* [PATCH v3] LoongArch: testsuite:Added support for vector object detection.
@ 2024-01-05 6:05 chenxiaolong
2024-01-05 12:30 ` [pushed][PATCH " chenglulu
2024-01-09 19:51 ` [PATCH " Andreas Schwab
0 siblings, 2 replies; 4+ messages in thread
From: chenxiaolong @ 2024-01-05 6:05 UTC (permalink / raw)
To: gcc-patches; +Cc: xry111, i, xuchenghua, chenglulu, chenxiaolong
- Change the default vectorization "-mlasx" option to "-mlsx" because there
are many non-aligned memory accesses when using 256-bit vectorization.
- The following detection procedure is added to the target-supports.exp file:
1.check_effective_target_scalar_all_fma
2.check_effective_target_vect_int
3.check_effective_target_vect_intfloat_cvt
4.check_effective_target_vect_doubleint_cvt
5.check_effective_target_vect_intdouble_cvt
6.check_effective_target_vect_uintfloat_cvt
7.check_effective_target_vect_floatint_cvt
8.check_effective_target_vect_floatuint_cvt
9.check_effective_target_vect_shift
10.check_effective_target_vect_var_shift
11.check_effective_target_whole_vector_shift
12.check_effective_target_vect_bswap
13.check_effective_target_vect_bool_cmp
14.check_effective_target_vect_char_add
15.check_effective_target_vect_shift_char
16.check_effective_target_vect_long
17.check_effective_target_vect_float
18.check_effective_target_vect_double
19.check_effective_target_vect_long_long
20.check_effective_target_vect_perm
21.check_effective_target_vect_perm_byte
22.check_effective_target_vect_perm_short
23.check_effective_target_vect_widen_sum_hi_to_si
24.check_effective_target_vect_widen_sum_qi_to_hi
25.check_effective_target_vect_widen_sum_qi_to_hi
26.check_effective_target_vect_widen_mult_qi_to_hi
27.check_effective_target_vect_widen_mult_hi_to_si
28.check_effective_target_vect_widen_mult_qi_to_hi_pattern
29.check_effective_target_vect_widen_mult_hi_to_si_pattern
30.check_effective_target_vect_widen_mult_si_to_di_pattern
31.check_effective_target_vect_sdot_qi
32.check_effective_target_vect_udot_qi
33.check_effective_target_vect_sdot_hi
34.check_effective_target_vect_udot_hi
35.check_effective_target_vect_usad_char
36.check_effective_target_vect_avg_qi
37.check_effective_target_vect_pack_trunc
38.check_effective_target_vect_unpack
39.check_effective_target_vect_hw_misalign
40.check_effective_target_vect_gather_load_ifn
40.check_effective_target_vect_condition
42.check_effective_target_vect_cond_mixed
43.check_effective_target_vect_char_mult
44.check_effective_target_vect_short_mult
45.check_effective_target_vect_int_mult
46.check_effective_target_vect_long_mult
47.check_effective_target_vect_int_mod
48.check_effective_target_vect_extract_even_odd
49.check_effective_target_vect_interleave
50.check_effective_target_vect_call_copysignf
51.check_effective_target_vect_call_sqrtf
52.check_effective_target_vect_call_lrint
53.check_effective_target_vect_call_btrunc
54.check_effective_target_vect_call_btruncf
55.check_effective_target_vect_call_ceil
56.check_effective_target_vect_call_ceilf
57.check_effective_target_vect_call_floor
58.check_effective_target_vect_call_floorf
59.check_effective_target_vect_call_lceil
60.check_effective_target_vect_call_lfloor
61.check_effective_target_vect_logical_reduc
62.check_effective_target_section_anchors
63.check_vect_support_and_set_flags
64.check_effective_target_vect_max_reduc
65.check_effective_target_loongarch_sx
66.check_effective_target_loongarch_sx_hw
gcc/testsuite/ChangeLog:
* lib/target-supports.exp: Add LoongArch to the list of supported
targets.
---
gcc/testsuite/lib/target-supports.exp | 217 +++++++++++++++++++-------
1 file changed, 162 insertions(+), 55 deletions(-)
diff --git a/gcc/testsuite/lib/target-supports.exp b/gcc/testsuite/lib/target-supports.exp
index 167e630f5a5..9addf35ade4 100644
--- a/gcc/testsuite/lib/target-supports.exp
+++ b/gcc/testsuite/lib/target-supports.exp
@@ -3815,7 +3815,11 @@ proc add_options_for_bfloat16 { flags } {
# (fma, fms, fnma, and fnms) for both float and double.
proc check_effective_target_scalar_all_fma { } {
- return [istarget aarch64*-*-*]
+ if { [istarget aarch64*-*-*]
+ || [istarget loongarch*-*-*]} {
+ return 1
+ }
+ return 0
}
# Return 1 if the target supports compiling fixed-point,
@@ -4051,6 +4055,8 @@ proc check_effective_target_vect_int { } {
&& [check_effective_target_s390_vx])
|| ([istarget riscv*-*-*]
&& [check_effective_target_riscv_v])
+ || ([istarget loongarch*-*-*]
+ && [check_effective_target_loongarch_sx])
}}]
}
@@ -4218,7 +4224,9 @@ proc check_effective_target_vect_intfloat_cvt { } {
|| ([istarget s390*-*-*]
&& [check_effective_target_s390_vxe2])
|| ([istarget riscv*-*-*]
- && [check_effective_target_riscv_v]) }}]
+ && [check_effective_target_riscv_v])
+ || ([istarget loongarch*-*-*]
+ && [check_effective_target_loongarch_sx]) }}]
}
# Return 1 if the target supports signed double->int conversion
@@ -4239,7 +4247,9 @@ proc check_effective_target_vect_doubleint_cvt { } {
|| ([istarget s390*-*-*]
&& [check_effective_target_s390_vx])
|| ([istarget riscv*-*-*]
- && [check_effective_target_riscv_v]) }}]
+ && [check_effective_target_riscv_v])
+ || ([istarget loongarch*-*-*]
+ && [check_effective_target_loongarch_sx]) }}]
}
# Return 1 if the target supports signed int->double conversion
@@ -4260,7 +4270,9 @@ proc check_effective_target_vect_intdouble_cvt { } {
|| ([istarget s390*-*-*]
&& [check_effective_target_s390_vx])
|| ([istarget riscv*-*-*]
- && [check_effective_target_riscv_v]) }}]
+ && [check_effective_target_riscv_v])
+ || ([istarget loongarch*-*-*]
+ && [check_effective_target_loongarch_sx]) }}]
}
#Return 1 if we're supporting __int128 for target, 0 otherwise.
@@ -4293,7 +4305,9 @@ proc check_effective_target_vect_uintfloat_cvt { } {
|| ([istarget s390*-*-*]
&& [check_effective_target_s390_vxe2])
|| ([istarget riscv*-*-*]
- && [check_effective_target_riscv_v]) }}]
+ && [check_effective_target_riscv_v])
+ || ([istarget loongarch*-*-*]
+ && [check_effective_target_loongarch_sx]) }}]
}
@@ -4312,7 +4326,9 @@ proc check_effective_target_vect_floatint_cvt { } {
|| ([istarget s390*-*-*]
&& [check_effective_target_s390_vxe2])
|| ([istarget riscv*-*-*]
- && [check_effective_target_riscv_v]) }}]
+ && [check_effective_target_riscv_v])
+ || ([istarget loongarch*-*-*]
+ && [check_effective_target_loongarch_sx]) }}]
}
# Return 1 if the target supports unsigned float->int conversion
@@ -4329,7 +4345,9 @@ proc check_effective_target_vect_floatuint_cvt { } {
|| ([istarget s390*-*-*]
&& [check_effective_target_s390_vxe2])
|| ([istarget riscv*-*-*]
- && [check_effective_target_riscv_v]) }}]
+ && [check_effective_target_riscv_v])
+ || ([istarget loongarch*-*-*]
+ && [check_effective_target_loongarch_sx]) }}]
}
# Return 1 if the target supports vector integer char -> long long extend optab
@@ -4338,7 +4356,9 @@ proc check_effective_target_vect_floatuint_cvt { } {
proc check_effective_target_vect_ext_char_longlong { } {
return [check_cached_effective_target_indexed vect_ext_char_longlong {
expr { ([istarget riscv*-*-*]
- && [check_effective_target_riscv_v]) }}]
+ && [check_effective_target_riscv_v])
+ || ([istarget loongarch*-*-*]
+ && [check_effective_target_loongarch_sx]) }}]
}
# Return 1 if peeling for alignment might be profitable on the target
@@ -7462,7 +7482,9 @@ proc check_effective_target_vect_shift { } {
&& [check_effective_target_s390_vx])
|| [istarget amdgcn-*-*]
|| ([istarget riscv*-*-*]
- && [check_effective_target_riscv_v]) }}]
+ && [check_effective_target_riscv_v])
+ || ([istarget loongarch*-*-*]
+ && [check_effective_target_loongarch_sx]) }}]
