Adding upstream version 6.1.76.upstream/6.1.76upstream

Signed-off-by: Daniel Baumann <daniel.baumann@progress-linux.org>

26 files changed, 5030 insertions, 0 deletions

diff --git a/lib/raid6/.gitignore b/lib/raid6/.gitignore
new file mode 100644
index 000000000..6be57745a
--- /dev/null
+++ b/lib/raid6/.gitignore
@@ -0,0 +1,8 @@
+# SPDX-License-Identifier: GPL-2.0-only
+mktables
+altivec*.c
+int*.c
+tables.c
+neon?.c
+s390vx?.c
+vpermxor*.c
diff --git a/lib/raid6/Makefile b/lib/raid6/Makefile
new file mode 100644
index 000000000..45e176194
--- /dev/null
+++ b/lib/raid6/Makefile
@@ -0,0 +1,94 @@
+# SPDX-License-Identifier: GPL-2.0
+obj-$(CONFIG_RAID6_PQ)	+= raid6_pq.o
+
+raid6_pq-y	+= algos.o recov.o tables.o int1.o int2.o int4.o \
+		   int8.o int16.o int32.o
+
+raid6_pq-$(CONFIG_X86) += recov_ssse3.o recov_avx2.o mmx.o sse1.o sse2.o avx2.o avx512.o recov_avx512.o
+raid6_pq-$(CONFIG_ALTIVEC) += altivec1.o altivec2.o altivec4.o altivec8.o \
+                              vpermxor1.o vpermxor2.o vpermxor4.o vpermxor8.o
+raid6_pq-$(CONFIG_KERNEL_MODE_NEON) += neon.o neon1.o neon2.o neon4.o neon8.o recov_neon.o recov_neon_inner.o
+raid6_pq-$(CONFIG_S390) += s390vx8.o recov_s390xc.o
+
+hostprogs	+= mktables
+
+ifeq ($(CONFIG_ALTIVEC),y)
+altivec_flags := -maltivec $(call cc-option,-mabi=altivec)
+# Enable <altivec.h>
+altivec_flags += -isystem $(shell $(CC) -print-file-name=include)
+
+ifdef CONFIG_CC_IS_CLANG
+# clang ppc port does not yet support -maltivec when -msoft-float is
+# enabled. A future release of clang will resolve this
+# https://bugs.llvm.org/show_bug.cgi?id=31177
+CFLAGS_REMOVE_altivec1.o  += -msoft-float
+CFLAGS_REMOVE_altivec2.o  += -msoft-float
+CFLAGS_REMOVE_altivec4.o  += -msoft-float
+CFLAGS_REMOVE_altivec8.o  += -msoft-float
+CFLAGS_REMOVE_vpermxor1.o += -msoft-float
+CFLAGS_REMOVE_vpermxor2.o += -msoft-float
+CFLAGS_REMOVE_vpermxor4.o += -msoft-float
+CFLAGS_REMOVE_vpermxor8.o += -msoft-float
+endif
+endif
+
+# The GCC option -ffreestanding is required in order to compile code containing
+# ARM/NEON intrinsics in a non C99-compliant environment (such as the kernel)
+ifeq ($(CONFIG_KERNEL_MODE_NEON),y)
+NEON_FLAGS := -ffreestanding
+# Enable <arm_neon.h>
+NEON_FLAGS += -isystem $(shell $(CC) -print-file-name=include)
+ifeq ($(ARCH),arm)
+NEON_FLAGS += -march=armv7-a -mfloat-abi=softfp -mfpu=neon
+endif
+CFLAGS_recov_neon_inner.o += $(NEON_FLAGS)
+ifeq ($(ARCH),arm64)
+CFLAGS_REMOVE_recov_neon_inner.o += -mgeneral-regs-only
+CFLAGS_REMOVE_neon1.o += -mgeneral-regs-only
+CFLAGS_REMOVE_neon2.o += -mgeneral-regs-only
+CFLAGS_REMOVE_neon4.o += -mgeneral-regs-only
+CFLAGS_REMOVE_neon8.o += -mgeneral-regs-only
+endif
+endif
+
+quiet_cmd_unroll = UNROLL  $@
+      cmd_unroll = $(AWK) -v N=$* -f $(srctree)/$(src)/unroll.awk < $< > $@
+
+targets += int1.c int2.c int4.c int8.c int16.c int32.c
+$(obj)/int%.c: $(src)/int.uc $(src)/unroll.awk FORCE
+	$(call if_changed,unroll)
+
+CFLAGS_altivec1.o += $(altivec_flags)
+CFLAGS_altivec2.o += $(altivec_flags)
+CFLAGS_altivec4.o += $(altivec_flags)
+CFLAGS_altivec8.o += $(altivec_flags)
+targets += altivec1.c altivec2.c altivec4.c altivec8.c
+$(obj)/altivec%.c: $(src)/altivec.uc $(src)/unroll.awk FORCE
+	$(call if_changed,unroll)
+
+CFLAGS_vpermxor1.o += $(altivec_flags)
+CFLAGS_vpermxor2.o += $(altivec_flags)
+CFLAGS_vpermxor4.o += $(altivec_flags)
+CFLAGS_vpermxor8.o += $(altivec_flags)
+targets += vpermxor1.c vpermxor2.c vpermxor4.c vpermxor8.c
+$(obj)/vpermxor%.c: $(src)/vpermxor.uc $(src)/unroll.awk FORCE
+	$(call if_changed,unroll)
+
+CFLAGS_neon1.o += $(NEON_FLAGS)
+CFLAGS_neon2.o += $(NEON_FLAGS)
+CFLAGS_neon4.o += $(NEON_FLAGS)
+CFLAGS_neon8.o += $(NEON_FLAGS)
+targets += neon1.c neon2.c neon4.c neon8.c
+$(obj)/neon%.c: $(src)/neon.uc $(src)/unroll.awk FORCE
+	$(call if_changed,unroll)
+
+targets += s390vx8.c
+$(obj)/s390vx%.c: $(src)/s390vx.uc $(src)/unroll.awk FORCE
+	$(call if_changed,unroll)
+
+quiet_cmd_mktable = TABLE   $@
+      cmd_mktable = $(obj)/mktables > $@
+
+targets += tables.c
+$(obj)/tables.c: $(obj)/mktables FORCE
+	$(call if_changed,mktable)
diff --git a/lib/raid6/algos.c b/lib/raid6/algos.c
new file mode 100644
index 000000000..39b74221f
--- /dev/null
+++ b/lib/raid6/algos.c
@@ -0,0 +1,281 @@
+// SPDX-License-Identifier: GPL-2.0-or-later
+/* -*- linux-c -*- ------------------------------------------------------- *
+ *
+ *   Copyright 2002 H. Peter Anvin - All Rights Reserved
+ *
+ * ----------------------------------------------------------------------- */
+
+/*
+ * raid6/algos.c
+ *
+ * Algorithm list and algorithm selection for RAID-6
+ */
+
+#include <linux/raid/pq.h>
+#ifndef __KERNEL__
+#include <sys/mman.h>
+#include <stdio.h>
+#else
+#include <linux/module.h>
+#include <linux/gfp.h>
+#if !RAID6_USE_EMPTY_ZERO_PAGE
+/* In .bss so it's zeroed */
+const char raid6_empty_zero_page[PAGE_SIZE] __attribute__((aligned(256)));
+EXPORT_SYMBOL(raid6_empty_zero_page);
+#endif
+#endif
+
+struct raid6_calls raid6_call;
+EXPORT_SYMBOL_GPL(raid6_call);
+
+const struct raid6_calls * const raid6_algos[] = {
+#if defined(__i386__) && !defined(__arch_um__)
+#ifdef CONFIG_AS_AVX512
+	&raid6_avx512x2,
+	&raid6_avx512x1,
+#endif
+	&raid6_avx2x2,
+	&raid6_avx2x1,
+	&raid6_sse2x2,
+	&raid6_sse2x1,
+	&raid6_sse1x2,
+	&raid6_sse1x1,
+	&raid6_mmxx2,
+	&raid6_mmxx1,
+#endif
+#if defined(__x86_64__) && !defined(__arch_um__)
+#ifdef CONFIG_AS_AVX512
+	&raid6_avx512x4,
+	&raid6_avx512x2,
+	&raid6_avx512x1,
+#endif
+	&raid6_avx2x4,
+	&raid6_avx2x2,
+	&raid6_avx2x1,
+	&raid6_sse2x4,
+	&raid6_sse2x2,
+	&raid6_sse2x1,
+#endif
+#ifdef CONFIG_ALTIVEC
+	&raid6_vpermxor8,
+	&raid6_vpermxor4,
+	&raid6_vpermxor2,
+	&raid6_vpermxor1,
+	&raid6_altivec8,
+	&raid6_altivec4,
+	&raid6_altivec2,
+	&raid6_altivec1,
+#endif
+#if defined(CONFIG_S390)
+	&raid6_s390vx8,
+#endif
+#ifdef CONFIG_KERNEL_MODE_NEON
+	&raid6_neonx8,
+	&raid6_neonx4,
+	&raid6_neonx2,
+	&raid6_neonx1,
+#endif
+#if defined(__ia64__)
+	&raid6_intx32,
+	&raid6_intx16,
+#endif
+	&raid6_intx8,
+	&raid6_intx4,
+	&raid6_intx2,
+	&raid6_intx1,
+	NULL
+};
+
+void (*raid6_2data_recov)(int, size_t, int, int, void **);
+EXPORT_SYMBOL_GPL(raid6_2data_recov);
+
+void (*raid6_datap_recov)(int, size_t, int, void **);
+EXPORT_SYMBOL_GPL(raid6_datap_recov);
+
+const struct raid6_recov_calls *const raid6_recov_algos[] = {
+#ifdef CONFIG_X86
+#ifdef CONFIG_AS_AVX512
+	&raid6_recov_avx512,
+#endif
+	&raid6_recov_avx2,
+	&raid6_recov_ssse3,
+#endif
+#ifdef CONFIG_S390
+	&raid6_recov_s390xc,
+#endif
+#if defined(CONFIG_KERNEL_MODE_NEON)
+	&raid6_recov_neon,
+#endif
+	&raid6_recov_intx1,
+	NULL
+};
+
+#ifdef __KERNEL__
+#define RAID6_TIME_JIFFIES_LG2	4
+#else
+/* Need more time to be stable in userspace */
+#define RAID6_TIME_JIFFIES_LG2	9
+#define time_before(x, y) ((x) < (y))
+#endif
+
+#define RAID6_TEST_DISKS	8
+#define RAID6_TEST_DISKS_ORDER	3
+
+static inline const struct raid6_recov_calls *raid6_choose_recov(void)
+{
+	const struct raid6_recov_calls *const *algo;
+	const struct raid6_recov_calls *best;
+
+	for (best = NULL, algo = raid6_recov_algos; *algo; algo++)
+		if (!best || (*algo)->priority > best->priority)
+			if (!(*algo)->valid || (*algo)->valid())
+				best = *algo;
+
+	if (best) {
+		raid6_2data_recov = best->data2;
+		raid6_datap_recov = best->datap;
+
+		pr_info("raid6: using %s recovery algorithm\n", best->name);
+	} else
+		pr_err("raid6: Yikes! No recovery algorithm found!\n");
+
+	return best;
+}
+
+static inline const struct raid6_calls *raid6_choose_gen(
+	void *(*const dptrs)[RAID6_TEST_DISKS], const int disks)
+{
+	unsigned long perf, bestgenperf, j0, j1;
+	int start = (disks>>1)-1, stop = disks-3;	/* work on the second half of the disks */
+	const struct raid6_calls *const *algo;
+	const struct raid6_calls *best;
+
+	for (bestgenperf = 0, best = NULL, algo = raid6_algos; *algo; algo++) {
+		if (!best || (*algo)->priority >= best->priority) {
+			if ((*algo)->valid && !(*algo)->valid())
+				continue;
+
+			if (!IS_ENABLED(CONFIG_RAID6_PQ_BENCHMARK)) {
+				best = *algo;
+				break;
+			}
+
+			perf = 0;
+
+			preempt_disable();
+			j0 = jiffies;
+			while ((j1 = jiffies) == j0)
+				cpu_relax();
+			while (time_before(jiffies,
+					    j1 + (1<<RAID6_TIME_JIFFIES_LG2))) {
+				(*algo)->gen_syndrome(disks, PAGE_SIZE, *dptrs);
+				perf++;
+			}
+			preempt_enable();
+
+			if (perf > bestgenperf) {
+				bestgenperf = perf;
+				best = *algo;
+			}
+			pr_info("raid6: %-8s gen() %5ld MB/s\n", (*algo)->name,
+				(perf * HZ * (disks-2)) >>
+				(20 - PAGE_SHIFT + RAID6_TIME_JIFFIES_LG2));
+		}
+	}
+
+	if (!best) {
+		pr_err("raid6: Yikes! No algorithm found!\n");
+		goto out;
+	}
+
+	raid6_call = *best;
+
+	if (!IS_ENABLED(CONFIG_RAID6_PQ_BENCHMARK)) {
+		pr_info("raid6: skipped pq benchmark and selected %s\n",
+			best->name);
+		goto out;
+	}
+
+	pr_info("raid6: using algorithm %s gen() %ld MB/s\n",
+		best->name,
+		(bestgenperf * HZ * (disks - 2)) >>
+		(20 - PAGE_SHIFT + RAID6_TIME_JIFFIES_LG2));
+
+	if (best->xor_syndrome) {
+		perf = 0;
+
+		preempt_disable();
+		j0 = jiffies;
+		while ((j1 = jiffies) == j0)
+			cpu_relax();
+		while (time_before(jiffies,
+				   j1 + (1 << RAID6_TIME_JIFFIES_LG2))) {
+			best->xor_syndrome(disks, start, stop,
+					   PAGE_SIZE, *dptrs);
+			perf++;
+		}
+		preempt_enable();
+
+		pr_info("raid6: .... xor() %ld MB/s, rmw enabled\n",
+			(perf * HZ * (disks - 2)) >>
+			(20 - PAGE_SHIFT + RAID6_TIME_JIFFIES_LG2 + 1));
+	}
+
+out:
+	return best;
+}
+
+
+/* Try to pick the best algorithm */
+/* This code uses the gfmul table as convenient data set to abuse */
+
+int __init raid6_select_algo(void)
+{
+	const int disks = RAID6_TEST_DISKS;
+
+	const struct raid6_calls *gen_best;
+	const struct raid6_recov_calls *rec_best;
+	char *disk_ptr, *p;
+	void *dptrs[RAID6_TEST_DISKS];
+	int i, cycle;
+
+	/* prepare the buffer and fill it circularly with gfmul table */
+	disk_ptr = (char *)__get_free_pages(GFP_KERNEL, RAID6_TEST_DISKS_ORDER);
+	if (!disk_ptr) {
+		pr_err("raid6: Yikes!  No memory available.\n");
+		return -ENOMEM;
+	}
+
+	p = disk_ptr;
+	for (i = 0; i < disks; i++)
+		dptrs[i] = p + PAGE_SIZE * i;
+
+	cycle = ((disks - 2) * PAGE_SIZE) / 65536;
+	for (i = 0; i < cycle; i++) {
+		memcpy(p, raid6_gfmul, 65536);
+		p += 65536;
+	}
+
+	if ((disks - 2) * PAGE_SIZE % 65536)
+		memcpy(p, raid6_gfmul, (disks - 2) * PAGE_SIZE % 65536);
+
+	/* select raid gen_syndrome function */
+	gen_best = raid6_choose_gen(&dptrs, disks);
+
+	/* select raid recover functions */
+	rec_best = raid6_choose_recov();
+
+	free_pages((unsigned long)disk_ptr, RAID6_TEST_DISKS_ORDER);
+
+	return gen_best && rec_best ? 0 : -EINVAL;
+}
+
+static void raid6_exit(void)
+{
+	do { } while (0);
+}
+
+subsys_initcall(raid6_select_algo);
+module_exit(raid6_exit);
+MODULE_LICENSE("GPL");
+MODULE_DESCRIPTION("RAID6 Q-syndrome calculations");
diff --git a/lib/raid6/altivec.uc b/lib/raid6/altivec.uc
new file mode 100644
index 000000000..d20ed0d11
--- /dev/null
+++ b/lib/raid6/altivec.uc
@@ -0,0 +1,132 @@
+/* -*- linux-c -*- ------------------------------------------------------- *
+ *
+ *   Copyright 2002-2004 H. Peter Anvin - All Rights Reserved
+ *
+ *   This program is free software; you can redistribute it and/or modify
+ *   it under the terms of the GNU General Public License as published by
+ *   the Free Software Foundation, Inc., 53 Temple Place Ste 330,
+ *   Boston MA 02111-1307, USA; either version 2 of the License, or
+ *   (at your option) any later version; incorporated herein by reference.
