similar to: [LLVMdev] Propagate clang attribute to IR

> On 24 Mar 2015, at 14:55, Aaron Ballman <aaron at aaronballman.com> wrote: > > On Tue, Mar 24, 2015 at 9:48 AM, Rinaldini Julien > <julien.rinaldini at heig-vd.ch> wrote: >> Hi, >> >> I want to *tag* some functions with some *flags*. I was using annotate((“myFlag”)) and everything was working fine until I tried on ObjC method. It seems that clang just

Hi all, I am newbie for llvm. I just create a global variable, there are some statements in my pass like: LoadInst* int64_64 = new LoadInst(pthreadPID, "", false, OptAplusOne); int64_64->setAlignment(8); int64_64->dump(); LoadInst* int32_65 = new LoadInst(gvar_int32_myFlag, "", false, OptAplusOne); int32_65->setAlignment(4);

Thanks for the suggestions Hal, So if I understand correctly, you're recommending we add a module flag <https://llvm.org/docs/LangRef.html#module-flags-metadata> to LLVM, something like: !llvm.module.flags = !{..., !123} !123 = !{i32 1, !"memeq_lib_function", !"user_memeq"} I've given it a try in the following patch: https://reviews.llvm.org/D56311 If this

Hi, all. I want to give programmer ability to tell LLVM that certain region of code is expected to get specialized optimization. So, I'm trying to make custom pragma to mark certain region of code and pass this information to LLVM, in the similar way that '#pragma clang loop unroll_count(N)' works. By tracking the framework of loop unroll pragma, I found out it works in the way below.

I've been investigating what is needed to ensure command line options are passed to the backend codegen passes during LTO and enable compiling different functions in a module with different command line options (see the links below for previous discussions). http://thread.gmane.org/gmane.comp.compilers.llvm.devel/78855 http://thread.gmane.org/gmane.comp.compilers.llvm.devel/80456 The command

Yes. I checked that file, but I wasn't sure whether that is right one to look at. What is *.td extension btw? When I google it, it says this is for Windows Debug Simulator. Why clang use this extension, which is limited to Windows environment? On Tue, Oct 25, 2016 at 7:00 PM, Vedant Kumar <vsk at apple.com> wrote: > That file is generated by the clang-tblgen tool: > >

> Yes. I checked that file, but I wasn't sure whether that is right one to look at. It seems promising. I imagine you'd want something like the LoopHint attr. > What is *.td extension btw? The llvm project uses that extension to identify TableGen inputs (http://llvm.org/docs/TableGen/). vedant > On Oct 25, 2016, at 4:12 PM, Sunghyun Park <sunggg at umich.edu> wrote:

LLVM's Attribute APIs need an overhaul. Current problems ================ First, testing for an attribute on an Argument is slow. llvm::AttributeSet::getAttributes(int) consumed 2% of cycles while optimizing llc during LTO. Our mid-level optimizations are constantly asking if a given argument has some attribute (nonnull, dereferencable, etc), and this is currently linear in the size of the

Hi, I'm trying to detect if a basicblock is part of a loop or not. I tried the llvm::LoopInfo like that (http://llvm.org/docs/doxygen/html/classllvm_1_1LoopInfo.html#a4abca289c73cd09487e05d11d9f7d877): LoopInfo *loop = new LoopInfo(); bool isLoop = loop->getLoopFor(myBB); // getLoopFor - Return the inner most loop that BB lives in. If a basic block is in no loop (for example the entry

Hi, I have a question about the llvm passes. I'm iterating over a basicblock and I can get an instruction and print it. Now, I want to iterate over the instruction and be able to modify the values of the instruction. For example, if my instruction is an add "<result> = add i32 4, %var" I want to transform it in a sub "<result> = sub i32 4, %var". I looked up

On Wed, Sep 28, 2016 at 11:13:35AM -0700, Mehdi Amini wrote: > > > On Sep 28, 2016, at 11:00 AM, Rinaldini Julien via llvm-dev <llvm-dev at lists.llvm.org> wrote: > > > > Hi, > > > > I’m using llvm+clang (+ a wrapper for swift that allow me to use my clang to compile swift files) inside Xcode and I have encountered two issues with the new clang 3.9 and

> De : 陳韋任 [chenwj at iis.sinica.edu.tw] > Date d'envoi : mardi 8 mai 2012 11:37 > À : Rinaldini Julien > Cc: LLVM Developers Mailing List > Objet : Re: [LLVMdev] RE : svn trunk comilation error > > Hi Rinaldini, > > You probably need to illustrate what your enviroment is, what revision you > checkout and how you build LLVM. I have no problem build LLVM svn here.

Hi Rinaldini, In order to find information about loops inside a given function you should use something like "LoopInfo *LI = P->getAnalysis<LoopInfo>()", remembering to add "AU.addRequired<LoopInfo>();" to your getAnalysisUsage method. If the function you are interested to is not located in the module being compiled (if you created it as an auxiliary function,

Rinaldini, What exactly did you run? Specifically, you may be missing some analysis passes that are necessary for LoopInfo to have the loop information you desire. -Hal On Thu, 26 Apr 2012 14:02:04 +0000 Rinaldini Julien <julien.rinaldini at heig-vd.ch> wrote: > Hi, > > I'm trying to detect if a basicblock is part of a loop or not. > > I tried the llvm::LoopInfo

> De : 陳韋任 [chenwj at iis.sinica.edu.tw] > Date d'envoi : mardi 8 mai 2012 11:37 > À : Rinaldini Julien > Cc: LLVM Developers Mailing List > Objet : Re: [LLVMdev] RE : svn trunk comilation error > > Hi Rinaldini, > > You probably need to illustrate what your enviroment is, what revision you > checkout and how you build LLVM. I have no problem build LLVM svn here.

Hi, Sorry I did not precise that, but I'm already doing that... I have, as you say, two different director, one for source and one for build (where I run my configure)... Cheers ________________________________________ De : 陳韋任 [chenwj at iis.sinica.edu.tw] Date d'envoi : mardi 8 mai 2012 05:21 À : Rinaldini Julien Cc: LLVM Developers Mailing List Objet : Re: [LLVMdev] svn trunk

>> // Increment loop variable and jmp >> BuildMI(*MBB_erase, MBB_erase->end(), db, >> TII->get(X86::ADD64ri32),reg).addReg(reg).addImm(8); > > It looks like this instruction is defining virtual register "reg" the second time. Thx for your answer... Why would it define it again? I just want to use this register and add something to it... Cheers

Hi, I’m using llvm+clang (+ a wrapper for swift that allow me to use my clang to compile swift files) inside Xcode and I have encountered two issues with the new clang 3.9 and Xcode 8: 1) when compiling with the deployment target set to 10.0, I got this error: clang-3.9: error: invalid version number in '-miphoneos-version-min=10.0’. If I set back the deployment target to 9.3 it works fine.

Hi, I'm writing a pass that implements a jump table with an array of blockaddress and an indirectbr instruction. It get a blockaddress in the array (via getelementptr and an index) and then jump to this basicblock via the indirectbr. I tried to compile several libraries to test my pass and the run their test-suite. It works fine with, for e.g, libTomCrypt (in -O0,1,2,3). With GMP or

Re-adding the list, below message was sent to me alone: On 11 August 2011 13:45, Rinaldini Julien <julien.rinaldini at heig-vd.ch> wrote: > Thx for all answers... > > I'll try that. But in a long term what I want to do will be a bit more complicated... It was just an example. In this case, the goal is to replace all add with sub that return the same result, like: > > var