This is the mail archive of the
binutils@sources.redhat.com
mailing list for the binutils project.
Re: PATCH: Support '*' in pattern and fix a typo
On Fri, Oct 24, 2003 at 06:16:04PM +0200, Jakub Jelinek wrote:
> On Fri, Oct 24, 2003 at 08:39:09AM -0700, H. J. Lu wrote:
> > On Fri, Oct 24, 2003 at 08:43:17AM +0200, Jakub Jelinek wrote:
> > > On Fri, Oct 24, 2003 at 12:07:44AM -0700, H. J. Lu wrote:
> > > > Here is another update to fix a typo:
> > > >
> > > > +- if (fnmatch (expr->pattern, sym, 0) == 0)
> > > > ++ if (fnmatch (expr->pattern, s, 0) == 0)
> > > >
> > > > Otherwise, extern "C++" won't work.
> > >
> > > Oops, sorry. Unless you check the whole patch RSN, please check this
> > > in separately so that the tree is not broken for too long.
> > >
> >
> > There are 2 other typos. I will check in the following to fix it.
> >
> >
> > H.J.
> > ----
> > 2003-10-24 H.J. Lu <hongjiu.lu@intel.com>
> >
> > * ldlang.c (lang_vers_match): Check demangled symbols.
> >
> > --- ldlang.c.dem 2003-10-24 00:01:57.000000000 -0700
> > +++ ldlang.c 2003-10-24 08:34:10.000000000 -0700
> > @@ -4993,7 +4993,7 @@ lang_vers_match (struct bfd_elf_version_
> > if (!cxx_sym)
> > cxx_sym = sym;
> > }
> > - if (head->mask & BFD_ELF_VERSION_JAVA_TYPE)
> > + else if (head->mask & BFD_ELF_VERSION_JAVA_TYPE)
> > {
> > java_sym = cplus_demangle (sym, DMGL_JAVA);
> > if (!java_sym)
>
> This hunk doesn't look right.
> Although each symbol can be either C, or C++, or Java, not a combination of
> them, head->mask can have more bits set and it needs to set up
> both cxx_sym and java_sym if both C++ and Java symbols need to be checked.
Ooops. Fixed.
H.J.