+ *
+ * ----------------------------------------------------------------------- */
+
+/*
+ * raid6altivec$#.c
+ *
+ * $#-way unrolled portable integer math RAID-6 instruction set
+ *
+ * This file is postprocessed using unroll.awk
+ *
+ * <benh> hpa: in process,
+ * you can just "steal" the vec unit with enable_kernel_altivec() (but
+ * bracked this with preempt_disable/enable or in a lock)
+ */
+
+#include <linux/raid/pq.h>
+
+#ifdef CONFIG_ALTIVEC
+
+#include <altivec.h>
+#ifdef __KERNEL__
+# include <asm/cputable.h>
+# include <asm/switch_to.h>
+#endif /* __KERNEL__ */
+
+/*
+ * This is the C data type to use.  We use a vector of
+ * signed char so vec_cmpgt() will generate the right
+ * instruction.
+ */
+
+typedef vector signed char unative_t;
+
+#define NBYTES(x) ((vector signed char) {x,x,x,x, x,x,x,x, x,x,x,x, x,x,x,x})
+#define NSIZE	sizeof(unative_t)
+
+/*
+ * The SHLBYTE() operation shifts each byte left by 1, *not*
+ * rolling over into the next byte
+ */
+static inline __attribute_const__ unative_t SHLBYTE(unative_t v)
+{
+	return vec_add(v,v);
+}
+
+/*
+ * The MASK() operation returns 0xFF in any byte for which the high
+ * bit is 1, 0x00 for any byte for which the high bit is 0.
+ */
+static inline __attribute_const__ unative_t MASK(unative_t v)
+{
+	unative_t zv = NBYTES(0);
+
+	/* vec_cmpgt returns a vector bool char; thus the need for the cast */
+	return (unative_t)vec_cmpgt(zv, v);
+}
+
+
+/* This is noinline to make damned sure that gcc doesn't move any of the
+   Altivec code around the enable/disable code */
+static void noinline
+raid6_altivec$#_gen_syndrome_real(int disks, size_t bytes, void **ptrs)
+{
+	u8 **dptr = (u8 **)ptrs;
+	u8 *p, *q;
+	int d, z, z0;
+
+	unative_t wd$$, wq$$, wp$$, w1$$, w2$$;
+	unative_t x1d = NBYTES(0x1d);
+
+	z0 = disks - 3;		/* Highest data disk */
+	p = dptr[z0+1];		/* XOR parity */
+	q = dptr[z0+2];		/* RS syndrome */
+
+	for ( d = 0 ; d < bytes ; d += NSIZE*$# ) {
+		wq$$ = wp$$ = *(unative_t *)&dptr[z0][d+$$*NSIZE];
+		for ( z = z0-1 ; z >= 0 ; z-- ) {
+			wd$$ = *(unative_t *)&dptr[z][d+$$*NSIZE];
+			wp$$ = vec_xor(wp$$, wd$$);
+			w2$$ = MASK(wq$$);
+			w1$$ = SHLBYTE(wq$$);
+			w2$$ = vec_and(w2$$, x1d);
+			w1$$ = vec_xor(w1$$, w2$$);
+			wq$$ = vec_xor(w1$$, wd$$);
+		}
+		*(unative_t *)&p[d+NSIZE*$$] = wp$$;
+		*(unative_t *)&q[d+NSIZE*$$] = wq$$;
+	}
+}
+
+static void raid6_altivec$#_gen_syndrome(int disks, size_t bytes, void **ptrs)
+{
+	preempt_disable();
+	enable_kernel_altivec();
+
+	raid6_altivec$#_gen_syndrome_real(disks, bytes, ptrs);
+
+	disable_kernel_altivec();
+	preempt_enable();
+}
+
+int raid6_have_altivec(void);
+#if $# == 1
+int raid6_have_altivec(void)
+{
+	/* This assumes either all CPUs have Altivec or none does */
+# ifdef __KERNEL__
+	return cpu_has_feature(CPU_FTR_ALTIVEC);
+# else
+	return 1;
+# endif
+}
+#endif
+
+const struct raid6_calls raid6_altivec$# = {
+	raid6_altivec$#_gen_syndrome,
+	NULL,			/* XOR not yet implemented */
+	raid6_have_altivec,
+	"altivecx$#",
+	0
+};
+
+#endif /* CONFIG_ALTIVEC */
diff --git a/lib/raid6/avx2.c b/lib/raid6/avx2.c
new file mode 100644
index 000000000..059024234
--- /dev/null
+++ b/lib/raid6/avx2.c
@@ -0,0 +1,470 @@
+// SPDX-License-Identifier: GPL-2.0-or-later
+/* -*- linux-c -*- ------------------------------------------------------- *
+ *
+ *   Copyright (C) 2012 Intel Corporation
+ *   Author: Yuanhan Liu <yuanhan.liu@linux.intel.com>
+ *
+ *   Based on sse2.c: Copyright 2002 H. Peter Anvin - All Rights Reserved
+ *
+ * ----------------------------------------------------------------------- */
+
+/*
+ * AVX2 implementation of RAID-6 syndrome functions
+ *
+ */
+
+#include <linux/raid/pq.h>
+#include "x86.h"
+
+static const struct raid6_avx2_constants {
+	u64 x1d[4];
+} raid6_avx2_constants __aligned(32) = {
+	{ 0x1d1d1d1d1d1d1d1dULL, 0x1d1d1d1d1d1d1d1dULL,
+	  0x1d1d1d1d1d1d1d1dULL, 0x1d1d1d1d1d1d1d1dULL,},
+};
+
+static int raid6_have_avx2(void)
+{
+	return boot_cpu_has(X86_FEATURE_AVX2) && boot_cpu_has(X86_FEATURE_AVX);
+}
+
+/*
+ * Plain AVX2 implementation
+ */
+static void raid6_avx21_gen_syndrome(int disks, size_t bytes, void **ptrs)
+{
+	u8 **dptr = (u8 **)ptrs;
+	u8 *p, *q;
+	int d, z, z0;
+
+	z0 = disks - 3;		/* Highest data disk */
+	p = dptr[z0+1];		/* XOR parity */
+	q = dptr[z0+2];		/* RS syndrome */
+
+	kernel_fpu_begin();
+
+	asm volatile("vmovdqa %0,%%ymm0" : : "m" (raid6_avx2_constants.x1d[0]));
+	asm volatile("vpxor %ymm3,%ymm3,%ymm3");	/* Zero temp */
+
+	for (d = 0; d < bytes; d += 32) {
+		asm volatile("prefetchnta %0" : : "m" (dptr[z0][d]));
+		asm volatile("vmovdqa %0,%%ymm2" : : "m" (dptr[z0][d]));/* P[0] */
+		asm volatile("prefetchnta %0" : : "m" (dptr[z0-1][d]));
+		asm volatile("vmovdqa %ymm2,%ymm4");/* Q[0] */
+		asm volatile("vmovdqa %0,%%ymm6" : : "m" (dptr[z0-1][d]));
+		for (z = z0-2; z >= 0; z--) {
+			asm volatile("prefetchnta %0" : : "m" (dptr[z][d]));
+			asm volatile("vpcmpgtb %ymm4,%ymm3,%ymm5");
+			asm volatile("vpaddb %ymm4,%ymm4,%ymm4");
+			asm volatile("vpand %ymm0,%ymm5,%ymm5");
+			asm volatile("vpxor %ymm5,%ymm4,%ymm4");
+			asm volatile("vpxor %ymm6,%ymm2,%ymm2");
+			asm volatile("vpxor %ymm6,%ymm4,%ymm4");
+			asm volatile("vmovdqa %0,%%ymm6" : : "m" (dptr[z][d]));
+		}
+		asm volatile("vpcmpgtb %ymm4,%ymm3,%ymm5");
+		asm volatile("vpaddb %ymm4,%ymm4,%ymm4");
+		asm volatile("vpand %ymm0,%ymm5,%ymm5");
+		asm volatile("vpxor %ymm5,%ymm4,%ymm4");
+		asm volatile("vpxor %ymm6,%ymm2,%ymm2");
+		asm volatile("vpxor %ymm6,%ymm4,%ymm4");
+
+		asm volatile("vmovntdq %%ymm2,%0" : "=m" (p[d]));
+		asm volatile("vpxor %ymm2,%ymm2,%ymm2");
+		asm volatile("vmovntdq %%ymm4,%0" : "=m" (q[d]));
+		asm volatile("vpxor %ymm4,%ymm4,%ymm4");
+	}
+
+	asm volatile("sfence" : : : "memory");
+	kernel_fpu_end();
+}
+
+static void raid6_avx21_xor_syndrome(int disks, int start, int stop,
+				     size_t bytes, void **ptrs)
+{
+	u8 **dptr = (u8 **)ptrs;
+	u8 *p, *q;
+	int d, z, z0;
+
+	z0 = stop;		/* P/Q right side optimization */
+	p = dptr[disks-2];	/* XOR parity */
+	q = dptr[disks-1];	/* RS syndrome */
+
+	kernel_fpu_begin();
+
+	asm volatile("vmovdqa %0,%%ymm0" : : "m" (raid6_avx2_constants.x1d[0]));
+
+	for (d = 0 ; d < bytes ; d += 32) {
+		asm volatile("vmovdqa %0,%%ymm4" :: "m" (dptr[z0][d]));
+		asm volatile("vmovdqa %0,%%ymm2" : : "m" (p[d]));
+		asm volatile("vpxor %ymm4,%ymm2,%ymm2");
+		/* P/Q data pages */
+		for (z = z0-1 ; z >= start ; z--) {
+			asm volatile("vpxor %ymm5,%ymm5,%ymm5");
+			asm volatile("vpcmpgtb %ymm4,%ymm5,%ymm5");
+			asm volatile("vpaddb %ymm4,%ymm4,%ymm4");
+			asm volatile("vpand %ymm0,%ymm5,%ymm5");
+			asm volatile("vpxor %ymm5,%ymm4,%ymm4");
+			asm volatile("vmovdqa %0,%%ymm5" :: "m" (dptr[z][d]));
+			asm volatile("vpxor %ymm5,%ymm2,%ymm2");
+			asm volatile("vpxor %ymm5,%ymm4,%ymm4");
+		}
+		/* P/Q left side optimization */
+		for (z = start-1 ; z >= 0 ; z--) {
+			asm volatile("vpxor %ymm5,%ymm5,%ymm5");
+			asm volatile("vpcmpgtb %ymm4,%ymm5,%ymm5");
+			asm volatile("vpaddb %ymm4,%ymm4,%ymm4");
+			asm volatile("vpand %ymm0,%ymm5,%ymm5");
+			asm volatile("vpxor %ymm5,%ymm4,%ymm4");
+		}
+		asm volatile("vpxor %0,%%ymm4,%%ymm4" : : "m" (q[d]));
+		/* Don't use movntdq for r/w memory area < cache line */
+		asm volatile("vmovdqa %%ymm4,%0" : "=m" (q[d]));
+		asm volatile("vmovdqa %%ymm2,%0" : "=m" (p[d]));
+	}
+
+	asm volatile("sfence" : : : "memory");
+	kernel_fpu_end();
+}
+
+const struct raid6_calls raid6_avx2x1 = {
+	raid6_avx21_gen_syndrome,
+	raid6_avx21_xor_syndrome,
+	raid6_have_avx2,
+	"avx2x1",
+	.priority = 2		/* Prefer AVX2 over priority 1 (SSE2 and others) */
+};
+
+/*
+ * Unrolled-by-2 AVX2 implementation
+ */
+static void raid6_avx22_gen_syndrome(int disks, size_t bytes, void **ptrs)
+{
+	u8 **dptr = (u8 **)ptrs;
+	u8 *p, *q;
+	int d, z, z0;
+
+	z0 = disks - 3;		/* Highest data disk */
+	p = dptr[z0+1];		/* XOR parity */
+	q = dptr[z0+2];		/* RS syndrome */
+
+	kernel_fpu_begin();
+
+	asm volatile("vmovdqa %0,%%ymm0" : : "m" (raid6_avx2_constants.x1d[0]));
+	asm volatile("vpxor %ymm1,%ymm1,%ymm1"); /* Zero temp */
+
+	/* We uniformly assume a single prefetch covers at least 32 bytes */
+	for (d = 0; d < bytes; d += 64) {
+		asm volatile("prefetchnta %0" : : "m" (dptr[z0][d]));
+		asm volatile("prefetchnta %0" : : "m" (dptr[z0][d+32]));
+		asm volatile("vmovdqa %0,%%ymm2" : : "m" (dptr[z0][d]));/* P[0] */
