Implementation of AVX2 optimized Fletcher-4

New functionality:
- Preserves existing scalar implementation.
- Adds AVX2 optimized Fletcher-4 computation.
- Fastest routines selected on module load (benchmark).
- Test case for Fletcher-4 added to ztest.

New zcommon module parameters:
-  zfs_fletcher_4_impl (str): selects the implementation to use.
    "fastest" - use the fastest version available
    "cycle"   - cycle trough all available impl for ztest
    "scalar"  - use the original version
    "avx2"    - new AVX2 implementation if available

Performance comparison (Intel i7 CPU, 1MB data buffers):
- Scalar:  4216 MB/s
- AVX2:   14499 MB/s

See contents of `/sys/module/zcommon/parameters/zfs_fletcher_4_impl`
to get list of supported values. If an implementation is not supported
on the system, it will not be shown. Currently selected option is
enclosed in `[]`.

Signed-off-by: Jinshan Xiong <[email protected]>
Signed-off-by: Andreas Dilger <[email protected]>
Signed-off-by: Brian Behlendorf <[email protected]>
Closes #4330

1 files changed, 1 insertions, 32 deletions

diff --git a/include/sys/spa.h b/include/sys/spa.h
index c80e8337e..7c09caf6c 100644
--- a/include/sys/spa.h
+++ b/include/sys/spa.h
@@ -35,6 +35,7 @@
 #include <sys/sysmacros.h>
 #include <sys/types.h>
 #include <sys/fs/zfs.h>
+#include <sys/spa_checksum.h>
 
 #ifdef	__cplusplus
 extern "C" {
@@ -142,12 +143,6 @@ typedef struct dva {
 	uint64_t	dva_word[2];
 } dva_t;
 
-/*
- * Each block has a 256-bit checksum -- strong enough for cryptographic hashes.
- */
-typedef struct zio_cksum {
-	uint64_t	zc_word[4];
-} zio_cksum_t;
 
 /*
  * Each block is described by its DVAs, time of birth, checksum, etc.
@@ -440,35 +435,9 @@ _NOTE(CONSTCOND) } while (0)
 	DVA_EQUAL(&(bp1)->blk_dva[1], &(bp2)->blk_dva[1]) &&	\
 	DVA_EQUAL(&(bp1)->blk_dva[2], &(bp2)->blk_dva[2]))
 
-#define	ZIO_CHECKSUM_EQUAL(zc1, zc2) \
-	(0 == (((zc1).zc_word[0] - (zc2).zc_word[0]) | \
-	((zc1).zc_word[1] - (zc2).zc_word[1]) | \
-	((zc1).zc_word[2] - (zc2).zc_word[2]) | \
-	((zc1).zc_word[3] - (zc2).zc_word[3])))
-
-#define	ZIO_CHECKSUM_IS_ZERO(zc) \
-	(0 == ((zc)->zc_word[0] | (zc)->zc_word[1] | \
-	(zc)->zc_word[2] | (zc)->zc_word[3]))
-
-#define	ZIO_CHECKSUM_BSWAP(zcp)					\
-{								\
-	(zcp)->zc_word[0] = BSWAP_64((zcp)->zc_word[0]);	\
-	(zcp)->zc_word[1] = BSWAP_64((zcp)->zc_word[1]);	\
-	(zcp)->zc_word[2] = BSWAP_64((zcp)->zc_word[2]);	\
-	(zcp)->zc_word[3] = BSWAP_64((zcp)->zc_word[3]);	\
-}
-
 
 #define	DVA_IS_VALID(dva)	(DVA_GET_ASIZE(dva) != 0)
 
-#define	ZIO_SET_CHECKSUM(zcp, w0, w1, w2, w3)	\
-{						\
-	(zcp)->zc_word[0] = w0;			\
-	(zcp)->zc_word[1] = w1;			\
-	(zcp)->zc_word[2] = w2;			\
-	(zcp)->zc_word[3] = w3;			\
-}
-
 #define	BP_IDENTITY(bp)		(ASSERT(!BP_IS_EMBEDDED(bp)), &(bp)->blk_dva[0])
 #define	BP_IS_GANG(bp)		\
 	(BP_IS_EMBEDDED(bp) ? B_FALSE : DVA_GET_GANG(BP_IDENTITY(bp)))