Revert "Implementation of SSE optimized Fletcher-4"

include/zfs_fletcher.h

@@ -61,14 +61,6 @@ typedef struct fletcher_4_func {
 	const char *name;
 } fletcher_4_ops_t;
 
-#if defined(HAVE_SSE2)
-extern const fletcher_4_ops_t fletcher_4_sse2_ops;
-#endif
-
-#if defined(HAVE_SSE2) && defined(HAVE_SSSE3)
-extern const fletcher_4_ops_t fletcher_4_ssse3_ops;
-#endif
-
 #if defined(HAVE_AVX) && defined(HAVE_AVX2)
 extern const fletcher_4_ops_t fletcher_4_avx2_ops;
 #endif

lib/libzpool/Makefile.am

@@ -23,7 +23,6 @@ KERNEL_C = \
 	zfs_deleg.c \
 	zfs_fletcher.c \
 	zfs_fletcher_intel.c \
-	zfs_fletcher_sse.c \
 	zfs_namecheck.c \
 	zfs_prop.c \
 	zfs_uio.c \

man/man5/zfs-module-parameters.5

@@ -838,15 +838,11 @@ Default value: \fB67,108,864\fR.
 .RS 12n
 Select a fletcher 4 implementation.
 .sp
-Supported selectors are: \fBfastest\fR, \fBscalar\fR, \fBsse2\fR, \fBssse3\fR,
-and \fBavx2\fR. All of the selectors except \fBfastest\fR and \fBscalar\fR
-require instruction set extensions to be available and will only appear if ZFS
-detects that they are present at runtime. If multiple implementations of
-fletcher 4 are available, the \fBfastest\fR will be chosen using a micro
-benchmark. Selecting \fBscalar\fR results in the original CPU based calculation
-being used. Selecting any option other than \fBfastest\fR and \fBscalar\fR
-results in vector instructions from the respective CPU instruction set being
-used.
+Supported selectors are: \fBfastest\fR, \fBscalar\fR, and \fBavx2\fR when
+AVX2 is supported by the processor.  If multiple implementations of fletcher 4
+are available the \fBfastest\fR will be chosen using a micro benchmark.
+Selecting \fBscalar\fR results in the original CPU based calculation being
+used, \fBavx2\fR uses the AVX2 vector instructions to compute a fletcher 4.
 .sp
 Default value: \fBfastest\fR.
 .RE

module/zcommon/Makefile.in

@@ -17,4 +17,3 @@ $(MODULE)-objs += zfs_uio.o
 $(MODULE)-objs += zpool_prop.o
 
 $(MODULE)-$(CONFIG_X86) += zfs_fletcher_intel.o
-$(MODULE)-$(CONFIG_X86) += zfs_fletcher_sse.o

module/zcommon/zfs_fletcher.c

@@ -149,12 +149,6 @@ static const fletcher_4_ops_t fletcher_4_scalar_ops = {
 
 static const fletcher_4_ops_t *fletcher_4_algos[] = {
 	&fletcher_4_scalar_ops,
-#if defined(HAVE_SSE2)
-	&fletcher_4_sse2_ops,
-#endif
-#if defined(HAVE_SSE2) && defined(HAVE_SSSE3)
-	&fletcher_4_ssse3_ops,
-#endif
 #if defined(HAVE_AVX) && defined(HAVE_AVX2)
 	&fletcher_4_avx2_ops,
 #endif
@@ -163,12 +157,6 @@ static const fletcher_4_ops_t *fletcher_4_algos[] = {
 static enum fletcher_selector {
 	FLETCHER_FASTEST = 0,
 	FLETCHER_SCALAR,
-#if defined(HAVE_SSE2)
-	FLETCHER_SSE2,
-#endif
-#if defined(HAVE_SSE2) && defined(HAVE_SSSE3)
-	FLETCHER_SSSE3,
-#endif
 #if defined(HAVE_AVX) && defined(HAVE_AVX2)
 	FLETCHER_AVX2,
 #endif
@@ -181,12 +169,6 @@ static struct fletcher_4_impl_selector {
 } fletcher_4_impl_selectors[] = {
 	[ FLETCHER_FASTEST ]	= { "fastest", NULL },
 	[ FLETCHER_SCALAR ]	= { "scalar", &fletcher_4_scalar_ops },
-#if defined(HAVE_SSE2)
-	[ FLETCHER_SSE2 ]	= { "sse2", &fletcher_4_sse2_ops },
-#endif
-#if defined(HAVE_SSE2) && defined(HAVE_SSSE3)
-	[ FLETCHER_SSSE3 ]	= { "ssse3", &fletcher_4_ssse3_ops },
-#endif
 #if defined(HAVE_AVX) && defined(HAVE_AVX2)
 	[ FLETCHER_AVX2 ]	= { "avx2", &fletcher_4_avx2_ops },
 #endif
@@ -425,7 +407,6 @@ fletcher_4_init(void)
 		ops->init(&zc);
 		do {
 			ops->compute(databuf, data_size, &zc);
-			ops->compute_byteswap(databuf, data_size, &zc);
 			run_count++;
 		} while (gethrtime() < start + bench_ns);
 		if (ops->fini != NULL)