This is the mail archive of the
binutils@sourceware.org
mailing list for the binutils project.
Re: [PATCHx3] LD plugins: matters outstanding.
- From: Richard Henderson <rth at redhat dot com>
- To: Dave Korn <dave dot korn dot cygwin at gmail dot com>
- Cc: "binutils at sourceware dot org" <binutils at sourceware dot org>, Tristan Gingold <gingold at adacore dot com>
- Date: Fri, 15 Oct 2010 08:52:34 -0700
- Subject: Re: [PATCHx3] LD plugins: matters outstanding.
- References: <4CB879B7.8080507@gmail.com>
On 10/15/2010 08:56 AM, Dave Korn wrote:
> * ld-plugin/plugin.exp: Don't error out if there is no target compiler
> available, make tests UNSUPPORTED instead.
Ok.
> * plugin.c (add_input_file): Take copy of input string.
> (add_input_library): Likewise.
> (set_extra_library_path): Likewise.
Ok.
> * configure.in: If <dlfcn.h> can't be found, try for <Windows.h>
> * configure: Regenerate.
> * config.in: Likewise.
> * plugin.c [!HAVE_DLFCN_H && HAVE_WINDOWS_H] (dlopen): Provide
> trival LoadLibrary-based replacement for Windows systems.
> [!HAVE_DLFCN_H && HAVE_WINDOWS_H] (dlsym): Likewise trivial
> replacement based on GetProcAddress.
> [!HAVE_DLFCN_H && HAVE_WINDOWS_H] (dlsym): Likewise FreeLibrary.
> * sysdep.h: Don't infer presence of <dlfcn.h> from ENABLE_PLUGINS
> anymore, use its own guard.
I think it's slightly confusing that HAVE_WINDOWS_H is defined only if
HAVE_DLFCN_H isn't. I think it might be a tad better to also test
HAVE_DLFCN_H in the ifdef around the include.
Ok with that change.
r~