Implementation notes: amd64, gcc123, crypto_dh/gls254

Computer: gcc123
Architecture: amd64
CPU ID: GenuineIntel-000206d7-bfebfbff
SUPERCOP version: 20220213
Operation: crypto_dh
Primitive: gls254
TimeObject sizeTest sizeImplementationCompilerBenchmark dateSUPERCOP version
298710129483 16 0151213 852 1744T:refclang_-march=native_-O3_-fomit-frame-pointer_-fwrapv_-Qunused-arguments_-fPIC_-fPIE2021011620210114
302526125595 16 0145757 852 1744T:refclang_-march=native_-O2_-fomit-frame-pointer_-fwrapv_-Qunused-arguments_-fPIC_-fPIE2021011620210114
302841125595 16 0145757 852 1744T:refclang_-march=native_-O_-fomit-frame-pointer_-fwrapv_-Qunused-arguments_-fPIC_-fPIE2021011620210114
316148113539 16 0130211 844 1712T:refclang_-march=native_-Os_-fomit-frame-pointer_-fwrapv_-Qunused-arguments_-fPIC_-fPIE2021011620210114
33039712833 98336 030566 99188 1744T:refgcc_-march=native_-mtune=native_-O_-fomit-frame-pointer_-fwrapv_-fPIC_-fPIE2021011620210114
33130414097 98336 032510 99188 1744T:refgcc_-march=native_-mtune=native_-O2_-fomit-frame-pointer_-fwrapv_-fPIC_-fPIE2021011620210114
34041811978 98336 028606 99180 1744T:refgcc_-march=native_-mtune=native_-Os_-fomit-frame-pointer_-fwrapv_-fPIC_-fPIE2021011620210114
430117143100 0 0162962 820 1744T:protclang_-march=native_-O_-fomit-frame-pointer_-fwrapv_-Qunused-arguments_-fPIC_-fPIE2021011620210114
430154143100 0 0162962 820 1744T:protclang_-march=native_-O2_-fomit-frame-pointer_-fwrapv_-Qunused-arguments_-fPIC_-fPIE2021011620210114
433376145812 0 0167418 820 1744T:protclang_-march=native_-O3_-fomit-frame-pointer_-fwrapv_-Qunused-arguments_-fPIC_-fPIE2021011620210114
45860340328 98304 060215 99148 1744T:protgcc_-march=native_-mtune=native_-O3_-fomit-frame-pointer_-fwrapv_-fPIC_-fPIE2021011620210114
45992340366 196608 062602 197436 1744T:optclang_-march=native_-O3_-fomit-frame-pointer_-fwrapv_-Qunused-arguments_-fPIC_-fPIE2021011620210114
46010140366 196608 060938 197436 1744T:optclang_-march=native_-O2_-fomit-frame-pointer_-fwrapv_-Qunused-arguments_-fPIC_-fPIE2021011620210114
46011740366 196608 060938 197436 1744T:optclang_-march=native_-O_-fomit-frame-pointer_-fwrapv_-Qunused-arguments_-fPIC_-fPIE2021011620210114
466185120640 0 0137072 812 1712T:protclang_-march=native_-Os_-fomit-frame-pointer_-fwrapv_-Qunused-arguments_-fPIC_-fPIE2021011620210114
47541020111 196608 039503 197452 1744T:optgcc_-march=native_-mtune=native_-O3_-fomit-frame-pointer_-fwrapv_-fPIC_-fPIE2021011620210114
47758015539 98304 033895 99148 1744T:protgcc_-march=native_-mtune=native_-O2_-fomit-frame-pointer_-fwrapv_-fPIC_-fPIE2021011620210114
47784218759 196608 036687 197452 1744T:optgcc_-march=native_-mtune=native_-O2_-fomit-frame-pointer_-fwrapv_-fPIC_-fPIE2021011620210114
47865313005 98304 029639 99140 1744T:protgcc_-march=native_-mtune=native_-Os_-fomit-frame-pointer_-fwrapv_-fPIC_-fPIE2021011620210114
48040414186 98304 031927 99148 1744T:protgcc_-march=native_-mtune=native_-O_-fomit-frame-pointer_-fwrapv_-fPIC_-fPIE2021011620210114
48591921992 196608 038992 197428 1712T:optclang_-march=native_-Os_-fomit-frame-pointer_-fwrapv_-Qunused-arguments_-fPIC_-fPIE2021011620210114
50195218408 196608 036047 197452 1744T:optgcc_-march=native_-mtune=native_-O_-fomit-frame-pointer_-fwrapv_-fPIC_-fPIE2021011620210114
50304517003 196608 033535 197444 1744T:optgcc_-march=native_-mtune=native_-Os_-fomit-frame-pointer_-fwrapv_-fPIC_-fPIE2021011620210114