+		asm volatile("vmovdqa %0,%%ymm3" : : "m" (dptr[z0][d+32]));/* P[1] */
+		asm volatile("vmovdqa %ymm2,%ymm4"); /* Q[0] */
+		asm volatile("vmovdqa %ymm3,%ymm6"); /* Q[1] */
+		for (z = z0-1; z >= 0; z--) {
+			asm volatile("prefetchnta %0" : : "m" (dptr[z][d]));
+			asm volatile("prefetchnta %0" : : "m" (dptr[z][d+32]));
+			asm volatile("vpcmpgtb %ymm4,%ymm1,%ymm5");
+			asm volatile("vpcmpgtb %ymm6,%ymm1,%ymm7");
+			asm volatile("vpaddb %ymm4,%ymm4,%ymm4");
+			asm volatile("vpaddb %ymm6,%ymm6,%ymm6");
+			asm volatile("vpand %ymm0,%ymm5,%ymm5");
+			asm volatile("vpand %ymm0,%ymm7,%ymm7");
+			asm volatile("vpxor %ymm5,%ymm4,%ymm4");
+			asm volatile("vpxor %ymm7,%ymm6,%ymm6");
+			asm volatile("vmovdqa %0,%%ymm5" : : "m" (dptr[z][d]));
+			asm volatile("vmovdqa %0,%%ymm7" : : "m" (dptr[z][d+32]));
+			asm volatile("vpxor %ymm5,%ymm2,%ymm2");
+			asm volatile("vpxor %ymm7,%ymm3,%ymm3");
+			asm volatile("vpxor %ymm5,%ymm4,%ymm4");
+			asm volatile("vpxor %ymm7,%ymm6,%ymm6");
+		}
+		asm volatile("vmovntdq %%ymm2,%0" : "=m" (p[d]));
+		asm volatile("vmovntdq %%ymm3,%0" : "=m" (p[d+32]));
+		asm volatile("vmovntdq %%ymm4,%0" : "=m" (q[d]));
+		asm volatile("vmovntdq %%ymm6,%0" : "=m" (q[d+32]));
+	}
+
+	asm volatile("sfence" : : : "memory");
+	kernel_fpu_end();
+}
+
+static void raid6_avx22_xor_syndrome(int disks, int start, int stop,
+				     size_t bytes, void **ptrs)
+{
+	u8 **dptr = (u8 **)ptrs;
+	u8 *p, *q;
+	int d, z, z0;
+
+	z0 = stop;		/* P/Q right side optimization */
+	p = dptr[disks-2];	/* XOR parity */
+	q = dptr[disks-1];	/* RS syndrome */
+
+	kernel_fpu_begin();
+
+	asm volatile("vmovdqa %0,%%ymm0" : : "m" (raid6_avx2_constants.x1d[0]));
+
+	for (d = 0 ; d < bytes ; d += 64) {
+		asm volatile("vmovdqa %0,%%ymm4" :: "m" (dptr[z0][d]));
+		asm volatile("vmovdqa %0,%%ymm6" :: "m" (dptr[z0][d+32]));
+		asm volatile("vmovdqa %0,%%ymm2" : : "m" (p[d]));
+		asm volatile("vmovdqa %0,%%ymm3" : : "m" (p[d+32]));
+		asm volatile("vpxor %ymm4,%ymm2,%ymm2");
+		asm volatile("vpxor %ymm6,%ymm3,%ymm3");
+		/* P/Q data pages */
+		for (z = z0-1 ; z >= start ; z--) {
+			asm volatile("vpxor %ymm5,%ymm5,%ymm5");
+			asm volatile("vpxor %ymm7,%ymm7,%ymm7");
+			asm volatile("vpcmpgtb %ymm4,%ymm5,%ymm5");
+			asm volatile("vpcmpgtb %ymm6,%ymm7,%ymm7");
+			asm volatile("vpaddb %ymm4,%ymm4,%ymm4");
+			asm volatile("vpaddb %ymm6,%ymm6,%ymm6");
+			asm volatile("vpand %ymm0,%ymm5,%ymm5");
+			asm volatile("vpand %ymm0,%ymm7,%ymm7");
+			asm volatile("vpxor %ymm5,%ymm4,%ymm4");
+			asm volatile("vpxor %ymm7,%ymm6,%ymm6");
+			asm volatile("vmovdqa %0,%%ymm5" :: "m" (dptr[z][d]));
+			asm volatile("vmovdqa %0,%%ymm7"
+				     :: "m" (dptr[z][d+32]));
+			asm volatile("vpxor %ymm5,%ymm2,%ymm2");
+			asm volatile("vpxor %ymm7,%ymm3,%ymm3");
+			asm volatile("vpxor %ymm5,%ymm4,%ymm4");
+			asm volatile("vpxor %ymm7,%ymm6,%ymm6");
+		}
+		/* P/Q left side optimization */
+		for (z = start-1 ; z >= 0 ; z--) {
+			asm volatile("vpxor %ymm5,%ymm5,%ymm5");
+			asm volatile("vpxor %ymm7,%ymm7,%ymm7");
+			asm volatile("vpcmpgtb %ymm4,%ymm5,%ymm5");
+			asm volatile("vpcmpgtb %ymm6,%ymm7,%ymm7");
+			asm volatile("vpaddb %ymm4,%ymm4,%ymm4");
+			asm volatile("vpaddb %ymm6,%ymm6,%ymm6");
+			asm volatile("vpand %ymm0,%ymm5,%ymm5");
+			asm volatile("vpand %ymm0,%ymm7,%ymm7");
+			asm volatile("vpxor %ymm5,%ymm4,%ymm4");
+			asm volatile("vpxor %ymm7,%ymm6,%ymm6");
+		}
+		asm volatile("vpxor %0,%%ymm4,%%ymm4" : : "m" (q[d]));
+		asm volatile("vpxor %0,%%ymm6,%%ymm6" : : "m" (q[d+32]));
+		/* Don't use movntdq for r/w memory area < cache line */
+		asm volatile("vmovdqa %%ymm4,%0" : "=m" (q[d]));
+		asm volatile("vmovdqa %%ymm6,%0" : "=m" (q[d+32]));
+		asm volatile("vmovdqa %%ymm2,%0" : "=m" (p[d]));
+		asm volatile("vmovdqa %%ymm3,%0" : "=m" (p[d+32]));
+	}
+
+	asm volatile("sfence" : : : "memory");
+	kernel_fpu_end();
+}
+
+const struct raid6_calls raid6_avx2x2 = {
+	raid6_avx22_gen_syndrome,
+	raid6_avx22_xor_syndrome,
+	raid6_have_avx2,
+	"avx2x2",
+	.priority = 2		/* Prefer AVX2 over priority 1 (SSE2 and others) */
+};
+
+#ifdef CONFIG_X86_64
+
+/*
+ * Unrolled-by-4 AVX2 implementation
+ */
+static void raid6_avx24_gen_syndrome(int disks, size_t bytes, void **ptrs)
+{
+	u8 **dptr = (u8 **)ptrs;
+	u8 *p, *q;
+	int d, z, z0;
+
+	z0 = disks - 3;		/* Highest data disk */
+	p = dptr[z0+1];		/* XOR parity */
+	q = dptr[z0+2];		/* RS syndrome */
+
+	kernel_fpu_begin();
+
+	asm volatile("vmovdqa %0,%%ymm0" : : "m" (raid6_avx2_constants.x1d[0]));
+	asm volatile("vpxor %ymm1,%ymm1,%ymm1");	/* Zero temp */
+	asm volatile("vpxor %ymm2,%ymm2,%ymm2");	/* P[0] */
+	asm volatile("vpxor %ymm3,%ymm3,%ymm3");	/* P[1] */
+	asm volatile("vpxor %ymm4,%ymm4,%ymm4");	/* Q[0] */
+	asm volatile("vpxor %ymm6,%ymm6,%ymm6");	/* Q[1] */
+	asm volatile("vpxor %ymm10,%ymm10,%ymm10");	/* P[2] */
+	asm volatile("vpxor %ymm11,%ymm11,%ymm11");	/* P[3] */
+	asm volatile("vpxor %ymm12,%ymm12,%ymm12");	/* Q[2] */
+	asm volatile("vpxor %ymm14,%ymm14,%ymm14");	/* Q[3] */
+
+	for (d = 0; d < bytes; d += 128) {
+		for (z = z0; z >= 0; z--) {
+			asm volatile("prefetchnta %0" : : "m" (dptr[z][d]));
+			asm volatile("prefetchnta %0" : : "m" (dptr[z][d+32]));
+			asm volatile("prefetchnta %0" : : "m" (dptr[z][d+64]));
+			asm volatile("prefetchnta %0" : : "m" (dptr[z][d+96]));
+			asm volatile("vpcmpgtb %ymm4,%ymm1,%ymm5");
+			asm volatile("vpcmpgtb %ymm6,%ymm1,%ymm7");
+			asm volatile("vpcmpgtb %ymm12,%ymm1,%ymm13");
+			asm volatile("vpcmpgtb %ymm14,%ymm1,%ymm15");
+			asm volatile("vpaddb %ymm4,%ymm4,%ymm4");
+			asm volatile("vpaddb %ymm6,%ymm6,%ymm6");
+			asm volatile("vpaddb %ymm12,%ymm12,%ymm12");
+			asm volatile("vpaddb %ymm14,%ymm14,%ymm14");
+			asm volatile("vpand %ymm0,%ymm5,%ymm5");
+			asm volatile("vpand %ymm0,%ymm7,%ymm7");
+			asm volatile("vpand %ymm0,%ymm13,%ymm13");
+			asm volatile("vpand %ymm0,%ymm15,%ymm15");
+			asm volatile("vpxor %ymm5,%ymm4,%ymm4");
+			asm volatile("vpxor %ymm7,%ymm6,%ymm6");
+			asm volatile("vpxor %ymm13,%ymm12,%ymm12");
+			asm volatile("vpxor %ymm15,%ymm14,%ymm14");
+			asm volatile("vmovdqa %0,%%ymm5" : : "m" (dptr[z][d]));
+			asm volatile("vmovdqa %0,%%ymm7" : : "m" (dptr[z][d+32]));
+			asm volatile("vmovdqa %0,%%ymm13" : : "m" (dptr[z][d+64]));
+			asm volatile("vmovdqa %0,%%ymm15" : : "m" (dptr[z][d+96]));
+			asm volatile("vpxor %ymm5,%ymm2,%ymm2");
+			asm volatile("vpxor %ymm7,%ymm3,%ymm3");
+			asm volatile("vpxor %ymm13,%ymm10,%ymm10");
+			asm volatile("vpxor %ymm15,%ymm11,%ymm11");
+			asm volatile("vpxor %ymm5,%ymm4,%ymm4");
+			asm volatile("vpxor %ymm7,%ymm6,%ymm6");
+			asm volatile("vpxor %ymm13,%ymm12,%ymm12");
+			asm volatile("vpxor %ymm15,%ymm14,%ymm14");
+		}
+		asm volatile("vmovntdq %%ymm2,%0" : "=m" (p[d]));
+		asm volatile("vpxor %ymm2,%ymm2,%ymm2");
+		asm volatile("vmovntdq %%ymm3,%0" : "=m" (p[d+32]));
+		asm volatile("vpxor %ymm3,%ymm3,%ymm3");
+		asm volatile("vmovntdq %%ymm10,%0" : "=m" (p[d+64]));
+		asm volatile("vpxor %ymm10,%ymm10,%ymm10");
+		asm volatile("vmovntdq %%ymm11,%0" : "=m" (p[d+96]));
+		asm volatile("vpxor %ymm11,%ymm11,%ymm11");
+		asm volatile("vmovntdq %%ymm4,%0" : "=m" (q[d]));
+		asm volatile("vpxor %ymm4,%ymm4,%ymm4");
+		asm volatile("vmovntdq %%ymm6,%0" : "=m" (q[d+32]));
+		asm volatile("vpxor %ymm6,%ymm6,%ymm6");
+		asm volatile("vmovntdq %%ymm12,%0" : "=m" (q[d+64]));
+		asm volatile("vpxor %ymm12,%ymm12,%ymm12");
+		asm volatile("vmovntdq %%ymm14,%0" : "=m" (q[d+96]));
+		asm volatile("vpxor %ymm14,%ymm14,%ymm14");
+	}
+
+	asm volatile("sfence" : : : "memory");
+	kernel_fpu_end();
+}
+
+static void raid6_avx24_xor_syndrome(int disks, int start, int stop,
+				     size_t bytes, void **ptrs)
+{
+	u8 **dptr = (u8 **)ptrs;
+	u8 *p, *q;
+	int d, z, z0;
+
+	z0 = stop;		/* P/Q right side optimization */
+	p = dptr[disks-2];	/* XOR parity */
+	q = dptr[disks-1];	/* RS syndrome */
+
+	kernel_fpu_begin();
+
+	asm volatile("vmovdqa %0,%%ymm0" :: "m" (raid6_avx2_constants.x1d[0]));
+
+	for (d = 0 ; d < bytes ; d += 128) {
+		asm volatile("vmovdqa %0,%%ymm4" :: "m" (dptr[z0][d]));
+		asm volatile("vmovdqa %0,%%ymm6" :: "m" (dptr[z0][d+32]));
+		asm volatile("vmovdqa %0,%%ymm12" :: "m" (dptr[z0][d+64]));
+		asm volatile("vmovdqa %0,%%ymm14" :: "m" (dptr[z0][d+96]));
+		asm volatile("vmovdqa %0,%%ymm2" : : "m" (p[d]));
+		asm volatile("vmovdqa %0,%%ymm3" : : "m" (p[d+32]));
