PowerPC64 little-endian fixes: AltiVec tests
authorUlrich Weigand <ulrich.weigand@de.ibm.com>
Tue, 4 Feb 2014 17:28:24 +0000 (18:28 +0100)
committerUlrich Weigand <ulrich.weigand@de.ibm.com>
Tue, 4 Feb 2014 17:28:24 +0000 (18:28 +0100)
A couple of AltiVec tests fail spuriously on powerpc64le-linux, because
they compare against an incorrect pattern.  Note that those tests already
contain little-endian variants of the patterns, but those seem to have
bit-rotted a bit: when outputting a vector, GDB no longer omits trailing
zero elements (as it used to do in the past).

This patch updates the pattern to the new GDB output behavior.

In addition, the patch updates the endian test to use the new
gdb_test_multiple logic instead of gdb_expect.

gdb/testsuite/ChangeLog:

* gdb.arch/altivec-regs.exp: Use gdb_test_multiple for endian test.
(decimal_vector): Fix for little-endian.

gdb/testsuite/ChangeLog
gdb/testsuite/gdb.arch/altivec-regs.exp

index e2c1bc8..9d886c5 100644 (file)
@@ -1,3 +1,8 @@
+2014-02-04  Ulrich Weigand  <uweigand@de.ibm.com>
+
+       * gdb.arch/altivec-regs.exp: Use gdb_test_multiple for endian test.
+       (decimal_vector): Fix for little-endian.
+
 2014-01-29  Jose E. Marchesi  <jose.marchesi@oracle.com>
 
        * gdb.arch/sparc-sysstep.exp: New file.
index 7d5049c..33c34b5 100644 (file)
@@ -79,17 +79,16 @@ gdb_test "set \$vrsave = 1" "" ""
 
 gdb_test "next" "" ""
 
-send_gdb "show endian\n"
 set endianness ""
-gdb_expect {
+set msg "detect endianness"
+gdb_test_multiple "show endian" "$msg" {
     -re "(The target endianness is set automatically .currently )(big|little)( endian.*)$gdb_prompt $" {
-        pass "endianness"
-       set endianness $expect_out(2,string)
+        pass "$msg"
+        set endianness $expect_out(2,string)
     }
     -re ".*$gdb_prompt $" {
-       fail "couldn't get endianness"
+        fail "$msg"
     }
-    timeout            { fail "(timeout) endianness" }
 }
 
 # And then read the AltiVec registers back, to see that
@@ -118,7 +117,7 @@ gdb_test "info reg vscr" "vscr.*0x1\t1" "info reg vscr"
 if {$endianness == "big"} {
      set decimal_vector ".uint128 = 0x00000001000000010000000100000001, v4_float = .1.*e-45, 1.*e-45, 1.*e-45, 1.*e-45., v4_int32 = .1, 1, 1, 1., v8_int16 = .0, 1, 0, 1, 0, 1, 0, 1., v16_int8 = .0, 0, 0, 1, 0, 0, 0, 1, 0, 0, 0, 1, 0, 0, 0, 1.."
 } else {
-     set decimal_vector ".uint128 = 0x00000001000000010000000100000001, v4_float = .1.*e-45, 1.*e-45, 1.*e-45, 1.*e-45., v4_int32 = .1, 1, 1, 1., v8_int16 = .1, 0, 1, 0, 1, 0, 1, 0., v16_int8 = .1, 0, 0, 0, 1, 0, 0, 0, 1, 0, 0, 0, 1, 0, 0.."
+     set decimal_vector ".uint128 = 0x00000001000000010000000100000001, v4_float = .1.*e-45, 1.*e-45, 1.*e-45, 1.*e-45., v4_int32 = .1, 1, 1, 1., v8_int16 = .1, 0, 1, 0, 1, 0, 1, 0., v16_int8 = .1, 0, 0, 0, 1, 0, 0, 0, 1, 0, 0, 0, 1, 0, 0, 0.."
 }
 
 for {set i 0} {$i < 32} {incr i 1} {