similar to: [LLVMdev] Disable optimization

On Thu, Nov 27, 2008 at 6:14 AM, RAAD B <raad_7007 at yahoo.com> wrote: > Hello together, > > although i use the "-O0" flag as follow, the llvm perform dead code > elimination. > How can i disable optimizing completely? Sorry, it's not possible; the gcc front-end does some optimizations which the LLVM backend can't do anything about, and dead code

Hi Eli, I want to disable optimizations because, i'm going to implement a framework in JAVA for educational purposes. I have planned to do followings: 1) Create LLVM-IR and export it as Assembly (without optimizing the source-code) 2) Transform the LLVM-Assembly to a data-structure, similar to LLVM data structure (Module, Function, BB ...) but implemented in Java 3) Optimizing

Hello together, how can i insert annotations in IR ? I want actually write something like: CallInst *call = CallInst::Create( ??? , aBasicBlock); in my pass. Regards Raad -------------- next part -------------- An HTML attachment was scrubbed... URL: <http://lists.llvm.org/pipermail/llvm-dev/attachments/20090305/cd1db761/attachment.html>

Ok, but what about a pass like alias-analysis. How are the results stored, so that other passes can use them. -Raad ________________________________ From: Eli Friedman <eli.friedman at gmail.com> To: LLVM Developers Mailing List <llvmdev at cs.uiuc.edu> Sent: Tuesday, February 17, 2009 9:52:58 AM Subject: Re: [LLVMdev] information-transfer between analysis-pases On Tue, Feb 17,

Hello together, I have seen that the analysis-results are stored in llvm-IR as annotations. For example <; preds = %entry> in basicBlock level and <; <i32*> [#uses=2]> for a variable. Is there any documentation about annotations? Regards Raad -------------- next part -------------- An HTML attachment was scrubbed... URL:

Hi, As far as I know, Instruction class does not inherit "Annotable" Class. Only Function inherits annotable according to the documentation of annotable class. What you are asking would require adding annotations at instruction level, which does not seem to be possible. You can maintain external maps between instructions and your annotations, --Kapil On Thu, Mar 5, 2009 at 7:17 PM,

Hi Duncan, thanks for the answer. But "opt -load ..../Decorator.so -dec <prog.bc> new_prog.bc" creates an invalid file. When i use "lli", i get "lli: error loading program 'new_prog.bc': Invalid bitcode signature" I have done the followings: 1) used "opt -p -load ..../Decorator.so -dec <prog.bc> new_prog.bc" 2) copied the output

On Tue, Feb 17, 2009 at 12:46 AM, RAAD B <raad_7007 at yahoo.com> wrote: > Hello together, > > I have seen that the analysis-results are stored in llvm-IR as annotations. > For example <; preds = %entry> in basicBlock level and <; <i32*> [#uses=2]> > for a variable. > > Is there any documentation about annotations? That isn't really analysis;

These are the contents of the generated file. Start >>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> ; ModuleID = '<stdin>' target datalayout =

Hi, how can I create the IR-representation from multiple source-files? Regards Raad -------------- next part -------------- An HTML attachment was scrubbed... URL: <http://lists.llvm.org/pipermail/llvm-dev/attachments/20090415/9040f657/attachment.html>

On Fri, Nov 28, 2008 at 12:36 AM, RAAD B <raad_7007 at yahoo.com> wrote: > Hi Eli, > > I want to disable optimizations because, i'm going to implement a framework > in JAVA for educational purposes. > > I have planned to do followings: > > 1) Create LLVM-IR and export it as Assembly (without optimizing the > source-code) > 2) Transform the

Hi, i need a sample-code, for which the llvm alias-analysis finds a *must-aliases*. I have tried codes like followings. In all cases, i see just *may-aliases* when i use "opt -aa-eval -print-all-alias-modref-info foo.bc": Regards Raad 1 ========================================== void foo() { int i = 2; int& r = i; } 2 =========================================== void

Hi, take a look inside the file new_prog.bc. How does it start? Ciao, Duncan.

Hello. I read following article in this ML. I'm implementing the library similar with RAAD's plan. - AST node classes for LLVM-IR - LLVM-Assembly writer class - 100% pure Java - Without any optimizations - Without any frontends (parsers) I would like to release my library publicly if all problems below could be solved. (Although it is in the early stage) My library is the mixture of my

Hello, I m trying to test individual O3 optimizations/ replicating O3 behavior on IR. I took unoptimized IR (O0), used disable-o0-optnone via (*clang -O0 -Xclang -disable-O0-optnone*). I read somewhere about *clang -O1 -Xclang -disable-O0-optnone,* so I also tested on this initial IR. I have observed by using individual optimizations, the performance (i.e time) is better when the base/initial

Hello together, I use the "opt" as following, to run my pass, which modifies the LLVM-IR by inserting instructions in basic-blocks. opt -load ..../Decorator.so -dec <prog.bc> /dev/null how can i store the modified bitcode on the disc (as bitcode or as human-readable assembly)? Regards Raad -------------- next part -------------- An HTML attachment was scrubbed...

Hi together, i want to create a map containing a set of aliases for each value. For example, for a code like: int main() { int i; int *p1 = &i; int *p2 = &i; return 0; } the map should contain something like: {i --> (p1, p2), ..... } For that, i do followings in my pass: AliasAnalysis &AA = getAnalysis<AliasAnalysis>(); AliasSetTracker *tracker =

Dropping opt level should not lead to ABI changes. Otherwise you won't be able to mix-match O2 and O0 objects either. David On Mon, Jun 17, 2013 at 10:59 AM, jahanian <fjahanian at apple.com> wrote: > Wouldn’t implementing this proposal be a red herring? By this I mean, it is > possible that > throughout the optimization phases, there is an implied assumption that all >

This showed up in RISKS and no one has mentioned it here yet, so.. OpenSSH contains lots of code like: char *password = read_passphrase(prompt, 0); [do stuff] memset(password, 0, strlen(password));

Wouldn’t implementing this proposal be a red herring? By this I mean, it is possible that throughout the optimization phases, there is an implied assumption that all functions are similarly optimized. An example would be under certain optimization flag, compiler changes calling convention of static functions. - Fariborz On Jun 17, 2013, at 8:58 AM, Andrea_DiBiagio at sn.scee.net wrote: >