Przemek:
I know but the problem seems to be inside OpenWatcom code and for me it
looks like OpenWatcom bug which probably can be quite easy fixed but
I need more information to be sure.
///
Thank you very much. So now we know that we can remove -s as
workaround. Unfortunately it causes very big speed overhead. Much
bigger then in Linux where removing -s reduce the speed only ~24
percent. It's 363% what is even bigger then in Windows (295% in my
test).
That increase in Windows is surprise for me, so problem are not in OS/2 only
Now I would like to check if -s is necessary as global switch or it
can be used locally only for GT2OS2 code.
Please restore -s switch (just like in clean SVN code) and then
add to src/rtl/gtos2/Makefile[5] this line:
HB_BUILD_OPTIM := no
It should disable optimization switches.
Then check if speedtst works in MT mode.
src\rtl\gtos2\Makefile does not have HB_BUILD_OPTIM
I used:
set HB_BUILD_OPTIM=no
Tests with
speedtst.exe --thread ...
fails as before with GPF:
-----------------
SYS1808:
The process has stopped. The software diagnostic
code (exception code) is 0001.
-----------------
Below are results using:
------------------------
hbmk2 -m -n -w -es2 -l -kmo -gc3 speedtst.prg
speedtst.exe
hbmk2 -m -n -w -es2 -l -kmo -gc3 -mt speedtst.prg
speedtst.exe
speedtst.exe --thread
speedtst.exe --thread=2 --scale
------------------------
David Macias
2009.11.28 08:22:22 OS/2 4.50
Harbour 2.0.0beta3 (Rev. 13035) Open Watcom C++ 12.80.8 (32-bit) x86
THREADS: 0
N_LOOPS: 1000000
[ T000: empty loop overhead ]...................................0.00
====================================================================
[ T001: x := L_C ]..............................................0.03
[ T002: x := L_N ]..............................................0.03
[ T003: x := L_D ]..............................................0.06
[ T004: x := S_C ]..............................................0.06
[ T005: x := S_N ]..............................................0.03
[ T006: x := S_D ]..............................................0.06
[ T007: x := M->M_C ]...........................................0.06
[ T008: x := M->M_N ]...........................................0.06
[ T009: x := M->M_D ]...........................................0.06
[ T010: x := M->P_C ]...........................................0.06
[ T011: x := M->P_N ]...........................................0.06
[ T012: x := M->P_D ]...........................................0.06
[ T013: x := F_C ]..............................................0.13
[ T014: x := F_N ]..............................................0.16
[ T015: x := F_D ]..............................................0.10
[ T016: x := o:Args ]...........................................0.13
[ T017: x := o[2] ].............................................0.10
[ T018: round( i / 1000, 2 ) ]..................................0.19
[ T019: str( i / 1000 ) ].......................................0.39
[ T020: val( s ) ]..............................................0.19
[ T021: val( a [ i % 16 + 1 ] ) ]...............................0.29
[ T022: dtos( d - i % 10000 ) ].................................0.32
[ T023: eval( { || i % 16 } ) ].................................0.26
[ T024: eval( bc := { || i % 16 } ) ]...........................0.16
[ T025: eval( { |x| x % 16 }, i ) ].............................0.23
[ T026: eval( bc := { |x| x % 16 }, i ) ].......................0.19
[ T027: eval( { |x| f1( x ) }, i ) ]............................0.26
[ T028: eval( bc := { |x| f1( x ) }, i ) ]......................0.23
[ T029: eval( bc := &("{ |x| f1( x ) }"), i ) ].................0.23
[ T030: x := &( "f1(" + str(i) + ")" ) ]........................1.87
[ T031: bc := &( "{|x|f1(x)}" ), eval( bc, i ) ]................2.00
[ T032: x := valtype( x ) + valtype( i ) ].....................0.32
[ T033: x := strzero( i % 100, 2 ) $ a[ i % 16 + 1 ] ]..........0.52
[ T034: x := a[ i % 16 + 1 ] == s ].............................0.19
[ T035: x := a[ i % 16 + 1 ] = s ]..............................0.23
[ T036: x := a[ i % 16 + 1 ] >= s ].............................0.23
[ T037: x := a[ i % 16 + 1 ] <= s ].............................0.23
[ T038: x := a[ i % 16 + 1 ] < s ]..............................0.23
[ T039: x := a[ i % 16 + 1 ] > s ]..............................0.23
[ T040: ascan( a, i % 16 ) ]....................................0.26
[ T041: ascan( a, { |x| x == i % 16 } ) ].......................1.87
[ T042: iif( i%1000==0, a:={}, ) , aadd(a,{i,1,.T.,s,s2,a2 ]....0.52
[ T043: x := a ]................................................0.06
[ T044: x := {} ]...............................................0.10
[ T045: f0() ]..................................................0.06
[ T046: f1( i ) ]...............................................0.13
[ T047: f2( c[1...8] ) ]........................................0.10
[ T048: f2( c[1...40000] ) ]....................................0.10
[ T049: f2( @c[1...40000] ) ]...................................0.10
[ T050: f2( @c[1...40000] ), c2 := c ]..........................0.10
[ T051: f3( a, a2, s, i, s2, bc, i, n, x ) ]....................0.29
[ T052: f2( a ) ]...............................................0.06
[ T053: x := f4() ].............................................0.61
[ T054: x := f5() ].............................................0.16