+		asm volatile("vmovdqa %0,%%ymm10" : : "m" (p[d+64]));
+		asm volatile("vmovdqa %0,%%ymm11" : : "m" (p[d+96]));
+		asm volatile("vpxor %ymm4,%ymm2,%ymm2");
+		asm volatile("vpxor %ymm6,%ymm3,%ymm3");
+		asm volatile("vpxor %ymm12,%ymm10,%ymm10");
+		asm volatile("vpxor %ymm14,%ymm11,%ymm11");
+		/* P/Q data pages */
+		for (z = z0-1 ; z >= start ; z--) {
+			asm volatile("prefetchnta %0" :: "m" (dptr[z][d]));
+			asm volatile("prefetchnta %0" :: "m" (dptr[z][d+64]));
+			asm volatile("vpxor %ymm5,%ymm5,%ymm5");
+			asm volatile("vpxor %ymm7,%ymm7,%ymm7");
+			asm volatile("vpxor %ymm13,%ymm13,%ymm13");
+			asm volatile("vpxor %ymm15,%ymm15,%ymm15");
+			asm volatile("vpcmpgtb %ymm4,%ymm5,%ymm5");
+			asm volatile("vpcmpgtb %ymm6,%ymm7,%ymm7");
+			asm volatile("vpcmpgtb %ymm12,%ymm13,%ymm13");
+			asm volatile("vpcmpgtb %ymm14,%ymm15,%ymm15");
+			asm volatile("vpaddb %ymm4,%ymm4,%ymm4");
+			asm volatile("vpaddb %ymm6,%ymm6,%ymm6");
+			asm volatile("vpaddb %ymm12,%ymm12,%ymm12");
+			asm volatile("vpaddb %ymm14,%ymm14,%ymm14");
+			asm volatile("vpand %ymm0,%ymm5,%ymm5");
+			asm volatile("vpand %ymm0,%ymm7,%ymm7");
+			asm volatile("vpand %ymm0,%ymm13,%ymm13");
+			asm volatile("vpand %ymm0,%ymm15,%ymm15");
+			asm volatile("vpxor %ymm5,%ymm4,%ymm4");
+			asm volatile("vpxor %ymm7,%ymm6,%ymm6");
+			asm volatile("vpxor %ymm13,%ymm12,%ymm12");
+			asm volatile("vpxor %ymm15,%ymm14,%ymm14");
+			asm volatile("vmovdqa %0,%%ymm5" :: "m" (dptr[z][d]));
+			asm volatile("vmovdqa %0,%%ymm7"
+				     :: "m" (dptr[z][d+32]));
+			asm volatile("vmovdqa %0,%%ymm13"
+				     :: "m" (dptr[z][d+64]));
+			asm volatile("vmovdqa %0,%%ymm15"
+				     :: "m" (dptr[z][d+96]));
+			asm volatile("vpxor %ymm5,%ymm2,%ymm2");
+			asm volatile("vpxor %ymm7,%ymm3,%ymm3");
+			asm volatile("vpxor %ymm13,%ymm10,%ymm10");
+			asm volatile("vpxor %ymm15,%ymm11,%ymm11");
+			asm volatile("vpxor %ymm5,%ymm4,%ymm4");
+			asm volatile("vpxor %ymm7,%ymm6,%ymm6");
+			asm volatile("vpxor %ymm13,%ymm12,%ymm12");
+			asm volatile("vpxor %ymm15,%ymm14,%ymm14");
+		}
+		asm volatile("prefetchnta %0" :: "m" (q[d]));
+		asm volatile("prefetchnta %0" :: "m" (q[d+64]));
+		/* P/Q left side optimization */
+		for (z = start-1 ; z >= 0 ; z--) {
+			asm volatile("vpxor %ymm5,%ymm5,%ymm5");
+			asm volatile("vpxor %ymm7,%ymm7,%ymm7");
+			asm volatile("vpxor %ymm13,%ymm13,%ymm13");
+			asm volatile("vpxor %ymm15,%ymm15,%ymm15");
+			asm volatile("vpcmpgtb %ymm4,%ymm5,%ymm5");
+			asm volatile("vpcmpgtb %ymm6,%ymm7,%ymm7");
+			asm volatile("vpcmpgtb %ymm12,%ymm13,%ymm13");
+			asm volatile("vpcmpgtb %ymm14,%ymm15,%ymm15");
+			asm volatile("vpaddb %ymm4,%ymm4,%ymm4");
+			asm volatile("vpaddb %ymm6,%ymm6,%ymm6");
+			asm volatile("vpaddb %ymm12,%ymm12,%ymm12");
+			asm volatile("vpaddb %ymm14,%ymm14,%ymm14");
+			asm volatile("vpand %ymm0,%ymm5,%ymm5");
+			asm volatile("vpand %ymm0,%ymm7,%ymm7");
+			asm volatile("vpand %ymm0,%ymm13,%ymm13");
+			asm volatile("vpand %ymm0,%ymm15,%ymm15");
+			asm volatile("vpxor %ymm5,%ymm4,%ymm4");
+			asm volatile("vpxor %ymm7,%ymm6,%ymm6");
+			asm volatile("vpxor %ymm13,%ymm12,%ymm12");
+			asm volatile("vpxor %ymm15,%ymm14,%ymm14");
+		}
+		asm volatile("vmovntdq %%ymm2,%0" : "=m" (p[d]));
+		asm volatile("vmovntdq %%ymm3,%0" : "=m" (p[d+32]));
+		asm volatile("vmovntdq %%ymm10,%0" : "=m" (p[d+64]));
+		asm volatile("vmovntdq %%ymm11,%0" : "=m" (p[d+96]));
+		asm volatile("vpxor %0,%%ymm4,%%ymm4" : : "m" (q[d]));
+		asm volatile("vpxor %0,%%ymm6,%%ymm6" : : "m" (q[d+32]));
+		asm volatile("vpxor %0,%%ymm12,%%ymm12" : : "m" (q[d+64]));
+		asm volatile("vpxor %0,%%ymm14,%%ymm14" : : "m" (q[d+96]));
+		asm volatile("vmovntdq %%ymm4,%0" : "=m" (q[d]));
+		asm volatile("vmovntdq %%ymm6,%0" : "=m" (q[d+32]));
+		asm volatile("vmovntdq %%ymm12,%0" : "=m" (q[d+64]));
+		asm volatile("vmovntdq %%ymm14,%0" : "=m" (q[d+96]));
+	}
+	asm volatile("sfence" : : : "memory");
+	kernel_fpu_end();
+}
+
+const struct raid6_calls raid6_avx2x4 = {
+	raid6_avx24_gen_syndrome,
+	raid6_avx24_xor_syndrome,
+	raid6_have_avx2,
+	"avx2x4",
+	.priority = 2		/* Prefer AVX2 over priority 1 (SSE2 and others) */
+};
+#endif /* CONFIG_X86_64 */
diff --git a/lib/raid6/avx512.c b/lib/raid6/avx512.c
new file mode 100644
index 000000000..9c3e822e1
--- /dev/null
+++ b/lib/raid6/avx512.c
@@ -0,0 +1,564 @@
+// SPDX-License-Identifier: GPL-2.0-or-later
+/* -*- linux-c -*- --------------------------------------------------------
+ *
+ *   Copyright (C) 2016 Intel Corporation
+ *
+ *   Author: Gayatri Kammela <gayatri.kammela@intel.com>
+ *   Author: Megha Dey <megha.dey@linux.intel.com>
+ *
+ *   Based on avx2.c: Copyright 2012 Yuanhan Liu All Rights Reserved
+ *   Based on sse2.c: Copyright 2002 H. Peter Anvin - All Rights Reserved
+ *
+ * -----------------------------------------------------------------------
+ */
+
+/*
+ * AVX512 implementation of RAID-6 syndrome functions
+ *
+ */
+
+#ifdef CONFIG_AS_AVX512
+
+#include <linux/raid/pq.h>
+#include "x86.h"
+
+static const struct raid6_avx512_constants {
+	u64 x1d[8];
+} raid6_avx512_constants __aligned(512/8) = {
+	{ 0x1d1d1d1d1d1d1d1dULL, 0x1d1d1d1d1d1d1d1dULL,
+	  0x1d1d1d1d1d1d1d1dULL, 0x1d1d1d1d1d1d1d1dULL,
+	  0x1d1d1d1d1d1d1d1dULL, 0x1d1d1d1d1d1d1d1dULL,
+	  0x1d1d1d1d1d1d1d1dULL, 0x1d1d1d1d1d1d1d1dULL,},
+};
+
+static int raid6_have_avx512(void)
+{
+	return boot_cpu_has(X86_FEATURE_AVX2) &&
+		boot_cpu_has(X86_FEATURE_AVX) &&
+		boot_cpu_has(X86_FEATURE_AVX512F) &&
+		boot_cpu_has(X86_FEATURE_AVX512BW) &&
+		boot_cpu_has(X86_FEATURE_AVX512VL) &&
+		boot_cpu_has(X86_FEATURE_AVX512DQ);
+}
+
+static void raid6_avx5121_gen_syndrome(int disks, size_t bytes, void **ptrs)
+{
+	u8 **dptr = (u8 **)ptrs;
+	u8 *p, *q;
+	int d, z, z0;
+
+	z0 = disks - 3;         /* Highest data disk */
+	p = dptr[z0+1];         /* XOR parity */
+	q = dptr[z0+2];         /* RS syndrome */
+
+	kernel_fpu_begin();
+
+	asm volatile("vmovdqa64 %0,%%zmm0\n\t"
+		     "vpxorq %%zmm1,%%zmm1,%%zmm1" /* Zero temp */
+		     :
+		     : "m" (raid6_avx512_constants.x1d[0]));
+
+	for (d = 0; d < bytes; d += 64) {
+		asm volatile("prefetchnta %0\n\t"
+			     "vmovdqa64 %0,%%zmm2\n\t"     /* P[0] */
+			     "prefetchnta %1\n\t"
+			     "vmovdqa64 %%zmm2,%%zmm4\n\t" /* Q[0] */
+			     "vmovdqa64 %1,%%zmm6"
+			     :
+			     : "m" (dptr[z0][d]), "m" (dptr[z0-1][d]));
+		for (z = z0-2; z >= 0; z--) {
+			asm volatile("prefetchnta %0\n\t"
+				     "vpcmpgtb %%zmm4,%%zmm1,%%k1\n\t"
+				     "vpmovm2b %%k1,%%zmm5\n\t"
+				     "vpaddb %%zmm4,%%zmm4,%%zmm4\n\t"
+				     "vpandq %%zmm0,%%zmm5,%%zmm5\n\t"
+				     "vpxorq %%zmm5,%%zmm4,%%zmm4\n\t"
+				     "vpxorq %%zmm6,%%zmm2,%%zmm2\n\t"
+				     "vpxorq %%zmm6,%%zmm4,%%zmm4\n\t"
+				     "vmovdqa64 %0,%%zmm6"
+				     :
+				     : "m" (dptr[z][d]));
+		}
+		asm volatile("vpcmpgtb %%zmm4,%%zmm1,%%k1\n\t"
+			     "vpmovm2b %%k1,%%zmm5\n\t"
+			     "vpaddb %%zmm4,%%zmm4,%%zmm4\n\t"
+			     "vpandq %%zmm0,%%zmm5,%%zmm5\n\t"
+			     "vpxorq %%zmm5,%%zmm4,%%zmm4\n\t"
+			     "vpxorq %%zmm6,%%zmm2,%%zmm2\n\t"
+			     "vpxorq %%zmm6,%%zmm4,%%zmm4\n\t"
+			     "vmovntdq %%zmm2,%0\n\t"
+			     "vpxorq %%zmm2,%%zmm2,%%zmm2\n\t"
+			     "vmovntdq %%zmm4,%1\n\t"
+			     "vpxorq %%zmm4,%%zmm4,%%zmm4"
+			     :
+			     : "m" (p[d]), "m" (q[d]));
+	}
+
+	asm volatile("sfence" : : : "memory");
+	kernel_fpu_end();
+}
+
+static void raid6_avx5121_xor_syndrome(int disks, int start, int stop,
+				       size_t bytes, void **ptrs)
+{
+	u8 **dptr = (u8 **)ptrs;
+	u8 *p, *q;
+	int d, z, z0;
+
+	z0 = stop;		/* P/Q right side optimization */
+	p = dptr[disks-2];	/* XOR parity */
+	q = dptr[disks-1];	/* RS syndrome */
+
+	kernel_fpu_begin();
+
+	asm volatile("vmovdqa64 %0,%%zmm0"
+		     : : "m" (raid6_avx512_constants.x1d[0]));
+
+	for (d = 0 ; d < bytes ; d += 64) {
+		asm volatile("vmovdqa64 %0,%%zmm4\n\t"
+			     "vmovdqa64 %1,%%zmm2\n\t"
+			     "vpxorq %%zmm4,%%zmm2,%%zmm2"
+			     :
+			     : "m" (dptr[z0][d]),  "m" (p[d]));
+		/* P/Q data pages */
+		for (z = z0-1 ; z >= start ; z--) {
+			asm volatile("vpxorq %%zmm5,%%zmm5,%%zmm5\n\t"
+				     "vpcmpgtb %%zmm4,%%zmm5,%%k1\n\t"
+				     "vpmovm2b %%k1,%%zmm5\n\t"
+				     "vpaddb %%zmm4,%%zmm4,%%zmm4\n\t"
+				     "vpandq %%zmm0,%%zmm5,%%zmm5\n\t"
+				     "vpxorq %%zmm5,%%zmm4,%%zmm4\n\t"
+				     "vmovdqa64 %0,%%zmm5\n\t"
+				     "vpxorq %%zmm5,%%zmm2,%%zmm2\n\t"