}
# Return 1 if the target supports hardware vector shift by register operation.
@@ -7474,6 +7496,8 @@ proc check_effective_target_vect_var_shift { } {
|| [istarget aarch64*-*-*]
|| ([istarget riscv*-*-*]
&& [check_effective_target_riscv_v])
+ || ([istarget loongarch*-*-*]
+ && [check_effective_target_loongarch_sx])
}}]
}
@@ -7490,7 +7514,9 @@ proc check_effective_target_whole_vector_shift { } {
&& [check_effective_target_s390_vx])
|| [istarget amdgcn-*-*]
|| ([istarget riscv*-*-*]
- && [check_effective_target_riscv_v]) } {
+ && [check_effective_target_riscv_v])
+ || ([istarget loongarch*-*-*]
+ && [check_effective_target_loongarch_sx]) } {
set answer 1
} else {
set answer 0
@@ -7507,6 +7533,7 @@ proc check_effective_target_vect_bswap { } {
expr { ([istarget aarch64*-*-*]
|| [is-effective-target arm_neon]
|| [istarget amdgcn-*-*])
+ || [istarget loongarch*-*-*]
|| ([istarget s390*-*-*]
&& [check_effective_target_s390_vx]) }}]
}
@@ -7520,7 +7547,9 @@ proc check_effective_target_vect_bool_cmp { } {
|| [istarget aarch64*-*-*]
|| [is-effective-target arm_neon]
|| ([istarget riscv*-*-*]
- && [check_effective_target_riscv_v]) }}]
+ && [check_effective_target_riscv_v])
+ || ([istarget loongarch*-*-*]
+ && [check_effective_target_loongarch_sx]) }}]
}
# Return 1 if the target supports addition of char vectors for at least
@@ -7543,6 +7572,8 @@ proc check_effective_target_vect_char_add { } {
&& [check_effective_target_s390_vx])
|| ([istarget riscv*-*-*]
&& [check_effective_target_riscv_v])
+ || ([istarget loongarch*-*-*]
+ && [check_effective_target_loongarch_sx])
}}]
}
@@ -7559,7 +7590,9 @@ proc check_effective_target_vect_shift_char { } {
&& [check_effective_target_s390_vx])
|| [istarget amdgcn-*-*]
|| ([istarget riscv*-*-*]
- && [check_effective_target_riscv_v]) }}]
+ && [check_effective_target_riscv_v])
+ || ([istarget loongarch*-*-*]
+ && [check_effective_target_loongarch_sx]) }}]
}
# Return 1 if the target supports hardware vectors of long, 0 otherwise.
@@ -7580,7 +7613,9 @@ proc check_effective_target_vect_long { } {
&& [check_effective_target_s390_vx])
|| [istarget amdgcn-*-*]
|| ([istarget riscv*-*-*]
- && [check_effective_target_riscv_v]) } {
+ && [check_effective_target_riscv_v])
+ || ([istarget loongarch*-*-*]
+ && [check_effective_target_loongarch_sx]) } {
set answer 1
} else {
set answer 0
@@ -7610,7 +7645,9 @@ proc check_effective_target_vect_float { } {
&& [check_effective_target_s390_vxe])
|| [istarget amdgcn-*-*]
|| ([istarget riscv*-*-*]
- && [check_effective_target_riscv_v]) }}]
+ && [check_effective_target_riscv_v])
+ || ([istarget loongarch*-*-*]
+ && [check_effective_target_loongarch_sx]) }}]
}
# Return 1 if the target supports hardware vectors of float without
@@ -7641,7 +7678,9 @@ proc check_effective_target_vect_double { } {
&& [check_effective_target_s390_vx])
|| [istarget amdgcn-*-*]
|| ([istarget riscv*-*-*]
- && [check_effective_target_riscv_v])} }]
+ && [check_effective_target_riscv_v])
+ || ([istarget loongarch*-*-*]
+ && [check_effective_target_loongarch_sx]) }}]
}
# Return 1 if the target supports conditional addition, subtraction,
@@ -7669,7 +7708,9 @@ proc check_effective_target_vect_long_long { } {
&& [check_effective_target_has_arch_pwr8])
|| [istarget aarch64*-*-*]
|| ([istarget riscv*-*-*]
- && [check_effective_target_riscv_v])}}]
+ && [check_effective_target_riscv_v])
+ || ([istarget loongarch*-*-*]
+ && [check_effective_target_loongarch_sx])}}]
}
@@ -7724,7 +7765,9 @@ proc check_effective_target_vect_perm { } {
&& [check_effective_target_s390_vx])
|| [istarget amdgcn-*-*]
|| ([istarget riscv*-*-*]
- && [check_effective_target_riscv_v]) }}]
+ && [check_effective_target_riscv_v])
+ || ([istarget loongarch*-*-*]
+ && [check_effective_target_loongarch_sx]) }}]
}
# Return 1 if, for some VF:
@@ -7819,7 +7862,9 @@ proc check_effective_target_vect_perm_byte { } {
&& [check_effective_target_s390_vx])
|| [istarget amdgcn-*-*]
|| ([istarget riscv*-*-*]
- && [check_effective_target_riscv_v]) }}]
+ && [check_effective_target_riscv_v])
+ || ([istarget loongarch*-*-*]
+ && [check_effective_target_loongarch_sx]) }}]
}
# Return 1 if the target supports SLP permutation of 3 vectors when each
@@ -7850,7 +7895,9 @@ proc check_effective_target_vect_perm_short { } {
&& [check_effective_target_s390_vx])
|| [istarget amdgcn-*-*]
|| ([istarget riscv*-*-*]
- && [check_effective_target_riscv_v]) }}]
+ && [check_effective_target_riscv_v])
+ || ([istarget loongarch*-*-*]
+ && [check_effective_target_loongarch_sx]) }}]
}
# Return 1 if the target supports SLP permutation of 3 vectors when each
@@ -7898,6 +7945,7 @@ proc check_effective_target_vect_widen_sum_hi_to_si { } {
expr { [check_effective_target_vect_unpack]
|| [istarget powerpc*-*-*]
|| [istarget ia64-*-*]
+ || [istarget loongarch*-*-*]
|| [istarget riscv*-*-*] }}]
}
@@ -7913,7 +7961,8 @@ proc check_effective_target_vect_widen_sum_qi_to_hi { } {
expr { [check_effective_target_vect_unpack]
|| [is-effective-target arm_neon]
|| [istarget ia64-*-*]
- || [istarget riscv*-*-*] }}]
+ || [istarget riscv*-*-*]
+ || [istarget loongarch*-*-*] }}]
}
# Return 1 if the target plus current options supports a vector
@@ -7924,6 +7973,7 @@ proc check_effective_target_vect_widen_sum_qi_to_hi { } {
proc check_effective_target_vect_widen_sum_qi_to_si { } {
return [check_cached_effective_target_indexed vect_widen_sum_qi_to_si {
expr { [istarget powerpc*-*-*]
+ || [istarget loongarch*-*-*]
|| [istarget riscv*-*-*] }}]
}
@@ -7944,6 +7994,7 @@ proc check_effective_target_vect_widen_mult_qi_to_hi { } {
|| ([istarget aarch64*-*-*]
&& ![check_effective_target_aarch64_sve])
|| [is-effective-target arm_neon]
+ || [is-effective-target loongarch*-*-*]
|| ([istarget s390*-*-*]
&& [check_effective_target_s390_vx]))
|| [istarget amdgcn-*-*] }}]
@@ -7968,6 +8019,7 @@ proc check_effective_target_vect_widen_mult_hi_to_si { } {
&& ![check_effective_target_aarch64_sve])
|| [istarget i?86-*-*] || [istarget x86_64-*-*]
|| [is-effective-target arm_neon]
+ || [is-effective-target loongarch*-*-*]
|| ([istarget s390*-*-*]
&& [check_effective_target_s390_vx]))
|| [istarget amdgcn-*-*] }}]
@@ -7985,6 +8037,7 @@ proc check_effective_target_vect_widen_mult_qi_to_hi_pattern { } {
&& [check_effective_target_arm_little_endian])
|| ([istarget s390*-*-*]
&& [check_effective_target_s390_vx])
+ || [istarget loongarch*-*-*]
|| [istarget amdgcn-*-*] }}]
}
@@ -7997,6 +8050,7 @@ proc check_effective_target_vect_widen_mult_hi_to_si_pattern { } {
return [check_cached_effective_target_indexed vect_widen_mult_hi_to_si_pattern {
expr { [istarget powerpc*-*-*]
|| [istarget ia64-*-*]
+ || [istarget loongarch*-*-*]
|| [istarget i?86-*-*] || [istarget x86_64-*-*]
|| ([is-effective-target arm_neon]
&& [check_effective_target_arm_little_endian])
@@ -8014,6 +8068,7 @@ proc check_effective_target_vect_widen_mult_si_to_di_pattern { } {
return [check_cached_effective_target_indexed vect_widen_mult_si_to_di_pattern {
expr { [istarget ia64-*-*]
|| [istarget i?86-*-*] || [istarget x86_64-*-*]
+ || [istarget loongarch*-*-*]
|| ([istarget s390*-*-*]
&& [check_effective_target_s390_vx]) }}]
}
@@ -8041,7 +8096,9 @@ proc check_effective_target_vect_sdot_qi { } {
|| ([istarget mips*-*-*]
&& [et-is-effective-target mips_msa])
|| ([istarget riscv*-*-*]
- && [check_effective_target_riscv_v]) }}]
+ && [check_effective_target_riscv_v])
+ || ([istarget loongarch*-*-*]
