On Thu, Jun 14, 2012 at 10:01 PM, Jerry James <loganjerry@xxxxxxxxx> wrote: > I'm having a problem with building the coq package for the new OCaml > 4.00.0, and I'm at my wits' end. There were some bad interactions > between the new OCaml, camlp5, and coq which I think I have > successfully worked around. (It isn't pretty, but it seems to do the > job.) But now, after the tools are built, the documentation building > step is failing due to seemingly random segfaults in the coqdoc tool. I patched a few more things in the coq build process for OCaml 4 today and got builds that went all the way to completion in my 64-bit and 32-bit Fedora Rawhide VMs. I rejoiced and kicked off a koji build, which failed with a segfault in coqdoc during the 32-bit build again: http://koji.fedoraproject.org/koji/taskinfo?taskID=4168975 The apparently random pattern of segfaults makes me suspect that the garbage collector is involved somehow. Has anybody had experience with a relatively long-running OCaml 4 program (one that would go through multiple rounds of collecting garbage)? -- Jerry James http://www.jamezone.org/ -- devel mailing list devel@xxxxxxxxxxxxxxxxxxxxxxx https://admin.fedoraproject.org/mailman/listinfo/devel