+				     "vpxorq %%zmm5,%%zmm4,%%zmm4"
+				     :
+				     : "m" (dptr[z][d]));
+		}
+		/* P/Q left side optimization */
+		for (z = start-1 ; z >= 0 ; z--) {
+			asm volatile("vpxorq %%zmm5,%%zmm5,%%zmm5\n\t"
+				     "vpcmpgtb %%zmm4,%%zmm5,%%k1\n\t"
+				     "vpmovm2b %%k1,%%zmm5\n\t"
+				     "vpaddb %%zmm4,%%zmm4,%%zmm4\n\t"
+				     "vpandq %%zmm0,%%zmm5,%%zmm5\n\t"
+				     "vpxorq %%zmm5,%%zmm4,%%zmm4"
+				     :
+				     : );
+		}
+		asm volatile("vpxorq %0,%%zmm4,%%zmm4\n\t"
+		/* Don't use movntdq for r/w memory area < cache line */
+			     "vmovdqa64 %%zmm4,%0\n\t"
+			     "vmovdqa64 %%zmm2,%1"
+			     :
+			     : "m" (q[d]), "m" (p[d]));
+	}
+
+	asm volatile("sfence" : : : "memory");
+	kernel_fpu_end();
+}
+
+const struct raid6_calls raid6_avx512x1 = {
+	raid6_avx5121_gen_syndrome,
+	raid6_avx5121_xor_syndrome,
+	raid6_have_avx512,
+	"avx512x1",
+	.priority = 2		/* Prefer AVX512 over priority 1 (SSE2 and others) */
+};
+
+/*
+ * Unrolled-by-2 AVX512 implementation
+ */
+static void raid6_avx5122_gen_syndrome(int disks, size_t bytes, void **ptrs)
+{
+	u8 **dptr = (u8 **)ptrs;
+	u8 *p, *q;
+	int d, z, z0;
+
+	z0 = disks - 3;         /* Highest data disk */
+	p = dptr[z0+1];         /* XOR parity */
+	q = dptr[z0+2];         /* RS syndrome */
+
+	kernel_fpu_begin();
+
+	asm volatile("vmovdqa64 %0,%%zmm0\n\t"
+		     "vpxorq %%zmm1,%%zmm1,%%zmm1" /* Zero temp */
+		     :
+		     : "m" (raid6_avx512_constants.x1d[0]));
+
+	/* We uniformly assume a single prefetch covers at least 64 bytes */
+	for (d = 0; d < bytes; d += 128) {
+		asm volatile("prefetchnta %0\n\t"
+			     "prefetchnta %1\n\t"
+			     "vmovdqa64 %0,%%zmm2\n\t"      /* P[0] */
+			     "vmovdqa64 %1,%%zmm3\n\t"      /* P[1] */
+			     "vmovdqa64 %%zmm2,%%zmm4\n\t"  /* Q[0] */
+			     "vmovdqa64 %%zmm3,%%zmm6"      /* Q[1] */
+			     :
+			     : "m" (dptr[z0][d]), "m" (dptr[z0][d+64]));
+		for (z = z0-1; z >= 0; z--) {
+			asm volatile("prefetchnta %0\n\t"
+				     "prefetchnta %1\n\t"
+				     "vpcmpgtb %%zmm4,%%zmm1,%%k1\n\t"
+				     "vpcmpgtb %%zmm6,%%zmm1,%%k2\n\t"
+				     "vpmovm2b %%k1,%%zmm5\n\t"
+				     "vpmovm2b %%k2,%%zmm7\n\t"
+				     "vpaddb %%zmm4,%%zmm4,%%zmm4\n\t"
+				     "vpaddb %%zmm6,%%zmm6,%%zmm6\n\t"
+				     "vpandq %%zmm0,%%zmm5,%%zmm5\n\t"
+				     "vpandq %%zmm0,%%zmm7,%%zmm7\n\t"
+				     "vpxorq %%zmm5,%%zmm4,%%zmm4\n\t"
+				     "vpxorq %%zmm7,%%zmm6,%%zmm6\n\t"
+				     "vmovdqa64 %0,%%zmm5\n\t"
+				     "vmovdqa64 %1,%%zmm7\n\t"
+				     "vpxorq %%zmm5,%%zmm2,%%zmm2\n\t"
+				     "vpxorq %%zmm7,%%zmm3,%%zmm3\n\t"
+				     "vpxorq %%zmm5,%%zmm4,%%zmm4\n\t"
+				     "vpxorq %%zmm7,%%zmm6,%%zmm6"
+				     :
+				     : "m" (dptr[z][d]), "m" (dptr[z][d+64]));
+		}
+		asm volatile("vmovntdq %%zmm2,%0\n\t"
+			     "vmovntdq %%zmm3,%1\n\t"
+			     "vmovntdq %%zmm4,%2\n\t"
+			     "vmovntdq %%zmm6,%3"
+			     :
+			     : "m" (p[d]), "m" (p[d+64]), "m" (q[d]),
+			       "m" (q[d+64]));
+	}
+
+	asm volatile("sfence" : : : "memory");
+	kernel_fpu_end();
+}
+
+static void raid6_avx5122_xor_syndrome(int disks, int start, int stop,
+				       size_t bytes, void **ptrs)
+{
+	u8 **dptr = (u8 **)ptrs;
+	u8 *p, *q;
+	int d, z, z0;
+
+	z0 = stop;		/* P/Q right side optimization */
+	p = dptr[disks-2];	/* XOR parity */
+	q = dptr[disks-1];	/* RS syndrome */
+
+	kernel_fpu_begin();
+
+	asm volatile("vmovdqa64 %0,%%zmm0"
+		     : : "m" (raid6_avx512_constants.x1d[0]));
+
+	for (d = 0 ; d < bytes ; d += 128) {
+		asm volatile("vmovdqa64 %0,%%zmm4\n\t"
+			     "vmovdqa64 %1,%%zmm6\n\t"
+			     "vmovdqa64 %2,%%zmm2\n\t"
+			     "vmovdqa64 %3,%%zmm3\n\t"
+			     "vpxorq %%zmm4,%%zmm2,%%zmm2\n\t"
+			     "vpxorq %%zmm6,%%zmm3,%%zmm3"
+			     :
+			     : "m" (dptr[z0][d]), "m" (dptr[z0][d+64]),
+			       "m" (p[d]), "m" (p[d+64]));
+		/* P/Q data pages */
+		for (z = z0-1 ; z >= start ; z--) {
+			asm volatile("vpxorq %%zmm5,%%zmm5,%%zmm5\n\t"
+				     "vpxorq %%zmm7,%%zmm7,%%zmm7\n\t"
+				     "vpcmpgtb %%zmm4,%%zmm5,%%k1\n\t"
+				     "vpcmpgtb %%zmm6,%%zmm7,%%k2\n\t"
+				     "vpmovm2b %%k1,%%zmm5\n\t"
+				     "vpmovm2b %%k2,%%zmm7\n\t"
+				     "vpaddb %%zmm4,%%zmm4,%%zmm4\n\t"
+				     "vpaddb %%zmm6,%%zmm6,%%zmm6\n\t"
+				     "vpandq %%zmm0,%%zmm5,%%zmm5\n\t"
+				     "vpandq %%zmm0,%%zmm7,%%zmm7\n\t"
+				     "vpxorq %%zmm5,%%zmm4,%%zmm4\n\t"
+				     "vpxorq %%zmm7,%%zmm6,%%zmm6\n\t"
+				     "vmovdqa64 %0,%%zmm5\n\t"
+				     "vmovdqa64 %1,%%zmm7\n\t"
+				     "vpxorq %%zmm5,%%zmm2,%%zmm2\n\t"
+				     "vpxorq %%zmm7,%%zmm3,%%zmm3\n\t"
+				     "vpxorq %%zmm5,%%zmm4,%%zmm4\n\t"
+				     "vpxorq %%zmm7,%%zmm6,%%zmm6"
+				     :
+				     : "m" (dptr[z][d]),  "m" (dptr[z][d+64]));
+		}
+		/* P/Q left side optimization */
+		for (z = start-1 ; z >= 0 ; z--) {
+			asm volatile("vpxorq %%zmm5,%%zmm5,%%zmm5\n\t"
+				     "vpxorq %%zmm7,%%zmm7,%%zmm7\n\t"
+				     "vpcmpgtb %%zmm4,%%zmm5,%%k1\n\t"
+				     "vpcmpgtb %%zmm6,%%zmm7,%%k2\n\t"
+				     "vpmovm2b %%k1,%%zmm5\n\t"
+				     "vpmovm2b %%k2,%%zmm7\n\t"
+				     "vpaddb %%zmm4,%%zmm4,%%zmm4\n\t"
+				     "vpaddb %%zmm6,%%zmm6,%%zmm6\n\t"
+				     "vpandq %%zmm0,%%zmm5,%%zmm5\n\t"
+				     "vpandq %%zmm0,%%zmm7,%%zmm7\n\t"
+				     "vpxorq %%zmm5,%%zmm4,%%zmm4\n\t"
+				     "vpxorq %%zmm7,%%zmm6,%%zmm6"
+				     :
+				     : );
+		}
+		asm volatile("vpxorq %0,%%zmm4,%%zmm4\n\t"
+			     "vpxorq %1,%%zmm6,%%zmm6\n\t"
+			     /* Don't use movntdq for r/w
+			      * memory area < cache line
+			      */
+			     "vmovdqa64 %%zmm4,%0\n\t"
+			     "vmovdqa64 %%zmm6,%1\n\t"
+			     "vmovdqa64 %%zmm2,%2\n\t"
+			     "vmovdqa64 %%zmm3,%3"
+			     :
+			     : "m" (q[d]), "m" (q[d+64]), "m" (p[d]),
+			       "m" (p[d+64]));
+	}
+
+	asm volatile("sfence" : : : "memory");
+	kernel_fpu_end();
+}
+
+const struct raid6_calls raid6_avx512x2 = {
+	raid6_avx5122_gen_syndrome,
+	raid6_avx5122_xor_syndrome,
+	raid6_have_avx512,
+	"avx512x2",
+	.priority = 2		/* Prefer AVX512 over priority 1 (SSE2 and others) */
+};
+
+#ifdef CONFIG_X86_64
+
+/*
+ * Unrolled-by-4 AVX2 implementation
+ */
+static void raid6_avx5124_gen_syndrome(int disks, size_t bytes, void **ptrs)
+{
+	u8 **dptr = (u8 **)ptrs;
+	u8 *p, *q;
+	int d, z, z0;
+
+	z0 = disks - 3;         /* Highest data disk */
+	p = dptr[z0+1];         /* XOR parity */
+	q = dptr[z0+2];         /* RS syndrome */
+
+	kernel_fpu_begin();
+
+	asm volatile("vmovdqa64 %0,%%zmm0\n\t"
+		     "vpxorq %%zmm1,%%zmm1,%%zmm1\n\t"       /* Zero temp */
+		     "vpxorq %%zmm2,%%zmm2,%%zmm2\n\t"       /* P[0] */
+		     "vpxorq %%zmm3,%%zmm3,%%zmm3\n\t"       /* P[1] */
+		     "vpxorq %%zmm4,%%zmm4,%%zmm4\n\t"       /* Q[0] */
+		     "vpxorq %%zmm6,%%zmm6,%%zmm6\n\t"       /* Q[1] */
+		     "vpxorq %%zmm10,%%zmm10,%%zmm10\n\t"    /* P[2] */
+		     "vpxorq %%zmm11,%%zmm11,%%zmm11\n\t"    /* P[3] */
+		     "vpxorq %%zmm12,%%zmm12,%%zmm12\n\t"    /* Q[2] */
+		     "vpxorq %%zmm14,%%zmm14,%%zmm14"        /* Q[3] */
+		     :
+		     : "m" (raid6_avx512_constants.x1d[0]));
+
+	for (d = 0; d < bytes; d += 256) {
+		for (z = z0; z >= 0; z--) {
+		asm volatile("prefetchnta %0\n\t"
+			     "prefetchnta %1\n\t"
+			     "prefetchnta %2\n\t"
+			     "prefetchnta %3\n\t"
+			     "vpcmpgtb %%zmm4,%%zmm1,%%k1\n\t"
+			     "vpcmpgtb %%zmm6,%%zmm1,%%k2\n\t"
+			     "vpcmpgtb %%zmm12,%%zmm1,%%k3\n\t"
+			     "vpcmpgtb %%zmm14,%%zmm1,%%k4\n\t"
+			     "vpmovm2b %%k1,%%zmm5\n\t"
+			     "vpmovm2b %%k2,%%zmm7\n\t"
+			     "vpmovm2b %%k3,%%zmm13\n\t"
+			     "vpmovm2b %%k4,%%zmm15\n\t"
+			     "vpaddb %%zmm4,%%zmm4,%%zmm4\n\t"
+			     "vpaddb %%zmm6,%%zmm6,%%zmm6\n\t"
+			     "vpaddb %%zmm12,%%zmm12,%%zmm12\n\t"
+			     "vpaddb %%zmm14,%%zmm14,%%zmm14\n\t"
+			     "vpandq %%zmm0,%%zmm5,%%zmm5\n\t"
+			     "vpandq %%zmm0,%%zmm7,%%zmm7\n\t"
+			     "vpandq %%zmm0,%%zmm13,%%zmm13\n\t"
+			     "vpandq %%zmm0,%%zmm15,%%zmm15\n\t"
+			     "vpxorq %%zmm5,%%zmm4,%%zmm4\n\t"
+			     "vpxorq %%zmm7,%%zmm6,%%zmm6\n\t"
+			     "vpxorq %%zmm13,%%zmm12,%%zmm12\n\t"
+			     "vpxorq %%zmm15,%%zmm14,%%zmm14\n\t"
+			     "vmovdqa64 %0,%%zmm5\n\t"
+			     "vmovdqa64 %1,%%zmm7\n\t"
+			     "vmovdqa64 %2,%%zmm13\n\t"
+			     "vmovdqa64 %3,%%zmm15\n\t"
+			     "vpxorq %%zmm5,%%zmm2,%%zmm2\n\t"
+			     "vpxorq %%zmm7,%%zmm3,%%zmm3\n\t"
+			     "vpxorq %%zmm13,%%zmm10,%%zmm10\n\t"