+ && [check_effective_target_loongarch_sx]) }}]
}
# Return 1 if the target plus current options supports a vector
@@ -8058,7 +8115,9 @@ proc check_effective_target_vect_udot_qi { } {
|| ([istarget mips*-*-*]
&& [et-is-effective-target mips_msa])
|| ([istarget riscv*-*-*]
- && [check_effective_target_riscv_v]) }}]
+ && [check_effective_target_riscv_v])
+ || ([istarget loongarch*-*-*]
+ && [check_effective_target_loongarch_sx]) }}]
}
# Return 1 if the target plus current options supports a vector
@@ -8087,7 +8146,9 @@ proc check_effective_target_vect_sdot_hi { } {
|| ([istarget mips*-*-*]
&& [et-is-effective-target mips_msa])
|| ([istarget riscv*-*-*]
- && [check_effective_target_riscv_v]) }}]
+ && [check_effective_target_riscv_v])
+ || ([istarget loongarch*-*-*]
+ && [check_effective_target_loongarch_sx]) }}]
}
# Return 1 if the target plus current options supports a vector
@@ -8101,7 +8162,9 @@ proc check_effective_target_vect_udot_hi { } {
|| ([istarget mips*-*-*]
&& [et-is-effective-target mips_msa])
|| ([istarget riscv*-*-*]
- && [check_effective_target_riscv_v]) }}]
+ && [check_effective_target_riscv_v])
+ || ([istarget loongarch*-*-*]
+ && [check_effective_target_loongarch_sx]) }}]
}
# Return 1 if the target plus current options supports a vector
@@ -8118,7 +8181,9 @@ proc check_effective_target_vect_usad_char { } {
|| ([istarget powerpc*-*-*]
&& [check_p9vector_hw_available])
|| ([istarget riscv*-*-*]
- && [check_effective_target_riscv_v]) }}]
+ && [check_effective_target_riscv_v])
+ || ([istarget loongarch*-*-*]
+ && [check_effective_target_loongarch_sx]) }}]
}
# Return 1 if the target plus current options supports both signed
@@ -8128,7 +8193,9 @@ proc check_effective_target_vect_avg_qi {} {
return [expr { ([istarget aarch64*-*-*]
&& ![check_effective_target_aarch64_sve1_only])
|| ([istarget riscv*-*-*]
- && [check_effective_target_riscv_v]) }]
+ && [check_effective_target_riscv_v])
+ || ([istarget loongarch*-*-*]
+ && [check_effective_target_loongarch_sx]) }]
}
# Return 1 if the target plus current options supports both signed
@@ -8167,7 +8234,9 @@ proc check_effective_target_vect_pack_trunc { } {
&& [check_effective_target_s390_vx])
|| [istarget amdgcn*-*-*]
|| ([istarget riscv*-*-*]
- && [check_effective_target_riscv_v]) }}]
+ && [check_effective_target_riscv_v])
+ || ([istarget loongarch*-*-*]
+ && [check_effective_target_loongarch_sx]) }}]
}
# Return 1 if the target plus current options supports a vector
@@ -8189,7 +8258,9 @@ proc check_effective_target_vect_unpack { } {
&& [check_effective_target_s390_vx])
|| [istarget amdgcn*-*-*]
|| ([istarget riscv*-*-*]
- && [check_effective_target_riscv_v]) }}]
+ && [check_effective_target_riscv_v])
+ || ([istarget loongarch*-*-*]
+ && [check_effective_target_loongarch_sx]) }}]
}
# Return 1 if the target plus current options does not guarantee
@@ -8230,7 +8301,8 @@ proc check_effective_target_vect_hw_misalign { } {
|| ([istarget mips*-*-*] && [et-is-effective-target mips_msa])
|| ([istarget s390*-*-*]
&& [check_effective_target_s390_vx])
- || ([istarget riscv*-*-*]) } {
+ || ([istarget riscv*-*-*])
+ || ([istarget loongarch*-*-*]) } {
return 1
}
if { [istarget arm*-*-*]
@@ -8849,7 +8921,8 @@ proc check_effective_target_vect_gather_load_ifn { } {
proc check_effective_target_vect_scatter_store { } {
return [expr { [check_effective_target_aarch64_sve]
|| [istarget amdgcn*-*-*]
- || [check_effective_target_riscv_v] }]
+ || [check_effective_target_riscv_v]
+ || [check_effective_target_loongarch_sx] }]
}
# Return 1 if the target supports vector conditional operations, 0 otherwise.
@@ -8868,7 +8941,9 @@ proc check_effective_target_vect_condition { } {
&& [check_effective_target_s390_vx])
|| [istarget amdgcn-*-*]
|| ([istarget riscv*-*-*]
- && [check_effective_target_riscv_v]) }}]
+ && [check_effective_target_riscv_v])
+ || ([istarget loongarch*-*-*]
+ && [check_effective_target_loongarch_sx]) }}]
}
# Return 1 if the target supports vector conditional operations where
@@ -8887,7 +8962,9 @@ proc check_effective_target_vect_cond_mixed { } {
&& [check_effective_target_s390_vx])
|| [istarget amdgcn-*-*]
|| ([istarget riscv*-*-*]
- && [check_effective_target_riscv_v]) }}]
+ && [check_effective_target_riscv_v])
+ || ([istarget loongarch*-*-*]
+ && [check_effective_target_loongarch_sx]) }}]
}
# Return 1 if the target supports vector char multiplication, 0 otherwise.
@@ -8905,7 +8982,9 @@ proc check_effective_target_vect_char_mult { } {
&& [check_effective_target_s390_vx])
|| [istarget amdgcn-*-*]
|| ([istarget riscv*-*-*]
- && [check_effective_target_riscv_v]) }}]
+ && [check_effective_target_riscv_v])
+ || ([istarget loongarch*-*-*]
+ && [check_effective_target_loongarch_sx]) }}]
}
# Return 1 if the target supports vector short multiplication, 0 otherwise.
@@ -8924,7 +9003,9 @@ proc check_effective_target_vect_short_mult { } {
&& [check_effective_target_s390_vx])
|| [istarget amdgcn-*-*]
|| ([istarget riscv*-*-*]
- && [check_effective_target_riscv_v]) }}]
+ && [check_effective_target_riscv_v])
+ || ([istarget loongarch*-*-*]
+ && [check_effective_target_loongarch_sx]) }}]
}
# Return 1 if the target supports vector int multiplication, 0 otherwise.
@@ -8942,7 +9023,9 @@ proc check_effective_target_vect_int_mult { } {
&& [check_effective_target_s390_vx])
|| [istarget amdgcn-*-*]
|| ([istarget riscv*-*-*]
- && [check_effective_target_riscv_v]) }}]
+ && [check_effective_target_riscv_v])
+ || ([istarget loongarch*-*-*]
+ && [check_effective_target_loongarch_sx]) }}]
}
# Return 1 if the target supports 64 bit hardware vector
@@ -8961,7 +9044,9 @@ proc check_effective_target_vect_long_mult { } {
|| ([istarget mips*-*-*]
&& [et-is-effective-target mips_msa])
|| ([istarget riscv*-*-*]
- && [check_effective_target_riscv_v]) } {
+ && [check_effective_target_riscv_v])
+ || ([istarget loongarch*-*-*]
+ && [check_effective_target_loongarch_sx]) } {
set answer 1
} else {
set answer 0
@@ -8999,7 +9084,9 @@ proc check_effective_target_vect_extract_even_odd { } {
|| ([istarget s390*-*-*]
&& [check_effective_target_s390_vx])
|| ([istarget riscv*-*-*]
- && [check_effective_target_riscv_v]) }}]
+ && [check_effective_target_riscv_v])
+ || ([istarget loongarch*-*-*]
+ && [check_effective_target_loongarch_sx]) }}]
}
# Return 1 if the target supports vector interleaving, 0 otherwise.
@@ -9017,7 +9104,9 @@ proc check_effective_target_vect_interleave { } {
|| ([istarget s390*-*-*]
&& [check_effective_target_s390_vx])
|| ([istarget riscv*-*-*]
- && [check_effective_target_riscv_v]) }}]
+ && [check_effective_target_riscv_v])
+ || ([istarget loongarch*-*-*]
+ && [check_effective_target_loongarch_sx]) }}]
}
foreach N {2 3 4 5 6 7 8} {
@@ -9142,7 +9231,9 @@ proc check_effective_target_vect_call_copysignf { } {
|| [istarget aarch64*-*-*]
|| [istarget amdgcn-*-*]
|| ([istarget riscv*-*-*]
- && [check_effective_target_riscv_v]) }}]
+ && [check_effective_target_riscv_v])
+ || ([istarget loongarch*-*-*]
+ && [check_effective_target_loongarch_sx]) }}]
}
# Return 1 if the target supports hardware square root instructions.
@@ -9181,7 +9272,9 @@ proc check_effective_target_vect_call_sqrtf { } {
&& [check_effective_target_s390_vx])
|| [istarget amdgcn-*-*]
|| ([istarget riscv*-*-*]
- && [check_effective_target_riscv_v]) }}]
+ && [check_effective_target_riscv_v])
+ || ([istarget loongarch*-*-*]
+ && [check_effective_target_loongarch_sx]) }}]
