May 2012 - Pharo-dev - lists.pharo.org

Re: [Pharo-project] Pharo Kernel 2.0

by Edgar J. De Cleene

On 5/4/12 2:04 PM, "Pavel Krivanek" <pavel.krivanek(a)gmail.com> wrote: > Hi, > > MarkusÂ added all packages to the Pharo 2.0 repository (THANKS!) so I was able > to bring Pharo Kernel 2.0 image to life. Few notices: > > - I had to add FileSystem to the kernel. I added 'FileSystem-AnsiStreams' > 'FileSystem-Core' 'FileSystem-Disk' and 'FileSystem-Memory' but maybe some are > not necessary. > - CommandLine has poor packaging because it includes tests too. > -Â SimpleCodeLoader is now not needed and can be removed > - CommandLine package will need some some next changes like to be independent > on Compiler (because of Fuel) > -Â KMPragmaKeymapBuilder needs better cleanup code. > - I added 'Ring-Core-Kernel' Â 'Ring-Core-Containers' and 'Fuel' > 'FuelMetalevel' 'FuelPackageLoader' packages (Fuel from external repository) > > I will open the issues. > > Cheers, > -- Pavel Great news !!! Edgar

May 4, 2012

Re: [Pharo-project] [Vm-dev] Re: Do we have the new primitive?? [WAS] Re: IdentitySet but using #hash rather than #identityHash ?

by Eliot Miranda

On Thu, Apr 26, 2012 at 8:19 AM, Mariano Martinez Peck < marianopeck(a)gmail.com> wrote: > > > >>>> Hi Eliot/Levente. What is the status of this? Do we have already the >>>> new primitive? If true, how can we adapt LargeIdentitySet to use such new >>>> primitive? >>>> >>> >>> AFAIK the new primitive is not implemented yet. Adding the primitive to >>> the interpreter VM is very easy, but it seems to be a lot more complicated >>> (to me) to add it to Cog, because the receiver can be a MethodContext which >>> needs special handling. >>> I'll rewrite both LargeIdentitySet and LargeIdentityDictionary when the >>> primitive is ready. >>> >> >> Thanks Levente. So we should wait Eliot. >> I will ping again in a couple of weeks/months ;) >> >> > Ping. So I did it :) > Eliot if you tell us what it is needed maybe I can push Esteban or Igor > (or me?) to do it ;) > Some form of accurate spec. e.g. a simulation in Smalltalk, along with a specification of the types. I'm not happy about the receiver being a MethodContext because that means the primitive has to check argument types. A primitive can assume the type of the receiver because the primitive can be put on a specific class in the hierarchy. Hence aBehavior adoptInstance: anObject is much better than anObject changeClassTo: aClass, because the former can know that the receiver is a valid behavior, and simply check that the argument is a suitable instance of the receiver, whereas the latter has to check both that aClass is a valid behavior (walking its hierarchy? checking it has a valid methodDictionary, etc, etc) /and/ that the receiver is a suitable instance of the argument. So if possible specify it as one or two primitives on LargeIdentitySet & LargeIdentityDictionary. > > Thanks! > > > >> >> >>> >>> >>> Levente >>> >>> >>> >>>> Thanks! >>>> >>>> >>>> >>>> >>>> >>>> >>>> >>>> >>>> Levente >>>> >>>> >>>> thanks >>>> >>>> On Fri, Dec 16, 2011 at 3:28 PM, Levente Uzonyi < >>>> leves(a)elte.hu> wrote: >>>> >>>> On Fri, 16 Dec 2011, Henrik Sperre Johansen wrote: >>>> >>>> On 16.12.2011 03:26, Levente Uzonyi wrote: >>>> >>>> >>>> How about my numbers? :) >>>> >>>> "Preallocate objects, so we won't count gc time." >>>> n := 1000000. >>>> objects := Array new: n streamContents: [ :stream | >>>> n timesRepeat: [ stream nextPut: Object new ] ]. >>>> >>>> set := IdentitySet new: n. >>>> Smalltalk garbageCollect. >>>> [1 to: n do: [ :i | set add: (objects at: i) ] ] timeToRun. >>>> "4949" >>>> >>>> set := LargeIdentitySet new. >>>> Smalltalk garbageCollect. >>>> [1 to: n do: [ :i | set add: (objects at: i) ] ] timeToRun. >>>> "331" >>>> >>>> set := (PluggableSet new: n) >>>> hashBlock: [ :object | object identityHash * 4096 + object >>>> class >>>> identityHash * 64 ]; "Change this to #basicIdentityHash in >>>> Pharo" >>>> equalBlock: [ :a :b | a == b ]; >>>> yourself. >>>> Smalltalk garbageCollect. >>>> [1 to: n do: [ :i | set add: (objects at: i) ] ] timeToRun. >>>> "5511" >>>> >>>> >>>> I also have a LargeIdentityDictionary, which is relatively >>>> fast, but not >>>> as fast as LargeIdentitySet, because (for some unknown >>>> reason) we don't >>>> have a primitive that could support it. If we had a >>>> primitive like >>>> primitive 132 which would return the index of the element if >>>> found or 0 if >>>> not, then we could have a really fast >>>> LargeIdentityDictionary. >>>> >>>> >>>> Levente >>>> >>>> Hehe yes, if writing a version fully exploiting the limited range, >>>> that's >>>> probably the approach I would go for as well. >>>> (IAssuming it's the version at http://leves.web.elte.hu/** >>>> squeak/LargeIdentitySet.st<htt**p://leves.web.elte.hu/squeak/** >>>> LargeIdentitySet.st<http://leves.web.elte.hu/squeak/LargeIdentitySet.st> >>>> > >>>> ) >>>> >>>> Mariano commented in the version at http://www.squeaksource.com/** >>>> FuelExperiments <http://www.squeaksource.com/**FuelExperiments<http://www.squeaksource.com/FuelExperiments>> >>>> that it's >>>> slow for them, which I guess is due to not adopting #identityHash calls >>>> to >>>> #basicIdentityHash calls for Pharo: >>>> ((0 to: 4095) collect: [:each | each << 22 \\ 4096 ]) asSet size -> 1 >>>> So it basically uses 1 bucket instead of 4096... Whoops. :) >>>> >>>> Uploaded a new version to the MC repository which is adapted for Pharo, >>>> on the same machine my numbers were taken from, it does the same test >>>> as I >>>> used above in 871 ms. (181 with preallocation). >>>> >>>> >>>> Cool. One more thing: in Squeak the method using primitive 132 directly >>>> was renamed to #instVarsInclude:, so now #pointsTo: works as expected. >>>> If >>>> this was also added to Pharo, then the #pointsTo: sends should be >>>> changed >>>> to #instVarsInclude:, otherwise Array can be reported as included even >>>> if >>>> it wasn't added. >>>> I'll upload my LargeIdentityDictionary implementation to the same place >>>> this evening, since it's still 2-3 factor faster than other solutionts >>>> and >>>> there seem to be demand for it. >>>> >>>> >>>> Levente >>>> >>>> >>>> Cheers, >>>> Henry >>>> >>>> >>>> >>>> >>>> >>>> >>>> -- >>>> Mariano >>>> http://marianopeck.wordpress.**com <http://marianopeck.wordpress.com> >>>> >>>> >>>> >>>> >>>> >>>> -- >>>> Mariano >>>> http://marianopeck.wordpress.**com <http://marianopeck.wordpress.com> >>>> >>>> >>>> >> >> >> -- >> Mariano >> http://marianopeck.wordpress.com >> >> > > > -- > Mariano > http://marianopeck.wordpress.com > > > -- best, Eliot

