
+** mops_za32_u8_3_p0_p1_z0_z1:
+** umops za3\.s, p0/m, p1/m, z0\.b, z1\.b
+** ret
+*/
+TEST_UNIFORM_ZA (mops_za32_u8_3_p0_p1_z0_z1, svuint8_t,
+ svmops_za32_u8_m (3, p0, p1, z0, z1),
+ svmops_za32_m (3, p0, p1, z0, z1))
+
+/*
+** mops_za32_bf16_0_p0_p1_z0_z1:
+** bfmops za0\.s, p0/m, p1/m, z0\.h, z1\.h
+** ret
+*/
+TEST_UNIFORM_ZA (mops_za32_bf16_0_p0_p1_z0_z1, svbfloat16_t,
+ svmops_za32_bf16_m (0, p0, p1, z0, z1),
+ svmops_za32_m (0, p0, p1, z0, z1))
+
+/*
+** mops_za32_bf16_3_p0_p1_z0_z1:
+** bfmops za3\.s, p0/m, p1/m, z0\.h, z1\.h
+** ret
+*/
+TEST_UNIFORM_ZA (mops_za32_bf16_3_p0_p1_z0_z1, svbfloat16_t,
+ svmops_za32_bf16_m (3, p0, p1, z0, z1),
+ svmops_za32_m (3, p0, p1, z0, z1))
+
+/*
+** mops_za32_f16_0_p0_p1_z0_z1:
+** fmops za0\.s, p0/m, p1/m, z0\.h, z1\.h
+** ret
+*/
+TEST_UNIFORM_ZA (mops_za32_f16_0_p0_p1_z0_z1, svfloat16_t,
+ svmops_za32_f16_m (0, p0, p1, z0, z1),
+ svmops_za32_m (0, p0, p1, z0, z1))
+
+/*
+** mops_za32_f16_3_p0_p1_z0_z1:
+** fmops za3\.s, p0/m, p1/m, z0\.h, z1\.h
+** ret
+*/
+TEST_UNIFORM_ZA (mops_za32_f16_3_p0_p1_z0_z1, svfloat16_t,
+ svmops_za32_f16_m (3, p0, p1, z0, z1),
+ svmops_za32_m (3, p0, p1, z0, z1))
+
+/*
+** mops_za32_f32_0_p0_p1_z0_z1:
+** fmops za0\.s, p0/m, p1/m, z0\.s, z1\.s
+** ret
+*/
+TEST_UNIFORM_ZA (mops_za32_f32_0_p0_p1_z0_z1, svfloat32_t,
+ svmops_za32_f32_m (0, p0, p1, z0, z1),
+ svmops_za32_m (0, p0, p1, z0, z1))
+
+/*
+** mops_za32_f32_3_p0_p1_z0_z1:
+** fmops za3\.s, p0/m, p1/m, z0\.s, z1\.s
+** ret
+*/
+TEST_UNIFORM_ZA (mops_za32_f32_3_p0_p1_z0_z1, svfloat32_t,
+ svmops_za32_f32_m (3, p0, p1, z0, z1),
+ svmops_za32_m (3, p0, p1, z0, z1))
--- /dev/null
+/* { dg-final { check-function-bodies "**" "" "-DCHECK_ASM" } } */
+
+#include "test_sme_acle.h"
+
+#pragma GCC target "+sme-i16i64"
+
+/*
+** mops_za64_s16_0_p0_p1_z0_z1:
+** smops za0\.d, p0/m, p1/m, z0\.h, z1\.h
+** ret
+*/
+TEST_UNIFORM_ZA (mops_za64_s16_0_p0_p1_z0_z1, svint16_t,
+ svmops_za64_s16_m (0, p0, p1, z0, z1),
+ svmops_za64_m (0, p0, p1, z0, z1))
+
+/*
+** mops_za64_s16_0_p1_p0_z1_z0:
+** smops za0\.d, p1/m, p0/m, z1\.h, z0\.h
+** ret
+*/
+TEST_UNIFORM_ZA (mops_za64_s16_0_p1_p0_z1_z0, svint16_t,
+ svmops_za64_s16_m (0, p1, p0, z1, z0),
+ svmops_za64_m (0, p1, p0, z1, z0))
+
+/*
+** mops_za64_s16_7_p0_p1_z0_z1:
+** smops za7\.d, p0/m, p1/m, z0\.h, z1\.h
+** ret
+*/
+TEST_UNIFORM_ZA (mops_za64_s16_7_p0_p1_z0_z1, svint16_t,
+ svmops_za64_s16_m (7, p0, p1, z0, z1),
+ svmops_za64_m (7, p0, p1, z0, z1))
+
+/*
+** mops_za64_u16_0_p0_p1_z0_z1:
+** umops za0\.d, p0/m, p1/m, z0\.h, z1\.h
+** ret
+*/
+TEST_UNIFORM_ZA (mops_za64_u16_0_p0_p1_z0_z1, svuint16_t,
+ svmops_za64_u16_m (0, p0, p1, z0, z1),
+ svmops_za64_m (0, p0, p1, z0, z1))
+
+/*
+** mops_za64_u16_7_p0_p1_z0_z1:
+** umops za7\.d, p0/m, p1/m, z0\.h, z1\.h
+** ret
+*/
+TEST_UNIFORM_ZA (mops_za64_u16_7_p0_p1_z0_z1, svuint16_t,
+ svmops_za64_u16_m (7, p0, p1, z0, z1),
+ svmops_za64_m (7, p0, p1, z0, z1))
+
+#pragma GCC target "+nosme-i16i64+sme-f64f64"
+
+/*
+** mops_za64_f64_0_p0_p1_z0_z1:
+** fmops za0\.d, p0/m, p1/m, z0\.d, z1\.d
+** ret
+*/
+TEST_UNIFORM_ZA (mops_za64_f64_0_p0_p1_z0_z1, svfloat64_t,
+ svmops_za64_f64_m (0, p0, p1, z0, z1),
+ svmops_za64_m (0, p0, p1, z0, z1))
+
+/*
+** mops_za64_f64_7_p0_p1_z0_z1:
+** fmops za7\.d, p0/m, p1/m, z0\.d, z1\.d
+** ret
+*/
+TEST_UNIFORM_ZA (mops_za64_f64_7_p0_p1_z0_z1, svfloat64_t,
+ svmops_za64_f64_m (7, p0, p1, z0, z1),
+ svmops_za64_m (7, p0, p1, z0, z1))
--- /dev/null
+/* { dg-final { check-function-bodies "**" "" "-DCHECK_ASM" } } */
+
+#include "test_sme_acle.h"
+
+/*
+** read_za128_s8_0_0_tied:
+** mov (w1[2-5]), (?:wzr|#?0)
+** mova z0\.q, p0/m, za0h\.q\[\1, 0\]
+** ret
+*/
+TEST_READ_ZA (read_za128_s8_0_0_tied, svint8_t,
+ z0 = svread_hor_za128_s8_m (z0, p0, 0, 0),
+ z0 = svread_hor_za128_m (z0, p0, 0, 0))
+
+/*
+** read_za128_s8_0_1_tied:
+** mov (w1[2-5]), #?1
+** mova z0\.q, p0/m, za0h\.q\[\1, 0\]
+** ret
+*/
+TEST_READ_ZA (read_za128_s8_0_1_tied, svint8_t,
+ z0 = svread_hor_za128_s8_m (z0, p0, 0, 1),
+ z0 = svread_hor_za128_m (z0, p0, 0, 1))
+
+/*
+** read_za128_s8_0_w0_tied:
+** mov (w1[2-5]), w0
+** mova z0\.q, p0/m, za0h\.q\[\1, 0\]
+** ret
+*/
+TEST_READ_ZA (read_za128_s8_0_w0_tied, svint8_t,
+ z0 = svread_hor_za128_s8_m (z0, p0, 0, w0),
+ z0 = svread_hor_za128_m (z0, p0, 0, w0))
+
+/*
+** read_za128_s8_0_w0p1_tied:
+** add (w1[2-5]), w0, #?1
+** mova z0\.q, p0/m, za0h\.q\[\1, 0\]
+** ret
+*/
+TEST_READ_ZA (read_za128_s8_0_w0p1_tied, svint8_t,
+ z0 = svread_hor_za128_s8_m (z0, p0, 0, w0 + 1),
+ z0 = svread_hor_za128_m (z0, p0, 0, w0 + 1))
+
+/*
+** read_za128_s8_0_w0m1_tied:
+** sub (w1[2-5]), w0, #?1
+** mova z0\.q, p0/m, za0h\.q\[\1, 0\]
+** ret
+*/
+TEST_READ_ZA (read_za128_s8_0_w0m1_tied, svint8_t,
+ z0 = svread_hor_za128_s8_m (z0, p0, 0, w0 - 1),
+ z0 = svread_hor_za128_m (z0, p0, 0, w0 - 1))
+
+/*
+** read_za128_s8_1_w0_tied:
+** mov (w1[2-5]), w0
+** mova z0\.q, p0/m, za1h\.q\[\1, 0\]
+** ret
+*/
+TEST_READ_ZA (read_za128_s8_1_w0_tied, svint8_t,
+ z0 = svread_hor_za128_s8_m (z0, p0, 1, w0),
+ z0 = svread_hor_za128_m (z0, p0, 1, w0))
+
+/*
+** read_za128_s8_15_w0_tied:
+** mov (w1[2-5]), w0
+** mova z0\.q, p0/m, za15h\.q\[\1, 0\]
+** ret
+*/
+TEST_READ_ZA (read_za128_s8_15_w0_tied, svint8_t,
+ z0 = svread_hor_za128_s8_m (z0, p0, 15, w0),
+ z0 = svread_hor_za128_m (z0, p0, 15, w0))
+
+/*
+** read_za128_s8_0_w0_untied:
+** (
+** mov (w1[2-5]), w0
+** mov z0\.d, z1\.d
+** mova z0\.q, p0/m, za0h\.q\[\1, 0\]
+** |
+** mov z0\.d, z1\.d
+** mov (w1[2-5]), w0
+** mova z0\.q, p0/m, za0h\.q\[\2, 0\]
+** |
+** mov (w1[2-5]), w0
+** mova z1\.q, p0/m, za0h\.q\[\3, 0\]
+** mov z0\.d, z1\.d
+** )
+** ret
+*/
+TEST_READ_ZA (read_za128_s8_0_w0_untied, svint8_t,
+ z0 = svread_hor_za128_s8_m (z1, p0, 0, w0),
+ z0 = svread_hor_za128_m (z1, p0, 0, w0))
+
+/*
+** read_za128_u8_0_w0_tied:
+** mov (w1[2-5]), w0
+** mova z0\.q, p0/m, za0h\.q\[\1, 0\]
+** ret
+*/
+TEST_READ_ZA (read_za128_u8_0_w0_tied, svuint8_t,
+ z0 = svread_hor_za128_u8_m (z0, p0, 0, w0),
+ z0 = svread_hor_za128_m (z0, p0, 0, w0))
+
+/*
+** read_za128_u8_0_w0_untied:
+** (
+** mov (w1[2-5]), w0
+** mov z0\.d, z1\.d
+** mova z0\.q, p0/m, za0h\.q\[\1, 0\]
+** |
+** mov z0\.d, z1\.d
+** mov (w1[2-5]), w0
+** mova z0\.q, p0/m, za0h\.q\[\2, 0\]
+** |
+** mov (w1[2-5]), w0
+** mova z1\.q, p0/m, za0h\.q\[\3, 0\]
+** mov z0\.d, z1\.d
+** )
+** ret
+*/
+TEST_READ_ZA (read_za128_u8_0_w0_untied, svuint8_t,
+ z0 = svread_hor_za128_u8_m (z1, p0, 0, w0),
+ z0 = svread_hor_za128_m (z1, p0, 0, w0))
+
+/*
+** read_za128_s16_0_w0_tied:
+** mov (w1[2-5]), w0
+** mova z0\.q, p0/m, za0h\.q\[\1, 0\]
+** ret
+*/
+TEST_READ_ZA (read_za128_s16_0_w0_tied, svint16_t,
+ z0 = svread_hor_za128_s16_m (z0, p0, 0, w0),
+ z0 = svread_hor_za128_m (z0, p0, 0, w0))
+
+/*
+** read_za128_s16_0_w0_untied:
+** (
+** mov (w1[2-5]), w0
+** mov z0\.d, z1\.d
+** mova z0\.q, p0/m, za0h\.q\[\1, 0\]
+** |
+** mov z0\.d, z1\.d
+** mov (w1[2-5]), w0
+** mova z0\.q, p0/m, za0h\.q\[\2, 0\]
+** |
+** mov (w1[2-5]), w0
+** mova z1\.q, p0/m, za0h\.q\[\3, 0\]
+** mov z0\.d, z1\.d
+** )
+** ret
+*/
+TEST_READ_ZA (read_za128_s16_0_w0_untied, svint16_t,
+ z0 = svread_hor_za128_s16_m (z1, p0, 0, w0),
+ z0 = svread_hor_za128_m (z1, p0, 0, w0))
+
+/*
+** read_za128_u16_0_w0_tied:
+** mov (w1[2-5]), w0
+** mova z0\.q, p0/m, za0h\.q\[\1, 0\]
+** ret
+*/
+TEST_READ_ZA (read_za128_u16_0_w0_tied, svuint16_t,
+ z0 = svread_hor_za128_u16_m (z0, p0, 0, w0),
+ z0 = svread_hor_za128_m (z0, p0, 0, w0))
+
+/*
+** read_za128_u16_0_w0_untied:
+** (
+** mov (w1[2-5]), w0
+** mov z0\.d, z1\.d
+** mova z0\.q, p0/m, za0h\.q\[\1, 0\]
+** |
+** mov z0\.d, z1\.d
+** mov (w1[2-5]), w0
+** mova z0\.q, p0/m, za0h\.q\[\2, 0\]
+** |
+** mov (w1[2-5]), w0
+** mova z1\.q, p0/m, za0h\.q\[\3, 0\]
+** mov z0\.d, z1\.d
+** )
+** ret
+*/
+TEST_READ_ZA (read_za128_u16_0_w0_untied, svuint16_t,
+ z0 = svread_hor_za128_u16_m (z1, p0, 0, w0),
+ z0 = svread_hor_za128_m (z1, p0, 0, w0))
+
+/*
+** read_za128_f16_0_w0_tied:
+** mov (w1[2-5]), w0
+** mova z0\.q, p0/m, za0h\.q\[\1, 0\]
+** ret
+*/
+TEST_READ_ZA (read_za128_f16_0_w0_tied, svfloat16_t,
+ z0 = svread_hor_za128_f16_m (z0, p0, 0, w0),
+ z0 = svread_hor_za128_m (z0, p0, 0, w0))
+
+/*
+** read_za128_f16_0_w0_untied:
+** (
+** mov (w1[2-5]), w0
+** mov z0\.d, z1\.d
+** mova z0\.q, p0/m, za0h\.q\[\1, 0\]
+** |
+** mov z0\.d, z1\.d
+** mov (w1[2-5]), w0
+** mova z0\.q, p0/m, za0h\.q\[\2, 0\]
+** |
+** mov (w1[2-5]), w0
+** mova z1\.q, p0/m, za0h\.q\[\3, 0\]
+** mov z0\.d, z1\.d
+** )
+** ret
+*/
+TEST_READ_ZA (read_za128_f16_0_w0_untied, svfloat16_t,
+ z0 = svread_hor_za128_f16_m (z1, p0, 0, w0),
+ z0 = svread_hor_za128_m (z1, p0, 0, w0))
+
+/*
+** read_za128_bf16_0_w0_tied:
+** mov (w1[2-5]), w0
+** mova z0\.q, p0/m, za0h\.q\[\1, 0\]
+** ret
+*/
+TEST_READ_ZA (read_za128_bf16_0_w0_tied, svbfloat16_t,
+ z0 = svread_hor_za128_bf16_m (z0, p0, 0, w0),
+ z0 = svread_hor_za128_m (z0, p0, 0, w0))
+
+/*
+** read_za128_bf16_0_w0_untied:
+** (
+** mov (w1[2-5]), w0
+** mov z0\.d, z1\.d
+** mova z0\.q, p0/m, za0h\.q\[\1, 0\]
+** |
+** mov z0\.d, z1\.d
+** mov (w1[2-5]), w0
+** mova z0\.q, p0/m, za0h\.q\[\2, 0\]
+** |
+** mov (w1[2-5]), w0
+** mova z1\.q, p0/m, za0h\.q\[\3, 0\]
+** mov z0\.d, z1\.d
+** )
+** ret
+*/
+TEST_READ_ZA (read_za128_bf16_0_w0_untied, svbfloat16_t,
+ z0 = svread_hor_za128_bf16_m (z1, p0, 0, w0),
+ z0 = svread_hor_za128_m (z1, p0, 0, w0))
+
+/*
+** read_za128_s32_0_w0_tied:
+** mov (w1[2-5]), w0
+** mova z0\.q, p0/m, za0h\.q\[\1, 0\]
+** ret
+*/
+TEST_READ_ZA (read_za128_s32_0_w0_tied, svint32_t,
+ z0 = svread_hor_za128_s32_m (z0, p0, 0, w0),
+ z0 = svread_hor_za128_m (z0, p0, 0, w0))
+
+/*
+** read_za128_s32_0_w0_untied:
+** (
+** mov (w1[2-5]), w0
+** mov z0\.d, z1\.d
+** mova z0\.q, p0/m, za0h\.q\[\1, 0\]
+** |
+** mov z0\.d, z1\.d
+** mov (w1[2-5]), w0
+** mova z0\.q, p0/m, za0h\.q\[\2, 0\]
+** |
+** mov (w1[2-5]), w0
+** mova z1\.q, p0/m, za0h\.q\[\3, 0\]
+** mov z0\.d, z1\.d
+** )
+** ret
+*/
+TEST_READ_ZA (read_za128_s32_0_w0_untied, svint32_t,
+ z0 = svread_hor_za128_s32_m (z1, p0, 0, w0),
+ z0 = svread_hor_za128_m (z1, p0, 0, w0))
+
+/*
+** read_za128_u32_0_w0_tied:
+** mov (w1[2-5]), w0
+** mova z0\.q, p0/m, za0h\.q\[\1, 0\]
+** ret
+*/
+TEST_READ_ZA (read_za128_u32_0_w0_tied, svuint32_t,
+ z0 = svread_hor_za128_u32_m (z0, p0, 0, w0),
+ z0 = svread_hor_za128_m (z0, p0, 0, w0))
+
+/*
+** read_za128_u32_0_w0_untied:
+** (
+** mov (w1[2-5]), w0
+** mov z0\.d, z1\.d
+** mova z0\.q, p0/m, za0h\.q\[\1, 0\]
+** |
+** mov z0\.d, z1\.d
+** mov (w1[2-5]), w0
+** mova z0\.q, p0/m, za0h\.q\[\2, 0\]
+** |
+** mov (w1[2-5]), w0
+** mova z1\.q, p0/m, za0h\.q\[\3, 0\]
+** mov z0\.d, z1\.d
+** )
+** ret
+*/
+TEST_READ_ZA (read_za128_u32_0_w0_untied, svuint32_t,
+ z0 = svread_hor_za128_u32_m (z1, p0, 0, w0),
+ z0 = svread_hor_za128_m (z1, p0, 0, w0))
+
+/*
+** read_za128_f32_0_w0_tied:
+** mov (w1[2-5]), w0
+** mova z0\.q, p0/m, za0h\.q\[\1, 0\]
+** ret
+*/
+TEST_READ_ZA (read_za128_f32_0_w0_tied, svfloat32_t,
+ z0 = svread_hor_za128_f32_m (z0, p0, 0, w0),
+ z0 = svread_hor_za128_m (z0, p0, 0, w0))
+
+/*
+** read_za128_f32_0_w0_untied:
+** (
+** mov (w1[2-5]), w0
+** mov z0\.d, z1\.d
+** mova z0\.q, p0/m, za0h\.q\[\1, 0\]
+** |
+** mov z0\.d, z1\.d
+** mov (w1[2-5]), w0
+** mova z0\.q, p0/m, za0h\.q\[\2, 0\]
+** |
+** mov (w1[2-5]), w0
+** mova z1\.q, p0/m, za0h\.q\[\3, 0\]
+** mov z0\.d, z1\.d
+** )
+** ret
+*/
+TEST_READ_ZA (read_za128_f32_0_w0_untied, svfloat32_t,
+ z0 = svread_hor_za128_f32_m (z1, p0, 0, w0),
+ z0 = svread_hor_za128_m (z1, p0, 0, w0))
+
+/*
+** read_za128_s64_0_w0_tied:
+** mov (w1[2-5]), w0
+** mova z0\.q, p0/m, za0h\.q\[\1, 0\]
+** ret
+*/
+TEST_READ_ZA (read_za128_s64_0_w0_tied, svint64_t,
+ z0 = svread_hor_za128_s64_m (z0, p0, 0, w0),
+ z0 = svread_hor_za128_m (z0, p0, 0, w0))
+
+/*
+** read_za128_s64_0_w0_untied:
+** (
+** mov (w1[2-5]), w0
+** mov z0\.d, z1\.d
+** mova z0\.q, p0/m, za0h\.q\[\1, 0\]
+** |
+** mov z0\.d, z1\.d
+** mov (w1[2-5]), w0
+** mova z0\.q, p0/m, za0h\.q\[\2, 0\]
+** |
+** mov (w1[2-5]), w0
+** mova z1\.q, p0/m, za0h\.q\[\3, 0\]
+** mov z0\.d, z1\.d
+** )
+** ret
+*/
+TEST_READ_ZA (read_za128_s64_0_w0_untied, svint64_t,
+ z0 = svread_hor_za128_s64_m (z1, p0, 0, w0),
+ z0 = svread_hor_za128_m (z1, p0, 0, w0))
+
+/*
+** read_za128_u64_0_w0_tied:
+** mov (w1[2-5]), w0
+** mova z0\.q, p0/m, za0h\.q\[\1, 0\]
+** ret
+*/
+TEST_READ_ZA (read_za128_u64_0_w0_tied, svuint64_t,
+ z0 = svread_hor_za128_u64_m (z0, p0, 0, w0),
+ z0 = svread_hor_za128_m (z0, p0, 0, w0))
+
+/*
+** read_za128_u64_0_w0_untied:
+** (
+** mov (w1[2-5]), w0
+** mov z0\.d, z1\.d
+** mova z0\.q, p0/m, za0h\.q\[\1, 0\]
+** |
+** mov z0\.d, z1\.d
+** mov (w1[2-5]), w0
+** mova z0\.q, p0/m, za0h\.q\[\2, 0\]
+** |
+** mov (w1[2-5]), w0
+** mova z1\.q, p0/m, za0h\.q\[\3, 0\]
+** mov z0\.d, z1\.d
+** )
+** ret
+*/
+TEST_READ_ZA (read_za128_u64_0_w0_untied, svuint64_t,
+ z0 = svread_hor_za128_u64_m (z1, p0, 0, w0),
+ z0 = svread_hor_za128_m (z1, p0, 0, w0))
+
+/*
+** read_za128_f64_0_w0_tied:
+** mov (w1[2-5]), w0
+** mova z0\.q, p0/m, za0h\.q\[\1, 0\]
+** ret
+*/
+TEST_READ_ZA (read_za128_f64_0_w0_tied, svfloat64_t,
+ z0 = svread_hor_za128_f64_m (z0, p0, 0, w0),
+ z0 = svread_hor_za128_m (z0, p0, 0, w0))
+
+/*
+** read_za128_f64_0_w0_untied:
+** (
+** mov (w1[2-5]), w0
+** mov z0\.d, z1\.d
+** mova z0\.q, p0/m, za0h\.q\[\1, 0\]
+** |
+** mov z0\.d, z1\.d
+** mov (w1[2-5]), w0
+** mova z0\.q, p0/m, za0h\.q\[\2, 0\]
+** |
+** mov (w1[2-5]), w0
+** mova z1\.q, p0/m, za0h\.q\[\3, 0\]
+** mov z0\.d, z1\.d
+** )
+** ret
+*/
+TEST_READ_ZA (read_za128_f64_0_w0_untied, svfloat64_t,
+ z0 = svread_hor_za128_f64_m (z1, p0, 0, w0),
+ z0 = svread_hor_za128_m (z1, p0, 0, w0))
--- /dev/null
+/* { dg-final { check-function-bodies "**" "" "-DCHECK_ASM" } } */
+
+#include "test_sme_acle.h"
+
+/*
+** read_za16_s16_0_0_tied:
+** mov (w1[2-5]), (?:wzr|#?0)
+** mova z0\.h, p0/m, za0h\.h\[\1, 0\]
+** ret
+*/
+TEST_READ_ZA (read_za16_s16_0_0_tied, svint16_t,
+ z0 = svread_hor_za16_s16_m (z0, p0, 0, 0),
+ z0 = svread_hor_za16_m (z0, p0, 0, 0))
+
+/*
+** read_za16_s16_0_1_tied:
+** mov (w1[2-5]), #?1
+** mova z0\.h, p0/m, za0h\.h\[\1, 0\]
+** ret
+*/
+TEST_READ_ZA (read_za16_s16_0_1_tied, svint16_t,
+ z0 = svread_hor_za16_s16_m (z0, p0, 0, 1),
+ z0 = svread_hor_za16_m (z0, p0, 0, 1))
+
+/*
+** read_za16_s16_0_w0_tied:
+** mov (w1[2-5]), w0
+** mova z0\.h, p0/m, za0h\.h\[\1, 0\]
+** ret
+*/
+TEST_READ_ZA (read_za16_s16_0_w0_tied, svint16_t,
+ z0 = svread_hor_za16_s16_m (z0, p0, 0, w0),
+ z0 = svread_hor_za16_m (z0, p0, 0, w0))
+
+/*
+** read_za16_s16_0_w0p1_tied:
+** mov (w1[2-5]), w0
+** mova z0\.h, p0/m, za0h\.h\[\1, 1\]
+** ret
+*/
+TEST_READ_ZA (read_za16_s16_0_w0p1_tied, svint16_t,
+ z0 = svread_hor_za16_s16_m (z0, p0, 0, w0 + 1),
+ z0 = svread_hor_za16_m (z0, p0, 0, w0 + 1))
+
+/*
+** read_za16_s16_0_w0p7_tied:
+** mov (w1[2-5]), w0
+** mova z0\.h, p0/m, za0h\.h\[\1, 7\]
+** ret
+*/
+TEST_READ_ZA (read_za16_s16_0_w0p7_tied, svint16_t,
+ z0 = svread_hor_za16_s16_m (z0, p0, 0, w0 + 7),
+ z0 = svread_hor_za16_m (z0, p0, 0, w0 + 7))
+
+/*
+** read_za16_s16_0_w0p8_tied:
+** add (w1[2-5]), w0, #?8
+** mova z0\.h, p0/m, za0h\.h\[\1, 0\]
+** ret
+*/
+TEST_READ_ZA (read_za16_s16_0_w0p8_tied, svint16_t,
+ z0 = svread_hor_za16_s16_m (z0, p0, 0, w0 + 8),
+ z0 = svread_hor_za16_m (z0, p0, 0, w0 + 8))
+
+/*
+** read_za16_s16_0_w0m1_tied:
+** sub (w1[2-5]), w0, #?1
+** mova z0\.h, p0/m, za0h\.h\[\1, 0\]
+** ret
+*/
+TEST_READ_ZA (read_za16_s16_0_w0m1_tied, svint16_t,
+ z0 = svread_hor_za16_s16_m (z0, p0, 0, w0 - 1),
+ z0 = svread_hor_za16_m (z0, p0, 0, w0 - 1))
+
+/*
+** read_za16_s16_1_w0_tied:
+** mov (w1[2-5]), w0
+** mova z0\.h, p0/m, za1h\.h\[\1, 0\]
+** ret
+*/
+TEST_READ_ZA (read_za16_s16_1_w0_tied, svint16_t,
+ z0 = svread_hor_za16_s16_m (z0, p0, 1, w0),
+ z0 = svread_hor_za16_m (z0, p0, 1, w0))
+
+/*
+** read_za16_s16_1_w0p7_tied:
+** mov (w1[2-5]), w0
+** mova z0\.h, p0/m, za1h\.h\[\1, 7\]
+** ret
+*/
+TEST_READ_ZA (read_za16_s16_1_w0p7_tied, svint16_t,
+ z0 = svread_hor_za16_s16_m (z0, p0, 1, w0 + 7),
+ z0 = svread_hor_za16_m (z0, p0, 1, w0 + 7))
+
+/*
+** read_za16_s16_0_w0_untied:
+** (
+** mov (w1[2-5]), w0
+** mov z0\.d, z1\.d
+** mova z0\.h, p0/m, za0h\.h\[\1, 0\]
+** |
+** mov z0\.d, z1\.d
+** mov (w1[2-5]), w0
+** mova z0\.h, p0/m, za0h\.h\[\2, 0\]
+** |
+** mov (w1[2-5]), w0
+** mova z1\.h, p0/m, za0h\.h\[\3, 0\]
+** mov z0\.d, z1\.d
+** )
+** ret
+*/
+TEST_READ_ZA (read_za16_s16_0_w0_untied, svint16_t,
+ z0 = svread_hor_za16_s16_m (z1, p0, 0, w0),
+ z0 = svread_hor_za16_m (z1, p0, 0, w0))
+
+/*
+** read_za16_u16_0_w0_tied:
+** mov (w1[2-5]), w0
+** mova z0\.h, p0/m, za0h\.h\[\1, 0\]
+** ret
+*/
+TEST_READ_ZA (read_za16_u16_0_w0_tied, svuint16_t,
+ z0 = svread_hor_za16_u16_m (z0, p0, 0, w0),
+ z0 = svread_hor_za16_m (z0, p0, 0, w0))
+
+/*
+** read_za16_u16_0_w0_untied:
+** (
+** mov (w1[2-5]), w0
+** mov z0\.d, z1\.d
+** mova z0\.h, p0/m, za0h\.h\[\1, 0\]
+** |
+** mov z0\.d, z1\.d
+** mov (w1[2-5]), w0
+** mova z0\.h, p0/m, za0h\.h\[\2, 0\]
+** |
+** mov (w1[2-5]), w0
+** mova z1\.h, p0/m, za0h\.h\[\3, 0\]
+** mov z0\.d, z1\.d
+** )
+** ret
+*/
+TEST_READ_ZA (read_za16_u16_0_w0_untied, svuint16_t,
+ z0 = svread_hor_za16_u16_m (z1, p0, 0, w0),
+ z0 = svread_hor_za16_m (z1, p0, 0, w0))
+
+/*
+** read_za16_f16_0_w0_tied:
+** mov (w1[2-5]), w0
+** mova z0\.h, p0/m, za0h\.h\[\1, 0\]
+** ret
+*/
+TEST_READ_ZA (read_za16_f16_0_w0_tied, svfloat16_t,
+ z0 = svread_hor_za16_f16_m (z0, p0, 0, w0),
+ z0 = svread_hor_za16_m (z0, p0, 0, w0))
+
+/*
+** read_za16_f16_0_w0_untied:
+** (
+** mov (w1[2-5]), w0
+** mov z0\.d, z1\.d
+** mova z0\.h, p0/m, za0h\.h\[\1, 0\]
+** |
+** mov z0\.d, z1\.d
+** mov (w1[2-5]), w0
+** mova z0\.h, p0/m, za0h\.h\[\2, 0\]
+** |
+** mov (w1[2-5]), w0
+** mova z1\.h, p0/m, za0h\.h\[\3, 0\]
+** mov z0\.d, z1\.d
+** )
+** ret
+*/
+TEST_READ_ZA (read_za16_f16_0_w0_untied, svfloat16_t,
+ z0 = svread_hor_za16_f16_m (z1, p0, 0, w0),
+ z0 = svread_hor_za16_m (z1, p0, 0, w0))
+
+/*
+** read_za16_bf16_0_w0_tied:
+** mov (w1[2-5]), w0
+** mova z0\.h, p0/m, za0h\.h\[\1, 0\]
+** ret
+*/
+TEST_READ_ZA (read_za16_bf16_0_w0_tied, svbfloat16_t,
+ z0 = svread_hor_za16_bf16_m (z0, p0, 0, w0),
+ z0 = svread_hor_za16_m (z0, p0, 0, w0))
+
+/*
+** read_za16_bf16_0_w0_untied:
+** (
+** mov (w1[2-5]), w0
+** mov z0\.d, z1\.d
+** mova z0\.h, p0/m, za0h\.h\[\1, 0\]
+** |
+** mov z0\.d, z1\.d
+** mov (w1[2-5]), w0
+** mova z0\.h, p0/m, za0h\.h\[\2, 0\]
+** |
+** mov (w1[2-5]), w0
+** mova z1\.h, p0/m, za0h\.h\[\3, 0\]
+** mov z0\.d, z1\.d
+** )
+** ret
+*/
+TEST_READ_ZA (read_za16_bf16_0_w0_untied, svbfloat16_t,
+ z0 = svread_hor_za16_bf16_m (z1, p0, 0, w0),
+ z0 = svread_hor_za16_m (z1, p0, 0, w0))
--- /dev/null
+/* { dg-final { check-function-bodies "**" "" "-DCHECK_ASM" } } */
+
+#include "test_sme_acle.h"
+
+/*
+** read_za32_s32_0_0_tied:
+** mov (w1[2-5]), (?:wzr|#?0)
+** mova z0\.s, p0/m, za0h\.s\[\1, 0\]
+** ret
+*/
+TEST_READ_ZA (read_za32_s32_0_0_tied, svint32_t,
+ z0 = svread_hor_za32_s32_m (z0, p0, 0, 0),
+ z0 = svread_hor_za32_m (z0, p0, 0, 0))
+
+/*
+** read_za32_s32_0_1_tied:
+** mov (w1[2-5]), #?1
+** mova z0\.s, p0/m, za0h\.s\[\1, 0\]
+** ret
+*/
+TEST_READ_ZA (read_za32_s32_0_1_tied, svint32_t,
+ z0 = svread_hor_za32_s32_m (z0, p0, 0, 1),
+ z0 = svread_hor_za32_m (z0, p0, 0, 1))
+
+/*
+** read_za32_s32_0_w0_tied:
+** mov (w1[2-5]), w0
+** mova z0\.s, p0/m, za0h\.s\[\1, 0\]
+** ret
+*/
+TEST_READ_ZA (read_za32_s32_0_w0_tied, svint32_t,
+ z0 = svread_hor_za32_s32_m (z0, p0, 0, w0),
+ z0 = svread_hor_za32_m (z0, p0, 0, w0))
+
+/*
+** read_za32_s32_0_w0p1_tied:
+** mov (w1[2-5]), w0
+** mova z0\.s, p0/m, za0h\.s\[\1, 1\]
+** ret
+*/
+TEST_READ_ZA (read_za32_s32_0_w0p1_tied, svint32_t,
+ z0 = svread_hor_za32_s32_m (z0, p0, 0, w0 + 1),
+ z0 = svread_hor_za32_m (z0, p0, 0, w0 + 1))
+
+/*
+** read_za32_s32_0_w0p3_tied:
+** mov (w1[2-5]), w0
+** mova z0\.s, p0/m, za0h\.s\[\1, 3\]
+** ret
+*/
+TEST_READ_ZA (read_za32_s32_0_w0p3_tied, svint32_t,
+ z0 = svread_hor_za32_s32_m (z0, p0, 0, w0 + 3),
+ z0 = svread_hor_za32_m (z0, p0, 0, w0 + 3))
+
+/*
+** read_za32_s32_0_w0p4_tied:
+** add (w1[2-5]), w0, #?4
+** mova z0\.s, p0/m, za0h\.s\[\1, 0\]
+** ret
+*/
+TEST_READ_ZA (read_za32_s32_0_w0p4_tied, svint32_t,
+ z0 = svread_hor_za32_s32_m (z0, p0, 0, w0 + 4),
+ z0 = svread_hor_za32_m (z0, p0, 0, w0 + 4))
+
+/*
+** read_za32_s32_0_w0m1_tied:
+** sub (w1[2-5]), w0, #?1
+** mova z0\.s, p0/m, za0h\.s\[\1, 0\]
+** ret
+*/
+TEST_READ_ZA (read_za32_s32_0_w0m1_tied, svint32_t,
+ z0 = svread_hor_za32_s32_m (z0, p0, 0, w0 - 1),
+ z0 = svread_hor_za32_m (z0, p0, 0, w0 - 1))
+
+/*
+** read_za32_s32_1_w0_tied:
+** mov (w1[2-5]), w0
+** mova z0\.s, p0/m, za1h\.s\[\1, 0\]
+** ret
+*/
+TEST_READ_ZA (read_za32_s32_1_w0_tied, svint32_t,
+ z0 = svread_hor_za32_s32_m (z0, p0, 1, w0),
+ z0 = svread_hor_za32_m (z0, p0, 1, w0))
+
+/*
+** read_za32_s32_1_w0p3_tied:
+** mov (w1[2-5]), w0
+** mova z0\.s, p0/m, za1h\.s\[\1, 3\]
+** ret
+*/
+TEST_READ_ZA (read_za32_s32_1_w0p3_tied, svint32_t,
+ z0 = svread_hor_za32_s32_m (z0, p0, 1, w0 + 3),
+ z0 = svread_hor_za32_m (z0, p0, 1, w0 + 3))
+
+/*
+** read_za32_s32_3_w0_tied:
+** mov (w1[2-5]), w0
+** mova z0\.s, p0/m, za3h\.s\[\1, 0\]
+** ret
+*/
+TEST_READ_ZA (read_za32_s32_3_w0_tied, svint32_t,
+ z0 = svread_hor_za32_s32_m (z0, p0, 3, w0),
+ z0 = svread_hor_za32_m (z0, p0, 3, w0))
+
+/*
+** read_za32_s32_3_w0p3_tied:
+** mov (w1[2-5]), w0
+** mova z0\.s, p0/m, za3h\.s\[\1, 3\]
+** ret
+*/
+TEST_READ_ZA (read_za32_s32_3_w0p3_tied, svint32_t,
+ z0 = svread_hor_za32_s32_m (z0, p0, 3, w0 + 3),
+ z0 = svread_hor_za32_m (z0, p0, 3, w0 + 3))
+
+/*
+** read_za32_s32_0_w0_untied:
+** (
+** mov (w1[2-5]), w0
+** mov z0\.d, z1\.d
+** mova z0\.s, p0/m, za0h\.s\[\1, 0\]
+** |
+** mov z0\.d, z1\.d
+** mov (w1[2-5]), w0
+** mova z0\.s, p0/m, za0h\.s\[\2, 0\]
+** |
+** mov (w1[2-5]), w0
+** mova z1\.s, p0/m, za0h\.s\[\3, 0\]
+** mov z0\.d, z1\.d
+** )
+** ret
+*/
+TEST_READ_ZA (read_za32_s32_0_w0_untied, svint32_t,
+ z0 = svread_hor_za32_s32_m (z1, p0, 0, w0),
+ z0 = svread_hor_za32_m (z1, p0, 0, w0))
+
+/*
+** read_za32_u32_0_w0_tied:
+** mov (w1[2-5]), w0
+** mova z0\.s, p0/m, za0h\.s\[\1, 0\]
+** ret
+*/
+TEST_READ_ZA (read_za32_u32_0_w0_tied, svuint32_t,
+ z0 = svread_hor_za32_u32_m (z0, p0, 0, w0),
+ z0 = svread_hor_za32_m (z0, p0, 0, w0))
+
+/*
+** read_za32_u32_0_w0_untied:
+** (
+** mov (w1[2-5]), w0
+** mov z0\.d, z1\.d
+** mova z0\.s, p0/m, za0h\.s\[\1, 0\]
+** |
+** mov z0\.d, z1\.d
+** mov (w1[2-5]), w0
+** mova z0\.s, p0/m, za0h\.s\[\2, 0\]
+** |
+** mov (w1[2-5]), w0
+** mova z1\.s, p0/m, za0h\.s\[\3, 0\]
+** mov z0\.d, z1\.d
+** )
+** ret
+*/
+TEST_READ_ZA (read_za32_u32_0_w0_untied, svuint32_t,
+ z0 = svread_hor_za32_u32_m (z1, p0, 0, w0),
+ z0 = svread_hor_za32_m (z1, p0, 0, w0))
+
+/*
+** read_za32_f32_0_w0_tied:
+** mov (w1[2-5]), w0
+** mova z0\.s, p0/m, za0h\.s\[\1, 0\]
+** ret
+*/
+TEST_READ_ZA (read_za32_f32_0_w0_tied, svfloat32_t,
+ z0 = svread_hor_za32_f32_m (z0, p0, 0, w0),
+ z0 = svread_hor_za32_m (z0, p0, 0, w0))
+
+/*
+** read_za32_f32_0_w0_untied:
+** (
+** mov (w1[2-5]), w0
+** mov z0\.d, z1\.d
+** mova z0\.s, p0/m, za0h\.s\[\1, 0\]
+** |
+** mov z0\.d, z1\.d
+** mov (w1[2-5]), w0
+** mova z0\.s, p0/m, za0h\.s\[\2, 0\]
+** |
+** mov (w1[2-5]), w0
