diff mbox

Add VxWorks options from specs to vxworks.opt

Message ID Pine.LNX.4.64.1012311641580.16354@digraph.polyomino.org.uk
State New
Headers show

Commit Message

Joseph Myers Dec. 31, 2010, 4:42 p.m. UTC 
In preparation for the driver only accepting options listed in .opt
files, and not other options that simply happen to match some spec,
this patch adds VxWorks options matched by specs in vxworks.h to
vxworks.opt.  None of the options are matched by specs in vxworksae.h
and so Condition settings are used as for the other options in
vxworks.opt.

Note that before this patch -Bdynamic and -Bstatic would have added
various search paths based on the strings "dynamic" and "static" in
addition to having the effects from specs passing them to the linker.
Now they will only have the effects from specs, without also being
handled as -B options; this seems in accord with the intended
semantics as I interpret them.

Tested building cc1 and xgcc for crosses to: arm-wrs-vxworks
i686-wrs-vxworks i686-wrs-vxworksae mips-wrs-vxworks
powerpc-wrs-vxworks powerpc-wrs-vxworksae sh-wrs-vxworks
sparc-wrs-vxworks.  OK to commit?

2010-12-31  Joseph Myers  <joseph@codesourcery.com>

	* config/vxworks.opt (Bdynamic, Bstatic, Xbind-lazy, Xbind-now,
	non-static): New Driver options.

Comments

Nathan Sidwell Jan. 4, 2011, 11:44 a.m. UTC | #1 
On 12/31/10 16:42, Joseph S. Myers wrote:
> In preparation for the driver only accepting options listed in .opt
> files, and not other options that simply happen to match some spec,
> this patch adds VxWorks options matched by specs in vxworks.h to
> vxworks.opt.  None of the options are matched by specs in vxworksae.h
> and so Condition settings are used as for the other options in
> vxworks.opt.
>
> Note that before this patch -Bdynamic and -Bstatic would have added
> various search paths based on the strings "dynamic" and "static" in
> addition to having the effects from specs passing them to the linker.
> Now they will only have the effects from specs, without also being
> handled as -B options; this seems in accord with the intended
> semantics as I interpret them.
>
> Tested building cc1 and xgcc for crosses to: arm-wrs-vxworks
> i686-wrs-vxworks i686-wrs-vxworksae mips-wrs-vxworks
> powerpc-wrs-vxworks powerpc-wrs-vxworksae sh-wrs-vxworks
> sparc-wrs-vxworks.  OK to commit?

Looks good to me. thanks

nathan
diff mbox

Patch

Index: gcc/config/vxworks.opt
===================================================================
--- gcc/config/vxworks.opt	(revision 168378)
+++ gcc/config/vxworks.opt	(working copy)
@@ -1,6 +1,6 @@ 
 ; Processor-independent options for VxWorks.
 ;
-; Copyright (C) 2005, 2007 Free Software Foundation, Inc.
+; Copyright (C) 2005, 2007, 2010 Free Software Foundation, Inc.
 ; Contributed by CodeSourcery, LLC.
 ;
 ; This file is part of GCC.
@@ -19,6 +19,18 @@ 
 ; along with GCC; see the file COPYING3.  If not see
 ; <http://www.gnu.org/licenses/>.
 
+Bdynamic
+Driver Condition(VXWORKS_KIND == VXWORKS_KIND_NORMAL)
+
+Bstatic
+Driver Condition(VXWORKS_KIND == VXWORKS_KIND_NORMAL)
+
+Xbind-lazy
+Driver Condition(VXWORKS_KIND == VXWORKS_KIND_NORMAL)
+
+Xbind-now
+Driver Condition(VXWORKS_KIND == VXWORKS_KIND_NORMAL)
+
 mrtp
 Target Report RejectNegative Mask(VXWORKS_RTP) Condition(VXWORKS_KIND == VXWORKS_KIND_NORMAL)
 Assume the VxWorks RTP environment
@@ -29,3 +41,6 @@  Assume the VxWorks RTP environment
 mvthreads
 Target RejectNegative Condition(VXWORKS_KIND == VXWORKS_KIND_AE)
 Assume the VxWorks vThreads environment
+
+non-static
+Driver Condition(VXWORKS_KIND == VXWORKS_KIND_NORMAL)