similar to: [LLVMdev] JIT, FFI

Hello, I'm using the llvm-c bindings for C, but I'm getting some problems, the LLVMCreateExecutionEngine is returning 1 but the error message is an empty string (it's !=NULL). The steps I'm doing is to create the module, then I call LLVMVerifyModule to verify and it works fine, later I call LLVMCreateModuleProviderForExistingModule and it returns !=NULL MP, so it's ok, but

You're right, that probably won't have very good performance. What we (unladen-swallow) do is to call getPointerToFunction which gives a function pointer of type (void*)(void*) which we then cast to the appropriate function type and then call from C/C++. That way the C/C++ compiler for your application can codegen the function call once instead having the JIT do it many times. Reid On

Does anyone knows if there is a patch for OpenSSH in order to make it work with 0.9.8r OpenSSL in FIPS Mode ? I'm having problem with the RSA_public_decrypt() function that is failing in FIPS Mode, I changed it to use RSA_verify instead and setting the flag "RSA_FLAG_NON_FIPS_ALLOW", and it's working fine now, but I'm not sure if this is allowed in FIPS Mode, does anyone

Hello, I'm facing some problems while building LLVM 2.8 shared libraries from source using CMake: cmake -DBUILD_SHARED_LIBS=true .. This is the output error: CMake Error: The inter-target dependency graph contains the following strongly connected component (cycle): "LLVMARMCodeGen" of type SHARED_LIBRARY depends on "LLVMARMAsmPrinter"

Hello, We generally use CentOS for some servers, and so do not use a GUI interface. However, I have recently installed CentOS 6.3 onto a PC with KDE. I am familiar with KDE as I use it with Fedora for my work PC. The problem is that we would like to configure the login screen, so that it does not show the user list, that it does not allow the shutdown or reboot commands (from the login screen),

I'm getting a segfault when I call `SDL_Init( SDL_INIT_VIDEO )` within the MCJIT.  If I compile the same program to an EXE I don't have a problem. Are there any general restrictions, or known issues, with loading a library like SDL2 via the MCJIT?  (LLVM-3.8 still) The stack trace, from GDB, is below. I noticed the error actually happens deep inside a transitively loaded GL drive module,

Sorry to be so slow in responding. I've done what you suggested. I created /etc/sysconfig/desktop and entered "$DISPLAYMANAGER=KDM". I rebooted but there was no change. GDM is still being used. Any other suggestions? mw ======================== Just create the file /etc/sysconfig/desktop and put in it: DESKTOP=KDM I believe it's DESKTOP - you can dbl check /etc/X11/prefdm

I've installed CentOS 6.3 on a quad-four box. The model only indicated dual core but CentOS is telling me there are four. This is the first multi-core Linux installation I've had. What is the best way to utilize the multi-core CPUs? I'd like to distribute the load but I'm unsure how to do that. mw -- -- "Lose not thy airspeed, lest the ground rises up and smites

Out of curiosity, I'm replacing the JIT with MCJIT on my compiler. As all "external" functions are provided by the language's FFI mechanism, it does MyExecutionEngine->DisableSymbolSearching(); MyExecutionEngine->InstallLazyFunctionCreator(&MyLazyFunctionCreator); which works fine with the JIT. However, MCJIT insists on resolving unknown symbols by searching them

I'm sorry, you can pinch me for that I'm trying to install qmailadmin and getting this error about not having vlimits.h file, I found that this is a system file, so I thought may be someone has it and can email it to me. Thanks Anton

I guess we'll have to add that to the list of things that needs to be done to replace the old JIT. In the projects I've worked on that use MCJIT, we've been using getPointerToFunction and then calling it from outside the MCJIT engine, but I suppose that the generic handling in runFunction would be useful in some cases. -Andy -----Original Message----- From: llvmdev-bounces at

Does anyone knows why in some OpenSSH patches for FIPS we have something like: SSLeay_add_all_algorithms(); if (FIPS_mode() && !FIPSCHECK_verify(NULL, NULL)) { fprintf(stderr, "FIPS integrity verification test failed.\n"); exit(3); } This block of code is always in main() soon after starting service/client. Why are they

(I will not make a joke about the streets of Carlyle) Jan 13 19:09:07 mail dovecot: lda(john at example.com): Panic: file mail-transaction-log-file.c: line 1148 (mail_transaction_log_file_get_highest_modseq_at): assertion failed: (offset <= file->sync_offset) Jan 13 19:09:08 mail kernel: pid 8435 (dovecot-lda), uid 89: exited on signal 6 (core dumped) Jan 13 19:14:16 mail dovecot: lda(john

I looked around for the status of OrcJIT FFI support. The last e-mail thread I could find was this one: Link <http://lists.llvm.org/pipermail/llvm-dev/2015-February/081679.html> Raw: http://lists.llvm.org/pipermail/llvm-dev/2015-February/081679.html Is OrcJIT now considered stable enough that there can be "official" exposed C APIs? If not, what's the standard approach if I

Óscar Fuentes <ofv at wanadoo.es> writes: > Out of curiosity, I'm replacing the JIT with MCJIT on my compiler. As > all "external" functions are provided by the language's FFI mechanism, > it does > > MyExecutionEngine->DisableSymbolSearching(); > MyExecutionEngine->InstallLazyFunctionCreator(&MyLazyFunctionCreator); > > which works fine

I'm having another play with LLVM using the OCaml bindings for a forthcoming OCaml Journal article and I have a couple of remarks: Firstly, I noticed that the execute engine is very slow, taking milliseconds to call a JIT compiled function. Is this an inherent overhead or am I calling it incorrectly or is this something that can be optimized in the OCaml bindings? Secondly, I happened to

Hi, I''ve been using FFI with a Ruby 1.9.3 built with MSVC++ and it''s been working well. One thing I''ve run into though is this: ffi_lib :msvcrt But that''s not the runtime I want. But I don''t want to hard code the runtime name either. I realize I could parse it out of RbConfig, but I was hoping for something nicer. Is there a way we could create an

Hi Gordon, I wrote a small example, but while running I get an error("Tied to execute an unknown external function"), where I am wrong? Thanks in advance. Kirill. int some_test_func( int ){ std::cout << "!!!!" << std::endl; return 8848; } int _tmain(int argc, _TCHAR* argv[]){ Module *M = new Module("test"); ExistingModuleProvider* MP = new

Hi all, How does this look as a potential FFI for our theoretical Ruby++ ? I''m not sure how to do mixins, though. Some sort of special directive? Or just add it to inheritance chain? Anyway, this is just a rough prototype. No attempt at an actual implementation yet. Regards, Dan // foo.h - Function prototypes #define _FOO_H_ using namespace ruby; // Subclass Object class Foo :

While browsing the configure generated llvm/Config/config.h, I'm noticed that the macros: HAVE_FFI_CALL HAVE_FFI_FFI_H HAVE_FFI_H are not defined on OS X (currently 10.6.5). Given that ffi exists on this platform (both headers and lib), is this result intended? Thanks Garrison