+			     "vpxorq %%zmm15,%%zmm11,%%zmm11\n"
+			     "vpxorq %%zmm5,%%zmm4,%%zmm4\n\t"
+			     "vpxorq %%zmm7,%%zmm6,%%zmm6\n\t"
+			     "vpxorq %%zmm13,%%zmm12,%%zmm12\n\t"
+			     "vpxorq %%zmm15,%%zmm14,%%zmm14"
+			     :
+			     : "m" (dptr[z][d]), "m" (dptr[z][d+64]),
+			       "m" (dptr[z][d+128]), "m" (dptr[z][d+192]));
+		}
+		asm volatile("vmovntdq %%zmm2,%0\n\t"
+			     "vpxorq %%zmm2,%%zmm2,%%zmm2\n\t"
+			     "vmovntdq %%zmm3,%1\n\t"
+			     "vpxorq %%zmm3,%%zmm3,%%zmm3\n\t"
+			     "vmovntdq %%zmm10,%2\n\t"
+			     "vpxorq %%zmm10,%%zmm10,%%zmm10\n\t"
+			     "vmovntdq %%zmm11,%3\n\t"
+			     "vpxorq %%zmm11,%%zmm11,%%zmm11\n\t"
+			     "vmovntdq %%zmm4,%4\n\t"
+			     "vpxorq %%zmm4,%%zmm4,%%zmm4\n\t"
+			     "vmovntdq %%zmm6,%5\n\t"
+			     "vpxorq %%zmm6,%%zmm6,%%zmm6\n\t"
+			     "vmovntdq %%zmm12,%6\n\t"
+			     "vpxorq %%zmm12,%%zmm12,%%zmm12\n\t"
+			     "vmovntdq %%zmm14,%7\n\t"
+			     "vpxorq %%zmm14,%%zmm14,%%zmm14"
+			     :
+			     : "m" (p[d]), "m" (p[d+64]), "m" (p[d+128]),
+			       "m" (p[d+192]), "m" (q[d]), "m" (q[d+64]),
+			       "m" (q[d+128]), "m" (q[d+192]));
+	}
+
+	asm volatile("sfence" : : : "memory");
+	kernel_fpu_end();
+}
+
+static void raid6_avx5124_xor_syndrome(int disks, int start, int stop,
+				       size_t bytes, void **ptrs)
+{
+	u8 **dptr = (u8 **)ptrs;
+	u8 *p, *q;
+	int d, z, z0;
+
+	z0 = stop;		/* P/Q right side optimization */
+	p = dptr[disks-2];	/* XOR parity */
+	q = dptr[disks-1];	/* RS syndrome */
+
+	kernel_fpu_begin();
+
+	asm volatile("vmovdqa64 %0,%%zmm0"
+		     :: "m" (raid6_avx512_constants.x1d[0]));
+
+	for (d = 0 ; d < bytes ; d += 256) {
+		asm volatile("vmovdqa64 %0,%%zmm4\n\t"
+			     "vmovdqa64 %1,%%zmm6\n\t"
+			     "vmovdqa64 %2,%%zmm12\n\t"
+			     "vmovdqa64 %3,%%zmm14\n\t"
+			     "vmovdqa64 %4,%%zmm2\n\t"
+			     "vmovdqa64 %5,%%zmm3\n\t"
+			     "vmovdqa64 %6,%%zmm10\n\t"
+			     "vmovdqa64 %7,%%zmm11\n\t"
+			     "vpxorq %%zmm4,%%zmm2,%%zmm2\n\t"
+			     "vpxorq %%zmm6,%%zmm3,%%zmm3\n\t"
+			     "vpxorq %%zmm12,%%zmm10,%%zmm10\n\t"
+			     "vpxorq %%zmm14,%%zmm11,%%zmm11"
+			     :
+			     : "m" (dptr[z0][d]), "m" (dptr[z0][d+64]),
+			       "m" (dptr[z0][d+128]), "m" (dptr[z0][d+192]),
+			       "m" (p[d]), "m" (p[d+64]), "m" (p[d+128]),
+			       "m" (p[d+192]));
+		/* P/Q data pages */
+		for (z = z0-1 ; z >= start ; z--) {
+			asm volatile("vpxorq %%zmm5,%%zmm5,%%zmm5\n\t"
+				     "vpxorq %%zmm7,%%zmm7,%%zmm7\n\t"
+				     "vpxorq %%zmm13,%%zmm13,%%zmm13\n\t"
+				     "vpxorq %%zmm15,%%zmm15,%%zmm15\n\t"
+				     "prefetchnta %0\n\t"
+				     "prefetchnta %2\n\t"
+				     "vpcmpgtb %%zmm4,%%zmm5,%%k1\n\t"
+				     "vpcmpgtb %%zmm6,%%zmm7,%%k2\n\t"
+				     "vpcmpgtb %%zmm12,%%zmm13,%%k3\n\t"
+				     "vpcmpgtb %%zmm14,%%zmm15,%%k4\n\t"
+				     "vpmovm2b %%k1,%%zmm5\n\t"
+				     "vpmovm2b %%k2,%%zmm7\n\t"
+				     "vpmovm2b %%k3,%%zmm13\n\t"
+				     "vpmovm2b %%k4,%%zmm15\n\t"
+				     "vpaddb %%zmm4,%%zmm4,%%zmm4\n\t"
+				     "vpaddb %%zmm6,%%zmm6,%%zmm6\n\t"
+				     "vpaddb %%zmm12,%%zmm12,%%zmm12\n\t"
+				     "vpaddb %%Zmm14,%%zmm14,%%zmm14\n\t"
+				     "vpandq %%zmm0,%%zmm5,%%zmm5\n\t"
+				     "vpandq %%zmm0,%%zmm7,%%zmm7\n\t"
+				     "vpandq %%zmm0,%%zmm13,%%zmm13\n\t"
+				     "vpandq %%zmm0,%%zmm15,%%zmm15\n\t"
+				     "vpxorq %%zmm5,%%zmm4,%%zmm4\n\t"
+				     "vpxorq %%zmm7,%%zmm6,%%zmm6\n\t"
+				     "vpxorq %%zmm13,%%zmm12,%%zmm12\n\t"
+				     "vpxorq %%zmm15,%%zmm14,%%zmm14\n\t"
+				     "vmovdqa64 %0,%%zmm5\n\t"
+				     "vmovdqa64 %1,%%zmm7\n\t"
+				     "vmovdqa64 %2,%%zmm13\n\t"
+				     "vmovdqa64 %3,%%zmm15\n\t"
+				     "vpxorq %%zmm5,%%zmm2,%%zmm2\n\t"
+				     "vpxorq %%zmm7,%%zmm3,%%zmm3\n\t"
+				     "vpxorq %%zmm13,%%zmm10,%%zmm10\n\t"
+				     "vpxorq %%zmm15,%%zmm11,%%zmm11\n\t"
+				     "vpxorq %%zmm5,%%zmm4,%%zmm4\n\t"
+				     "vpxorq %%zmm7,%%zmm6,%%zmm6\n\t"
+				     "vpxorq %%zmm13,%%zmm12,%%zmm12\n\t"
+				     "vpxorq %%zmm15,%%zmm14,%%zmm14"
+				     :
+				     : "m" (dptr[z][d]), "m" (dptr[z][d+64]),
+				       "m" (dptr[z][d+128]),
+				       "m" (dptr[z][d+192]));
+		}
+		asm volatile("prefetchnta %0\n\t"
+			     "prefetchnta %1\n\t"
+			     :
+			     : "m" (q[d]), "m" (q[d+128]));
+		/* P/Q left side optimization */
+		for (z = start-1 ; z >= 0 ; z--) {
+			asm volatile("vpxorq %%zmm5,%%zmm5,%%zmm5\n\t"
+				     "vpxorq %%zmm7,%%zmm7,%%zmm7\n\t"
+				     "vpxorq %%zmm13,%%zmm13,%%zmm13\n\t"
+				     "vpxorq %%zmm15,%%zmm15,%%zmm15\n\t"
+				     "vpcmpgtb %%zmm4,%%zmm5,%%k1\n\t"
+				     "vpcmpgtb %%zmm6,%%zmm7,%%k2\n\t"
+				     "vpcmpgtb %%zmm12,%%zmm13,%%k3\n\t"
+				     "vpcmpgtb %%zmm14,%%zmm15,%%k4\n\t"
+				     "vpmovm2b %%k1,%%zmm5\n\t"
+				     "vpmovm2b %%k2,%%zmm7\n\t"
+				     "vpmovm2b %%k3,%%zmm13\n\t"
+				     "vpmovm2b %%k4,%%zmm15\n\t"
+				     "vpaddb %%zmm4,%%zmm4,%%zmm4\n\t"
+				     "vpaddb %%zmm6,%%zmm6,%%zmm6\n\t"
+				     "vpaddb %%zmm12,%%zmm12,%%zmm12\n\t"
+				     "vpaddb %%zmm14,%%zmm14,%%zmm14\n\t"
+				     "vpandq %%zmm0,%%zmm5,%%zmm5\n\t"
+				     "vpandq %%zmm0,%%zmm7,%%zmm7\n\t"
+				     "vpandq %%zmm0,%%zmm13,%%zmm13\n\t"
+				     "vpandq %%zmm0,%%zmm15,%%zmm15\n\t"
+				     "vpxorq %%zmm5,%%zmm4,%%zmm4\n\t"
+				     "vpxorq %%zmm7,%%zmm6,%%zmm6\n\t"
+				     "vpxorq %%zmm13,%%zmm12,%%zmm12\n\t"
+				     "vpxorq %%zmm15,%%zmm14,%%zmm14"
+				     :
+				     : );
+		}
+		asm volatile("vmovntdq %%zmm2,%0\n\t"
+			     "vmovntdq %%zmm3,%1\n\t"
+			     "vmovntdq %%zmm10,%2\n\t"
+			     "vmovntdq %%zmm11,%3\n\t"
+			     "vpxorq %4,%%zmm4,%%zmm4\n\t"
+			     "vpxorq %5,%%zmm6,%%zmm6\n\t"
+			     "vpxorq %6,%%zmm12,%%zmm12\n\t"
+			     "vpxorq %7,%%zmm14,%%zmm14\n\t"
+			     "vmovntdq %%zmm4,%4\n\t"
+			     "vmovntdq %%zmm6,%5\n\t"
+			     "vmovntdq %%zmm12,%6\n\t"
+			     "vmovntdq %%zmm14,%7"
+			     :
+			     : "m" (p[d]),  "m" (p[d+64]), "m" (p[d+128]),
+			       "m" (p[d+192]), "m" (q[d]),  "m" (q[d+64]),
+			       "m" (q[d+128]), "m" (q[d+192]));
+	}
+	asm volatile("sfence" : : : "memory");
+	kernel_fpu_end();
+}
+const struct raid6_calls raid6_avx512x4 = {
+	raid6_avx5124_gen_syndrome,
+	raid6_avx5124_xor_syndrome,
+	raid6_have_avx512,
+	"avx512x4",
+	.priority = 2		/* Prefer AVX512 over priority 1 (SSE2 and others) */
+};
+#endif
+
+#endif /* CONFIG_AS_AVX512 */
diff --git a/lib/raid6/int.uc b/lib/raid6/int.uc
new file mode 100644
index 000000000..558aeac93
--- /dev/null
+++ b/lib/raid6/int.uc
@@ -0,0 +1,156 @@
+/* -*- linux-c -*- ------------------------------------------------------- *
+ *
+ *   Copyright 2002-2004 H. Peter Anvin - All Rights Reserved
+ *
+ *   This program is free software; you can redistribute it and/or modify
+ *   it under the terms of the GNU General Public License as published by
+ *   the Free Software Foundation, Inc., 53 Temple Place Ste 330,
+ *   Boston MA 02111-1307, USA; either version 2 of the License, or
+ *   (at your option) any later version; incorporated herein by reference.
+ *
+ * ----------------------------------------------------------------------- */
+
+/*
+ * int$#.c
+ *
+ * $#-way unrolled portable integer math RAID-6 instruction set
+ *
+ * This file is postprocessed using unroll.awk
+ */
+
+#include <linux/raid/pq.h>
+
+/*
+ * This is the C data type to use
+ */
+
+/* Change this from BITS_PER_LONG if there is something better... */
+#if BITS_PER_LONG == 64
+# define NBYTES(x) ((x) * 0x0101010101010101UL)
+# define NSIZE  8
+# define NSHIFT 3
+# define NSTRING "64"
+typedef u64 unative_t;
+#else
+# define NBYTES(x) ((x) * 0x01010101U)
+# define NSIZE  4
+# define NSHIFT 2
+# define NSTRING "32"
+typedef u32 unative_t;
+#endif
+
+
+
+/*
+ * IA-64 wants insane amounts of unrolling.  On other architectures that
+ * is just a waste of space.
+ */
+#if ($# <= 8) || defined(__ia64__)
+
+
+/*
+ * These sub-operations are separate inlines since they can sometimes be
+ * specially optimized using architecture-specific hacks.
+ */
+
+/*
+ * The SHLBYTE() operation shifts each byte left by 1, *not*
+ * rolling over into the next byte
+ */
+static inline __attribute_const__ unative_t SHLBYTE(unative_t v)
+{
+	unative_t vv;
+
+	vv = (v << 1) & NBYTES(0xfe);
+	return vv;
+}
+
+/*
+ * The MASK() operation returns 0xFF in any byte for which the high
+ * bit is 1, 0x00 for any byte for which the high bit is 0.
