Merge lp:~ams-codesourcery/gcc-linaro/lp675347-4.6 into lp:gcc-linaro/4.6

[Andrew Stubbs]

Julian submitted this patch upstream here, but it has not yet been accepted:
  http://gcc.gnu.org/ml/gcc-patches/2010-11/msg02245.html

I've submitted this merge request so we don't lose the patch while we wait for review upstream.

[Andrew Stubbs (ams-codesourcery)]

I've updated this branch to the patch version posted upstream.

[Linaro Toolchain Builder (cbuild)]

cbuild has taken a snapshot of this branch at r106730 and queued it for
build.

The snapshot is available at:
 http://ex.seabright.co.nz/snapshots/

and named something like gcc-linaro-4.6+bzr106730~ams-codesourcery~lp675347-4.6.*

You can track the build queue at:
 http://ex.seabright.co.nz/helpers/scheduler

cbuild-snapshot: gcc-linaro-4.6+bzr106730~ams-codesourcery~lp675347-4.6
cbuild-state: check

[Linaro Toolchain Builder (cbuild)]

cbuild successfully built this on i686-lucid-cbuild93-scorpius-i686r1.

The build results are available at:
 http://ex.seabright.co.nz/build/gcc-linaro-4.6+bzr106730~ams-codesourcery~lp675347-4.6/logs/i686-lucid-cbuild93-scorpius-i686r1

cbuild-checked: i686-lucid-cbuild93-scorpius-i686r1

review: Approve

[Linaro Toolchain Builder (cbuild)]

cbuild successfully built this on x86_64-maverick-cbuild93-crucis-x86_64r1.

The build results are available at:
 http://ex.seabright.co.nz/build/gcc-linaro-4.6+bzr106730~ams-codesourcery~lp675347-4.6/logs/x86_64-maverick-cbuild93-crucis-x86_64r1

cbuild-checked: x86_64-maverick-cbuild93-crucis-x86_64r1

review: Approve

[Linaro Toolchain Builder (cbuild)]

cbuild successfully built this on armv7l-maverick-cbuild93-ursa1-cortexa9r1.

The build results are available at:
 http://ex.seabright.co.nz/build/gcc-linaro-4.6+bzr106730~ams-codesourcery~lp675347-4.6/logs/armv7l-maverick-cbuild93-ursa1-cortexa9r1

cbuild-checked: armv7l-maverick-cbuild93-ursa1-cortexa9r1

review: Approve

[Andrew Stubbs (ams-codesourcery)]

The test results all seem perfectly fine.

review: Approve

=== modified file 'ChangeLog.linaro'
--- ChangeLog.linaro	2011-03-22 10:59:10 +0000
+++ ChangeLog.linaro	2011-03-23 15:06:20 +0000
@@ -1,3 +1,13 @@
[phone]  Andrew Stubbs  <[e-mail]>
+
+	Backport from FSF:
+
+	2011-03-23  Julian Brown  <[e-mail]>
+
+	gcc/
+	* expr.c (expand_expr_real_1): Only use BLKmode for volatile
+	accesses which are not naturally aligned.
+
 2011-03-22  Andrew Stubbs  <[e-mail]>
 
 	Backport from FSF:

=== modified file 'gcc/expr.c'
--- gcc/expr.c	2011-03-09 20:49:00 +0000
+++ gcc/expr.c	2011-03-23 15:06:20 +0000
@@ -9147,8 +9147,11 @@
 		&& modifier != EXPAND_CONST_ADDRESS
 		&& modifier != EXPAND_INITIALIZER)
 	    /* If the field is volatile, we always want an aligned
-	       access.  */
-	    || (volatilep && flag_strict_volatile_bitfields > 0)
+	       access.  Only do this if the access is not already naturally
+	       aligned, otherwise "normal" (non-bitfield) volatile fields
+	       become non-addressable.  */
+	    || (volatilep && flag_strict_volatile_bitfields > 0
+		&& (bitpos % GET_MODE_ALIGNMENT (mode) != 0))
 	    /* If the field isn't aligned enough to fetch as a memref,
 	       fetch it as a bit field.  */
 	    || (mode1 != BLKmode