Test failure

Implementation: T:ref
Security model: timingleaks
Compiler: gcc -march=native -mtune=native -O3 -fomit-frame-pointer -fwrapv -fPIC -fPIE
error 111

Number of similar (compiler,implementation) pairs: 1, namely:
CompilerImplementations
gcc -march=native -mtune=native -O3 -fomit-frame-pointer -fwrapv -fPIC -fPIE T:ref

Compiler output

Implementation: T:opt
Security model: timingleaks
Compiler: clang -mcpu=native -O3 -fomit-frame-pointer -fwrapv -Qunused-arguments -fPIC -fPIE
dh.c: In file included from dh.c:10:
dh.c: ./ffa.h:94:5: error: '__builtin_ia32_pclmulqdq128' needs target feature pclmul
dh.c: low_kts_stp(op00,op10,k00,k02,op01,op11,r00,r01,r02,0x00); /* a0xb0 */
dh.c: ^
dh.c: ./ffa.h:62:11: note: expanded from macro 'low_kts_stp'
dh.c: re0 = _mm_clmulepi64_si128(op0, op1, ord);\
dh.c: ^
dh.c: /usr/lib/llvm-7/lib/clang/7.0.1/include/__wmmintrin_pclmul.h:59:13: note: expanded from macro '_mm_clmulepi64_si128'
dh.c: ((__m128i)__builtin_ia32_pclmulqdq128((__v2di)(__m128i)(X), \
dh.c: ^
dh.c: In file included from dh.c:10:
dh.c: ./ffa.h:94:5: error: '__builtin_ia32_pclmulqdq128' needs target feature pclmul
dh.c: ./ffa.h:63:11: note: expanded from macro 'low_kts_stp'
dh.c: re1 = _mm_clmulepi64_si128(op2, op3, ord);\
dh.c: ^
dh.c: /usr/lib/llvm-7/lib/clang/7.0.1/include/__wmmintrin_pclmul.h:59:13: note: expanded from macro '_mm_clmulepi64_si128'
dh.c: ((__m128i)__builtin_ia32_pclmulqdq128((__v2di)(__m128i)(X), \
dh.c: ^
dh.c: In file included from dh.c:10:
dh.c: ./ffa.h:94:5: error: '__builtin_ia32_pclmulqdq128' needs target feature pclmul
dh.c: ./ffa.h:64:11: note: expanded from macro 'low_kts_stp'
dh.c: re2 = _mm_clmulepi64_si128(op4, op5, ord);\
dh.c: ^
dh.c: /usr/lib/llvm-7/lib/clang/7.0.1/include/__wmmintrin_pclmul.h:59:13: note: expanded from macro '_mm_clmulepi64_si128'
dh.c: ((__m128i)__builtin_ia32_pclmulqdq128((__v2di)(__m128i)(X), \
dh.c: ...

Number of similar (compiler,implementation) pairs: 1, namely:
CompilerImplementations
clang -mcpu=native -O3 -fomit-frame-pointer -fwrapv -Qunused-arguments -fPIC -fPIE T:opt

Compiler output

Implementation: T:prot
Security model: timingleaks
Compiler: clang -march=native -O2 -fomit-frame-pointer -fwrapv -Qunused-arguments -fPIC -fPIE
dh.c: In file included from dh.c:9:
dh.c: ./scmul.incl:273:17: warning: implicit declaration of function 'types_cond_copy2' is invalid in C99 [-Wimplicit-function-declaration]
dh.c: types_cond_copy2(tab0, tab4, tabx, n1, n2);
dh.c: ^
dh.c: ./scmul.incl:284:9: warning: implicit declaration of function 'types_cond_copy' is invalid in C99 [-Wimplicit-function-declaration]
dh.c: types_cond_copy(QX, tab0, o1);
dh.c: ^
dh.c: 2 warnings generated.

