sh64 build error - sh, kbuild or binutils problem?

sh64 build error - sh, kbuild or binutils problem?

[Adrian Bunk]

cayman_defconfig fails to build with the following error
(using binutils 2.18.50.0.7):

<--  snip  -->

...
  LD      drivers/macintosh/built-in.o
  LD      drivers/media/common/tuners/built-in.o
  CC [M]  drivers/media/common/tuners/tuner-xc2028.o
  CC [M]  drivers/media/common/tuners/tuner-simple.o
  CC [M]  drivers/media/common/tuners/tuner-types.o
  CC [M]  drivers/media/common/tuners/mt20xx.o
  CC [M]  drivers/media/common/tuners/tda8290.o
  CC [M]  drivers/media/common/tuners/tea5767.o
  CC [M]  drivers/media/common/tuners/tea5761.o
  CC [M]  drivers/media/common/tuners/tda9887.o
  CC [M]  drivers/media/common/tuners/xc5000.o
  LD      drivers/media/common/built-in.o
...
  LD      drivers/media/built-in.o
sh64-linux-ld: sh3 architecture of input file `drivers/media/common/built-in.o' is incompatible with sh5 outputmake[3]: *** [drivers/media/built-in.o] Error 1

<--  snip  -->


Minimal example demonstrating the problem:

$ sh64-linux-ar rcs tmp.o
$ sh64-linux-ld -EL -r -o tmp2.o tmp.o
$ sh64-linux-ld -EL -r -o out.o tmp2.o 
sh64-linux-ld: sh3 architecture of input file `tmp2.o' is incompatible with sh5 output
$ 


cu
Adrian

-- 

       "Is there not promise of rain?" Ling Tan asked suddenly out
        of the darkness. There had been need of rain for many days.
       "Only a promise," Lao Er said.
                                       Pearl S. Buck - Dragon Seed

Re: sh64 build error - sh, kbuild or binutils problem?

[Paul Mundt]

On Sat, May 31, 2008 at 11:22:39PM +0300, Adrian Bunk wrote:
> Minimal example demonstrating the problem:
> 
> $ sh64-linux-ar rcs tmp.o
> $ sh64-linux-ld -EL -r -o tmp2.o tmp.o
> $ sh64-linux-ld -EL -r -o out.o tmp2.o 
> sh64-linux-ld: sh3 architecture of input file `tmp2.o' is incompatible with sh5 output
> $ 
> 
Yes, it's a binutils bug. I've hit this on SH-2A zImage builds too, and
hacked together a patch for it. I'll see about fixing it up fo sh64 too
and then posting it to the list.

Re: sh64 build error - sh, kbuild or binutils problem?

[Adrian Bunk]

On Sun, Jun 01, 2008 at 02:30:47PM +0900, Paul Mundt wrote:
> On Sat, May 31, 2008 at 11:22:39PM +0300, Adrian Bunk wrote:
> > Minimal example demonstrating the problem:
> > 
> > $ sh64-linux-ar rcs tmp.o
> > $ sh64-linux-ld -EL -r -o tmp2.o tmp.o
> > $ sh64-linux-ld -EL -r -o out.o tmp2.o 
> > sh64-linux-ld: sh3 architecture of input file `tmp2.o' is incompatible with sh5 output
> > $ 
> > 
> Yes, it's a binutils bug. I've hit this on SH-2A zImage builds too, and
> hacked together a patch for it. I'll see about fixing it up fo sh64 too
> and then posting it to the list.

Thanks  :-)
Adrian

-- 

       "Is there not promise of rain?" Ling Tan asked suddenly out
        of the darkness. There had been need of rain for many days.
       "Only a promise," Lao Er said.
                                       Pearl S. Buck - Dragon Seed

Re: sh64 build error - sh, kbuild or binutils problem?

[Yoshinori Sato]

At Sun, 1 Jun 2008 14:30:47 +0900,
Paul Mundt wrote:
> 
> On Sat, May 31, 2008 at 11:22:39PM +0300, Adrian Bunk wrote:
> > Minimal example demonstrating the problem:
> > 
> > $ sh64-linux-ar rcs tmp.o
> > $ sh64-linux-ld -EL -r -o tmp2.o tmp.o
> > $ sh64-linux-ld -EL -r -o out.o tmp2.o 
> > sh64-linux-ld: sh3 architecture of input file `tmp2.o' is incompatible with sh5 output
> > $ 
> > 
> Yes, it's a binutils bug. I've hit this on SH-2A zImage builds too, and
> hacked together a patch for it. I'll see about fixing it up fo sh64 too
> and then posting it to the list.

I think ld is strict machine type check.

