[PATCH 02/23] arm: use the new byteorder headers

[PATCH 02/23] arm: use the new byteorder headers

[Harvey Harrison]

Signed-off-by: Harvey Harrison <harvey.harrison@gmail.com>
---
 arch/arm/include/asm/byteorder.h |   25 +++++++++++--------------
 1 files changed, 11 insertions(+), 14 deletions(-)

diff --git a/arch/arm/include/asm/byteorder.h b/arch/arm/include/asm/byteorder.h
index 4fbfb22..d04a7a2 100644
--- a/arch/arm/include/asm/byteorder.h
+++ b/arch/arm/include/asm/byteorder.h
@@ -18,7 +18,15 @@
 #include <linux/compiler.h>
 #include <asm/types.h>
 
-static inline __attribute_const__ __u32 ___arch__swab32(__u32 x)
+#ifdef __ARMEB__
+# define __BIG_ENDIAN
+#else
+# define __LITTLE_ENDIAN
+#endif
+
+#define __SWAB_64_THRU_32__
+
+static inline __attribute_const__ __u32 __arch_swab32(__u32 x)
 {
 	__u32 t;
 
@@ -40,19 +48,8 @@ static inline __attribute_const__ __u32 ___arch__swab32(__u32 x)
 
 	return x;
 }
+#define __arch_swab32 __arch_swab32
 
-#define __arch__swab32(x) ___arch__swab32(x)
-
-#if !defined(__STRICT_ANSI__) || defined(__KERNEL__)
-#  define __BYTEORDER_HAS_U64__
-#  define __SWAB_64_THRU_32__
-#endif
-
-#ifdef __ARMEB__
-#include <linux/byteorder/big_endian.h>
-#else
-#include <linux/byteorder/little_endian.h>
-#endif
+#include <linux/byteorder.h>
 
 #endif
-
-- 
1.6.0.274.g8aacc

Re: [PATCH 02/23] arm: use the new byteorder headers

[Russell King]

On Mon, Aug 18, 2008 at 05:47:55PM -0700, Harvey Harrison wrote:
> Signed-off-by: Harvey Harrison <harvey.harrison@gmail.com>
> ---
>  arch/arm/include/asm/byteorder.h |   25 +++++++++++--------------
>  1 files changed, 11 insertions(+), 14 deletions(-)

You keep posting this, and it's the only bit I see, but... what about
the other bits that it depends on, which without them, it's impossible
to test this patch with?

-- 
Russell King
 Linux kernel    2.6 ARM Linux   - http://www.arm.linux.org.uk/
 maintainer of:

Re: [PATCH 02/23] arm: use the new byteorder headers

[Harvey Harrison]

On Tue, 2008-08-19 at 08:27 +0100, Russell King wrote:
> On Mon, Aug 18, 2008 at 05:47:55PM -0700, Harvey Harrison wrote:
> > Signed-off-by: Harvey Harrison <harvey.harrison@gmail.com>
> > ---
> >  arch/arm/include/asm/byteorder.h |   25 +++++++++++--------------
> >  1 files changed, 11 insertions(+), 14 deletions(-)
> 
> You keep posting this, and it's the only bit I see, but... what about
> the other bits that it depends on, which without them, it's impossible
> to test this patch with?
> 

include/linux/swab.h, include/linux/byteorder.h are in current mainline
which is all that these patches depend on.

Harvey

Re: [PATCH 02/23] arm: use the new byteorder headers

[David Woodhouse]

On Tue, 2008-08-19 at 12:03 -0700, Harvey Harrison wrote:
> include/linux/swab.h, include/linux/byteorder.h are in current
> mainline which is all that these patches depend on.

Er, it also depends on the patch which exports those headers to
userspace, which isn't in Linus' tree yet.

So merging this for ARM (between -rc4 and rc5!) has broken the exported
headers on ARM -- as noted by Khem.

-- 
David Woodhouse                            Open Source Technology Centre
David.Woodhouse@intel.com                              Intel Corporation

Re: [PATCH 02/23] arm: use the new byteorder headers

[Russell King]

On Fri, Sep 05, 2008 at 10:49:29AM +0100, David Woodhouse wrote:
> On Tue, 2008-08-19 at 12:03 -0700, Harvey Harrison wrote:
> > include/linux/swab.h, include/linux/byteorder.h are in current
> > mainline which is all that these patches depend on.
> 
> Er, it also depends on the patch which exports those headers to
> userspace, which isn't in Linus' tree yet.

