search for: linearizable

Hello, I intend to create a tool for detecting thread-safety problems within libraries. Therefore I would like to use the "Linearizability" correctness principle. Were there any efforts within the KLEE project regarding to that topic? Would it be interesting for the community at all? Best regards, Andreas

Hi Andreas, Cloud9 is a project that has an extensive POSIX environment model which also lets you symbolically execute multithreaded programs. It is based on KLEE. http://dslab.epfl.ch/pubs/cloud9.pdf Portend builds on top of Cloud9 to perform race detection/classification: http://dslab.epfl.ch/pubs/portend.pdf Cloud9 code base is also public, it might be of interest to you:

...l void @llvm.atomic.membarrier( i1 true, i1 true, i1 true, > i1 true ) > %result = call <ty> @llvm.atomic.las( <ty>* %ptr, <ty> %value ) Shouldn't you have a second membar after the las() to be very conservative (i.e., if las() is supposed to really be linearizable)? Otherwise, the effects of the las() can be reordered with respect to effects of subsequent instructions. This also shows that you get additional overhead in the codegen if barriers are not associated with an operation: To emit efficient code, codegen would have to check whether membars are n...

Hello, This is the second major revision of the atomic proposal for LLVM. I will try and give a brief overview of the motivating changes, but a greater portion of the text has changed, along with some changes to the proposed additions. http://chandlerc.net/llvm_atomics.html - The proposal has been rewritten to better delineate the goals and purposes of LLVM, and these additions to LLVM. The why

I have a model Q=K*A*(R^r)*(S^s) A, R, and S are data I have and K is a curve fitting parameter. I have linearized as log(Q)=log(K)+log(A)+r*log(R)+s*log(S) I have taken the log of the data that I have and this is the model formula without the K part lm(Q~offset(A)+R+S, data=x) What is the formula that I should use? Thanks for all of your help. I can provide a subset of data if necessary.

...mbarrier( i1 true, i1 true, i1 true, > > i1 true ) > > %result = call <ty> @llvm.atomic.las( <ty>* %ptr, <ty> %value ) > > Shouldn't you have a second membar after the las() to be very conservative > (i.e., if las() is supposed to really be linearizable)? Otherwise, the > effects of the las() can be reordered with respect to effects of subsequent > instructions. You are probably right here. It was very late, and as mentioned, the GCC spec is extremely ambiguous on the precise semantics for these intrinsics. I'm going to move to what I t...