This is workaround.
SH-2A build success.

diff -uN ../org/arch/sh/boot/compressed/Makefile_32 arch/sh/boot/compressed/Makefile_32
--- org/arch/sh/boot/compressed/Makefile_32	2008-07-01 00:34:41.000000000 -0400
+++ arch/sh/boot/compressed/Makefile_32	2008-07-02 21:21:57.000000000 -0400
@@ -35,8 +35,7 @@
 $(obj)/vmlinux.bin.gz: $(obj)/vmlinux.bin FORCE
 	$(call if_changed,gzip)
 
-LDFLAGS_piggy.o := -r --format binary --oformat elf32-sh-linux -T
 OBJCOPYFLAGS += -R .empty_zero_page
 
-$(obj)/piggy.o: $(obj)/vmlinux.scr $(obj)/vmlinux.bin.gz FORCE
-	$(call if_changed,ld)
+$(obj)/piggy.o: $(obj)/piggy.S $(obj)/vmlinux.bin.gz FORCE
+	$(call if_changed,as_o_S)
diff -uN ../org/arch/sh/boot/compressed/piggy.S arch/sh/boot/compressed/piggy.S
--- ../org/arch/sh/boot/compressed/piggy.S	1969-12-31 19:00:00.000000000 -0500
+++ arch/sh/boot/compressed/piggy.S	2008-07-02 21:15:05.000000000 -0400
@@ -0,0 +1,8 @@
+	.global	input_len, input_data
+	.data
+input_len:
+	.long	input_data_end - input_data
+input_data:
+	.incbin	"arch/sh/boot/compressed/vmlinux.bin.gz"
+input_data_end:
+	.end

> --
> To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html
> Please read the FAQ at  http://www.tux.org/lkml/

-- 
Yoshinori Sato
<ysato@users.sourceforge.jp>

Re: sh64 build error - sh, kbuild or binutils problem?

[Paul Mundt]

On Wed, Jul 02, 2008 at 09:45:27PM -0400, Yoshinori Sato wrote:
> At Sun, 1 Jun 2008 14:30:47 +0900,
> Paul Mundt wrote:
> > 
> > On Sat, May 31, 2008 at 11:22:39PM +0300, Adrian Bunk wrote:
> > > Minimal example demonstrating the problem:
> > > 
> > > $ sh64-linux-ar rcs tmp.o
> > > $ sh64-linux-ld -EL -r -o tmp2.o tmp.o
> > > $ sh64-linux-ld -EL -r -o out.o tmp2.o 
> > > sh64-linux-ld: sh3 architecture of input file `tmp2.o' is incompatible with sh5 output
> > > $ 
> > > 
> > Yes, it's a binutils bug. I've hit this on SH-2A zImage builds too, and
> > hacked together a patch for it. I'll see about fixing it up fo sh64 too
> > and then posting it to the list.
> 
> I think ld is strict machine type check.
> 
> This is workaround.
> SH-2A build success.
> 
This is probably a reasonable workaround until fixed versions of binutils
are more readily available. Please re-send this with a Signed-off-by and
I'll add it to my queue.

Re: sh64 build error - sh, kbuild or binutils problem?

[Yoshinori Sato]

At Thu, 3 Jul 2008 19:03:23 +0900,
Paul Mundt wrote:
> 
> On Wed, Jul 02, 2008 at 09:45:27PM -0400, Yoshinori Sato wrote:
> > At Sun, 1 Jun 2008 14:30:47 +0900,
> > Paul Mundt wrote:
> > > 
> > > On Sat, May 31, 2008 at 11:22:39PM +0300, Adrian Bunk wrote:
> > > > Minimal example demonstrating the problem:
> > > > 
> > > > $ sh64-linux-ar rcs tmp.o
> > > > $ sh64-linux-ld -EL -r -o tmp2.o tmp.o
> > > > $ sh64-linux-ld -EL -r -o out.o tmp2.o 
> > > > sh64-linux-ld: sh3 architecture of input file `tmp2.o' is incompatible with sh5 output
> > > > $ 
> > > > 
> > > Yes, it's a binutils bug. I've hit this on SH-2A zImage builds too, and
> > > hacked together a patch for it. I'll see about fixing it up fo sh64 too
> > > and then posting it to the list.
> > 
> > I think ld is strict machine type check.
> > 
> > This is workaround.
> > SH-2A build success.
> > 
> This is probably a reasonable workaround until fixed versions of binutils
> are more readily available. Please re-send this with a Signed-off-by and
> I'll add it to my queue.

