Merge lp:~michaelh1/gcc-linaro/rearnshaw-vector-1 into lp:gcc-linaro/4.7

[Michael Hope]

Don't allow arbitrary constants into Neon vector compare.

Backport from mainline r185603:

gcc/
2012-03-20 Richard Earnshaw <email address hidden>

Don't force vget_lane returning a 64-bit result to transfer to core registers.

Backport from mainline r185603:

gcc/
2012-03-21 Richard Earnshaw <email address hidden>

[Michael Hope (michaelh1)]

cbuild has taken a snapshot of this branch at r114974 and queued it for build.

The diff against the ancestor r114972 is available at:
 http://builds.linaro.org/toolchain/snapshots/gcc-linaro-4.7+bzr114974~michaelh1~rearnshaw-vector-1.diff

and will be built on the following builders:
 a9-builder armv5-builder i686 x86_64

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

cbuild-snapshot: gcc-linaro-4.7+bzr114974~michaelh1~rearnshaw-vector-1
cbuild-ancestor: lp:gcc-linaro/4.7+bzr114972
cbuild-state: check

[Michael Hope (michaelh1)]

cbuild successfully built this on x86_64-natty-cbuild276-oort1-x86_64r1.

The build results are available at:
 http://ex.seabright.co.nz/build/gcc-linaro-4.7+bzr114974~michaelh1~rearnshaw-vector-1/logs/x86_64-natty-cbuild276-oort1-x86_64r1

The testsuite results are the same as the branch point lp:gcc-linaro/4.7+bzr114972

The full diff is at:
 http://ex.seabright.co.nz/build/gcc-linaro-4.7+bzr114974~michaelh1~rearnshaw-vector-1/logs/x86_64-natty-cbuild276-oort1-x86_64r1/testsuite-diff.txt

The full testsuite results are at:
 http://ex.seabright.co.nz/build/gcc-linaro-4.7+bzr114974~michaelh1~rearnshaw-vector-1/logs/x86_64-natty-cbuild276-oort1-x86_64r1/gcc-testsuite.txt

cbuild-checked: x86_64-natty-cbuild276-oort1-x86_64r1

[Michael Hope (michaelh1)]

cbuild successfully built this on i686-natty-cbuild276-oort2-i686r1.

The build results are available at:
 http://ex.seabright.co.nz/build/gcc-linaro-4.7+bzr114974~michaelh1~rearnshaw-vector-1/logs/i686-natty-cbuild276-oort2-i686r1

-PASS: abi/demangle/abi_examples/26.cc (test for excess errors)
-PASS: abi/demangle/abi_examples/26.cc execution test
+FAIL: abi/demangle/abi_examples/26.cc (test for excess errors)
+UNRESOLVED: abi/demangle/abi_examples/26.cc compilation failed to produce executable

The full diff is at:
 http://ex.seabright.co.nz/build/gcc-linaro-4.7+bzr114974~michaelh1~rearnshaw-vector-1/logs/i686-natty-cbuild276-oort2-i686r1/testsuite-diff.txt

The full testsuite results are at:
 http://ex.seabright.co.nz/build/gcc-linaro-4.7+bzr114974~michaelh1~rearnshaw-vector-1/logs/i686-natty-cbuild276-oort2-i686r1/gcc-testsuite.txt

cbuild-checked: i686-natty-cbuild276-oort2-i686r1

[Michael Hope (michaelh1)]

cbuild successfully built this on armv7l-natty-cbuild276-tcpanda04-armv5r2.

The build results are available at:
 http://ex.seabright.co.nz/build/gcc-linaro-4.7+bzr114974~michaelh1~rearnshaw-vector-1/logs/armv7l-natty-cbuild276-tcpanda04-armv5r2

-PASS: gcc.target/arm/neon/vgetQ_lanes64.c scan-assembler vmov[ \t]+[rR][0-9]+, [rR][0-9]+, [dD][0-9]+!?([ \t]+@[a-zA-Z0-9 ]+)?\n
+PASS: gcc.target/arm/neon/vgetQ_lanes64.c scan-assembler ((vmov)|(fmrrd))[ \t]+[rR][0-9]+, [rR][0-9]+, [dD][0-9]+!?([ \t]+@[a-zA-Z0-9 ]+)?\n
-PASS: gcc.target/arm/neon/vgetQ_laneu64.c scan-assembler vmov[ \t]+[rR][0-9]+, [rR][0-9]+, [dD][0-9]+!?([ \t]+@[a-zA-Z0-9 ]+)?\n
+PASS: gcc.target/arm/neon/vgetQ_laneu64.c scan-assembler ((vmov)|(fmrrd))[ \t]+[rR][0-9]+, [rR][0-9]+, [dD][0-9]+!?([ \t]+@[a-zA-Z0-9 ]+)?\n