May 4, 2012

[Pharo-project] better audio on meta-how-to video tutorial

by Lawson English

Here's a better version (better audio) of my "meta howto" video on how to deconstruct doing something in a class/package (Connectors) when the documentation isn't obvious: http://www.youtube.com/watch?v=xC6jcLw25Go&list=PL6601A198DF14788D&index=65… Lawson

May 4, 2012

Re: [Pharo-project] NativeBoost help

by Nicolas Cellier

I found this link quite detailed on the subject, but restricted to C++ (not entirely sure C compiler produces exactly the same...) http://www.shell-storm.org/papers/files/657.pdf See table 7 for the case of returning objects... See also stack cleanup by, 32 bit gcc use hybrid method for structure passed by values thru a pointer... Nicolas 2012/5/4 Igor Stasenko <siguctua(a)gmail.com>: > On 4 May 2012 19:18, Nicolas Cellier <nicolas.cellier.aka.nice(a)gmail.com> wrote: >> Yes, on max too >> https://developer.apple.com/library/mac/#documentation/developertools/Conce… >> >> Structures. The called function returns structures according to their >> aligned size. >> >> Â Â Structures 1 or 2 bytes in size are placed in EAX. >> >> Â Â Structures 4 or 8 bytes in size are placed in: EAX and EDX. >> > > okay.. i fixed it in Â NativeBoost-Core-IgorStasenko.58 > >> Â Â Structures of other sizes are placed at the address supplied by >> the caller. For example, the C++ language occasionally forces the >> compiler to return a value in memory when it would normally be >> returned in registers. See âPassing Argumentsâ for more information. >> >> I wonder if a structure of size 3 bytes can exist or is automatically >> extended on 4 bytes... >> > yes.. struct alignment is what i fear of.. > IMO it is actually silly to not return struct in registers if it less > than 8 bytes > but not power of 2.. > > sounds like an artefact of establishing the standards :) > >> $ cat > test_call.c <<END >> >> #include <stdio.h> >> typedef struct { char a,b,c; } abc; >> int main() { >> Â Â abc x; >> Â Â printf("sizeof(abc)=%d\n",sizeof x); >> Â Â return 0; >> } >> END >> >> $ gcc test_call.c >> $ ./a.exe >> sizeof(abc)=3 >> >> And what would be the calling conventions of cygwin gcc ? >> >> $ cat > test_call.c <<END >> >> #include <stdio.h> >> typedef struct { char a,b,c; } abc; >> abc sub() { >> Â abc result; >> Â result.a='a'; >> Â result.b='b'; >> Â result.c='c'; >> Â return result; >> } >> int main() { >> Â Â abc x=sub(); >> Â Â printf("sizeof(abc)=%d abc.a=$c\n",sizeof x,x.a); >> Â Â return 0; >> } >> END >> >> $ gcc test_call.c >> $ ./a.exe >> sizeof(abc)=3 abc.a=a >> >> $ gcc -S test_call.c >> $ cat test_call.S >> >> Â Â Â Â .file Â "test_call.c" >> Â Â Â Â .text >> .globl _sub >> Â Â Â Â .def Â Â _sub; Â .scl Â Â 2; Â Â Â .type Â 32; Â Â .endef >> _sub: >> Â Â Â Â pushl Â %ebp >> Â Â Â Â movl Â Â %esp, %ebp >> Â Â Â Â subl Â Â $16, %esp >> Â Â Â Â movl Â Â 8(%ebp), %eax >> Â Â Â Â movb Â Â $97, -3(%ebp) >> Â Â Â Â movb Â Â $98, -2(%ebp) >> Â Â Â Â movb Â Â $99, -1(%ebp) >> Â Â Â Â movzwl Â -3(%ebp), %edx >> Â Â Â Â movw Â Â %dx, (%eax) >> Â Â Â Â movzbl Â -1(%ebp), %edx >> Â Â Â Â movb Â Â %dl, 2(%eax) >> Â Â Â Â leave >> Â Â Â Â ret Â Â $4 >> Â Â Â Â .def Â Â ___main; Â Â Â Â .scl Â Â 2; Â Â Â .type Â 32; Â Â .endef >> Â Â Â Â .section .rdata,"dr" >> LC0: >> Â Â Â Â .ascii "sizeof(abc)=%d\12\0" >> Â Â Â Â .text >> .globl _main >> Â Â Â Â .def Â Â _main; Â .scl Â Â 2; Â Â Â .type Â 32; Â Â .endef >> _main: >> Â Â Â Â leal Â Â 4(%esp), %ecx >> Â Â Â Â andl Â Â $-16, %esp >> Â Â Â Â pushl Â -4(%ecx) >> Â Â Â Â pushl Â %ebp >> Â Â Â Â movl Â Â %esp, %ebp >> Â Â Â Â pushl Â %ecx >> Â Â Â Â subl Â Â $36, %esp >> Â Â Â Â call Â Â ___main >> Â Â Â Â leal Â Â -11(%ebp), %eax >> Â Â Â Â movl Â Â %eax, (%esp) >> Â Â Â Â call Â Â _sub >> Â Â Â Â subl Â Â $4, %esp >> Â Â Â Â movl Â Â $3, 4(%esp) >> Â Â Â Â movl Â Â $LC0, (%esp) >> Â Â Â Â call Â Â _printf >> Â Â Â Â movl Â Â $0, %eax >> Â Â Â Â movl Â Â -4(%ebp), %ecx >> Â Â Â Â leave >> Â Â Â Â leal Â Â -4(%ecx), %esp >> Â Â Â Â ret >> Â Â Â Â .def Â Â _printf; Â Â Â Â .scl Â Â 2; Â Â Â .type Â 32; Â Â .endef >> >> But of course, you already were faster than me... >> >> Nicolas >> >> 2012/5/4 Igor Stasenko <siguctua(a)gmail.com>: >>> On 4 May 2012 18:37, Eliot Miranda <eliot.miranda(a)gmail.com> wrote: >>>> >>>> >>>> On Fri, May 4, 2012 at 6:13 AM, Igor Stasenko <siguctua(a)gmail.com> wrote: >>>>> >>>>> On 4 May 2012 10:38, Luc Fabresse <luc.fabresse(a)gmail.com> wrote: >>>>> > Hi Igor, >>>>> > >>>>> >> >>>>> >> there is a workaround. >>>>> >> all structs in C is returned via pointer. >>>>> >> so, actually a true return type of a function is CvSize* >>>>> > >>>>> > >>>>> > I agree that usually structs are returned through pointers. >>>>> > But here it is not the case. >>>>> > >>>>> > I want to wrap this function: >>>>> > CvSize cvGetSize( const CvArr* arr ); >>>>> > >>>>> > and the CvSize struct is returned by value. Am I wrong? >>>>> > >>>>> >>>>> i don't know.. this is a bit shady part of C ABI. >>>>> Some compilers use return-by-value, some return by pointer >>>>> by forcing a caller to pass a hidden implicit argument as a pointer to >>>>> the struct, which then filled with values by >>>>> the function you call. >>>> >>>> >>>> IIRC x86 ABIs (all the ones I encountered at least, windows, mac, linux, >>>> solaris) Â support structure return (by-value structure return) by passing a >>>> hidden first argument which is a pointer to a structure into which to copy >>>> the result. Â The callee copies the result back through this pointer. >>>> >>>> Where these ABIs differ is in returning small structures. Â On win32 >>>> structures a power-of-two in size of 8 bytes or less are returned in >>>> EAX:EDX. Â IIRC on linux and mac all structures are returned through the >>>> hidden first argument pointer. Â See >>>> ThreadedIA32FFIPlugin>>returnStructInRegisters: >>>> >>> on macs (at least what gcc generates) it also returns structs <=8 >>> bytes size in registers: >>> >>> >>> so , currently i implemented it like that: >>> if sizeof(struct) <= 8 then i assume function returns it in EAX: EDX >>> otherwise via hidden pointer. >>> >>>>> >>>>> >>>>> It is easy to add support for that, but to make it work correctly can >>>>> be tricky. >>>>> See this, for example: >>>>> >>>>> http://stackoverflow.com/questions/161788/are-there-any-downsides-to-passin… >>>>> >>>>> >>>>> >> >>>>> >> >>>>> >> so, what you do is to change the return value to simple void* >>>>> >> and then copy the contents of struct into instance of corresponding >>>>> >> NBExternalStructure subclass >>>>> >> using #fromPointer: method.. to conveniently access its fields values. >>>>> > >>>>> > >>>>> > yes I saw that in NB tests. >>>>> > >>>>> >> >>>>> >> >>>>> >> >>>>> >> but this is in general, in your case this function is actually inlined >>>>> >> by compiler (as CV_INLINE suggests), so there could be even no >>>>> >> exported symbol for it in dynamic library. >>>>> >> and the function is actually so simple, that you don't need to wrap it >>>>> >> at >>>>> >> all. >>>>> >> you can just create a subclass of NBExternalStructure , named CvSize, >>>>> >> implement fieldDesc method as: >>>>> >> >>>>> >> fieldsDesc >>>>> >> ^ #( >>>>> >> int width; >>>>> >> int height >>>>> >> ) >>>>> >> >>>>> >> >>>>> >> and if you invoke: >>>>> >> >>>>> >> CvSize new width: 1 height:2 >>>>> >> >>>>> >> it will be equivalent as if you will call cvSize(1,2) function. >>>>> > >>>>> > >>>>> > yes it is exactly what I did. >>>>> > >>>>> > So my problem is: >>>>> > >>>>> > NBCvSize implemented as a subclass of NBExternalStruct with fieldDesc >>>>> > correctly set up. >>>>> > Then wrap the function cvGetSize: >>>>> > >>>>> > getSize >>>>> > Â Â Â self nbCall: #(NBCvSizeÂ cvGetSize(IplImage self)) module: LIB_CV. >>>>> > >>>>> > But I get a debugger because: >>>>> > >>>>> > NBExternalStruct>>coerceReturn: gen >>>>> > " ... should we support handling return of external structures? " >>>>> > self error: 'returning pointer to structure?' >>>>> > >>>>> > And I have no idea how to implement this (black magic EAX pop push ;-)) >>>>> > to >>>>> > coerce result of struct. >>>>> > >>>>> >>>>> >>>>> you don't. if function takes an implicit struct pointer, you can just >>>>> rewrite the function signature from >>>>> NBCvSize cvGetSize(IplImage self) >>>>> to: >>>>> >>>>> void cvGetSize(NBCvSize hidden, IplImage self) >>>>> >>>>> and so, you pass a fresh instance of NBCvSize as argument, which >>>>> function then fill. >>>>> But if C compiler using registers to return struct.. then you're screwed >>>>> :) >>>>> >>>>> Anyways, Luc i will try to read docs about C ABI and see if we can >>>>> easily add support for that. >>>>> >>>>> -- >>>>> Best regards, >>>>> Igor Stasenko. >>>>> >>>> >>>> >>>> >>>> -- >>>> best, >>>> Eliot >>>> >>> >>> >>> >>> -- >>> Best regards, >>> Igor Stasenko. >> > > > > -- > Best regards, > Igor Stasenko. >

