| Message ID | 20101129203628.GA5115@intel.com |
|---|---|
| State | New |
| Headers | show |
* H.J. Lu wrote on Mon, Nov 29, 2010 at 09:36:28PM CET: > exec-tool.in uses LT_RCU to check recursive calls on as/ld in a > gcc+binutils combined tree. However, it doesn't work with > -fuse-linker-plugin, which calls ld, which calls as. This patch > checks and exports LT_RCU_$id instead of LT_RCU. OK for trunk? I'm not sure if a build maintainer can approve this, but it looks good to me with one nit below. Thanks, Ralf > 2010-11-29 H.J. Lu <hongjiu.lu@intel.com> > > PR driver/46712 > * exec-tool.in (id): New. > Check and export LT_RCU_$id instead of LT_RCU. > --- a/gcc/exec-tool.in > +++ b/gcc/exec-tool.in > @@ -28,6 +28,7 @@ fast_install=@enable_fast_install@ > objdir=@objdir@ > > invoked=`basename "$0"` > +id=$invoked > case "$invoked" in > as) > original=$ORIGINAL_AS_FOR_TARGET > @@ -44,6 +45,7 @@ case "$invoked" in > fi > prog=ld-new$exeext > dir=ld > + id=ld > ;; > nm) > original=$ORIGINAL_NM_FOR_TARGET > @@ -68,9 +70,11 @@ case "$original" in > # libtool has not relinked ld-new yet, but we cannot just use the > # previous stage (because then the relinking would just never happen!). > # So we take extra care to use prev-ld/ld-new *on recursive calls*. > + eval LT_RCU="\${LT_RCU_$id}" > test x"$LT_RCU" = x"1" && exec $scriptdir/../prev-$dir/$prog ${1+"$@"} > > - LT_RCU=1; export LT_RCU > + eval LT_RCU_$id=1 > + eval export LT_RCU_$id FWIW, this last eval is not necessary. > $scriptdir/../$dir/$prog ${1+"$@"} > result=$? > exit $result
On 11/29/2010 11:00 PM, Ralf Wildenhues wrote: > I'm not sure if a build maintainer can approve this, but it looks good > to me with one nit below. It is part of the build scripts, and it's okay with the eval removed. Paolo
diff --git a/gcc/exec-tool.in b/gcc/exec-tool.in index 0ae8f2a..e6ae329 100644 --- a/gcc/exec-tool.in +++ b/gcc/exec-tool.in @@ -28,6 +28,7 @@ fast_install=@enable_fast_install@ objdir=@objdir@ invoked=`basename "$0"` +id=$invoked case "$invoked" in as) original=$ORIGINAL_AS_FOR_TARGET @@ -44,6 +45,7 @@ case "$invoked" in fi prog=ld-new$exeext dir=ld + id=ld ;; nm) original=$ORIGINAL_NM_FOR_TARGET @@ -68,9 +70,11 @@ case "$original" in # libtool has not relinked ld-new yet, but we cannot just use the # previous stage (because then the relinking would just never happen!). # So we take extra care to use prev-ld/ld-new *on recursive calls*. + eval LT_RCU="\${LT_RCU_$id}" test x"$LT_RCU" = x"1" && exec $scriptdir/../prev-$dir/$prog ${1+"$@"} - LT_RCU=1; export LT_RCU + eval LT_RCU_$id=1 + eval export LT_RCU_$id $scriptdir/../$dir/$prog ${1+"$@"} result=$? exit $result @@ -83,5 +87,3 @@ case "$original" in exec $original ${1+"$@"} ;; esac - -
exec-tool.in uses LT_RCU to check recursive calls on as/ld in a gcc+binutils combined tree. However, it doesn't work with -fuse-linker-plugin, which calls ld, which calls as. This patch checks and exports LT_RCU_$id instead of LT_RCU. OK for trunk? Thanks. H.J. --- 2010-11-29 H.J. Lu <hongjiu.lu@intel.com> PR driver/46712 * exec-tool.in (id): New. Check and export LT_RCU_$id instead of LT_RCU.