}
# Return 1 if the target supports vector lrint calls.
@@ -9190,7 +9283,8 @@ proc check_effective_target_vect_call_lrint { } {
set et_vect_call_lrint 0
if { (([istarget i?86-*-*] || [istarget x86_64-*-*])
&& [check_effective_target_ilp32])
- || [istarget amdgcn-*-*] } {
+ || [istarget amdgcn-*-*]
+ || [istarget loongarch*-*-*] } {
set et_vect_call_lrint 1
}
@@ -9203,7 +9297,8 @@ proc check_effective_target_vect_call_lrint { } {
proc check_effective_target_vect_call_btrunc { } {
return [check_cached_effective_target_indexed vect_call_btrunc {
expr { [istarget aarch64*-*-*]
- || [istarget amdgcn-*-*] }}]
+ || [istarget amdgcn-*-*]
+ || [istarget loongarch*-*-*] }}]
}
# Return 1 if the target supports vector btruncf calls.
@@ -9211,7 +9306,8 @@ proc check_effective_target_vect_call_btrunc { } {
proc check_effective_target_vect_call_btruncf { } {
return [check_cached_effective_target_indexed vect_call_btruncf {
expr { [istarget aarch64*-*-*]
- || [istarget amdgcn-*-*] }}]
+ || [istarget amdgcn-*-*]
+ || [istarget loongarch*-*-*] }}]
}
# Return 1 if the target supports vector ceil calls.
@@ -9219,7 +9315,8 @@ proc check_effective_target_vect_call_btruncf { } {
proc check_effective_target_vect_call_ceil { } {
return [check_cached_effective_target_indexed vect_call_ceil {
expr { [istarget aarch64*-*-*]
- || [istarget amdgcn-*-*] }}]
+ || [istarget amdgcn-*-*]
+ || [istarget loongarch*-*-*] }}]
}
# Return 1 if the target supports vector ceilf calls.
@@ -9227,7 +9324,8 @@ proc check_effective_target_vect_call_ceil { } {
proc check_effective_target_vect_call_ceilf { } {
return [check_cached_effective_target_indexed vect_call_ceilf {
expr { [istarget aarch64*-*-*]
- || [istarget amdgcn-*-*] }}]
+ || [istarget amdgcn-*-*]
+ || [istarget loongarch*-*-*] }}]
}
# Return 1 if the target supports vector floor calls.
@@ -9235,7 +9333,8 @@ proc check_effective_target_vect_call_ceilf { } {
proc check_effective_target_vect_call_floor { } {
return [check_cached_effective_target_indexed vect_call_floor {
expr { [istarget aarch64*-*-*]
- || [istarget amdgcn-*-*] }}]
+ || [istarget amdgcn-*-*]
+ || [istarget loongarch*-*-*] }}]
}
# Return 1 if the target supports vector floorf calls.
@@ -9243,21 +9342,24 @@ proc check_effective_target_vect_call_floor { } {
proc check_effective_target_vect_call_floorf { } {
return [check_cached_effective_target_indexed vect_call_floorf {
expr { [istarget aarch64*-*-*]
- || [istarget amdgcn-*-*] }}]
+ || [istarget amdgcn-*-*]
+ || [istarget loongarch*-*-*] }}]
}
# Return 1 if the target supports vector lceil calls.
proc check_effective_target_vect_call_lceil { } {
return [check_cached_effective_target_indexed vect_call_lceil {
- expr { [istarget aarch64*-*-*] }}]
+ expr { [istarget aarch64*-*-*]
+ || [istarget loongarch*-*-*] }}]
}
# Return 1 if the target supports vector lfloor calls.
proc check_effective_target_vect_call_lfloor { } {
return [check_cached_effective_target_indexed vect_call_lfloor {
- expr { [istarget aarch64*-*-*] }}]
+ expr { [istarget aarch64*-*-*]
+ || [istarget loongarch*-*-*] }}]
}
# Return 1 if the target supports vector nearbyint calls.
@@ -9294,6 +9396,7 @@ proc check_effective_target_vect_logical_reduc { } {
return [expr { [check_effective_target_aarch64_sve]
|| [istarget amdgcn-*-*]
|| [check_effective_target_riscv_v]
+ || [check_effective_target_loongarch_sx]
|| [istarget i?86-*-*] || [istarget x86_64-*-*]}]
}
@@ -9311,7 +9414,8 @@ proc check_effective_target_section_anchors { } {
return [check_cached_effective_target section_anchors {
expr { [istarget powerpc*-*-*]
|| [istarget arm*-*-*]
- || [istarget aarch64*-*-*] }}]
+ || [istarget aarch64*-*-*]
+ || [istarget loongarch*-*-*] }}]
}
# Return 1 if the target supports atomic operations on "int_128" values.
@@ -11623,8 +11727,10 @@ proc check_vect_support_and_set_flags { } {
set dg-do-what-default compile
}
} elseif [istarget loongarch*-*-*] {
- lappend DEFAULT_VECTCFLAGS "-mdouble-float" "-mlasx"
- if [check_effective_target_loongarch_asx_hw] {
+ # Set the default vectorization option to "-mlsx" due to the problem
+ # of non-aligned memory access when using 256-bit vectorization.
+ lappend DEFAULT_VECTCFLAGS "-mdouble-float" "-mlsx"
+ if [check_effective_target_loongarch_sx_hw] {
set dg-do-what-default run
} else {
set dg-do-what-default compile
@@ -12190,7 +12296,8 @@ proc check_effective_target_builtin_eh_return { } {
proc check_effective_target_vect_max_reduc { } {
if { [istarget aarch64*-*-*] || [is-effective-target arm_neon]
- || [check_effective_target_riscv_v] } {
+ || [check_effective_target_riscv_v]
+ || [check_effective_target_loongarch_sx] } {
return 1
}
return 0
@@ -13205,7 +13312,7 @@ proc check_effective_target_loongarch_sx { } {
#if !defined(__loongarch_sx)
#error "LSX not defined"
#endif
- }]
+ } "-mlsx"]
}
proc check_effective_target_loongarch_sx_hw { } {
@@ -13225,7 +13332,7 @@ proc check_effective_target_loongarch_asx { } {
#if !defined(__loongarch_asx)
#error "LASX not defined"
#endif
- }]
+ } "-mlasx"]
}
proc check_effective_target_loongarch_asx_hw { } {
--
2.20.1
^ permalink raw reply [flat|nested] 4+ messages in thread
* Re: [pushed][PATCH v3] LoongArch: testsuite:Added support for vector object detection.
2024-01-05 6:05 [PATCH v3] LoongArch: testsuite:Added support for vector object detection chenxiaolong
@ 2024-01-05 12:30 ` chenglulu
2024-01-09 19:51 ` [PATCH " Andreas Schwab
1 sibling, 0 replies; 4+ messages in thread
From: chenglulu @ 2024-01-05 12:30 UTC (permalink / raw)
To: chenxiaolong, gcc-patches; +Cc: xry111, i, xuchenghua
pushed to r14-6954.
在 2024/1/5 下午2:05, chenxiaolong 写道:
> - Change the default vectorization "-mlasx" option to "-mlsx" because there
> are many non-aligned memory accesses when using 256-bit vectorization.
>
> - The following detection procedure is added to the target-supports.exp file:
>
> 1.check_effective_target_scalar_all_fma
> 2.check_effective_target_vect_int
> 3.check_effective_target_vect_intfloat_cvt
> 4.check_effective_target_vect_doubleint_cvt
> 5.check_effective_target_vect_intdouble_cvt
> 6.check_effective_target_vect_uintfloat_cvt
> 7.check_effective_target_vect_floatint_cvt
> 8.check_effective_target_vect_floatuint_cvt
> 9.check_effective_target_vect_shift
> 10.check_effective_target_vect_var_shift
> 11.check_effective_target_whole_vector_shift
> 12.check_effective_target_vect_bswap
> 13.check_effective_target_vect_bool_cmp
> 14.check_effective_target_vect_char_add
> 15.check_effective_target_vect_shift_char
> 16.check_effective_target_vect_long
> 17.check_effective_target_vect_float
> 18.check_effective_target_vect_double
> 19.check_effective_target_vect_long_long
> 20.check_effective_target_vect_perm
> 21.check_effective_target_vect_perm_byte
> 22.check_effective_target_vect_perm_short
> 23.check_effective_target_vect_widen_sum_hi_to_si
> 24.check_effective_target_vect_widen_sum_qi_to_hi
> 25.check_effective_target_vect_widen_sum_qi_to_hi
> 26.check_effective_target_vect_widen_mult_qi_to_hi
> 27.check_effective_target_vect_widen_mult_hi_to_si
> 28.check_effective_target_vect_widen_mult_qi_to_hi_pattern
> 29.check_effective_target_vect_widen_mult_hi_to_si_pattern
> 30.check_effective_target_vect_widen_mult_si_to_di_pattern
> 31.check_effective_target_vect_sdot_qi
> 32.check_effective_target_vect_udot_qi
> 33.check_effective_target_vect_sdot_hi
> 34.check_effective_target_vect_udot_hi
> 35.check_effective_target_vect_usad_char
> 36.check_effective_target_vect_avg_qi
> 37.check_effective_target_vect_pack_trunc
> 38.check_effective_target_vect_unpack
> 39.check_effective_target_vect_hw_misalign
> 40.check_effective_target_vect_gather_load_ifn
> 40.check_effective_target_vect_condition
> 42.check_effective_target_vect_cond_mixed
> 43.check_effective_target_vect_char_mult
> 44.check_effective_target_vect_short_mult
> 45.check_effective_target_vect_int_mult
> 46.check_effective_target_vect_long_mult
> 47.check_effective_target_vect_int_mod
> 48.check_effective_target_vect_extract_even_odd
> 49.check_effective_target_vect_interleave
> 50.check_effective_target_vect_call_copysignf
> 51.check_effective_target_vect_call_sqrtf
> 52.check_effective_target_vect_call_lrint
> 53.check_effective_target_vect_call_btrunc
> 54.check_effective_target_vect_call_btruncf
> 55.check_effective_target_vect_call_ceil
> 56.check_effective_target_vect_call_ceilf
> 57.check_effective_target_vect_call_floor
> 58.check_effective_target_vect_call_floorf
> 59.check_effective_target_vect_call_lceil
> 60.check_effective_target_vect_call_lfloor
> 61.check_effective_target_vect_logical_reduc
> 62.check_effective_target_section_anchors
> 63.check_vect_support_and_set_flags
> 64.check_effective_target_vect_max_reduc
> 65.check_effective_target_loongarch_sx
> 66.check_effective_target_loongarch_sx_hw
>
> gcc/testsuite/ChangeLog:
>
> * lib/target-supports.exp: Add LoongArch to the list of supported
> targets.