May 4, 2012

Re: [Pharo-project] NativeBoost help

by Igor Stasenko

On 4 May 2012 19:18, Nicolas Cellier <nicolas.cellier.aka.nice(a)gmail.com> wrote: > Yes, on max too > https://developer.apple.com/library/mac/#documentation/developertools/Conce… > > Structures. The called function returns structures according to their > aligned size. > > Â Â Structures 1 or 2 bytes in size are placed in EAX. > > Â Â Structures 4 or 8 bytes in size are placed in: EAX and EDX. > okay.. i fixed it in NativeBoost-Core-IgorStasenko.58 > Â Â Structures of other sizes are placed at the address supplied by > the caller. For example, the C++ language occasionally forces the > compiler to return a value in memory when it would normally be > returned in registers. See âPassing Argumentsâ for more information. > > I wonder if a structure of size 3 bytes can exist or is automatically > extended on 4 bytes... > yes.. struct alignment is what i fear of.. IMO it is actually silly to not return struct in registers if it less than 8 bytes but not power of 2.. sounds like an artefact of establishing the standards :) > $ cat > test_call.c <<END > > #include <stdio.h> > typedef struct { char a,b,c; } abc; > int main() { > Â Â abc x; > Â Â printf("sizeof(abc)=%d\n",sizeof x); > Â Â return 0; > } > END > > $ gcc test_call.c > $ ./a.exe > sizeof(abc)=3 > > And what would be the calling conventions of cygwin gcc ? > > $ cat > test_call.c <<END > > #include <stdio.h> > typedef struct { char a,b,c; } abc; > abc sub() { > Â abc result; > Â result.a='a'; > Â result.b='b'; > Â result.c='c'; > Â return result; > } > int main() { > Â Â abc x=sub(); > Â Â printf("sizeof(abc)=%d abc.a=$c\n",sizeof x,x.a); > Â Â return 0; > } > END > > $ gcc test_call.c > $ ./a.exe > sizeof(abc)=3 abc.a=a > > $ gcc -S test_call.c > $ cat test_call.S > > Â Â Â Â .file Â "test_call.c" > Â Â Â Â .text > .globl _sub > Â Â Â Â .def Â Â _sub; Â .scl Â Â 2; Â Â Â .type Â 32; Â Â .endef > _sub: > Â Â Â Â pushl Â %ebp > Â Â Â Â movl Â Â %esp, %ebp > Â Â Â Â subl Â Â $16, %esp > Â Â Â Â movl Â Â 8(%ebp), %eax > Â Â Â Â movb Â Â $97, -3(%ebp) > Â Â Â Â movb Â Â $98, -2(%ebp) > Â Â Â Â movb Â Â $99, -1(%ebp) > Â Â Â Â movzwl Â -3(%ebp), %edx > Â Â Â Â movw Â Â %dx, (%eax) > Â Â Â Â movzbl Â -1(%ebp), %edx > Â Â Â Â movb Â Â %dl, 2(%eax) > Â Â Â Â leave > Â Â Â Â ret Â Â $4 > Â Â Â Â .def Â Â ___main; Â Â Â Â .scl Â Â 2; Â Â Â .type Â 32; Â Â .endef > Â Â Â Â .section .rdata,"dr" > LC0: > Â Â Â Â .ascii "sizeof(abc)=%d\12\0" > Â Â Â Â .text > .globl _main > Â Â Â Â .def Â Â _main; Â .scl Â Â 2; Â Â Â .type Â 32; Â Â .endef > _main: > Â Â Â Â leal Â Â 4(%esp), %ecx > Â Â Â Â andl Â Â $-16, %esp > Â Â Â Â pushl Â -4(%ecx) > Â Â Â Â pushl Â %ebp > Â Â Â Â movl Â Â %esp, %ebp > Â Â Â Â pushl Â %ecx > Â Â Â Â subl Â Â $36, %esp > Â Â Â Â call Â Â ___main > Â Â Â Â leal Â Â -11(%ebp), %eax > Â Â Â Â movl Â Â %eax, (%esp) > Â Â Â Â call Â Â _sub > Â Â Â Â subl Â Â $4, %esp > Â Â Â Â movl Â Â $3, 4(%esp) > Â Â Â Â movl Â Â $LC0, (%esp) > Â Â Â Â call Â Â _printf > Â Â Â Â movl Â Â $0, %eax > Â Â Â Â movl Â Â -4(%ebp), %ecx > Â Â Â Â leave > Â Â Â Â leal Â Â -4(%ecx), %esp > Â Â Â Â ret > Â Â Â Â .def Â Â _printf; Â Â Â Â .scl Â Â 2; Â Â Â .type Â 32; Â Â .endef > > But of course, you already were faster than me... > > Nicolas > > 2012/5/4 Igor Stasenko <siguctua(a)gmail.com>: >> On 4 May 2012 18:37, Eliot Miranda <eliot.miranda(a)gmail.com> wrote: >>> >>> >>> On Fri, May 4, 2012 at 6:13 AM, Igor Stasenko <siguctua(a)gmail.com> wrote: >>>> >>>> On 4 May 2012 10:38, Luc Fabresse <luc.fabresse(a)gmail.com> wrote: >>>> > Hi Igor, >>>> > >>>> >> >>>> >> there is a workaround. >>>> >> all structs in C is returned via pointer. >>>> >> so, actually a true return type of a function is CvSize* >>>> > >>>> > >>>> > I agree that usually structs are returned through pointers. >>>> > But here it is not the case. >>>> > >>>> > I want to wrap this function: >>>> > CvSize cvGetSize( const CvArr* arr ); >>>> > >>>> > and the CvSize struct is returned by value. Am I wrong? >>>> > >>>> >>>> i don't know.. this is a bit shady part of C ABI. >>>> Some compilers use return-by-value, some return by pointer >>>> by forcing a caller to pass a hidden implicit argument as a pointer to >>>> the struct, which then filled with values by >>>> the function you call. >>> >>> >>> IIRC x86 ABIs (all the ones I encountered at least, windows, mac, linux, >>> solaris) Â support structure return (by-value structure return) by passing a >>> hidden first argument which is a pointer to a structure into which to copy >>> the result. Â The callee copies the result back through this pointer. >>> >>> Where these ABIs differ is in returning small structures. Â On win32 >>> structures a power-of-two in size of 8 bytes or less are returned in >>> EAX:EDX. Â IIRC on linux and mac all structures are returned through the >>> hidden first argument pointer. Â See >>> ThreadedIA32FFIPlugin>>returnStructInRegisters: >>> >> on macs (at least what gcc generates) it also returns structs <=8 >> bytes size in registers: >> >> >> so , currently i implemented it like that: >> if sizeof(struct) <= 8 then i assume function returns it in EAX: EDX >> otherwise via hidden pointer. >> >>>> >>>> >>>> It is easy to add support for that, but to make it work correctly can >>>> be tricky. >>>> See this, for example: >>>> >>>> http://stackoverflow.com/questions/161788/are-there-any-downsides-to-passin… >>>> >>>> >>>> >> >>>> >> >>>> >> so, what you do is to change the return value to simple void* >>>> >> and then copy the contents of struct into instance of corresponding >>>> >> NBExternalStructure subclass >>>> >> using #fromPointer: method.. to conveniently access its fields values. >>>> > >>>> > >>>> > yes I saw that in NB tests. >>>> > >>>> >> >>>> >> >>>> >> >>>> >> but this is in general, in your case this function is actually inlined >>>> >> by compiler (as CV_INLINE suggests), so there could be even no >>>> >> exported symbol for it in dynamic library. >>>> >> and the function is actually so simple, that you don't need to wrap it >>>> >> at >>>> >> all. >>>> >> you can just create a subclass of NBExternalStructure , named CvSize, >>>> >> implement fieldDesc method as: >>>> >> >>>> >> fieldsDesc >>>> >> ^ #( >>>> >> int width; >>>> >> int height >>>> >> ) >>>> >> >>>> >> >>>> >> and if you invoke: >>>> >> >>>> >> CvSize new width: 1 height:2 >>>> >> >>>> >> it will be equivalent as if you will call cvSize(1,2) function. >>>> > >>>> > >>>> > yes it is exactly what I did. >>>> > >>>> > So my problem is: >>>> > >>>> > NBCvSize implemented as a subclass of NBExternalStruct with fieldDesc >>>> > correctly set up. >>>> > Then wrap the function cvGetSize: >>>> > >>>> > getSize >>>> > Â Â Â self nbCall: #(NBCvSizeÂ cvGetSize(IplImage self)) module: LIB_CV. >>>> > >>>> > But I get a debugger because: >>>> > >>>> > NBExternalStruct>>coerceReturn: gen >>>> > " ... should we support handling return of external structures? " >>>> > self error: 'returning pointer to structure?' >>>> > >>>> > And I have no idea how to implement this (black magic EAX pop push ;-)) >>>> > to >>>> > coerce result of struct. >>>> > >>>> >>>> >>>> you don't. if function takes an implicit struct pointer, you can just >>>> rewrite the function signature from >>>> NBCvSize cvGetSize(IplImage self) >>>> to: >>>> >>>> void cvGetSize(NBCvSize hidden, IplImage self) >>>> >>>> and so, you pass a fresh instance of NBCvSize as argument, which >>>> function then fill. >>>> But if C compiler using registers to return struct.. then you're screwed >>>> :) >>>> >>>> Anyways, Luc i will try to read docs about C ABI and see if we can >>>> easily add support for that. >>>> >>>> -- >>>> Best regards, >>>> Igor Stasenko. >>>> >>> >>> >>> >>> -- >>> best, >>> Eliot >>> >> >> >> >> -- >> Best regards, >> Igor Stasenko. > -- Best regards, Igor Stasenko.

