search for: rangeanalysi

Hi, Andrey, sorry for the delay: I made a page with the code available for download: http://homepages.dcc.ufmg.br/~douglas/projects/RangeAnalysis/RangeAnalysis.html Feel free to get it, and if you need some help, I will be happy to tell you how to set the analysis up, in case the explanation in the page is not good. I also have a report describing the implementation here: ( http://homepages.dcc.ufmg.br/~douglas/projects/RangeAnalysi...

Can't read your paper because the permission is not set. Chuck On 2/25/2011 12:16 PM, Douglas do Couto Teixeira wrote: > Hi, Andrey, > > sorry for the delay: I made a page with the code available for > download: > http://homepages.dcc.ufmg.br/~douglas/projects/RangeAnalysis/RangeAnalysis.html > <http://homepages.dcc.ufmg.br/%7Edouglas/projects/RangeAnalysis/RangeAnalysis.html> > Feel free to get it, and if you need some help, I will be happy to > tell you how to set the analysis up, in case the explanation in the > page is not good. >...

Hi Douglas, On 21.02.2011 20:27, Douglas do Couto Teixeira wrote: > My work is not part of the LLVM mainline yet. But I would be happy to > contribute with the code of my range analysis implementation if it can help > you in something else. We were thinking of adding VRP to LLVM too, though we were mostly interested in Patterson's approach (i.e. not connected with SSI form). It

...(cmake ../llvm -G "Eclipse CDT4 - Unix Makefiles"), re-open it in Eclipse. 4. Fix some minor issues with range-analysis (some symbols were renamed in newer LLVM releases). 5. Add the necessary code to my own pass, almost verbatim from https://code.google.com/p/range-analysis/wiki/HowToUseRangeAnalysisInAnotherPass. Everything compiles fine, but loading the pass in LLVM gives me the following error: > /opt/llvm/bin/opt: symbol lookup error: ./build/lib/RangeAnalysis.so: undefined symbol: _ZN4llvm13ConstantRangeC1ENS_5APIntES1_ > error: unable to interface with target machine I use LLVM...

dw <dw.dev.mailing at gmail.com> writes: [snip] > Everything compiles fine, but loading the pass in LLVM gives me the > following error: > >> /opt/llvm/bin/opt: symbol lookup error: ./build/lib/RangeAnalysis.so: undefined symbol: _ZN4llvm13ConstantRangeC1ENS_5APIntES1_ >> error: unable to interface with target machine > > I use LLVM 3.4svn on Ubuntu x64. > > Do I miss a specific library? $ c++filt _ZN4llvm13ConstantRangeC1ENS_5APIntES1_ llvm::ConstantRange::ConstantRange(llvm::APInt...

...not tell if this part of the code is already completely done. Would it be possible to print the ranges of the values computed by LVI? Third, besides JumpThreading.cpp is there any other client that uses LVI? I would like to couple my range analysis ( http://homepages.dcc.ufmg.br/~douglas/projects/RangeAnalysis/RangeAnalysis.paper.pdf) with the LVI interface, so that is why I am interested on it. Kind regards, Douglas On Mon, Mar 14, 2011 at 6:55 PM, Douglas do Couto Teixeira < douglasdocouto at gmail.com> wrote: > Hi guys, > > I have an analysis that is able to answer questions like t...

...rôme Leroux, Jan Reineke, Helmut Seidl, Grégoire Sutre, Reinhard > Wilhelm: Efficient Algorithms 2009: 422-437 > > 4. Linear Time Range Analysis with Affine Constraints.Douglas do Couto > Teixeira and Fernando Magno Quintao Pereira > (http://homepages.dcc.ufmg.br/~douglas/projects/RangeAnalysis/RangeAnalysis.paper.pdf > <http://homepages.dcc.ufmg.br/%7Edouglas/projects/RangeAnalysis/RangeAnalysis.paper.pdf>) > > 5. Constant propagation with conditional branches. Mark N. Wegman and > F. Kenneth Zadeck, In ACM Transactions on Programming Languages and > Systems (TOP...

