2008-10-08 06:39:11

by Harvey Harrison

Subject: [PATCH 3/4] xtensa: Change unaligned access to use packed struct implementation

Signed-off-by: Harvey Harrison <[email protected]>
---
There are only 3 arches that use the memmove-based implementation, there
doesn't seem to be any obvious reason not to use the struct implementation.

Chris?

include/asm-xtensa/unaligned.h | 4 ++--
1 files changed, 2 insertions(+), 2 deletions(-)

diff --git a/include/asm-xtensa/unaligned.h b/include/asm-xtensa/unaligned.h
index 8f3424f..1a242b2 100644
--- a/include/asm-xtensa/unaligned.h
+++ b/include/asm-xtensa/unaligned.h
@@ -11,13 +11,13 @@
#define _ASM_XTENSA_UNALIGNED_H

#ifdef __XTENSA_EL__
-# include <linux/unaligned/le_memmove.h>
+# include <linux/unaligned/le_struct.h>
# include <linux/unaligned/be_byteshift.h>
# include <linux/unaligned/generic.h>
# define get_unaligned __get_unaligned_le
# define put_unaligned __put_unaligned_le
#elif defined(__XTENSA_EB__)
-# include <linux/unaligned/be_memmove.h>
+# include <linux/unaligned/be_struct.h>
# include <linux/unaligned/le_byteshift.h>
# include <linux/unaligned/generic.h>
# define get_unaligned __get_unaligned_be
--
1.6.0.2.471.g47a76

2008-10-10 05:26:21

by Chris Zankel

Subject: Re: [PATCH 3/4] xtensa: Change unaligned access to use packed struct implementation

Hi Harvey,

Please allow me a couple of days to look into it. I don't know the
historic reasons why the memmove variant was chosen.

Thanks,
-Chris


Harvey Harrison wrote:
> Signed-off-by: Harvey Harrison <[email protected]>
> ---
> There are only 3 arches that use the memmove-based implementation, there
> doesn't seem to be any obvious reason not to use the struct implementation.
>
> Chris?
>
> include/asm-xtensa/unaligned.h | 4 ++--
> 1 files changed, 2 insertions(+), 2 deletions(-)
>
> diff --git a/include/asm-xtensa/unaligned.h b/include/asm-xtensa/unaligned.h
> index 8f3424f..1a242b2 100644
> --- a/include/asm-xtensa/unaligned.h
> +++ b/include/asm-xtensa/unaligned.h
> @@ -11,13 +11,13 @@
> #define _ASM_XTENSA_UNALIGNED_H
>
> #ifdef __XTENSA_EL__
> -# include <linux/unaligned/le_memmove.h>
> +# include <linux/unaligned/le_struct.h>
> # include <linux/unaligned/be_byteshift.h>
> # include <linux/unaligned/generic.h>
> # define get_unaligned __get_unaligned_le
> # define put_unaligned __put_unaligned_le
> #elif defined(__XTENSA_EB__)
> -# include <linux/unaligned/be_memmove.h>
> +# include <linux/unaligned/be_struct.h>
> # include <linux/unaligned/le_byteshift.h>
> # include <linux/unaligned/generic.h>
> # define get_unaligned __get_unaligned_be

2008-10-10 05:39:59

by Harvey Harrison

Subject: Re: [PATCH 3/4] xtensa: Change unaligned access to use packed struct implementation

On Thu, 2008-10-09 at 22:16 -0700, Chris Zankel wrote:
> Hi Harvey,
>
> Please allow me a couple of days to look into it. I don't know the
> historic reasons why the memmove variant was chosen.
>
> Thanks,
> -Chris
>

Sure, from the comments in the original version, it looks like it was done
this way to avoid __builtin_memcpy from being generated. But maybe there was another
reason?

Harvey