+ */
+static inline __attribute_const__ unative_t MASK(unative_t v)
+{
+	unative_t vv;
+
+	vv = v & NBYTES(0x80);
+	vv = (vv << 1) - (vv >> 7); /* Overflow on the top bit is OK */
+	return vv;
+}
+
+
+static void raid6_int$#_gen_syndrome(int disks, size_t bytes, void **ptrs)
+{
+	u8 **dptr = (u8 **)ptrs;
+	u8 *p, *q;
+	int d, z, z0;
+
+	unative_t wd$$, wq$$, wp$$, w1$$, w2$$;
+
+	z0 = disks - 3;		/* Highest data disk */
+	p = dptr[z0+1];		/* XOR parity */
+	q = dptr[z0+2];		/* RS syndrome */
+
+	for ( d = 0 ; d < bytes ; d += NSIZE*$# ) {
+		wq$$ = wp$$ = *(unative_t *)&dptr[z0][d+$$*NSIZE];
+		for ( z = z0-1 ; z >= 0 ; z-- ) {
+			wd$$ = *(unative_t *)&dptr[z][d+$$*NSIZE];
+			wp$$ ^= wd$$;
+			w2$$ = MASK(wq$$);
+			w1$$ = SHLBYTE(wq$$);
+			w2$$ &= NBYTES(0x1d);
+			w1$$ ^= w2$$;
+			wq$$ = w1$$ ^ wd$$;
+		}
+		*(unative_t *)&p[d+NSIZE*$$] = wp$$;
+		*(unative_t *)&q[d+NSIZE*$$] = wq$$;
+	}
+}
+
+static void raid6_int$#_xor_syndrome(int disks, int start, int stop,
+				     size_t bytes, void **ptrs)
+{
+	u8 **dptr = (u8 **)ptrs;
+	u8 *p, *q;
+	int d, z, z0;
+
+	unative_t wd$$, wq$$, wp$$, w1$$, w2$$;
+
+	z0 = stop;		/* P/Q right side optimization */
+	p = dptr[disks-2];	/* XOR parity */
+	q = dptr[disks-1];	/* RS syndrome */
+
+	for ( d = 0 ; d < bytes ; d += NSIZE*$# ) {
+		/* P/Q data pages */
+		wq$$ = wp$$ = *(unative_t *)&dptr[z0][d+$$*NSIZE];
+		for ( z = z0-1 ; z >= start ; z-- ) {
+			wd$$ = *(unative_t *)&dptr[z][d+$$*NSIZE];
+			wp$$ ^= wd$$;
+			w2$$ = MASK(wq$$);
+			w1$$ = SHLBYTE(wq$$);
+			w2$$ &= NBYTES(0x1d);
+			w1$$ ^= w2$$;
+			wq$$ = w1$$ ^ wd$$;
+		}
+		/* P/Q left side optimization */
+		for ( z = start-1 ; z >= 0 ; z-- ) {
+			w2$$ = MASK(wq$$);
+			w1$$ = SHLBYTE(wq$$);
+			w2$$ &= NBYTES(0x1d);
+			wq$$ = w1$$ ^ w2$$;
+		}
+		*(unative_t *)&p[d+NSIZE*$$] ^= wp$$;
+		*(unative_t *)&q[d+NSIZE*$$] ^= wq$$;
+	}
+
+}
+
+const struct raid6_calls raid6_intx$# = {
+	raid6_int$#_gen_syndrome,
+	raid6_int$#_xor_syndrome,
+	NULL,			/* always valid */
+	"int" NSTRING "x$#",
+	0
+};
+
+#endif
diff --git a/lib/raid6/mktables.c b/lib/raid6/mktables.c
new file mode 100644
index 000000000..f02e10fa6
--- /dev/null
+++ b/lib/raid6/mktables.c
@@ -0,0 +1,175 @@
+// SPDX-License-Identifier: GPL-2.0-or-later
+/* -*- linux-c -*- ------------------------------------------------------- *
+ *
+ *   Copyright 2002-2007 H. Peter Anvin - All Rights Reserved
+ *
+ * ----------------------------------------------------------------------- */
+
+/*
+ * mktables.c
+ *
+ * Make RAID-6 tables.  This is a host user space program to be run at
+ * compile time.
+ */
+
+#include <stdio.h>
+#include <string.h>
+#include <inttypes.h>
+#include <stdlib.h>
+#include <time.h>
+
+static uint8_t gfmul(uint8_t a, uint8_t b)
+{
+	uint8_t v = 0;
+
+	while (b) {
+		if (b & 1)
+			v ^= a;
+		a = (a << 1) ^ (a & 0x80 ? 0x1d : 0);
+		b >>= 1;
+	}
+
+	return v;
+}
+
+static uint8_t gfpow(uint8_t a, int b)
+{
+	uint8_t v = 1;
+
+	b %= 255;
+	if (b < 0)
+		b += 255;
+
+	while (b) {
+		if (b & 1)
+			v = gfmul(v, a);
+		a = gfmul(a, a);
+		b >>= 1;
+	}
+
+	return v;
+}
+
+int main(int argc, char *argv[])
+{
+	int i, j, k;
+	uint8_t v;
+	uint8_t exptbl[256], invtbl[256];
+
+	printf("#include <linux/export.h>\n");
+	printf("#include <linux/raid/pq.h>\n");
+
+	/* Compute multiplication table */
+	printf("\nconst u8  __attribute__((aligned(256)))\n"
+		"raid6_gfmul[256][256] =\n"
+		"{\n");
+	for (i = 0; i < 256; i++) {
+		printf("\t{\n");
+		for (j = 0; j < 256; j += 8) {
+			printf("\t\t");
+			for (k = 0; k < 8; k++)
+				printf("0x%02x,%c", gfmul(i, j + k),
+				       (k == 7) ? '\n' : ' ');
+		}
+		printf("\t},\n");
+	}
+	printf("};\n");
+	printf("#ifdef __KERNEL__\n");
+	printf("EXPORT_SYMBOL(raid6_gfmul);\n");
+	printf("#endif\n");
+
+	/* Compute vector multiplication table */
+	printf("\nconst u8  __attribute__((aligned(256)))\n"
+		"raid6_vgfmul[256][32] =\n"
+		"{\n");
+	for (i = 0; i < 256; i++) {
+		printf("\t{\n");
+		for (j = 0; j < 16; j += 8) {
+			printf("\t\t");
+			for (k = 0; k < 8; k++)
+				printf("0x%02x,%c", gfmul(i, j + k),
+				       (k == 7) ? '\n' : ' ');
+		}
+		for (j = 0; j < 16; j += 8) {
+			printf("\t\t");
+			for (k = 0; k < 8; k++)
+				printf("0x%02x,%c", gfmul(i, (j + k) << 4),
+				       (k == 7) ? '\n' : ' ');
+		}
+		printf("\t},\n");
+	}
+	printf("};\n");
+	printf("#ifdef __KERNEL__\n");
+	printf("EXPORT_SYMBOL(raid6_vgfmul);\n");
+	printf("#endif\n");
+
+	/* Compute power-of-2 table (exponent) */
+	v = 1;
+	printf("\nconst u8 __attribute__((aligned(256)))\n"
+	       "raid6_gfexp[256] =\n" "{\n");
+	for (i = 0; i < 256; i += 8) {
+		printf("\t");
+		for (j = 0; j < 8; j++) {
+			exptbl[i + j] = v;
+			printf("0x%02x,%c", v, (j == 7) ? '\n' : ' ');
+			v = gfmul(v, 2);
+			if (v == 1)
+				v = 0;	/* For entry 255, not a real entry */
+		}
+	}
+	printf("};\n");
+	printf("#ifdef __KERNEL__\n");
+	printf("EXPORT_SYMBOL(raid6_gfexp);\n");
+	printf("#endif\n");
+
+	/* Compute log-of-2 table */
+	printf("\nconst u8 __attribute__((aligned(256)))\n"
+	       "raid6_gflog[256] =\n" "{\n");
+	for (i = 0; i < 256; i += 8) {
+		printf("\t");
+		for (j = 0; j < 8; j++) {
+			v = 255;
+			for (k = 0; k < 256; k++)
+				if (exptbl[k] == (i + j)) {
+					v = k;
+					break;
+				}
+			printf("0x%02x,%c", v, (j == 7) ? '\n' : ' ');
+		}
+	}
+	printf("};\n");
+	printf("#ifdef __KERNEL__\n");
+	printf("EXPORT_SYMBOL(raid6_gflog);\n");
+	printf("#endif\n");
+
+	/* Compute inverse table x^-1 == x^254 */
+	printf("\nconst u8 __attribute__((aligned(256)))\n"
+	       "raid6_gfinv[256] =\n" "{\n");
+	for (i = 0; i < 256; i += 8) {
+		printf("\t");
+		for (j = 0; j < 8; j++) {
+			invtbl[i + j] = v = gfpow(i + j, 254);
+			printf("0x%02x,%c", v, (j == 7) ? '\n' : ' ');
+		}
+	}
+	printf("};\n");
+	printf("#ifdef __KERNEL__\n");
+	printf("EXPORT_SYMBOL(raid6_gfinv);\n");
+	printf("#endif\n");
+
+	/* Compute inv(2^x + 1) (exponent-xor-inverse) table */
+	printf("\nconst u8 __attribute__((aligned(256)))\n"
+	       "raid6_gfexi[256] =\n" "{\n");
+	for (i = 0; i < 256; i += 8) {
+		printf("\t");
+		for (j = 0; j < 8; j++)
+			printf("0x%02x,%c", invtbl[exptbl[i + j] ^ 1],
+			       (j == 7) ? '\n' : ' ');
+	}
+	printf("};\n");
+	printf("#ifdef __KERNEL__\n");
+	printf("EXPORT_SYMBOL(raid6_gfexi);\n");
+	printf("#endif\n");
+
+	return 0;
+}
diff --git a/lib/raid6/mmx.c b/lib/raid6/mmx.c
new file mode 100644
index 000000000..3a5bf53a2
--- /dev/null
+++ b/lib/raid6/mmx.c
@@ -0,0 +1,139 @@
+// SPDX-License-Identifier: GPL-2.0-or-later
+/* -*- linux-c -*- ------------------------------------------------------- *
+ *
+ *   Copyright 2002 H. Peter Anvin - All Rights Reserved
+ *
+ * ----------------------------------------------------------------------- */
+
+/*
+ * raid6/mmx.c
+ *
+ * MMX implementation of RAID-6 syndrome functions
+ */
+
+#ifdef CONFIG_X86_32
+
+#include <linux/raid/pq.h>
+#include "x86.h"
+
+/* Shared with raid6/sse1.c */
+const struct raid6_mmx_constants {
+	u64 x1d;
+} raid6_mmx_constants = {
+	0x1d1d1d1d1d1d1d1dULL,
+};
+
+static int raid6_have_mmx(void)
+{
+	/* Not really "boot_cpu" but "all_cpus" */
+	return boot_cpu_has(X86_FEATURE_MMX);
+}
+
+/*
+ * Plain MMX implementation
+ */
+static void raid6_mmx1_gen_syndrome(int disks, size_t bytes, void **ptrs)
+{
+	u8 **dptr = (u8 **)ptrs;
+	u8 *p, *q;
+	int d, z, z0;
+
+	z0 = disks - 3;		/* Highest data disk */
+	p = dptr[z0+1];		/* XOR parity */
+	q = dptr[z0+2];		/* RS syndrome */
+
+	kernel_fpu_begin();
+
+	asm volatile("movq %0,%%mm0" : : "m" (raid6_mmx_constants.x1d));
+	asm volatile("pxor %mm5,%mm5");	/* Zero temp */
+
+	for ( d = 0 ; d < bytes ; d += 8 ) {
+		asm volatile("movq %0,%%mm2" : : "m" (dptr[z0][d])); /* P[0] */
+		asm volatile("movq %mm2,%mm4");	/* Q[0] */
+		for ( z = z0-1 ; z >= 0 ; z-- ) {
+			asm volatile("movq %0,%%mm6" : : "m" (dptr[z][d]));
+			asm volatile("pcmpgtb %mm4,%mm5");
+			asm volatile("paddb %mm4,%mm4");
+			asm volatile("pand %mm0,%mm5");
+			asm volatile("pxor %mm5,%mm4");
+			asm volatile("pxor %mm5,%mm5");
+			asm volatile("pxor %mm6,%mm2");
+			asm volatile("pxor %mm6,%mm4");
+		}
+		asm volatile("movq %%mm2,%0" : "=m" (p[d]));
+		asm volatile("pxor %mm2,%mm2");
+		asm volatile("movq %%mm4,%0" : "=m" (q[d]));
+		asm volatile("pxor %mm4,%mm4");
+	}
+
+	kernel_fpu_end();
+}
+
+const struct raid6_calls raid6_mmxx1 = {
+	raid6_mmx1_gen_syndrome,
+	NULL,			/* XOR not yet implemented */
+	raid6_have_mmx,
+	"mmxx1",
+	0
+};
+
+/*
+ * Unrolled-by-2 MMX implementation
+ */
+static void raid6_mmx2_gen_syndrome(int disks, size_t bytes, void **ptrs)
+{
+	u8 **dptr = (u8 **)ptrs;
+	u8 *p, *q;
+	int d, z, z0;
+
+	z0 = disks - 3;		/* Highest data disk */
+	p = dptr[z0+1];		/* XOR parity */
+	q = dptr[z0+2];		/* RS syndrome */
+
+	kernel_fpu_begin();
+
+	asm volatile("movq %0,%%mm0" : : "m" (raid6_mmx_constants.x1d));
+	asm volatile("pxor %mm5,%mm5");	/* Zero temp */
+	asm volatile("pxor %mm7,%mm7"); /* Zero temp */
+
+	for ( d = 0 ; d < bytes ; d += 16 ) {
+		asm volatile("movq %0,%%mm2" : : "m" (dptr[z0][d])); /* P[0] */
+		asm volatile("movq %0,%%mm3" : : "m" (dptr[z0][d+8]));
+		asm volatile("movq %mm2,%mm4"); /* Q[0] */
+		asm volatile("movq %mm3,%mm6"); /* Q[1] */
+		for ( z = z0-1 ; z >= 0 ; z-- ) {
+			asm volatile("pcmpgtb %mm4,%mm5");
+			asm volatile("pcmpgtb %mm6,%mm7");
+			asm volatile("paddb %mm4,%mm4");
+			asm volatile("paddb %mm6,%mm6");
+			asm volatile("pand %mm0,%mm5");
+			asm volatile("pand %mm0,%mm7");
+			asm volatile("pxor %mm5,%mm4");
+			asm volatile("pxor %mm7,%mm6");
+			asm volatile("movq %0,%%mm5" : : "m" (dptr[z][d]));
+			asm volatile("movq %0,%%mm7" : : "m" (dptr[z][d+8]));
+			asm volatile("pxor %mm5,%mm2");
+			asm volatile("pxor %mm7,%mm3");
+			asm volatile("pxor %mm5,%mm4");
+			asm volatile("pxor %mm7,%mm6");
+			asm volatile("pxor %mm5,%mm5");
+			asm volatile("pxor %mm7,%mm7");
+		}
+		asm volatile("movq %%mm2,%0" : "=m" (p[d]));
+		asm volatile("movq %%mm3,%0" : "=m" (p[d+8]));
+		asm volatile("movq %%mm4,%0" : "=m" (q[d]));
+		asm volatile("movq %%mm6,%0" : "=m" (q[d+8]));
+	}
+
+	kernel_fpu_end();
+}
+
+const struct raid6_calls raid6_mmxx2 = {
+	raid6_mmx2_gen_syndrome,
+	NULL,			/* XOR not yet implemented */
+	raid6_have_mmx,
+	"mmxx2",
+	0
+};
+
+#endif
diff --git a/lib/raid6/neon.c b/lib/raid6/neon.c
new file mode 100644
index 000000000..0a2e76035
--- /dev/null
+++ b/lib/raid6/neon.c
@@ -0,0 +1,67 @@
+// SPDX-License-Identifier: GPL-2.0-only
+/*
+ * linux/lib/raid6/neon.c - RAID6 syndrome calculation using ARM NEON intrinsics
+ *
+ * Copyright (C) 2013 Linaro Ltd <ard.biesheuvel@linaro.org>
+ */
+
+#include <linux/raid/pq.h>
+
+#ifdef __KERNEL__
+#include <asm/neon.h>
+#else
+#define kernel_neon_begin()
+#define kernel_neon_end()
+#define cpu_has_neon()		(1)
+#endif
+
+/*
+ * There are 2 reasons these wrappers are kept in a separate compilation unit
+ * from the actual implementations in neonN.c (generated from neon.uc by
+ * unroll.awk):
+ * - the actual implementations use NEON intrinsics, and the GCC support header
+ *   (arm_neon.h) is not fully compatible (type wise) with the kernel;
+ * - the neonN.c files are compiled with -mfpu=neon and optimization enabled,
+ *   and we have to make sure that we never use *any* NEON/VFP instructions
+ *   outside a kernel_neon_begin()/kernel_neon_end() pair.
