I hope this is the correct avenue to contact the poolalloc developers. I'm trying to use an alias analysis from the poolalloc repository and can't get it to compile with the latest LLVM. CMake is now required for LLVM, I'm pretty sure at least, but poolalloc does not seem to use it correctly. The README in the project refers to the old Makefiles. I corrected a minor CMake error and a bunch of compiler errors in poolalloc when compiling with the current version of LLVM. I'd be glad to submit them to wherever. But poolalloc doesn't seem to be on reviews.llvm.org.
Dear Jon, I'm not sure that anyone is currently keeping poolalloc trunk updated with LLVM mainline. There were some updates around the LLVM 3.7 release, but I don't think there's been any activity since then. If you want to use DSA, you can fetch a copy from https://github.com/jtcriswell/llvm-dsa. Kevin Hu used this version for a project (though he was only using the call graph analysis within DSA). I think this was updated to work with a version of LLVM mainline shortly before LLVM 3.7 was released. I think https://github.com/jtcriswell/safecode-llvm37 has DSA updated to the LLVM 3.7 release (though I don't think it was tested). Regards, John Criswell On 1/19/16 9:00 AM, Jon Eyolfson via llvm-dev wrote:> I hope this is the correct avenue to contact the poolalloc developers. > > I'm trying to use an alias analysis from the poolalloc repository and can't get > it to compile with the latest LLVM. CMake is now required for LLVM, I'm pretty > sure at least, but poolalloc does not seem to use it correctly. The README in > the project refers to the old Makefiles. > > I corrected a minor CMake error and a bunch of compiler errors in poolalloc > when compiling with the current version of LLVM. I'd be glad to submit them to > wherever. But poolalloc doesn't seem to be on reviews.llvm.org. > _______________________________________________ > LLVM Developers mailing list > llvm-dev at lists.llvm.org > http://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-dev-- John Criswell Assistant Professor Department of Computer Science, University of Rochester http://www.cs.rochester.edu/u/criswell
Hi John, Do you know of anyone that's responsible for the official 'poolalloc' repository? or are there just multiple forks floating around? Most of the changes I made were trivial, except for the linking stage that's failing right now. Thanks, Jon On Tue, Jan 19, 2016 at 09:41:00AM -0500, John Criswell wrote:> Dear Jon, > > I'm not sure that anyone is currently keeping poolalloc trunk updated with > LLVM mainline. There were some updates around the LLVM 3.7 release, but I > don't think there's been any activity since then. > > If you want to use DSA, you can fetch a copy from > https://github.com/jtcriswell/llvm-dsa. Kevin Hu used this version for a > project (though he was only using the call graph analysis within DSA). I > think this was updated to work with a version of LLVM mainline shortly > before LLVM 3.7 was released. > > I think https://github.com/jtcriswell/safecode-llvm37 has DSA updated to the > LLVM 3.7 release (though I don't think it was tested). > > Regards, > > John Criswell > > > On 1/19/16 9:00 AM, Jon Eyolfson via llvm-dev wrote: > >I hope this is the correct avenue to contact the poolalloc developers. > > > >I'm trying to use an alias analysis from the poolalloc repository and can't get > >it to compile with the latest LLVM. CMake is now required for LLVM, I'm pretty > >sure at least, but poolalloc does not seem to use it correctly. The README in > >the project refers to the old Makefiles. > > > >I corrected a minor CMake error and a bunch of compiler errors in poolalloc > >when compiling with the current version of LLVM. I'd be glad to submit them to > >wherever. But poolalloc doesn't seem to be on reviews.llvm.org. > >_______________________________________________ > >LLVM Developers mailing list > >llvm-dev at lists.llvm.org > >http://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-dev > > > -- > John Criswell > Assistant Professor > Department of Computer Science, University of Rochester > http://www.cs.rochester.edu/u/criswell >