Teste de velocidade Ming x VS x BCC

Mensagem por **Itamar M. Lins Jr.** » 03 Abr 2017 11:15

Ola!
Mingw melhor desempenho, mais ainda usando 64Bits.

2017-03-29 17:58:35 Windows 7 6.1.7601 Service Pack 1
M: Harbour 3.2.0dev (r1703291750) MinGW GNU C 6.3 (32-bit) x86
V: Harbour 3.2.0dev (r1703291750) Microsoft Visual C++ 19.0.24215 (32-bit) x86
B: Harbour 3.2.0dev (r1703291750) Borland/Embarcadero C++ 7.2 (32-bit) x86
THREADS: 0 M V B
N_LOOPS: 1000000 ---- ---- ----
[ T000: empty loop overhead ]...................................0.03 0.05 0.05
==================================================================== ===== =====
[ T001: x := L_C ]..............................................0.03 .0.02 .0.06
[ T002: x := L_N ]..............................................0.02 .0.00 .0.05
[ T003: x := L_D ]..............................................0.02 .0.00 .0.03
[ T004: x := S_C ]..............................................0.02 .0.00 .0.05
[ T005: x := S_N ]..............................................0.03 .0.02 .0.06
[ T006: x := S_D ]..............................................0.02 .0.02 .0.06
[ T007: x := M->M_C ]...........................................0.03 .0.02 .0.05
[ T008: x := M->M_N ]...........................................0.03 .0.00 .0.05
[ T009: x := M->M_D ]...........................................0.02 .0.02 .0.05
[ T010: x := M->P_C ]...........................................0.03 .0.02 .0.06
[ T011: x := M->P_N ]...........................................0.03 .0.03 .0.03
[ T012: x := M->P_D ]...........................................0.03 .0.02 .0.06
[ T013: x := F_C ]..............................................0.09 .0.08 .0.09
[ T014: x := F_N ]..............................................0.09 .0.09 .0.16
[ T015: x := F_D ]..............................................0.06 .0.05 .0.08
[ T016: x := o:Args ]...........................................0.08 .0.08 .0.14
[ T017: x := o[2] ].............................................0.06 .0.02 .0.11
[ T018: round( i / 1000, 2 ) ]..................................0.09 .0.11 .0.19
[ T019: str( i / 1000 ) ].......................................0.27 .0.25 .0.55
[ T020: val( s ) ]..............................................0.12 .0.12 .0.16
[ T021: val( a [ i % 16 + 1 ] ) ]...............................0.23 .0.23 .0.31
[ T022: dtos( d - i % 10000 ) ].................................0.25 .0.34 .0.34
[ T023: eval( {|| i % 16 } ) ]..................................0.17 .0.14 .0.22
[ T024: eval( bc := {|| i % 16 } ) ]............................0.09 .0.11 .0.16
[ T025: eval( {| x | x % 16 }, i ) ]............................0.19 .0.17 .0.23
[ T026: eval( bc := {| x | x % 16 }, i ) ]......................0.12 .0.12 .0.19
[ T027: eval( {| x | f1( x ) }, i ) ]...........................0.23 .0.19 .0.31
[ T028: eval( bc := {| x | f1( x ) }, i ) ].....................0.19 .0.17 .0.25
[ T029: eval( bc := &("{| x | f1( x ) }"), i ) ]................0.19 .0.20 .0.28
[ T030: x := &( "f1(" + str(i) + ")" ) ]........................1.34 .1.12 .1.76
[ T031: bc := &( "{| x |f1(x)}" ), eval( bc, i ) ]..............1.51 .1.47 .2.28
[ T032: x := valtype( x ) + valtype( i ) ].....................0.22 .0.19 .0.25
[ T033: x := strzero( i % 100, 2 ) $ a[ i % 16 + 1 ] ]..........0.34 .0.37 .0.61
[ T034: x := a[ i % 16 + 1 ] == s ].............................0.16 .0.17 .0.22
[ T035: x := a[ i % 16 + 1 ] = s ]..............................0.20 .0.19 .0.27
[ T036: x := a[ i % 16 + 1 ] >= s ].............................0.20 .0.17 .0.27
[ T037: x := a[ i % 16 + 1 ] <= s ].............................0.19 .0.17 .0.27
[ T038: x := a[ i % 16 + 1 ] < s ]..............................0.17 .0.17 .0.27
[ T039: x := a[ i % 16 + 1 ] > s ]..............................0.17 .0.19 .0.27
[ T040: ascan( a, i % 16 ) ]....................................0.14 .0.20 .0.20
[ T041: ascan( a, {| x | x == i % 16 } ) ]......................1.44 .1.36 .1.95
[ T042: iif( i%1000==0, a:={}, ) , aadd(a,{i,1,.T.,s,s2,a2 ]....0.44 .0.47 .0.64
[ T043: x := a ]................................................0.03 .0.02 .0.05
[ T044: x := {} ]...............................................0.05 .0.05 .0.08
[ T045: f0() ]..................................................0.03 .0.05 .0.06
[ T046: f1( i ) ]...............................................0.09 .0.11 .0.12
[ T047: f2( c[1...8] ) ]........................................0.09 .0.09 .0.09
[ T048: f2( c[1...40000] ) ]....................................0.06 .0.06 .0.11
[ T049: f2( @c[1...40000] ) ]...................................0.08 .0.09 .0.09
[ T050: f2( @c[1...40000] ), c2 := c ]..........................0.12 .0.11 .0.14
[ T051: f3( a, a2, s, i, s2, bc, i, n, x ) ]....................0.22 .0.20 .0.33
[ T052: f2( a ) ]...............................................0.05 .0.08 .0.09
[ T053: x := f4() ].............................................0.27 .0.31 .0.42
[ T054: x := f5() ].............................................0.12 .0.16 .0.22
[ T055: x := space(16) ]........................................0.06 .0.08 .0.12
[ T056: f_prv( c ) ]............................................0.20 .0.28 .0.37
==================================================================== ===== =====
[ total application time: ]....................................12.60 13.17 18.55
[ total real time: ]...........................................13.06 13.71 19.20

