Fixed: the new test ld-elf/pr14156c fails for arm-eabi

Fixed: the new test ld-elf/pr14156c fails for arm-eabi

[Hans-Peter Nilsson]

I saw for my autotester, ever since this test was added:

Running /tmp/hpautotest-binutils/bsrc/src/ld/testsuite/ld-elf/elf.exp ...
FAIL: ld-elf/pr14156c

and in the ld.log:
/tmp/hpautotest-binutils/arm-unknown-eabi/ld/../gas/as-new    -o tmpdir/dump0.o /tmp/hpautotest-binutils/bsrc/src/ld/tes
tsuite/ld-elf/foo0.s
Executing on host: sh -c {/tmp/hpautotest-binutils/arm-unknown-eabi/ld/../gas/as-new    -o tmpdir/dump0.o /tmp/hpautotes
t-binutils/bsrc/src/ld/testsuite/ld-elf/foo0.s 2>&1}  /dev/null ld.tmp (timeout = 300)
/tmp/hpautotest-binutils/bsrc/src/ld/testsuite/ld-elf/foo0.s: Assembler messages:
/tmp/hpautotest-binutils/bsrc/src/ld/testsuite/ld-elf/foo0.s:12: Error: junk at end of line, first unrecognized characte
r is `,'
/tmp/hpautotest-binutils/bsrc/src/ld/testsuite/ld-elf/foo0.s: Assembler messages:
/tmp/hpautotest-binutils/bsrc/src/ld/testsuite/ld-elf/foo0.s:12: Error: junk at end of line, first unrecognized characte
r is `,'

failed with: </tmp/hpautotest-binutils/bsrc/src/ld/testsuite/ld-elf/foo0.s: Assembler messages:
/tmp/hpautotest-binutils/bsrc/src/ld/testsuite/ld-elf/foo0.s:12: Error: junk at end of line, first unrecognized characte
r is `,'
>
FAIL: ld-elf/pr14156c

This was rather easily fixable, so I committed the following as
obvious (after a sanity check for cris-elf too).  I'll suggest
that people please use the list of targets posted somewhat
recently as a sanity check when adding new tests, and besides,
it's rather known that ARM has an odd .section syntax, and that
there is a compatible syntax different from what e.g. GCC emits
for most targets.  The only other .s files in ld-elf with the
@-syntax are assembled just for x86_64.

ld/testsuite:
	* ld-elf/foo0.s: Use ARM-compatible .section syntax.
	* ld-elf/foo1.s: Likewise.
	* ld-elf/foo2.s: Likewise.
	* ld-elf/foo3.s: Likewise.
	* ld-elf/foon.s: Likewise.

Index: ld-elf/foo0.s
===================================================================
RCS file: /cvs/src/src/ld/testsuite/ld-elf/foo0.s,v
retrieving revision 1.1
diff -p -u -r1.1 foo0.s
--- ld-elf/foo0.s	10 Jul 2012 06:50:57 -0000	1.1
+++ ld-elf/foo0.s	13 Jul 2012 21:36:47 -0000
@@ -9,7 +9,7 @@ __start:
 main:
 	.dc.a 0
 