The full diff is at:
 http://ex.seabright.co.nz/build/gcc-linaro-4.7+bzr114974~michaelh1~rearnshaw-vector-1/logs/armv7l-natty-cbuild276-tcpanda04-armv5r2/testsuite-diff.txt

The full testsuite results are at:
 http://ex.seabright.co.nz/build/gcc-linaro-4.7+bzr114974~michaelh1~rearnshaw-vector-1/logs/armv7l-natty-cbuild276-tcpanda04-armv5r2/gcc-testsuite.txt

cbuild-checked: armv7l-natty-cbuild276-tcpanda04-armv5r2

[Ulrich Weigand (uweigand)]

The failure on i686 is weird -- since the patch doesn't change anything at all for non-ARM platforms, we shouldn't see any differences. Can this be some external effect?

Otherwise the backport looks good to me.

review: Approve

[Michael Hope (michaelh1)]

cbuild successfully built this on i686-natty-cbuild278-oort2-i686r1.

The build results are available at:
 http://ex.seabright.co.nz/build/gcc-linaro-4.7+bzr114974~michaelh1~rearnshaw-vector-1/logs/i686-natty-cbuild278-oort2-i686r1

The testsuite results are the same as the branch point lp:gcc-linaro/4.7+bzr114972

The full diff is at:
 http://ex.seabright.co.nz/build/gcc-linaro-4.7+bzr114974~michaelh1~rearnshaw-vector-1/logs/i686-natty-cbuild278-oort2-i686r1/testsuite-diff.txt

The full testsuite results are at:
 http://ex.seabright.co.nz/build/gcc-linaro-4.7+bzr114974~michaelh1~rearnshaw-vector-1/logs/i686-natty-cbuild278-oort2-i686r1/gcc-testsuite.txt

cbuild-checked: i686-natty-cbuild278-oort2-i686r1

[Michael Hope (michaelh1)]

The i686 results seem to be a glitch due to that instance. A re-run (above) clears the fault and a similar random fault on the longlong-4.7b.

[Michael Hope (michaelh1)]

cbuild successfully built this on armv7l-natty-cbuild276-tcpanda03-cortexa9r1.

The build results are available at:
 http://ex.seabright.co.nz/build/gcc-linaro-4.7+bzr114974~michaelh1~rearnshaw-vector-1/logs/armv7l-natty-cbuild276-tcpanda03-cortexa9r1

-PASS: gcc.target/arm/neon/vgetQ_lanes64.c scan-assembler vmov[ \t]+[rR][0-9]+, [rR][0-9]+, [dD][0-9]+!?([ \t]+@[a-zA-Z0-9 ]+)?\n
+PASS: gcc.target/arm/neon/vgetQ_lanes64.c scan-assembler ((vmov)|(fmrrd))[ \t]+[rR][0-9]+, [rR][0-9]+, [dD][0-9]+!?([ \t]+@[a-zA-Z0-9 ]+)?\n
-PASS: gcc.target/arm/neon/vgetQ_laneu64.c scan-assembler vmov[ \t]+[rR][0-9]+, [rR][0-9]+, [dD][0-9]+!?([ \t]+@[a-zA-Z0-9 ]+)?\n
+PASS: gcc.target/arm/neon/vgetQ_laneu64.c scan-assembler ((vmov)|(fmrrd))[ \t]+[rR][0-9]+, [rR][0-9]+, [dD][0-9]+!?([ \t]+@[a-zA-Z0-9 ]+)?\n

The full diff is at:
 http://ex.seabright.co.nz/build/gcc-linaro-4.7+bzr114974~michaelh1~rearnshaw-vector-1/logs/armv7l-natty-cbuild276-tcpanda03-cortexa9r1/testsuite-diff.txt

The full testsuite results are at:
 http://ex.seabright.co.nz/build/gcc-linaro-4.7+bzr114974~michaelh1~rearnshaw-vector-1/logs/armv7l-natty-cbuild276-tcpanda03-cortexa9r1/gcc-testsuite.txt

cbuild-checked: armv7l-natty-cbuild276-tcpanda03-cortexa9r1