Number of similar (compiler,implementation) pairs: 4, namely:
CompilerImplementations
clang -march=native -O2 -fomit-frame-pointer -fwrapv -Qunused-arguments -fPIC -fPIE T:prot
clang -march=native -O3 -fomit-frame-pointer -fwrapv -Qunused-arguments -fPIC -fPIE T:prot
clang -march=native -O -fomit-frame-pointer -fwrapv -Qunused-arguments -fPIC -fPIE T:prot
clang -march=native -Os -fomit-frame-pointer -fwrapv -Qunused-arguments -fPIC -fPIE T:prot

Compiler output

Implementation: T:prot
Security model: timingleaks
Compiler: clang -mcpu=native -O3 -fomit-frame-pointer -fwrapv -Qunused-arguments -fPIC -fPIE
dh.c: In file included from dh.c:7:
dh.c: ./low.incl:26:2: error: '__builtin_ia32_pclmulqdq128' needs target feature pclmul
dh.c: MUL4(m1, m0, ma, mb);
dh.c: ^
dh.c: ./gls254.h:155:7: note: expanded from macro 'MUL4'
dh.c: r0 = _mm_clmulepi64_si128(ma, mb, 0x00);\
dh.c: ^
dh.c: /usr/lib/llvm-7/lib/clang/7.0.1/include/__wmmintrin_pclmul.h:59:13: note: expanded from macro '_mm_clmulepi64_si128'
dh.c: ((__m128i)__builtin_ia32_pclmulqdq128((__v2di)(__m128i)(X), \
dh.c: ^
dh.c: In file included from dh.c:7:
dh.c: ./low.incl:26:2: error: '__builtin_ia32_pclmulqdq128' needs target feature pclmul
dh.c: ./gls254.h:156:7: note: expanded from macro 'MUL4'
dh.c: r1 = _mm_clmulepi64_si128(ma, mb, 0x11);\
dh.c: ^
dh.c: /usr/lib/llvm-7/lib/clang/7.0.1/include/__wmmintrin_pclmul.h:59:13: note: expanded from macro '_mm_clmulepi64_si128'
dh.c: ((__m128i)__builtin_ia32_pclmulqdq128((__v2di)(__m128i)(X), \
dh.c: ^
dh.c: In file included from dh.c:7:
dh.c: ./low.incl:26:2: error: '__builtin_ia32_pclmulqdq128' needs target feature pclmul
dh.c: ./gls254.h:157:7: note: expanded from macro 'MUL4'
dh.c: t0 = _mm_clmulepi64_si128(t0, t0, 0x10);\
dh.c: ^
dh.c: /usr/lib/llvm-7/lib/clang/7.0.1/include/__wmmintrin_pclmul.h:59:13: note: expanded from macro '_mm_clmulepi64_si128'
dh.c: ((__m128i)__builtin_ia32_pclmulqdq128((__v2di)(__m128i)(X), \
dh.c: ...

Number of similar (compiler,implementation) pairs: 1, namely:
CompilerImplementations
clang -mcpu=native -O3 -fomit-frame-pointer -fwrapv -Qunused-arguments -fPIC -fPIE T:prot

Compiler output

Implementation: T:prot
Security model: timingleaks
Compiler: gcc -march=native -mtune=native -O2 -fomit-frame-pointer -fwrapv -fPIC -fPIE
dh.c: In file included from dh.c:9:
dh.c: scmul.incl: In function 'scmul_rand_wreg':
dh.c: scmul.incl:273:17: warning: implicit declaration of function 'types_cond_copy2'; did you mean 'types_copy'? [-Wimplicit-function-declaration]
dh.c: types_cond_copy2(tab0, tab4, tabx, n1, n2);
dh.c: ^~~~~~~~~~~~~~~~
dh.c: types_copy
dh.c: scmul.incl:284:9: warning: implicit declaration of function 'types_cond_copy'; did you mean 'types_copy'? [-Wimplicit-function-declaration]
dh.c: types_cond_copy(QX, tab0, o1);
dh.c: ^~~~~~~~~~~~~~~
dh.c: types_copy

Number of similar (compiler,implementation) pairs: 4, namely:
CompilerImplementations
gcc -march=native -mtune=native -O2 -fomit-frame-pointer -fwrapv -fPIC -fPIE T:prot
gcc -march=native -mtune=native -O3 -fomit-frame-pointer -fwrapv -fPIC -fPIE T:prot
gcc -march=native -mtune=native -O -fomit-frame-pointer -fwrapv -fPIC -fPIE T:prot
gcc -march=native -mtune=native -Os -fomit-frame-pointer -fwrapv -fPIC -fPIE T:prot

Compiler output

Implementation: T:ref
Security model: timingleaks
Compiler: clang -march=native -O2 -fomit-frame-pointer -fwrapv -Qunused-arguments -fPIC -fPIE
dh.c: In file included from dh.c:9:
dh.c: ./scmul.incl:204:14: warning: & has lower precedence than ==; == will be evaluated first [-Wparentheses]
dh.c: if (num[0] & 0x1 == 1) {
dh.c: ^~~~~~~~~~
dh.c: ./scmul.incl:204:14: note: place parentheses around the '==' expression to silence this warning
dh.c: if (num[0] & 0x1 == 1) {
dh.c: ^
dh.c: ( )
dh.c: ./scmul.incl:204:14: note: place parentheses around the & expression to evaluate it first
dh.c: if (num[0] & 0x1 == 1) {
dh.c: ^
dh.c: ( )
dh.c: 1 warning generated.

Number of similar (compiler,implementation) pairs: 4, namely:
CompilerImplementations
clang -march=native -O2 -fomit-frame-pointer -fwrapv -Qunused-arguments -fPIC -fPIE T:ref
clang -march=native -O3 -fomit-frame-pointer -fwrapv -Qunused-arguments -fPIC -fPIE T:ref
clang -march=native -O -fomit-frame-pointer -fwrapv -Qunused-arguments -fPIC -fPIE T:ref
clang -march=native -Os -fomit-frame-pointer -fwrapv -Qunused-arguments -fPIC -fPIE T:ref

Compiler output

Implementation: T:ref
Security model: timingleaks
Compiler: clang -mcpu=native -O3 -fomit-frame-pointer -fwrapv -Qunused-arguments -fPIC -fPIE
dh.c: In file included from dh.c:6:
dh.c: ./curves.incl:53:2: error: '__builtin_ia32_aeskeygenassist128' needs target feature aes
dh.c: STEP(k, t, tmp, 16*1, 0x01);
dh.c: ^
dh.c: ./gls254.h:272:6: note: expanded from macro 'STEP'
dh.c: t = _mm_aeskeygenassist_si128(k, rcon);\
dh.c: ^
dh.c: /usr/lib/llvm-7/lib/clang/7.0.1/include/__wmmintrin_aes.h:150:12: note: expanded from macro '_mm_aeskeygenassist_si128'
dh.c: (__m128i)__builtin_ia32_aeskeygenassist128((__v2di)(__m128i)(C), (int)(R))
dh.c: ^
dh.c: In file included from dh.c:6:
dh.c: ./curves.incl:54:2: error: '__builtin_ia32_aeskeygenassist128' needs target feature aes
dh.c: STEP(k, t, tmp, 16*2, 0x02);
dh.c: ^
dh.c: ./gls254.h:272:6: note: expanded from macro 'STEP'
dh.c: t = _mm_aeskeygenassist_si128(k, rcon);\
dh.c: ^
dh.c: /usr/lib/llvm-7/lib/clang/7.0.1/include/__wmmintrin_aes.h:150:12: note: expanded from macro '_mm_aeskeygenassist_si128'
dh.c: (__m128i)__builtin_ia32_aeskeygenassist128((__v2di)(__m128i)(C), (int)(R))
dh.c: ^
dh.c: In file included from dh.c:6:
dh.c: ./curves.incl:55:2: error: '__builtin_ia32_aeskeygenassist128' needs target feature aes
dh.c: STEP(k, t, tmp, 16*3, 0x04);
dh.c: ^
dh.c: ./gls254.h:272:6: note: expanded from macro 'STEP'
dh.c: ...