> ---
> gcc/testsuite/lib/target-supports.exp | 217 +++++++++++++++++++-------
> 1 file changed, 162 insertions(+), 55 deletions(-)
>
> diff --git a/gcc/testsuite/lib/target-supports.exp b/gcc/testsuite/lib/target-supports.exp
> index 167e630f5a5..9addf35ade4 100644
> --- a/gcc/testsuite/lib/target-supports.exp
> +++ b/gcc/testsuite/lib/target-supports.exp
> @@ -3815,7 +3815,11 @@ proc add_options_for_bfloat16 { flags } {
> # (fma, fms, fnma, and fnms) for both float and double.
>
> proc check_effective_target_scalar_all_fma { } {
> - return [istarget aarch64*-*-*]
> + if { [istarget aarch64*-*-*]
> + || [istarget loongarch*-*-*]} {
> + return 1
> + }
> + return 0
> }
>
> # Return 1 if the target supports compiling fixed-point,
> @@ -4051,6 +4055,8 @@ proc check_effective_target_vect_int { } {
> && [check_effective_target_s390_vx])
> || ([istarget riscv*-*-*]
> && [check_effective_target_riscv_v])
> + || ([istarget loongarch*-*-*]
> + && [check_effective_target_loongarch_sx])
> }}]
> }
>
> @@ -4218,7 +4224,9 @@ proc check_effective_target_vect_intfloat_cvt { } {
> || ([istarget s390*-*-*]
> && [check_effective_target_s390_vxe2])
> || ([istarget riscv*-*-*]
> - && [check_effective_target_riscv_v]) }}]
> + && [check_effective_target_riscv_v])
> + || ([istarget loongarch*-*-*]
> + && [check_effective_target_loongarch_sx]) }}]
> }
>
> # Return 1 if the target supports signed double->int conversion
> @@ -4239,7 +4247,9 @@ proc check_effective_target_vect_doubleint_cvt { } {
> || ([istarget s390*-*-*]
> && [check_effective_target_s390_vx])
> || ([istarget riscv*-*-*]
> - && [check_effective_target_riscv_v]) }}]
> + && [check_effective_target_riscv_v])
> + || ([istarget loongarch*-*-*]
> + && [check_effective_target_loongarch_sx]) }}]
> }
>
> # Return 1 if the target supports signed int->double conversion
> @@ -4260,7 +4270,9 @@ proc check_effective_target_vect_intdouble_cvt { } {
> || ([istarget s390*-*-*]
> && [check_effective_target_s390_vx])
> || ([istarget riscv*-*-*]
> - && [check_effective_target_riscv_v]) }}]
> + && [check_effective_target_riscv_v])
> + || ([istarget loongarch*-*-*]
> + && [check_effective_target_loongarch_sx]) }}]
> }
>
> #Return 1 if we're supporting __int128 for target, 0 otherwise.
> @@ -4293,7 +4305,9 @@ proc check_effective_target_vect_uintfloat_cvt { } {
> || ([istarget s390*-*-*]
> && [check_effective_target_s390_vxe2])
> || ([istarget riscv*-*-*]
> - && [check_effective_target_riscv_v]) }}]
> + && [check_effective_target_riscv_v])
> + || ([istarget loongarch*-*-*]
> + && [check_effective_target_loongarch_sx]) }}]
> }
>
>
> @@ -4312,7 +4326,9 @@ proc check_effective_target_vect_floatint_cvt { } {
> || ([istarget s390*-*-*]
> && [check_effective_target_s390_vxe2])
> || ([istarget riscv*-*-*]
> - && [check_effective_target_riscv_v]) }}]
> + && [check_effective_target_riscv_v])
> + || ([istarget loongarch*-*-*]
> + && [check_effective_target_loongarch_sx]) }}]
> }
>
> # Return 1 if the target supports unsigned float->int conversion
> @@ -4329,7 +4345,9 @@ proc check_effective_target_vect_floatuint_cvt { } {
> || ([istarget s390*-*-*]
> && [check_effective_target_s390_vxe2])
> || ([istarget riscv*-*-*]
> - && [check_effective_target_riscv_v]) }}]
> + && [check_effective_target_riscv_v])
> + || ([istarget loongarch*-*-*]
> + && [check_effective_target_loongarch_sx]) }}]
> }
>
> # Return 1 if the target supports vector integer char -> long long extend optab
> @@ -4338,7 +4356,9 @@ proc check_effective_target_vect_floatuint_cvt { } {
> proc check_effective_target_vect_ext_char_longlong { } {
> return [check_cached_effective_target_indexed vect_ext_char_longlong {
> expr { ([istarget riscv*-*-*]
> - && [check_effective_target_riscv_v]) }}]
> + && [check_effective_target_riscv_v])
> + || ([istarget loongarch*-*-*]
> + && [check_effective_target_loongarch_sx]) }}]
> }
>
> # Return 1 if peeling for alignment might be profitable on the target
> @@ -7462,7 +7482,9 @@ proc check_effective_target_vect_shift { } {
> && [check_effective_target_s390_vx])
> || [istarget amdgcn-*-*]
> || ([istarget riscv*-*-*]
> - && [check_effective_target_riscv_v]) }}]
> + && [check_effective_target_riscv_v])
> + || ([istarget loongarch*-*-*]
> + && [check_effective_target_loongarch_sx]) }}]
> }
>
> # Return 1 if the target supports hardware vector shift by register operation.
> @@ -7474,6 +7496,8 @@ proc check_effective_target_vect_var_shift { } {
> || [istarget aarch64*-*-*]
> || ([istarget riscv*-*-*]
> && [check_effective_target_riscv_v])
> + || ([istarget loongarch*-*-*]
> + && [check_effective_target_loongarch_sx])
> }}]
> }
>
> @@ -7490,7 +7514,9 @@ proc check_effective_target_whole_vector_shift { } {
> && [check_effective_target_s390_vx])
> || [istarget amdgcn-*-*]
> || ([istarget riscv*-*-*]
> - && [check_effective_target_riscv_v]) } {
> + && [check_effective_target_riscv_v])
> + || ([istarget loongarch*-*-*]
> + && [check_effective_target_loongarch_sx]) } {
> set answer 1
> } else {
> set answer 0
> @@ -7507,6 +7533,7 @@ proc check_effective_target_vect_bswap { } {
> expr { ([istarget aarch64*-*-*]
> || [is-effective-target arm_neon]
> || [istarget amdgcn-*-*])
> + || [istarget loongarch*-*-*]
> || ([istarget s390*-*-*]
> && [check_effective_target_s390_vx]) }}]
> }
> @@ -7520,7 +7547,9 @@ proc check_effective_target_vect_bool_cmp { } {
> || [istarget aarch64*-*-*]
> || [is-effective-target arm_neon]
> || ([istarget riscv*-*-*]
> - && [check_effective_target_riscv_v]) }}]
> + && [check_effective_target_riscv_v])
> + || ([istarget loongarch*-*-*]
> + && [check_effective_target_loongarch_sx]) }}]
> }
>
> # Return 1 if the target supports addition of char vectors for at least
> @@ -7543,6 +7572,8 @@ proc check_effective_target_vect_char_add { } {
> && [check_effective_target_s390_vx])
> || ([istarget riscv*-*-*]
> && [check_effective_target_riscv_v])
> + || ([istarget loongarch*-*-*]
> + && [check_effective_target_loongarch_sx])
> }}]
> }
>
> @@ -7559,7 +7590,9 @@ proc check_effective_target_vect_shift_char { } {
> && [check_effective_target_s390_vx])
> || [istarget amdgcn-*-*]
> || ([istarget riscv*-*-*]
> - && [check_effective_target_riscv_v]) }}]
> + && [check_effective_target_riscv_v])
> + || ([istarget loongarch*-*-*]
> + && [check_effective_target_loongarch_sx]) }}]
> }
>
> # Return 1 if the target supports hardware vectors of long, 0 otherwise.