+** mova z1\.s, p0/m, za0h\.s\[\3, 0\]
+** mov z0\.d, z1\.d
+** )
+** ret
+*/
+TEST_READ_ZA (read_za32_f32_0_w0_untied, svfloat32_t,
+ z0 = svread_hor_za32_f32_m (z1, p0, 0, w0),
+ z0 = svread_hor_za32_m (z1, p0, 0, w0))
--- /dev/null
+/* { dg-final { check-function-bodies "**" "" "-DCHECK_ASM" } } */
+
+#include "test_sme_acle.h"
+
+/*
+** read_za64_s64_0_0_tied:
+** mov (w1[2-5]), (?:wzr|#?0)
+** mova z0\.d, p0/m, za0h\.d\[\1, 0\]
+** ret
+*/
+TEST_READ_ZA (read_za64_s64_0_0_tied, svint64_t,
+ z0 = svread_hor_za64_s64_m (z0, p0, 0, 0),
+ z0 = svread_hor_za64_m (z0, p0, 0, 0))
+
+/*
+** read_za64_s64_0_1_tied:
+** mov (w1[2-5]), #?1
+** mova z0\.d, p0/m, za0h\.d\[\1, 0\]
+** ret
+*/
+TEST_READ_ZA (read_za64_s64_0_1_tied, svint64_t,
+ z0 = svread_hor_za64_s64_m (z0, p0, 0, 1),
+ z0 = svread_hor_za64_m (z0, p0, 0, 1))
+
+/*
+** read_za64_s64_0_w0_tied:
+** mov (w1[2-5]), w0
+** mova z0\.d, p0/m, za0h\.d\[\1, 0\]
+** ret
+*/
+TEST_READ_ZA (read_za64_s64_0_w0_tied, svint64_t,
+ z0 = svread_hor_za64_s64_m (z0, p0, 0, w0),
+ z0 = svread_hor_za64_m (z0, p0, 0, w0))
+
+/*
+** read_za64_s64_0_w0p1_tied:
+** mov (w1[2-5]), w0
+** mova z0\.d, p0/m, za0h\.d\[\1, 1\]
+** ret
+*/
+TEST_READ_ZA (read_za64_s64_0_w0p1_tied, svint64_t,
+ z0 = svread_hor_za64_s64_m (z0, p0, 0, w0 + 1),
+ z0 = svread_hor_za64_m (z0, p0, 0, w0 + 1))
+
+/*
+** read_za64_s64_0_w0p2_tied:
+** add (w1[2-5]), w0, #?2
+** mova z0\.d, p0/m, za0h\.d\[\1, 0\]
+** ret
+*/
+TEST_READ_ZA (read_za64_s64_0_w0p2_tied, svint64_t,
+ z0 = svread_hor_za64_s64_m (z0, p0, 0, w0 + 2),
+ z0 = svread_hor_za64_m (z0, p0, 0, w0 + 2))
+
+/*
+** read_za64_s64_0_w0m1_tied:
+** sub (w1[2-5]), w0, #?1
+** mova z0\.d, p0/m, za0h\.d\[\1, 0\]
+** ret
+*/
+TEST_READ_ZA (read_za64_s64_0_w0m1_tied, svint64_t,
+ z0 = svread_hor_za64_s64_m (z0, p0, 0, w0 - 1),
+ z0 = svread_hor_za64_m (z0, p0, 0, w0 - 1))
+
+/*
+** read_za64_s64_1_w0_tied:
+** mov (w1[2-5]), w0
+** mova z0\.d, p0/m, za1h\.d\[\1, 0\]
+** ret
+*/
+TEST_READ_ZA (read_za64_s64_1_w0_tied, svint64_t,
+ z0 = svread_hor_za64_s64_m (z0, p0, 1, w0),
+ z0 = svread_hor_za64_m (z0, p0, 1, w0))
+
+/*
+** read_za64_s64_1_w0p1_tied:
+** mov (w1[2-5]), w0
+** mova z0\.d, p0/m, za1h\.d\[\1, 1\]
+** ret
+*/
+TEST_READ_ZA (read_za64_s64_1_w0p1_tied, svint64_t,
+ z0 = svread_hor_za64_s64_m (z0, p0, 1, w0 + 1),
+ z0 = svread_hor_za64_m (z0, p0, 1, w0 + 1))
+
+/*
+** read_za64_s64_7_w0_tied:
+** mov (w1[2-5]), w0
+** mova z0\.d, p0/m, za7h\.d\[\1, 0\]
+** ret
+*/
+TEST_READ_ZA (read_za64_s64_7_w0_tied, svint64_t,
+ z0 = svread_hor_za64_s64_m (z0, p0, 7, w0),
+ z0 = svread_hor_za64_m (z0, p0, 7, w0))
+
+/*
+** read_za64_s64_7_w0p1_tied:
+** mov (w1[2-5]), w0
+** mova z0\.d, p0/m, za7h\.d\[\1, 1\]
+** ret
+*/
+TEST_READ_ZA (read_za64_s64_7_w0p1_tied, svint64_t,
+ z0 = svread_hor_za64_s64_m (z0, p0, 7, w0 + 1),
+ z0 = svread_hor_za64_m (z0, p0, 7, w0 + 1))
+
+/*
+** read_za64_s64_0_w0_untied:
+** (
+** mov (w1[2-5]), w0
+** mov z0\.d, z1\.d
+** mova z0\.d, p0/m, za0h\.d\[\1, 0\]
+** |
+** mov z0\.d, z1\.d
+** mov (w1[2-5]), w0
+** mova z0\.d, p0/m, za0h\.d\[\2, 0\]
+** |
+** mov (w1[2-5]), w0
+** mova z1\.d, p0/m, za0h\.d\[\3, 0\]
+** mov z0\.d, z1\.d
+** )
+** ret
+*/
+TEST_READ_ZA (read_za64_s64_0_w0_untied, svint64_t,
+ z0 = svread_hor_za64_s64_m (z1, p0, 0, w0),
+ z0 = svread_hor_za64_m (z1, p0, 0, w0))
+
+/*
+** read_za64_u64_0_w0_tied:
+** mov (w1[2-5]), w0
+** mova z0\.d, p0/m, za0h\.d\[\1, 0\]
+** ret
+*/
+TEST_READ_ZA (read_za64_u64_0_w0_tied, svuint64_t,
+ z0 = svread_hor_za64_u64_m (z0, p0, 0, w0),
+ z0 = svread_hor_za64_m (z0, p0, 0, w0))
+
+/*
+** read_za64_u64_0_w0_untied:
+** (
+** mov (w1[2-5]), w0
+** mov z0\.d, z1\.d
+** mova z0\.d, p0/m, za0h\.d\[\1, 0\]
+** |
+** mov z0\.d, z1\.d
+** mov (w1[2-5]), w0
+** mova z0\.d, p0/m, za0h\.d\[\2, 0\]
+** |
+** mov (w1[2-5]), w0
+** mova z1\.d, p0/m, za0h\.d\[\3, 0\]
+** mov z0\.d, z1\.d
+** )
+** ret
+*/
+TEST_READ_ZA (read_za64_u64_0_w0_untied, svuint64_t,
+ z0 = svread_hor_za64_u64_m (z1, p0, 0, w0),
+ z0 = svread_hor_za64_m (z1, p0, 0, w0))
+
+/*
+** read_za64_f64_0_w0_tied:
+** mov (w1[2-5]), w0
+** mova z0\.d, p0/m, za0h\.d\[\1, 0\]
+** ret
+*/
+TEST_READ_ZA (read_za64_f64_0_w0_tied, svfloat64_t,
+ z0 = svread_hor_za64_f64_m (z0, p0, 0, w0),
+ z0 = svread_hor_za64_m (z0, p0, 0, w0))
+
+/*
+** read_za64_f64_0_w0_untied:
+** (
+** mov (w1[2-5]), w0
+** mov z0\.d, z1\.d
+** mova z0\.d, p0/m, za0h\.d\[\1, 0\]
+** |
+** mov z0\.d, z1\.d
+** mov (w1[2-5]), w0
+** mova z0\.d, p0/m, za0h\.d\[\2, 0\]
+** |
+** mov (w1[2-5]), w0
+** mova z1\.d, p0/m, za0h\.d\[\3, 0\]
+** mov z0\.d, z1\.d
+** )
+** ret
+*/
+TEST_READ_ZA (read_za64_f64_0_w0_untied, svfloat64_t,
+ z0 = svread_hor_za64_f64_m (z1, p0, 0, w0),
+ z0 = svread_hor_za64_m (z1, p0, 0, w0))
--- /dev/null
+/* { dg-final { check-function-bodies "**" "" "-DCHECK_ASM" } } */
+
+#include "test_sme_acle.h"
+
+/*
+** read_za8_s8_0_0_tied:
+** mov (w1[2-5]), (?:wzr|#?0)
+** mova z0\.b, p0/m, za0h\.b\[\1, 0\]
+** ret
+*/
+TEST_READ_ZA (read_za8_s8_0_0_tied, svint8_t,
+ z0 = svread_hor_za8_s8_m (z0, p0, 0, 0),
+ z0 = svread_hor_za8_m (z0, p0, 0, 0))
+
+/*
+** read_za8_s8_0_1_tied:
+** mov (w1[2-5]), #?1
+** mova z0\.b, p0/m, za0h\.b\[\1, 0\]
+** ret
+*/
+TEST_READ_ZA (read_za8_s8_0_1_tied, svint8_t,
+ z0 = svread_hor_za8_s8_m (z0, p0, 0, 1),
+ z0 = svread_hor_za8_m (z0, p0, 0, 1))
+
+/*
+** read_za8_s8_0_w0_tied:
+** mov (w1[2-5]), w0
+** mova z0\.b, p0/m, za0h\.b\[\1, 0\]
+** ret
+*/
+TEST_READ_ZA (read_za8_s8_0_w0_tied, svint8_t,
+ z0 = svread_hor_za8_s8_m (z0, p0, 0, w0),
+ z0 = svread_hor_za8_m (z0, p0, 0, w0))
+
+/*
+** read_za8_s8_0_w0p1_tied:
+** mov (w1[2-5]), w0
+** mova z0\.b, p0/m, za0h\.b\[\1, 1\]
+** ret
+*/
+TEST_READ_ZA (read_za8_s8_0_w0p1_tied, svint8_t,
+ z0 = svread_hor_za8_s8_m (z0, p0, 0, w0 + 1),
+ z0 = svread_hor_za8_m (z0, p0, 0, w0 + 1))
+
+/*
+** read_za8_s8_0_w0p15_tied:
+** mov (w1[2-5]), w0
+** mova z0\.b, p0/m, za0h\.b\[\1, 15\]
+** ret
+*/
+TEST_READ_ZA (read_za8_s8_0_w0p15_tied, svint8_t,
+ z0 = svread_hor_za8_s8_m (z0, p0, 0, w0 + 15),
+ z0 = svread_hor_za8_m (z0, p0, 0, w0 + 15))
+
+/*
+** read_za8_s8_0_w0p16_tied:
+** add (w1[2-5]), w0, #?16
+** mova z0\.b, p0/m, za0h\.b\[\1, 0\]
+** ret
+*/
+TEST_READ_ZA (read_za8_s8_0_w0p16_tied, svint8_t,
+ z0 = svread_hor_za8_s8_m (z0, p0, 0, w0 + 16),
+ z0 = svread_hor_za8_m (z0, p0, 0, w0 + 16))
+
+/*
+** read_za8_s8_0_w0m1_tied:
+** sub (w1[2-5]), w0, #?1
+** mova z0\.b, p0/m, za0h\.b\[\1, 0\]
+** ret
+*/
+TEST_READ_ZA (read_za8_s8_0_w0m1_tied, svint8_t,
+ z0 = svread_hor_za8_s8_m (z0, p0, 0, w0 - 1),
+ z0 = svread_hor_za8_m (z0, p0, 0, w0 - 1))
+
+/*
+** read_za8_s8_0_w0_untied:
+** (
+** mov (w1[2-5]), w0
+** mov z0\.d, z1\.d
+** mova z0\.b, p0/m, za0h\.b\[\1, 0\]
+** |
+** mov z0\.d, z1\.d
+** mov (w1[2-5]), w0
+** mova z0\.b, p0/m, za0h\.b\[\2, 0\]
+** |
+** mov (w1[2-5]), w0
+** mova z1\.b, p0/m, za0h\.b\[\3, 0\]
+** mov z0\.d, z1\.d
+** )
+** ret
+*/
+TEST_READ_ZA (read_za8_s8_0_w0_untied, svint8_t,
+ z0 = svread_hor_za8_s8_m (z1, p0, 0, w0),
+ z0 = svread_hor_za8_m (z1, p0, 0, w0))
+
+/*
+** read_za8_u8_0_w0_tied:
+** mov (w1[2-5]), w0
+** mova z0\.b, p0/m, za0h\.b\[\1, 0\]
+** ret
+*/
+TEST_READ_ZA (read_za8_u8_0_w0_tied, svuint8_t,
+ z0 = svread_hor_za8_u8_m (z0, p0, 0, w0),
+ z0 = svread_hor_za8_m (z0, p0, 0, w0))
+
+/*
+** read_za8_u8_0_w0_untied:
+** (
+** mov (w1[2-5]), w0
+** mov z0\.d, z1\.d
+** mova z0\.b, p0/m, za0h\.b\[\1, 0\]
+** |
+** mov z0\.d, z1\.d
+** mov (w1[2-5]), w0
+** mova z0\.b, p0/m, za0h\.b\[\2, 0\]
+** |
+** mov (w1[2-5]), w0
+** mova z1\.b, p0/m, za0h\.b\[\3, 0\]
+** mov z0\.d, z1\.d
+** )
+** ret
+*/
+TEST_READ_ZA (read_za8_u8_0_w0_untied, svuint8_t,
+ z0 = svread_hor_za8_u8_m (z1, p0, 0, w0),
+ z0 = svread_hor_za8_m (z1, p0, 0, w0))
--- /dev/null
+/* { dg-final { check-function-bodies "**" "" "-DCHECK_ASM" } } */
+
+#include "test_sme_acle.h"
+
+/*
+** read_za128_s8_0_0_tied:
+** mov (w1[2-5]), (?:wzr|#?0)
+** mova z0\.q, p0/m, za0v\.q\[\1, 0\]
+** ret
+*/
+TEST_READ_ZA (read_za128_s8_0_0_tied, svint8_t,
+ z0 = svread_ver_za128_s8_m (z0, p0, 0, 0),
+ z0 = svread_ver_za128_m (z0, p0, 0, 0))
+
+/*
+** read_za128_s8_0_1_tied:
+** mov (w1[2-5]), #?1
+** mova z0\.q, p0/m, za0v\.q\[\1, 0\]
+** ret
+*/
+TEST_READ_ZA (read_za128_s8_0_1_tied, svint8_t,
+ z0 = svread_ver_za128_s8_m (z0, p0, 0, 1),
+ z0 = svread_ver_za128_m (z0, p0, 0, 1))
+
+/*
+** read_za128_s8_0_w0_tied:
+** mov (w1[2-5]), w0
+** mova z0\.q, p0/m, za0v\.q\[\1, 0\]
+** ret
+*/
+TEST_READ_ZA (read_za128_s8_0_w0_tied, svint8_t,
+ z0 = svread_ver_za128_s8_m (z0, p0, 0, w0),
+ z0 = svread_ver_za128_m (z0, p0, 0, w0))
+
+/*
+** read_za128_s8_0_w0p1_tied:
+** add (w1[2-5]), w0, #?1
+** mova z0\.q, p0/m, za0v\.q\[\1, 0\]
+** ret
+*/
+TEST_READ_ZA (read_za128_s8_0_w0p1_tied, svint8_t,
+ z0 = svread_ver_za128_s8_m (z0, p0, 0, w0 + 1),
+ z0 = svread_ver_za128_m (z0, p0, 0, w0 + 1))
+
+/*
+** read_za128_s8_0_w0m1_tied:
+** sub (w1[2-5]), w0, #?1
+** mova z0\.q, p0/m, za0v\.q\[\1, 0\]
+** ret
+*/
+TEST_READ_ZA (read_za128_s8_0_w0m1_tied, svint8_t,
+ z0 = svread_ver_za128_s8_m (z0, p0, 0, w0 - 1),
+ z0 = svread_ver_za128_m (z0, p0, 0, w0 - 1))
+
+/*
+** read_za128_s8_1_w0_tied:
+** mov (w1[2-5]), w0
+** mova z0\.q, p0/m, za1v\.q\[\1, 0\]
+** ret
+*/
+TEST_READ_ZA (read_za128_s8_1_w0_tied, svint8_t,
+ z0 = svread_ver_za128_s8_m (z0, p0, 1, w0),
+ z0 = svread_ver_za128_m (z0, p0, 1, w0))
+
+/*
+** read_za128_s8_15_w0_tied:
+** mov (w1[2-5]), w0
+** mova z0\.q, p0/m, za15v\.q\[\1, 0\]
+** ret
+*/
+TEST_READ_ZA (read_za128_s8_15_w0_tied, svint8_t,
+ z0 = svread_ver_za128_s8_m (z0, p0, 15, w0),
+ z0 = svread_ver_za128_m (z0, p0, 15, w0))
+
+/*
+** read_za128_s8_0_w0_untied:
+** (
+** mov (w1[2-5]), w0
+** mov z0\.d, z1\.d
+** mova z0\.q, p0/m, za0v\.q\[\1, 0\]
+** |
+** mov z0\.d, z1\.d
+** mov (w1[2-5]), w0
+** mova z0\.q, p0/m, za0v\.q\[\2, 0\]
+** |
+** mov (w1[2-5]), w0
+** mova z1\.q, p0/m, za0v\.q\[\3, 0\]
+** mov z0\.d, z1\.d
+** )
+** ret
+*/
+TEST_READ_ZA (read_za128_s8_0_w0_untied, svint8_t,
+ z0 = svread_ver_za128_s8_m (z1, p0, 0, w0),
+ z0 = svread_ver_za128_m (z1, p0, 0, w0))
+
+/*
+** read_za128_u8_0_w0_tied:
+** mov (w1[2-5]), w0
+** mova z0\.q, p0/m, za0v\.q\[\1, 0\]
+** ret
+*/
+TEST_READ_ZA (read_za128_u8_0_w0_tied, svuint8_t,
+ z0 = svread_ver_za128_u8_m (z0, p0, 0, w0),
+ z0 = svread_ver_za128_m (z0, p0, 0, w0))
+
+/*
+** read_za128_u8_0_w0_untied:
+** (
+** mov (w1[2-5]), w0
+** mov z0\.d, z1\.d
+** mova z0\.q, p0/m, za0v\.q\[\1, 0\]
+** |
+** mov z0\.d, z1\.d
+** mov (w1[2-5]), w0
+** mova z0\.q, p0/m, za0v\.q\[\2, 0\]
+** |
+** mov (w1[2-5]), w0
+** mova z1\.q, p0/m, za0v\.q\[\3, 0\]
+** mov z0\.d, z1\.d
+** )
+** ret
+*/
+TEST_READ_ZA (read_za128_u8_0_w0_untied, svuint8_t,
+ z0 = svread_ver_za128_u8_m (z1, p0, 0, w0),
+ z0 = svread_ver_za128_m (z1, p0, 0, w0))
+
+/*
+** read_za128_s16_0_w0_tied:
+** mov (w1[2-5]), w0
+** mova z0\.q, p0/m, za0v\.q\[\1, 0\]
+** ret
+*/
+TEST_READ_ZA (read_za128_s16_0_w0_tied, svint16_t,
+ z0 = svread_ver_za128_s16_m (z0, p0, 0, w0),
+ z0 = svread_ver_za128_m (z0, p0, 0, w0))
+
+/*
+** read_za128_s16_0_w0_untied:
+** (
+** mov (w1[2-5]), w0
+** mov z0\.d, z1\.d
+** mova z0\.q, p0/m, za0v\.q\[\1, 0\]
+** |
+** mov z0\.d, z1\.d
+** mov (w1[2-5]), w0
+** mova z0\.q, p0/m, za0v\.q\[\2, 0\]
+** |
+** mov (w1[2-5]), w0
+** mova z1\.q, p0/m, za0v\.q\[\3, 0\]
+** mov z0\.d, z1\.d
+** )
+** ret
+*/
+TEST_READ_ZA (read_za128_s16_0_w0_untied, svint16_t,
+ z0 = svread_ver_za128_s16_m (z1, p0, 0, w0),
+ z0 = svread_ver_za128_m (z1, p0, 0, w0))
+
+/*
+** read_za128_u16_0_w0_tied:
+** mov (w1[2-5]), w0
+** mova z0\.q, p0/m, za0v\.q\[\1, 0\]
+** ret
+*/
+TEST_READ_ZA (read_za128_u16_0_w0_tied, svuint16_t,
+ z0 = svread_ver_za128_u16_m (z0, p0, 0, w0),
+ z0 = svread_ver_za128_m (z0, p0, 0, w0))
+
+/*
+** read_za128_u16_0_w0_untied:
+** (
+** mov (w1[2-5]), w0
+** mov z0\.d, z1\.d
+** mova z0\.q, p0/m, za0v\.q\[\1, 0\]
+** |
+** mov z0\.d, z1\.d
+** mov (w1[2-5]), w0
+** mova z0\.q, p0/m, za0v\.q\[\2, 0\]
+** |
+** mov (w1[2-5]), w0
+** mova z1\.q, p0/m, za0v\.q\[\3, 0\]
+** mov z0\.d, z1\.d
+** )
+** ret
+*/
+TEST_READ_ZA (read_za128_u16_0_w0_untied, svuint16_t,
+ z0 = svread_ver_za128_u16_m (z1, p0, 0, w0),
+ z0 = svread_ver_za128_m (z1, p0, 0, w0))
+
+/*
+** read_za128_f16_0_w0_tied:
+** mov (w1[2-5]), w0
+** mova z0\.q, p0/m, za0v\.q\[\1, 0\]
+** ret
+*/
+TEST_READ_ZA (read_za128_f16_0_w0_tied, svfloat16_t,
+ z0 = svread_ver_za128_f16_m (z0, p0, 0, w0),
+ z0 = svread_ver_za128_m (z0, p0, 0, w0))
+
+/*
+** read_za128_f16_0_w0_untied:
+** (
+** mov (w1[2-5]), w0
+** mov z0\.d, z1\.d
+** mova z0\.q, p0/m, za0v\.q\[\1, 0\]
+** |
+** mov z0\.d, z1\.d
+** mov (w1[2-5]), w0
+** mova z0\.q, p0/m, za0v\.q\[\2, 0\]
+** |
+** mov (w1[2-5]), w0
+** mova z1\.q, p0/m, za0v\.q\[\3, 0\]
+** mov z0\.d, z1\.d
+** )
+** ret
+*/
+TEST_READ_ZA (read_za128_f16_0_w0_untied, svfloat16_t,
+ z0 = svread_ver_za128_f16_m (z1, p0, 0, w0),
+ z0 = svread_ver_za128_m (z1, p0, 0, w0))
+
+/*
+** read_za128_bf16_0_w0_tied:
+** mov (w1[2-5]), w0
+** mova z0\.q, p0/m, za0v\.q\[\1, 0\]
+** ret
+*/
+TEST_READ_ZA (read_za128_bf16_0_w0_tied, svbfloat16_t,
+ z0 = svread_ver_za128_bf16_m (z0, p0, 0, w0),
+ z0 = svread_ver_za128_m (z0, p0, 0, w0))
+
+/*
+** read_za128_bf16_0_w0_untied:
+** (
+** mov (w1[2-5]), w0
+** mov z0\.d, z1\.d
+** mova z0\.q, p0/m, za0v\.q\[\1, 0\]
+** |
+** mov z0\.d, z1\.d
+** mov (w1[2-5]), w0
+** mova z0\.q, p0/m, za0v\.q\[\2, 0\]
+** |
+** mov (w1[2-5]), w0
+** mova z1\.q, p0/m, za0v\.q\[\3, 0\]
+** mov z0\.d, z1\.d
+** )
+** ret
+*/
+TEST_READ_ZA (read_za128_bf16_0_w0_untied, svbfloat16_t,
+ z0 = svread_ver_za128_bf16_m (z1, p0, 0, w0),
+ z0 = svread_ver_za128_m (z1, p0, 0, w0))
+
+/*
+** read_za128_s32_0_w0_tied:
+** mov (w1[2-5]), w0
+** mova z0\.q, p0/m, za0v\.q\[\1, 0\]
+** ret
+*/
+TEST_READ_ZA (read_za128_s32_0_w0_tied, svint32_t,
+ z0 = svread_ver_za128_s32_m (z0, p0, 0, w0),
+ z0 = svread_ver_za128_m (z0, p0, 0, w0))
+
+/*
+** read_za128_s32_0_w0_untied:
+** (
+** mov (w1[2-5]), w0
+** mov z0\.d, z1\.d
+** mova z0\.q, p0/m, za0v\.q\[\1, 0\]
+** |
+** mov z0\.d, z1\.d
+** mov (w1[2-5]), w0
+** mova z0\.q, p0/m, za0v\.q\[\2, 0\]
+** |
+** mov (w1[2-5]), w0
+** mova z1\.q, p0/m, za0v\.q\[\3, 0\]
+** mov z0\.d, z1\.d
+** )
+** ret
+*/
+TEST_READ_ZA (read_za128_s32_0_w0_untied, svint32_t,
+ z0 = svread_ver_za128_s32_m (z1, p0, 0, w0),
+ z0 = svread_ver_za128_m (z1, p0, 0, w0))
+
+/*
+** read_za128_u32_0_w0_tied:
+** mov (w1[2-5]), w0
+** mova z0\.q, p0/m, za0v\.q\[\1, 0\]
+** ret
+*/
+TEST_READ_ZA (read_za128_u32_0_w0_tied, svuint32_t,
+ z0 = svread_ver_za128_u32_m (z0, p0, 0, w0),
+ z0 = svread_ver_za128_m (z0, p0, 0, w0))
+
+/*
+** read_za128_u32_0_w0_untied:
+** (
+** mov (w1[2-5]), w0
+** mov z0\.d, z1\.d
+** mova z0\.q, p0/m, za0v\.q\[\1, 0\]
+** |
+** mov z0\.d, z1\.d
+** mov (w1[2-5]), w0
+** mova z0\.q, p0/m, za0v\.q\[\2, 0\]
+** |
+** mov (w1[2-5]), w0
+** mova z1\.q, p0/m, za0v\.q\[\3, 0\]
+** mov z0\.d, z1\.d
+** )
+** ret
+*/
+TEST_READ_ZA (read_za128_u32_0_w0_untied, svuint32_t,
+ z0 = svread_ver_za128_u32_m (z1, p0, 0, w0),
+ z0 = svread_ver_za128_m (z1, p0, 0, w0))
+
+/*
+** read_za128_f32_0_w0_tied:
+** mov (w1[2-5]), w0
+** mova z0\.q, p0/m, za0v\.q\[\1, 0\]
+** ret
+*/
+TEST_READ_ZA (read_za128_f32_0_w0_tied, svfloat32_t,
+ z0 = svread_ver_za128_f32_m (z0, p0, 0, w0),
+ z0 = svread_ver_za128_m (z0, p0, 0, w0))
+
+/*
+** read_za128_f32_0_w0_untied:
+** (
+** mov (w1[2-5]), w0
+** mov z0\.d, z1\.d
+** mova z0\.q, p0/m, za0v\.q\[\1, 0\]
+** |
+** mov z0\.d, z1\.d
+** mov (w1[2-5]), w0
+** mova z0\.q, p0/m, za0v\.q\[\2, 0\]
+** |
+** mov (w1[2-5]), w0
+** mova z1\.q, p0/m, za0v\.q\[\3, 0\]
+** mov z0\.d, z1\.d
+** )
+** ret
+*/
+TEST_READ_ZA (read_za128_f32_0_w0_untied, svfloat32_t,
+ z0 = svread_ver_za128_f32_m (z1, p0, 0, w0),
+ z0 = svread_ver_za128_m (z1, p0, 0, w0))
+
+/*
+** read_za128_s64_0_w0_tied:
+** mov (w1[2-5]), w0
+** mova z0\.q, p0/m, za0v\.q\[\1, 0\]
+** ret
+*/
+TEST_READ_ZA (read_za128_s64_0_w0_tied, svint64_t,
+ z0 = svread_ver_za128_s64_m (z0, p0, 0, w0),
+ z0 = svread_ver_za128_m (z0, p0, 0, w0))
+
+/*
+** read_za128_s64_0_w0_untied:
+** (
+** mov (w1[2-5]), w0
+** mov z0\.d, z1\.d
+** mova z0\.q, p0/m, za0v\.q\[\1, 0\]
+** |
+** mov z0\.d, z1\.d
+** mov (w1[2-5]), w0
+** mova z0\.q, p0/m, za0v\.q\[\2, 0\]
+** |
+** mov (w1[2-5]), w0
+** mova z1\.q, p0/m, za0v\.q\[\3, 0\]
+** mov z0\.d, z1\.d
+** )
+** ret
+*/
+TEST_READ_ZA (read_za128_s64_0_w0_untied, svint64_t,
+ z0 = svread_ver_za128_s64_m (z1, p0, 0, w0),
+ z0 = svread_ver_za128_m (z1, p0, 0, w0))
+
+/*
+** read_za128_u64_0_w0_tied:
+** mov (w1[2-5]), w0
+** mova z0\.q, p0/m, za0v\.q\[\1, 0\]
+** ret
+*/
+TEST_READ_ZA (read_za128_u64_0_w0_tied, svuint64_t,
+ z0 = svread_ver_za128_u64_m (z0, p0, 0, w0),
+ z0 = svread_ver_za128_m (z0, p0, 0, w0))
+
+/*
+** read_za128_u64_0_w0_untied:
+** (
+** mov (w1[2-5]), w0
+** mov z0\.d, z1\.d
+** mova z0\.q, p0/m, za0v\.q\[\1, 0\]
+** |
+** mov z0\.d, z1\.d
+** mov (w1[2-5]), w0
+** mova z0\.q, p0/m, za0v\.q\[\2, 0\]
+** |
+** mov (w1[2-5]), w0
+** mova z1\.q, p0/m, za0v\.q\[\3, 0\]
+** mov z0\.d, z1\.d
+** )
+** ret
+*/
+TEST_READ_ZA (read_za128_u64_0_w0_untied, svuint64_t,
+ z0 = svread_ver_za128_u64_m (z1, p0, 0, w0),
+ z0 = svread_ver_za128_m (z1, p0, 0, w0))
+
+/*
+** read_za128_f64_0_w0_tied:
+** mov (w1[2-5]), w0
+** mova z0\.q, p0/m, za0v\.q\[\1, 0\]
+** ret
+*/
+TEST_READ_ZA (read_za128_f64_0_w0_tied, svfloat64_t,
+ z0 = svread_ver_za128_f64_m (z0, p0, 0, w0),
+ z0 = svread_ver_za128_m (z0, p0, 0, w0))
+
+/*
+** read_za128_f64_0_w0_untied:
+** (
+** mov (w1[2-5]), w0
+** mov z0\.d, z1\.d
+** mova z0\.q, p0/m, za0v\.q\[\1, 0\]
+** |
+** mov z0\.d, z1\.d
+** mov (w1[2-5]), w0
+** mova z0\.q, p0/m, za0v\.q\[\2, 0\]
+** |
+** mov (w1[2-5]), w0
+** mova z1\.q, p0/m, za0v\.q\[\3, 0\]
+** mov z0\.d, z1\.d
+** )
+** ret
+*/
+TEST_READ_ZA (read_za128_f64_0_w0_untied, svfloat64_t,
+ z0 = svread_ver_za128_f64_m (z1, p0, 0, w0),
+ z0 = svread_ver_za128_m (z1, p0, 0, w0))
--- /dev/null
+/* { dg-final { check-function-bodies "**" "" "-DCHECK_ASM" } } */
+
+#include "test_sme_acle.h"
+
+/*
+** read_za16_s16_0_0_tied:
+** mov (w1[2-5]), (?:wzr|#?0)
+** mova z0\.h, p0/m, za0v\.h\[\1, 0\]
+** ret
+*/
+TEST_READ_ZA (read_za16_s16_0_0_tied, svint16_t,
+ z0 = svread_ver_za16_s16_m (z0, p0, 0, 0),
+ z0 = svread_ver_za16_m (z0, p0, 0, 0))
+
+/*
+** read_za16_s16_0_1_tied:
+** mov (w1[2-5]), #?1
+** mova z0\.h, p0/m, za0v\.h\[\1, 0\]
+** ret
+*/
+TEST_READ_ZA (read_za16_s16_0_1_tied, svint16_t,
+ z0 = svread_ver_za16_s16_m (z0, p0, 0, 1),
+ z0 = svread_ver_za16_m (z0, p0, 0, 1))
+
+/*
+** read_za16_s16_0_w0_tied:
+** mov (w1[2-5]), w0
+** mova z0\.h, p0/m, za0v\.h\[\1, 0\]
+** ret
+*/
+TEST_READ_ZA (read_za16_s16_0_w0_tied, svint16_t,
+ z0 = svread_ver_za16_s16_m (z0, p0, 0, w0),
+ z0 = svread_ver_za16_m (z0, p0, 0, w0))
+
+/*
+** read_za16_s16_0_w0p1_tied:
+** mov (w1[2-5]), w0
+** mova z0\.h, p0/m, za0v\.h\[\1, 1\]
+** ret
+*/
+TEST_READ_ZA (read_za16_s16_0_w0p1_tied, svint16_t,
+ z0 = svread_ver_za16_s16_m (z0, p0, 0, w0 + 1),
+ z0 = svread_ver_za16_m (z0, p0, 0, w0 + 1))
+
+/*
+** read_za16_s16_0_w0p7_tied:
+** mov (w1[2-5]), w0
+** mova z0\.h, p0/m, za0v\.h\[\1, 7\]
+** ret
+*/
+TEST_READ_ZA (read_za16_s16_0_w0p7_tied, svint16_t,
+ z0 = svread_ver_za16_s16_m (z0, p0, 0, w0 + 7),
+ z0 = svread_ver_za16_m (z0, p0, 0, w0 + 7))
+
+/*
+** read_za16_s16_0_w0p8_tied:
+** add (w1[2-5]), w0, #?8
+** mova z0\.h, p0/m, za0v\.h\[\1, 0\]
+** ret
+*/
+TEST_READ_ZA (read_za16_s16_0_w0p8_tied, svint16_t,
+ z0 = svread_ver_za16_s16_m (z0, p0, 0, w0 + 8),
+ z0 = svread_ver_za16_m (z0, p0, 0, w0 + 8))
+
+/*
+** read_za16_s16_0_w0m1_tied:
+** sub (w1[2-5]), w0, #?1
+** mova z0\.h, p0/m, za0v\.h\[\1, 0\]
+** ret
+*/
+TEST_READ_ZA (read_za16_s16_0_w0m1_tied, svint16_t,
+ z0 = svread_ver_za16_s16_m (z0, p0, 0, w0 - 1),
+ z0 = svread_ver_za16_m (z0, p0, 0, w0 - 1))
+
+/*
+** read_za16_s16_1_w0_tied:
+** mov (w1[2-5]), w0
+** mova z0\.h, p0/m, za1v\.h\[\1, 0\]
+** ret
+*/
+TEST_READ_ZA (read_za16_s16_1_w0_tied, svint16_t,
+ z0 = svread_ver_za16_s16_m (z0, p0, 1, w0),
+ z0 = svread_ver_za16_m (z0, p0, 1, w0))
+
+/*
+** read_za16_s16_1_w0p7_tied:
+** mov (w1[2-5]), w0
+** mova z0\.h, p0/m, za1v\.h\[\1, 7\]
+** ret
+*/
+TEST_READ_ZA (read_za16_s16_1_w0p7_tied, svint16_t,
+ z0 = svread_ver_za16_s16_m (z0, p0, 1, w0 + 7),
+ z0 = svread_ver_za16_m (z0, p0, 1, w0 + 7))
+
+/*
+** read_za16_s16_0_w0_untied:
+** (
+** mov (w1[2-5]), w0
+** mov z0\.d, z1\.d
+** mova z0\.h, p0/m, za0v\.h\[\1, 0\]
+** |
+** mov z0\.d, z1\.d
+** mov (w1[2-5]), w0
+** mova z0\.h, p0/m, za0v\.h\[\2, 0\]
+** |
+** mov (w1[2-5]), w0
+** mova z1\.h, p0/m, za0v\.h\[\3, 0\]
+** mov z0\.d, z1\.d
+** )
+** ret
+*/
+TEST_READ_ZA (read_za16_s16_0_w0_untied, svint16_t,
+ z0 = svread_ver_za16_s16_m (z1, p0, 0, w0),
+ z0 = svread_ver_za16_m (z1, p0, 0, w0))
+
+/*
+** read_za16_u16_0_w0_tied:
+** mov (w1[2-5]), w0
+** mova z0\.h, p0/m, za0v\.h\[\1, 0\]
+** ret
+*/
+TEST_READ_ZA (read_za16_u16_0_w0_tied, svuint16_t,
+ z0 = svread_ver_za16_u16_m (z0, p0, 0, w0),
+ z0 = svread_ver_za16_m (z0, p0, 0, w0))
+
+/*
+** read_za16_u16_0_w0_untied:
+** (
+** mov (w1[2-5]), w0
+** mov z0\.d, z1\.d
+** mova z0\.h, p0/m, za0v\.h\[\1, 0\]
+** |
+** mov z0\.d, z1\.d
+** mov (w1[2-5]), w0
+** mova z0\.h, p0/m, za0v\.h\[\2, 0\]
+** |
+** mov (w1[2-5]), w0
+** mova z1\.h, p0/m, za0v\.h\[\3, 0\]
+** mov z0\.d, z1\.d
+** )
+** ret
+*/
+TEST_READ_ZA (read_za16_u16_0_w0_untied, svuint16_t,
+ z0 = svread_ver_za16_u16_m (z1, p0, 0, w0),
+ z0 = svread_ver_za16_m (z1, p0, 0, w0))
+
+/*
+** read_za16_f16_0_w0_tied:
+** mov (w1[2-5]), w0
+** mova z0\.h, p0/m, za0v\.h\[\1, 0\]
+** ret
+*/
+TEST_READ_ZA (read_za16_f16_0_w0_tied, svfloat16_t,
+ z0 = svread_ver_za16_f16_m (z0, p0, 0, w0),
+ z0 = svread_ver_za16_m (z0, p0, 0, w0))
+
+/*
+** read_za16_f16_0_w0_untied:
+** (
+** mov (w1[2-5]), w0
+** mov z0\.d, z1\.d
+** mova z0\.h, p0/m, za0v\.h\[\1, 0\]
+** |
+** mov z0\.d, z1\.d
+** mov (w1[2-5]), w0
+** mova z0\.h, p0/m, za0v\.h\[\2, 0\]
+** |
+** mov (w1[2-5]), w0
+** mova z1\.h, p0/m, za0v\.h\[\3, 0\]
+** mov z0\.d, z1\.d
+** )
+** ret
+*/
+TEST_READ_ZA (read_za16_f16_0_w0_untied, svfloat16_t,
+ z0 = svread_ver_za16_f16_m (z1, p0, 0, w0),
+ z0 = svread_ver_za16_m (z1, p0, 0, w0))
+
+/*
+** read_za16_bf16_0_w0_tied:
+** mov (w1[2-5]), w0
+** mova z0\.h, p0/m, za0v\.h\[\1, 0\]
+** ret
+*/
+TEST_READ_ZA (read_za16_bf16_0_w0_tied, svbfloat16_t,
+ z0 = svread_ver_za16_bf16_m (z0, p0, 0, w0),
+ z0 = svread_ver_za16_m (z0, p0, 0, w0))
+
+/*
+** read_za16_bf16_0_w0_untied:
+** (
+** mov (w1[2-5]), w0
+** mov z0\.d, z1\.d
+** mova z0\.h, p0/m, za0v\.h\[\1, 0\]
+** |
+** mov z0\.d, z1\.d
+** mov (w1[2-5]), w0
+** mova z0\.h, p0/m, za0v\.h\[\2, 0\]
+** |
+** mov (w1[2-5]), w0
+** mova z1\.h, p0/m, za0v\.h\[\3, 0\]
+** mov z0\.d, z1\.d
+** )
+** ret
+*/
+TEST_READ_ZA (read_za16_bf16_0_w0_untied, svbfloat16_t,
+ z0 = svread_ver_za16_bf16_m (z1, p0, 0, w0),
+ z0 = svread_ver_za16_m (z1, p0, 0, w0))
--- /dev/null
+/* { dg-final { check-function-bodies "**" "" "-DCHECK_ASM" } } */
+
+#include "test_sme_acle.h"
+
+/*
+** read_za32_s32_0_0_tied:
+** mov (w1[2-5]), (?:wzr|#?0)
+** mova z0\.s, p0/m, za0v\.s\[\1, 0\]
+** ret
+*/
+TEST_READ_ZA (read_za32_s32_0_0_tied, svint32_t,