Signed-off-by: Yoshinori Sato <ysato@users.sourceforge.jp>

diff --git a/arch/sh/boot/compressed/Makefile_32 b/arch/sh/boot/compressed/Makefile_32
index c0d25fb..47685f6 100644
--- a/arch/sh/boot/compressed/Makefile_32
+++ b/arch/sh/boot/compressed/Makefile_32
@@ -35,8 +35,7 @@ $(obj)/vmlinux.bin: vmlinux FORCE
 $(obj)/vmlinux.bin.gz: $(obj)/vmlinux.bin FORCE
 	$(call if_changed,gzip)
 
-LDFLAGS_piggy.o := -r --format binary --oformat elf32-sh-linux -T
 OBJCOPYFLAGS += -R .empty_zero_page
 
-$(obj)/piggy.o: $(obj)/vmlinux.scr $(obj)/vmlinux.bin.gz FORCE
-	$(call if_changed,ld)
+$(obj)/piggy.o: $(obj)/piggy.S $(obj)/vmlinux.bin.gz FORCE
+	$(call if_changed,as_o_S)
diff --git a/arch/sh/boot/compressed/Makefile_64 b/arch/sh/boot/compressed/Makefile_64
index 912f3e2..658d4f9 100644
--- a/arch/sh/boot/compressed/Makefile_64
+++ b/arch/sh/boot/compressed/Makefile_64
@@ -37,8 +37,7 @@ $(obj)/vmlinux.bin: vmlinux FORCE
 $(obj)/vmlinux.bin.gz: $(obj)/vmlinux.bin FORCE
 	$(call if_changed,gzip)
 
-LDFLAGS_piggy.o := -r --format binary --oformat elf32-sh64-linux -T
 OBJCOPYFLAGS += -R .empty_zero_page
 
-$(obj)/piggy.o: $(obj)/vmlinux.scr $(obj)/vmlinux.bin.gz FORCE
-	$(call if_changed,ld)
+$(obj)/piggy.o: $(obj)/piggy.S $(obj)/vmlinux.bin.gz FORCE
+	$(call if_changed,as_o_S)
diff --git a/arch/sh/boot/compressed/piggy.S b/arch/sh/boot/compressed/piggy.S
new file mode 100644
index 0000000..5660719
--- /dev/null
+++ b/arch/sh/boot/compressed/piggy.S
@@ -0,0 +1,8 @@
+	.global	input_len, input_data
+	.data
+input_len:
+	.long	input_data_end - input_data
+input_data:
+	.incbin	"arch/sh/boot/compressed/vmlinux.bin.gz"
+input_data_end:
+	.end
diff --git a/arch/sh/boot/compressed/vmlinux.scr b/arch/sh/boot/compressed/vmlinux.scr
deleted file mode 100644
index 1ed9d79..0000000
--- a/arch/sh/boot/compressed/vmlinux.scr
+++ /dev/null
@@ -1,9 +0,0 @@
-SECTIONS
-{
-  .data : { 
-	input_len = .;
-	LONG(input_data_end - input_data) input_data = .; 
-	*(.data) 
-	input_data_end = .; 
-	}
-}

-- 
Yoshinori Sato
<ysato@users.sourceforge.jp>

Re: sh64 build error - sh, kbuild or binutils problem?

[Paul Mundt]

On Thu, Jul 03, 2008 at 08:24:52PM -0400, Yoshinori Sato wrote:
> At Thu, 3 Jul 2008 19:03:23 +0900,
> Paul Mundt wrote:
> > 
> > On Wed, Jul 02, 2008 at 09:45:27PM -0400, Yoshinori Sato wrote:
> > > At Sun, 1 Jun 2008 14:30:47 +0900,
> > > Paul Mundt wrote:
> > > > 
> > > > On Sat, May 31, 2008 at 11:22:39PM +0300, Adrian Bunk wrote:
> > > > > Minimal example demonstrating the problem:
> > > > > 
> > > > > $ sh64-linux-ar rcs tmp.o
> > > > > $ sh64-linux-ld -EL -r -o tmp2.o tmp.o
> > > > > $ sh64-linux-ld -EL -r -o out.o tmp2.o 
> > > > > sh64-linux-ld: sh3 architecture of input file `tmp2.o' is incompatible with sh5 output
> > > > > $ 
> > > > > 
> > > > Yes, it's a binutils bug. I've hit this on SH-2A zImage builds too, and
> > > > hacked together a patch for it. I'll see about fixing it up fo sh64 too
> > > > and then posting it to the list.
> > > 
> > > I think ld is strict machine type check.
> > > 
> > > This is workaround.
> > > SH-2A build success.
> > > 
> > This is probably a reasonable workaround until fixed versions of binutils
> > are more readily available. Please re-send this with a Signed-off-by and
> > I'll add it to my queue.
> 
> Signed-off-by: Yoshinori Sato <ysato@users.sourceforge.jp>
> 
Applied, thanks.