Number of similar (compiler,implementation) pairs: 1, namely:
CompilerImplementations
clang -mcpu=native -O3 -fomit-frame-pointer -fwrapv -Qunused-arguments -fPIC -fPIE T:ref

Namespace violations

Implementation: T:opt
Security model: timingleaks
Compiler: clang -march=native -O2 -fomit-frame-pointer -fwrapv -Qunused-arguments -fPIC -fPIE
dh.o eca_add_add_dbl T
dh.o eca_add_dbl T
dh.o eca_add_ful T
dh.o eca_add_mix T
dh.o eca_add_mma T
dh.o eca_add_sub_dbl T
dh.o eca_dbl_ful T
dh.o low_inv T
dh.o low_inv_bas T
dh.o low_msq_015 T
dh.o low_msq_030 T
dh.o low_msq_063 T
dh.o low_mul T
dh.o low_mul_00u T
dh.o low_mul_01u T
dh.o low_mul_27u T
dh.o low_mul_bas T
dh.o low_mul_fq1 T
dh.o low_sqr T
dh.o low_sqr_003_bas T
dh.o low_sqr_006_bas T
dh.o low_sqr_bas T
dh.o mtx_msq_015_000 D
dh.o mtx_msq_015_001 D
dh.o mtx_msq_015_002 D
dh.o mtx_msq_015_003 D
dh.o mtx_msq_015_004 D
dh.o mtx_msq_015_005 D
dh.o mtx_msq_015_006 D
dh.o mtx_msq_015_007 D
dh.o mtx_msq_015_008 D
dh.o mtx_msq_015_009 D
dh.o mtx_msq_015_010 D
dh.o mtx_msq_015_011 D
dh.o mtx_msq_015_012 D
dh.o mtx_msq_015_013 D
dh.o mtx_msq_015_014 D
dh.o mtx_msq_015_015 D
dh.o mtx_msq_030_000 D
dh.o mtx_msq_030_001 D
dh.o mtx_msq_030_002 D
dh.o mtx_msq_030_003 D
dh.o mtx_msq_030_004 D
dh.o mtx_msq_030_005 D
dh.o mtx_msq_030_006 D
dh.o mtx_msq_030_007 D
dh.o mtx_msq_030_008 D
dh.o mtx_msq_030_009 D
dh.o mtx_msq_030_010 D
dh.o mtx_msq_030_011 D
dh.o mtx_msq_030_012 D
dh.o mtx_msq_030_013 D
dh.o mtx_msq_030_014 D
dh.o mtx_msq_030_015 D
dh.o mtx_msq_063_000 D
dh.o mtx_msq_063_001 D
dh.o mtx_msq_063_002 D
dh.o mtx_msq_063_003 D
dh.o mtx_msq_063_004 D
dh.o mtx_msq_063_005 D
dh.o mtx_msq_063_006 D
dh.o mtx_msq_063_007 D
dh.o mtx_msq_063_008 D
dh.o mtx_msq_063_009 D
dh.o mtx_msq_063_010 D
dh.o mtx_msq_063_011 D
dh.o mtx_msq_063_012 D
dh.o mtx_msq_063_013 D
dh.o mtx_msq_063_014 D
dh.o mtx_msq_063_015 D
dh.o smu_5nf_dna_ltr T
dh.o smu_pre_5nf T
dh.o smu_reg_rec T

Number of similar (compiler,implementation) pairs: 8, namely:
CompilerImplementations
clang -march=native -O2 -fomit-frame-pointer -fwrapv -Qunused-arguments -fPIC -fPIE T:opt
clang -march=native -O3 -fomit-frame-pointer -fwrapv -Qunused-arguments -fPIC -fPIE T:opt
clang -march=native -O -fomit-frame-pointer -fwrapv -Qunused-arguments -fPIC -fPIE T:opt
clang -march=native -Os -fomit-frame-pointer -fwrapv -Qunused-arguments -fPIC -fPIE T:opt
gcc -march=native -mtune=native -O2 -fomit-frame-pointer -fwrapv -fPIC -fPIE T:opt
gcc -march=native -mtune=native -O3 -fomit-frame-pointer -fwrapv -fPIC -fPIE T:opt
gcc -march=native -mtune=native -O -fomit-frame-pointer -fwrapv -fPIC -fPIE T:opt
gcc -march=native -mtune=native -Os -fomit-frame-pointer -fwrapv -fPIC -fPIE T:opt

