http://gcc.gnu.org/bugzilla/show_bug.cgi?id=45841
--- Comment #29 from Paolo Carlini <paolo.carlini at oracle dot com> 2010-10-05 20:39:11 UTC --- Can I ask a simple question? At this point in the analysis, do we have a testcase failing on Linux or Darwin, thus showing a clear regression in the generic code? If the answer is yes, do we believe a fix is reasonably simple? Otherwise, if the situation is too confused maybe we have to humbly take out the improvements to filebuf, allow HP to fix the other issues on the target, get back to zero fails there, and on top of that clean status reapply the filebuf changes.