May 4, 2012

Re: [Pharo-project] NativeBoost help

by Nicolas Cellier

2012/5/4 Nicolas Cellier <nicolas.cellier.aka.nice(a)gmail.com>: > Yes, on max too > https://developer.apple.com/library/mac/#documentation/developertools/Conce… > > Structures. The called function returns structures according to their > aligned size. > > Â Â Structures 1 or 2 bytes in size are placed in EAX. > > Â Â Structures 4 or 8 bytes in size are placed in: EAX and EDX. > > Â Â Structures of other sizes are placed at the address supplied by > the caller. For example, the C++ language occasionally forces the > compiler to return a value in memory when it would normally be > returned in registers. See âPassing Argumentsâ for more information. > > I wonder if a structure of size 3 bytes can exist or is automatically > extended on 4 bytes... > > $ cat > test_call.c <<END > > #include <stdio.h> > typedef struct { char a,b,c; } abc; > int main() { > Â Â abc x; > Â Â printf("sizeof(abc)=%d\n",sizeof x); > Â Â return 0; > } > END > > $ gcc test_call.c > $ ./a.exe > sizeof(abc)=3 > > And what would be the calling conventions of cygwin gcc ? > > $ cat > test_call.c <<END > > #include <stdio.h> > typedef struct { char a,b,c; } abc; > abc sub() { > Â abc result; > Â result.a='a'; > Â result.b='b'; > Â result.c='c'; > Â return result; > } > int main() { > Â Â abc x=sub(); > Â Â printf("sizeof(abc)=%d abc.a=$c\n",sizeof x,x.a); > Â Â return 0; > } > END > > $ gcc test_call.c > $ ./a.exe > sizeof(abc)=3 abc.a=a > > $ gcc -S test_call.c > $ cat test_call.S > > Â Â Â Â .file Â "test_call.c" > Â Â Â Â .text > .globl _sub > Â Â Â Â .def Â Â _sub; Â .scl Â Â 2; Â Â Â .type Â 32; Â Â .endef > _sub: > Â Â Â Â pushl Â %ebp > Â Â Â Â movl Â Â %esp, %ebp > Â Â Â Â subl Â Â $16, %esp > Â Â Â Â movl Â Â 8(%ebp), %eax > Â Â Â Â movb Â Â $97, -3(%ebp) > Â Â Â Â movb Â Â $98, -2(%ebp) > Â Â Â Â movb Â Â $99, -1(%ebp) > Â Â Â Â movzwl Â -3(%ebp), %edx > Â Â Â Â movw Â Â %dx, (%eax) > Â Â Â Â movzbl Â -1(%ebp), %edx > Â Â Â Â movb Â Â %dl, 2(%eax) > Â Â Â Â leave > Â Â Â Â ret Â Â $4 > Â Â Â Â .def Â Â ___main; Â Â Â Â .scl Â Â 2; Â Â Â .type Â 32; Â Â .endef > Â Â Â Â .section .rdata,"dr" > LC0: > Â Â Â Â .ascii "sizeof(abc)=%d\12\0" > Â Â Â Â .text > .globl _main > Â Â Â Â .def Â Â _main; Â .scl Â Â 2; Â Â Â .type Â 32; Â Â .endef > _main: > Â Â Â Â leal Â Â 4(%esp), %ecx > Â Â Â Â andl Â Â $-16, %esp > Â Â Â Â pushl Â -4(%ecx) > Â Â Â Â pushl Â %ebp > Â Â Â Â movl Â Â %esp, %ebp > Â Â Â Â pushl Â %ecx > Â Â Â Â subl Â Â $36, %esp > Â Â Â Â call Â Â ___main > Â Â Â Â leal Â Â -11(%ebp), %eax > Â Â Â Â movl Â Â %eax, (%esp) > Â Â Â Â call Â Â _sub > Â Â Â Â subl Â Â $4, %esp > Â Â Â Â movl Â Â $3, 4(%esp) > Â Â Â Â movl Â Â $LC0, (%esp) > Â Â Â Â call Â Â _printf > Â Â Â Â movl Â Â $0, %eax > Â Â Â Â movl Â Â -4(%ebp), %ecx > Â Â Â Â leave > Â Â Â Â leal Â Â -4(%ecx), %esp > Â Â Â Â ret > Â Â Â Â .def Â Â _printf; Â Â Â Â .scl Â Â 2; Â Â Â .type Â 32; Â Â .endef > > But of course, you already were faster than me... > > Nicolas > Argh, no, copy paste error... $ cat test_call.S .file "test_call.c" .text .globl _sub .def _sub; .scl 2; .type 32; .endef _sub: pushl %ebp movl %esp, %ebp subl $16, %esp movl 8(%ebp), %eax movb $97, -3(%ebp) movb $98, -2(%ebp) movb $99, -1(%ebp) movzwl -3(%ebp), %edx movw %dx, (%eax) movzbl -1(%ebp), %edx movb %dl, 2(%eax) leave ret $4 .def ___main; .scl 2; .type 32; .endef .section .rdata,"dr" LC0: .ascii "sizeof(abc)=%d abc.a=%c\12\0" .text .globl _main .def _main; .scl 2; .type 32; .endef _main: leal 4(%esp), %ecx andl $-16, %esp pushl -4(%ecx) pushl %ebp movl %esp, %ebp pushl %ecx subl $36, %esp call ___main leal -11(%ebp), %eax movl %eax, (%esp) call _sub subl $4, %esp movzbl -11(%ebp), %eax movsbl %al, %eax movl %eax, 8(%esp) movl $3, 4(%esp) movl $LC0, (%esp) call _printf movl $0, %eax movl -4(%ebp), %ecx leave leal -4(%ecx), %esp ret .def _printf; .scl 2; .type 32; .endef > 2012/5/4 Igor Stasenko <siguctua(a)gmail.com>: >> On 4 May 2012 18:37, Eliot Miranda <eliot.miranda(a)gmail.com> wrote: >>> >>> >>> On Fri, May 4, 2012 at 6:13 AM, Igor Stasenko <siguctua(a)gmail.com> wrote: >>>> >>>> On 4 May 2012 10:38, Luc Fabresse <luc.fabresse(a)gmail.com> wrote: >>>> > Hi Igor, >>>> > >>>> >> >>>> >> there is a workaround. >>>> >> all structs in C is returned via pointer. >>>> >> so, actually a true return type of a function is CvSize* >>>> > >>>> > >>>> > I agree that usually structs are returned through pointers. >>>> > But here it is not the case. >>>> > >>>> > I want to wrap this function: >>>> > CvSize cvGetSize( const CvArr* arr ); >>>> > >>>> > and the CvSize struct is returned by value. Am I wrong? >>>> > >>>> >>>> i don't know.. this is a bit shady part of C ABI. >>>> Some compilers use return-by-value, some return by pointer >>>> by forcing a caller to pass a hidden implicit argument as a pointer to >>>> the struct, which then filled with values by >>>> the function you call. >>> >>> >>> IIRC x86 ABIs (all the ones I encountered at least, windows, mac, linux, >>> solaris) Â support structure return (by-value structure return) by passing a >>> hidden first argument which is a pointer to a structure into which to copy >>> the result. Â The callee copies the result back through this pointer. >>> >>> Where these ABIs differ is in returning small structures. Â On win32 >>> structures a power-of-two in size of 8 bytes or less are returned in >>> EAX:EDX. Â IIRC on linux and mac all structures are returned through the >>> hidden first argument pointer. Â See >>> ThreadedIA32FFIPlugin>>returnStructInRegisters: >>> >> on macs (at least what gcc generates) it also returns structs <=8 >> bytes size in registers: >> >> >> so , currently i implemented it like that: >> if sizeof(struct) <= 8 then i assume function returns it in EAX: EDX >> otherwise via hidden pointer. >> >>>> >>>> >>>> It is easy to add support for that, but to make it work correctly can >>>> be tricky. >>>> See this, for example: >>>> >>>> http://stackoverflow.com/questions/161788/are-there-any-downsides-to-passin… >>>> >>>> >>>> >> >>>> >> >>>> >> so, what you do is to change the return value to simple void* >>>> >> and then copy the contents of struct into instance of corresponding >>>> >> NBExternalStructure subclass >>>> >> using #fromPointer: method.. to conveniently access its fields values. >>>> > >>>> > >>>> > yes I saw that in NB tests. >>>> > >>>> >> >>>> >> >>>> >> >>>> >> but this is in general, in your case this function is actually inlined >>>> >> by compiler (as CV_INLINE suggests), so there could be even no >>>> >> exported symbol for it in dynamic library. >>>> >> and the function is actually so simple, that you don't need to wrap it >>>> >> at >>>> >> all. >>>> >> you can just create a subclass of NBExternalStructure , named CvSize, >>>> >> implement fieldDesc method as: >>>> >> >>>> >> fieldsDesc >>>> >> ^ #( >>>> >> int width; >>>> >> int height >>>> >> ) >>>> >> >>>> >> >>>> >> and if you invoke: >>>> >> >>>> >> CvSize new width: 1 height:2 >>>> >> >>>> >> it will be equivalent as if you will call cvSize(1,2) function. >>>> > >>>> > >>>> > yes it is exactly what I did. >>>> > >>>> > So my problem is: >>>> > >>>> > NBCvSize implemented as a subclass of NBExternalStruct with fieldDesc >>>> > correctly set up. >>>> > Then wrap the function cvGetSize: >>>> > >>>> > getSize >>>> > Â Â Â self nbCall: #(NBCvSizeÂ cvGetSize(IplImage self)) module: LIB_CV. >>>> > >>>> > But I get a debugger because: >>>> > >>>> > NBExternalStruct>>coerceReturn: gen >>>> > " ... should we support handling return of external structures? " >>>> > self error: 'returning pointer to structure?' >>>> > >>>> > And I have no idea how to implement this (black magic EAX pop push ;-)) >>>> > to >>>> > coerce result of struct. >>>> > >>>> >>>> >>>> you don't. if function takes an implicit struct pointer, you can just >>>> rewrite the function signature from >>>> NBCvSize cvGetSize(IplImage self) >>>> to: >>>> >>>> void cvGetSize(NBCvSize hidden, IplImage self) >>>> >>>> and so, you pass a fresh instance of NBCvSize as argument, which >>>> function then fill. >>>> But if C compiler using registers to return struct.. then you're screwed >>>> :) >>>> >>>> Anyways, Luc i will try to read docs about C ABI and see if we can >>>> easily add support for that. >>>> >>>> -- >>>> Best regards, >>>> Igor Stasenko. >>>> >>> >>> >>> >>> -- >>> best, >>> Eliot >>> >> >> >> >> -- >> Best regards, >> Igor Stasenko.

