On Thu, Nov 08, 2012 at 12:29:37PM -0800, Jason Evans wrote: > On Nov 7, 2012, at 10:40 PM, Mike Hommey wrote: > > From: Mike Hommey <[email protected]> > > > > When statically linking jemalloc, it may be beneficial not to export its > > symbols if it makes sense, which allows the compiler and the linker to do > > some further optimizations. > > --- > > configure.ac | 7 +++++++ > > 1 file changed, 7 insertions(+) > > > > diff --git a/configure.ac b/configure.ac > > index 1c52439..8558961 100644 > > --- a/configure.ac > > +++ b/configure.ac > > @@ -471,6 +471,13 @@ for stem in ${public_syms}; do > > AC_DEFINE_UNQUOTED([${n}], [${m}]) > > done > > > > +AC_ARG_WITH([export], > > + [AS_HELP_STRING([--without-export], [disable exporting jemalloc public > > APIs])], > > + [if test "x$with_export" = "xno"; then > > + AC_DEFINE([JEMALLOC_EXPORT],[])] > > +fi] > > +) > > + > > dnl Do not mangle library-private APIs by default. > > AC_ARG_WITH([private_namespace], > > [AS_HELP_STRING([--with-private-namespace=<prefix>], [Prefix to prepend > > to all library-private APIs])], > > -- > > 1.7.10.4 > > This looks to me like it will cause redefined symbol warnings. Does it?
Why would it? Mike _______________________________________________ jemalloc-discuss mailing list [email protected] http://www.canonware.com/mailman/listinfo/jemalloc-discuss