Com 64

Código: Selecionar todos

2017-04-03 07:49:09 Windows 7 6.1.7601 Service Pack 1
Harbour 3.2.0dev (r1703291750) MinGW GNU C 6.3 (64-bit) x86-64
====================================================================
2017-04-03 08:03:08 Windows 7 6.1.7601 Service Pack 1
Harbour 3.2.0dev (r1703291750) Microsoft Visual C 19.0.24215 (64-bit) x86-64
====================================================================
2017-04-03 07:48:15 Windows 7 6.1.7601 Service Pack 1
Harbour 3.2.0dev (r1703291750) LLVM/Clang C 3.3.1 (35759.1709ea1.58602a0) (64-bit) x86-64
====================================================================

THREADS: 0                                                      GCC   VC    BCC
N_LOOPS: 1000000                                               ===== ===== =====
[ T000: empty loop overhead ]...................................0.03  0.05  0.03
==================================================================== ===== =====
[ T001: x := L_C ]..............................................0.00  0.02  0.02
[ T002: x := L_N ]..............................................0.00  0.00  0.02
[ T003: x := L_D ]..............................................0.00  0.02  0.02
[ T004: x := S_C ]..............................................0.00  0.03  0.03
[ T005: x := S_N ]..............................................0.00  0.02  0.02
[ T006: x := S_D ]..............................................0.00  0.02  0.02
[ T007: x := M->M_C ]...........................................0.02  0.03  0.03
[ T008: x := M->M_N ]...........................................0.00  0.03  0.02
[ T009: x := M->M_D ]...........................................0.00  0.02  0.02
[ T010: x := M->P_C ]...........................................0.00  0.03  0.03
[ T011: x := M->P_N ]...........................................0.00  0.02  0.03
[ T012: x := M->P_D ]...........................................0.00  0.03  0.03
[ T013: x := F_C ]..............................................0.06  0.12  0.11
[ T014: x := F_N ]..............................................0.08  0.11  0.09
[ T015: x := F_D ]..............................................0.03  0.06  0.06
[ T016: x := o:Args ]...........................................0.06  0.12  0.09
[ T017: x := o[2] ].............................................0.02  0.06  0.06
[ T018: round( i / 1000, 2 ) ]..................................0.09  0.14  0.12
[ T019: str( i / 1000 ) ].......................................0.27  0.27  0.33
[ T020: val( s ) ]..............................................0.08  0.14  0.11
[ T021: val( a [ i % 16 + 1 ] ) ]...............................0.12  0.25  0.19
[ T022: dtos( d - i % 10000 ) ].................................0.14  0.23  0.19
[ T023: eval( {|| i % 16 } ) ]..................................0.14  0.28  0.19
[ T024: eval( bc := {|| i % 16 } ) ]............................0.09  0.17  0.14
[ T025: eval( {| x | x % 16 }, i ) ]............................0.12  0.22  0.19
[ T026: eval( bc := {| x | x % 16 }, i ) ]......................0.09  0.16  0.16
[ T027: eval( {| x | f1( x ) }, i ) ]...........................0.14  0.27  0.22
[ T028: eval( bc := {| x | f1( x ) }, i ) ].....................0.09  0.22  0.19
