Re: [PATCH] D21081: MPI-Checker patch for Clang Static Analyzer

This revision was automatically updated to reflect the committed changes. Closed by commit rL272529: [analyzer] Add checker to verify the correct usage of the MPI API (authored by dcoughlin). Changed prior to commit: http://reviews.llvm.org/D21081?vs=60472=60486#toc Repository: rL LLVM

Re: [PATCH] D21081: MPI-Checker patch for Clang Static Analyzer

Alexander_Droste added a comment. Hi Devin, this is much cleaner. Looks good for me! http://reviews.llvm.org/D21081 ___ cfe-commits mailing list cfe-commits@lists.llvm.org http://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

Re: [PATCH] D21081: MPI-Checker patch for Clang Static Analyzer

dcoughlin updated this revision to Diff 60472. dcoughlin added a comment. Alexander, a drawback with using keeping a vector of MPIBugReporters in the checker is that this will essentially "leak" all the created bug reporters. How about this instead: - Remove MPIBugReporter's

Re: [PATCH] D21081: MPI-Checker patch for Clang Static Analyzer

Alexander_Droste added a comment. F2057622: MPI-Checker.diff Hi, I now solved this, by using a `small_vector` of `MPIBugReporter`s (MPIChecker.h line: 101). So each time a report is generated, a new `MPIBugReporter` is added to that vector, in order to refer

Re: [PATCH] D21081: MPI-Checker patch for Clang Static Analyzer

Alexander_Droste added a comment. Hi Devin, thanks for fixing the GCC build errors and setting up the commit! Unfortunately, creating an `MPIBugReporter` instance only when a bug occurs does cause a separate issue. The instance seems to be freed, before the reports get flushed. void

Re: [PATCH] D21081: MPI-Checker patch for Clang Static Analyzer

dcoughlin added a comment. Here is the ASan report. It looks to me like the issue is that MPIChecker is holding on a reference to a clang::ento::BugReporter after the lifetime of the BugReporter has ended. The BugReporter lives in ExprEngine and the analyzer creates a new ExprEngine for each

Re: [PATCH] D21081: MPI-Checker patch for Clang Static Analyzer

dcoughlin added a comment. Alexander, When I committed your patch, the AddressSanitizer bots found a memory corruption issue so I reverted it. This will need to be fixed before we can commit the patch. Since Phabricator automatically closed http://reviews.llvm.org/D12761, I've created a new

[PATCH] D21081: MPI-Checker patch for Clang Static Analyzer

dcoughlin created this revision. dcoughlin added reviewers: zaks.anna, dcoughlin, Alexander_Droste. dcoughlin added a subscriber: cfe-commits. (Cloning the revision for the ASan issue since the old one is closed and can't be re-opened) This patch adds a static analysis checker to verify the