May 4, 2012

Re: [Pharo-project] NativeBoost help

by Nicolas Cellier

Yes, on max too https://developer.apple.com/library/mac/#documentation/developertools/Conce… Structures. The called function returns structures according to their aligned size. Structures 1 or 2 bytes in size are placed in EAX. Structures 4 or 8 bytes in size are placed in: EAX and EDX. Structures of other sizes are placed at the address supplied by the caller. For example, the C++ language occasionally forces the compiler to return a value in memory when it would normally be returned in registers. See âPassing Argumentsâ for more information. I wonder if a structure of size 3 bytes can exist or is automatically extended on 4 bytes... $ cat > test_call.c <<END #include <stdio.h> typedef struct { char a,b,c; } abc; int main() { abc x; printf("sizeof(abc)=%d\n",sizeof x); return 0; } END $ gcc test_call.c $ ./a.exe sizeof(abc)=3 And what would be the calling conventions of cygwin gcc ? $ cat > test_call.c <<END #include <stdio.h> typedef struct { char a,b,c; } abc; abc sub() { abc result; result.a='a'; result.b='b'; result.c='c'; return result; } int main() { abc x=sub(); printf("sizeof(abc)=%d abc.a=$c\n",sizeof x,x.a); return 0; } END $ gcc test_call.c $ ./a.exe sizeof(abc)=3 abc.a=a $ gcc -S test_call.c $ cat test_call.S .file "test_call.c" .text .globl _sub .def _sub; .scl 2; .type 32; .endef _sub: pushl %ebp movl %esp, %ebp subl $16, %esp movl 8(%ebp), %eax movb $97, -3(%ebp) movb $98, -2(%ebp) movb $99, -1(%ebp) movzwl -3(%ebp), %edx movw %dx, (%eax) movzbl -1(%ebp), %edx movb %dl, 2(%eax) leave ret $4 .def ___main; .scl 2; .type 32; .endef .section .rdata,"dr" LC0: .ascii "sizeof(abc)=%d\12\0" .text .globl _main .def _main; .scl 2; .type 32; .endef _main: leal 4(%esp), %ecx andl $-16, %esp pushl -4(%ecx) pushl %ebp movl %esp, %ebp pushl %ecx subl $36, %esp call ___main leal -11(%ebp), %eax movl %eax, (%esp) call _sub subl $4, %esp movl $3, 4(%esp) movl $LC0, (%esp) call _printf movl $0, %eax movl -4(%ebp), %ecx leave leal -4(%ecx), %esp ret .def _printf; .scl 2; .type 32; .endef But of course, you already were faster than me... Nicolas 2012/5/4 Igor Stasenko <siguctua(a)gmail.com>: > On 4 May 2012 18:37, Eliot Miranda <eliot.miranda(a)gmail.com> wrote: >> >> >> On Fri, May 4, 2012 at 6:13 AM, Igor Stasenko <siguctua(a)gmail.com> wrote: >>> >>> On 4 May 2012 10:38, Luc Fabresse <luc.fabresse(a)gmail.com> wrote: >>> > Hi Igor, >>> > >>> >> >>> >> there is a workaround. >>> >> all structs in C is returned via pointer. >>> >> so, actually a true return type of a function is CvSize* >>> > >>> > >>> > I agree that usually structs are returned through pointers. >>> > But here it is not the case. >>> > >>> > I want to wrap this function: >>> > CvSize cvGetSize( const CvArr* arr ); >>> > >>> > and the CvSize struct is returned by value. Am I wrong? >>> > >>> >>> i don't know.. this is a bit shady part of C ABI. >>> Some compilers use return-by-value, some return by pointer >>> by forcing a caller to pass a hidden implicit argument as a pointer to >>> the struct, which then filled with values by >>> the function you call. >> >> >> IIRC x86 ABIs (all the ones I encountered at least, windows, mac, linux, >> solaris) Â support structure return (by-value structure return) by passing a >> hidden first argument which is a pointer to a structure into which to copy >> the result. Â The callee copies the result back through this pointer. >> >> Where these ABIs differ is in returning small structures. Â On win32 >> structures a power-of-two in size of 8 bytes or less are returned in >> EAX:EDX. Â IIRC on linux and mac all structures are returned through the >> hidden first argument pointer. Â See >> ThreadedIA32FFIPlugin>>returnStructInRegisters: >> > on macs (at least what gcc generates) it also returns structs <=8 > bytes size in registers: > > > so , currently i implemented it like that: > if sizeof(struct) <= 8 then i assume function returns it in EAX: EDX > otherwise via hidden pointer. > >>> >>> >>> It is easy to add support for that, but to make it work correctly can >>> be tricky. >>> See this, for example: >>> >>> http://stackoverflow.com/questions/161788/are-there-any-downsides-to-passin… >>> >>> >>> >> >>> >> >>> >> so, what you do is to change the return value to simple void* >>> >> and then copy the contents of struct into instance of corresponding >>> >> NBExternalStructure subclass >>> >> using #fromPointer: method.. to conveniently access its fields values. >>> > >>> > >>> > yes I saw that in NB tests. >>> > >>> >> >>> >> >>> >> >>> >> but this is in general, in your case this function is actually inlined >>> >> by compiler (as CV_INLINE suggests), so there could be even no >>> >> exported symbol for it in dynamic library. >>> >> and the function is actually so simple, that you don't need to wrap it >>> >> at >>> >> all. >>> >> you can just create a subclass of NBExternalStructure , named CvSize, >>> >> implement fieldDesc method as: >>> >> >>> >> fieldsDesc >>> >> ^ #( >>> >> int width; >>> >> int height >>> >> ) >>> >> >>> >> >>> >> and if you invoke: >>> >> >>> >> CvSize new width: 1 height:2 >>> >> >>> >> it will be equivalent as if you will call cvSize(1,2) function. >>> > >>> > >>> > yes it is exactly what I did. >>> > >>> > So my problem is: >>> > >>> > NBCvSize implemented as a subclass of NBExternalStruct with fieldDesc >>> > correctly set up. >>> > Then wrap the function cvGetSize: >>> > >>> > getSize >>> > Â Â Â self nbCall: #(NBCvSizeÂ cvGetSize(IplImage self)) module: LIB_CV. >>> > >>> > But I get a debugger because: >>> > >>> > NBExternalStruct>>coerceReturn: gen >>> > " ... should we support handling return of external structures? " >>> > self error: 'returning pointer to structure?' >>> > >>> > And I have no idea how to implement this (black magic EAX pop push ;-)) >>> > to >>> > coerce result of struct. >>> > >>> >>> >>> you don't. if function takes an implicit struct pointer, you can just >>> rewrite the function signature from >>> NBCvSize cvGetSize(IplImage self) >>> to: >>> >>> void cvGetSize(NBCvSize hidden, IplImage self) >>> >>> and so, you pass a fresh instance of NBCvSize as argument, which >>> function then fill. >>> But if C compiler using registers to return struct.. then you're screwed >>> :) >>> >>> Anyways, Luc i will try to read docs about C ABI and see if we can >>> easily add support for that. >>> >>> -- >>> Best regards, >>> Igor Stasenko. >>> >> >> >> >> -- >> best, >> Eliot >> > > > > -- > Best regards, > Igor Stasenko.