[ T029: eval( bc := &("{| x | f1( x ) }"), i ) ]................0.11  0.22  0.19
[ T030: x := &( "f1(" + str(i) + ")" ) ]........................1.00  1.62  1.53
[ T031: bc := &( "{| x |f1(x)}" ), eval( bc, i ) ]..............1.26  2.07  1.95
[ T032: x := valtype( x ) +  valtype( i ) ].....................0.12  0.25  0.22
[ T033: x := strzero( i % 100, 2 ) $ a[ i % 16 + 1 ] ]..........0.28  0.42  0.39
[ T034: x := a[ i % 16 + 1 ] == s ].............................0.09  0.19  0.14
[ T035: x := a[ i % 16 + 1 ] = s ]..............................0.09  0.22  0.16
[ T036: x := a[ i % 16 + 1 ] >= s ].............................0.09  0.20  0.16
[ T037: x := a[ i % 16 + 1 ] <= s ].............................0.11  0.20  0.14
[ T038: x := a[ i % 16 + 1 ] < s ]..............................0.09  0.20  0.16
[ T039: x := a[ i % 16 + 1 ] > s ]..............................0.09  0.20  0.16
[ T040: ascan( a, i % 16 ) ]....................................0.11  0.23  0.17
[ T041: ascan( a, {| x | x == i % 16 } ) ]......................1.12  2.00  1.53
[ T042: iif( i%1000==0, a:={}, ) , aadd(a,{i,1,.T.,s,s2,a2 ]....0.33  0.64  0.51
[ T043: x := a ]................................................0.00  0.03  0.03
[ T044: x := {} ]...............................................0.05  0.08  0.06
[ T045: f0() ]..................................................0.02  0.03  0.03
[ T046: f1( i ) ]...............................................0.03  0.09  0.06
[ T047: f2( c[1...8] ) ]........................................0.02  0.06  0.05
[ T048: f2( c[1...40000] ) ]....................................0.03  0.06  0.06
[ T049: f2( @c[1...40000] ) ]...................................0.02  0.06  0.05
[ T050: f2( @c[1...40000] ), c2 := c ]..........................0.03  0.11  0.08
[ T051: f3( a, a2, s, i, s2, bc, i, n, x ) ]....................0.12  0.41  0.28
[ T052: f2( a ) ]...............................................0.02  0.08  0.06
[ T053: x := f4() ].............................................0.28  0.27  0.50
[ T054: x := f5() ].............................................0.08  0.16  0.14
[ T055: x := space(16) ]........................................0.06  0.11  0.11
[ T056: f_prv( c ) ]............................................0.11  0.27  0.19
==================================================================== ===== =====
[ total application time: ].....................................9.19 16.19 13.54
[ total real time: ]............................................9.27 16.26 13.75

Saudações,
Itamar M. Lins Jr.

Mensagem por **JoséQuintas** » 10 Abr 2017 17:42

Independente disso, pode ser que acabemos usando o mingw ao invés do msvc.

Motivo: o run-time do Microsoft Visual C.

Como atualmente se misturam Windows XP e posteriores, o msvc pode acabar exigindo instalar run-time, o que vai acabar tornando mais prático distribuir com mingw mesmo.

Até saiu o MSVC 2017, e o Harbour teve ajustes pra ele, mas nem sei se funcionaria no XP....

Pra quem não sabe:
run-time seria algo como uma LIB para o executável, contendo partes principais de execução.
Sem isso, o programa não roda.
Para as llnguagens Microsoft já vém pré-instalada no Windows, mas lógico, uma determinada versão.
Versões diferentes, só instalando... se o Windows aceitar.