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

[Andrew Stubbs]

This is merged from lp:gcc-linaro/4.5 r99321. Actually, this version is a backport from FSF where it was committed by Tom de Vries.

[Ulrich Weigand (uweigand)]

OK.

review: Approve

[Linaro Toolchain Builder (cbuild)]

cbuild has taken a snapshot of this branch at r106738 and queued it for build.

The snapshot is available at:
 http://ex.seabright.co.nz/snapshots/gcc-linaro-4.6+bzr106738~ams-codesourcery~branch-cost-4.6.tar.xdelta3.xz

and will be built on the following builders:
 a9-builder i686 x86_64

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

cbuild-snapshot: gcc-linaro-4.6+bzr106738~ams-codesourcery~branch-cost-4.6
cbuild-ancestor: lp:gcc-linaro/4.6+bzr106737
cbuild-state: check

[Linaro Toolchain Builder (cbuild)]

cbuild successfully built this on x86_64-maverick-cbuild112-crucis-x86_64r1.

The build results are available at:
 http://ex.seabright.co.nz/build/gcc-linaro-4.6+bzr106738~ams-codesourcery~branch-cost-4.6/logs/x86_64-maverick-cbuild112-crucis-x86_64r1

The test suite results were unchanged compared to the branch point lp:gcc-linaro/4.6+bzr106737.

The full testsuite results are at:
 http://ex.seabright.co.nz/build/gcc-linaro-4.6+bzr106738~ams-codesourcery~branch-cost-4.6/logs/x86_64-maverick-cbuild112-crucis-x86_64r1/gcc-testsuite.txt

cbuild-checked: x86_64-maverick-cbuild112-crucis-x86_64r1

[Linaro Toolchain Builder (cbuild)]

cbuild successfully built this on i686-lucid-cbuild113-scorpius-i686r1.

The build results are available at:
 http://ex.seabright.co.nz/build/gcc-linaro-4.6+bzr106738~ams-codesourcery~branch-cost-4.6/logs/i686-lucid-cbuild113-scorpius-i686r1

The test suite results were unchanged compared to the branch point lp:gcc-linaro/4.6+bzr106737.

The full testsuite results are at:
 http://ex.seabright.co.nz/build/gcc-linaro-4.6+bzr106738~ams-codesourcery~branch-cost-4.6/logs/i686-lucid-cbuild113-scorpius-i686r1/gcc-testsuite.txt

cbuild-checked: i686-lucid-cbuild113-scorpius-i686r1

[Linaro Toolchain Builder (cbuild)]

cbuild successfully built this on armv7l-maverick-cbuild113-ursa4-cortexa9r1.

The build results are available at:
 http://ex.seabright.co.nz/build/gcc-linaro-4.6+bzr106738~ams-codesourcery~branch-cost-4.6/logs/armv7l-maverick-cbuild113-ursa4-cortexa9r1

The test suite results changed compared to the branch point lp:gcc-linaro/4.6+bzr106737:
 +WARNING: program timed out.

The full testsuite results are at:
 http://ex.seabright.co.nz/build/gcc-linaro-4.6+bzr106738~ams-codesourcery~branch-cost-4.6/logs/armv7l-maverick-cbuild113-ursa4-cortexa9r1/gcc-testsuite.txt

cbuild-checked: armv7l-maverick-cbuild113-ursa4-cortexa9r1

[Michael Hope (michaelh1)]

The difference on armv7l is a timeout in one of the limits tests and changed due to running the testsuite in parallel. Looks OK.

=== modified file 'ChangeLog.linaro'
--- ChangeLog.linaro	2011-04-21 13:51:06 +0000
+++ ChangeLog.linaro	2011-04-26 15:02:35 +0000
@@ -1,3 +1,14 @@
[phone]  Andrew Stubbs  <[e-mail]>
+
+	Backport from FSF:
+
+	2011-04-05  Tom de Vries  <[e-mail]>
+
+	PR target/43920
+	gcc/
+	* config/arm/arm.h (BRANCH_COST): Set to 1 for Thumb-2 when optimizing
+	for size.
+
 2011-04-21  Andrew Stubbs  <[e-mail]>
 
 	Backport from FSF:

=== modified file 'gcc/config/arm/arm.h'
--- gcc/config/arm/arm.h	2011-02-21 14:04:51 +0000
+++ gcc/config/arm/arm.h	2011-04-26 15:02:35 +0000
@@ -2039,7 +2039,8 @@
 /* Try to generate sequences that don't involve branches, we can then use
    conditional instructions */
 #define BRANCH_COST(speed_p, predictable_p) \
-  (TARGET_32BIT ? 4 : (optimize > 0 ? 2 : 0))
+  (TARGET_32BIT ? (TARGET_THUMB2 && !speed_p ? 1 : 4) \
+		: (optimize > 0 ? 2 : 0))
 

 /* Position Independent Code.  */
 /* We decide which register to use based on the compilation options and