+ z0 = svread_ver_za32_s32_m (z0, p0, 0, 0),
+ z0 = svread_ver_za32_m (z0, p0, 0, 0))
+
+/*
+** read_za32_s32_0_1_tied:
+** mov (w1[2-5]), #?1
+** mova z0\.s, p0/m, za0v\.s\[\1, 0\]
+** ret
+*/
+TEST_READ_ZA (read_za32_s32_0_1_tied, svint32_t,
+ z0 = svread_ver_za32_s32_m (z0, p0, 0, 1),
+ z0 = svread_ver_za32_m (z0, p0, 0, 1))
+
+/*
+** read_za32_s32_0_w0_tied:
+** mov (w1[2-5]), w0
+** mova z0\.s, p0/m, za0v\.s\[\1, 0\]
+** ret
+*/
+TEST_READ_ZA (read_za32_s32_0_w0_tied, svint32_t,
+ z0 = svread_ver_za32_s32_m (z0, p0, 0, w0),
+ z0 = svread_ver_za32_m (z0, p0, 0, w0))
+
+/*
+** read_za32_s32_0_w0p1_tied:
+** mov (w1[2-5]), w0
+** mova z0\.s, p0/m, za0v\.s\[\1, 1\]
+** ret
+*/
+TEST_READ_ZA (read_za32_s32_0_w0p1_tied, svint32_t,
+ z0 = svread_ver_za32_s32_m (z0, p0, 0, w0 + 1),
+ z0 = svread_ver_za32_m (z0, p0, 0, w0 + 1))
+
+/*
+** read_za32_s32_0_w0p3_tied:
+** mov (w1[2-5]), w0
+** mova z0\.s, p0/m, za0v\.s\[\1, 3\]
+** ret
+*/
+TEST_READ_ZA (read_za32_s32_0_w0p3_tied, svint32_t,
+ z0 = svread_ver_za32_s32_m (z0, p0, 0, w0 + 3),
+ z0 = svread_ver_za32_m (z0, p0, 0, w0 + 3))
+
+/*
+** read_za32_s32_0_w0p4_tied:
+** add (w1[2-5]), w0, #?4
+** mova z0\.s, p0/m, za0v\.s\[\1, 0\]
+** ret
+*/
+TEST_READ_ZA (read_za32_s32_0_w0p4_tied, svint32_t,
+ z0 = svread_ver_za32_s32_m (z0, p0, 0, w0 + 4),
+ z0 = svread_ver_za32_m (z0, p0, 0, w0 + 4))
+
+/*
+** read_za32_s32_0_w0m1_tied:
+** sub (w1[2-5]), w0, #?1
+** mova z0\.s, p0/m, za0v\.s\[\1, 0\]
+** ret
+*/
+TEST_READ_ZA (read_za32_s32_0_w0m1_tied, svint32_t,
+ z0 = svread_ver_za32_s32_m (z0, p0, 0, w0 - 1),
+ z0 = svread_ver_za32_m (z0, p0, 0, w0 - 1))
+
+/*
+** read_za32_s32_1_w0_tied:
+** mov (w1[2-5]), w0
+** mova z0\.s, p0/m, za1v\.s\[\1, 0\]
+** ret
+*/
+TEST_READ_ZA (read_za32_s32_1_w0_tied, svint32_t,
+ z0 = svread_ver_za32_s32_m (z0, p0, 1, w0),
+ z0 = svread_ver_za32_m (z0, p0, 1, w0))
+
+/*
+** read_za32_s32_1_w0p3_tied:
+** mov (w1[2-5]), w0
+** mova z0\.s, p0/m, za1v\.s\[\1, 3\]
+** ret
+*/
+TEST_READ_ZA (read_za32_s32_1_w0p3_tied, svint32_t,
+ z0 = svread_ver_za32_s32_m (z0, p0, 1, w0 + 3),
+ z0 = svread_ver_za32_m (z0, p0, 1, w0 + 3))
+
+/*
+** read_za32_s32_3_w0_tied:
+** mov (w1[2-5]), w0
+** mova z0\.s, p0/m, za3v\.s\[\1, 0\]
+** ret
+*/
+TEST_READ_ZA (read_za32_s32_3_w0_tied, svint32_t,
+ z0 = svread_ver_za32_s32_m (z0, p0, 3, w0),
+ z0 = svread_ver_za32_m (z0, p0, 3, w0))
+
+/*
+** read_za32_s32_3_w0p3_tied:
+** mov (w1[2-5]), w0
+** mova z0\.s, p0/m, za3v\.s\[\1, 3\]
+** ret
+*/
+TEST_READ_ZA (read_za32_s32_3_w0p3_tied, svint32_t,
+ z0 = svread_ver_za32_s32_m (z0, p0, 3, w0 + 3),
+ z0 = svread_ver_za32_m (z0, p0, 3, w0 + 3))
+
+/*
+** read_za32_s32_0_w0_untied:
+** (
+** mov (w1[2-5]), w0
+** mov z0\.d, z1\.d
+** mova z0\.s, p0/m, za0v\.s\[\1, 0\]
+** |
+** mov z0\.d, z1\.d
+** mov (w1[2-5]), w0
+** mova z0\.s, p0/m, za0v\.s\[\2, 0\]
+** |
+** mov (w1[2-5]), w0
+** mova z1\.s, p0/m, za0v\.s\[\3, 0\]
+** mov z0\.d, z1\.d
+** )
+** ret
+*/
+TEST_READ_ZA (read_za32_s32_0_w0_untied, svint32_t,
+ z0 = svread_ver_za32_s32_m (z1, p0, 0, w0),
+ z0 = svread_ver_za32_m (z1, p0, 0, w0))
+
+/*
+** read_za32_u32_0_w0_tied:
+** mov (w1[2-5]), w0
+** mova z0\.s, p0/m, za0v\.s\[\1, 0\]
+** ret
+*/
+TEST_READ_ZA (read_za32_u32_0_w0_tied, svuint32_t,
+ z0 = svread_ver_za32_u32_m (z0, p0, 0, w0),
+ z0 = svread_ver_za32_m (z0, p0, 0, w0))
+
+/*
+** read_za32_u32_0_w0_untied:
+** (
+** mov (w1[2-5]), w0
+** mov z0\.d, z1\.d
+** mova z0\.s, p0/m, za0v\.s\[\1, 0\]
+** |
+** mov z0\.d, z1\.d
+** mov (w1[2-5]), w0
+** mova z0\.s, p0/m, za0v\.s\[\2, 0\]
+** |
+** mov (w1[2-5]), w0
+** mova z1\.s, p0/m, za0v\.s\[\3, 0\]
+** mov z0\.d, z1\.d
+** )
+** ret
+*/
+TEST_READ_ZA (read_za32_u32_0_w0_untied, svuint32_t,
+ z0 = svread_ver_za32_u32_m (z1, p0, 0, w0),
+ z0 = svread_ver_za32_m (z1, p0, 0, w0))
+
+/*
+** read_za32_f32_0_w0_tied:
+** mov (w1[2-5]), w0
+** mova z0\.s, p0/m, za0v\.s\[\1, 0\]
+** ret
+*/
+TEST_READ_ZA (read_za32_f32_0_w0_tied, svfloat32_t,
+ z0 = svread_ver_za32_f32_m (z0, p0, 0, w0),
+ z0 = svread_ver_za32_m (z0, p0, 0, w0))
+
+/*
+** read_za32_f32_0_w0_untied:
+** (
+** mov (w1[2-5]), w0
+** mov z0\.d, z1\.d
+** mova z0\.s, p0/m, za0v\.s\[\1, 0\]
+** |
+** mov z0\.d, z1\.d
+** mov (w1[2-5]), w0
+** mova z0\.s, p0/m, za0v\.s\[\2, 0\]
+** |
+** mov (w1[2-5]), w0
+** mova z1\.s, p0/m, za0v\.s\[\3, 0\]
+** mov z0\.d, z1\.d
+** )
+** ret
+*/
+TEST_READ_ZA (read_za32_f32_0_w0_untied, svfloat32_t,
+ z0 = svread_ver_za32_f32_m (z1, p0, 0, w0),
+ z0 = svread_ver_za32_m (z1, p0, 0, w0))
--- /dev/null
+/* { dg-final { check-function-bodies "**" "" "-DCHECK_ASM" } } */
+
+#include "test_sme_acle.h"
+
+/*
+** read_za64_s64_0_0_tied:
+** mov (w1[2-5]), (?:wzr|#?0)
+** mova z0\.d, p0/m, za0v\.d\[\1, 0\]
+** ret
+*/
+TEST_READ_ZA (read_za64_s64_0_0_tied, svint64_t,
+ z0 = svread_ver_za64_s64_m (z0, p0, 0, 0),
+ z0 = svread_ver_za64_m (z0, p0, 0, 0))
+
+/*
+** read_za64_s64_0_1_tied:
+** mov (w1[2-5]), #?1
+** mova z0\.d, p0/m, za0v\.d\[\1, 0\]
+** ret
+*/
+TEST_READ_ZA (read_za64_s64_0_1_tied, svint64_t,
+ z0 = svread_ver_za64_s64_m (z0, p0, 0, 1),
+ z0 = svread_ver_za64_m (z0, p0, 0, 1))
+
+/*
+** read_za64_s64_0_w0_tied:
+** mov (w1[2-5]), w0
+** mova z0\.d, p0/m, za0v\.d\[\1, 0\]
+** ret
+*/
+TEST_READ_ZA (read_za64_s64_0_w0_tied, svint64_t,
+ z0 = svread_ver_za64_s64_m (z0, p0, 0, w0),
+ z0 = svread_ver_za64_m (z0, p0, 0, w0))
+
+/*
+** read_za64_s64_0_w0p1_tied:
+** mov (w1[2-5]), w0
+** mova z0\.d, p0/m, za0v\.d\[\1, 1\]
+** ret
+*/
+TEST_READ_ZA (read_za64_s64_0_w0p1_tied, svint64_t,
+ z0 = svread_ver_za64_s64_m (z0, p0, 0, w0 + 1),
+ z0 = svread_ver_za64_m (z0, p0, 0, w0 + 1))
+
+/*
+** read_za64_s64_0_w0p2_tied:
+** add (w1[2-5]), w0, #?2
+** mova z0\.d, p0/m, za0v\.d\[\1, 0\]
+** ret
+*/
+TEST_READ_ZA (read_za64_s64_0_w0p2_tied, svint64_t,
+ z0 = svread_ver_za64_s64_m (z0, p0, 0, w0 + 2),
+ z0 = svread_ver_za64_m (z0, p0, 0, w0 + 2))
+
+/*
+** read_za64_s64_0_w0m1_tied:
+** sub (w1[2-5]), w0, #?1
+** mova z0\.d, p0/m, za0v\.d\[\1, 0\]
+** ret
+*/
+TEST_READ_ZA (read_za64_s64_0_w0m1_tied, svint64_t,
+ z0 = svread_ver_za64_s64_m (z0, p0, 0, w0 - 1),
+ z0 = svread_ver_za64_m (z0, p0, 0, w0 - 1))
+
+/*
+** read_za64_s64_1_w0_tied:
+** mov (w1[2-5]), w0
+** mova z0\.d, p0/m, za1v\.d\[\1, 0\]
+** ret
+*/
+TEST_READ_ZA (read_za64_s64_1_w0_tied, svint64_t,
+ z0 = svread_ver_za64_s64_m (z0, p0, 1, w0),
+ z0 = svread_ver_za64_m (z0, p0, 1, w0))
+
+/*
+** read_za64_s64_1_w0p1_tied:
+** mov (w1[2-5]), w0
+** mova z0\.d, p0/m, za1v\.d\[\1, 1\]
+** ret
+*/
+TEST_READ_ZA (read_za64_s64_1_w0p1_tied, svint64_t,
+ z0 = svread_ver_za64_s64_m (z0, p0, 1, w0 + 1),
+ z0 = svread_ver_za64_m (z0, p0, 1, w0 + 1))
+
+/*
+** read_za64_s64_7_w0_tied:
+** mov (w1[2-5]), w0
+** mova z0\.d, p0/m, za7v\.d\[\1, 0\]
+** ret
+*/
+TEST_READ_ZA (read_za64_s64_7_w0_tied, svint64_t,
+ z0 = svread_ver_za64_s64_m (z0, p0, 7, w0),
+ z0 = svread_ver_za64_m (z0, p0, 7, w0))
+
+/*
+** read_za64_s64_7_w0p1_tied:
+** mov (w1[2-5]), w0
+** mova z0\.d, p0/m, za7v\.d\[\1, 1\]
+** ret
+*/
+TEST_READ_ZA (read_za64_s64_7_w0p1_tied, svint64_t,
+ z0 = svread_ver_za64_s64_m (z0, p0, 7, w0 + 1),
+ z0 = svread_ver_za64_m (z0, p0, 7, w0 + 1))
+
+/*
+** read_za64_s64_0_w0_untied:
+** (
+** mov (w1[2-5]), w0
+** mov z0\.d, z1\.d
+** mova z0\.d, p0/m, za0v\.d\[\1, 0\]
+** |
+** mov z0\.d, z1\.d
+** mov (w1[2-5]), w0
+** mova z0\.d, p0/m, za0v\.d\[\2, 0\]
+** |
+** mov (w1[2-5]), w0
+** mova z1\.d, p0/m, za0v\.d\[\3, 0\]
+** mov z0\.d, z1\.d
+** )
+** ret
+*/
+TEST_READ_ZA (read_za64_s64_0_w0_untied, svint64_t,
+ z0 = svread_ver_za64_s64_m (z1, p0, 0, w0),
+ z0 = svread_ver_za64_m (z1, p0, 0, w0))
+
+/*
+** read_za64_u64_0_w0_tied:
+** mov (w1[2-5]), w0
+** mova z0\.d, p0/m, za0v\.d\[\1, 0\]
+** ret
+*/
+TEST_READ_ZA (read_za64_u64_0_w0_tied, svuint64_t,
+ z0 = svread_ver_za64_u64_m (z0, p0, 0, w0),
+ z0 = svread_ver_za64_m (z0, p0, 0, w0))
+
+/*
+** read_za64_u64_0_w0_untied:
+** (
+** mov (w1[2-5]), w0
+** mov z0\.d, z1\.d
+** mova z0\.d, p0/m, za0v\.d\[\1, 0\]
+** |
+** mov z0\.d, z1\.d
+** mov (w1[2-5]), w0
+** mova z0\.d, p0/m, za0v\.d\[\2, 0\]
+** |
+** mov (w1[2-5]), w0
+** mova z1\.d, p0/m, za0v\.d\[\3, 0\]
+** mov z0\.d, z1\.d
+** )
+** ret
+*/
+TEST_READ_ZA (read_za64_u64_0_w0_untied, svuint64_t,
+ z0 = svread_ver_za64_u64_m (z1, p0, 0, w0),
+ z0 = svread_ver_za64_m (z1, p0, 0, w0))
+
+/*
+** read_za64_f64_0_w0_tied:
+** mov (w1[2-5]), w0
+** mova z0\.d, p0/m, za0v\.d\[\1, 0\]
+** ret
+*/
+TEST_READ_ZA (read_za64_f64_0_w0_tied, svfloat64_t,
+ z0 = svread_ver_za64_f64_m (z0, p0, 0, w0),
+ z0 = svread_ver_za64_m (z0, p0, 0, w0))
+
+/*
+** read_za64_f64_0_w0_untied:
+** (
+** mov (w1[2-5]), w0
+** mov z0\.d, z1\.d
+** mova z0\.d, p0/m, za0v\.d\[\1, 0\]
+** |
+** mov z0\.d, z1\.d
+** mov (w1[2-5]), w0
+** mova z0\.d, p0/m, za0v\.d\[\2, 0\]
+** |
+** mov (w1[2-5]), w0
+** mova z1\.d, p0/m, za0v\.d\[\3, 0\]
+** mov z0\.d, z1\.d
+** )
+** ret
+*/
+TEST_READ_ZA (read_za64_f64_0_w0_untied, svfloat64_t,
+ z0 = svread_ver_za64_f64_m (z1, p0, 0, w0),
+ z0 = svread_ver_za64_m (z1, p0, 0, w0))
--- /dev/null
+/* { dg-final { check-function-bodies "**" "" "-DCHECK_ASM" } } */
+
+#include "test_sme_acle.h"
+
+/*
+** read_za8_s8_0_0_tied:
+** mov (w1[2-5]), (?:wzr|#?0)
+** mova z0\.b, p0/m, za0v\.b\[\1, 0\]
+** ret
+*/
+TEST_READ_ZA (read_za8_s8_0_0_tied, svint8_t,
+ z0 = svread_ver_za8_s8_m (z0, p0, 0, 0),
+ z0 = svread_ver_za8_m (z0, p0, 0, 0))
+
+/*
+** read_za8_s8_0_1_tied:
+** mov (w1[2-5]), #?1
+** mova z0\.b, p0/m, za0v\.b\[\1, 0\]
+** ret
+*/
+TEST_READ_ZA (read_za8_s8_0_1_tied, svint8_t,
+ z0 = svread_ver_za8_s8_m (z0, p0, 0, 1),
+ z0 = svread_ver_za8_m (z0, p0, 0, 1))
+
+/*
+** read_za8_s8_0_w0_tied:
+** mov (w1[2-5]), w0
+** mova z0\.b, p0/m, za0v\.b\[\1, 0\]
+** ret
+*/
+TEST_READ_ZA (read_za8_s8_0_w0_tied, svint8_t,
+ z0 = svread_ver_za8_s8_m (z0, p0, 0, w0),
+ z0 = svread_ver_za8_m (z0, p0, 0, w0))
+
+/*
+** read_za8_s8_0_w0p1_tied:
+** mov (w1[2-5]), w0
+** mova z0\.b, p0/m, za0v\.b\[\1, 1\]
+** ret
+*/
+TEST_READ_ZA (read_za8_s8_0_w0p1_tied, svint8_t,
+ z0 = svread_ver_za8_s8_m (z0, p0, 0, w0 + 1),
+ z0 = svread_ver_za8_m (z0, p0, 0, w0 + 1))
+
+/*
+** read_za8_s8_0_w0p15_tied:
+** mov (w1[2-5]), w0
+** mova z0\.b, p0/m, za0v\.b\[\1, 15\]
+** ret
+*/
+TEST_READ_ZA (read_za8_s8_0_w0p15_tied, svint8_t,
+ z0 = svread_ver_za8_s8_m (z0, p0, 0, w0 + 15),
+ z0 = svread_ver_za8_m (z0, p0, 0, w0 + 15))
+
+/*
+** read_za8_s8_0_w0p16_tied:
+** add (w1[2-5]), w0, #?16
+** mova z0\.b, p0/m, za0v\.b\[\1, 0\]
+** ret
+*/
+TEST_READ_ZA (read_za8_s8_0_w0p16_tied, svint8_t,
+ z0 = svread_ver_za8_s8_m (z0, p0, 0, w0 + 16),
+ z0 = svread_ver_za8_m (z0, p0, 0, w0 + 16))
+
+/*
+** read_za8_s8_0_w0m1_tied:
+** sub (w1[2-5]), w0, #?1
+** mova z0\.b, p0/m, za0v\.b\[\1, 0\]
+** ret
+*/
+TEST_READ_ZA (read_za8_s8_0_w0m1_tied, svint8_t,
+ z0 = svread_ver_za8_s8_m (z0, p0, 0, w0 - 1),
+ z0 = svread_ver_za8_m (z0, p0, 0, w0 - 1))
+
+/*
+** read_za8_s8_0_w0_untied:
+** (
+** mov (w1[2-5]), w0
+** mov z0\.d, z1\.d
+** mova z0\.b, p0/m, za0v\.b\[\1, 0\]
+** |
+** mov z0\.d, z1\.d
+** mov (w1[2-5]), w0
+** mova z0\.b, p0/m, za0v\.b\[\2, 0\]
+** |
+** mov (w1[2-5]), w0
+** mova z1\.b, p0/m, za0v\.b\[\3, 0\]
+** mov z0\.d, z1\.d
+** )
+** ret
+*/
+TEST_READ_ZA (read_za8_s8_0_w0_untied, svint8_t,
+ z0 = svread_ver_za8_s8_m (z1, p0, 0, w0),
+ z0 = svread_ver_za8_m (z1, p0, 0, w0))
+
+/*
+** read_za8_u8_0_w0_tied:
+** mov (w1[2-5]), w0
+** mova z0\.b, p0/m, za0v\.b\[\1, 0\]
+** ret
+*/
+TEST_READ_ZA (read_za8_u8_0_w0_tied, svuint8_t,
+ z0 = svread_ver_za8_u8_m (z0, p0, 0, w0),
+ z0 = svread_ver_za8_m (z0, p0, 0, w0))
+
+/*
+** read_za8_u8_0_w0_untied:
+** (
+** mov (w1[2-5]), w0
+** mov z0\.d, z1\.d
+** mova z0\.b, p0/m, za0v\.b\[\1, 0\]
+** |
+** mov z0\.d, z1\.d
+** mov (w1[2-5]), w0
+** mova z0\.b, p0/m, za0v\.b\[\2, 0\]
+** |
+** mov (w1[2-5]), w0
+** mova z1\.b, p0/m, za0v\.b\[\3, 0\]
+** mov z0\.d, z1\.d
+** )
+** ret
+*/
+TEST_READ_ZA (read_za8_u8_0_w0_untied, svuint8_t,
+ z0 = svread_ver_za8_u8_m (z1, p0, 0, w0),
+ z0 = svread_ver_za8_m (z1, p0, 0, w0))
--- /dev/null
+/* { dg-final { check-function-bodies "**" "" "-DCHECK_ASM" } } */
+
+#include "test_sme_acle.h"
+
+/*
+** st1_vnum_za128_0_0_0:
+** mov (w1[2-5]), (?:wzr|#?0)
+** st1q { za0h\.q\[\1, 0\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_vnum_za128_0_0_0,
+ svst1_hor_vnum_za128 (0, 0, p0, x1, 0),
+ svst1_hor_vnum_za128 (0, 0, p0, x1, 0))
+
+/*
+** st1_vnum_za128_7_1_0:
+** mov (w1[2-5]), #?1
+** st1q { za7h\.q\[\1, 0\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_vnum_za128_7_1_0,
+ svst1_hor_vnum_za128 (7, 1, p0, x1, 0),
+ svst1_hor_vnum_za128 (7, 1, p0, x1, 0))
+
+/*
+** st1_vnum_za128_11_1_5:
+** incb x1, all, mul #5
+** mov (w1[2-5]), #?6
+** st1q { za11h\.q\[\1, 0\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_vnum_za128_11_1_5,
+ svst1_hor_vnum_za128 (11, 1, p0, x1, 5),
+ svst1_hor_vnum_za128 (11, 1, p0, x1, 5))
+
+/*
+** st1_vnum_za128_3_w0_0:
+** mov (w1[2-5]), w0
+** st1q { za3h\.q\[\1, 0\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_vnum_za128_3_w0_0,
+ svst1_hor_vnum_za128 (3, w0, p0, x1, 0),
+ svst1_hor_vnum_za128 (3, w0, p0, x1, 0))
+
+/*
+** st1_vnum_za128_5_w0_0:
+** incb x1, all, mul #13
+** add (w1[2-5]), w0, #?13
+** st1q { za5h\.q\[\1, 0\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_vnum_za128_5_w0_0,
+ svst1_hor_vnum_za128 (5, w0, p0, x1, 13),
+ svst1_hor_vnum_za128 (5, w0, p0, x1, 13))
+
+/*
+** st1_vnum_za128_11_w0_0:
+** cntb (x[0-9]+)
+** madd (x[0-9]+), (?:\1, x2|x2, \1), x1
+** add (w1[2-5]), (?:w0, w2|w2, w0)
+** st1q { za11h\.q\[\3, 0\] }, p0, \[\2\]
+** ret
+*/
+TEST_STORE_ZA (st1_vnum_za128_11_w0_0,
+ svst1_hor_vnum_za128 (11, w0, p0, x1, x2),
+ svst1_hor_vnum_za128 (11, w0, p0, x1, x2))
+
+/*
+** st1_vnum_za128_15_w0p1_0:
+** add (w1[2-5]), w0, #?1
+** st1q { za15h\.q\[\1, 0\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_vnum_za128_15_w0p1_0,
+ svst1_hor_vnum_za128 (15, w0 + 1, p0, x1, 0),
+ svst1_hor_vnum_za128 (15, w0 + 1, p0, x1, 0))
--- /dev/null
+/* { dg-final { check-function-bodies "**" "" "-DCHECK_ASM" } } */
+
+#include "test_sme_acle.h"
+
+/*
+** st1_vnum_za16_1_0_1:
+** incb x1
+** mov (w1[2-5]), (?:wzr|#?0)
+** st1h { za1h\.h\[\1, 1\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_vnum_za16_1_0_1,
+ svst1_hor_vnum_za16 (1, 0, p0, x1, 1),
+ svst1_hor_vnum_za16 (1, 0, p0, x1, 1))
+
+/*
+** st1_vnum_za16_1_1_1:
+** incb x1
+** mov (w1[2-5]), #?1
+** st1h { za1h\.h\[\1, 1\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_vnum_za16_1_1_1,
+ svst1_hor_vnum_za16 (1, 1, p0, x1, 1),
+ svst1_hor_vnum_za16 (1, 1, p0, x1, 1))
+
+/*
+** st1_vnum_za16_0_0_8:
+** incb x1, all, mul #8
+** mov (w1[2-5]), #?8
+** st1h { za0h\.h\[\1, 0\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_vnum_za16_0_0_8,
+ svst1_hor_vnum_za16 (0, 0, p0, x1, 8),
+ svst1_hor_vnum_za16 (0, 0, p0, x1, 8))
+
+/*
+** st1_vnum_za16_0_1_8:
+** incb x1, all, mul #8
+** mov (w1[2-5]), #?9
+** st1h { za0h\.h\[\1, 0\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_vnum_za16_0_1_8,
+ svst1_hor_vnum_za16 (0, 1, p0, x1, 8),
+ svst1_hor_vnum_za16 (0, 1, p0, x1, 8))
+
+/*
+** st1_vnum_za16_0_w0_0:
+** mov (w1[2-5]), w0
+** st1h { za0h\.h\[\1, 0\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_vnum_za16_0_w0_0,
+ svst1_hor_vnum_za16 (0, w0, p0, x1, 0),
+ svst1_hor_vnum_za16 (0, w0, p0, x1, 0))
+
+/*
+** st1_vnum_za16_0_w0_1:
+** incb x1
+** mov (w1[2-5]), w0
+** st1h { za0h\.h\[\1, 1\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_vnum_za16_0_w0_1,
+ svst1_hor_vnum_za16 (0, w0, p0, x1, 1),
+ svst1_hor_vnum_za16 (0, w0, p0, x1, 1))
+
+/*
+** st1_vnum_za16_0_w0_7:
+** incb x1, all, mul #7
+** mov (w1[2-5]), w0
+** st1h { za0h\.h\[\1, 7\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_vnum_za16_0_w0_7,
+ svst1_hor_vnum_za16 (0, w0, p0, x1, 7),
+ svst1_hor_vnum_za16 (0, w0, p0, x1, 7))
+
+/*
+** st1_vnum_za16_1_w0_8:
+** incb x1, all, mul #8
+** add (w1[2-5]), w0, #?8
+** st1h { za1h\.h\[\1, 0\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_vnum_za16_1_w0_8,
+ svst1_hor_vnum_za16 (1, w0, p0, x1, 8),
+ svst1_hor_vnum_za16 (1, w0, p0, x1, 8))
+
+/*
+** st1_vnum_za16_1_w0_13:
+** incb x1, all, mul #13
+** add (w1[2-5]), w0, #?13
+** st1h { za1h\.h\[\1, 0\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_vnum_za16_1_w0_13,
+ svst1_hor_vnum_za16 (1, w0, p0, x1, 13),
+ svst1_hor_vnum_za16 (1, w0, p0, x1, 13))
+
+/*
+** st1_vnum_za16_0_w0_x2:
+** cntb (x[0-9]+)
+** madd (x[0-9]+), (?:\1, x2|x2, \1), x1
+** add (w1[2-5]), (?:w0, w2|w2, w0)
+** st1h { za0h\.h\[\3, 0\] }, p0, \[\2\]
+** ret
+*/
+TEST_STORE_ZA (st1_vnum_za16_0_w0_x2,
+ svst1_hor_vnum_za16 (0, w0, p0, x1, x2),
+ svst1_hor_vnum_za16 (0, w0, p0, x1, x2))
+
+/*
+** st1_vnum_za16_1_w0p1_0:
+** mov (w1[2-5]), w0
+** st1h { za1h\.h\[\1, 1\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_vnum_za16_1_w0p1_0,
+ svst1_hor_vnum_za16 (1, w0 + 1, p0, x1, 0),
+ svst1_hor_vnum_za16 (1, w0 + 1, p0, x1, 0))
--- /dev/null
+/* { dg-final { check-function-bodies "**" "" "-DCHECK_ASM" } } */
+
+#include "test_sme_acle.h"
+
+/*
+** st1_vnum_za32_3_0_1:
+** incb x1
+** mov (w1[2-5]), (?:wzr|#?0)
+** st1w { za3h\.s\[\1, 1\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_vnum_za32_3_0_1,
+ svst1_hor_vnum_za32 (3, 0, p0, x1, 1),
+ svst1_hor_vnum_za32 (3, 0, p0, x1, 1))
+
+/*
+** st1_vnum_za32_2_1_1:
+** incb x1
+** mov (w1[2-5]), #?1
+** st1w { za2h\.s\[\1, 1\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_vnum_za32_2_1_1,
+ svst1_hor_vnum_za32 (2, 1, p0, x1, 1),
+ svst1_hor_vnum_za32 (2, 1, p0, x1, 1))
+
+/*
+** st1_vnum_za32_0_0_4:
+** incb x1, all, mul #4
+** mov (w1[2-5]), #?4
+** st1w { za0h\.s\[\1, 0\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_vnum_za32_0_0_4,
+ svst1_hor_vnum_za32 (0, 0, p0, x1, 4),
+ svst1_hor_vnum_za32 (0, 0, p0, x1, 4))
+
+/*
+** st1_vnum_za32_2_1_4:
+** incb x1, all, mul #4
+** mov (w1[2-5]), #?5
+** st1w { za2h\.s\[\1, 0\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_vnum_za32_2_1_4,
+ svst1_hor_vnum_za32 (2, 1, p0, x1, 4),
+ svst1_hor_vnum_za32 (2, 1, p0, x1, 4))
+
+/*
+** st1_vnum_za32_0_w0_0:
+** mov (w1[2-5]), w0
+** st1w { za0h\.s\[\1, 0\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_vnum_za32_0_w0_0,
+ svst1_hor_vnum_za32 (0, w0, p0, x1, 0),
+ svst1_hor_vnum_za32 (0, w0, p0, x1, 0))
+
+/*
+** st1_vnum_za32_0_w0_1:
+** incb x1
+** mov (w1[2-5]), w0
+** st1w { za0h\.s\[\1, 1\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_vnum_za32_0_w0_1,
+ svst1_hor_vnum_za32 (0, w0, p0, x1, 1),
+ svst1_hor_vnum_za32 (0, w0, p0, x1, 1))
+
+/*
+** st1_vnum_za32_0_w0_3:
+** incb x1, all, mul #3
+** mov (w1[2-5]), w0
+** st1w { za0h\.s\[\1, 3\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_vnum_za32_0_w0_3,
+ svst1_hor_vnum_za32 (0, w0, p0, x1, 3),
+ svst1_hor_vnum_za32 (0, w0, p0, x1, 3))
+
+/*
+** st1_vnum_za32_1_w0_4:
+** incb x1, all, mul #4
+** add (w1[2-5]), w0, #?4
+** st1w { za1h\.s\[\1, 0\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_vnum_za32_1_w0_4,
+ svst1_hor_vnum_za32 (1, w0, p0, x1, 4),
+ svst1_hor_vnum_za32 (1, w0, p0, x1, 4))
+
+/*
+** st1_vnum_za32_3_w0_13:
+** incb x1, all, mul #13
+** add (w1[2-5]), w0, #?13
+** st1w { za3h\.s\[\1, 0\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_vnum_za32_3_w0_13,
+ svst1_hor_vnum_za32 (3, w0, p0, x1, 13),
+ svst1_hor_vnum_za32 (3, w0, p0, x1, 13))
+
+/*
+** st1_vnum_za32_0_w0_x2:
+** cntb (x[0-9]+)
+** madd (x[0-9]+), (?:\1, x2|x2, \1), x1
+** add (w1[2-5]), (?:w0, w2|w2, w0)
+** st1w { za0h\.s\[\3, 0\] }, p0, \[\2\]
+** ret
+*/
+TEST_STORE_ZA (st1_vnum_za32_0_w0_x2,
+ svst1_hor_vnum_za32 (0, w0, p0, x1, x2),
+ svst1_hor_vnum_za32 (0, w0, p0, x1, x2))
+
+/*
+** st1_vnum_za32_1_w0p1_0:
+** mov (w1[2-5]), w0
+** st1w { za1h\.s\[\1, 1\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_vnum_za32_1_w0p1_0,
+ svst1_hor_vnum_za32 (1, w0 + 1, p0, x1, 0),
+ svst1_hor_vnum_za32 (1, w0 + 1, p0, x1, 0))
--- /dev/null
+/* { dg-final { check-function-bodies "**" "" "-DCHECK_ASM" } } */
+
+#include "test_sme_acle.h"
+
+/*
+** st1_vnum_za64_3_0_1:
+** incb x1
+** mov (w1[2-5]), (?:wzr|#?0)
+** st1d { za3h\.d\[\1, 1\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_vnum_za64_3_0_1,
+ svst1_hor_vnum_za64 (3, 0, p0, x1, 1),
+ svst1_hor_vnum_za64 (3, 0, p0, x1, 1))
+
+/*
+** st1_vnum_za64_7_1_1:
+** incb x1
+** mov (w1[2-5]), #?1
+** st1d { za7h\.d\[\1, 1\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_vnum_za64_7_1_1,
+ svst1_hor_vnum_za64 (7, 1, p0, x1, 1),
+ svst1_hor_vnum_za64 (7, 1, p0, x1, 1))
+
+/*
+** st1_vnum_za64_0_0_2:
+** incb x1, all, mul #2
+** mov (w1[2-5]), #?2
+** st1d { za0h\.d\[\1, 0\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_vnum_za64_0_0_2,
+ svst1_hor_vnum_za64 (0, 0, p0, x1, 2),
+ svst1_hor_vnum_za64 (0, 0, p0, x1, 2))
+
+/*
+** st1_vnum_za64_5_1_2:
+** incb x1, all, mul #2
+** mov (w1[2-5]), #?3
+** st1d { za5h\.d\[\1, 0\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_vnum_za64_5_1_2,
+ svst1_hor_vnum_za64 (5, 1, p0, x1, 2),
+ svst1_hor_vnum_za64 (5, 1, p0, x1, 2))
+
+/*
+** st1_vnum_za64_0_w0_0:
+** mov (w1[2-5]), w0
+** st1d { za0h\.d\[\1, 0\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_vnum_za64_0_w0_0,
+ svst1_hor_vnum_za64 (0, w0, p0, x1, 0),
+ svst1_hor_vnum_za64 (0, w0, p0, x1, 0))
+
+/*
+** st1_vnum_za64_0_w0_1:
+** incb x1
+** mov (w1[2-5]), w0
+** st1d { za0h\.d\[\1, 1\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_vnum_za64_0_w0_1,
+ svst1_hor_vnum_za64 (0, w0, p0, x1, 1),
+ svst1_hor_vnum_za64 (0, w0, p0, x1, 1))
+
+/*
+** st1_vnum_za64_6_w0_2:
+** incb x1, all, mul #2
+** add (w1[2-5]), w0, #?2
+** st1d { za6h\.d\[\1, 0\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_vnum_za64_6_w0_2,
+ svst1_hor_vnum_za64 (6, w0, p0, x1, 2),
+ svst1_hor_vnum_za64 (6, w0, p0, x1, 2))
+
+/*
+** st1_vnum_za64_2_w0_13:
+** incb x1, all, mul #13
+** add (w1[2-5]), w0, #?13
+** st1d { za2h\.d\[\1, 0\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_vnum_za64_2_w0_13,
+ svst1_hor_vnum_za64 (2, w0, p0, x1, 13),
+ svst1_hor_vnum_za64 (2, w0, p0, x1, 13))
+
+/*
+** st1_vnum_za64_4_w0_x2:
+** cntb (x[0-9]+)
+** madd (x[0-9]+), (?:\1, x2|x2, \1), x1
+** add (w1[2-5]), (?:w0, w2|w2, w0)
+** st1d { za4h\.d\[\3, 0\] }, p0, \[\2\]
+** ret
+*/
+TEST_STORE_ZA (st1_vnum_za64_4_w0_x2,
+ svst1_hor_vnum_za64 (4, w0, p0, x1, x2),
+ svst1_hor_vnum_za64 (4, w0, p0, x1, x2))
+
+/*
+** st1_vnum_za64_1_w0p1_0:
+** mov (w1[2-5]), w0
+** st1d { za1h\.d\[\1, 1\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_vnum_za64_1_w0p1_0,
+ svst1_hor_vnum_za64 (1, w0 + 1, p0, x1, 0),
+ svst1_hor_vnum_za64 (1, w0 + 1, p0, x1, 0))
--- /dev/null
+/* { dg-final { check-function-bodies "**" "" "-DCHECK_ASM" } } */
+
+#include "test_sme_acle.h"
+
+/*
+** st1_vnum_za8_0_0_1:
+** incb x1
+** mov (w1[2-5]), (?:wzr|#?0)
+** st1b { za0h\.b\[\1, 1\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_vnum_za8_0_0_1,
+ svst1_hor_vnum_za8 (0, 0, p0, x1, 1),
+ svst1_hor_vnum_za8 (0, 0, p0, x1, 1))
+
+/*
+** st1_vnum_za8_0_1_1:
+** incb x1
+** mov (w1[2-5]), #?1
+** st1b { za0h\.b\[\1, 1\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_vnum_za8_0_1_1,
+ svst1_hor_vnum_za8 (0, 1, p0, x1, 1),
+ svst1_hor_vnum_za8 (0, 1, p0, x1, 1))
+
+/*
+** st1_vnum_za8_0_0_16:
+** incb x1, all, mul #16
+** mov (w1[2-5]), #?16
+** st1b { za0h\.b\[\1, 0\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_vnum_za8_0_0_16,
+ svst1_hor_vnum_za8 (0, 0, p0, x1, 16),
+ svst1_hor_vnum_za8 (0, 0, p0, x1, 16))
+
+/*
+** st1_vnum_za8_0_1_16:
+** incb x1, all, mul #16
+** mov (w1[2-5]), #?17
+** st1b { za0h\.b\[\1, 0\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_vnum_za8_0_1_16,
+ svst1_hor_vnum_za8 (0, 1, p0, x1, 16),
+ svst1_hor_vnum_za8 (0, 1, p0, x1, 16))
+
+/*
+** st1_vnum_za8_0_w0_0:
+** mov (w1[2-5]), w0
+** st1b { za0h\.b\[\1, 0\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_vnum_za8_0_w0_0,
+ svst1_hor_vnum_za8 (0, w0, p0, x1, 0),
+ svst1_hor_vnum_za8 (0, w0, p0, x1, 0))
+
+/*
+** st1_vnum_za8_0_w0_1:
+** incb x1
+** mov (w1[2-5]), w0
+** st1b { za0h\.b\[\1, 1\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_vnum_za8_0_w0_1,
+ svst1_hor_vnum_za8 (0, w0, p0, x1, 1),
+ svst1_hor_vnum_za8 (0, w0, p0, x1, 1))
+
+/*
+** st1_vnum_za8_0_w0_15:
+** incb x1, all, mul #15
+** mov (w1[2-5]), w0
+** st1b { za0h\.b\[\1, 15\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_vnum_za8_0_w0_15,
+ svst1_hor_vnum_za8 (0, w0, p0, x1, 15),
+ svst1_hor_vnum_za8 (0, w0, p0, x1, 15))
+
+/*
+** st1_vnum_za8_0_w0_16:
+** incb x1, all, mul #16
+** add (w1[2-5]), w0, #?16
+** st1b { za0h\.b\[\1, 0\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_vnum_za8_0_w0_16,
+ svst1_hor_vnum_za8 (0, w0, p0, x1, 16),
+ svst1_hor_vnum_za8 (0, w0, p0, x1, 16))
+
+/*
+** st1_vnum_za8_0_w0_x2:
+** cntb (x[0-9]+)
+** mul (x[0-9]+), (?:\1, x2|x2, \1)
+** add (w1[2-5]), (?:w0, w2|w2, w0)
+** st1b { za0h\.b\[\3, 0\] }, p0, \[x1, \2\]
+** ret
+*/
+TEST_STORE_ZA (st1_vnum_za8_0_w0_x2,
+ svst1_hor_vnum_za8 (0, w0, p0, x1, x2),
+ svst1_hor_vnum_za8 (0, w0, p0, x1, x2))
+
+/*
+** st1_vnum_za8_0_w0p1_0:
+** mov (w1[2-5]), w0
+** st1b { za0h\.b\[\1, 1\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_vnum_za8_0_w0p1_0,
+ svst1_hor_vnum_za8 (0, w0 + 1, p0, x1, 0),
+ svst1_hor_vnum_za8 (0, w0 + 1, p0, x1, 0))
--- /dev/null
+/* { dg-final { check-function-bodies "**" "" "-DCHECK_ASM" } } */
+
+#include "test_sme_acle.h"
+
+/*
+** st1_za128_0_0:
+** mov (w1[2-5]), (?:wzr|#?0)
+** st1q { za0h\.q\[\1, 0\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_za128_0_0,
+ svst1_hor_za128 (0, 0, p0, x1),
+ svst1_hor_za128 (0, 0, p0, x1))
+
+/*
+** st1_za128_0_1:
+** mov (w1[2-5]), #?1
+** st1q { za0h\.q\[\1, 0\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_za128_0_1,
+ svst1_hor_za128 (0, 1, p0, x1),
+ svst1_hor_za128 (0, 1, p0, x1))
+
+/*
+** st1_za128_0_w0:
+** mov (w1[2-5]), w0
+** st1q { za0h\.q\[\1, 0\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_za128_0_w0,
+ svst1_hor_za128 (0, w0, p0, x1),
+ svst1_hor_za128 (0, w0, p0, x1))
+
+/*
+** st1_za128_0_w0_p1:
+** add (w1[2-5]), w0, #?1
+** st1q { za0h\.q\[\1, 0\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_za128_0_w0_p1,
+ svst1_hor_za128 (0, w0 + 1, p0, x1),
+ svst1_hor_za128 (0, w0 + 1, p0, x1))
+
+/*
+** st1_za128_7_w0:
+** mov (w1[2-5]), w0
+** st1q { za7h\.q\[\1, 0\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_za128_7_w0,
+ svst1_hor_za128 (7, w0, p0, x1),
+ svst1_hor_za128 (7, w0, p0, x1))
+
+/*
+** st1_za128_13_w0:
+** mov (w1[2-5]), w0
+** st1q { za13h\.q\[\1, 0\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_za128_13_w0,
+ svst1_hor_za128 (13, w0, p0, x1),
+ svst1_hor_za128 (13, w0, p0, x1))
+
+/*
+** st1_za128_15_w0:
+** mov (w1[2-5]), w0
+** st1q { za15h\.q\[\1, 0\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_za128_15_w0,
+ svst1_hor_za128 (15, w0, p0, x1),
+ svst1_hor_za128 (15, w0, p0, x1))
+
+/*
+** st1_za128_9_w0_index:
+** mov (w1[2-5]), w0
+** st1q { za9h\.q\[\1, 0\] }, p0, \[x1, x2, lsl #?4\]
+** ret
+*/
+TEST_STORE_ZA (st1_za128_9_w0_index,
+ svst1_hor_za128 (9, w0, p0, x1 + x2 * 16),
+ svst1_hor_za128 (9, w0, p0, x1 + x2 * 16))
--- /dev/null
+/* { dg-final { check-function-bodies "**" "" "-DCHECK_ASM" } } */
+
+#include "test_sme_acle.h"
+
+/*
+** st1_za16_0_0:
+** mov (w1[2-5]), (?:wzr|#?0)
+** st1h { za0h\.h\[\1, 0\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_za16_0_0,
+ svst1_hor_za16 (0, 0, p0, x1),
+ svst1_hor_za16 (0, 0, p0, x1))
+
+/* It would also be OK (and perhaps better) to move 0 into a register
+ and use an offset of 7. */
+/*
+** st1_za16_0_7:
+** mov (w1[2-5]), #?7
+** st1h { za0h\.h\[\1, 0\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_za16_0_7,
+ svst1_hor_za16 (0, 7, p0, x1),
+ svst1_hor_za16 (0, 7, p0, x1))
+
+/*
+** st1_za16_0_8:
+** mov (w1[2-5]), #?8
+** st1h { za0h\.h\[\1, 0\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_za16_0_8,
+ svst1_hor_za16 (0, 8, p0, x1),
+ svst1_hor_za16 (0, 8, p0, x1))
+
+/*
+** st1_za16_0_w0:
+** mov (w1[2-5]), w0
+** st1h { za0h\.h\[\1, 0\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_za16_0_w0,
+ svst1_hor_za16 (0, w0, p0, x1),
+ svst1_hor_za16 (0, w0, p0, x1))
+
+/*
+** st1_za16_0_w0_p1:
+** mov (w1[2-5]), w0
+** st1h { za0h\.h\[\1, 1\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_za16_0_w0_p1,
+ svst1_hor_za16 (0, w0 + 1, p0, x1),
+ svst1_hor_za16 (0, w0 + 1, p0, x1))
+
+/*
+** st1_za16_0_w0_p7:
+** mov (w1[2-5]), w0
+** st1h { za0h\.h\[\1, 7\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_za16_0_w0_p7,
+ svst1_hor_za16 (0, w0 + 7, p0, x1),
+ svst1_hor_za16 (0, w0 + 7, p0, x1))
+
+/*
+** st1_za16_1_w0:
+** mov (w1[2-5]), w0
+** st1h { za1h\.h\[\1, 0\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_za16_1_w0,
+ svst1_hor_za16 (1, w0, p0, x1),
+ svst1_hor_za16 (1, w0, p0, x1))
+
+
+/*
+** st1_za16_1_w0_p1:
+** mov (w1[2-5]), w0
+** st1h { za1h\.h\[\1, 1\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_za16_1_w0_p1,
+ svst1_hor_za16 (1, w0 + 1, p0, x1),
+ svst1_hor_za16 (1, w0 + 1, p0, x1))
+
+/*
+** st1_za16_1_w0_p7:
+** mov (w1[2-5]), w0
+** st1h { za1h\.h\[\1, 7\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_za16_1_w0_p7,
+ svst1_hor_za16 (1, w0 + 7, p0, x1),
+ svst1_hor_za16 (1, w0 + 7, p0, x1))
+
+/*
+** st1_za16_1_w0_p5_index:
+** mov (w1[2-5]), w0
+** st1h { za1h\.h\[\1, 5\] }, p0, \[x1, x2, lsl #?1\]
+** ret
+*/
+TEST_STORE_ZA (st1_za16_1_w0_p5_index,
+ svst1_hor_za16 (1, w0 + 5, p0, x1 + x2 * 2),
+ svst1_hor_za16 (1, w0 + 5, p0, x1 + x2 * 2))
+
+/*
+** st1_za16_0_w0_p8:
+** add (w1[2-5]), w0, #?8
+** st1h { za0h\.h\[\1, 0\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_za16_0_w0_p8,
+ svst1_hor_za16 (0, w0 + 8, p0, x1),
+ svst1_hor_za16 (0, w0 + 8, p0, x1))
+
+/*
+** st1_za16_0_w0_m1:
+** sub (w1[2-5]), w0, #?1
+** st1h { za0h\.h\[\1, 0\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_za16_0_w0_m1,
+ svst1_hor_za16 (0, w0 - 1, p0, x1),
+ svst1_hor_za16 (0, w0 - 1, p0, x1))
--- /dev/null
+/* { dg-final { check-function-bodies "**" "" "-DCHECK_ASM" } } */
+
+#include "test_sme_acle.h"
+
+/*
+** st1_za32_0_0:
+** mov (w1[2-5]), (?:w0|#?0)
+** st1w { za0h\.s\[\1, 0\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_za32_0_0,
+ svst1_hor_za32 (0, 0, p0, x1),
+ svst1_hor_za32 (0, 0, p0, x1))
+
+/* It would also be OK (and perhaps better) to move 0 into a register
+ and use an offset of 3. */
+/*
+** st1_za32_0_3:
+** mov (w1[2-5]), #?3
+** st1w { za0h\.s\[\1, 0\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_za32_0_3,
+ svst1_hor_za32 (0, 3, p0, x1),
+ svst1_hor_za32 (0, 3, p0, x1))
+
+/*
+** st1_za32_0_4:
+** mov (w1[2-5]), #?4
+** st1w { za0h\.s\[\1, 0\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_za32_0_4,
+ svst1_hor_za32 (0, 4, p0, x1),
+ svst1_hor_za32 (0, 4, p0, x1))
+
+/*
+** st1_za32_0_w0:
+** mov (w1[2-5]), w0
+** st1w { za0h\.s\[\1, 0\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_za32_0_w0,
+ svst1_hor_za32 (0, w0, p0, x1),
+ svst1_hor_za32 (0, w0, p0, x1))
+
+/*
+** st1_za32_0_w0_p1:
+** mov (w1[2-5]), w0
+** st1w { za0h\.s\[\1, 1\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_za32_0_w0_p1,
+ svst1_hor_za32 (0, w0 + 1, p0, x1),
+ svst1_hor_za32 (0, w0 + 1, p0, x1))
+
+/*
+** st1_za32_0_w0_p3:
+** mov (w1[2-5]), w0
+** st1w { za0h\.s\[\1, 3\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_za32_0_w0_p3,
+ svst1_hor_za32 (0, w0 + 3, p0, x1),
+ svst1_hor_za32 (0, w0 + 3, p0, x1))
+
+/*
+** st1_za32_3_w0:
+** mov (w1[2-5]), w0
+** st1w { za3h\.s\[\1, 0\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_za32_3_w0,
+ svst1_hor_za32 (3, w0, p0, x1),
+ svst1_hor_za32 (3, w0, p0, x1))
+
+/*
+** st1_za32_3_w0_p1:
+** mov (w1[2-5]), w0
+** st1w { za3h\.s\[\1, 1\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_za32_3_w0_p1,
+ svst1_hor_za32 (3, w0 + 1, p0, x1),
+ svst1_hor_za32 (3, w0 + 1, p0, x1))
+
+/*
+** st1_za32_3_w0_p3:
+** mov (w1[2-5]), w0
+** st1w { za3h\.s\[\1, 3\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_za32_3_w0_p3,
+ svst1_hor_za32 (3, w0 + 3, p0, x1),
+ svst1_hor_za32 (3, w0 + 3, p0, x1))
+
+/*
+** st1_za32_1_w0_p2_index:
+** mov (w1[2-5]), w0
+** st1w { za1h\.s\[\1, 2\] }, p0, \[x1, x2, lsl #?2\]
+** ret
+*/
+TEST_STORE_ZA (st1_za32_1_w0_p2_index,
+ svst1_hor_za32 (1, w0 + 2, p0, x1 + x2 * 4),
+ svst1_hor_za32 (1, w0 + 2, p0, x1 + x2 * 4))
+
+/*
+** st1_za32_0_w0_p4:
+** add (w1[2-5]), w0, #?4
+** st1w { za0h\.s\[\1, 0\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_za32_0_w0_p4,
+ svst1_hor_za32 (0, w0 + 4, p0, x1),
+ svst1_hor_za32 (0, w0 + 4, p0, x1))
+
+/*
+** st1_za32_0_w0_m1:
+** sub (w1[2-5]), w0, #?1
+** st1w { za0h\.s\[\1, 0\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_za32_0_w0_m1,
+ svst1_hor_za32 (0, w0 - 1, p0, x1),
+ svst1_hor_za32 (0, w0 - 1, p0, x1))
--- /dev/null
+/* { dg-final { check-function-bodies "**" "" "-DCHECK_ASM" } } */
+
+#include "test_sme_acle.h"
+
+/*
+** st1_za64_0_0:
+** mov (w1[2-5]), (?:wzr|#?0)
+** st1d { za0h\.d\[\1, 0\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_za64_0_0,
+ svst1_hor_za64 (0, 0, p0, x1),
+ svst1_hor_za64 (0, 0, p0, x1))
+
+/* It would also be OK (and perhaps better) to move 0 into a register
+ and use an offset of 1. */
+/*
+** st1_za64_0_1:
+** mov (w1[2-5]), #?1
+** st1d { za0h\.d\[\1, 0\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_za64_0_1,
+ svst1_hor_za64 (0, 1, p0, x1),
+ svst1_hor_za64 (0, 1, p0, x1))
+
+/*
+** st1_za64_0_2:
+** mov (w1[2-5]), #?2
+** st1d { za0h\.d\[\1, 0\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_za64_0_2,
+ svst1_hor_za64 (0, 2, p0, x1),
+ svst1_hor_za64 (0, 2, p0, x1))
+
+/*
+** st1_za64_0_w0:
+** mov (w1[2-5]), w0
+** st1d { za0h\.d\[\1, 0\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_za64_0_w0,
+ svst1_hor_za64 (0, w0, p0, x1),
+ svst1_hor_za64 (0, w0, p0, x1))
+
+/*
+** st1_za64_0_w0_p1:
+** mov (w1[2-5]), w0
+** st1d { za0h\.d\[\1, 1\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_za64_0_w0_p1,
+ svst1_hor_za64 (0, w0 + 1, p0, x1),
+ svst1_hor_za64 (0, w0 + 1, p0, x1))
+
+/*
+** st1_za64_7_w0:
+** mov (w1[2-5]), w0
+** st1d { za7h\.d\[\1, 0\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_za64_7_w0,
+ svst1_hor_za64 (7, w0, p0, x1),
+ svst1_hor_za64 (7, w0, p0, x1))
+
+/*
+** st1_za64_7_w0_p1:
+** mov (w1[2-5]), w0
+** st1d { za7h\.d\[\1, 1\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_za64_7_w0_p1,
+ svst1_hor_za64 (7, w0 + 1, p0, x1),
+ svst1_hor_za64 (7, w0 + 1, p0, x1))
+
+/*
+** st1_za64_5_w0_p1_index:
+** mov (w1[2-5]), w0
+** st1d { za5h\.d\[\1, 1\] }, p0, \[x1, x2, lsl #?3\]
+** ret
+*/
+TEST_STORE_ZA (st1_za64_5_w0_p1_index,
+ svst1_hor_za64 (5, w0 + 1, p0, x1 + x2 * 8),
+ svst1_hor_za64 (5, w0 + 1, p0, x1 + x2 * 8))
+
+/*
+** st1_za64_0_w0_p2:
+** add (w1[2-5]), w0, #?2
+** st1d { za0h\.d\[\1, 0\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_za64_0_w0_p2,
+ svst1_hor_za64 (0, w0 + 2, p0, x1),
+ svst1_hor_za64 (0, w0 + 2, p0, x1))
+
+/*
+** st1_za64_0_w0_m1:
+** sub (w1[2-5]), w0, #?1
+** st1d { za0h\.d\[\1, 0\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_za64_0_w0_m1,
+ svst1_hor_za64 (0, w0 - 1, p0, x1),
+ svst1_hor_za64 (0, w0 - 1, p0, x1))
--- /dev/null
+/* { dg-final { check-function-bodies "**" "" "-DCHECK_ASM" } } */
+
+#include "test_sme_acle.h"
+
+/*
+** st1_za8_0_0:
+** mov (w1[2-5]), (?:wzr|#?0)
+** st1b { za0h\.b\[\1, 0\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_za8_0_0,
+ svst1_hor_za8 (0, 0, p0, x1),
+ svst1_hor_za8 (0, 0, p0, x1))
+
+/* It would also be OK (and perhaps better) to move 0 into a register
+ and use an offset of 15. */
+/*
+** st1_za8_0_15:
+** mov (w1[2-5]), #?15
+** st1b { za0h\.b\[\1, 0\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_za8_0_15,
+ svst1_hor_za8 (0, 15, p0, x1),
+ svst1_hor_za8 (0, 15, p0, x1))
+
+/*
+** st1_za8_0_16:
+** mov (w1[2-5]), #?16
+** st1b { za0h\.b\[\1, 0\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_za8_0_16,
+ svst1_hor_za8 (0, 16, p0, x1),
+ svst1_hor_za8 (0, 16, p0, x1))
+
+/*
+** st1_za8_0_w0:
+** mov (w1[2-5]), w0
+** st1b { za0h\.b\[\1, 0\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_za8_0_w0,
+ svst1_hor_za8 (0, w0, p0, x1),
+ svst1_hor_za8 (0, w0, p0, x1))
+
+/*
+** st1_za8_0_w0_p1:
+** mov (w1[2-5]), w0
+** st1b { za0h\.b\[\1, 1\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_za8_0_w0_p1,
+ svst1_hor_za8 (0, w0 + 1, p0, x1),
+ svst1_hor_za8 (0, w0 + 1, p0, x1))
+
+/*
+** st1_za8_0_w0_p15:
+** mov (w1[2-5]), w0
+** st1b { za0h\.b\[\1, 15\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_za8_0_w0_p15,
+ svst1_hor_za8 (0, w0 + 15, p0, x1),
+ svst1_hor_za8 (0, w0 + 15, p0, x1))
+
+/*
+** st1_za8_0_w0_p13_index:
+** mov (w1[2-5]), w0
+** st1b { za0h\.b\[\1, 15\] }, p0, \[x1, x2\]
+** ret
+*/
+TEST_STORE_ZA (st1_za8_0_w0_p13_index,
+ svst1_hor_za8 (0, w0 + 15, p0, x1 + x2),
+ svst1_hor_za8 (0, w0 + 15, p0, x1 + x2))
+
+/*
+** st1_za8_0_w0_p16:
+** add (w1[2-5]), w0, #?16
+** st1b { za0h\.b\[\1, 0\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_za8_0_w0_p16,
+ svst1_hor_za8 (0, w0 + 16, p0, x1),
+ svst1_hor_za8 (0, w0 + 16, p0, x1))
+
+/*
+** st1_za8_0_w0_m1:
+** sub (w1[2-5]), w0, #?1
+** st1b { za0h\.b\[\1, 0\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_za8_0_w0_m1,
+ svst1_hor_za8 (0, w0 - 1, p0, x1),
+ svst1_hor_za8 (0, w0 - 1, p0, x1))
--- /dev/null
+/* { dg-final { check-function-bodies "**" "" "-DCHECK_ASM" } } */
+
+#include "test_sme_acle.h"
+
+/*
+** st1_vnum_za128_0_0_0:
+** mov (w1[2-5]), (?:wzr|#?0)
+** st1q { za0v\.q\[\1, 0\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_vnum_za128_0_0_0,
+ svst1_ver_vnum_za128 (0, 0, p0, x1, 0),
+ svst1_ver_vnum_za128 (0, 0, p0, x1, 0))
+
+/*
+** st1_vnum_za128_7_1_0:
+** mov (w1[2-5]), #?1
+** st1q { za7v\.q\[\1, 0\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_vnum_za128_7_1_0,
+ svst1_ver_vnum_za128 (7, 1, p0, x1, 0),
+ svst1_ver_vnum_za128 (7, 1, p0, x1, 0))
+
+/*
+** st1_vnum_za128_11_1_5:
+** incb x1, all, mul #5
+** mov (w1[2-5]), #?6
+** st1q { za11v\.q\[\1, 0\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_vnum_za128_11_1_5,
+ svst1_ver_vnum_za128 (11, 1, p0, x1, 5),
+ svst1_ver_vnum_za128 (11, 1, p0, x1, 5))
+
+/*
+** st1_vnum_za128_3_w0_0:
+** mov (w1[2-5]), w0
+** st1q { za3v\.q\[\1, 0\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_vnum_za128_3_w0_0,
+ svst1_ver_vnum_za128 (3, w0, p0, x1, 0),
+ svst1_ver_vnum_za128 (3, w0, p0, x1, 0))
+
+/*
+** st1_vnum_za128_5_w0_0:
+** incb x1, all, mul #13
+** add (w1[2-5]), w0, #?13
+** st1q { za5v\.q\[\1, 0\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_vnum_za128_5_w0_0,
+ svst1_ver_vnum_za128 (5, w0, p0, x1, 13),
+ svst1_ver_vnum_za128 (5, w0, p0, x1, 13))
+
+/*
+** st1_vnum_za128_11_w0_0:
+** cntb (x[0-9]+)
+** madd (x[0-9]+), (?:\1, x2|x2, \1), x1
+** add (w1[2-5]), (?:w0, w2|w2, w0)
+** st1q { za11v\.q\[\3, 0\] }, p0, \[\2\]
+** ret
+*/
+TEST_STORE_ZA (st1_vnum_za128_11_w0_0,
+ svst1_ver_vnum_za128 (11, w0, p0, x1, x2),
+ svst1_ver_vnum_za128 (11, w0, p0, x1, x2))
+
+/*
+** st1_vnum_za128_15_w0p1_0:
+** add (w1[2-5]), w0, #?1
+** st1q { za15v\.q\[\1, 0\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_vnum_za128_15_w0p1_0,
+ svst1_ver_vnum_za128 (15, w0 + 1, p0, x1, 0),
+ svst1_ver_vnum_za128 (15, w0 + 1, p0, x1, 0))
--- /dev/null
+/* { dg-final { check-function-bodies "**" "" "-DCHECK_ASM" } } */
+
+#include "test_sme_acle.h"
+
+/*
+** st1_vnum_za16_1_0_1:
+** incb x1
+** mov (w1[2-5]), (?:wzr|#?0)
+** st1h { za1v\.h\[\1, 1\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_vnum_za16_1_0_1,
+ svst1_ver_vnum_za16 (1, 0, p0, x1, 1),
+ svst1_ver_vnum_za16 (1, 0, p0, x1, 1))
+
+/*
+** st1_vnum_za16_1_1_1:
+** incb x1
+** mov (w1[2-5]), #?1
+** st1h { za1v\.h\[\1, 1\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_vnum_za16_1_1_1,
+ svst1_ver_vnum_za16 (1, 1, p0, x1, 1),
+ svst1_ver_vnum_za16 (1, 1, p0, x1, 1))
+
+/*
+** st1_vnum_za16_0_0_8:
+** incb x1, all, mul #8
+** mov (w1[2-5]), #?8
+** st1h { za0v\.h\[\1, 0\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_vnum_za16_0_0_8,
+ svst1_ver_vnum_za16 (0, 0, p0, x1, 8),
+ svst1_ver_vnum_za16 (0, 0, p0, x1, 8))
+
+/*
+** st1_vnum_za16_0_1_8:
+** incb x1, all, mul #8
+** mov (w1[2-5]), #?9
+** st1h { za0v\.h\[\1, 0\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_vnum_za16_0_1_8,
+ svst1_ver_vnum_za16 (0, 1, p0, x1, 8),
+ svst1_ver_vnum_za16 (0, 1, p0, x1, 8))
+
+/*
+** st1_vnum_za16_0_w0_0:
+** mov (w1[2-5]), w0
+** st1h { za0v\.h\[\1, 0\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_vnum_za16_0_w0_0,
+ svst1_ver_vnum_za16 (0, w0, p0, x1, 0),
+ svst1_ver_vnum_za16 (0, w0, p0, x1, 0))
+
+/*
+** st1_vnum_za16_0_w0_1:
+** incb x1
+** mov (w1[2-5]), w0
+** st1h { za0v\.h\[\1, 1\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_vnum_za16_0_w0_1,
+ svst1_ver_vnum_za16 (0, w0, p0, x1, 1),
+ svst1_ver_vnum_za16 (0, w0, p0, x1, 1))
+
+/*
+** st1_vnum_za16_0_w0_7:
+** incb x1, all, mul #7
+** mov (w1[2-5]), w0
+** st1h { za0v\.h\[\1, 7\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_vnum_za16_0_w0_7,
+ svst1_ver_vnum_za16 (0, w0, p0, x1, 7),
+ svst1_ver_vnum_za16 (0, w0, p0, x1, 7))
+
+/*
+** st1_vnum_za16_1_w0_8:
+** incb x1, all, mul #8
+** add (w1[2-5]), w0, #?8
+** st1h { za1v\.h\[\1, 0\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_vnum_za16_1_w0_8,
+ svst1_ver_vnum_za16 (1, w0, p0, x1, 8),
+ svst1_ver_vnum_za16 (1, w0, p0, x1, 8))
+
+/*
+** st1_vnum_za16_1_w0_13:
+** incb x1, all, mul #13
+** add (w1[2-5]), w0, #?13
+** st1h { za1v\.h\[\1, 0\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_vnum_za16_1_w0_13,
+ svst1_ver_vnum_za16 (1, w0, p0, x1, 13),
+ svst1_ver_vnum_za16 (1, w0, p0, x1, 13))
+
+/*
+** st1_vnum_za16_0_w0_x2:
+** cntb (x[0-9]+)
+** madd (x[0-9]+), (?:\1, x2|x2, \1), x1
+** add (w1[2-5]), (?:w0, w2|w2, w0)
+** st1h { za0v\.h\[\3, 0\] }, p0, \[\2\]
+** ret
+*/
+TEST_STORE_ZA (st1_vnum_za16_0_w0_x2,
+ svst1_ver_vnum_za16 (0, w0, p0, x1, x2),
+ svst1_ver_vnum_za16 (0, w0, p0, x1, x2))
+
+/*
+** st1_vnum_za16_1_w0p1_0:
+** mov (w1[2-5]), w0
+** st1h { za1v\.h\[\1, 1\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_vnum_za16_1_w0p1_0,
+ svst1_ver_vnum_za16 (1, w0 + 1, p0, x1, 0),
+ svst1_ver_vnum_za16 (1, w0 + 1, p0, x1, 0))
--- /dev/null
+/* { dg-final { check-function-bodies "**" "" "-DCHECK_ASM" } } */
+
+#include "test_sme_acle.h"
+
+/*
+** st1_vnum_za32_3_0_1:
+** incb x1
+** mov (w1[2-5]), (?:wzr|#?0)
+** st1w { za3v\.s\[\1, 1\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_vnum_za32_3_0_1,
+ svst1_ver_vnum_za32 (3, 0, p0, x1, 1),
+ svst1_ver_vnum_za32 (3, 0, p0, x1, 1))
+
+/*
+** st1_vnum_za32_2_1_1:
+** incb x1
+** mov (w1[2-5]), #?1
+** st1w { za2v\.s\[\1, 1\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_vnum_za32_2_1_1,
+ svst1_ver_vnum_za32 (2, 1, p0, x1, 1),
+ svst1_ver_vnum_za32 (2, 1, p0, x1, 1))
+
+/*
+** st1_vnum_za32_0_0_4:
+** incb x1, all, mul #4
+** mov (w1[2-5]), #?4
+** st1w { za0v\.s\[\1, 0\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_vnum_za32_0_0_4,
+ svst1_ver_vnum_za32 (0, 0, p0, x1, 4),
+ svst1_ver_vnum_za32 (0, 0, p0, x1, 4))
+
+/*
+** st1_vnum_za32_2_1_4:
+** incb x1, all, mul #4
+** mov (w1[2-5]), #?5
+** st1w { za2v\.s\[\1, 0\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_vnum_za32_2_1_4,
+ svst1_ver_vnum_za32 (2, 1, p0, x1, 4),
+ svst1_ver_vnum_za32 (2, 1, p0, x1, 4))
+
+/*
+** st1_vnum_za32_0_w0_0:
+** mov (w1[2-5]), w0
+** st1w { za0v\.s\[\1, 0\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_vnum_za32_0_w0_0,
+ svst1_ver_vnum_za32 (0, w0, p0, x1, 0),
+ svst1_ver_vnum_za32 (0, w0, p0, x1, 0))
+
+/*
+** st1_vnum_za32_0_w0_1:
+** incb x1
+** mov (w1[2-5]), w0
+** st1w { za0v\.s\[\1, 1\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_vnum_za32_0_w0_1,
+ svst1_ver_vnum_za32 (0, w0, p0, x1, 1),
+ svst1_ver_vnum_za32 (0, w0, p0, x1, 1))
+
+/*
+** st1_vnum_za32_0_w0_3:
+** incb x1, all, mul #3
+** mov (w1[2-5]), w0
+** st1w { za0v\.s\[\1, 3\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_vnum_za32_0_w0_3,
+ svst1_ver_vnum_za32 (0, w0, p0, x1, 3),
+ svst1_ver_vnum_za32 (0, w0, p0, x1, 3))
+
+/*
+** st1_vnum_za32_1_w0_4:
+** incb x1, all, mul #4
+** add (w1[2-5]), w0, #?4
+** st1w { za1v\.s\[\1, 0\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_vnum_za32_1_w0_4,
+ svst1_ver_vnum_za32 (1, w0, p0, x1, 4),
+ svst1_ver_vnum_za32 (1, w0, p0, x1, 4))
+
+/*
+** st1_vnum_za32_3_w0_13:
+** incb x1, all, mul #13
+** add (w1[2-5]), w0, #?13
+** st1w { za3v\.s\[\1, 0\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_vnum_za32_3_w0_13,
+ svst1_ver_vnum_za32 (3, w0, p0, x1, 13),
+ svst1_ver_vnum_za32 (3, w0, p0, x1, 13))
+
+/*
+** st1_vnum_za32_0_w0_x2:
+** cntb (x[0-9]+)
+** madd (x[0-9]+), (?:\1, x2|x2, \1), x1
+** add (w1[2-5]), (?:w0, w2|w2, w0)
+** st1w { za0v\.s\[\3, 0\] }, p0, \[\2\]
+** ret
+*/
+TEST_STORE_ZA (st1_vnum_za32_0_w0_x2,
+ svst1_ver_vnum_za32 (0, w0, p0, x1, x2),
+ svst1_ver_vnum_za32 (0, w0, p0, x1, x2))
+
+/*
+** st1_vnum_za32_1_w0p1_0:
+** mov (w1[2-5]), w0
+** st1w { za1v\.s\[\1, 1\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_vnum_za32_1_w0p1_0,
+ svst1_ver_vnum_za32 (1, w0 + 1, p0, x1, 0),
+ svst1_ver_vnum_za32 (1, w0 + 1, p0, x1, 0))
--- /dev/null
+/* { dg-final { check-function-bodies "**" "" "-DCHECK_ASM" } } */
+
+#include "test_sme_acle.h"
+
+/*
+** st1_vnum_za64_3_0_1:
+** incb x1
+** mov (w1[2-5]), (?:wzr|#?0)
+** st1d { za3v\.d\[\1, 1\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_vnum_za64_3_0_1,
+ svst1_ver_vnum_za64 (3, 0, p0, x1, 1),
+ svst1_ver_vnum_za64 (3, 0, p0, x1, 1))
+
+/*
+** st1_vnum_za64_7_1_1:
+** incb x1
+** mov (w1[2-5]), #?1
+** st1d { za7v\.d\[\1, 1\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_vnum_za64_7_1_1,
+ svst1_ver_vnum_za64 (7, 1, p0, x1, 1),
+ svst1_ver_vnum_za64 (7, 1, p0, x1, 1))
+
+/*
+** st1_vnum_za64_0_0_2:
+** incb x1, all, mul #2
+** mov (w1[2-5]), #?2
+** st1d { za0v\.d\[\1, 0\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_vnum_za64_0_0_2,
+ svst1_ver_vnum_za64 (0, 0, p0, x1, 2),
+ svst1_ver_vnum_za64 (0, 0, p0, x1, 2))
+
+/*
+** st1_vnum_za64_5_1_2:
+** incb x1, all, mul #2
+** mov (w1[2-5]), #?3
+** st1d { za5v\.d\[\1, 0\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_vnum_za64_5_1_2,
+ svst1_ver_vnum_za64 (5, 1, p0, x1, 2),
+ svst1_ver_vnum_za64 (5, 1, p0, x1, 2))
+
+/*
+** st1_vnum_za64_0_w0_0:
+** mov (w1[2-5]), w0
+** st1d { za0v\.d\[\1, 0\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_vnum_za64_0_w0_0,
