similar to: [LLVMdev] patch: MSIL backend - interfacing vararg pinvoke functions with Mono

Hello, I work in Kalray (Montbonnot, France) and I'm PhD student at Universite Joseph Fourier in Grenoble. We want to use LLVM framework for MSIL code generation, which is part of my thesis. Currently I'm still reading LLVM's documentation and I've started completing the MSIL backend for running on Mono. Things that need to be fixed include pointers initialization, call to

Hi, Artur > So again, one of them makes the global pointers initialization work. > The second one allows executing code with vararg pinvoke functions under > Mono. It generates separate function declaration for each call signature. Is this mono-only feature? Or pretty "universal" one? -- With best regards, Anton Korobeynikov Faculty of Mathematics and Mechanics, Saint

Anton Korobeynikov wrote: > Hi, Artur > > >> I'm working on that backend now, so probably I'll send some more patches >> soon. I'd be grateful if you could give me some suggestions how to add >> some test for that backend to the test-suite. On Linux the output code >> could be run on Mono and compared with outputs for other backends but >>

Hi Artur, Artur Pietrek wrote: > Hello, > > I work in Kalray (Montbonnot, France) and I'm PhD student at Universite > Joseph Fourier in Grenoble. > We want to use LLVM framework for MSIL code generation, which is part of > my thesis. > Currently I'm still reading LLVM's documentation and I've started > completing the MSIL backend for running on Mono.

Hi, Here's a patch that fixes initialization of global pointers (also function pointers). I'm working on that backend now, so probably I'll send some more patches soon. I'd be grateful if you could give me some suggestions how to add some test for that backend to the test-suite. On Linux the output code could be run on Mono and compared with outputs for other backends but

Hi, After discussion with Anton, I did the whole thing in a little bit different way (please see the attached patch). It again looks like reinvented wheel for LLVM experts, I suppose, but it works. What I want to achieve is to generate real (with all variadic arguments) call signature for each variadic and external function, without the "vararg" keyword and without the "...",

Hi, Artur > I'm working on that backend now, so probably I'll send some more patches > soon. I'd be grateful if you could give me some suggestions how to add > some test for that backend to the test-suite. On Linux the output code > could be run on Mono and compared with outputs for other backends but > I'm not sure if it's preferable to involve third party

Hi, Artur > // CallSites have equal signatures > bool MSILWriter::cmpCallSite(CallSite A, CallSite B) { >   return (getCallSiteFType(A)==getCallSiteFType(B) && >     A.getAttributes()==B.getAttributes()); > } As it is impossible to honour argument attributes in MSIL I don't see why you should compare attributes. You seems to have the same MSIL call signature for calls

Hi, Artur Minor comments: > +// Comparision for std::lower_bound used in > MSILWriter::printExternals() > +static bool CompareInstructions(Instruction *A,Instruction *B) > +{ Put brace on the same line as function def. > + if ( !F->use_empty() ) // Print only if used > + { Likewise. Plus use "if (foo)" instead of "if ( foo )". All code

Hi, Artur > + if (isa<ConstantExpr>(I->constant)){ > + const ConstantExpr *CE = > dyn_cast<ConstantExpr>(I->constant); > + printConstantExpr(CE); > + ty = CE->getType(); > + } else { > + const Function * F = dyn_cast<Function>(I->constant); > + printValueLoad(F); > + ty =

Hi, Artur > %1 = tail call i32 (i8*, ...)* @printf(i8* noalias getelementptr ([16 x > i8]* @.str, i32 0, i32 0), i32 %0) nounwind > %10 = call i32 (i8*, ...)* @printf(i8* noalias getelementptr ([11 x i8]* > @.str2, i32 0, i32 0), i32 5) nounwind > > Instruction::isSameOperationAs() returns false for those two. Is it a > bug or I misunderstood something? These are two

Hi, Artur > The interesting for me part of the CallInst is printf(i8* noalias %0, i32 > 123). > I was diging in doxygen documentation but I really can't see the easy way to > compare those instructions and again finish with reinvented (but working) > wheel ;). Ah, sorry. I missed that you're doing variadic calls, not casting variadic function to definite ones. I think you

Hi Anton, Anton Korobeynikov wrote: > Minor comments: > Thanks for your comments and your patience, I'll now check the style four times before I send anything ;) >> + Instruction *instr = >> + const_cast<Instruction*>(dynamic_cast<const >> Instruction*>(*i)); >> > Sounds hacky. Why do you need to cast away const?

Hello, > So, looking for type of callee (not result, but function type!) you'll > obtain the > real "signature" of callee and if you'll strip all pointer cast you'll > obtain the "declaration" (=variadic) type of the callee. Maybe I misunderstood something but I just get the variadic declaration not the real "signature", like this: const

Hi Anton > > > The interesting for me part of the CallInst is printf(i8* noalias %0, i32 > > 123). > > I was diging in doxygen documentation but I really can't see the easy way > to > > compare those instructions and again finish with reinvented (but working) > > wheel ;). > Ah, sorry. I missed that you're doing variadic calls, not casting >

Artur, > OK, I just need the same signature for both of those instructions. Both are callinsts of same function, isn't that enough? Since it's a variadic function there is also a bitcast to proper type. So, looking for type of callee (not result, but function type!) you'll obtain the real "signature" of callee and if you'll strip all pointer cast you'll obtain the

Hello Anton > %1 = tail call i32 (i8*, ...)* @printf(i8* noalias getelementptr ([16 x > > i8]* @.str, i32 0, i32 0), i32 %0) nounwind > > %10 = call i32 (i8*, ...)* @printf(i8* noalias getelementptr ([11 x i8]* > > @.str2, i32 0, i32 0), i32 5) nounwind > > > > Instruction::isSameOperationAs() returns false for those two. Is it a > > bug or I misunderstood

Hi All, Since few days I observe weird change. Consider the following C code char array[] = "0123456789"; extern int test(char arr[], int size); int main(void) { return test(array-1, sizeof(array)-1); } using clang frontend i get this: %call = call i32 @test(i8* getelementptr inbounds ([11 x i8]* @array, i32 0, i32 -1), i32 10) ; <i32> [#uses=1] and using LLVM-GCC this: %1 =

On Fri, Aug 21, 2009 at 2:02 AM, Artur Pietrek<pietreka at gmail.com> wrote: > Hi All, > Since few days I observe weird change. > Consider the following C code > > char array[] = "0123456789"; > extern int test(char arr[], int size); > > int main(void) { >   return test(array-1, sizeof(array)-1); > } > > using clang frontend i get this: > >

On Fri, Aug 21, 2009 at 12:33 PM, Eli Friedman <eli.friedman at gmail.com>wrote: > On Fri, Aug 21, 2009 at 2:02 AM, Artur Pietrek<pietreka at gmail.com> wrote: > > Hi All, > > Since few days I observe weird change. > > Consider the following C code > > > > char array[] = "0123456789"; > > extern int test(char arr[], int size); > >