...ited. Thomas Gawlitza, Jérôme Leroux, Jan Reineke, Helmut Seidl, Grégoire Sutre, Reinhard Wilhelm: Efficient Algorithms 2009: 422-437 4. Linear Time Range Analysis with Affine Constraints. Douglas do Couto Teixeira and Fernando Magno Quintao Pereira ( http://homepages.dcc.ufmg.br/~douglas/projects/RangeAnalysis/RangeAnalysis.paper.pdf ) 5. Constant propagation with conditional branches. Mark N. Wegman and F. Kenneth Zadeck, In ACM Transactions on Programming Languages and Systems (TOPLAS), 181-210, 1991. 6. Efficient SSI Conversion. André Luiz C. Tavares, Fernando Magno Quintão Pereira, Mariza A. S. Bi...

Hi guys, I have an analysis that is able to answer questions like this: given an integer variable, what is the interval of values that this variable can assume during the program's execution? I want to integrate this analysis into LLVM and it seems LVI (Lazy Value Info) is the best place to do this kind of stuff. Can someone give some hints about what I have to do to integrate my analysis

Does llvm have a pass that minimizes the bitwidth of llvm instructions? For instance:   %8 = and i32 %7, 63 63 is 111111 in binary. So the 'and' instruction only requires 6 bits. We could rewrite the above code as:   %8 = trunc i32 %7 to i6   %9 = and i6 %8, 63 Since we only need the lower 6 bits we could also propagate this change backwards to reduce the bitwidth of prior

...39;s not even implemented there. Short answer, without special casing this in magic ways, i wouldn't expect this to get fixed anytime soon. If we fixed assume in one of the ways i thought about, like bodik's extended ssa: http://homepages.dcc.ufmg.br/~fernando/classes/dcc888/ementa/slides/RangeAnalysis.pdf You would at least see that the load result is used by an assume, and could go look at that assume and so something with it. Currently, it's a few steps away. In the current scheme, assumes just float in air, and so it can be hard to see what their effects touch :) On Fri, Jun 10, 201...

...ut special casing this in magic ways, i wouldn't >> expect this to get fixed anytime soon. >> >> If we fixed assume in one of the ways i thought about, like bodik's >> extended ssa: >> >> http://homepages.dcc.ufmg.br/~fernando/classes/dcc888/ementa/slides/RangeAnalysis.pdf >> >> You would at least see that the load result is used by an assume, and >> could go look at that assume and so something with it. Currently, it's a >> few steps away. >> >> In the current scheme, assumes just float in air, and so it can be hard >...

On Wed, Mar 23, 2011 at 6:24 PM, John Criswell <criswell at illinois.edu> wrote: > > > On 3/23/11 8:06 AM, Douglas do Couto Teixeira wrote: > > > > the execution of a program. Thus, for each integer variable, a range > > analysis determines its lower and upper limits. A very simple range analysis > > > would, for instance, map each variable to the limits

...enson, which gives a nice description of an iterative analysis. Although I have not compared it with an implementation of Stephenson's algo yet, I believe Su's technique would converge much faster. I have an explanation about our implementation here: ( http://homepages.dcc.ufmg.br/~douglas/RangeAnalysis.html). We have been able to apply it on the whole LLVM test suite, and also on the SPEC CPU 2006 programs. Results are good for small programs (about 40% bitwidth reduction in Stanford benchmark, for instance), but are a bit disappoint for big programs (around 8% reduction for SPEC CPU 2006). In a...

Hi! I'm a student who would like to participate on Google SOC for LLVM, and was thinking about what project to pick. I saw on the "Open projects" page that Value Range Propagation is not implemented and thought about doing it, based on a paper by Patterson (it's also used by GCC). But then I saw that last year someone did a Range Analysis pass that seems to do pretty much the

As of llvm 3.8, the early CSE pass seems to remove llvm.assume intrinsics. Is this the expected behavior? I've attached as small-ish example of this happening in my production code. $ opt -early-cse before-early-cse.ll -S > after-early-cse.ll Note the use of the assume intrinsic indicating that the loaded value %channels equals 3. In a later pass I replace the load instruction with

Hi All, We have recently found some optimization opportunities created by replicating code into branches in order to enable optimization. In general, the optimization opportunity we are pursuing is like the following. Given pseudo-code: // block A if (some condition) // block B // block C If it can be efficiently proven that some portion of block C can be simplified had control flow not