[google] Backport patch r175881 from gcc-4_6-branch to google/gcc-4_6

[carrot, 2011-07-13 07:12:53]

Hi

This patch fixes a testing error on arm backend. It has been tested on both
x86 and arm target with following commands.

make check-g++ RUNTESTFLAGS="--target_board=arm-sim/thumb/arch=armv7-a
dg.exp=anon-ns1.C"
make check-g++ RUNTESTFLAGS="dg.exp=anon-ns1.C"

thanks
Carrot

2011-07-13  Guozhi Wei  <carrot@google.com>

	Backport from gcc-4_6-branch

	2011-07-05  Jason Merrill  <jason@redhat.com>

		PR testsuite/49643
		* g++.dg/rtti/anon-ns1.C: Allow for null-termination.


 	Backport from gcc-4_6-branch
Index: ChangeLog
===================================================================
--- ChangeLog	(revision 176194)
+++ ChangeLog	(working copy)
@@ -1,3 +1,8 @@
+2011-07-05  Jason Merrill  <jason@redhat.com>
+
+	PR testsuite/49643
+	* g++.dg/rtti/anon-ns1.C: Allow for null-termination.
+
 2011-07-05  Rainer Orth  <ro@CeBiTec.Uni-Bielefeld.DE>
 
 	* ada/acats/run_acats (which): Extract last field from type -p,
Index: g++.dg/rtti/anon-ns1.C
===================================================================
--- g++.dg/rtti/anon-ns1.C	(revision 176194)
+++ g++.dg/rtti/anon-ns1.C	(working copy)
@@ -2,7 +2,7 @@
 // The typeinfo name for A should start with * so we compare
 // it by address rather than contents.
 
-// { dg-final { scan-assembler "\"\*N\[^\"\]+1AE\"" } }
+// { dg-final { scan-assembler "\"\*N\[^\"\]+1AE" } }
 
 namespace
 {

--
This patch is available for review at http://codereview.appspot.com/4695051

[Diego Novillo, 2011-07-13 11:37:38]

On Wed, Jul 13, 2011 at 03:12, Guozhi Wei <carrot@google.com> wrote:
> Hi
>
> This patch fixes a testing error on arm backend. It has been tested on both
> x86 and arm target with following commands.
>
> make check-g++ RUNTESTFLAGS="--target_board=arm-sim/thumb/arch=armv7-a
dg.exp=anon-ns1.C"
> make check-g++ RUNTESTFLAGS="dg.exp=anon-ns1.C"

Carrot, did you backport this patch with svnmerge.py?


Thanks.  Diego.

[carrot, 2011-07-14 02:39:21]

Hi Diego

The previous patch was done with svn merge.

This new version is done with svnmerge.py. Again tested with

make check-g++ RUNTESTFLAGS="--target_board=arm-sim/thumb/arch=armv7-a
dg.exp=anon-ns1.C"
make check-g++ RUNTESTFLAGS="dg.exp=anon-ns1.C"

BTW, there are some unexpected property changes after merge, I don't
how did they come out and how should I deal with them?

thanks
Carrot


2011-07-14   Guozhi Wei  <carrot@google.com>

	Backport r175881 from gcc-4_6-branch

	2011-07-05  Jason Merrill  <jason@redhat.com>

		PR testsuite/49643
		* g++.dg/rtti/anon-ns1.C: Allow for null-termination.


Property changes on: .
___________________________________________________________________
Modified: svnmerge-integrated
   - /branches/gcc-4_6-branch:1-175849
/branches/google/integration:1-170988,173923,173959
/branches/google/main:1-175816
   + /branches/gcc-4_6-branch:1-175849,175881
/branches/google/integration:1-170988,173923,173959
/branches/google/main:1-175816
Modified: svn:mergeinfo
   Merged /branches/gcc-4_6-branch:r175881


Property changes on: libjava/classpath
___________________________________________________________________
Modified: svn:mergeinfo
   Merged /branches/gcc-4_6-branch/libjava/classpath:r175881


Property changes on: gcc/testsuite/gcc.target/powerpc/ppc-round.c
___________________________________________________________________
Modified: svn:mergeinfo
   Merged
/branches/gcc-4_6-branch/gcc/testsuite/gcc.target/powerpc/ppc-round.c:r175881

Index: gcc/testsuite/g++.dg/rtti/anon-ns1.C
===================================================================
--- gcc/testsuite/g++.dg/rtti/anon-ns1.C	(revision 176259)
+++ gcc/testsuite/g++.dg/rtti/anon-ns1.C	(working copy)
@@ -2,7 +2,7 @@
 // The typeinfo name for A should start with * so we compare
 // it by address rather than contents.

-// { dg-final { scan-assembler "\"\*N\[^\"\]+1AE\"" } }
+// { dg-final { scan-assembler "\"\*N\[^\"\]+1AE" } }

 namespace
 {

Property changes on: gcc/config/rs6000/rs6000.c
___________________________________________________________________
Modified: svn:mergeinfo
   Merged /branches/gcc-4_6-branch/gcc/config/rs6000/rs6000.c:r175881


Property changes on: gcc/config/rs6000/rs6000.h
___________________________________________________________________
Modified: svn:mergeinfo
   Merged /branches/gcc-4_6-branch/gcc/config/rs6000/rs6000.h:r175881



On Wed, Jul 13, 2011 at 7:37 PM, Diego Novillo <dnovillo@google.com> wrote:
> On Wed, Jul 13, 2011 at 03:12, Guozhi Wei <carrot@google.com> wrote:
>> Hi
>>
>> This patch fixes a testing error on arm backend. It has been tested on both
>> x86 and arm target with following commands.
>>
>> make check-g++ RUNTESTFLAGS="--target_board=arm-sim/thumb/arch=armv7-a
dg.exp=anon-ns1.C"
>> make check-g++ RUNTESTFLAGS="dg.exp=anon-ns1.C"
>
> Carrot, did you backport this patch with svnmerge.py?
>
>
> Thanks.  Diego.
>

[Diego Novillo, 2011-07-14 19:48:40]

On Wed, Jul 13, 2011 at 22:39, Carrot Wei <carrot@google.com> wrote:
> Hi Diego
>
> The previous patch was done with svn merge.
>
> This new version is done with svnmerge.py. Again tested with

Great, thanks.  This simplifies regular merges, since svnmerge.py will
know that this rev was merged already.  I think svnmerge.py also reads
'svn merge' markers, but I'm not sure.

> make check-g++ RUNTESTFLAGS="--target_board=arm-sim/thumb/arch=armv7-a
> dg.exp=anon-ns1.C"
> make check-g++ RUNTESTFLAGS="dg.exp=anon-ns1.C"
>
> BTW, there are some unexpected property changes after merge, I don't
> how did they come out and how should I deal with them?

Don't worry about those.  The branch used to merge from multiple
sources that had different attributes for those files.  It will be
gone after the next merge I'm doing next week.

The backport is OK, of course.


Diego.