How useful.  If those generic headers are already in Linus' tree, and
are required for architectures to convert, why hasn't the patch which
exports them to userspace already been submitted?

However, another question: why should userspace be using a kernel header
file for byteswapping?

-- 
Russell King
 Linux kernel    2.6 ARM Linux   - http://www.arm.linux.org.uk/
 maintainer of:

Re: [PATCH 02/23] arm: use the new byteorder headers

[David Woodhouse]

On Fri, 2008-09-05 at 11:25 +0100, Russell King wrote:
> On Fri, Sep 05, 2008 at 10:49:29AM +0100, David Woodhouse wrote:
> > On Tue, 2008-08-19 at 12:03 -0700, Harvey Harrison wrote:
> > > include/linux/swab.h, include/linux/byteorder.h are in current
> > > mainline which is all that these patches depend on.
> > 
> > Er, it also depends on the patch which exports those headers to
> > userspace, which isn't in Linus' tree yet.
> 
> How useful.  If those generic headers are already in Linus' tree, and
> are required for architectures to convert, why hasn't the patch which
> exports them to userspace already been submitted?

Because they weren't being used by anything exported to userspace yet.
The patch has been seen, but wasn't destined for 2.6.27...

> However, another question: why should userspace be using a kernel header
> file for byteswapping?

... mostly (on my part, at least) because of that question.

Given my druthers, I'd revert the ARM part of the patch for now, and we
can have a _serious_ think about whether we really need to export these
functions to userspace.

A quick survey shows that the following user-visible headers currently
use cpu_to_xxx() and xxx_to_cpu():

linux/romfs_fs.h
linux/bfs_fs.h
linux/ext2_fs.h
linux/reiserfs_fs.h
linux/msdos_fs.h
linux/ncp.h
linux/ncp_no.h

Of those, most of them use the form without underscores, which isn't
visible in userspace _anyway_ and would be broken even if the new header
is exported. I suspect that they should all just be dropped from the
list of exported headers.

-- 
David Woodhouse                            Open Source Technology Centre
David.Woodhouse@intel.com                              Intel Corporation

Re: [PATCH 02/23] arm: use the new byteorder headers

[Russell King]

On Fri, Sep 05, 2008 at 03:06:06PM +0100, David Woodhouse wrote:
> On Fri, 2008-09-05 at 11:25 +0100, Russell King wrote:
> > On Fri, Sep 05, 2008 at 10:49:29AM +0100, David Woodhouse wrote:
> > > On Tue, 2008-08-19 at 12:03 -0700, Harvey Harrison wrote:
> > > > include/linux/swab.h, include/linux/byteorder.h are in current
> > > > mainline which is all that these patches depend on.
> > > 
> > > Er, it also depends on the patch which exports those headers to
> > > userspace, which isn't in Linus' tree yet.
> > 
> > How useful.  If those generic headers are already in Linus' tree, and
> > are required for architectures to convert, why hasn't the patch which
> > exports them to userspace already been submitted?
> 
> Because they weren't being used by anything exported to userspace yet.
> The patch has been seen, but wasn't destined for 2.6.27...
>
> > However, another question: why should userspace be using a kernel header
> > file for byteswapping?
> 
> ... mostly (on my part, at least) because of that question.
> 
> Given my druthers, I'd revert the ARM part of the patch for now, and we
> can have a _serious_ think about whether we really need to export these
> functions to userspace.

I'll revert it, and if it's resubmitted, I'll ignore it until after the
next merge window - since my tree for the upcoming merge window currently
pre-dates the revert, so effectively continues to carry it until it's
merged.

I'm not sure what went wrong, other than being continuously bugged by
Harvey to apply his patch.  If his patches are inappropriate, surely
that should've been pointed out to Harvey before he got the first of
his series merged?

Maybe more of his patches need reverting if some have been applied?

-- 
Russell King
 Linux kernel    2.6 ARM Linux   - http://www.arm.linux.org.uk/
 maintainer of: