[PATCH] dm vdo: use kernel byteswapping routines instead of GCC ones

[PATCH] dm vdo: use kernel byteswapping routines instead of GCC ones

[Matthew Sakai]

From: Ken Raeburn <raeburn@redhat.com>

Reported-by: Guenter Roeck <linux@roeck-us.net>
Signed-off-by: Ken Raeburn <raeburn@redhat.com>
Signed-off-by: Matthew Sakai <msakai@redhat.com>
---
 drivers/md/dm-vdo/murmurhash3.c | 18 ++++--------------
 1 file changed, 4 insertions(+), 14 deletions(-)

diff --git a/drivers/md/dm-vdo/murmurhash3.c b/drivers/md/dm-vdo/murmurhash3.c
index 00c9b9c05001..124026cb676a 100644
--- a/drivers/md/dm-vdo/murmurhash3.c
+++ b/drivers/md/dm-vdo/murmurhash3.c
@@ -8,6 +8,8 @@
 
 #include "murmurhash3.h"
 
+#include <asm/byteorder.h>
+
 static inline u64 rotl64(u64 x, s8 r)
 {
 	return (x << r) | (x >> (64 - r));
@@ -16,24 +18,12 @@ static inline u64 rotl64(u64 x, s8 r)
 #define ROTL64(x, y) rotl64(x, y)
 static __always_inline u64 getblock64(const u64 *p, int i)
 {
-#if __BYTE_ORDER__ == __ORDER_LITTLE_ENDIAN__
-	return p[i];
-#elif __BYTE_ORDER__ == __ORDER_BIG_ENDIAN__
-	return __builtin_bswap64(p[i]);
-#else
-#error "can't figure out byte order"
-#endif
+	return le64_to_cpup(&p[i]);
 }
 
 static __always_inline void putblock64(u64 *p, int i, u64 value)
 {
-#if __BYTE_ORDER__ == __ORDER_LITTLE_ENDIAN__
-	p[i] = value;
-#elif __BYTE_ORDER__ == __ORDER_BIG_ENDIAN__
-	p[i] = __builtin_bswap64(value);
-#else
-#error "can't figure out byte order"
-#endif
+	p[i] = cpu_to_le64(value);
 }
 
 /* Finalization mix - force all bits of a hash block to avalanche */
-- 
2.42.0

Re: [PATCH] dm vdo: use kernel byteswapping routines instead of GCC ones

[Guenter Roeck]

On Wed, Mar 20, 2024 at 05:44:05PM -0400, Matthew Sakai wrote:
> From: Ken Raeburn <raeburn@redhat.com>
> 
> Reported-by: Guenter Roeck <linux@roeck-us.net>
> Signed-off-by: Ken Raeburn <raeburn@redhat.com>
> Signed-off-by: Matthew Sakai <msakai@redhat.com>

Tested-by: Guenter Roeck <linux@roeck-us.net>

> ---
>  drivers/md/dm-vdo/murmurhash3.c | 18 ++++--------------
>  1 file changed, 4 insertions(+), 14 deletions(-)
> 
> diff --git a/drivers/md/dm-vdo/murmurhash3.c b/drivers/md/dm-vdo/murmurhash3.c
> index 00c9b9c05001..124026cb676a 100644
> --- a/drivers/md/dm-vdo/murmurhash3.c
> +++ b/drivers/md/dm-vdo/murmurhash3.c
> @@ -8,6 +8,8 @@
>  
>  #include "murmurhash3.h"
>  
> +#include <asm/byteorder.h>
> +
>  static inline u64 rotl64(u64 x, s8 r)
>  {
>  	return (x << r) | (x >> (64 - r));
> @@ -16,24 +18,12 @@ static inline u64 rotl64(u64 x, s8 r)
>  #define ROTL64(x, y) rotl64(x, y)
>  static __always_inline u64 getblock64(const u64 *p, int i)
>  {
> -#if __BYTE_ORDER__ == __ORDER_LITTLE_ENDIAN__
> -	return p[i];
> -#elif __BYTE_ORDER__ == __ORDER_BIG_ENDIAN__
> -	return __builtin_bswap64(p[i]);
> -#else
> -#error "can't figure out byte order"
> -#endif
> +	return le64_to_cpup(&p[i]);
>  }
>  
>  static __always_inline void putblock64(u64 *p, int i, u64 value)
>  {
> -#if __BYTE_ORDER__ == __ORDER_LITTLE_ENDIAN__
> -	p[i] = value;
> -#elif __BYTE_ORDER__ == __ORDER_BIG_ENDIAN__
> -	p[i] = __builtin_bswap64(value);
> -#else
> -#error "can't figure out byte order"
> -#endif
> +	p[i] = cpu_to_le64(value);
>  }
>  
>  /* Finalization mix - force all bits of a hash block to avalanche */
> -- 
> 2.42.0
> 

Re: [PATCH] dm vdo: use kernel byteswapping routines instead of GCC ones

[Eric Biggers]

On Wed, Mar 20, 2024 at 05:44:05PM -0400, Matthew Sakai wrote:
>  static __always_inline u64 getblock64(const u64 *p, int i)
>  {
> -#if __BYTE_ORDER__ == __ORDER_LITTLE_ENDIAN__
> -	return p[i];
> -#elif __BYTE_ORDER__ == __ORDER_BIG_ENDIAN__
> -	return __builtin_bswap64(p[i]);
> -#else
> -#error "can't figure out byte order"
> -#endif
> +	return le64_to_cpup(&p[i]);
>  }
>  
>  static __always_inline void putblock64(u64 *p, int i, u64 value)
>  {
> -#if __BYTE_ORDER__ == __ORDER_LITTLE_ENDIAN__
> -	p[i] = value;
> -#elif __BYTE_ORDER__ == __ORDER_BIG_ENDIAN__
> -	p[i] = __builtin_bswap64(value);
> -#else
> -#error "can't figure out byte order"
> -#endif
> +	p[i] = cpu_to_le64(value);
>  }

This is very broken.  What you're actually looking for is get_unaligned_le64()
and put_unaligned_le64().  And they should be folded directly into the caller.

- Eric

Re: [PATCH] dm vdo: use kernel byteswapping routines instead of GCC ones

[Matthew Sakai]

On 3/20/24 23:52, Eric Biggers wrote:
> On Wed, Mar 20, 2024 at 05:44:05PM -0400, Matthew Sakai wrote:
>>   static __always_inline u64 getblock64(const u64 *p, int i)
>>   {
>> -#if __BYTE_ORDER__ == __ORDER_LITTLE_ENDIAN__
>> -	return p[i];
>> -#elif __BYTE_ORDER__ == __ORDER_BIG_ENDIAN__
>> -	return __builtin_bswap64(p[i]);
>> -#else
>> -#error "can't figure out byte order"
>> -#endif
>> +	return le64_to_cpup(&p[i]);
>>   }
>>   
>>   static __always_inline void putblock64(u64 *p, int i, u64 value)
>>   {
>> -#if __BYTE_ORDER__ == __ORDER_LITTLE_ENDIAN__
>> -	p[i] = value;
>> -#elif __BYTE_ORDER__ == __ORDER_BIG_ENDIAN__
>> -	p[i] = __builtin_bswap64(value);
>> -#else
>> -#error "can't figure out byte order"
>> -#endif
>> +	p[i] = cpu_to_le64(value);
>>   }
> 
> This is very broken.  What you're actually looking for is get_unaligned_le64()
> and put_unaligned_le64().  And they should be folded directly into the caller.
> 
> - Eric
> 

Thanks for the suggestion. We'll work on a v2 to include this.

Matt