May 4, 2012

Re: [Pharo-project] NativeBoost help

by Igor Stasenko

On 4 May 2012 18:52, Eliot Miranda <eliot.miranda(a)gmail.com> wrote: > > > On Fri, May 4, 2012 at 9:44 AM, Igor Stasenko <siguctua(a)gmail.com> wrote: >> >> On 4 May 2012 18:37, Eliot Miranda <eliot.miranda(a)gmail.com> wrote: >> > >> > >> > On Fri, May 4, 2012 at 6:13 AM, Igor Stasenko <siguctua(a)gmail.com> >> > wrote: >> >> >> >> On 4 May 2012 10:38, Luc Fabresse <luc.fabresse(a)gmail.com> wrote: >> >> > Hi Igor, >> >> > >> >> >> >> >> >> there is a workaround. >> >> >> all structs in C is returned via pointer. >> >> >> so, actually a true return type of a function is CvSize* >> >> > >> >> > >> >> > I agree that usually structs are returned through pointers. >> >> > But here it is not the case. >> >> > >> >> > I want to wrap this function: >> >> > CvSize cvGetSize( const CvArr* arr ); >> >> > >> >> > and the CvSize struct is returned by value. Am I wrong? >> >> > >> >> >> >> i don't know.. this is a bit shady part of C ABI. >> >> Some compilers use return-by-value, some return by pointer >> >> by forcing a caller to pass a hidden implicit argument as a pointer to >> >> the struct, which then filled with values by >> >> the function you call. >> > >> > >> > IIRC x86 ABIs (all the ones I encountered at least, windows, mac, linux, >> > solaris) Â support structure return (by-value structure return) by >> > passing a >> > hidden first argument which is a pointer to a structure into which to >> > copy >> > the result. Â The callee copies the result back through this pointer. >> > >> > Where these ABIs differ is in returning small structures. Â On win32 >> > structures a power-of-two in size of 8 bytes or less are returned in >> > EAX:EDX. Â IIRC on linux and mac all structures are returned through the >> > hidden first argument pointer. Â See >> > ThreadedIA32FFIPlugin>>returnStructInRegisters: >> > >> on macs (at least what gcc generates) it also returns structs <=8 >> bytes size in registers: >> >> >> so , currently i implemented it like that: >> if sizeof(struct) <= 8 then i assume function returns it in EAX: EDX >> otherwise via hidden pointer. > > > add a test for a structs like > Â Â typedef struct { char a;Â charÂ b; char c; } threeBytes; > Â Â typedef struct { short a; short b; short c; } sixBytes; > > (and 5 & 7). Â You should see on Mac and Windows these are not passed back in > registers. > omg.. what a mess.. so it actually uses for structs with 1, 2, 4 and 8 bytes size... >> >> >> >> >> >> >> It is easy to add support for that, but to make it work correctly can >> >> be tricky. >> >> See this, for example: >> >> >> >> >> >> http://stackoverflow.com/questions/161788/are-there-any-downsides-to-passin… >> >> >> >> >> >> >> >> >> >> >> >> >> so, what you do is to change the return value to simple void* >> >> >> and then copy the contents of struct into instance of corresponding >> >> >> NBExternalStructure subclass >> >> >> using #fromPointer: method.. to conveniently access its fields >> >> >> values. >> >> > >> >> > >> >> > yes I saw that in NB tests. >> >> > >> >> >> >> >> >> >> >> >> >> >> >> but this is in general, in your case this function is actually >> >> >> inlined >> >> >> by compiler (as CV_INLINE suggests), so there could be even no >> >> >> exported symbol for it in dynamic library. >> >> >> and the function is actually so simple, that you don't need to wrap >> >> >> it >> >> >> at >> >> >> all. >> >> >> you can just create a subclass of NBExternalStructure , named >> >> >> CvSize, >> >> >> implement fieldDesc method as: >> >> >> >> >> >> fieldsDesc >> >> >> ^ #( >> >> >> int width; >> >> >> int height >> >> >> ) >> >> >> >> >> >> >> >> >> and if you invoke: >> >> >> >> >> >> CvSize new width: 1 height:2 >> >> >> >> >> >> it will be equivalent as if you will call cvSize(1,2) function. >> >> > >> >> > >> >> > yes it is exactly what I did. >> >> > >> >> > So my problem is: >> >> > >> >> > NBCvSize implemented as a subclass of NBExternalStruct with fieldDesc >> >> > correctly set up. >> >> > Then wrap the function cvGetSize: >> >> > >> >> > getSize >> >> > Â Â Â self nbCall: #(NBCvSizeÂ cvGetSize(IplImage self)) module: >> >> > LIB_CV. >> >> > >> >> > But I get a debugger because: >> >> > >> >> > NBExternalStruct>>coerceReturn: gen >> >> > " ... should we support handling return of external structures? " >> >> > self error: 'returning pointer to structure?' >> >> > >> >> > And I have no idea how to implement this (black magic EAX pop push >> >> > ;-)) >> >> > to >> >> > coerce result of struct. >> >> > >> >> >> >> >> >> you don't. if function takes an implicit struct pointer, you can just >> >> rewrite the function signature from >> >> NBCvSize cvGetSize(IplImage self) >> >> to: >> >> >> >> void cvGetSize(NBCvSize hidden, IplImage self) >> >> >> >> and so, you pass a fresh instance of NBCvSize as argument, which >> >> function then fill. >> >> But if C compiler using registers to return struct.. then you're >> >> screwed >> >> :) >> >> >> >> Anyways, Luc i will try to read docs about C ABI and see if we can >> >> easily add support for that. >> >> >> >> -- >> >> Best regards, >> >> Igor Stasenko. >> >> >> > >> > >> > >> > -- >> > best, >> > Eliot >> > >> >> >> >> -- >> Best regards, >> Igor Stasenko. > > > > > -- > best, > Eliot > -- Best regards, Igor Stasenko.