> @@ -7580,7 +7613,9 @@ proc check_effective_target_vect_long { } {
> && [check_effective_target_s390_vx])
> || [istarget amdgcn-*-*]
> || ([istarget riscv*-*-*]
> - && [check_effective_target_riscv_v]) } {
> + && [check_effective_target_riscv_v])
> + || ([istarget loongarch*-*-*]
> + && [check_effective_target_loongarch_sx]) } {
> set answer 1
> } else {
> set answer 0
> @@ -7610,7 +7645,9 @@ proc check_effective_target_vect_float { } {
> && [check_effective_target_s390_vxe])
> || [istarget amdgcn-*-*]
> || ([istarget riscv*-*-*]
> - && [check_effective_target_riscv_v]) }}]
> + && [check_effective_target_riscv_v])
> + || ([istarget loongarch*-*-*]
> + && [check_effective_target_loongarch_sx]) }}]
> }
>
> # Return 1 if the target supports hardware vectors of float without
> @@ -7641,7 +7678,9 @@ proc check_effective_target_vect_double { } {
> && [check_effective_target_s390_vx])
> || [istarget amdgcn-*-*]
> || ([istarget riscv*-*-*]
> - && [check_effective_target_riscv_v])} }]
> + && [check_effective_target_riscv_v])
> + || ([istarget loongarch*-*-*]
> + && [check_effective_target_loongarch_sx]) }}]
> }
>
> # Return 1 if the target supports conditional addition, subtraction,
> @@ -7669,7 +7708,9 @@ proc check_effective_target_vect_long_long { } {
> && [check_effective_target_has_arch_pwr8])
> || [istarget aarch64*-*-*]
> || ([istarget riscv*-*-*]
> - && [check_effective_target_riscv_v])}}]
> + && [check_effective_target_riscv_v])
> + || ([istarget loongarch*-*-*]
> + && [check_effective_target_loongarch_sx])}}]
> }
>
>
> @@ -7724,7 +7765,9 @@ proc check_effective_target_vect_perm { } {
> && [check_effective_target_s390_vx])
> || [istarget amdgcn-*-*]
> || ([istarget riscv*-*-*]
> - && [check_effective_target_riscv_v]) }}]
> + && [check_effective_target_riscv_v])
> + || ([istarget loongarch*-*-*]
> + && [check_effective_target_loongarch_sx]) }}]
> }
>
> # Return 1 if, for some VF:
> @@ -7819,7 +7862,9 @@ proc check_effective_target_vect_perm_byte { } {
> && [check_effective_target_s390_vx])
> || [istarget amdgcn-*-*]
> || ([istarget riscv*-*-*]
> - && [check_effective_target_riscv_v]) }}]
> + && [check_effective_target_riscv_v])
> + || ([istarget loongarch*-*-*]
> + && [check_effective_target_loongarch_sx]) }}]
> }
>
> # Return 1 if the target supports SLP permutation of 3 vectors when each
> @@ -7850,7 +7895,9 @@ proc check_effective_target_vect_perm_short { } {
> && [check_effective_target_s390_vx])
> || [istarget amdgcn-*-*]
> || ([istarget riscv*-*-*]
> - && [check_effective_target_riscv_v]) }}]
> + && [check_effective_target_riscv_v])
> + || ([istarget loongarch*-*-*]
> + && [check_effective_target_loongarch_sx]) }}]
> }
>
> # Return 1 if the target supports SLP permutation of 3 vectors when each
> @@ -7898,6 +7945,7 @@ proc check_effective_target_vect_widen_sum_hi_to_si { } {
> expr { [check_effective_target_vect_unpack]
> || [istarget powerpc*-*-*]
> || [istarget ia64-*-*]
> + || [istarget loongarch*-*-*]
> || [istarget riscv*-*-*] }}]
> }
>
> @@ -7913,7 +7961,8 @@ proc check_effective_target_vect_widen_sum_qi_to_hi { } {
> expr { [check_effective_target_vect_unpack]
> || [is-effective-target arm_neon]
> || [istarget ia64-*-*]
> - || [istarget riscv*-*-*] }}]
> + || [istarget riscv*-*-*]
> + || [istarget loongarch*-*-*] }}]
> }
>
> # Return 1 if the target plus current options supports a vector
> @@ -7924,6 +7973,7 @@ proc check_effective_target_vect_widen_sum_qi_to_hi { } {
> proc check_effective_target_vect_widen_sum_qi_to_si { } {
> return [check_cached_effective_target_indexed vect_widen_sum_qi_to_si {
> expr { [istarget powerpc*-*-*]
> + || [istarget loongarch*-*-*]
> || [istarget riscv*-*-*] }}]
> }
>
> @@ -7944,6 +7994,7 @@ proc check_effective_target_vect_widen_mult_qi_to_hi { } {
> || ([istarget aarch64*-*-*]
> && ![check_effective_target_aarch64_sve])
> || [is-effective-target arm_neon]
> + || [is-effective-target loongarch*-*-*]
> || ([istarget s390*-*-*]
> && [check_effective_target_s390_vx]))
> || [istarget amdgcn-*-*] }}]
> @@ -7968,6 +8019,7 @@ proc check_effective_target_vect_widen_mult_hi_to_si { } {
> && ![check_effective_target_aarch64_sve])
> || [istarget i?86-*-*] || [istarget x86_64-*-*]
> || [is-effective-target arm_neon]
> + || [is-effective-target loongarch*-*-*]
> || ([istarget s390*-*-*]
> && [check_effective_target_s390_vx]))
> || [istarget amdgcn-*-*] }}]
> @@ -7985,6 +8037,7 @@ proc check_effective_target_vect_widen_mult_qi_to_hi_pattern { } {
> && [check_effective_target_arm_little_endian])
> || ([istarget s390*-*-*]
> && [check_effective_target_s390_vx])
> + || [istarget loongarch*-*-*]
> || [istarget amdgcn-*-*] }}]
> }
>
> @@ -7997,6 +8050,7 @@ proc check_effective_target_vect_widen_mult_hi_to_si_pattern { } {
> return [check_cached_effective_target_indexed vect_widen_mult_hi_to_si_pattern {
> expr { [istarget powerpc*-*-*]
> || [istarget ia64-*-*]
> + || [istarget loongarch*-*-*]
> || [istarget i?86-*-*] || [istarget x86_64-*-*]
> || ([is-effective-target arm_neon]
> && [check_effective_target_arm_little_endian])
> @@ -8014,6 +8068,7 @@ proc check_effective_target_vect_widen_mult_si_to_di_pattern { } {
> return [check_cached_effective_target_indexed vect_widen_mult_si_to_di_pattern {
> expr { [istarget ia64-*-*]
> || [istarget i?86-*-*] || [istarget x86_64-*-*]
> + || [istarget loongarch*-*-*]
> || ([istarget s390*-*-*]
> && [check_effective_target_s390_vx]) }}]
> }
> @@ -8041,7 +8096,9 @@ proc check_effective_target_vect_sdot_qi { } {
> || ([istarget mips*-*-*]
> && [et-is-effective-target mips_msa])
> || ([istarget riscv*-*-*]
> - && [check_effective_target_riscv_v]) }}]
> + && [check_effective_target_riscv_v])
> + || ([istarget loongarch*-*-*]
> + && [check_effective_target_loongarch_sx]) }}]
> }
>
> # Return 1 if the target plus current options supports a vector
> @@ -8058,7 +8115,9 @@ proc check_effective_target_vect_udot_qi { } {
> || ([istarget mips*-*-*]
> && [et-is-effective-target mips_msa])
> || ([istarget riscv*-*-*]
> - && [check_effective_target_riscv_v]) }}]
> + && [check_effective_target_riscv_v])
> + || ([istarget loongarch*-*-*]
> + && [check_effective_target_loongarch_sx]) }}]
> }
>
> # Return 1 if the target plus current options supports a vector
> @@ -8087,7 +8146,9 @@ proc check_effective_target_vect_sdot_hi { } {
> || ([istarget mips*-*-*]
> && [et-is-effective-target mips_msa])
> || ([istarget riscv*-*-*]
> - && [check_effective_target_riscv_v]) }}]
> + && [check_effective_target_riscv_v])
> + || ([istarget loongarch*-*-*]
> + && [check_effective_target_loongarch_sx]) }}]
> }
>
> # Return 1 if the target plus current options supports a vector
> @@ -8101,7 +8162,9 @@ proc check_effective_target_vect_udot_hi { } {
> || ([istarget mips*-*-*]
> && [et-is-effective-target mips_msa])
> || ([istarget riscv*-*-*]
> - && [check_effective_target_riscv_v]) }}]
> + && [check_effective_target_riscv_v])
> + || ([istarget loongarch*-*-*]
> + && [check_effective_target_loongarch_sx]) }}]
> }
>
> # Return 1 if the target plus current options supports a vector
> @@ -8118,7 +8181,9 @@ proc check_effective_target_vect_usad_char { } {
> || ([istarget powerpc*-*-*]
> && [check_p9vector_hw_available])
> || ([istarget riscv*-*-*]
> - && [check_effective_target_riscv_v]) }}]
> + && [check_effective_target_riscv_v])
> + || ([istarget loongarch*-*-*]
> + && [check_effective_target_loongarch_sx]) }}]
> }
>
> # Return 1 if the target plus current options supports both signed
> @@ -8128,7 +8193,9 @@ proc check_effective_target_vect_avg_qi {} {
> return [expr { ([istarget aarch64*-*-*]
> && ![check_effective_target_aarch64_sve1_only])
> || ([istarget riscv*-*-*]
> - && [check_effective_target_riscv_v]) }]
> + && [check_effective_target_riscv_v])
> + || ([istarget loongarch*-*-*]
> + && [check_effective_target_loongarch_sx]) }]
> }
>
> # Return 1 if the target plus current options supports both signed
> @@ -8167,7 +8234,9 @@ proc check_effective_target_vect_pack_trunc { } {
> && [check_effective_target_s390_vx])
> || [istarget amdgcn*-*-*]
> || ([istarget riscv*-*-*]
> - && [check_effective_target_riscv_v]) }}]
> + && [check_effective_target_riscv_v])
> + || ([istarget loongarch*-*-*]
> + && [check_effective_target_loongarch_sx]) }}]
> }
>
> # Return 1 if the target plus current options supports a vector
> @@ -8189,7 +8258,9 @@ proc check_effective_target_vect_unpack { } {
> && [check_effective_target_s390_vx])
> || [istarget amdgcn*-*-*]
> || ([istarget riscv*-*-*]
> - && [check_effective_target_riscv_v]) }}]
> + && [check_effective_target_riscv_v])
> + || ([istarget loongarch*-*-*]
> + && [check_effective_target_loongarch_sx]) }}]
> }
>
> # Return 1 if the target plus current options does not guarantee
> @@ -8230,7 +8301,8 @@ proc check_effective_target_vect_hw_misalign { } {
> || ([istarget mips*-*-*] && [et-is-effective-target mips_msa])
> || ([istarget s390*-*-*]
> && [check_effective_target_s390_vx])
> - || ([istarget riscv*-*-*]) } {
> + || ([istarget riscv*-*-*])
> + || ([istarget loongarch*-*-*]) } {
> return 1
> }
> if { [istarget arm*-*-*]
> @@ -8849,7 +8921,8 @@ proc check_effective_target_vect_gather_load_ifn { } {
> proc check_effective_target_vect_scatter_store { } {
> return [expr { [check_effective_target_aarch64_sve]
> || [istarget amdgcn*-*-*]
> - || [check_effective_target_riscv_v] }]
> + || [check_effective_target_riscv_v]
> + || [check_effective_target_loongarch_sx] }]
> }
>
> # Return 1 if the target supports vector conditional operations, 0 otherwise.
> @@ -8868,7 +8941,9 @@ proc check_effective_target_vect_condition { } {
> && [check_effective_target_s390_vx])
> || [istarget amdgcn-*-*]
> || ([istarget riscv*-*-*]
> - && [check_effective_target_riscv_v]) }}]
> + && [check_effective_target_riscv_v])
> + || ([istarget loongarch*-*-*]
> + && [check_effective_target_loongarch_sx]) }}]
> }
>
> # Return 1 if the target supports vector conditional operations where
> @@ -8887,7 +8962,9 @@ proc check_effective_target_vect_cond_mixed { } {
> && [check_effective_target_s390_vx])
> || [istarget amdgcn-*-*]
> || ([istarget riscv*-*-*]
> - && [check_effective_target_riscv_v]) }}]
> + && [check_effective_target_riscv_v])
> + || ([istarget loongarch*-*-*]
> + && [check_effective_target_loongarch_sx]) }}]
> }
>
> # Return 1 if the target supports vector char multiplication, 0 otherwise.
> @@ -8905,7 +8982,9 @@ proc check_effective_target_vect_char_mult { } {
> && [check_effective_target_s390_vx])
> || [istarget amdgcn-*-*]
> || ([istarget riscv*-*-*]
> - && [check_effective_target_riscv_v]) }}]
> + && [check_effective_target_riscv_v])
> + || ([istarget loongarch*-*-*]
> + && [check_effective_target_loongarch_sx]) }}]
> }
>
> # Return 1 if the target supports vector short multiplication, 0 otherwise.
> @@ -8924,7 +9003,9 @@ proc check_effective_target_vect_short_mult { } {
> && [check_effective_target_s390_vx])
> || [istarget amdgcn-*-*]
> || ([istarget riscv*-*-*]
> - && [check_effective_target_riscv_v]) }}]
> + && [check_effective_target_riscv_v])
> + || ([istarget loongarch*-*-*]
> + && [check_effective_target_loongarch_sx]) }}]
> }
>
> # Return 1 if the target supports vector int multiplication, 0 otherwise.
> @@ -8942,7 +9023,9 @@ proc check_effective_target_vect_int_mult { } {
> && [check_effective_target_s390_vx])
> || [istarget amdgcn-*-*]
> || ([istarget riscv*-*-*]
> - && [check_effective_target_riscv_v]) }}]
> + && [check_effective_target_riscv_v])
> + || ([istarget loongarch*-*-*]
> + && [check_effective_target_loongarch_sx]) }}]
> }
>
> # Return 1 if the target supports 64 bit hardware vector
> @@ -8961,7 +9044,9 @@ proc check_effective_target_vect_long_mult { } {
> || ([istarget mips*-*-*]
> && [et-is-effective-target mips_msa])
> || ([istarget riscv*-*-*]
> - && [check_effective_target_riscv_v]) } {
> + && [check_effective_target_riscv_v])
> + || ([istarget loongarch*-*-*]
> + && [check_effective_target_loongarch_sx]) } {
> set answer 1
> } else {
> set answer 0
> @@ -8999,7 +9084,9 @@ proc check_effective_target_vect_extract_even_odd { } {
> || ([istarget s390*-*-*]
> && [check_effective_target_s390_vx])
> || ([istarget riscv*-*-*]
> - && [check_effective_target_riscv_v]) }}]
> + && [check_effective_target_riscv_v])
> + || ([istarget loongarch*-*-*]
> + && [check_effective_target_loongarch_sx]) }}]
> }
>
> # Return 1 if the target supports vector interleaving, 0 otherwise.
> @@ -9017,7 +9104,9 @@ proc check_effective_target_vect_interleave { } {
> || ([istarget s390*-*-*]
> && [check_effective_target_s390_vx])
> || ([istarget riscv*-*-*]
> - && [check_effective_target_riscv_v]) }}]
> + && [check_effective_target_riscv_v])
> + || ([istarget loongarch*-*-*]
> + && [check_effective_target_loongarch_sx]) }}]
> }
>
> foreach N {2 3 4 5 6 7 8} {
> @@ -9142,7 +9231,9 @@ proc check_effective_target_vect_call_copysignf { } {
> || [istarget aarch64*-*-*]
> || [istarget amdgcn-*-*]
> || ([istarget riscv*-*-*]
> - && [check_effective_target_riscv_v]) }}]
> + && [check_effective_target_riscv_v])
> + || ([istarget loongarch*-*-*]
> + && [check_effective_target_loongarch_sx]) }}]
> }
>
> # Return 1 if the target supports hardware square root instructions.
> @@ -9181,7 +9272,9 @@ proc check_effective_target_vect_call_sqrtf { } {
> && [check_effective_target_s390_vx])
> || [istarget amdgcn-*-*]
> || ([istarget riscv*-*-*]
> - && [check_effective_target_riscv_v]) }}]
> + && [check_effective_target_riscv_v])
> + || ([istarget loongarch*-*-*]
> + && [check_effective_target_loongarch_sx]) }}]
> }
>
> # Return 1 if the target supports vector lrint calls.
> @@ -9190,7 +9283,8 @@ proc check_effective_target_vect_call_lrint { } {
> set et_vect_call_lrint 0
> if { (([istarget i?86-*-*] || [istarget x86_64-*-*])
> && [check_effective_target_ilp32])
> - || [istarget amdgcn-*-*] } {
> + || [istarget amdgcn-*-*]
> + || [istarget loongarch*-*-*] } {
> set et_vect_call_lrint 1
> }
>
> @@ -9203,7 +9297,8 @@ proc check_effective_target_vect_call_lrint { } {
> proc check_effective_target_vect_call_btrunc { } {
> return [check_cached_effective_target_indexed vect_call_btrunc {
> expr { [istarget aarch64*-*-*]
> - || [istarget amdgcn-*-*] }}]
> + || [istarget amdgcn-*-*]
> + || [istarget loongarch*-*-*] }}]
> }
>
> # Return 1 if the target supports vector btruncf calls.
> @@ -9211,7 +9306,8 @@ proc check_effective_target_vect_call_btrunc { } {
> proc check_effective_target_vect_call_btruncf { } {
> return [check_cached_effective_target_indexed vect_call_btruncf {
> expr { [istarget aarch64*-*-*]
> - || [istarget amdgcn-*-*] }}]
> + || [istarget amdgcn-*-*]
> + || [istarget loongarch*-*-*] }}]
> }
>
> # Return 1 if the target supports vector ceil calls.
> @@ -9219,7 +9315,8 @@ proc check_effective_target_vect_call_btruncf { } {
> proc check_effective_target_vect_call_ceil { } {
> return [check_cached_effective_target_indexed vect_call_ceil {
> expr { [istarget aarch64*-*-*]
> - || [istarget amdgcn-*-*] }}]
> + || [istarget amdgcn-*-*]
> + || [istarget loongarch*-*-*] }}]
> }
>
> # Return 1 if the target supports vector ceilf calls.
> @@ -9227,7 +9324,8 @@ proc check_effective_target_vect_call_ceil { } {
> proc check_effective_target_vect_call_ceilf { } {
> return [check_cached_effective_target_indexed vect_call_ceilf {
> expr { [istarget aarch64*-*-*]
> - || [istarget amdgcn-*-*] }}]
> + || [istarget amdgcn-*-*]
> + || [istarget loongarch*-*-*] }}]
> }
>
> # Return 1 if the target supports vector floor calls.
> @@ -9235,7 +9333,8 @@ proc check_effective_target_vect_call_ceilf { } {
> proc check_effective_target_vect_call_floor { } {
> return [check_cached_effective_target_indexed vect_call_floor {
> expr { [istarget aarch64*-*-*]
> - || [istarget amdgcn-*-*] }}]
> + || [istarget amdgcn-*-*]
> + || [istarget loongarch*-*-*] }}]
> }
>
> # Return 1 if the target supports vector floorf calls.
> @@ -9243,21 +9342,24 @@ proc check_effective_target_vect_call_floor { } {
> proc check_effective_target_vect_call_floorf { } {
> return [check_cached_effective_target_indexed vect_call_floorf {
> expr { [istarget aarch64*-*-*]
> - || [istarget amdgcn-*-*] }}]
> + || [istarget amdgcn-*-*]
> + || [istarget loongarch*-*-*] }}]
> }
>
> # Return 1 if the target supports vector lceil calls.
>
> proc check_effective_target_vect_call_lceil { } {
> return [check_cached_effective_target_indexed vect_call_lceil {
> - expr { [istarget aarch64*-*-*] }}]
> + expr { [istarget aarch64*-*-*]
> + || [istarget loongarch*-*-*] }}]
> }
>
> # Return 1 if the target supports vector lfloor calls.
>
> proc check_effective_target_vect_call_lfloor { } {
> return [check_cached_effective_target_indexed vect_call_lfloor {
> - expr { [istarget aarch64*-*-*] }}]
> + expr { [istarget aarch64*-*-*]
> + || [istarget loongarch*-*-*] }}]
> }
>
> # Return 1 if the target supports vector nearbyint calls.
> @@ -9294,6 +9396,7 @@ proc check_effective_target_vect_logical_reduc { } {
> return [expr { [check_effective_target_aarch64_sve]
> || [istarget amdgcn-*-*]
> || [check_effective_target_riscv_v]
> + || [check_effective_target_loongarch_sx]
> || [istarget i?86-*-*] || [istarget x86_64-*-*]}]
> }
>
> @@ -9311,7 +9414,8 @@ proc check_effective_target_section_anchors { } {
> return [check_cached_effective_target section_anchors {
> expr { [istarget powerpc*-*-*]
> || [istarget arm*-*-*]
> - || [istarget aarch64*-*-*] }}]
> + || [istarget aarch64*-*-*]
> + || [istarget loongarch*-*-*] }}]
> }
>
> # Return 1 if the target supports atomic operations on "int_128" values.
> @@ -11623,8 +11727,10 @@ proc check_vect_support_and_set_flags { } {
> set dg-do-what-default compile
> }
> } elseif [istarget loongarch*-*-*] {
> - lappend DEFAULT_VECTCFLAGS "-mdouble-float" "-mlasx"
> - if [check_effective_target_loongarch_asx_hw] {
> + # Set the default vectorization option to "-mlsx" due to the problem
> + # of non-aligned memory access when using 256-bit vectorization.
> + lappend DEFAULT_VECTCFLAGS "-mdouble-float" "-mlsx"
> + if [check_effective_target_loongarch_sx_hw] {
> set dg-do-what-default run
> } else {
> set dg-do-what-default compile
> @@ -12190,7 +12296,8 @@ proc check_effective_target_builtin_eh_return { } {
>
> proc check_effective_target_vect_max_reduc { } {
> if { [istarget aarch64*-*-*] || [is-effective-target arm_neon]
> - || [check_effective_target_riscv_v] } {
> + || [check_effective_target_riscv_v]
> + || [check_effective_target_loongarch_sx] } {
> return 1
> }
> return 0
> @@ -13205,7 +13312,7 @@ proc check_effective_target_loongarch_sx { } {
> #if !defined(__loongarch_sx)
> #error "LSX not defined"
> #endif
> - }]
> + } "-mlsx"]
> }
>
> proc check_effective_target_loongarch_sx_hw { } {
> @@ -13225,7 +13332,7 @@ proc check_effective_target_loongarch_asx { } {
> #if !defined(__loongarch_asx)
> #error "LASX not defined"
> #endif
> - }]
> + } "-mlasx"]
> }
>
> proc check_effective_target_loongarch_asx_hw { } {
^ permalink raw reply [flat|nested] 4+ messages in thread
* Re: [PATCH v3] LoongArch: testsuite:Added support for vector object detection.
2024-01-05 6:05 [PATCH v3] LoongArch: testsuite:Added support for vector object detection chenxiaolong
2024-01-05 12:30 ` [pushed][PATCH " chenglulu
@ 2024-01-09 19:51 ` Andreas Schwab
2024-01-10 1:37 ` chenglulu
1 sibling, 1 reply; 4+ messages in thread
From: Andreas Schwab @ 2024-01-09 19:51 UTC (permalink / raw)
To: chenxiaolong; +Cc: gcc-patches, xry111, i, xuchenghua, chenglulu
gcc: gcc.dg/vect/vect-outer-4a-big-array.c -flto -ffat-lto-objects: error executing dg-final: unknown effective target keyword `loongarch*-*-*'
gcc: gcc.dg/vect/vect-outer-4a-big-array.c: error executing dg-final: unknown effective target keyword `loongarch*-*-*'
gcc: gcc.dg/vect/vect-outer-4a.c -flto -ffat-lto-objects: error executing dg-final: unknown effective target keyword `loongarch*-*-*'
gcc: gcc.dg/vect/vect-outer-4a.c: error executing dg-final: unknown effective target keyword `loongarch*-*-*'
gcc: gcc.dg/vect/vect-outer-4b-big-array.c -flto -ffat-lto-objects: error executing dg-final: unknown effective target keyword `loongarch*-*-*'
gcc: gcc.dg/vect/vect-outer-4b-big-array.c: error executing dg-final: unknown effective target keyword `loongarch*-*-*'
gcc: gcc.dg/vect/vect-outer-4b.c -flto -ffat-lto-objects: error executing dg-final: unknown effective target keyword `loongarch*-*-*'
gcc: gcc.dg/vect/vect-outer-4b.c: error executing dg-final: unknown effective target keyword `loongarch*-*-*'
--
Andreas Schwab, schwab@linux-m68k.org
GPG Key fingerprint = 7578 EB47 D4E5 4D69 2510 2552 DF73 E780 A9DA AEC1
"And now for something completely different."
^ permalink raw reply [flat|nested] 4+ messages in thread
* Re: [PATCH v3] LoongArch: testsuite:Added support for vector object detection.
2024-01-09 19:51 ` [PATCH " Andreas Schwab
@ 2024-01-10 1:37 ` chenglulu
0 siblings, 0 replies; 4+ messages in thread
From: chenglulu @ 2024-01-10 1:37 UTC (permalink / raw)
To: Andreas Schwab, chenxiaolong; +Cc: gcc-patches, xry111, i, xuchenghua
在 2024/1/10 上午3:51, Andreas Schwab 写道:
> gcc: gcc.dg/vect/vect-outer-4a-big-array.c -flto -ffat-lto-objects: error executing dg-final: unknown effective target keyword `loongarch*-*-*'
> gcc: gcc.dg/vect/vect-outer-4a-big-array.c: error executing dg-final: unknown effective target keyword `loongarch*-*-*'
> gcc: gcc.dg/vect/vect-outer-4a.c -flto -ffat-lto-objects: error executing dg-final: unknown effective target keyword `loongarch*-*-*'
> gcc: gcc.dg/vect/vect-outer-4a.c: error executing dg-final: unknown effective target keyword `loongarch*-*-*'
> gcc: gcc.dg/vect/vect-outer-4b-big-array.c -flto -ffat-lto-objects: error executing dg-final: unknown effective target keyword `loongarch*-*-*'
> gcc: gcc.dg/vect/vect-outer-4b-big-array.c: error executing dg-final: unknown effective target keyword `loongarch*-*-*'
> gcc: gcc.dg/vect/vect-outer-4b.c -flto -ffat-lto-objects: error executing dg-final: unknown effective target keyword `loongarch*-*-*'
> gcc: gcc.dg/vect/vect-outer-4b.c: error executing dg-final: unknown effective target keyword `loongarch*-*-*'
>
Sorry, we'll fix it as soon as possible.
^ permalink raw reply [flat|nested] 4+ messages in thread
end of thread, other threads:[~2024-01-10 1:37 UTC | newest]
Thread overview: 4+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2024-01-05 6:05 [PATCH v3] LoongArch: testsuite:Added support for vector object detection chenxiaolong
2024-01-05 12:30 ` [pushed][PATCH " chenglulu
2024-01-09 19:51 ` [PATCH " Andreas Schwab
2024-01-10 1:37 ` chenglulu
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).