+ svst1_ver_vnum_za64 (0, w0, p0, x1, 0),
+ svst1_ver_vnum_za64 (0, w0, p0, x1, 0))
+
+/*
+** st1_vnum_za64_0_w0_1:
+** incb x1
+** mov (w1[2-5]), w0
+** st1d { za0v\.d\[\1, 1\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_vnum_za64_0_w0_1,
+ svst1_ver_vnum_za64 (0, w0, p0, x1, 1),
+ svst1_ver_vnum_za64 (0, w0, p0, x1, 1))
+
+/*
+** st1_vnum_za64_6_w0_2:
+** incb x1, all, mul #2
+** add (w1[2-5]), w0, #?2
+** st1d { za6v\.d\[\1, 0\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_vnum_za64_6_w0_2,
+ svst1_ver_vnum_za64 (6, w0, p0, x1, 2),
+ svst1_ver_vnum_za64 (6, w0, p0, x1, 2))
+
+/*
+** st1_vnum_za64_2_w0_13:
+** incb x1, all, mul #13
+** add (w1[2-5]), w0, #?13
+** st1d { za2v\.d\[\1, 0\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_vnum_za64_2_w0_13,
+ svst1_ver_vnum_za64 (2, w0, p0, x1, 13),
+ svst1_ver_vnum_za64 (2, w0, p0, x1, 13))
+
+/*
+** st1_vnum_za64_4_w0_x2:
+** cntb (x[0-9]+)
+** madd (x[0-9]+), (?:\1, x2|x2, \1), x1
+** add (w1[2-5]), (?:w0, w2|w2, w0)
+** st1d { za4v\.d\[\3, 0\] }, p0, \[\2\]
+** ret
+*/
+TEST_STORE_ZA (st1_vnum_za64_4_w0_x2,
+ svst1_ver_vnum_za64 (4, w0, p0, x1, x2),
+ svst1_ver_vnum_za64 (4, w0, p0, x1, x2))
+
+/*
+** st1_vnum_za64_1_w0p1_0:
+** mov (w1[2-5]), w0
+** st1d { za1v\.d\[\1, 1\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_vnum_za64_1_w0p1_0,
+ svst1_ver_vnum_za64 (1, w0 + 1, p0, x1, 0),
+ svst1_ver_vnum_za64 (1, w0 + 1, p0, x1, 0))
--- /dev/null
+/* { dg-final { check-function-bodies "**" "" "-DCHECK_ASM" } } */
+
+#include "test_sme_acle.h"
+
+/*
+** st1_vnum_za8_0_0_1:
+** incb x1
+** mov (w1[2-5]), (?:wzr|#?0)
+** st1b { za0v\.b\[\1, 1\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_vnum_za8_0_0_1,
+ svst1_ver_vnum_za8 (0, 0, p0, x1, 1),
+ svst1_ver_vnum_za8 (0, 0, p0, x1, 1))
+
+/*
+** st1_vnum_za8_0_1_1:
+** incb x1
+** mov (w1[2-5]), #?1
+** st1b { za0v\.b\[\1, 1\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_vnum_za8_0_1_1,
+ svst1_ver_vnum_za8 (0, 1, p0, x1, 1),
+ svst1_ver_vnum_za8 (0, 1, p0, x1, 1))
+
+/*
+** st1_vnum_za8_0_0_16:
+** incb x1, all, mul #16
+** mov (w1[2-5]), #?16
+** st1b { za0v\.b\[\1, 0\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_vnum_za8_0_0_16,
+ svst1_ver_vnum_za8 (0, 0, p0, x1, 16),
+ svst1_ver_vnum_za8 (0, 0, p0, x1, 16))
+
+/*
+** st1_vnum_za8_0_1_16:
+** incb x1, all, mul #16
+** mov (w1[2-5]), #?17
+** st1b { za0v\.b\[\1, 0\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_vnum_za8_0_1_16,
+ svst1_ver_vnum_za8 (0, 1, p0, x1, 16),
+ svst1_ver_vnum_za8 (0, 1, p0, x1, 16))
+
+/*
+** st1_vnum_za8_0_w0_0:
+** mov (w1[2-5]), w0
+** st1b { za0v\.b\[\1, 0\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_vnum_za8_0_w0_0,
+ svst1_ver_vnum_za8 (0, w0, p0, x1, 0),
+ svst1_ver_vnum_za8 (0, w0, p0, x1, 0))
+
+/*
+** st1_vnum_za8_0_w0_1:
+** incb x1
+** mov (w1[2-5]), w0
+** st1b { za0v\.b\[\1, 1\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_vnum_za8_0_w0_1,
+ svst1_ver_vnum_za8 (0, w0, p0, x1, 1),
+ svst1_ver_vnum_za8 (0, w0, p0, x1, 1))
+
+/*
+** st1_vnum_za8_0_w0_15:
+** incb x1, all, mul #15
+** mov (w1[2-5]), w0
+** st1b { za0v\.b\[\1, 15\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_vnum_za8_0_w0_15,
+ svst1_ver_vnum_za8 (0, w0, p0, x1, 15),
+ svst1_ver_vnum_za8 (0, w0, p0, x1, 15))
+
+/*
+** st1_vnum_za8_0_w0_16:
+** incb x1, all, mul #16
+** add (w1[2-5]), w0, #?16
+** st1b { za0v\.b\[\1, 0\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_vnum_za8_0_w0_16,
+ svst1_ver_vnum_za8 (0, w0, p0, x1, 16),
+ svst1_ver_vnum_za8 (0, w0, p0, x1, 16))
+
+/*
+** st1_vnum_za8_0_w0_x2:
+** cntb (x[0-9]+)
+** mul (x[0-9]+), (?:\1, x2|x2, \1)
+** add (w1[2-5]), (?:w0, w2|w2, w0)
+** st1b { za0v\.b\[\3, 0\] }, p0, \[x1, \2\]
+** ret
+*/
+TEST_STORE_ZA (st1_vnum_za8_0_w0_x2,
+ svst1_ver_vnum_za8 (0, w0, p0, x1, x2),
+ svst1_ver_vnum_za8 (0, w0, p0, x1, x2))
+
+/*
+** st1_vnum_za8_0_w0p1_0:
+** mov (w1[2-5]), w0
+** st1b { za0v\.b\[\1, 1\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_vnum_za8_0_w0p1_0,
+ svst1_ver_vnum_za8 (0, w0 + 1, p0, x1, 0),
+ svst1_ver_vnum_za8 (0, w0 + 1, p0, x1, 0))
--- /dev/null
+/* { dg-final { check-function-bodies "**" "" "-DCHECK_ASM" } } */
+
+#include "test_sme_acle.h"
+
+/*
+** st1_za128_0_0:
+** mov (w1[2-5]), (?:wzr|#?0)
+** st1q { za0v\.q\[\1, 0\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_za128_0_0,
+ svst1_ver_za128 (0, 0, p0, x1),
+ svst1_ver_za128 (0, 0, p0, x1))
+
+/*
+** st1_za128_0_1:
+** mov (w1[2-5]), #?1
+** st1q { za0v\.q\[\1, 0\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_za128_0_1,
+ svst1_ver_za128 (0, 1, p0, x1),
+ svst1_ver_za128 (0, 1, p0, x1))
+
+/*
+** st1_za128_0_w0:
+** mov (w1[2-5]), w0
+** st1q { za0v\.q\[\1, 0\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_za128_0_w0,
+ svst1_ver_za128 (0, w0, p0, x1),
+ svst1_ver_za128 (0, w0, p0, x1))
+
+/*
+** st1_za128_0_w0_p1:
+** add (w1[2-5]), w0, #?1
+** st1q { za0v\.q\[\1, 0\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_za128_0_w0_p1,
+ svst1_ver_za128 (0, w0 + 1, p0, x1),
+ svst1_ver_za128 (0, w0 + 1, p0, x1))
+
+/*
+** st1_za128_7_w0:
+** mov (w1[2-5]), w0
+** st1q { za7v\.q\[\1, 0\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_za128_7_w0,
+ svst1_ver_za128 (7, w0, p0, x1),
+ svst1_ver_za128 (7, w0, p0, x1))
+
+/*
+** st1_za128_13_w0:
+** mov (w1[2-5]), w0
+** st1q { za13v\.q\[\1, 0\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_za128_13_w0,
+ svst1_ver_za128 (13, w0, p0, x1),
+ svst1_ver_za128 (13, w0, p0, x1))
+
+/*
+** st1_za128_15_w0:
+** mov (w1[2-5]), w0
+** st1q { za15v\.q\[\1, 0\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_za128_15_w0,
+ svst1_ver_za128 (15, w0, p0, x1),
+ svst1_ver_za128 (15, w0, p0, x1))
+
+/*
+** st1_za128_9_w0_index:
+** mov (w1[2-5]), w0
+** st1q { za9v\.q\[\1, 0\] }, p0, \[x1, x2, lsl #?4\]
+** ret
+*/
+TEST_STORE_ZA (st1_za128_9_w0_index,
+ svst1_ver_za128 (9, w0, p0, x1 + x2 * 16),
+ svst1_ver_za128 (9, w0, p0, x1 + x2 * 16))
--- /dev/null
+/* { dg-final { check-function-bodies "**" "" "-DCHECK_ASM" } } */
+
+#include "test_sme_acle.h"
+
+/*
+** st1_za16_0_0:
+** mov (w1[2-5]), (?:wzr|#?0)
+** st1h { za0v\.h\[\1, 0\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_za16_0_0,
+ svst1_ver_za16 (0, 0, p0, x1),
+ svst1_ver_za16 (0, 0, p0, x1))
+
+/* It would also be OK (and perhaps better) to move 0 into a register
+ and use an offset of 7. */
+/*
+** st1_za16_0_7:
+** mov (w1[2-5]), #?7
+** st1h { za0v\.h\[\1, 0\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_za16_0_7,
+ svst1_ver_za16 (0, 7, p0, x1),
+ svst1_ver_za16 (0, 7, p0, x1))
+
+/*
+** st1_za16_0_8:
+** mov (w1[2-5]), #?8
+** st1h { za0v\.h\[\1, 0\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_za16_0_8,
+ svst1_ver_za16 (0, 8, p0, x1),
+ svst1_ver_za16 (0, 8, p0, x1))
+
+/*
+** st1_za16_0_w0:
+** mov (w1[2-5]), w0
+** st1h { za0v\.h\[\1, 0\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_za16_0_w0,
+ svst1_ver_za16 (0, w0, p0, x1),
+ svst1_ver_za16 (0, w0, p0, x1))
+
+/*
+** st1_za16_0_w0_p1:
+** mov (w1[2-5]), w0
+** st1h { za0v\.h\[\1, 1\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_za16_0_w0_p1,
+ svst1_ver_za16 (0, w0 + 1, p0, x1),
+ svst1_ver_za16 (0, w0 + 1, p0, x1))
+
+/*
+** st1_za16_0_w0_p7:
+** mov (w1[2-5]), w0
+** st1h { za0v\.h\[\1, 7\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_za16_0_w0_p7,
+ svst1_ver_za16 (0, w0 + 7, p0, x1),
+ svst1_ver_za16 (0, w0 + 7, p0, x1))
+
+/*
+** st1_za16_1_w0:
+** mov (w1[2-5]), w0
+** st1h { za1v\.h\[\1, 0\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_za16_1_w0,
+ svst1_ver_za16 (1, w0, p0, x1),
+ svst1_ver_za16 (1, w0, p0, x1))
+
+
+/*
+** st1_za16_1_w0_p1:
+** mov (w1[2-5]), w0
+** st1h { za1v\.h\[\1, 1\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_za16_1_w0_p1,
+ svst1_ver_za16 (1, w0 + 1, p0, x1),
+ svst1_ver_za16 (1, w0 + 1, p0, x1))
+
+/*
+** st1_za16_1_w0_p7:
+** mov (w1[2-5]), w0
+** st1h { za1v\.h\[\1, 7\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_za16_1_w0_p7,
+ svst1_ver_za16 (1, w0 + 7, p0, x1),
+ svst1_ver_za16 (1, w0 + 7, p0, x1))
+
+/*
+** st1_za16_1_w0_p5_index:
+** mov (w1[2-5]), w0
+** st1h { za1v\.h\[\1, 5\] }, p0, \[x1, x2, lsl #?1\]
+** ret
+*/
+TEST_STORE_ZA (st1_za16_1_w0_p5_index,
+ svst1_ver_za16 (1, w0 + 5, p0, x1 + x2 * 2),
+ svst1_ver_za16 (1, w0 + 5, p0, x1 + x2 * 2))
+
+/*
+** st1_za16_0_w0_p8:
+** add (w1[2-5]), w0, #?8
+** st1h { za0v\.h\[\1, 0\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_za16_0_w0_p8,
+ svst1_ver_za16 (0, w0 + 8, p0, x1),
+ svst1_ver_za16 (0, w0 + 8, p0, x1))
+
+/*
+** st1_za16_0_w0_m1:
+** sub (w1[2-5]), w0, #?1
+** st1h { za0v\.h\[\1, 0\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_za16_0_w0_m1,
+ svst1_ver_za16 (0, w0 - 1, p0, x1),
+ svst1_ver_za16 (0, w0 - 1, p0, x1))
--- /dev/null
+/* { dg-final { check-function-bodies "**" "" "-DCHECK_ASM" } } */
+
+#include "test_sme_acle.h"
+
+/*
+** st1_za32_0_0:
+** mov (w1[2-5]), (?:w0|#?0)
+** st1w { za0v\.s\[\1, 0\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_za32_0_0,
+ svst1_ver_za32 (0, 0, p0, x1),
+ svst1_ver_za32 (0, 0, p0, x1))
+
+/* It would also be OK (and perhaps better) to move 0 into a register
+ and use an offset of 3. */
+/*
+** st1_za32_0_3:
+** mov (w1[2-5]), #?3
+** st1w { za0v\.s\[\1, 0\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_za32_0_3,
+ svst1_ver_za32 (0, 3, p0, x1),
+ svst1_ver_za32 (0, 3, p0, x1))
+
+/*
+** st1_za32_0_4:
+** mov (w1[2-5]), #?4
+** st1w { za0v\.s\[\1, 0\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_za32_0_4,
+ svst1_ver_za32 (0, 4, p0, x1),
+ svst1_ver_za32 (0, 4, p0, x1))
+
+/*
+** st1_za32_0_w0:
+** mov (w1[2-5]), w0
+** st1w { za0v\.s\[\1, 0\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_za32_0_w0,
+ svst1_ver_za32 (0, w0, p0, x1),
+ svst1_ver_za32 (0, w0, p0, x1))
+
+/*
+** st1_za32_0_w0_p1:
+** mov (w1[2-5]), w0
+** st1w { za0v\.s\[\1, 1\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_za32_0_w0_p1,
+ svst1_ver_za32 (0, w0 + 1, p0, x1),
+ svst1_ver_za32 (0, w0 + 1, p0, x1))
+
+/*
+** st1_za32_0_w0_p3:
+** mov (w1[2-5]), w0
+** st1w { za0v\.s\[\1, 3\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_za32_0_w0_p3,
+ svst1_ver_za32 (0, w0 + 3, p0, x1),
+ svst1_ver_za32 (0, w0 + 3, p0, x1))
+
+/*
+** st1_za32_3_w0:
+** mov (w1[2-5]), w0
+** st1w { za3v\.s\[\1, 0\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_za32_3_w0,
+ svst1_ver_za32 (3, w0, p0, x1),
+ svst1_ver_za32 (3, w0, p0, x1))
+
+/*
+** st1_za32_3_w0_p1:
+** mov (w1[2-5]), w0
+** st1w { za3v\.s\[\1, 1\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_za32_3_w0_p1,
+ svst1_ver_za32 (3, w0 + 1, p0, x1),
+ svst1_ver_za32 (3, w0 + 1, p0, x1))
+
+/*
+** st1_za32_3_w0_p3:
+** mov (w1[2-5]), w0
+** st1w { za3v\.s\[\1, 3\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_za32_3_w0_p3,
+ svst1_ver_za32 (3, w0 + 3, p0, x1),
+ svst1_ver_za32 (3, w0 + 3, p0, x1))
+
+/*
+** st1_za32_1_w0_p2_index:
+** mov (w1[2-5]), w0
+** st1w { za1v\.s\[\1, 2\] }, p0, \[x1, x2, lsl #?2\]
+** ret
+*/
+TEST_STORE_ZA (st1_za32_1_w0_p2_index,
+ svst1_ver_za32 (1, w0 + 2, p0, x1 + x2 * 4),
+ svst1_ver_za32 (1, w0 + 2, p0, x1 + x2 * 4))
+
+/*
+** st1_za32_0_w0_p4:
+** add (w1[2-5]), w0, #?4
+** st1w { za0v\.s\[\1, 0\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_za32_0_w0_p4,
+ svst1_ver_za32 (0, w0 + 4, p0, x1),
+ svst1_ver_za32 (0, w0 + 4, p0, x1))
+
+/*
+** st1_za32_0_w0_m1:
+** sub (w1[2-5]), w0, #?1
+** st1w { za0v\.s\[\1, 0\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_za32_0_w0_m1,
+ svst1_ver_za32 (0, w0 - 1, p0, x1),
+ svst1_ver_za32 (0, w0 - 1, p0, x1))
--- /dev/null
+/* { dg-final { check-function-bodies "**" "" "-DCHECK_ASM" } } */
+
+#include "test_sme_acle.h"
+
+/*
+** st1_za64_0_0:
+** mov (w1[2-5]), (?:wzr|#?0)
+** st1d { za0v\.d\[\1, 0\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_za64_0_0,
+ svst1_ver_za64 (0, 0, p0, x1),
+ svst1_ver_za64 (0, 0, p0, x1))
+
+/* It would also be OK (and perhaps better) to move 0 into a register
+ and use an offset of 1. */
+/*
+** st1_za64_0_1:
+** mov (w1[2-5]), #?1
+** st1d { za0v\.d\[\1, 0\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_za64_0_1,
+ svst1_ver_za64 (0, 1, p0, x1),
+ svst1_ver_za64 (0, 1, p0, x1))
+
+/*
+** st1_za64_0_2:
+** mov (w1[2-5]), #?2
+** st1d { za0v\.d\[\1, 0\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_za64_0_2,
+ svst1_ver_za64 (0, 2, p0, x1),
+ svst1_ver_za64 (0, 2, p0, x1))
+
+/*
+** st1_za64_0_w0:
+** mov (w1[2-5]), w0
+** st1d { za0v\.d\[\1, 0\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_za64_0_w0,
+ svst1_ver_za64 (0, w0, p0, x1),
+ svst1_ver_za64 (0, w0, p0, x1))
+
+/*
+** st1_za64_0_w0_p1:
+** mov (w1[2-5]), w0
+** st1d { za0v\.d\[\1, 1\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_za64_0_w0_p1,
+ svst1_ver_za64 (0, w0 + 1, p0, x1),
+ svst1_ver_za64 (0, w0 + 1, p0, x1))
+
+/*
+** st1_za64_7_w0:
+** mov (w1[2-5]), w0
+** st1d { za7v\.d\[\1, 0\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_za64_7_w0,
+ svst1_ver_za64 (7, w0, p0, x1),
+ svst1_ver_za64 (7, w0, p0, x1))
+
+/*
+** st1_za64_7_w0_p1:
+** mov (w1[2-5]), w0
+** st1d { za7v\.d\[\1, 1\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_za64_7_w0_p1,
+ svst1_ver_za64 (7, w0 + 1, p0, x1),
+ svst1_ver_za64 (7, w0 + 1, p0, x1))
+
+/*
+** st1_za64_5_w0_p1_index:
+** mov (w1[2-5]), w0
+** st1d { za5v\.d\[\1, 1\] }, p0, \[x1, x2, lsl #?3\]
+** ret
+*/
+TEST_STORE_ZA (st1_za64_5_w0_p1_index,
+ svst1_ver_za64 (5, w0 + 1, p0, x1 + x2 * 8),
+ svst1_ver_za64 (5, w0 + 1, p0, x1 + x2 * 8))
+
+/*
+** st1_za64_0_w0_p2:
+** add (w1[2-5]), w0, #?2
+** st1d { za0v\.d\[\1, 0\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_za64_0_w0_p2,
+ svst1_ver_za64 (0, w0 + 2, p0, x1),
+ svst1_ver_za64 (0, w0 + 2, p0, x1))
+
+/*
+** st1_za64_0_w0_m1:
+** sub (w1[2-5]), w0, #?1
+** st1d { za0v\.d\[\1, 0\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_za64_0_w0_m1,
+ svst1_ver_za64 (0, w0 - 1, p0, x1),
+ svst1_ver_za64 (0, w0 - 1, p0, x1))
--- /dev/null
+/* { dg-final { check-function-bodies "**" "" "-DCHECK_ASM" } } */
+
+#include "test_sme_acle.h"
+
+/*
+** st1_za8_0_0:
+** mov (w1[2-5]), (?:wzr|#?0)
+** st1b { za0v\.b\[\1, 0\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_za8_0_0,
+ svst1_ver_za8 (0, 0, p0, x1),
+ svst1_ver_za8 (0, 0, p0, x1))
+
+/* It would also be OK (and perhaps better) to move 0 into a register
+ and use an offset of 15. */
+/*
+** st1_za8_0_15:
+** mov (w1[2-5]), #?15
+** st1b { za0v\.b\[\1, 0\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_za8_0_15,
+ svst1_ver_za8 (0, 15, p0, x1),
+ svst1_ver_za8 (0, 15, p0, x1))
+
+/*
+** st1_za8_0_16:
+** mov (w1[2-5]), #?16
+** st1b { za0v\.b\[\1, 0\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_za8_0_16,
+ svst1_ver_za8 (0, 16, p0, x1),
+ svst1_ver_za8 (0, 16, p0, x1))
+
+/*
+** st1_za8_0_w0:
+** mov (w1[2-5]), w0
+** st1b { za0v\.b\[\1, 0\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_za8_0_w0,
+ svst1_ver_za8 (0, w0, p0, x1),
+ svst1_ver_za8 (0, w0, p0, x1))
+
+/*
+** st1_za8_0_w0_p1:
+** mov (w1[2-5]), w0
+** st1b { za0v\.b\[\1, 1\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_za8_0_w0_p1,
+ svst1_ver_za8 (0, w0 + 1, p0, x1),
+ svst1_ver_za8 (0, w0 + 1, p0, x1))
+
+/*
+** st1_za8_0_w0_p15:
+** mov (w1[2-5]), w0
+** st1b { za0v\.b\[\1, 15\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_za8_0_w0_p15,
+ svst1_ver_za8 (0, w0 + 15, p0, x1),
+ svst1_ver_za8 (0, w0 + 15, p0, x1))
+
+/*
+** st1_za8_0_w0_p13_index:
+** mov (w1[2-5]), w0
+** st1b { za0v\.b\[\1, 15\] }, p0, \[x1, x2\]
+** ret
+*/
+TEST_STORE_ZA (st1_za8_0_w0_p13_index,
+ svst1_ver_za8 (0, w0 + 15, p0, x1 + x2),
+ svst1_ver_za8 (0, w0 + 15, p0, x1 + x2))
+
+/*
+** st1_za8_0_w0_p16:
+** add (w1[2-5]), w0, #?16
+** st1b { za0v\.b\[\1, 0\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_za8_0_w0_p16,
+ svst1_ver_za8 (0, w0 + 16, p0, x1),
+ svst1_ver_za8 (0, w0 + 16, p0, x1))
+
+/*
+** st1_za8_0_w0_m1:
+** sub (w1[2-5]), w0, #?1
+** st1b { za0v\.b\[\1, 0\] }, p0, \[x1\]
+** ret
+*/
+TEST_STORE_ZA (st1_za8_0_w0_m1,
+ svst1_ver_za8 (0, w0 - 1, p0, x1),
+ svst1_ver_za8 (0, w0 - 1, p0, x1))
--- /dev/null
+/* { dg-final { check-function-bodies "**" "" "-DCHECK_ASM" } } */
+
+#include "test_sme_acle.h"
+
+/*
+** str_vnum_za_0_0:
+** mov (w1[2-5]), (?:wzr|#?0)
+** str za\[\1, 0\], \[x1(?:, #0, mul vl)?\]
+** ret
+*/
+TEST_STORE_ZA (str_vnum_za_0_0,
+ svstr_vnum_za (0, x1, 0),
+ svstr_vnum_za (0, x1, 0))
+
+/*
+** str_vnum_za_0_1:
+** mov (w1[2-5]), (?:wzr|#?0)
+** str za\[\1, 1\], \[x1(?:, #1, mul vl)?\]
+** ret
+*/
+TEST_STORE_ZA (str_vnum_za_0_1,
+ svstr_vnum_za (0, x1, 1),
+ svstr_vnum_za (0, x1, 1))
+
+/*
+** str_vnum_za_1_0:
+** mov (w1[2-5]), #?1
+** str za\[\1, 0\], \[x1(?:, #0, mul vl)?\]
+** ret
+*/
+TEST_STORE_ZA (str_vnum_za_1_0,
+ svstr_vnum_za (1, x1, 0),
+ svstr_vnum_za (1, x1, 0))
+
+/*
+** str_vnum_za_1_2:
+** mov (w1[2-5]), #?1
+** str za\[\1, 2\], \[x1(?:, #2, mul vl)?\]
+** ret
+*/
+TEST_STORE_ZA (str_vnum_za_1_2,
+ svstr_vnum_za (1, x1, 2),
+ svstr_vnum_za (1, x1, 2))
+
+/*
+** str_vnum_za_w0_0:
+** mov (w1[2-5]), w0
+** str za\[\1, 0\], \[x1(?:, #0, mul vl)?\]
+** ret
+*/
+TEST_STORE_ZA (str_vnum_za_w0_0,
+ svstr_vnum_za (w0, x1, 0),
+ svstr_vnum_za (w0, x1, 0))
+
+/*
+** str_vnum_za_w0_1:
+** mov (w1[2-5]), w0
+** str za\[\1, 1\], \[x1, #1, mul vl\]
+** ret
+*/
+TEST_STORE_ZA (str_vnum_za_w0_1,
+ svstr_vnum_za (w0, x1, 1),
+ svstr_vnum_za (w0, x1, 1))
+
+/*
+** str_vnum_za_w0_13:
+** mov (w1[2-5]), w0
+** str za\[\1, 13\], \[x1, #13, mul vl\]
+** ret
+*/
+TEST_STORE_ZA (str_vnum_za_w0_13,
+ svstr_vnum_za (w0, x1, 13),
+ svstr_vnum_za (w0, x1, 13))
+
+/*
+** str_vnum_za_w0_15:
+** mov (w1[2-5]), w0
+** str za\[\1, 15\], \[x1, #15, mul vl\]
+** ret
+*/
+TEST_STORE_ZA (str_vnum_za_w0_15,
+ svstr_vnum_za (w0, x1, 15),
+ svstr_vnum_za (w0, x1, 15))
+
+/*
+** str_vnum_za_w0_16:
+** (
+** add (w1[2-5]), w0, #?16
+** incb x1, all, mul #16
+** str za\[\1, 0\], \[x1(?:, #0, mul vl)?\]
+** |
+** incb x1, all, mul #16
+** add (w1[2-5]), w0, #?16
+** str za\[\2, 0\], \[x1(?:, #0, mul vl)?\]
+** )
+** ret
+*/
+TEST_STORE_ZA (str_vnum_za_w0_16,
+ svstr_vnum_za (w0, x1, 16),
+ svstr_vnum_za (w0, x1, 16))
+
+/*
+** str_vnum_za_w0_m1:
+** (
+** sub (w1[2-5]), w0, #?1
+** decb x1
+** str za\[\1, 0\], \[x1(?:, #0, mul vl)?\]
+** |
+** decb x1
+** sub (w1[2-5]), w0, #?1
+** str za\[\2, 0\], \[x1(?:, #0, mul vl)?\]
+** )
+** ret
+*/
+TEST_STORE_ZA (str_vnum_za_w0_m1,
+ svstr_vnum_za (w0, x1, -1),
+ svstr_vnum_za (w0, x1, -1))
+
+/*
+** str_vnum_za_w0p1_0:
+** add (w1[2-5]), w0, #?1
+** str za\[\1, 0\], \[x1(?:, #0, mul vl)?\]
+** ret
+*/
+TEST_STORE_ZA (str_vnum_za_w0p1_0,
+ svstr_vnum_za (w0 + 1, x1, 0),
+ svstr_vnum_za (w0 + 1, x1, 0))
+
+/*
+** str_vnum_za_w0m1_1:
+** sub (w1[2-5]), w0, #?1
+** str za\[\1, 1\], \[x1(?:, #1, mul vl)?\]
+** ret
+*/
+TEST_STORE_ZA (str_vnum_za_w0m1_1,
+ svstr_vnum_za (w0 - 1, x1, 1),
+ svstr_vnum_za (w0 - 1, x1, 1))
+
+/*
+** str_vnum_za_w0p2_3:
+** add (w1[2-5]), w0, #?2
+** str za\[\1, 3\], \[x1(?:, #3, mul vl)?\]
+** ret
+*/
+TEST_STORE_ZA (str_vnum_za_w0p2_3,
+ svstr_vnum_za (w0 + 2, x1, 3),
+ svstr_vnum_za (w0 + 2, x1, 3))
--- /dev/null
+/* { dg-final { check-function-bodies "**" "" "-DCHECK_ASM" } } */
+
+#define STREAMING_COMPATIBLE
+#include "test_sme_acle.h"
+
+/*
+** str_vnum_za_0_0:
+** mov (w1[2-5]), (?:wzr|#?0)
+** str za\[\1, 0\], \[x1(?:, #0, mul vl)?\]
+** ret
+*/
+TEST_STORE_ZA (str_vnum_za_0_0,
+ svstr_vnum_za (0, x1, 0),
+ svstr_vnum_za (0, x1, 0))
+
+/*
+** str_vnum_za_0_1:
+** mov (w1[2-5]), (?:wzr|#?0)
+** str za\[\1, 1\], \[x1(?:, #1, mul vl)?\]
+** ret
+*/
+TEST_STORE_ZA (str_vnum_za_0_1,
+ svstr_vnum_za (0, x1, 1),
+ svstr_vnum_za (0, x1, 1))
+
+/*
+** str_vnum_za_1_0:
+** mov (w1[2-5]), #?1
+** str za\[\1, 0\], \[x1(?:, #0, mul vl)?\]
+** ret
+*/
+TEST_STORE_ZA (str_vnum_za_1_0,
+ svstr_vnum_za (1, x1, 0),
+ svstr_vnum_za (1, x1, 0))
+
+/*
+** str_vnum_za_1_2:
+** mov (w1[2-5]), #?1
+** str za\[\1, 2\], \[x1(?:, #2, mul vl)?\]
+** ret
+*/
+TEST_STORE_ZA (str_vnum_za_1_2,
+ svstr_vnum_za (1, x1, 2),
+ svstr_vnum_za (1, x1, 2))
+
+/*
+** str_vnum_za_w0_0:
+** mov (w1[2-5]), w0
+** str za\[\1, 0\], \[x1(?:, #0, mul vl)?\]
+** ret
+*/
+TEST_STORE_ZA (str_vnum_za_w0_0,
+ svstr_vnum_za (w0, x1, 0),
+ svstr_vnum_za (w0, x1, 0))
+
+/*
+** str_vnum_za_w0_1:
+** mov (w1[2-5]), w0
+** str za\[\1, 1\], \[x1, #1, mul vl\]
+** ret
+*/
+TEST_STORE_ZA (str_vnum_za_w0_1,
+ svstr_vnum_za (w0, x1, 1),
+ svstr_vnum_za (w0, x1, 1))
+
+/*
+** str_vnum_za_w0_13:
+** mov (w1[2-5]), w0
+** str za\[\1, 13\], \[x1, #13, mul vl\]
+** ret
+*/
+TEST_STORE_ZA (str_vnum_za_w0_13,
+ svstr_vnum_za (w0, x1, 13),
+ svstr_vnum_za (w0, x1, 13))
+
+/*
+** str_vnum_za_w0_15:
+** mov (w1[2-5]), w0
+** str za\[\1, 15\], \[x1, #15, mul vl\]
+** ret
+*/
+TEST_STORE_ZA (str_vnum_za_w0_15,
+ svstr_vnum_za (w0, x1, 15),
+ svstr_vnum_za (w0, x1, 15))
+
+/*
+** str_vnum_za_w0_16:
+** (
+** add (w1[2-5]), w0, #?16
+** addsvl (x[0-9]+), x1, #16
+** str za\[\1, 0\], \[\2(?:, #0, mul vl)?\]
+** |
+** addsvl (x[0-9]+), x1, #16
+** add (w1[2-5]), w0, #?16
+** str za\[\4, 0\], \[\3(?:, #0, mul vl)?\]
+** )
+** ret
+*/
+TEST_STORE_ZA (str_vnum_za_w0_16,
+ svstr_vnum_za (w0, x1, 16),
+ svstr_vnum_za (w0, x1, 16))
+
+/*
+** str_vnum_za_w0_m1:
+** (
+** sub (w1[2-5]), w0, #?1
+** addsvl (x[0-9]+), x1, #-1
+** str za\[\1, 0\], \[\2(?:, #0, mul vl)?\]
+** |
+** addsvl (x[0-9]+), x1, #-1
+** sub (w1[2-5]), w0, #?1
+** str za\[\4, 0\], \[\3(?:, #0, mul vl)?\]
+** )
+** ret
+*/
+TEST_STORE_ZA (str_vnum_za_w0_m1,
+ svstr_vnum_za (w0, x1, -1),
+ svstr_vnum_za (w0, x1, -1))
+
+/*
+** str_vnum_za_w0p1_0:
+** add (w1[2-5]), w0, #?1
+** str za\[\1, 0\], \[x1(?:, #0, mul vl)?\]
+** ret
+*/
+TEST_STORE_ZA (str_vnum_za_w0p1_0,
+ svstr_vnum_za (w0 + 1, x1, 0),
+ svstr_vnum_za (w0 + 1, x1, 0))
+
+/*
+** str_vnum_za_w0m1_1:
+** sub (w1[2-5]), w0, #?1
+** str za\[\1, 1\], \[x1(?:, #1, mul vl)?\]
+** ret
+*/
+TEST_STORE_ZA (str_vnum_za_w0m1_1,
+ svstr_vnum_za (w0 - 1, x1, 1),
+ svstr_vnum_za (w0 - 1, x1, 1))
+
+/*
+** str_vnum_za_w0p2_3:
+** add (w1[2-5]), w0, #?2
+** str za\[\1, 3\], \[x1(?:, #3, mul vl)?\]
+** ret
+*/
+TEST_STORE_ZA (str_vnum_za_w0p2_3,
+ svstr_vnum_za (w0 + 2, x1, 3),
+ svstr_vnum_za (w0 + 2, x1, 3))
--- /dev/null
+/* { dg-final { check-function-bodies "**" "" "-DCHECK_ASM" } } */
+
+#include "test_sme_acle.h"
+
+/*
+** str_za_0:
+** mov (w1[2-5]), (?:wzr|#?0)
+** str za\[\1, 0\], \[x1(?:, #0, mul vl)?\]
+** ret
+*/
+TEST_STORE_ZA (str_za_0,
+ svstr_za (0, x1),
+ svstr_za (0, x1))
+
+/*
+** str_za_1:
+** mov (w1[2-5]), #?1
+** str za\[\1, 0\], \[x1(?:, #0, mul vl)?\]
+** ret
+*/
+TEST_STORE_ZA (str_za_1,
+ svstr_za (1, x1),
+ svstr_za (1, x1))
+
+/*
+** str_za_w0:
+** mov (w1[2-5]), w0
+** str za\[\1, 0\], \[x1(?:, #0, mul vl)?\]
+** ret
+*/
+TEST_STORE_ZA (str_za_w0,
+ svstr_za (w0, x1),
+ svstr_za (w0, x1))
+
+/*
+** str_za_w0_1_vnum:
+** mov (w1[2-5]), w0
+** str za\[\1, 1\], \[x1, #1, mul vl\]
+** ret
+*/
+TEST_STORE_ZA (str_za_w0_1_vnum,
+ svstr_za (w0 + 1, x1 + svcntsb ()),
+ svstr_za (w0 + 1, x1 + svcntsb ()))
+
+/*
+** str_za_w0_13_vnum:
+** mov (w1[2-5]), w0
+** str za\[\1, 13\], \[x1, #13, mul vl\]
+** ret
+*/
+TEST_STORE_ZA (str_za_w0_13_vnum,
