The inline-data test checks the specific bytes laid down by the
assembler, and so relies on the endianness of the target. I could
change the expected results to be endian agnostic, however, I worried
that a bug in the assembler that gets the endianness wrong would then
slip through. Instead I add a new test for big-endian arc, and restrict
the existing test to little-endian arc.
gas/ChangeLog:
* testsuite/gas/arc/inline-data-1.d: Add target restriction.
* testsuite/gas/arc/inline-data-2.d: New file.
+2016-03-21 Andrew Burgess <andrew.burgess@embecosm.com>
+
+ * testsuite/gas/arc/inline-data-1.d: Add target restriction.
+ * testsuite/gas/arc/inline-data-2.d: New file.
+
2016-03-21 Nick Clifton <nickc@redhat.com>
* atof-generic.c: Replace use of alloca with call to xmalloc.
#as: -mcpu=arc700
+#target: arc-*
#objdump: -sj .text
.*: +file format .*arc.*
--- /dev/null
+#as: -mcpu=arc700
+#target: arceb-*
+#objdump: -sj .text
+#source: inline-data-1.s
+
+.*: +file format .*arc.*
+
+Contents of section .text:
+ [0-9a-f]+ aabbccdd eeff .*