-	.section .foo,"ax",@progbits
+	.section .foo,"ax",%progbits
 	.p2align 2
 .globl foo
 	.type	foo,%function
Index: ld-elf/foo1.s
===================================================================
RCS file: /cvs/src/src/ld/testsuite/ld-elf/foo1.s,v
retrieving revision 1.1
diff -p -u -r1.1 foo1.s
--- ld-elf/foo1.s	10 Jul 2012 06:50:57 -0000	1.1
+++ ld-elf/foo1.s	13 Jul 2012 21:36:47 -0000
@@ -1,4 +1,4 @@
-	.section .foo,"ax",@progbits
+	.section .foo,"ax",%progbits
 	.p2align 2
 	.type	foo1,%function
 foo1:
Index: ld-elf/foo2.s
===================================================================
RCS file: /cvs/src/src/ld/testsuite/ld-elf/foo2.s,v
retrieving revision 1.1
diff -p -u -r1.1 foo2.s
--- ld-elf/foo2.s	10 Jul 2012 06:50:57 -0000	1.1
+++ ld-elf/foo2.s	13 Jul 2012 21:36:47 -0000
@@ -1,4 +1,4 @@
-	.section .foo,"ax",@progbits
+	.section .foo,"ax",%progbits
 	.p2align 6
 	.type	foo2,%function
 foo2:
Index: ld-elf/foo3.s
===================================================================
RCS file: /cvs/src/src/ld/testsuite/ld-elf/foo3.s,v
retrieving revision 1.1
diff -p -u -r1.1 foo3.s
--- ld-elf/foo3.s	10 Jul 2012 06:50:57 -0000	1.1
+++ ld-elf/foo3.s	13 Jul 2012 21:36:47 -0000
@@ -1,4 +1,4 @@
-	.section .foo,"ax",@progbits
+	.section .foo,"ax",%progbits
 	.p2align 4
 	.type	foo3,%function
 foo3:
Index: ld-elf/foon.s
===================================================================
RCS file: /cvs/src/src/ld/testsuite/ld-elf/foon.s,v
retrieving revision 1.1
diff -p -u -r1.1 foon.s
--- ld-elf/foon.s	10 Jul 2012 06:50:57 -0000	1.1
+++ ld-elf/foon.s	13 Jul 2012 21:36:47 -0000
@@ -1,4 +1,4 @@
-	.section .foo,"ax",@progbits
+	.section .foo,"ax",%progbits
 	.p2align 8
 	.type	last,%function
 last:

brgds, H-P

Re: Fixed: the new test ld-elf/pr14156c fails for arm-eabi

[Andrew Pinski]

On Fri, Jul 13, 2012 at 2:45 PM, Hans-Peter Nilsson
<hans-peter.nilsson@axis.com> wrote:
> I saw for my autotester, ever since this test was added:
>
> Running /tmp/hpautotest-binutils/bsrc/src/ld/testsuite/ld-elf/elf.exp ...
> FAIL: ld-elf/pr14156c
>
> and in the ld.log:
> /tmp/hpautotest-binutils/arm-unknown-eabi/ld/../gas/as-new    -o tmpdir/dump0.o /tmp/hpautotest-binutils/bsrc/src/ld/tes
> tsuite/ld-elf/foo0.s
> Executing on host: sh -c {/tmp/hpautotest-binutils/arm-unknown-eabi/ld/../gas/as-new    -o tmpdir/dump0.o /tmp/hpautotes
> t-binutils/bsrc/src/ld/testsuite/ld-elf/foo0.s 2>&1}  /dev/null ld.tmp (timeout = 300)
> /tmp/hpautotest-binutils/bsrc/src/ld/testsuite/ld-elf/foo0.s: Assembler messages:
> /tmp/hpautotest-binutils/bsrc/src/ld/testsuite/ld-elf/foo0.s:12: Error: junk at end of line, first unrecognized characte
> r is `,'
> /tmp/hpautotest-binutils/bsrc/src/ld/testsuite/ld-elf/foo0.s: Assembler messages:
> /tmp/hpautotest-binutils/bsrc/src/ld/testsuite/ld-elf/foo0.s:12: Error: junk at end of line, first unrecognized characte
> r is `,'
>
> failed with: </tmp/hpautotest-binutils/bsrc/src/ld/testsuite/ld-elf/foo0.s: Assembler messages:
> /tmp/hpautotest-binutils/bsrc/src/ld/testsuite/ld-elf/foo0.s:12: Error: junk at end of line, first unrecognized characte
> r is `,'
>>
> FAIL: ld-elf/pr14156c
>
> This was rather easily fixable, so I committed the following as
> obvious (after a sanity check for cris-elf too).  I'll suggest
> that people please use the list of targets posted somewhat
> recently as a sanity check when adding new tests, and besides,
> it's rather known that ARM has an odd .section syntax, and that
> there is a compatible syntax different from what e.g. GCC emits
> for most targets.  The only other .s files in ld-elf with the
> @-syntax are assembled just for x86_64.

It is not that ARM has an odd .section syntax but rather @ is the
comment character in ARM.

Thanks,
Andrew Pinski