+ svstr_za (w0 + 13, x1 + svcntsb () * 13),
+ svstr_za (w0 + 13, x1 + svcntsb () * 13))
+
+/*
+** str_za_w0_15_vnum:
+** mov (w1[2-5]), w0
+** str za\[\1, 15\], \[x1, #15, mul vl\]
+** ret
+*/
+TEST_STORE_ZA (str_za_w0_15_vnum,
+ svstr_za (w0 + 15, x1 + svcntsb () * 15),
+ svstr_za (w0 + 15, x1 + svcntsb () * 15))
+
+/*
+** str_za_w0_16_vnum:
+** (
+** add (w1[2-5]), w0, #?16
+** incb x1, all, mul #16
+** str za\[\1, 0\], \[x1(?:, #0, mul vl)?\]
+** |
+** incb x1, all, mul #16
+** add (w1[2-5]), w0, #?16
+** str za\[\2, 0\], \[x1(?:, #0, mul vl)?\]
+** )
+** ret
+*/
+TEST_STORE_ZA (str_za_w0_16_vnum,
+ svstr_za (w0 + 16, x1 + svcntsb () * 16),
+ svstr_za (w0 + 16, x1 + svcntsb () * 16))
+
+/*
+** str_za_w0_m1_vnum:
+** (
+** sub (w1[2-5]), w0, #?1
+** decb x1
+** str za\[\1, 0\], \[x1(?:, #0, mul vl)?\]
+** |
+** decb x1
+** sub (w1[2-5]), w0, #?1
+** str za\[\2, 0\], \[x1(?:, #0, mul vl)?\]
+** )
+** ret
+*/
+TEST_STORE_ZA (str_za_w0_m1_vnum,
+ svstr_za (w0 - 1, x1 - svcntsb ()),
+ svstr_za (w0 - 1, x1 - svcntsb ()))
+
+/*
+** str_za_w0p2:
+** add (w1[2-5]), w0, #?2
+** str za\[\1, 0\], \[x1(?:, #0, mul vl)?\]
+** ret
+*/
+TEST_STORE_ZA (str_za_w0p2,
+ svstr_za (w0 + 2, x1),
+ svstr_za (w0 + 2, x1))
+
+/*
+** str_za_offset:
+** (
+** mov (w1[2-5]), w0
+** add (x[0-9]+), x1, #?1
+** str za\[\1, 0\], \[\2(?:, #0, mul vl)?\]
+** |
+** add (x[0-9]+), x1, #?1
+** mov (w1[2-5]), w0
+** str za\[\4, 0\], \[\3(?:, #0, mul vl)?\]
+** )
+** ret
+*/
+TEST_STORE_ZA (str_za_offset,
+ svstr_za (w0, x1 + 1),
+ svstr_za (w0, x1 + 1))
--- /dev/null
+/* { dg-final { check-function-bodies "**" "" "-DCHECK_ASM" } } */
+
+#define STREAMING_COMPATIBLE
+#include "test_sme_acle.h"
+
+/*
+** str_za_0:
+** mov (w1[2-5]), (?:wzr|#?0)
+** str za\[\1, 0\], \[x1(?:, #0, mul vl)?\]
+** ret
+*/
+TEST_STORE_ZA (str_za_0,
+ svstr_za (0, x1),
+ svstr_za (0, x1))
+
+/*
+** str_za_1:
+** mov (w1[2-5]), #?1
+** str za\[\1, 0\], \[x1(?:, #0, mul vl)?\]
+** ret
+*/
+TEST_STORE_ZA (str_za_1,
+ svstr_za (1, x1),
+ svstr_za (1, x1))
+
+/*
+** str_za_w0:
+** mov (w1[2-5]), w0
+** str za\[\1, 0\], \[x1(?:, #0, mul vl)?\]
+** ret
+*/
+TEST_STORE_ZA (str_za_w0,
+ svstr_za (w0, x1),
+ svstr_za (w0, x1))
+
+/*
+** str_za_w0_1_vnum:
+** mov (w1[2-5]), w0
+** str za\[\1, 1\], \[x1, #1, mul vl\]
+** ret
+*/
+TEST_STORE_ZA (str_za_w0_1_vnum,
+ svstr_za (w0 + 1, x1 + svcntsb ()),
+ svstr_za (w0 + 1, x1 + svcntsb ()))
+
+/*
+** str_za_w0p2:
+** add (w1[2-5]), w0, #?2
+** str za\[\1, 0\], \[x1(?:, #0, mul vl)?\]
+** ret
+*/
+TEST_STORE_ZA (str_za_w0p2,
+ svstr_za (w0 + 2, x1),
+ svstr_za (w0 + 2, x1))
+
+/*
+** str_za_offset:
+** (
+** mov (w1[2-5]), w0
+** add (x[0-9]+), x1, #?1
+** str za\[\1, 0\], \[\2(?:, #0, mul vl)?\]
+** |
+** add (x[0-9]+), x1, #?1
+** mov (w1[2-5]), w0
+** str za\[\4, 0\], \[\3(?:, #0, mul vl)?\]
+** )
+** ret
+*/
+TEST_STORE_ZA (str_za_offset,
+ svstr_za (w0, x1 + 1),
+ svstr_za (w0, x1 + 1))
--- /dev/null
+/* { dg-final { check-function-bodies "**" "" "-DCHECK_ASM" } } */
+
+#include "test_sme_acle.h"
+
+/*
+** sumopa_za32_s8_0_p0_p1_z0_z4:
+** sumopa za0\.s, p0/m, p1/m, z0\.b, z4\.b
+** ret
+*/
+TEST_DUAL_ZA (sumopa_za32_s8_0_p0_p1_z0_z4, svint8_t, svuint8_t,
+ svsumopa_za32_s8_m (0, p0, p1, z0, z4),
+ svsumopa_za32_m (0, p0, p1, z0, z4))
+
+/*
+** sumopa_za32_s8_0_p1_p0_z4_z0:
+** sumopa za0\.s, p1/m, p0/m, z4\.b, z0\.b
+** ret
+*/
+TEST_DUAL_ZA (sumopa_za32_s8_0_p1_p0_z4_z0, svuint8_t, svint8_t,
+ svsumopa_za32_s8_m (0, p1, p0, z4, z0),
+ svsumopa_za32_m (0, p1, p0, z4, z0))
+
+/*
+** sumopa_za32_s8_3_p0_p1_z0_z4:
+** sumopa za3\.s, p0/m, p1/m, z0\.b, z4\.b
+** ret
+*/
+TEST_DUAL_ZA (sumopa_za32_s8_3_p0_p1_z0_z4, svint8_t, svuint8_t,
+ svsumopa_za32_s8_m (3, p0, p1, z0, z4),
+ svsumopa_za32_m (3, p0, p1, z0, z4))
--- /dev/null
+/* { dg-final { check-function-bodies "**" "" "-DCHECK_ASM" } } */
+
+#include "test_sme_acle.h"
+
+#pragma GCC target "+sme-i16i64"
+
+/*
+** sumopa_za64_s16_0_p0_p1_z0_z4:
+** sumopa za0\.d, p0/m, p1/m, z0\.h, z4\.h
+** ret
+*/
+TEST_DUAL_ZA (sumopa_za64_s16_0_p0_p1_z0_z4, svint16_t, svuint16_t,
+ svsumopa_za64_s16_m (0, p0, p1, z0, z4),
+ svsumopa_za64_m (0, p0, p1, z0, z4))
+
+/*
+** sumopa_za64_s16_0_p1_p0_z4_z0:
+** sumopa za0\.d, p1/m, p0/m, z4\.h, z0\.h
+** ret
+*/
+TEST_DUAL_ZA (sumopa_za64_s16_0_p1_p0_z4_z0, svuint16_t, svint16_t,
+ svsumopa_za64_s16_m (0, p1, p0, z4, z0),
+ svsumopa_za64_m (0, p1, p0, z4, z0))
+
+/*
+** sumopa_za64_s16_7_p0_p1_z0_z4:
+** sumopa za7\.d, p0/m, p1/m, z0\.h, z4\.h
+** ret
+*/
+TEST_DUAL_ZA (sumopa_za64_s16_7_p0_p1_z0_z4, svint16_t, svuint16_t,
+ svsumopa_za64_s16_m (7, p0, p1, z0, z4),
+ svsumopa_za64_m (7, p0, p1, z0, z4))
--- /dev/null
+/* { dg-final { check-function-bodies "**" "" "-DCHECK_ASM" } } */
+
+#include "test_sme_acle.h"
+
+/*
+** sumops_za32_s8_0_p0_p1_z0_z4:
+** sumops za0\.s, p0/m, p1/m, z0\.b, z4\.b
+** ret
+*/
+TEST_DUAL_ZA (sumops_za32_s8_0_p0_p1_z0_z4, svint8_t, svuint8_t,
+ svsumops_za32_s8_m (0, p0, p1, z0, z4),
+ svsumops_za32_m (0, p0, p1, z0, z4))
+
+/*
+** sumops_za32_s8_0_p1_p0_z4_z0:
+** sumops za0\.s, p1/m, p0/m, z4\.b, z0\.b
+** ret
+*/
+TEST_DUAL_ZA (sumops_za32_s8_0_p1_p0_z4_z0, svuint8_t, svint8_t,
+ svsumops_za32_s8_m (0, p1, p0, z4, z0),
+ svsumops_za32_m (0, p1, p0, z4, z0))
+
+/*
+** sumops_za32_s8_3_p0_p1_z0_z4:
+** sumops za3\.s, p0/m, p1/m, z0\.b, z4\.b
+** ret
+*/
+TEST_DUAL_ZA (sumops_za32_s8_3_p0_p1_z0_z4, svint8_t, svuint8_t,
+ svsumops_za32_s8_m (3, p0, p1, z0, z4),
+ svsumops_za32_m (3, p0, p1, z0, z4))
--- /dev/null
+/* { dg-final { check-function-bodies "**" "" "-DCHECK_ASM" } } */
+
+#include "test_sme_acle.h"
+
+#pragma GCC target "+sme-i16i64"
+
+/*
+** sumops_za64_s16_0_p0_p1_z0_z4:
+** sumops za0\.d, p0/m, p1/m, z0\.h, z4\.h
+** ret
+*/
+TEST_DUAL_ZA (sumops_za64_s16_0_p0_p1_z0_z4, svint16_t, svuint16_t,
+ svsumops_za64_s16_m (0, p0, p1, z0, z4),
+ svsumops_za64_m (0, p0, p1, z0, z4))
+
+/*
+** sumops_za64_s16_0_p1_p0_z4_z0:
+** sumops za0\.d, p1/m, p0/m, z4\.h, z0\.h
+** ret
+*/
+TEST_DUAL_ZA (sumops_za64_s16_0_p1_p0_z4_z0, svuint16_t, svint16_t,
+ svsumops_za64_s16_m (0, p1, p0, z4, z0),
+ svsumops_za64_m (0, p1, p0, z4, z0))
+
+/*
+** sumops_za64_s16_7_p0_p1_z0_z4:
+** sumops za7\.d, p0/m, p1/m, z0\.h, z4\.h
+** ret
+*/
+TEST_DUAL_ZA (sumops_za64_s16_7_p0_p1_z0_z4, svint16_t, svuint16_t,
+ svsumops_za64_s16_m (7, p0, p1, z0, z4),
+ svsumops_za64_m (7, p0, p1, z0, z4))
--- /dev/null
+#ifndef TEST_SME_ACLE_H
+#define TEST_SME_ACLE_H 1
+
+#if (!defined(STREAMING_COMPATIBLE) \
+ && !defined(NON_STREAMING) \
+ && !defined(STREAMING))
+#define STREAMING
+#endif
+
+#if !defined(NO_SHARED_ZA)
+#define SHARED_ZA
+#endif
+
+#include "../../sve/acle/asm/test_sve_acle.h"
+
+#include <arm_sme.h>
+
+#define TEST_LOAD_ZA(NAME, CODE1, CODE2) \
+ PROTO (NAME, void, (svbool_t p0, int32_t w0, const char *x1, \
+ uint64_t x2)) \
+ { \
+ INVOKE (CODE1, CODE2); \
+ }
+
+#define TEST_STORE_ZA(NAME, CODE1, CODE2) \
+ PROTO (NAME, void, (svbool_t p0, int32_t w0, char *x1, \
+ uint64_t x2)) \
+ { \
+ INVOKE (CODE1, CODE2); \
+ }
+
+#define TEST_READ_ZA(NAME, TYPE, CODE1, CODE2) \
+ PROTO (NAME, TYPE, (TYPE z0, TYPE z1, svbool_t p0, \
+ int32_t w0)) \
+ { \
+ INVOKE (CODE1, CODE2); \
+ return z0; \
+ }
+
+#define TEST_WRITE_ZA(NAME, TYPE, CODE1, CODE2) \
+ PROTO (NAME, void, (TYPE z0, TYPE z1, svbool_t p0, \
+ int32_t w0)) \
+ { \
+ INVOKE (CODE1, CODE2); \
+ }
+
+#define TEST_UNIFORM_ZA(NAME, TYPE, CODE1, CODE2) \
+ PROTO (NAME, void, (TYPE z0, TYPE z1, svbool_t p0, \
+ svbool_t p1)) \
+ { \
+ INVOKE (CODE1, CODE2); \
+ }
+
+#define TEST_DUAL_ZA(NAME, TYPE1, TYPE2, CODE1, CODE2) \
+ PROTO (NAME, void, (TYPE1 z0, TYPE1 z1, TYPE1 z2, TYPE1 z3, \
+ TYPE2 z4, TYPE2 z5, TYPE2 z6, TYPE2 z7, \
+ svbool_t p0, svbool_t p1)) \
+ { \
+ INVOKE (CODE1, CODE2); \
+ }
+
+#endif
--- /dev/null
+/* { dg-final { check-function-bodies "**" "" "-DCHECK_ASM" } } */
+
+#define STREAMING_COMPATIBLE
+#include "test_sme_acle.h"
+
+/*
+** undef_za_1:
+** ret
+*/
+PROTO (undef_za_1, void, ()) { svundef_za (); }
+
+/*
+** undef_za_2:
+** ret
+*/
+PROTO (undef_za_2, void, ())
+{
+ svzero_za ();
+ svundef_za ();
+}
+
+/*
+** undef_za_3:
+** mov (w1[2-5]), (?:wzr|#?0)
+** str za\[\1, 0\], \[x0(?:, #0, mul vl)\]
+** ret
+*/
+PROTO (undef_za_3, void, (void *ptr))
+{
+ svzero_za ();
+ svundef_za ();
+ svstr_za (0, ptr);
+}
--- /dev/null
+/* { dg-final { check-function-bodies "**" "" "-DCHECK_ASM" } } */
+
+#include "test_sme_acle.h"
+
+/*
+** usmopa_za32_u8_0_p0_p1_z0_z4:
+** usmopa za0\.s, p0/m, p1/m, z0\.b, z4\.b
+** ret
+*/
+TEST_DUAL_ZA (usmopa_za32_u8_0_p0_p1_z0_z4, svuint8_t, svint8_t,
+ svusmopa_za32_u8_m (0, p0, p1, z0, z4),
+ svusmopa_za32_m (0, p0, p1, z0, z4))
+
+/*
+** usmopa_za32_u8_0_p1_p0_z4_z0:
+** usmopa za0\.s, p1/m, p0/m, z4\.b, z0\.b
+** ret
+*/
+TEST_DUAL_ZA (usmopa_za32_u8_0_p1_p0_z4_z0, svint8_t, svuint8_t,
+ svusmopa_za32_u8_m (0, p1, p0, z4, z0),
+ svusmopa_za32_m (0, p1, p0, z4, z0))
+
+/*
+** usmopa_za32_u8_3_p0_p1_z0_z4:
+** usmopa za3\.s, p0/m, p1/m, z0\.b, z4\.b
+** ret
+*/
+TEST_DUAL_ZA (usmopa_za32_u8_3_p0_p1_z0_z4, svuint8_t, svint8_t,
+ svusmopa_za32_u8_m (3, p0, p1, z0, z4),
+ svusmopa_za32_m (3, p0, p1, z0, z4))
--- /dev/null
+/* { dg-final { check-function-bodies "**" "" "-DCHECK_ASM" } } */
+
+#include "test_sme_acle.h"
+
+#pragma GCC target "+sme-i16i64"
+
+/*
+** usmopa_za64_u16_0_p0_p1_z0_z4:
+** usmopa za0\.d, p0/m, p1/m, z0\.h, z4\.h
+** ret
+*/
+TEST_DUAL_ZA (usmopa_za64_u16_0_p0_p1_z0_z4, svuint16_t, svint16_t,
+ svusmopa_za64_u16_m (0, p0, p1, z0, z4),
+ svusmopa_za64_m (0, p0, p1, z0, z4))
+
+/*
+** usmopa_za64_u16_0_p1_p0_z4_z0:
+** usmopa za0\.d, p1/m, p0/m, z4\.h, z0\.h
+** ret
+*/
+TEST_DUAL_ZA (usmopa_za64_u16_0_p1_p0_z4_z0, svint16_t, svuint16_t,
+ svusmopa_za64_u16_m (0, p1, p0, z4, z0),
+ svusmopa_za64_m (0, p1, p0, z4, z0))
+
+/*
+** usmopa_za64_u16_7_p0_p1_z0_z4:
+** usmopa za7\.d, p0/m, p1/m, z0\.h, z4\.h
+** ret
+*/
+TEST_DUAL_ZA (usmopa_za64_u16_7_p0_p1_z0_z4, svuint16_t, svint16_t,
+ svusmopa_za64_u16_m (7, p0, p1, z0, z4),
+ svusmopa_za64_m (7, p0, p1, z0, z4))
--- /dev/null
+/* { dg-final { check-function-bodies "**" "" "-DCHECK_ASM" } } */
+
+#include "test_sme_acle.h"
+
+/*
+** usmops_za32_u8_0_p0_p1_z0_z4:
+** usmops za0\.s, p0/m, p1/m, z0\.b, z4\.b
+** ret
+*/
+TEST_DUAL_ZA (usmops_za32_u8_0_p0_p1_z0_z4, svuint8_t, svint8_t,
+ svusmops_za32_u8_m (0, p0, p1, z0, z4),
+ svusmops_za32_m (0, p0, p1, z0, z4))
+
+/*
+** usmops_za32_u8_0_p1_p0_z4_z0:
+** usmops za0\.s, p1/m, p0/m, z4\.b, z0\.b
+** ret
+*/
+TEST_DUAL_ZA (usmops_za32_u8_0_p1_p0_z4_z0, svint8_t, svuint8_t,
+ svusmops_za32_u8_m (0, p1, p0, z4, z0),
+ svusmops_za32_m (0, p1, p0, z4, z0))
+
+/*
+** usmops_za32_u8_3_p0_p1_z0_z4:
+** usmops za3\.s, p0/m, p1/m, z0\.b, z4\.b
+** ret
+*/
+TEST_DUAL_ZA (usmops_za32_u8_3_p0_p1_z0_z4, svuint8_t, svint8_t,
+ svusmops_za32_u8_m (3, p0, p1, z0, z4),
+ svusmops_za32_m (3, p0, p1, z0, z4))
--- /dev/null
+/* { dg-final { check-function-bodies "**" "" "-DCHECK_ASM" } } */
+
+#include "test_sme_acle.h"
+
+#pragma GCC target "+sme-i16i64"
+
+/*
+** usmops_za64_u16_0_p0_p1_z0_z4:
+** usmops za0\.d, p0/m, p1/m, z0\.h, z4\.h
+** ret
+*/
+TEST_DUAL_ZA (usmops_za64_u16_0_p0_p1_z0_z4, svuint16_t, svint16_t,
+ svusmops_za64_u16_m (0, p0, p1, z0, z4),
+ svusmops_za64_m (0, p0, p1, z0, z4))
+
+/*
+** usmops_za64_u16_0_p1_p0_z4_z0:
+** usmops za0\.d, p1/m, p0/m, z4\.h, z0\.h
+** ret
+*/
+TEST_DUAL_ZA (usmops_za64_u16_0_p1_p0_z4_z0, svint16_t, svuint16_t,
+ svusmops_za64_u16_m (0, p1, p0, z4, z0),
+ svusmops_za64_m (0, p1, p0, z4, z0))
+
+/*
+** usmops_za64_u16_7_p0_p1_z0_z4:
+** usmops za7\.d, p0/m, p1/m, z0\.h, z4\.h
+** ret
+*/
+TEST_DUAL_ZA (usmops_za64_u16_7_p0_p1_z0_z4, svuint16_t, svint16_t,
+ svusmops_za64_u16_m (7, p0, p1, z0, z4),
+ svusmops_za64_m (7, p0, p1, z0, z4))
--- /dev/null
+/* { dg-final { check-function-bodies "**" "" "-DCHECK_ASM" } } */
+
+#include "test_sme_acle.h"
+
+/*
+** write_za128_s8_0_0_z0:
+** mov (w1[2-5]), (?:wzr|#?0)
+** mova za0h\.q\[\1, 0\], p0/m, z0\.q
+** ret
+*/
+TEST_WRITE_ZA (write_za128_s8_0_0_z0, svint8_t,
+ svwrite_hor_za128_s8_m (0, 0, p0, z0),
+ svwrite_hor_za128_m (0, 0, p0, z0))
+
+/*
+** write_za128_s8_0_1_z0:
+** mov (w1[2-5]), #?1
+** mova za0h\.q\[\1, 0\], p0/m, z0\.q
+** ret
+*/
+TEST_WRITE_ZA (write_za128_s8_0_1_z0, svint8_t,
+ svwrite_hor_za128_s8_m (0, 1, p0, z0),
+ svwrite_hor_za128_m (0, 1, p0, z0))
+
+/*
+** write_za128_s8_0_w0_z0:
+** mov (w1[2-5]), w0
+** mova za0h\.q\[\1, 0\], p0/m, z0\.q
+** ret
+*/
+TEST_WRITE_ZA (write_za128_s8_0_w0_z0, svint8_t,
+ svwrite_hor_za128_s8_m (0, w0, p0, z0),
+ svwrite_hor_za128_m (0, w0, p0, z0))
+
+/*
+** write_za128_s8_0_w0p1_z0:
+** add (w1[2-5]), w0, #?1
+** mova za0h\.q\[\1, 0\], p0/m, z0\.q
+** ret
+*/
+TEST_WRITE_ZA (write_za128_s8_0_w0p1_z0, svint8_t,
+ svwrite_hor_za128_s8_m (0, w0 + 1, p0, z0),
+ svwrite_hor_za128_m (0, w0 + 1, p0, z0))
+
+/*
+** write_za128_s8_0_w0m1_z0:
+** sub (w1[2-5]), w0, #?1
+** mova za0h\.q\[\1, 0\], p0/m, z0\.q
+** ret
+*/
+TEST_WRITE_ZA (write_za128_s8_0_w0m1_z0, svint8_t,
+ svwrite_hor_za128_s8_m (0, w0 - 1, p0, z0),
+ svwrite_hor_za128_m (0, w0 - 1, p0, z0))
+
+/*
+** write_za128_s8_1_w0_z0:
+** mov (w1[2-5]), w0
+** mova za1h\.q\[\1, 0\], p0/m, z0\.q
+** ret
+*/
+TEST_WRITE_ZA (write_za128_s8_1_w0_z0, svint8_t,
+ svwrite_hor_za128_s8_m (1, w0, p0, z0),
+ svwrite_hor_za128_m (1, w0, p0, z0))
+
+/*
+** write_za128_s8_15_w0_z0:
+** mov (w1[2-5]), w0
+** mova za15h\.q\[\1, 0\], p0/m, z0\.q
+** ret
+*/
+TEST_WRITE_ZA (write_za128_s8_15_w0_z0, svint8_t,
+ svwrite_hor_za128_s8_m (15, w0, p0, z0),
+ svwrite_hor_za128_m (15, w0, p0, z0))
+
+/*
+** write_za128_s8_0_w0_z1:
+** mov (w1[2-5]), w0
+** mova za0h\.q\[\1, 0\], p0/m, z1\.q
+** ret
+*/
+TEST_WRITE_ZA (write_za128_s8_0_w0_z1, svint8_t,
+ svwrite_hor_za128_s8_m (0, w0, p0, z1),
+ svwrite_hor_za128_m (0, w0, p0, z1))
+
+/*
+** write_za128_u8_0_w0_z0:
+** mov (w1[2-5]), w0
+** mova za0h\.q\[\1, 0\], p0/m, z0\.q
+** ret
+*/
+TEST_WRITE_ZA (write_za128_u8_0_w0_z0, svuint8_t,
+ svwrite_hor_za128_u8_m (0, w0, p0, z0),
+ svwrite_hor_za128_m (0, w0, p0, z0))
+
+/*
+** write_za128_s16_0_w0_z0:
+** mov (w1[2-5]), w0
+** mova za0h\.q\[\1, 0\], p0/m, z0\.q
+** ret
+*/
+TEST_WRITE_ZA (write_za128_s16_0_w0_z0, svint16_t,
+ svwrite_hor_za128_s16_m (0, w0, p0, z0),
+ svwrite_hor_za128_m (0, w0, p0, z0))
+
+/*
+** write_za128_u16_0_w0_z0:
+** mov (w1[2-5]), w0
+** mova za0h\.q\[\1, 0\], p0/m, z0\.q
+** ret
+*/
+TEST_WRITE_ZA (write_za128_u16_0_w0_z0, svuint16_t,
+ svwrite_hor_za128_u16_m (0, w0, p0, z0),
+ svwrite_hor_za128_m (0, w0, p0, z0))
+
+/*
+** write_za128_f16_0_w0_z0:
+** mov (w1[2-5]), w0
+** mova za0h\.q\[\1, 0\], p0/m, z0\.q
+** ret
+*/
+TEST_WRITE_ZA (write_za128_f16_0_w0_z0, svfloat16_t,
+ svwrite_hor_za128_f16_m (0, w0, p0, z0),
+ svwrite_hor_za128_m (0, w0, p0, z0))
+
+/*
+** write_za128_bf16_0_w0_z0:
+** mov (w1[2-5]), w0
+** mova za0h\.q\[\1, 0\], p0/m, z0\.q
+** ret
+*/
+TEST_WRITE_ZA (write_za128_bf16_0_w0_z0, svbfloat16_t,
+ svwrite_hor_za128_bf16_m (0, w0, p0, z0),
+ svwrite_hor_za128_m (0, w0, p0, z0))
+
+/*
+** write_za128_s32_0_w0_z0:
+** mov (w1[2-5]), w0
+** mova za0h\.q\[\1, 0\], p0/m, z0\.q
+** ret
+*/
+TEST_WRITE_ZA (write_za128_s32_0_w0_z0, svint32_t,
+ svwrite_hor_za128_s32_m (0, w0, p0, z0),
+ svwrite_hor_za128_m (0, w0, p0, z0))
+
+/*
+** write_za128_u32_0_w0_z0:
+** mov (w1[2-5]), w0
+** mova za0h\.q\[\1, 0\], p0/m, z0\.q
+** ret
+*/
+TEST_WRITE_ZA (write_za128_u32_0_w0_z0, svuint32_t,
+ svwrite_hor_za128_u32_m (0, w0, p0, z0),
+ svwrite_hor_za128_m (0, w0, p0, z0))
+
+/*
+** write_za128_f32_0_w0_z0:
+** mov (w1[2-5]), w0
+** mova za0h\.q\[\1, 0\], p0/m, z0\.q
+** ret
+*/
+TEST_WRITE_ZA (write_za128_f32_0_w0_z0, svfloat32_t,
+ svwrite_hor_za128_f32_m (0, w0, p0, z0),
+ svwrite_hor_za128_m (0, w0, p0, z0))
+
+/*
+** write_za128_s64_0_w0_z0:
+** mov (w1[2-5]), w0
+** mova za0h\.q\[\1, 0\], p0/m, z0\.q
+** ret
+*/
+TEST_WRITE_ZA (write_za128_s64_0_w0_z0, svint64_t,
+ svwrite_hor_za128_s64_m (0, w0, p0, z0),
+ svwrite_hor_za128_m (0, w0, p0, z0))
+
+/*
+** write_za128_u64_0_w0_z0:
+** mov (w1[2-5]), w0
+** mova za0h\.q\[\1, 0\], p0/m, z0\.q
+** ret
+*/
+TEST_WRITE_ZA (write_za128_u64_0_w0_z0, svuint64_t,
+ svwrite_hor_za128_u64_m (0, w0, p0, z0),
+ svwrite_hor_za128_m (0, w0, p0, z0))
+
+/*
+** write_za128_f64_0_w0_z0:
+** mov (w1[2-5]), w0
+** mova za0h\.q\[\1, 0\], p0/m, z0\.q
+** ret
+*/
+TEST_WRITE_ZA (write_za128_f64_0_w0_z0, svfloat64_t,
+ svwrite_hor_za128_f64_m (0, w0, p0, z0),
+ svwrite_hor_za128_m (0, w0, p0, z0))
--- /dev/null
+/* { dg-final { check-function-bodies "**" "" "-DCHECK_ASM" } } */
+
+#include "test_sme_acle.h"
+
+/*
+** write_za16_s16_0_0_z0:
+** mov (w1[2-5]), (?:wzr|#?0)
+** mova za0h\.h\[\1, 0\], p0/m, z0\.h
+** ret
+*/
+TEST_WRITE_ZA (write_za16_s16_0_0_z0, svint16_t,
+ svwrite_hor_za16_s16_m (0, 0, p0, z0),
+ svwrite_hor_za16_m (0, 0, p0, z0))
+
+/*
+** write_za16_s16_0_1_z0:
+** mov (w1[2-5]), #?1
+** mova za0h\.h\[\1, 0\], p0/m, z0\.h
+** ret
+*/
+TEST_WRITE_ZA (write_za16_s16_0_1_z0, svint16_t,
+ svwrite_hor_za16_s16_m (0, 1, p0, z0),
+ svwrite_hor_za16_m (0, 1, p0, z0))
+
+/*
+** write_za16_s16_0_w0_z0:
+** mov (w1[2-5]), w0
+** mova za0h\.h\[\1, 0\], p0/m, z0\.h
+** ret
+*/
+TEST_WRITE_ZA (write_za16_s16_0_w0_z0, svint16_t,
+ svwrite_hor_za16_s16_m (0, w0, p0, z0),
+ svwrite_hor_za16_m (0, w0, p0, z0))
+
+/*
+** write_za16_s16_0_w0p1_z0:
+** mov (w1[2-5]), w0
+** mova za0h\.h\[\1, 1\], p0/m, z0\.h
+** ret
+*/
+TEST_WRITE_ZA (write_za16_s16_0_w0p1_z0, svint16_t,
+ svwrite_hor_za16_s16_m (0, w0 + 1, p0, z0),
+ svwrite_hor_za16_m (0, w0 + 1, p0, z0))
+
+/*
+** write_za16_s16_0_w0p7_z0:
+** mov (w1[2-5]), w0
+** mova za0h\.h\[\1, 7\], p0/m, z0\.h
+** ret
+*/
+TEST_WRITE_ZA (write_za16_s16_0_w0p7_z0, svint16_t,
+ svwrite_hor_za16_s16_m (0, w0 + 7, p0, z0),
+ svwrite_hor_za16_m (0, w0 + 7, p0, z0))
+
+/*
+** write_za16_s16_0_w0p8_z0:
+** add (w1[2-5]), w0, #?8
+** mova za0h\.h\[\1, 0\], p0/m, z0\.h
+** ret
+*/
+TEST_WRITE_ZA (write_za16_s16_0_w0p8_z0, svint16_t,
+ svwrite_hor_za16_s16_m (0, w0 + 8, p0, z0),
+ svwrite_hor_za16_m (0, w0 + 8, p0, z0))
+
+/*
+** write_za16_s16_0_w0m1_z0:
+** sub (w1[2-5]), w0, #?1
+** mova za0h\.h\[\1, 0\], p0/m, z0\.h
+** ret
+*/
+TEST_WRITE_ZA (write_za16_s16_0_w0m1_z0, svint16_t,
+ svwrite_hor_za16_s16_m (0, w0 - 1, p0, z0),
+ svwrite_hor_za16_m (0, w0 - 1, p0, z0))
+
+/*
+** write_za16_s16_1_w0_z0:
+** mov (w1[2-5]), w0
+** mova za1h\.h\[\1, 0\], p0/m, z0\.h
+** ret
+*/
+TEST_WRITE_ZA (write_za16_s16_1_w0_z0, svint16_t,
+ svwrite_hor_za16_s16_m (1, w0, p0, z0),
+ svwrite_hor_za16_m (1, w0, p0, z0))
+
+/*
+** write_za16_s16_1_w0p7_z0:
+** mov (w1[2-5]), w0
+** mova za1h\.h\[\1, 7\], p0/m, z0\.h
+** ret
+*/
+TEST_WRITE_ZA (write_za16_s16_1_w0p7_z0, svint16_t,
+ svwrite_hor_za16_s16_m (1, w0 + 7, p0, z0),
+ svwrite_hor_za16_m (1, w0 + 7, p0, z0))
+
+/*
+** write_za16_s16_0_w0_z1:
+** mov (w1[2-5]), w0
+** mova za0h\.h\[\1, 0\], p0/m, z1\.h
+** ret
+*/
+TEST_WRITE_ZA (write_za16_s16_0_w0_z1, svint16_t,
+ svwrite_hor_za16_s16_m (0, w0, p0, z1),
+ svwrite_hor_za16_m (0, w0, p0, z1))
+
+/*
+** write_za16_u16_0_w0_z0:
+** mov (w1[2-5]), w0
+** mova za0h\.h\[\1, 0\], p0/m, z0\.h
+** ret
+*/
+TEST_WRITE_ZA (write_za16_u16_0_w0_z0, svuint16_t,
+ svwrite_hor_za16_u16_m (0, w0, p0, z0),
+ svwrite_hor_za16_m (0, w0, p0, z0))
+
+/*
+** write_za16_f16_0_w0_z0:
+** mov (w1[2-5]), w0
+** mova za0h\.h\[\1, 0\], p0/m, z0\.h
+** ret
+*/
+TEST_WRITE_ZA (write_za16_f16_0_w0_z0, svfloat16_t,
+ svwrite_hor_za16_f16_m (0, w0, p0, z0),
+ svwrite_hor_za16_m (0, w0, p0, z0))
+
+/*
+** write_za16_bf16_0_w0_z0:
+** mov (w1[2-5]), w0
+** mova za0h\.h\[\1, 0\], p0/m, z0\.h
+** ret
+*/
+TEST_WRITE_ZA (write_za16_bf16_0_w0_z0, svbfloat16_t,
+ svwrite_hor_za16_bf16_m (0, w0, p0, z0),
+ svwrite_hor_za16_m (0, w0, p0, z0))
--- /dev/null
+/* { dg-final { check-function-bodies "**" "" "-DCHECK_ASM" } } */
+
+#include "test_sme_acle.h"
+
+/*
+** write_za32_s32_0_0_z0:
+** mov (w1[2-5]), (?:wzr|#?0)
+** mova za0h\.s\[\1, 0\], p0/m, z0\.s
+** ret
+*/
+TEST_WRITE_ZA (write_za32_s32_0_0_z0, svint32_t,
+ svwrite_hor_za32_s32_m (0, 0, p0, z0),
+ svwrite_hor_za32_m (0, 0, p0, z0))
+
+/*
+** write_za32_s32_0_1_z0:
+** mov (w1[2-5]), #?1
+** mova za0h\.s\[\1, 0\], p0/m, z0\.s
+** ret
+*/
+TEST_WRITE_ZA (write_za32_s32_0_1_z0, svint32_t,
+ svwrite_hor_za32_s32_m (0, 1, p0, z0),
+ svwrite_hor_za32_m (0, 1, p0, z0))
+
+/*
+** write_za32_s32_0_w0_z0:
+** mov (w1[2-5]), w0
+** mova za0h\.s\[\1, 0\], p0/m, z0\.s
+** ret
+*/
+TEST_WRITE_ZA (write_za32_s32_0_w0_z0, svint32_t,
+ svwrite_hor_za32_s32_m (0, w0, p0, z0),
+ svwrite_hor_za32_m (0, w0, p0, z0))
+
+/*
+** write_za32_s32_0_w0p1_z0:
+** mov (w1[2-5]), w0
+** mova za0h\.s\[\1, 1\], p0/m, z0\.s
+** ret
+*/
+TEST_WRITE_ZA (write_za32_s32_0_w0p1_z0, svint32_t,
+ svwrite_hor_za32_s32_m (0, w0 + 1, p0, z0),
+ svwrite_hor_za32_m (0, w0 + 1, p0, z0))
+
+/*
+** write_za32_s32_0_w0p3_z0:
+** mov (w1[2-5]), w0
+** mova za0h\.s\[\1, 3\], p0/m, z0\.s
+** ret
+*/
+TEST_WRITE_ZA (write_za32_s32_0_w0p3_z0, svint32_t,
+ svwrite_hor_za32_s32_m (0, w0 + 3, p0, z0),
+ svwrite_hor_za32_m (0, w0 + 3, p0, z0))
+
+/*
+** write_za32_s32_0_w0p4_z0:
+** add (w1[2-5]), w0, #?4
+** mova za0h\.s\[\1, 0\], p0/m, z0\.s
+** ret
+*/
+TEST_WRITE_ZA (write_za32_s32_0_w0p4_z0, svint32_t,
+ svwrite_hor_za32_s32_m (0, w0 + 4, p0, z0),
+ svwrite_hor_za32_m (0, w0 + 4, p0, z0))
+
+/*
+** write_za32_s32_0_w0m1_z0:
+** sub (w1[2-5]), w0, #?1
+** mova za0h\.s\[\1, 0\], p0/m, z0\.s
+** ret
+*/
+TEST_WRITE_ZA (write_za32_s32_0_w0m1_z0, svint32_t,
+ svwrite_hor_za32_s32_m (0, w0 - 1, p0, z0),
+ svwrite_hor_za32_m (0, w0 - 1, p0, z0))
+
+/*
+** write_za32_s32_1_w0_z0:
+** mov (w1[2-5]), w0
+** mova za1h\.s\[\1, 0\], p0/m, z0\.s
+** ret
+*/
+TEST_WRITE_ZA (write_za32_s32_1_w0_z0, svint32_t,
+ svwrite_hor_za32_s32_m (1, w0, p0, z0),
+ svwrite_hor_za32_m (1, w0, p0, z0))
+
+/*
+** write_za32_s32_1_w0p3_z0:
+** mov (w1[2-5]), w0
+** mova za1h\.s\[\1, 3\], p0/m, z0\.s
+** ret
+*/
+TEST_WRITE_ZA (write_za32_s32_1_w0p3_z0, svint32_t,
+ svwrite_hor_za32_s32_m (1, w0 + 3, p0, z0),
+ svwrite_hor_za32_m (1, w0 + 3, p0, z0))
+
+/*
+** write_za32_s32_3_w0_z0:
+** mov (w1[2-5]), w0
+** mova za3h\.s\[\1, 0\], p0/m, z0\.s
+** ret
+*/
+TEST_WRITE_ZA (write_za32_s32_3_w0_z0, svint32_t,
+ svwrite_hor_za32_s32_m (3, w0, p0, z0),
+ svwrite_hor_za32_m (3, w0, p0, z0))
+
+/*
+** write_za32_s32_3_w0p3_z0:
+** mov (w1[2-5]), w0
+** mova za3h\.s\[\1, 3\], p0/m, z0\.s
+** ret
+*/
+TEST_WRITE_ZA (write_za32_s32_3_w0p3_z0, svint32_t,
+ svwrite_hor_za32_s32_m (3, w0 + 3, p0, z0),
+ svwrite_hor_za32_m (3, w0 + 3, p0, z0))
+
+/*
+** write_za32_s32_0_w0_z1:
+** mov (w1[2-5]), w0
+** mova za0h\.s\[\1, 0\], p0/m, z1\.s
+** ret
+*/
+TEST_WRITE_ZA (write_za32_s32_0_w0_z1, svint32_t,
+ svwrite_hor_za32_s32_m (0, w0, p0, z1),
+ svwrite_hor_za32_m (0, w0, p0, z1))
+
+/*
+** write_za32_u32_0_w0_z0:
+** mov (w1[2-5]), w0
+** mova za0h\.s\[\1, 0\], p0/m, z0\.s
+** ret
+*/
+TEST_WRITE_ZA (write_za32_u32_0_w0_z0, svuint32_t,
+ svwrite_hor_za32_u32_m (0, w0, p0, z0),
+ svwrite_hor_za32_m (0, w0, p0, z0))
+
+/*
+** write_za32_f32_0_w0_z0:
+** mov (w1[2-5]), w0
+** mova za0h\.s\[\1, 0\], p0/m, z0\.s
+** ret
+*/
+TEST_WRITE_ZA (write_za32_f32_0_w0_z0, svfloat32_t,