May 4, 2012

[Pharo-project] Pharo Kernel 2.0

by Pavel Krivanek

Hi, Markus added all packages to the Pharo 2.0 repository (THANKS!) so I was able to bring Pharo Kernel 2.0 image to life. Few notices: - I had to add FileSystem to the kernel. I added 'FileSystem-AnsiStreams' 'FileSystem-Core' 'FileSystem-Disk' and 'FileSystem-Memory' but maybe some are not necessary. - CommandLine has poor packaging because it includes tests too. - SimpleCodeLoader is now not needed and can be removed - CommandLine package will need some some next changes like to be independent on Compiler (because of Fuel) - KMPragmaKeymapBuilder needs better cleanup code. - I added 'Ring-Core-Kernel' 'Ring-Core-Containers' and 'Fuel' 'FuelMetalevel' 'FuelPackageLoader' packages (Fuel from external repository) I will open the issues. Cheers, -- Pavel

May 4, 2012

Re: [Pharo-project] NativeBoost help

by Eliot Miranda

On Fri, May 4, 2012 at 9:44 AM, Igor Stasenko <siguctua(a)gmail.com> wrote: > On 4 May 2012 18:37, Eliot Miranda <eliot.miranda(a)gmail.com> wrote: > > > > > > On Fri, May 4, 2012 at 6:13 AM, Igor Stasenko <siguctua(a)gmail.com> > wrote: > >> > >> On 4 May 2012 10:38, Luc Fabresse <luc.fabresse(a)gmail.com> wrote: > >> > Hi Igor, > >> > > >> >> > >> >> there is a workaround. > >> >> all structs in C is returned via pointer. > >> >> so, actually a true return type of a function is CvSize* > >> > > >> > > >> > I agree that usually structs are returned through pointers. > >> > But here it is not the case. > >> > > >> > I want to wrap this function: > >> > CvSize cvGetSize( const CvArr* arr ); > >> > > >> > and the CvSize struct is returned by value. Am I wrong? > >> > > >> > >> i don't know.. this is a bit shady part of C ABI. > >> Some compilers use return-by-value, some return by pointer > >> by forcing a caller to pass a hidden implicit argument as a pointer to > >> the struct, which then filled with values by > >> the function you call. > > > > > > IIRC x86 ABIs (all the ones I encountered at least, windows, mac, linux, > > solaris) support structure return (by-value structure return) by > passing a > > hidden first argument which is a pointer to a structure into which to > copy > > the result. The callee copies the result back through this pointer. > > > > Where these ABIs differ is in returning small structures. On win32 > > structures a power-of-two in size of 8 bytes or less are returned in > > EAX:EDX. IIRC on linux and mac all structures are returned through the > > hidden first argument pointer. See > > ThreadedIA32FFIPlugin>>returnStructInRegisters: > > > on macs (at least what gcc generates) it also returns structs <=8 > bytes size in registers: > > > so , currently i implemented it like that: > if sizeof(struct) <= 8 then i assume function returns it in EAX: EDX > otherwise via hidden pointer. > add a test for a structs like typedef struct { char a; char b; char c; } threeBytes; typedef struct { short a; short b; short c; } sixBytes; (and 5 & 7). You should see on Mac and Windows these are not passed back in registers. > >> > >> > >> It is easy to add support for that, but to make it work correctly can > >> be tricky. > >> See this, for example: > >> > >> > http://stackoverflow.com/questions/161788/are-there-any-downsides-to-passin… > >> > >> > >> >> > >> >> > >> >> so, what you do is to change the return value to simple void* > >> >> and then copy the contents of struct into instance of corresponding > >> >> NBExternalStructure subclass > >> >> using #fromPointer: method.. to conveniently access its fields > values. > >> > > >> > > >> > yes I saw that in NB tests. > >> > > >> >> > >> >> > >> >> > >> >> but this is in general, in your case this function is actually > inlined > >> >> by compiler (as CV_INLINE suggests), so there could be even no > >> >> exported symbol for it in dynamic library. > >> >> and the function is actually so simple, that you don't need to wrap > it > >> >> at > >> >> all. > >> >> you can just create a subclass of NBExternalStructure , named CvSize, > >> >> implement fieldDesc method as: > >> >> > >> >> fieldsDesc > >> >> ^ #( > >> >> int width; > >> >> int height > >> >> ) > >> >> > >> >> > >> >> and if you invoke: > >> >> > >> >> CvSize new width: 1 height:2 > >> >> > >> >> it will be equivalent as if you will call cvSize(1,2) function. > >> > > >> > > >> > yes it is exactly what I did. > >> > > >> > So my problem is: > >> > > >> > NBCvSize implemented as a subclass of NBExternalStruct with fieldDesc > >> > correctly set up. > >> > Then wrap the function cvGetSize: > >> > > >> > getSize > >> > self nbCall: #(NBCvSize cvGetSize(IplImage self)) module: > LIB_CV. > >> > > >> > But I get a debugger because: > >> > > >> > NBExternalStruct>>coerceReturn: gen > >> > " ... should we support handling return of external structures? " > >> > self error: 'returning pointer to structure?' > >> > > >> > And I have no idea how to implement this (black magic EAX pop push > ;-)) > >> > to > >> > coerce result of struct. > >> > > >> > >> > >> you don't. if function takes an implicit struct pointer, you can just > >> rewrite the function signature from > >> NBCvSize cvGetSize(IplImage self) > >> to: > >> > >> void cvGetSize(NBCvSize hidden, IplImage self) > >> > >> and so, you pass a fresh instance of NBCvSize as argument, which > >> function then fill. > >> But if C compiler using registers to return struct.. then you're screwed > >> :) > >> > >> Anyways, Luc i will try to read docs about C ABI and see if we can > >> easily add support for that. > >> > >> -- > >> Best regards, > >> Igor Stasenko. > >> > > > > > > > > -- > > best, > > Eliot > > > > > > -- > Best regards, > Igor Stasenko. > -- best, Eliot

May 4, 2012