Namespace violations

Implementation: T:prot
Security model: timingleaks
Compiler: clang -march=native -O2 -fomit-frame-pointer -fwrapv -Qunused-arguments -fPIC -fPIE
dh.o bn_rsh T
dh.o curves_domParam T
dh.o ec_add_mix T
dh.o ec_doub T
dh.o ec_doub_add T
dh.o ec_doub_add2 T
dh.o ec_doub_add_mix T
dh.o ec_doub_addsub T
dh.o low_htr T
dh.o low_htr254 T
dh.o low_inv T
dh.o low_inv254 T
dh.o low_inv254_const T
dh.o low_inv254_sim T
dh.o low_inv_const T
dh.o low_mul T
dh.o low_mul254 T
dh.o low_mul_a T
dh.o low_mul_a_2 T
dh.o low_mul_aplus1 T
dh.o low_mul_nr254 T
dh.o low_sq T
dh.o low_sq254 T
dh.o low_sq_nr254 T
dh.o low_sqi T
dh.o low_sqr06 T
dh.o low_sqr12 T
dh.o low_sqr24 T
dh.o low_sqr48 T
dh.o low_sqrt254 T
dh.o scmul_protected_direct_recoding T
dh.o scmul_rand_wreg T
dh.o scmul_wreg T
types.o types_cond_copy T
types.o types_cond_copy1 T
types.o types_cond_copy2 T
types.o types_cond_save T

Number of similar (compiler,implementation) pairs: 8, namely:
CompilerImplementations
clang -march=native -O2 -fomit-frame-pointer -fwrapv -Qunused-arguments -fPIC -fPIE T:prot
clang -march=native -O3 -fomit-frame-pointer -fwrapv -Qunused-arguments -fPIC -fPIE T:prot
clang -march=native -O -fomit-frame-pointer -fwrapv -Qunused-arguments -fPIC -fPIE T:prot
clang -march=native -Os -fomit-frame-pointer -fwrapv -Qunused-arguments -fPIC -fPIE T:prot
gcc -march=native -mtune=native -O2 -fomit-frame-pointer -fwrapv -fPIC -fPIE T:prot
gcc -march=native -mtune=native -O3 -fomit-frame-pointer -fwrapv -fPIC -fPIE T:prot
gcc -march=native -mtune=native -O -fomit-frame-pointer -fwrapv -fPIC -fPIE T:prot
gcc -march=native -mtune=native -Os -fomit-frame-pointer -fwrapv -fPIC -fPIE T:prot

Namespace violations

Implementation: T:ref
Security model: timingleaks
Compiler: clang -march=native -O2 -fomit-frame-pointer -fwrapv -Qunused-arguments -fPIC -fPIE
dh.o AESenc T
dh.o AESkeyExp T
dh.o curves_domParam T
dh.o ec_add_full_lambda T
dh.o ec_add_mix_lambda_opt T
dh.o ec_doub_lambda T
dh.o ec_halv_opt T
dh.o gen253 T
dh.o low_htr T
dh.o low_htr254 T
dh.o low_inv T
dh.o low_inv254 T
dh.o low_mul T
dh.o low_mul254 T
dh.o low_mul_a T
dh.o low_mul_a_2 T
dh.o low_mul_aplus1 T
dh.o low_mul_nr254 T
dh.o low_sq T
dh.o low_sq254 T
dh.o low_sq_nr254 T
dh.o low_sqi T
dh.o low_sqr06 T
dh.o low_sqr12 T
dh.o low_sqr24 T
dh.o low_sqr48 T
dh.o low_sqrt254 T
dh.o scmul_4wnafend T
dh.o scmul_cvhalv_end T
dh.o scmul_direct_coding_doub T
dh.o scmul_end_halv_direct T
dh.o tbl4NAF D

Number of similar (compiler,implementation) pairs: 7, namely:
CompilerImplementations
clang -march=native -O2 -fomit-frame-pointer -fwrapv -Qunused-arguments -fPIC -fPIE T:ref
clang -march=native -O3 -fomit-frame-pointer -fwrapv -Qunused-arguments -fPIC -fPIE T:ref
clang -march=native -O -fomit-frame-pointer -fwrapv -Qunused-arguments -fPIC -fPIE T:ref
clang -march=native -Os -fomit-frame-pointer -fwrapv -Qunused-arguments -fPIC -fPIE T:ref
gcc -march=native -mtune=native -O2 -fomit-frame-pointer -fwrapv -fPIC -fPIE T:ref
gcc -march=native -mtune=native -O -fomit-frame-pointer -fwrapv -fPIC -fPIE T:ref
gcc -march=native -mtune=native -Os -fomit-frame-pointer -fwrapv -fPIC -fPIE T:ref