+ svwrite_hor_za32_f32_m (0, w0, p0, z0),
+ svwrite_hor_za32_m (0, w0, p0, z0))
--- /dev/null
+/* { dg-final { check-function-bodies "**" "" "-DCHECK_ASM" } } */
+
+#include "test_sme_acle.h"
+
+/*
+** write_za64_s64_0_0_z0:
+** mov (w1[2-5]), (?:wzr|#?0)
+** mova za0h\.d\[\1, 0\], p0/m, z0\.d
+** ret
+*/
+TEST_WRITE_ZA (write_za64_s64_0_0_z0, svint64_t,
+ svwrite_hor_za64_s64_m (0, 0, p0, z0),
+ svwrite_hor_za64_m (0, 0, p0, z0))
+
+/*
+** write_za64_s64_0_1_z0:
+** mov (w1[2-5]), #?1
+** mova za0h\.d\[\1, 0\], p0/m, z0\.d
+** ret
+*/
+TEST_WRITE_ZA (write_za64_s64_0_1_z0, svint64_t,
+ svwrite_hor_za64_s64_m (0, 1, p0, z0),
+ svwrite_hor_za64_m (0, 1, p0, z0))
+
+/*
+** write_za64_s64_0_w0_z0:
+** mov (w1[2-5]), w0
+** mova za0h\.d\[\1, 0\], p0/m, z0\.d
+** ret
+*/
+TEST_WRITE_ZA (write_za64_s64_0_w0_z0, svint64_t,
+ svwrite_hor_za64_s64_m (0, w0, p0, z0),
+ svwrite_hor_za64_m (0, w0, p0, z0))
+
+/*
+** write_za64_s64_0_w0p1_z0:
+** mov (w1[2-5]), w0
+** mova za0h\.d\[\1, 1\], p0/m, z0\.d
+** ret
+*/
+TEST_WRITE_ZA (write_za64_s64_0_w0p1_z0, svint64_t,
+ svwrite_hor_za64_s64_m (0, w0 + 1, p0, z0),
+ svwrite_hor_za64_m (0, w0 + 1, p0, z0))
+
+/*
+** write_za64_s64_0_w0p2_z0:
+** add (w1[2-5]), w0, #?2
+** mova za0h\.d\[\1, 0\], p0/m, z0\.d
+** ret
+*/
+TEST_WRITE_ZA (write_za64_s64_0_w0p2_z0, svint64_t,
+ svwrite_hor_za64_s64_m (0, w0 + 2, p0, z0),
+ svwrite_hor_za64_m (0, w0 + 2, p0, z0))
+
+/*
+** write_za64_s64_0_w0m1_z0:
+** sub (w1[2-5]), w0, #?1
+** mova za0h\.d\[\1, 0\], p0/m, z0\.d
+** ret
+*/
+TEST_WRITE_ZA (write_za64_s64_0_w0m1_z0, svint64_t,
+ svwrite_hor_za64_s64_m (0, w0 - 1, p0, z0),
+ svwrite_hor_za64_m (0, w0 - 1, p0, z0))
+
+/*
+** write_za64_s64_1_w0_z0:
+** mov (w1[2-5]), w0
+** mova za1h\.d\[\1, 0\], p0/m, z0\.d
+** ret
+*/
+TEST_WRITE_ZA (write_za64_s64_1_w0_z0, svint64_t,
+ svwrite_hor_za64_s64_m (1, w0, p0, z0),
+ svwrite_hor_za64_m (1, w0, p0, z0))
+
+/*
+** write_za64_s64_1_w0p1_z0:
+** mov (w1[2-5]), w0
+** mova za1h\.d\[\1, 1\], p0/m, z0\.d
+** ret
+*/
+TEST_WRITE_ZA (write_za64_s64_1_w0p1_z0, svint64_t,
+ svwrite_hor_za64_s64_m (1, w0 + 1, p0, z0),
+ svwrite_hor_za64_m (1, w0 + 1, p0, z0))
+
+/*
+** write_za64_s64_7_w0_z0:
+** mov (w1[2-5]), w0
+** mova za7h\.d\[\1, 0\], p0/m, z0\.d
+** ret
+*/
+TEST_WRITE_ZA (write_za64_s64_7_w0_z0, svint64_t,
+ svwrite_hor_za64_s64_m (7, w0, p0, z0),
+ svwrite_hor_za64_m (7, w0, p0, z0))
+
+/*
+** write_za64_s64_7_w0p1_z0:
+** mov (w1[2-5]), w0
+** mova za7h\.d\[\1, 1\], p0/m, z0\.d
+** ret
+*/
+TEST_WRITE_ZA (write_za64_s64_7_w0p1_z0, svint64_t,
+ svwrite_hor_za64_s64_m (7, w0 + 1, p0, z0),
+ svwrite_hor_za64_m (7, w0 + 1, p0, z0))
+
+/*
+** write_za64_s64_0_w0_z1:
+** mov (w1[2-5]), w0
+** mova za0h\.d\[\1, 0\], p0/m, z1\.d
+** ret
+*/
+TEST_WRITE_ZA (write_za64_s64_0_w0_z1, svint64_t,
+ svwrite_hor_za64_s64_m (0, w0, p0, z1),
+ svwrite_hor_za64_m (0, w0, p0, z1))
+
+/*
+** write_za64_u64_0_w0_z0:
+** mov (w1[2-5]), w0
+** mova za0h\.d\[\1, 0\], p0/m, z0\.d
+** ret
+*/
+TEST_WRITE_ZA (write_za64_u64_0_w0_z0, svuint64_t,
+ svwrite_hor_za64_u64_m (0, w0, p0, z0),
+ svwrite_hor_za64_m (0, w0, p0, z0))
+
+/*
+** write_za64_f64_0_w0_z0:
+** mov (w1[2-5]), w0
+** mova za0h\.d\[\1, 0\], p0/m, z0\.d
+** ret
+*/
+TEST_WRITE_ZA (write_za64_f64_0_w0_z0, svfloat64_t,
+ svwrite_hor_za64_f64_m (0, w0, p0, z0),
+ svwrite_hor_za64_m (0, w0, p0, z0))
--- /dev/null
+/* { dg-final { check-function-bodies "**" "" "-DCHECK_ASM" } } */
+
+#include "test_sme_acle.h"
+
+/*
+** write_za8_s8_0_0_z0:
+** mov (w1[2-5]), (?:wzr|#?0)
+** mova za0h\.b\[\1, 0\], p0/m, z0\.b
+** ret
+*/
+TEST_WRITE_ZA (write_za8_s8_0_0_z0, svint8_t,
+ svwrite_hor_za8_s8_m (0, 0, p0, z0),
+ svwrite_hor_za8_m (0, 0, p0, z0))
+
+/*
+** write_za8_s8_0_1_z0:
+** mov (w1[2-5]), #?1
+** mova za0h\.b\[\1, 0\], p0/m, z0\.b
+** ret
+*/
+TEST_WRITE_ZA (write_za8_s8_0_1_z0, svint8_t,
+ svwrite_hor_za8_s8_m (0, 1, p0, z0),
+ svwrite_hor_za8_m (0, 1, p0, z0))
+
+/*
+** write_za8_s8_0_w0_z0:
+** mov (w1[2-5]), w0
+** mova za0h\.b\[\1, 0\], p0/m, z0\.b
+** ret
+*/
+TEST_WRITE_ZA (write_za8_s8_0_w0_z0, svint8_t,
+ svwrite_hor_za8_s8_m (0, w0, p0, z0),
+ svwrite_hor_za8_m (0, w0, p0, z0))
+
+/*
+** write_za8_s8_0_w0p1_z0:
+** mov (w1[2-5]), w0
+** mova za0h\.b\[\1, 1\], p0/m, z0\.b
+** ret
+*/
+TEST_WRITE_ZA (write_za8_s8_0_w0p1_z0, svint8_t,
+ svwrite_hor_za8_s8_m (0, w0 + 1, p0, z0),
+ svwrite_hor_za8_m (0, w0 + 1, p0, z0))
+
+/*
+** write_za8_s8_0_w0p15_z0:
+** mov (w1[2-5]), w0
+** mova za0h\.b\[\1, 15\], p0/m, z0\.b
+** ret
+*/
+TEST_WRITE_ZA (write_za8_s8_0_w0p15_z0, svint8_t,
+ svwrite_hor_za8_s8_m (0, w0 + 15, p0, z0),
+ svwrite_hor_za8_m (0, w0 + 15, p0, z0))
+
+/*
+** write_za8_s8_0_w0p16_z0:
+** add (w1[2-5]), w0, #?16
+** mova za0h\.b\[\1, 0\], p0/m, z0\.b
+** ret
+*/
+TEST_WRITE_ZA (write_za8_s8_0_w0p16_z0, svint8_t,
+ svwrite_hor_za8_s8_m (0, w0 + 16, p0, z0),
+ svwrite_hor_za8_m (0, w0 + 16, p0, z0))
+
+/*
+** write_za8_s8_0_w0m1_z0:
+** sub (w1[2-5]), w0, #?1
+** mova za0h\.b\[\1, 0\], p0/m, z0\.b
+** ret
+*/
+TEST_WRITE_ZA (write_za8_s8_0_w0m1_z0, svint8_t,
+ svwrite_hor_za8_s8_m (0, w0 - 1, p0, z0),
+ svwrite_hor_za8_m (0, w0 - 1, p0, z0))
+
+/*
+** write_za8_s8_0_w0_z1:
+** mov (w1[2-5]), w0
+** mova za0h\.b\[\1, 0\], p0/m, z1\.b
+** ret
+*/
+TEST_WRITE_ZA (write_za8_s8_0_w0_z1, svint8_t,
+ svwrite_hor_za8_s8_m (0, w0, p0, z1),
+ svwrite_hor_za8_m (0, w0, p0, z1))
+
+/*
+** write_za8_u8_0_w0_z0:
+** mov (w1[2-5]), w0
+** mova za0h\.b\[\1, 0\], p0/m, z0\.b
+** ret
+*/
+TEST_WRITE_ZA (write_za8_u8_0_w0_z0, svuint8_t,
+ svwrite_hor_za8_u8_m (0, w0, p0, z0),
+ svwrite_hor_za8_m (0, w0, p0, z0))
--- /dev/null
+/* { dg-final { check-function-bodies "**" "" "-DCHECK_ASM" } } */
+
+#include "test_sme_acle.h"
+
+/*
+** write_za128_s8_0_0_z0:
+** mov (w1[2-5]), (?:wzr|#?0)
+** mova za0v\.q\[\1, 0\], p0/m, z0\.q
+** ret
+*/
+TEST_WRITE_ZA (write_za128_s8_0_0_z0, svint8_t,
+ svwrite_ver_za128_s8_m (0, 0, p0, z0),
+ svwrite_ver_za128_m (0, 0, p0, z0))
+
+/*
+** write_za128_s8_0_1_z0:
+** mov (w1[2-5]), #?1
+** mova za0v\.q\[\1, 0\], p0/m, z0\.q
+** ret
+*/
+TEST_WRITE_ZA (write_za128_s8_0_1_z0, svint8_t,
+ svwrite_ver_za128_s8_m (0, 1, p0, z0),
+ svwrite_ver_za128_m (0, 1, p0, z0))
+
+/*
+** write_za128_s8_0_w0_z0:
+** mov (w1[2-5]), w0
+** mova za0v\.q\[\1, 0\], p0/m, z0\.q
+** ret
+*/
+TEST_WRITE_ZA (write_za128_s8_0_w0_z0, svint8_t,
+ svwrite_ver_za128_s8_m (0, w0, p0, z0),
+ svwrite_ver_za128_m (0, w0, p0, z0))
+
+/*
+** write_za128_s8_0_w0p1_z0:
+** add (w1[2-5]), w0, #?1
+** mova za0v\.q\[\1, 0\], p0/m, z0\.q
+** ret
+*/
+TEST_WRITE_ZA (write_za128_s8_0_w0p1_z0, svint8_t,
+ svwrite_ver_za128_s8_m (0, w0 + 1, p0, z0),
+ svwrite_ver_za128_m (0, w0 + 1, p0, z0))
+
+/*
+** write_za128_s8_0_w0m1_z0:
+** sub (w1[2-5]), w0, #?1
+** mova za0v\.q\[\1, 0\], p0/m, z0\.q
+** ret
+*/
+TEST_WRITE_ZA (write_za128_s8_0_w0m1_z0, svint8_t,
+ svwrite_ver_za128_s8_m (0, w0 - 1, p0, z0),
+ svwrite_ver_za128_m (0, w0 - 1, p0, z0))
+
+/*
+** write_za128_s8_1_w0_z0:
+** mov (w1[2-5]), w0
+** mova za1v\.q\[\1, 0\], p0/m, z0\.q
+** ret
+*/
+TEST_WRITE_ZA (write_za128_s8_1_w0_z0, svint8_t,
+ svwrite_ver_za128_s8_m (1, w0, p0, z0),
+ svwrite_ver_za128_m (1, w0, p0, z0))
+
+/*
+** write_za128_s8_15_w0_z0:
+** mov (w1[2-5]), w0
+** mova za15v\.q\[\1, 0\], p0/m, z0\.q
+** ret
+*/
+TEST_WRITE_ZA (write_za128_s8_15_w0_z0, svint8_t,
+ svwrite_ver_za128_s8_m (15, w0, p0, z0),
+ svwrite_ver_za128_m (15, w0, p0, z0))
+
+/*
+** write_za128_s8_0_w0_z1:
+** mov (w1[2-5]), w0
+** mova za0v\.q\[\1, 0\], p0/m, z1\.q
+** ret
+*/
+TEST_WRITE_ZA (write_za128_s8_0_w0_z1, svint8_t,
+ svwrite_ver_za128_s8_m (0, w0, p0, z1),
+ svwrite_ver_za128_m (0, w0, p0, z1))
+
+/*
+** write_za128_u8_0_w0_z0:
+** mov (w1[2-5]), w0
+** mova za0v\.q\[\1, 0\], p0/m, z0\.q
+** ret
+*/
+TEST_WRITE_ZA (write_za128_u8_0_w0_z0, svuint8_t,
+ svwrite_ver_za128_u8_m (0, w0, p0, z0),
+ svwrite_ver_za128_m (0, w0, p0, z0))
+
+/*
+** write_za128_s16_0_w0_z0:
+** mov (w1[2-5]), w0
+** mova za0v\.q\[\1, 0\], p0/m, z0\.q
+** ret
+*/
+TEST_WRITE_ZA (write_za128_s16_0_w0_z0, svint16_t,
+ svwrite_ver_za128_s16_m (0, w0, p0, z0),
+ svwrite_ver_za128_m (0, w0, p0, z0))
+
+/*
+** write_za128_u16_0_w0_z0:
+** mov (w1[2-5]), w0
+** mova za0v\.q\[\1, 0\], p0/m, z0\.q
+** ret
+*/
+TEST_WRITE_ZA (write_za128_u16_0_w0_z0, svuint16_t,
+ svwrite_ver_za128_u16_m (0, w0, p0, z0),
+ svwrite_ver_za128_m (0, w0, p0, z0))
+
+/*
+** write_za128_f16_0_w0_z0:
+** mov (w1[2-5]), w0
+** mova za0v\.q\[\1, 0\], p0/m, z0\.q
+** ret
+*/
+TEST_WRITE_ZA (write_za128_f16_0_w0_z0, svfloat16_t,
+ svwrite_ver_za128_f16_m (0, w0, p0, z0),
+ svwrite_ver_za128_m (0, w0, p0, z0))
+
+/*
+** write_za128_bf16_0_w0_z0:
+** mov (w1[2-5]), w0
+** mova za0v\.q\[\1, 0\], p0/m, z0\.q
+** ret
+*/
+TEST_WRITE_ZA (write_za128_bf16_0_w0_z0, svbfloat16_t,
+ svwrite_ver_za128_bf16_m (0, w0, p0, z0),
+ svwrite_ver_za128_m (0, w0, p0, z0))
+
+/*
+** write_za128_s32_0_w0_z0:
+** mov (w1[2-5]), w0
+** mova za0v\.q\[\1, 0\], p0/m, z0\.q
+** ret
+*/
+TEST_WRITE_ZA (write_za128_s32_0_w0_z0, svint32_t,
+ svwrite_ver_za128_s32_m (0, w0, p0, z0),
+ svwrite_ver_za128_m (0, w0, p0, z0))
+
+/*
+** write_za128_u32_0_w0_z0:
+** mov (w1[2-5]), w0
+** mova za0v\.q\[\1, 0\], p0/m, z0\.q
+** ret
+*/
+TEST_WRITE_ZA (write_za128_u32_0_w0_z0, svuint32_t,
+ svwrite_ver_za128_u32_m (0, w0, p0, z0),
+ svwrite_ver_za128_m (0, w0, p0, z0))
+
+/*
+** write_za128_f32_0_w0_z0:
+** mov (w1[2-5]), w0
+** mova za0v\.q\[\1, 0\], p0/m, z0\.q
+** ret
+*/
+TEST_WRITE_ZA (write_za128_f32_0_w0_z0, svfloat32_t,
+ svwrite_ver_za128_f32_m (0, w0, p0, z0),
+ svwrite_ver_za128_m (0, w0, p0, z0))
+
+/*
+** write_za128_s64_0_w0_z0:
+** mov (w1[2-5]), w0
+** mova za0v\.q\[\1, 0\], p0/m, z0\.q
+** ret
+*/
+TEST_WRITE_ZA (write_za128_s64_0_w0_z0, svint64_t,
+ svwrite_ver_za128_s64_m (0, w0, p0, z0),
+ svwrite_ver_za128_m (0, w0, p0, z0))
+
+/*
+** write_za128_u64_0_w0_z0:
+** mov (w1[2-5]), w0
+** mova za0v\.q\[\1, 0\], p0/m, z0\.q
+** ret
+*/
+TEST_WRITE_ZA (write_za128_u64_0_w0_z0, svuint64_t,
+ svwrite_ver_za128_u64_m (0, w0, p0, z0),
+ svwrite_ver_za128_m (0, w0, p0, z0))
+
+/*
+** write_za128_f64_0_w0_z0:
+** mov (w1[2-5]), w0
+** mova za0v\.q\[\1, 0\], p0/m, z0\.q
+** ret
+*/
+TEST_WRITE_ZA (write_za128_f64_0_w0_z0, svfloat64_t,
+ svwrite_ver_za128_f64_m (0, w0, p0, z0),
+ svwrite_ver_za128_m (0, w0, p0, z0))
--- /dev/null
+/* { dg-final { check-function-bodies "**" "" "-DCHECK_ASM" } } */
+
+#include "test_sme_acle.h"
+
+/*
+** write_za16_s16_0_0_z0:
+** mov (w1[2-5]), (?:wzr|#?0)
+** mova za0v\.h\[\1, 0\], p0/m, z0\.h
+** ret
+*/
+TEST_WRITE_ZA (write_za16_s16_0_0_z0, svint16_t,
+ svwrite_ver_za16_s16_m (0, 0, p0, z0),
+ svwrite_ver_za16_m (0, 0, p0, z0))
+
+/*
+** write_za16_s16_0_1_z0:
+** mov (w1[2-5]), #?1
+** mova za0v\.h\[\1, 0\], p0/m, z0\.h
+** ret
+*/
+TEST_WRITE_ZA (write_za16_s16_0_1_z0, svint16_t,
+ svwrite_ver_za16_s16_m (0, 1, p0, z0),
+ svwrite_ver_za16_m (0, 1, p0, z0))
+
+/*
+** write_za16_s16_0_w0_z0:
+** mov (w1[2-5]), w0
+** mova za0v\.h\[\1, 0\], p0/m, z0\.h
+** ret
+*/
+TEST_WRITE_ZA (write_za16_s16_0_w0_z0, svint16_t,
+ svwrite_ver_za16_s16_m (0, w0, p0, z0),
+ svwrite_ver_za16_m (0, w0, p0, z0))
+
+/*
+** write_za16_s16_0_w0p1_z0:
+** mov (w1[2-5]), w0
+** mova za0v\.h\[\1, 1\], p0/m, z0\.h
+** ret
+*/
+TEST_WRITE_ZA (write_za16_s16_0_w0p1_z0, svint16_t,
+ svwrite_ver_za16_s16_m (0, w0 + 1, p0, z0),
+ svwrite_ver_za16_m (0, w0 + 1, p0, z0))
+
+/*
+** write_za16_s16_0_w0p7_z0:
+** mov (w1[2-5]), w0
+** mova za0v\.h\[\1, 7\], p0/m, z0\.h
+** ret
+*/
+TEST_WRITE_ZA (write_za16_s16_0_w0p7_z0, svint16_t,
+ svwrite_ver_za16_s16_m (0, w0 + 7, p0, z0),
+ svwrite_ver_za16_m (0, w0 + 7, p0, z0))
+
+/*
+** write_za16_s16_0_w0p8_z0:
+** add (w1[2-5]), w0, #?8
+** mova za0v\.h\[\1, 0\], p0/m, z0\.h
+** ret
+*/
+TEST_WRITE_ZA (write_za16_s16_0_w0p8_z0, svint16_t,
+ svwrite_ver_za16_s16_m (0, w0 + 8, p0, z0),
+ svwrite_ver_za16_m (0, w0 + 8, p0, z0))
+
+/*
+** write_za16_s16_0_w0m1_z0:
+** sub (w1[2-5]), w0, #?1
+** mova za0v\.h\[\1, 0\], p0/m, z0\.h
+** ret
+*/
+TEST_WRITE_ZA (write_za16_s16_0_w0m1_z0, svint16_t,
+ svwrite_ver_za16_s16_m (0, w0 - 1, p0, z0),
+ svwrite_ver_za16_m (0, w0 - 1, p0, z0))
+
+/*
+** write_za16_s16_1_w0_z0:
+** mov (w1[2-5]), w0
+** mova za1v\.h\[\1, 0\], p0/m, z0\.h
+** ret
+*/
+TEST_WRITE_ZA (write_za16_s16_1_w0_z0, svint16_t,
+ svwrite_ver_za16_s16_m (1, w0, p0, z0),
+ svwrite_ver_za16_m (1, w0, p0, z0))
+
+/*
+** write_za16_s16_1_w0p7_z0:
+** mov (w1[2-5]), w0
+** mova za1v\.h\[\1, 7\], p0/m, z0\.h
+** ret
+*/
+TEST_WRITE_ZA (write_za16_s16_1_w0p7_z0, svint16_t,
+ svwrite_ver_za16_s16_m (1, w0 + 7, p0, z0),
+ svwrite_ver_za16_m (1, w0 + 7, p0, z0))
+
+/*
+** write_za16_s16_0_w0_z1:
+** mov (w1[2-5]), w0
+** mova za0v\.h\[\1, 0\], p0/m, z1\.h
+** ret
+*/
+TEST_WRITE_ZA (write_za16_s16_0_w0_z1, svint16_t,
+ svwrite_ver_za16_s16_m (0, w0, p0, z1),
+ svwrite_ver_za16_m (0, w0, p0, z1))
+
+/*
+** write_za16_u16_0_w0_z0:
+** mov (w1[2-5]), w0
+** mova za0v\.h\[\1, 0\], p0/m, z0\.h
+** ret
+*/
+TEST_WRITE_ZA (write_za16_u16_0_w0_z0, svuint16_t,
+ svwrite_ver_za16_u16_m (0, w0, p0, z0),
+ svwrite_ver_za16_m (0, w0, p0, z0))
+
+/*
+** write_za16_f16_0_w0_z0:
+** mov (w1[2-5]), w0
+** mova za0v\.h\[\1, 0\], p0/m, z0\.h
+** ret
+*/
+TEST_WRITE_ZA (write_za16_f16_0_w0_z0, svfloat16_t,
+ svwrite_ver_za16_f16_m (0, w0, p0, z0),
+ svwrite_ver_za16_m (0, w0, p0, z0))
+
+/*
+** write_za16_bf16_0_w0_z0:
+** mov (w1[2-5]), w0
+** mova za0v\.h\[\1, 0\], p0/m, z0\.h
+** ret
+*/
+TEST_WRITE_ZA (write_za16_bf16_0_w0_z0, svbfloat16_t,
+ svwrite_ver_za16_bf16_m (0, w0, p0, z0),
+ svwrite_ver_za16_m (0, w0, p0, z0))
--- /dev/null
+/* { dg-final { check-function-bodies "**" "" "-DCHECK_ASM" } } */
+
+#include "test_sme_acle.h"
+
+/*
+** write_za32_s32_0_0_z0:
+** mov (w1[2-5]), (?:wzr|#?0)
+** mova za0v\.s\[\1, 0\], p0/m, z0\.s
+** ret
+*/
+TEST_WRITE_ZA (write_za32_s32_0_0_z0, svint32_t,
+ svwrite_ver_za32_s32_m (0, 0, p0, z0),
+ svwrite_ver_za32_m (0, 0, p0, z0))
+
+/*
+** write_za32_s32_0_1_z0:
+** mov (w1[2-5]), #?1
+** mova za0v\.s\[\1, 0\], p0/m, z0\.s
+** ret
+*/
+TEST_WRITE_ZA (write_za32_s32_0_1_z0, svint32_t,
+ svwrite_ver_za32_s32_m (0, 1, p0, z0),
+ svwrite_ver_za32_m (0, 1, p0, z0))
+
+/*
+** write_za32_s32_0_w0_z0:
+** mov (w1[2-5]), w0
+** mova za0v\.s\[\1, 0\], p0/m, z0\.s
+** ret
+*/
+TEST_WRITE_ZA (write_za32_s32_0_w0_z0, svint32_t,
+ svwrite_ver_za32_s32_m (0, w0, p0, z0),
+ svwrite_ver_za32_m (0, w0, p0, z0))
+
+/*
+** write_za32_s32_0_w0p1_z0:
+** mov (w1[2-5]), w0
+** mova za0v\.s\[\1, 1\], p0/m, z0\.s
+** ret
+*/
+TEST_WRITE_ZA (write_za32_s32_0_w0p1_z0, svint32_t,
+ svwrite_ver_za32_s32_m (0, w0 + 1, p0, z0),
+ svwrite_ver_za32_m (0, w0 + 1, p0, z0))
+
+/*
+** write_za32_s32_0_w0p3_z0:
+** mov (w1[2-5]), w0
+** mova za0v\.s\[\1, 3\], p0/m, z0\.s
+** ret
+*/
+TEST_WRITE_ZA (write_za32_s32_0_w0p3_z0, svint32_t,
+ svwrite_ver_za32_s32_m (0, w0 + 3, p0, z0),
+ svwrite_ver_za32_m (0, w0 + 3, p0, z0))
+
+/*
+** write_za32_s32_0_w0p4_z0:
+** add (w1[2-5]), w0, #?4
+** mova za0v\.s\[\1, 0\], p0/m, z0\.s
+** ret
+*/
+TEST_WRITE_ZA (write_za32_s32_0_w0p4_z0, svint32_t,
+ svwrite_ver_za32_s32_m (0, w0 + 4, p0, z0),
+ svwrite_ver_za32_m (0, w0 + 4, p0, z0))
+
+/*
+** write_za32_s32_0_w0m1_z0:
+** sub (w1[2-5]), w0, #?1
+** mova za0v\.s\[\1, 0\], p0/m, z0\.s
+** ret
+*/
+TEST_WRITE_ZA (write_za32_s32_0_w0m1_z0, svint32_t,
+ svwrite_ver_za32_s32_m (0, w0 - 1, p0, z0),
+ svwrite_ver_za32_m (0, w0 - 1, p0, z0))
+
+/*
+** write_za32_s32_1_w0_z0:
+** mov (w1[2-5]), w0
+** mova za1v\.s\[\1, 0\], p0/m, z0\.s
+** ret
+*/
+TEST_WRITE_ZA (write_za32_s32_1_w0_z0, svint32_t,
+ svwrite_ver_za32_s32_m (1, w0, p0, z0),
+ svwrite_ver_za32_m (1, w0, p0, z0))
+
+/*
+** write_za32_s32_1_w0p3_z0:
+** mov (w1[2-5]), w0
+** mova za1v\.s\[\1, 3\], p0/m, z0\.s
+** ret
+*/
+TEST_WRITE_ZA (write_za32_s32_1_w0p3_z0, svint32_t,
+ svwrite_ver_za32_s32_m (1, w0 + 3, p0, z0),
+ svwrite_ver_za32_m (1, w0 + 3, p0, z0))
+
+/*
+** write_za32_s32_3_w0_z0:
+** mov (w1[2-5]), w0
+** mova za3v\.s\[\1, 0\], p0/m, z0\.s
+** ret
+*/
+TEST_WRITE_ZA (write_za32_s32_3_w0_z0, svint32_t,
+ svwrite_ver_za32_s32_m (3, w0, p0, z0),
+ svwrite_ver_za32_m (3, w0, p0, z0))
+
+/*
+** write_za32_s32_3_w0p3_z0:
+** mov (w1[2-5]), w0
+** mova za3v\.s\[\1, 3\], p0/m, z0\.s
+** ret
+*/
+TEST_WRITE_ZA (write_za32_s32_3_w0p3_z0, svint32_t,
+ svwrite_ver_za32_s32_m (3, w0 + 3, p0, z0),
+ svwrite_ver_za32_m (3, w0 + 3, p0, z0))
+
+/*
+** write_za32_s32_0_w0_z1:
+** mov (w1[2-5]), w0
+** mova za0v\.s\[\1, 0\], p0/m, z1\.s
+** ret
+*/
+TEST_WRITE_ZA (write_za32_s32_0_w0_z1, svint32_t,
+ svwrite_ver_za32_s32_m (0, w0, p0, z1),
+ svwrite_ver_za32_m (0, w0, p0, z1))
+
+/*
+** write_za32_u32_0_w0_z0:
+** mov (w1[2-5]), w0
+** mova za0v\.s\[\1, 0\], p0/m, z0\.s
+** ret
+*/
+TEST_WRITE_ZA (write_za32_u32_0_w0_z0, svuint32_t,
+ svwrite_ver_za32_u32_m (0, w0, p0, z0),
+ svwrite_ver_za32_m (0, w0, p0, z0))
+
+/*
+** write_za32_f32_0_w0_z0:
+** mov (w1[2-5]), w0
+** mova za0v\.s\[\1, 0\], p0/m, z0\.s
+** ret
+*/
+TEST_WRITE_ZA (write_za32_f32_0_w0_z0, svfloat32_t,
+ svwrite_ver_za32_f32_m (0, w0, p0, z0),
+ svwrite_ver_za32_m (0, w0, p0, z0))
--- /dev/null
+/* { dg-final { check-function-bodies "**" "" "-DCHECK_ASM" } } */
+
+#include "test_sme_acle.h"
+
+/*
+** write_za64_s64_0_0_z0:
+** mov (w1[2-5]), (?:wzr|#?0)
+** mova za0v\.d\[\1, 0\], p0/m, z0\.d
+** ret
+*/
+TEST_WRITE_ZA (write_za64_s64_0_0_z0, svint64_t,
+ svwrite_ver_za64_s64_m (0, 0, p0, z0),
+ svwrite_ver_za64_m (0, 0, p0, z0))
+
+/*
+** write_za64_s64_0_1_z0:
+** mov (w1[2-5]), #?1
+** mova za0v\.d\[\1, 0\], p0/m, z0\.d
+** ret
+*/
+TEST_WRITE_ZA (write_za64_s64_0_1_z0, svint64_t,
+ svwrite_ver_za64_s64_m (0, 1, p0, z0),
+ svwrite_ver_za64_m (0, 1, p0, z0))
+
+/*
+** write_za64_s64_0_w0_z0:
+** mov (w1[2-5]), w0
+** mova za0v\.d\[\1, 0\], p0/m, z0\.d
+** ret
+*/
+TEST_WRITE_ZA (write_za64_s64_0_w0_z0, svint64_t,
+ svwrite_ver_za64_s64_m (0, w0, p0, z0),
+ svwrite_ver_za64_m (0, w0, p0, z0))
+
+/*
+** write_za64_s64_0_w0p1_z0:
+** mov (w1[2-5]), w0
+** mova za0v\.d\[\1, 1\], p0/m, z0\.d
+** ret
+*/
+TEST_WRITE_ZA (write_za64_s64_0_w0p1_z0, svint64_t,
+ svwrite_ver_za64_s64_m (0, w0 + 1, p0, z0),
+ svwrite_ver_za64_m (0, w0 + 1, p0, z0))
+
+/*
+** write_za64_s64_0_w0p2_z0:
+** add (w1[2-5]), w0, #?2
+** mova za0v\.d\[\1, 0\], p0/m, z0\.d
+** ret
+*/
+TEST_WRITE_ZA (write_za64_s64_0_w0p2_z0, svint64_t,
+ svwrite_ver_za64_s64_m (0, w0 + 2, p0, z0),
+ svwrite_ver_za64_m (0, w0 + 2, p0, z0))
+
+/*
+** write_za64_s64_0_w0m1_z0:
+** sub (w1[2-5]), w0, #?1
+** mova za0v\.d\[\1, 0\], p0/m, z0\.d
+** ret
+*/
+TEST_WRITE_ZA (write_za64_s64_0_w0m1_z0, svint64_t,
+ svwrite_ver_za64_s64_m (0, w0 - 1, p0, z0),
+ svwrite_ver_za64_m (0, w0 - 1, p0, z0))
+
+/*
+** write_za64_s64_1_w0_z0:
+** mov (w1[2-5]), w0
+** mova za1v\.d\[\1, 0\], p0/m, z0\.d
+** ret
+*/
+TEST_WRITE_ZA (write_za64_s64_1_w0_z0, svint64_t,
+ svwrite_ver_za64_s64_m (1, w0, p0, z0),
+ svwrite_ver_za64_m (1, w0, p0, z0))
+
+/*
+** write_za64_s64_1_w0p1_z0:
+** mov (w1[2-5]), w0
+** mova za1v\.d\[\1, 1\], p0/m, z0\.d
+** ret
+*/
+TEST_WRITE_ZA (write_za64_s64_1_w0p1_z0, svint64_t,
+ svwrite_ver_za64_s64_m (1, w0 + 1, p0, z0),
+ svwrite_ver_za64_m (1, w0 + 1, p0, z0))
+
+/*
+** write_za64_s64_7_w0_z0:
+** mov (w1[2-5]), w0
+** mova za7v\.d\[\1, 0\], p0/m, z0\.d
+** ret
+*/
+TEST_WRITE_ZA (write_za64_s64_7_w0_z0, svint64_t,
+ svwrite_ver_za64_s64_m (7, w0, p0, z0),
+ svwrite_ver_za64_m (7, w0, p0, z0))
+
+/*
+** write_za64_s64_7_w0p1_z0:
+** mov (w1[2-5]), w0
+** mova za7v\.d\[\1, 1\], p0/m, z0\.d
+** ret
+*/
+TEST_WRITE_ZA (write_za64_s64_7_w0p1_z0, svint64_t,
+ svwrite_ver_za64_s64_m (7, w0 + 1, p0, z0),
+ svwrite_ver_za64_m (7, w0 + 1, p0, z0))
+
+/*
+** write_za64_s64_0_w0_z1:
+** mov (w1[2-5]), w0
+** mova za0v\.d\[\1, 0\], p0/m, z1\.d
+** ret
+*/
+TEST_WRITE_ZA (write_za64_s64_0_w0_z1, svint64_t,
+ svwrite_ver_za64_s64_m (0, w0, p0, z1),
+ svwrite_ver_za64_m (0, w0, p0, z1))
+
+/*
+** write_za64_u64_0_w0_z0:
+** mov (w1[2-5]), w0
+** mova za0v\.d\[\1, 0\], p0/m, z0\.d
+** ret
+*/
+TEST_WRITE_ZA (write_za64_u64_0_w0_z0, svuint64_t,
+ svwrite_ver_za64_u64_m (0, w0, p0, z0),
+ svwrite_ver_za64_m (0, w0, p0, z0))
+
+/*
+** write_za64_f64_0_w0_z0:
+** mov (w1[2-5]), w0
+** mova za0v\.d\[\1, 0\], p0/m, z0\.d
+** ret
+*/
+TEST_WRITE_ZA (write_za64_f64_0_w0_z0, svfloat64_t,
+ svwrite_ver_za64_f64_m (0, w0, p0, z0),
+ svwrite_ver_za64_m (0, w0, p0, z0))
--- /dev/null
+/* { dg-final { check-function-bodies "**" "" "-DCHECK_ASM" } } */
+
+#include "test_sme_acle.h"
+
+/*
+** write_za8_s8_0_0_z0:
+** mov (w1[2-5]), (?:wzr|#?0)
+** mova za0v\.b\[\1, 0\], p0/m, z0\.b
+** ret
+*/
+TEST_WRITE_ZA (write_za8_s8_0_0_z0, svint8_t,
+ svwrite_ver_za8_s8_m (0, 0, p0, z0),
+ svwrite_ver_za8_m (0, 0, p0, z0))
+
+/*
+** write_za8_s8_0_1_z0:
+** mov (w1[2-5]), #?1
+** mova za0v\.b\[\1, 0\], p0/m, z0\.b
+** ret
+*/
+TEST_WRITE_ZA (write_za8_s8_0_1_z0, svint8_t,
+ svwrite_ver_za8_s8_m (0, 1, p0, z0),
+ svwrite_ver_za8_m (0, 1, p0, z0))
+
+/*
+** write_za8_s8_0_w0_z0:
+** mov (w1[2-5]), w0
+** mova za0v\.b\[\1, 0\], p0/m, z0\.b
+** ret
+*/
+TEST_WRITE_ZA (write_za8_s8_0_w0_z0, svint8_t,
+ svwrite_ver_za8_s8_m (0, w0, p0, z0),
+ svwrite_ver_za8_m (0, w0, p0, z0))
+
+/*
+** write_za8_s8_0_w0p1_z0:
+** mov (w1[2-5]), w0
+** mova za0v\.b\[\1, 1\], p0/m, z0\.b
+** ret
+*/
+TEST_WRITE_ZA (write_za8_s8_0_w0p1_z0, svint8_t,
+ svwrite_ver_za8_s8_m (0, w0 + 1, p0, z0),
+ svwrite_ver_za8_m (0, w0 + 1, p0, z0))
+
+/*
+** write_za8_s8_0_w0p15_z0:
+** mov (w1[2-5]), w0
+** mova za0v\.b\[\1, 15\], p0/m, z0\.b
+** ret
+*/
+TEST_WRITE_ZA (write_za8_s8_0_w0p15_z0, svint8_t,
+ svwrite_ver_za8_s8_m (0, w0 + 15, p0, z0),
+ svwrite_ver_za8_m (0, w0 + 15, p0, z0))
+
+/*
+** write_za8_s8_0_w0p16_z0:
+** add (w1[2-5]), w0, #?16
+** mova za0v\.b\[\1, 0\], p0/m, z0\.b
+** ret
+*/
+TEST_WRITE_ZA (write_za8_s8_0_w0p16_z0, svint8_t,
+ svwrite_ver_za8_s8_m (0, w0 + 16, p0, z0),
+ svwrite_ver_za8_m (0, w0 + 16, p0, z0))
+
+/*
+** write_za8_s8_0_w0m1_z0:
+** sub (w1[2-5]), w0, #?1
+** mova za0v\.b\[\1, 0\], p0/m, z0\.b
+** ret
+*/
+TEST_WRITE_ZA (write_za8_s8_0_w0m1_z0, svint8_t,
+ svwrite_ver_za8_s8_m (0, w0 - 1, p0, z0),
+ svwrite_ver_za8_m (0, w0 - 1, p0, z0))
+
+/*
+** write_za8_s8_0_w0_z1:
+** mov (w1[2-5]), w0
+** mova za0v\.b\[\1, 0\], p0/m, z1\.b
+** ret
+*/
+TEST_WRITE_ZA (write_za8_s8_0_w0_z1, svint8_t,
+ svwrite_ver_za8_s8_m (0, w0, p0, z1),
+ svwrite_ver_za8_m (0, w0, p0, z1))
+
+/*
+** write_za8_u8_0_w0_z0:
+** mov (w1[2-5]), w0
+** mova za0v\.b\[\1, 0\], p0/m, z0\.b
+** ret
+*/
+TEST_WRITE_ZA (write_za8_u8_0_w0_z0, svuint8_t,
+ svwrite_ver_za8_u8_m (0, w0, p0, z0),
+ svwrite_ver_za8_m (0, w0, p0, z0))
--- /dev/null
+/* { dg-final { check-function-bodies "**" "" "-DCHECK_ASM" } } */
+
+#define STREAMING_COMPATIBLE
+#include "test_sme_acle.h"
+
+/*
+** zero_mask_za_0:
+** zero { *}
+** ret
+*/
+PROTO (zero_mask_za_0, void, ()) { svzero_mask_za (0); }
+
+/*
+** zero_mask_za_01:
+** zero { za0\.d }
+** ret
+*/
+PROTO (zero_mask_za_01, void, ()) { svzero_mask_za (0x01); }
+
+/*
+** zero_mask_za_80:
+** zero { za7\.d }
+** ret
+*/
+PROTO (zero_mask_za_80, void, ()) { svzero_mask_za (0x80); }
+
+/*
+** zero_mask_za_03:
+** zero { za0\.d, za1\.d }
+** ret
+*/