[ T055: x := space(16) ]........................................0.13
[ T056: f_prv( c ) ]............................................0.23
====================================================================
[ total application time: ]....................................15.35
[ total real time: ]...........................................15.35
2009.11.28 08:22:41 OS/2 4.50
Harbour 2.0.0beta3 (Rev. 13035) (MT) Open Watcom C++ 12.80.8 (32-bit) x86
THREADS: 0
N_LOOPS: 1000000
[ T000: empty loop overhead ]...................................0.03
====================================================================
[ T001: x := L_C ]..............................................0.06
[ T002: x := L_N ]..............................................0.03
[ T003: x := L_D ]..............................................0.03
[ T004: x := S_C ]..............................................0.06
[ T005: x := S_N ]..............................................0.03
[ T006: x := S_D ]..............................................0.03
[ T007: x := M->M_C ]...........................................0.06
[ T008: x := M->M_N ]...........................................0.03
[ T009: x := M->M_D ]...........................................0.03
[ T010: x := M->P_C ]...........................................0.10
[ T011: x := M->P_N ]...........................................0.03
[ T012: x := M->P_D ]...........................................0.03
[ T013: x := F_C ]..............................................0.19
[ T014: x := F_N ]..............................................0.13
[ T015: x := F_D ]..............................................0.10
[ T016: x := o:Args ]...........................................0.16
[ T017: x := o[2] ].............................................0.10
[ T018: round( i / 1000, 2 ) ]..................................0.19
[ T019: str( i / 1000 ) ].......................................0.45
[ T020: val( s ) ]..............................................0.23
[ T021: val( a [ i % 16 + 1 ] ) ]...............................0.35
[ T022: dtos( d - i % 10000 ) ].................................0.35
[ T023: eval( { || i % 16 } ) ].................................0.48
[ T024: eval( bc := { || i % 16 } ) ]...........................0.19
[ T025: eval( { |x| x % 16 }, i ) ].............................0.32
[ T026: eval( bc := { |x| x % 16 }, i ) ].......................0.23
[ T027: eval( { |x| f1( x ) }, i ) ]............................0.39
[ T028: eval( bc := { |x| f1( x ) }, i ) ]......................0.29
[ T029: eval( bc := &("{ |x| f1( x ) }"), i ) ].................0.29
[ T030: x := &( "f1(" + str(i) + ")" ) ]........................3.06
[ T031: bc := &( "{|x|f1(x)}" ), eval( bc, i ) ]................3.26
[ T032: x := valtype( x ) + valtype( i ) ].....................0.39
[ T033: x := strzero( i % 100, 2 ) $ a[ i % 16 + 1 ] ]..........0.61
[ T034: x := a[ i % 16 + 1 ] == s ].............................0.32
[ T035: x := a[ i % 16 + 1 ] = s ]..............................0.35
[ T036: x := a[ i % 16 + 1 ] >= s ].............................0.32
[ T037: x := a[ i % 16 + 1 ] <= s ].............................0.32
[ T038: x := a[ i % 16 + 1 ] < s ]..............................0.35
[ T039: x := a[ i % 16 + 1 ] > s ]..............................0.32
[ T040: ascan( a, i % 16 ) ]....................................0.29
[ T041: ascan( a, { |x| x == i % 16 } ) ].......................2.87
[ T042: iif( i%1000==0, a:={}, ) , aadd(a,{i,1,.T.,s,s2,a2 ]....0.94
[ T043: x := a ]................................................0.06
[ T044: x := {} ]...............................................0.16
[ T045: f0() ]..................................................0.06
[ T046: f1( i ) ]...............................................0.10
[ T047: f2( c[1...8] ) ]........................................0.13
[ T048: f2( c[1...40000] ) ]....................................0.13
[ T049: f2( @c[1...40000] ) ]...................................0.06
[ T050: f2( @c[1...40000] ), c2 := c ]..........................0.16
[ T051: f3( a, a2, s, i, s2, bc, i, n, x ) ]....................0.42
[ T052: f2( a ) ]...............................................0.13
[ T053: x := f4() ].............................................0.68
[ T054: x := f5() ].............................................0.23
[ T055: x := space(16) ]........................................0.19
[ T056: f_prv( c ) ]............................................0.48
====================================================================
[ total application time: ]....................................23.35
[ total real time: ]...........................................23.43
2009.11.28 08:23:05 OS/2 4.50
Harbour 2.0.0beta3 (Rev. 13035) (MT)+ Open Watcom C++ 12.80.8 (32-bit) x86
THREADS: all->56
N_LOOPS: 1000000
[ T000: empty loop overhead ]...................................0.03
====================================================================
2009.11.28 08:23:05 OS/2 4.50
Harbour 2.0.0beta3 (Rev. 13035) (MT)+ Open Watcom C++ 12.80.8 (32-bit) x86
THREADS: 2
N_LOOPS: 1000000
1 th. 2 th.
factor
============================================================================
_______________________________________________
Harbour mailing list (attachment size limit: 40KB)
Harbour@harbour-project.org
http://lists.harbour-project.org/mailman/listinfo/harbour
