similar to: [LLVMdev] General LLVM question

Hello, Chris. > It depends on what sort of attribute it is. If it is something specific > to a particular function, adding it to the Function object makes sense > (e.g. see alignment and section info). If it affects the calling > convention, adding it to FunctionType makes sense. I'm wanting some "general" attribute interface which can be applied to concrete

Hello llvm-dev@ I have a pass that puts a global constant data array into the module. And another pass that gets runned by lld during LTO stage. When I look at bytecode files corresponding to source TUs produced by `clang -flto`, I can see GVs inserted by my pass #1. However, when pass #2 gets to run, these GV's are gone. I figured out that passing -mllvm -compute-dead=0 to the linker fixes

Hi Rafael, Yes, merging gv prevents linker to do garbage collection. Should it be implemented as a peephole pass? If we do it too early, the distance between GVs are not fixed yet. PS: Below is the GCC output with "extern" hidden: ldr r2, .L2 stmfd sp!, {r3, lr} .save {r3, lr} .LPIC0: add r0, pc, r2 bl _Z4initPv(PLT) ldr r1, .L2+4 .LPIC1: add r0, pc, r1 bl _Z4initPv(PLT) ldr

> On Mar 11, 2016, at 12:16 PM, Rafael Espíndola <rafael.espindola at gmail.com> wrote: > On 11 March 2016 at 14:58, John McCall <rjmccall at apple.com> wrote: >>> On Mar 11, 2016, at 11:26 AM, Rafael Espíndola <rafael.espindola at gmail.com> wrote: >>>> Now, there are a number of things about linkage that are kindof orthogonal, >>>> and it

Hi Tim, The global merge pass puts the GVs into a sturcture to guarantee their address are contiguous. It works for static GVs but for global hidden GVs, this will cause name resoltion fail during linking .o into .so Any thoughs? Thanks, Weiming > Hi Weiming, > > On 12 March 2014 17:43, Weiming Zhao <weimingz at codeaurora.org> wrote: >> Clang will emit 1 GOT entry for

Hi, all, I managed to insert the assertion in my bytecode, but the result is not really what I expected :( Let me do a quick recap: I am trying to instrument the bytecode with some assertions, and to do this, I want to insert the abort() function into the bytecode. The problem is that LLVM is creating a new abort() function, instead of using the one that already exists in libc. Let me

On Jul 27, 2009, at 3:05 PM, Devang Patel wrote: > On Mon, Jul 27, 2009 at 1:13 PM, Dan Gohman<gohman at apple.com> wrote: > >> >> >> On Jul 27, 2009, at 10:10 AM, Devang Patel wrote: >> >> >> >> >> >>> In LLVM IR metadata is used to attach auxiliary information with >>> >>> various IR constructs. Currently

On Mon, Jul 27, 2009 at 1:13 PM, Dan Gohman<gohman at apple.com> wrote: > > On Jul 27, 2009, at 10:10 AM, Devang Patel wrote: > > >> In LLVM IR metadata is used to attach auxiliary information with >> various IR constructs. Currently metadata information is represented >> using MDNode and MDString. The metadata can refer to LLVM values but >> these

Yest. thanks. I just resolved this error by installing MPFR, MPC and GMP(by the way, these are not listed as prerequisites in the website.). But other errors come: /home/xxx/llvm/tools/dragonegg/src/TypeConversion.cpp: In function > ‘llvm::FunctionType* ConvertArgListToFnType(tree_node*, > llvm::ArrayRef<tree_node*>, tree_node*, bool, llvm::CallingConv::ID&, >

Victor Campos wrote: > Guys, > > I would like to instrument the bytecode that LLVM produces with > assertions. I have written the instrumentation code manually, but I do > not know how to halt the program in case the assertion is false. I took > a look into the bytecode that LLVM produces for a program like: > > #include <stdlib.h> > int main() { >

Hello, Jon > What is the correct way to do this? FunctionType constructor takes extra boolean for marking a varargs function. There is some trick: feed your bytecode to llvm2cpp tool and you'll get C++ source, which will construct the module you provided. From this source you can see, how different things can be created via LLVM API. -- With best regards, Anton Korobeynikov. Faculty

In the following scrap of code (pared down from actually useful code), the func->dump() command segfaults iff the commented-out line is uncommented. This is with llvm 3.0. I'm only dipping my toes into the waters of llvm for the first time, and have no idea what I am doing incorrectly. In actual code, I would be wanting to call the memcpy intrinsic eventually, of course. int main(void)

Hello, can anyone help me calling native functions from LLVM-Bytecode functions run in the JIT? I have a program which creates an LLVM execution engine and adds modules and functions to it on the fly. I need to call some native functions of my program from bytecode functions which causes some troubles as it appears not to be documented. My test scenario works like the following: I have

Are you able make calls to well known external functions such as printf? As far as I known, this capability is well tested on x86 / Linux. I am wondering if there is some name mangling issue? Evan On Jun 7, 2007, at 8:38 AM, Jan Rehders wrote: > Hello, > > can anyone help me calling native functions from LLVM-Bytecode > functions run in the JIT? > > I have a program

> On Mar 11, 2016, at 1:40 PM, Rafael Espíndola <rafael.espindola at gmail.com> wrote: >>> And you can't also just produce STV_PROTECTED for every symbol. I >>> would love for that to be the case, but while most ELF systems support >>> copy relocations and related PLT hacks for functions it is not >>> practical to do it. >> >> I’m

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

On Wednesday 30 July 2008 18:13:27 Duncan Sands wrote: > On x86-64 linux, in stage 2, I get: > > c++ -g -O2 -DIN_GCC -W -Wall -Wwrite-strings -Wstrict-prototypes -Wmissing-prototypes -pedantic -Wno-long-long -Wno-variadic-macros -Wno-overlength-strings -Wold-style-definition -Wmissing-format-attribute -fno-common -DHAVE_CONFIG_H -o f951 \ > fortran/arith.o

Thank you Chris, for your hint, but I am still too stupid. I tried two versions asm_arguments.push_back(Type::VoidTy); FunctionType *asm_type = FunctionType::get(Type::VoidTy, asm_arguments, false); Alternatively FunctionType *asm_type = FunctionType::get(Type::VoidTy, std::vector<const Type*>(), false); . Can you give me a snippet of example code, or somebody else?

I am getting a timeout on running totem .... searching led me to think its gvs related.... Error creating proxy: Error calling StartServiceByName for org.gtk.vfs.GoaVolumeMonitor: Timeout was reached (g-io-error-quark, 24) Then totem does start - but with like a 30 second delay. I ran this command and it says its inactive systemctl --user status gvfs-mtp-volume-monitor.service ?

Hi all, Recently I have implemented a transformation pass based on LLVM3.5 and its function is to duplicate the function's argument list in a bytecode file and replace all use of original function with modified function. In LLVM3.5, the pass can work properly. However, when I tried to transplant the pass to LLVM3.9, the error "Argument value does not match function argument type!"