+PROTO (zero_mask_za_03, void, ()) { svzero_mask_za (0x03); }
+
+/*
+** zero_mask_za_09:
+** zero { za0\.d, za3\.d }
+** ret
+*/
+PROTO (zero_mask_za_09, void, ()) { svzero_mask_za (0x09); }
+
+/*
+** zero_mask_za_0d:
+** zero { za0\.d, za2\.d, za3\.d }
+** ret
+*/
+PROTO (zero_mask_za_0d, void, ()) { svzero_mask_za (0x0d); }
+
+/*
+** zero_mask_za_3c:
+** zero { za2\.d, za3\.d, za4\.d, za5\.d }
+** ret
+*/
+PROTO (zero_mask_za_3c, void, ()) { svzero_mask_za (0x3c); }
+
+/*
+** zero_mask_za_5a:
+** zero { za1\.d, za3\.d, za4\.d, za6\.d }
+** ret
+*/
+PROTO (zero_mask_za_5a, void, ()) { svzero_mask_za (0x5a); }
+
+/*
+** zero_mask_za_11:
+** zero { za0\.s }
+** ret
+*/
+PROTO (zero_mask_za_11, void, ()) { svzero_mask_za (0x11); }
+
+/*
+** zero_mask_za_88:
+** zero { za3\.s }
+** ret
+*/
+PROTO (zero_mask_za_88, void, ()) { svzero_mask_za (0x88); }
+
+/*
+** zero_mask_za_33:
+** zero { za0\.s, za1\.s }
+** ret
+*/
+PROTO (zero_mask_za_33, void, ()) { svzero_mask_za (0x33); }
+
+/*
+** zero_mask_za_cc:
+** zero { za2\.s, za3\.s }
+** ret
+*/
+PROTO (zero_mask_za_cc, void, ()) { svzero_mask_za (0xcc); }
+
+/*
+** zero_mask_za_55:
+** zero { za0\.h }
+** ret
+*/
+PROTO (zero_mask_za_55, void, ()) { svzero_mask_za (0x55); }
+
+/*
+** zero_mask_za_aa:
+** zero { za1\.h }
+** ret
+*/
+PROTO (zero_mask_za_aa, void, ()) { svzero_mask_za (0xaa); }
+
+/*
+** zero_mask_za_ab:
+** zero { za1\.h, za0\.d }
+** ret
+*/
+PROTO (zero_mask_za_ab, void, ()) { svzero_mask_za (0xab); }
+
+/*
+** zero_mask_za_d7:
+** zero { za0\.h, za1\.d, za7\.d }
+** ret
+*/
+PROTO (zero_mask_za_d7, void, ()) { svzero_mask_za (0xd7); }
+
+/*
+** zero_mask_za_bf:
+** zero { za1\.h, za0\.s, za2\.d }
+** ret
+*/
+PROTO (zero_mask_za_bf, void, ()) { svzero_mask_za (0xbf); }
+
+/*
+** zero_mask_za_ff:
+** zero { za }
+** ret
+*/
+PROTO (zero_mask_za_ff, void, ()) { svzero_mask_za (0xff); }
--- /dev/null
+/* { dg-final { check-function-bodies "**" "" "-DCHECK_ASM" } } */
+
+#define STREAMING_COMPATIBLE
+#include "test_sme_acle.h"
+
+/*
+** zero_za:
+** zero { za }
+** ret
+*/
+PROTO (zero_za, void, ()) { svzero_za (); }
#endif
#ifdef STREAMING_COMPATIBLE
-#define ATTR __arm_streaming_compatible
+#define SM_ATTR __arm_streaming_compatible
+#elif defined(STREAMING)
+#define SM_ATTR __arm_streaming
#else
-#define ATTR
+#define SM_ATTR
#endif
+#ifdef SHARED_ZA
+#define ZA_ATTR __arm_inout("za")
+#else
+#define ZA_ATTR
+#endif
+
+#define ATTR SM_ATTR ZA_ATTR
+
#ifdef __cplusplus
#define PROTO(NAME, RET, ARGS) \
extern "C" RET NAME ARGS ATTR; RET NAME ARGS ATTR
--- /dev/null
+/* { dg-do compile } */
+
+#include <arm_sme.h>
+
+#pragma GCC target ("arch=armv9-a+sme")
+
+void
+f1 (svbool_t pg, svint8_t s8, svuint8_t u8,
+ svint16_t s16, svuint16_t u16, svfloat16_t f16, uint32_t tile)
+ __arm_streaming __arm_inout("za")
+{
+ svusmopa_za32_m (0, pg, pg, u8); /* { dg-error {too few arguments to function 'svusmopa_za32_m'} } */
+ svusmopa_za32_m (0, pg, pg, u8, s8, 0); /* { dg-error {too many arguments to function 'svusmopa_za32_m'} } */
+ svusmopa_za32_m (tile, pg, pg, u8, s8); /* { dg-error {argument 1 of 'svusmopa_za32_m' must be an integer constant expression} } */
+ svusmopa_za32_m (-1, pg, pg, u8, s8); /* { dg-error {passing -1 to argument 1 of 'svusmopa_za32_m', which expects a value in the range \[0, 3\]} } */
+ svusmopa_za32_m (4, pg, pg, u8, s8); /* { dg-error {passing 4 to argument 1 of 'svusmopa_za32_m', which expects a value in the range \[0, 3\]} } */
+ svusmopa_za32_m (0, u8, pg, u8, s8); /* { dg-error {passing 'svuint8_t' to argument 2 of 'svusmopa_za32_m', which expects 'svbool_t'} } */
+ svusmopa_za32_m (0, pg, u8, u8, s8); /* { dg-error {passing 'svuint8_t' to argument 3 of 'svusmopa_za32_m', which expects 'svbool_t'} } */
+ svusmopa_za32_m (0, pg, pg, tile, s8); /* { dg-error {passing 'uint32_t'.* to argument 4 of 'svusmopa_za32_m', which expects an SVE type} } */
+ svusmopa_za32_m (0, pg, pg, s8, s8); /* { dg-error {'svusmopa_za32_m' has no form that takes 'svint8_t' arguments} } */
+ svusmopa_za32_m (0, pg, pg, pg, s8); /* { dg-error {'svusmopa_za32_m' has no form that takes 'svbool_t' arguments} } */
+ svusmopa_za32_m (0, pg, pg, f16, s8); /* { dg-error {'svusmopa_za32_m' has no form that takes 'svfloat16_t' arguments} } */
+ svusmopa_za32_m (0, pg, pg, u8, u8); /* { dg-error {passing 'svuint8_t' to argument 5 of 'svusmopa_za32_m', which expects a vector of signed integers} } */
+ svusmopa_za32_m (0, pg, pg, u8, s16); /* { dg-error {arguments 4 and 5 of 'svusmopa_za32_m' must have the same element size, but the values passed here have type 'svuint8_t' and 'svint16_t' respectively} } */
+ svusmopa_za32_m (0, pg, pg, u16, s16); /* { dg-error {'svusmopa_za32_m' has no form that takes 'svuint16_t' arguments} } */
+
+ svusmopa_za64_m (0, pg, pg, u16, s16); /* { dg-error {ACLE function 'svusmopa_za64_u16_m' requires ISA extension 'sme-i16i64'} } */
+}
+
+void
+f2 (svbool_t pg, svint8_t s8, svuint8_t u8) __arm_streaming
+{
+ svusmopa_za32_m (0, pg, pg, u8, s8); /* { dg-error {ACLE function 'svusmopa_za32_u8_m' can only be called from a function that has 'za' state} } */
+}
+
+void
+f3 (svbool_t pg, svint8_t s8, svuint8_t u8) __arm_inout("za")
+{
+ svusmopa_za32_m (0, pg, pg, u8, s8); /* { dg-error {ACLE function 'svusmopa_za32_u8_m' can only be called when SME streaming mode is enabled} } */
+}
+
+#pragma GCC target ("arch=armv9-a+sme-i16i64")
+
+void
+f4 (svbool_t pg, svint16_t s16, svuint16_t u16)
+ __arm_streaming __arm_inout("za")
+{
+ svusmopa_za64_m (-1, pg, pg, u16, s16); /* { dg-error {passing -1 to argument 1 of 'svusmopa_za64_m', which expects a value in the range \[0, 7\]} } */
+ svusmopa_za64_m (8, pg, pg, u16, s16); /* { dg-error {passing 8 to argument 1 of 'svusmopa_za64_m', which expects a value in the range \[0, 7\]} } */
+}
--- /dev/null
+/* { dg-do compile } */
+
+#include <arm_sme.h>
+
+#pragma GCC target ("arch=armv9-a+sme")
+
+void
+f1 (svbool_t pg, svint8_t s8, svuint8_t u8, svint16_t s16, svint32_t s32,
+ svfloat16_t f16, svfloat32_t f32, svfloat64_t f64, uint32_t tile)
+ __arm_streaming __arm_inout("za")
+{
+ svmopa_za32_m (0, pg, pg, s8); /* { dg-error {too few arguments to function 'svmopa_za32_m'} } */
+ svmopa_za32_m (0, pg, pg, s8, s8, 0); /* { dg-error {too many arguments to function 'svmopa_za32_m'} } */
+ svmopa_za32_m (tile, pg, pg, s8, s8); /* { dg-error {argument 1 of 'svmopa_za32_m' must be an integer constant expression} } */
+ svmopa_za32_m (-1, pg, pg, s8, s8); /* { dg-error {passing -1 to argument 1 of 'svmopa_za32_m', which expects a value in the range \[0, 3\]} } */
+ svmopa_za32_m (4, pg, pg, s8, s8); /* { dg-error {passing 4 to argument 1 of 'svmopa_za32_m', which expects a value in the range \[0, 3\]} } */
+ svmopa_za32_m (0, u8, pg, s8, s8); /* { dg-error {passing 'svuint8_t' to argument 2 of 'svmopa_za32_m', which expects 'svbool_t'} } */
+ svmopa_za32_m (0, pg, u8, s8, s8); /* { dg-error {passing 'svuint8_t' to argument 3 of 'svmopa_za32_m', which expects 'svbool_t'} } */
+ svmopa_za32_m (0, pg, pg, tile, s8); /* { dg-error {passing 'uint32_t'.* to argument 4 of 'svmopa_za32_m', which expects an SVE type} } */
+ svmopa_za32_m (0, pg, pg, u8, s8); /* { dg-error {passing 'svint8_t'.* to argument 5 of 'svmopa_za32_m', but argument 4 had type 'svuint8_t'} } */
+ svmopa_za32_m (0, pg, pg, s8, f16); /* { dg-error {passing 'svfloat16_t'.* to argument 5 of 'svmopa_za32_m', but argument 4 had type 'svint8_t'} } */
+ svmopa_za32_m (0, pg, pg, pg, pg); /* { dg-error {'svmopa_za32_m' has no form that takes 'svbool_t' arguments} } */
+ svmopa_za32_m (0, pg, pg, s16, s16); /* { dg-error {'svmopa_za32_m' has no form that takes 'svint16_t' arguments} } */
+ svmopa_za32_m (0, pg, pg, s32, s32); /* { dg-error {'svmopa_za32_m' has no form that takes 'svint32_t' arguments} } */
+ svmopa_za32_m (0, pg, pg, f64, f64); /* { dg-error {'svmopa_za32_m' has no form that takes 'svfloat64_t' arguments} } */
+
+ svmopa_za64_m (0, pg, pg, s16, s16); /* { dg-error {ACLE function 'svmopa_za64_s16_m' requires ISA extension 'sme-i16i64'} } */
+}
+
+void
+f2 (svbool_t pg, svint8_t s8) __arm_streaming
+{
+ svmopa_za32_m (0, pg, pg, s8, s8); /* { dg-error {ACLE function 'svmopa_za32_s8_m' can only be called from a function that has 'za' state} } */
+}
+
+void
+f3 (svbool_t pg, svint8_t s8) __arm_inout("za")
+{
+ svmopa_za32_m (0, pg, pg, s8, s8); /* { dg-error {ACLE function 'svmopa_za32_s8_m' can only be called when SME streaming mode is enabled} } */
+}
+
+#pragma GCC target ("arch=armv9-a+sme-i16i64")
+
+void
+f4 (svbool_t pg, svint16_t s16) __arm_streaming __arm_inout("za")
+{
+ svmopa_za64_m (-1, pg, pg, s16, s16); /* { dg-error {passing -1 to argument 1 of 'svmopa_za64_m', which expects a value in the range \[0, 7\]} } */
+ svmopa_za64_m (8, pg, pg, s16, s16); /* { dg-error {passing 8 to argument 1 of 'svmopa_za64_m', which expects a value in the range \[0, 7\]} } */
+}
--- /dev/null
+/* { dg-do compile } */
+
+#include <arm_sme.h>
+
+#pragma GCC target ("arch=armv9-a+sme")
+
+void
+f1 (svbool_t pg, svfloat64_t f64) __arm_streaming __arm_inout("za")
+{
+ svmopa_za64_m (0, pg, pg, f64, f64); /* { dg-error {ACLE function 'svmopa_za64_f64_m' requires ISA extension 'sme-f64f64'} } */
+}
--- /dev/null
+/* { dg-do compile } */
+
+#include <arm_sme.h>
+
+#pragma GCC target ("arch=armv9-a+sme")
+
+void
+f1 (svbool_t pg, svint8_t s8, svuint8_t u8,
+ svint16_t s16, svuint16_t u16, svfloat16_t f16, uint32_t tile)
+ __arm_streaming __arm_inout("za")
+{
+ svsumopa_za32_m (0, pg, pg, s8); /* { dg-error {too few arguments to function 'svsumopa_za32_m'} } */
+ svsumopa_za32_m (0, pg, pg, s8, u8, 0); /* { dg-error {too many arguments to function 'svsumopa_za32_m'} } */
+ svsumopa_za32_m (tile, pg, pg, s8, u8); /* { dg-error {argument 1 of 'svsumopa_za32_m' must be an integer constant expression} } */
+ svsumopa_za32_m (-1, pg, pg, s8, u8); /* { dg-error {passing -1 to argument 1 of 'svsumopa_za32_m', which expects a value in the range \[0, 3\]} } */
+ svsumopa_za32_m (4, pg, pg, s8, u8); /* { dg-error {passing 4 to argument 1 of 'svsumopa_za32_m', which expects a value in the range \[0, 3\]} } */
+ svsumopa_za32_m (0, u8, pg, s8, u8); /* { dg-error {passing 'svuint8_t' to argument 2 of 'svsumopa_za32_m', which expects 'svbool_t'} } */
+ svsumopa_za32_m (0, pg, u8, s8, u8); /* { dg-error {passing 'svuint8_t' to argument 3 of 'svsumopa_za32_m', which expects 'svbool_t'} } */
+ svsumopa_za32_m (0, pg, pg, tile, s8); /* { dg-error {passing 'uint32_t'.* to argument 4 of 'svsumopa_za32_m', which expects an SVE type} } */
+ svsumopa_za32_m (0, pg, pg, u8, u8); /* { dg-error {'svsumopa_za32_m' has no form that takes 'svuint8_t' arguments} } */
+ svsumopa_za32_m (0, pg, pg, pg, u8); /* { dg-error {'svsumopa_za32_m' has no form that takes 'svbool_t' arguments} } */
+ svsumopa_za32_m (0, pg, pg, f16, u8); /* { dg-error {'svsumopa_za32_m' has no form that takes 'svfloat16_t' arguments} } */
+ svsumopa_za32_m (0, pg, pg, s8, s8); /* { dg-error {passing 'svint8_t' to argument 5 of 'svsumopa_za32_m', which expects a vector of unsigned integers} } */
+ svsumopa_za32_m (0, pg, pg, s8, u16); /* { dg-error {arguments 4 and 5 of 'svsumopa_za32_m' must have the same element size, but the values passed here have type 'svint8_t' and 'svuint16_t' respectively} } */
+ svsumopa_za32_m (0, pg, pg, s16, u16); /* { dg-error {'svsumopa_za32_m' has no form that takes 'svint16_t' arguments} } */
+
+ svsumopa_za64_m (0, pg, pg, s16, u16); /* { dg-error {ACLE function 'svsumopa_za64_s16_m' requires ISA extension 'sme-i16i64'} } */
+}
+
+void
+f2 (svbool_t pg, svint8_t s8, svuint8_t u8) __arm_streaming
+{
+ svsumopa_za32_m (0, pg, pg, s8, u8); /* { dg-error {ACLE function 'svsumopa_za32_s8_m' can only be called from a function that has 'za' state} } */
+}
+
+void
+f3 (svbool_t pg, svint8_t s8, svuint8_t u8) __arm_inout("za")
+{
+ svsumopa_za32_m (0, pg, pg, s8, u8); /* { dg-error {ACLE function 'svsumopa_za32_s8_m' can only be called when SME streaming mode is enabled} } */
+}
+
+#pragma GCC target ("arch=armv9-a+sme-i16i64")
+
+void
+f4 (svbool_t pg, svint16_t s16, svuint16_t u16)
+ __arm_streaming __arm_inout("za")
+{
+ svsumopa_za64_m (-1, pg, pg, s16, u16); /* { dg-error {passing -1 to argument 1 of 'svsumopa_za64_m', which expects a value in the range \[0, 7\]} } */
+ svsumopa_za64_m (8, pg, pg, s16, u16); /* { dg-error {passing 8 to argument 1 of 'svsumopa_za64_m', which expects a value in the range \[0, 7\]} } */
+}
to be diagnosed. Any attempt to call the function before including
arm_sve.h will lead to a link failure. (Same for taking its address,
etc.) */
-extern __SVUint8_t svadd_u8_x (__SVBool_t, __SVUint8_t, __SVUint8_t);
+extern __SVUint8_t svadd_u8_x (__SVBool_t, __SVUint8_t, __SVUint8_t)
+ __arm_streaming_compatible;
#pragma GCC aarch64 "arm_sve.h"
explicit definition "wins". This isn't supported behavior though. */
__SVUint8_t
svadd_u8_x (__SVBool_t pg, __SVUint8_t x, __SVUint8_t y)
+ __arm_streaming_compatible
{
return x;
}
--- /dev/null
+/* { dg-do compile } */
+
+#include <arm_sme.h>
+
+#pragma GCC target ("arch=armv9-a+sme")
+
+void
+f1 (svbool_t pg, svint8_t s8, svint64_t s64, svuint8_t u8, svuint16_t u16,
+ svfloat32_t f32, uint32_t tile)
+ __arm_streaming __arm_inout("za")
+{
+ svread_hor_za8_m (s8, pg, 0); /* { dg-error {too few arguments to function 'svread_hor_za8_m'} } */
+ svread_hor_za8_m (s8, pg, 0, 0, 0); /* { dg-error {too many arguments to function 'svread_hor_za8_m'} } */
+ svread_hor_za8_m (tile, pg, 0, 0); /* { dg-error {passing 'uint32_t'.* to argument 1 of 'svread_hor_za8_m', which expects an SVE type} } */
+ svread_hor_za8_m (pg, pg, 0, 0); /* { dg-error {'svread_hor_za8_m' has no form that takes 'svbool_t' arguments} } */
+ svread_hor_za8_m (u16, pg, 0, 0); /* { dg-error {'svread_hor_za8_m' has no form that takes 'svuint16_t' arguments} } */
+ svread_hor_za8_m (s8, s8, 0, 0); /* { dg-error {passing 'svint8_t' to argument 2 of 'svread_hor_za8_m', which expects 'svbool_t'} } */
+ svread_hor_za8_m (s8, pg, tile, 0); /* { dg-error {argument 3 of 'svread_hor_za8_m' must be an integer constant expression} } */
+ svread_hor_za8_m (s8, pg, -1, 0); /* { dg-error {passing -1 to argument 3 of 'svread_hor_za8_m', which expects the value 0} } */
+ svread_hor_za8_m (s8, pg, 1, 0); /* { dg-error {passing 1 to argument 3 of 'svread_hor_za8_m', which expects the value 0} } */
+ svread_hor_za8_m (s8, pg, 0, u8); /* { dg-error {passing 'svuint8_t' to argument 4 of 'svread_hor_za8_m', which expects 'uint32_t'} } */
+
+ svread_hor_za16_m (u16, pg, -1, 0); /* { dg-error {passing -1 to argument 3 of 'svread_hor_za16_m', which expects a value in the range \[0, 1\]} } */
+ svread_hor_za16_m (u16, pg, 2, 0); /* { dg-error {passing 2 to argument 3 of 'svread_hor_za16_m', which expects a value in the range \[0, 1\]} } */
+
+ svread_hor_za32_m (f32, pg, -1, 0); /* { dg-error {passing -1 to argument 3 of 'svread_hor_za32_m', which expects a value in the range \[0, 3\]} } */
+ svread_hor_za32_m (f32, pg, 4, 0); /* { dg-error {passing 4 to argument 3 of 'svread_hor_za32_m', which expects a value in the range \[0, 3\]} } */
+
+ svread_hor_za64_m (s64, pg, -1, 0); /* { dg-error {passing -1 to argument 3 of 'svread_hor_za64_m', which expects a value in the range \[0, 7\]} } */
+ svread_hor_za64_m (s64, pg, 8, 0); /* { dg-error {passing 8 to argument 3 of 'svread_hor_za64_m', which expects a value in the range \[0, 7\]} } */
+
+ svread_hor_za128_m (s8, pg, -1, 0); /* { dg-error {passing -1 to argument 3 of 'svread_hor_za128_m', which expects a value in the range \[0, 15\]} } */
+ svread_hor_za128_m (s8, pg, 16, 0); /* { dg-error {passing 16 to argument 3 of 'svread_hor_za128_m', which expects a value in the range \[0, 15\]} } */
+ svread_hor_za128_m (f32, pg, -1, 0); /* { dg-error {passing -1 to argument 3 of 'svread_hor_za128_m', which expects a value in the range \[0, 15\]} } */
+ svread_hor_za128_m (f32, pg, 16, 0); /* { dg-error {passing 16 to argument 3 of 'svread_hor_za128_m', which expects a value in the range \[0, 15\]} } */
+}
+
+void
+f2 (svbool_t pg, svint8_t s8) __arm_streaming
+{
+ svread_hor_za8_m (s8, pg, 0, 0); /* { dg-error {ACLE function 'svread_hor_za8_s8_m' can only be called from a function that has 'za' state} } */
+}
+
+void
+f3 (svbool_t pg, svint8_t s8) __arm_inout("za")
+{
+ svread_hor_za8_m (s8, pg, 0, 0); /* { dg-error {ACLE function 'svread_hor_za8_s8_m' can only be called when SME streaming mode is enabled} } */
+}
--- /dev/null
+/* { dg-do compile } */
+
+#include <arm_sme.h>
+
+#pragma GCC target ("arch=armv9-a+sme")
+
+void
+f1 (svbool_t pg, svuint8_t u8, svint16_t s16, svint32_t s32, svint64_t s64,
+ svfloat32_t f32, uint32_t tile)
+ __arm_streaming __arm_inout("za")
+{
+ svaddha_za32_m (0, pg, pg); /* { dg-error {too few arguments to function 'svaddha_za32_m'} } */
+ svaddha_za32_m (0, pg, pg, s32, s32); /* { dg-error {too many arguments to function 'svaddha_za32_m'} } */
+ svaddha_za32_m (tile, pg, pg, s32); /* { dg-error {argument 1 of 'svaddha_za32_m' must be an integer constant expression} } */
+ svaddha_za32_m (-1, pg, pg, s32); /* { dg-error {passing -1 to argument 1 of 'svaddha_za32_m', which expects a value in the range \[0, 3\]} } */
+ svaddha_za32_m (4, pg, pg, s32); /* { dg-error {passing 4 to argument 1 of 'svaddha_za32_m', which expects a value in the range \[0, 3\]} } */
+ svaddha_za32_m (0, u8, pg, s32); /* { dg-error {passing 'svuint8_t' to argument 2 of 'svaddha_za32_m', which expects 'svbool_t'} } */
+ svaddha_za32_m (0, pg, u8, s32); /* { dg-error {passing 'svuint8_t' to argument 3 of 'svaddha_za32_m', which expects 'svbool_t'} } */
+ svaddha_za32_m (0, pg, pg, tile); /* { dg-error {passing 'uint32_t'.* to argument 4 of 'svaddha_za32_m', which expects an SVE type} } */
+ svaddha_za32_m (0, pg, pg, pg); /* { dg-error {'svaddha_za32_m' has no form that takes 'svbool_t' arguments} } */
+ svaddha_za32_m (0, pg, pg, u8); /* { dg-error {'svaddha_za32_m' has no form that takes 'svuint8_t' arguments} } */
+ svaddha_za32_m (0, pg, pg, s16); /* { dg-error {'svaddha_za32_m' has no form that takes 'svint16_t' arguments} } */
+ svaddha_za32_m (0, pg, pg, f32); /* { dg-error {'svaddha_za32_m' has no form that takes 'svfloat32_t' arguments} } */
+ svaddha_za32_m (0, pg, pg, s64); /* { dg-error {'svaddha_za32_m' has no form that takes 'svint64_t' arguments} } */
+
+ svaddha_za64_m (0, pg, pg, s64); /* { dg-error {ACLE function 'svaddha_za64_s64_m' requires ISA extension 'sme-i16i64'} } */
+}
+
+void
+f2 (svbool_t pg, svint32_t s32) __arm_streaming
+{
+ svaddha_za32_m (0, pg, pg, s32); /* { dg-error {ACLE function 'svaddha_za32_s32_m' can only be called from a function that has 'za' state} } */
+}
+
+void
+f3 (svbool_t pg, svint32_t s32) __arm_inout("za")
+{
+ svaddha_za32_m (0, pg, pg, s32); /* { dg-error {ACLE function 'svaddha_za32_s32_m' can only be called when SME streaming mode is enabled} } */
+}
+
+#pragma GCC target ("arch=armv9-a+sme-i16i64")
+
+void
+f4 (svbool_t pg, svint64_t s64)
+ __arm_streaming __arm_inout("za")
+{
+ svaddha_za64_m (-1, pg, pg, s64); /* { dg-error {passing -1 to argument 1 of 'svaddha_za64_m', which expects a value in the range \[0, 7\]} } */
+ svaddha_za64_m (8, pg, pg, s64); /* { dg-error {passing 8 to argument 1 of 'svaddha_za64_m', which expects a value in the range \[0, 7\]} } */
+}
--- /dev/null
+/* { dg-do compile } */
+
+#include <arm_sme.h>
+
+#pragma GCC target ("arch=armv9-a+sme")
+
+void
+f1 (svbool_t pg, svint8_t s8, svint64_t s64, svuint8_t u8, svuint16_t u16,
+ svfloat32_t f32, uint32_t tile)
+ __arm_streaming __arm_inout("za")
+{
+ svwrite_ver_za8_m (0, 0, pg); /* { dg-error {too few arguments to function 'svwrite_ver_za8_m'} } */
+ svwrite_ver_za8_m (0, 0, pg, s8, 0); /* { dg-error {too many arguments to function 'svwrite_ver_za8_m'} } */
+ svwrite_ver_za8_m (tile, 0, pg, s8); /* { dg-error {argument 1 of 'svwrite_ver_za8_m' must be an integer constant expression} } */
+ svwrite_ver_za8_m (-1, 0, pg, s8); /* { dg-error {passing -1 to argument 1 of 'svwrite_ver_za8_m', which expects the value 0} } */
+ svwrite_ver_za8_m (1, 0, pg, s8); /* { dg-error {passing 1 to argument 1 of 'svwrite_ver_za8_m', which expects the value 0} } */
+ svwrite_ver_za8_m (0, u8, pg, s8); /* { dg-error {passing 'svuint8_t' to argument 2 of 'svwrite_ver_za8_m', which expects 'uint32_t'} } */
+ svwrite_ver_za8_m (0, 0, s8, s8); /* { dg-error {passing 'svint8_t' to argument 3 of 'svwrite_ver_za8_m', which expects 'svbool_t'} } */
+ svwrite_ver_za8_m (0, 0, pg, tile); /* { dg-error {passing 'uint32_t'.* to argument 4 of 'svwrite_ver_za8_m', which expects an SVE type} } */
+ svwrite_ver_za8_m (0, 0, pg, pg); /* { dg-error {'svwrite_ver_za8_m' has no form that takes 'svbool_t' arguments} } */
+ svwrite_ver_za8_m (0, 0, pg, u16); /* { dg-error {'svwrite_ver_za8_m' has no form that takes 'svuint16_t' arguments} } */
+
+ svwrite_ver_za16_m (-1, 0, pg, u16); /* { dg-error {passing -1 to argument 1 of 'svwrite_ver_za16_m', which expects a value in the range \[0, 1\]} } */
+ svwrite_ver_za16_m (2, 0, pg, u16); /* { dg-error {passing 2 to argument 1 of 'svwrite_ver_za16_m', which expects a value in the range \[0, 1\]} } */
+
+ svwrite_ver_za32_m (-1, 0, pg, f32); /* { dg-error {passing -1 to argument 1 of 'svwrite_ver_za32_m', which expects a value in the range \[0, 3\]} } */
+ svwrite_ver_za32_m (4, 0, pg, f32); /* { dg-error {passing 4 to argument 1 of 'svwrite_ver_za32_m', which expects a value in the range \[0, 3\]} } */
+
+ svwrite_ver_za64_m (-1, 0, pg, s64); /* { dg-error {passing -1 to argument 1 of 'svwrite_ver_za64_m', which expects a value in the range \[0, 7\]} } */
+ svwrite_ver_za64_m (8, 0, pg, s64); /* { dg-error {passing 8 to argument 1 of 'svwrite_ver_za64_m', which expects a value in the range \[0, 7\]} } */
+
+ svwrite_ver_za128_m (-1, 0, pg, s8); /* { dg-error {passing -1 to argument 1 of 'svwrite_ver_za128_m', which expects a value in the range \[0, 15\]} } */
+ svwrite_ver_za128_m (16, 0, pg, s8); /* { dg-error {passing 16 to argument 1 of 'svwrite_ver_za128_m', which expects a value in the range \[0, 15\]} } */
+ svwrite_ver_za128_m (-1, 0, pg, f32); /* { dg-error {passing -1 to argument 1 of 'svwrite_ver_za128_m', which expects a value in the range \[0, 15\]} } */
+ svwrite_ver_za128_m (16, 0, pg, f32); /* { dg-error {passing 16 to argument 1 of 'svwrite_ver_za128_m', which expects a value in the range \[0, 15\]} } */
+}
+
+void
+f2 (svbool_t pg, svint8_t s8) __arm_streaming
+{
+ svwrite_ver_za8_m (0, 0, pg, s8); /* { dg-error {ACLE function 'svwrite_ver_za8_s8_m' can only be called from a function that has 'za' state} } */
+}
+
+void
+f3 (svbool_t pg, svint8_t s8) __arm_inout("za")
+{
+ svwrite_ver_za8_m (0, 0, pg, s8); /* { dg-error {ACLE function 'svwrite_ver_za8_s8_m' can only be called when SME streaming mode is enabled} } */
+}
# various architecture extensions via the .arch_extension pseudo-op.
foreach { aarch64_ext } { "fp" "simd" "crypto" "crc" "lse" "dotprod" "sve"
- "i8mm" "f32mm" "f64mm" "bf16" "sb" "sve2" "ls64" } {
+ "i8mm" "f32mm" "f64mm" "bf16" "sb" "sve2" "ls64"
+ "sme" "sme-i16i64" } {
eval [string map [list FUNC $aarch64_ext] {
proc check_effective_target_aarch64_asm_FUNC_ok { } {
if { [istarget aarch64*-*-*] } {