+ */
+
+#define RAID6_NEON_WRAPPER(_n)						\
+	static void raid6_neon ## _n ## _gen_syndrome(int disks,	\
+					size_t bytes, void **ptrs)	\
+	{								\
+		void raid6_neon ## _n  ## _gen_syndrome_real(int,	\
+						unsigned long, void**);	\
+		kernel_neon_begin();					\
+		raid6_neon ## _n ## _gen_syndrome_real(disks,		\
+					(unsigned long)bytes, ptrs);	\
+		kernel_neon_end();					\
+	}								\
+	static void raid6_neon ## _n ## _xor_syndrome(int disks,	\
+					int start, int stop, 		\
+					size_t bytes, void **ptrs)	\
+	{								\
+		void raid6_neon ## _n  ## _xor_syndrome_real(int,	\
+				int, int, unsigned long, void**);	\
+		kernel_neon_begin();					\
+		raid6_neon ## _n ## _xor_syndrome_real(disks,		\
+			start, stop, (unsigned long)bytes, ptrs);	\
+		kernel_neon_end();					\
+	}								\
+	struct raid6_calls const raid6_neonx ## _n = {			\
+		raid6_neon ## _n ## _gen_syndrome,			\
+		raid6_neon ## _n ## _xor_syndrome,			\
+		raid6_have_neon,					\
+		"neonx" #_n,						\
+		0							\
+	}
+
+static int raid6_have_neon(void)
+{
+	return cpu_has_neon();
+}
+
+RAID6_NEON_WRAPPER(1);
+RAID6_NEON_WRAPPER(2);
+RAID6_NEON_WRAPPER(4);
+RAID6_NEON_WRAPPER(8);
diff --git a/lib/raid6/neon.uc b/lib/raid6/neon.uc
new file mode 100644
index 000000000..b7c68030d
--- /dev/null
+++ b/lib/raid6/neon.uc
@@ -0,0 +1,152 @@
+/* -----------------------------------------------------------------------
+ *
+ *   neon.uc - RAID-6 syndrome calculation using ARM NEON instructions
+ *
+ *   Copyright (C) 2012 Rob Herring
+ *   Copyright (C) 2015 Linaro Ltd. <ard.biesheuvel@linaro.org>
+ *
+ *   Based on altivec.uc:
+ *     Copyright 2002-2004 H. Peter Anvin - All Rights Reserved
+ *
+ *   This program is free software; you can redistribute it and/or modify
+ *   it under the terms of the GNU General Public License as published by
+ *   the Free Software Foundation, Inc., 53 Temple Place Ste 330,
+ *   Boston MA 02111-1307, USA; either version 2 of the License, or
+ *   (at your option) any later version; incorporated herein by reference.
+ *
+ * ----------------------------------------------------------------------- */
+
+/*
+ * neon$#.c
+ *
+ * $#-way unrolled NEON intrinsics math RAID-6 instruction set
+ *
+ * This file is postprocessed using unroll.awk
+ */
+
+#include <arm_neon.h>
+
+typedef uint8x16_t unative_t;
+
+#define NSIZE	sizeof(unative_t)
+
+/*
+ * The SHLBYTE() operation shifts each byte left by 1, *not*
+ * rolling over into the next byte
+ */
+static inline unative_t SHLBYTE(unative_t v)
+{
+	return vshlq_n_u8(v, 1);
+}
+
+/*
+ * The MASK() operation returns 0xFF in any byte for which the high
+ * bit is 1, 0x00 for any byte for which the high bit is 0.
+ */
+static inline unative_t MASK(unative_t v)
+{
+	return (unative_t)vshrq_n_s8((int8x16_t)v, 7);
+}
+
+static inline unative_t PMUL(unative_t v, unative_t u)
+{
+	return (unative_t)vmulq_p8((poly8x16_t)v, (poly8x16_t)u);
+}
+
+void raid6_neon$#_gen_syndrome_real(int disks, unsigned long bytes, void **ptrs)
+{
+	uint8_t **dptr = (uint8_t **)ptrs;
+	uint8_t *p, *q;
+	int d, z, z0;
+
+	register unative_t wd$$, wq$$, wp$$, w1$$, w2$$;
+	const unative_t x1d = vdupq_n_u8(0x1d);
+
+	z0 = disks - 3;		/* Highest data disk */
+	p = dptr[z0+1];		/* XOR parity */
+	q = dptr[z0+2];		/* RS syndrome */
+
+	for ( d = 0 ; d < bytes ; d += NSIZE*$# ) {
+		wq$$ = wp$$ = vld1q_u8(&dptr[z0][d+$$*NSIZE]);
+		for ( z = z0-1 ; z >= 0 ; z-- ) {
+			wd$$ = vld1q_u8(&dptr[z][d+$$*NSIZE]);
+			wp$$ = veorq_u8(wp$$, wd$$);
+			w2$$ = MASK(wq$$);
+			w1$$ = SHLBYTE(wq$$);
+
+			w2$$ = vandq_u8(w2$$, x1d);
+			w1$$ = veorq_u8(w1$$, w2$$);
+			wq$$ = veorq_u8(w1$$, wd$$);
+		}
+		vst1q_u8(&p[d+NSIZE*$$], wp$$);
+		vst1q_u8(&q[d+NSIZE*$$], wq$$);
+	}
+}
+
+void raid6_neon$#_xor_syndrome_real(int disks, int start, int stop,
+				    unsigned long bytes, void **ptrs)
+{
+	uint8_t **dptr = (uint8_t **)ptrs;
+	uint8_t *p, *q;
+	int d, z, z0;
+
+	register unative_t wd$$, wq$$, wp$$, w1$$, w2$$;
+	const unative_t x1d = vdupq_n_u8(0x1d);
+
+	z0 = stop;		/* P/Q right side optimization */
+	p = dptr[disks-2];	/* XOR parity */
+	q = dptr[disks-1];	/* RS syndrome */
+
+	for ( d = 0 ; d < bytes ; d += NSIZE*$# ) {
+		wq$$ = vld1q_u8(&dptr[z0][d+$$*NSIZE]);
+		wp$$ = veorq_u8(vld1q_u8(&p[d+$$*NSIZE]), wq$$);
+
+		/* P/Q data pages */
+		for ( z = z0-1 ; z >= start ; z-- ) {
+			wd$$ = vld1q_u8(&dptr[z][d+$$*NSIZE]);
+			wp$$ = veorq_u8(wp$$, wd$$);
+			w2$$ = MASK(wq$$);
+			w1$$ = SHLBYTE(wq$$);
+
+			w2$$ = vandq_u8(w2$$, x1d);
+			w1$$ = veorq_u8(w1$$, w2$$);
+			wq$$ = veorq_u8(w1$$, wd$$);
+		}
+		/* P/Q left side optimization */
+		for ( z = start-1 ; z >= 3 ; z -= 4 ) {
+			w2$$ = vshrq_n_u8(wq$$, 4);
+			w1$$ = vshlq_n_u8(wq$$, 4);
+
+			w2$$ = PMUL(w2$$, x1d);
+			wq$$ = veorq_u8(w1$$, w2$$);
+		}
+
+		switch (z) {
+		case 2:
+			w2$$ = vshrq_n_u8(wq$$, 5);
+			w1$$ = vshlq_n_u8(wq$$, 3);
+
+			w2$$ = PMUL(w2$$, x1d);
+			wq$$ = veorq_u8(w1$$, w2$$);
+			break;
+		case 1:
+			w2$$ = vshrq_n_u8(wq$$, 6);
+			w1$$ = vshlq_n_u8(wq$$, 2);
+
+			w2$$ = PMUL(w2$$, x1d);
+			wq$$ = veorq_u8(w1$$, w2$$);
+			break;
+		case 0:
+			w2$$ = MASK(wq$$);
+			w1$$ = SHLBYTE(wq$$);
+
+			w2$$ = vandq_u8(w2$$, x1d);
+			wq$$ = veorq_u8(w1$$, w2$$);
+		}
+		w1$$ = vld1q_u8(&q[d+NSIZE*$$]);
+		wq$$ = veorq_u8(wq$$, w1$$);
+
+		vst1q_u8(&p[d+NSIZE*$$], wp$$);
+		vst1q_u8(&q[d+NSIZE*$$], wq$$);
+	}
+}
diff --git a/lib/raid6/recov.c b/lib/raid6/recov.c
new file mode 100644
index 000000000..e49d519de
--- /dev/null
+++ b/lib/raid6/recov.c
@@ -0,0 +1,136 @@
+// SPDX-License-Identifier: GPL-2.0-or-later
+/* -*- linux-c -*- ------------------------------------------------------- *
+ *
+ *   Copyright 2002 H. Peter Anvin - All Rights Reserved
+ *
+ * ----------------------------------------------------------------------- */
+
+/*
+ * raid6/recov.c
+ *
+ * RAID-6 data recovery in dual failure mode.  In single failure mode,
+ * use the RAID-5 algorithm (or, in the case of Q failure, just reconstruct
+ * the syndrome.)
+ */
+
+#include <linux/export.h>
+#include <linux/raid/pq.h>
+
+/* Recover two failed data blocks. */
+static void raid6_2data_recov_intx1(int disks, size_t bytes, int faila,
+		int failb, void **ptrs)
+{
+	u8 *p, *q, *dp, *dq;
+	u8 px, qx, db;
+	const u8 *pbmul;	/* P multiplier table for B data */
+	const u8 *qmul;		/* Q multiplier table (for both) */
+
+	p = (u8 *)ptrs[disks-2];
+	q = (u8 *)ptrs[disks-1];
+
+	/* Compute syndrome with zero for the missing data pages
+	   Use the dead data pages as temporary storage for
+	   delta p and delta q */
+	dp = (u8 *)ptrs[faila];
+	ptrs[faila] = (void *)raid6_empty_zero_page;
+	ptrs[disks-2] = dp;
+	dq = (u8 *)ptrs[failb];
+	ptrs[failb] = (void *)raid6_empty_zero_page;
+	ptrs[disks-1] = dq;
+
+	raid6_call.gen_syndrome(disks, bytes, ptrs);
+
+	/* Restore pointer table */
+	ptrs[faila]   = dp;
+	ptrs[failb]   = dq;
+	ptrs[disks-2] = p;
+	ptrs[disks-1] = q;
+
+	/* Now, pick the proper data tables */
+	pbmul = raid6_gfmul[raid6_gfexi[failb-faila]];
+	qmul  = raid6_gfmul[raid6_gfinv[raid6_gfexp[faila]^raid6_gfexp[failb]]];
+
+	/* Now do it... */
+	while ( bytes-- ) {
+		px    = *p ^ *dp;
+		qx    = qmul[*q ^ *dq];
+		*dq++ = db = pbmul[px] ^ qx; /* Reconstructed B */
+		*dp++ = db ^ px; /* Reconstructed A */
+		p++; q++;
+	}
+}
+
+/* Recover failure of one data block plus the P block */
+static void raid6_datap_recov_intx1(int disks, size_t bytes, int faila,
+		void **ptrs)
+{
+	u8 *p, *q, *dq;
+	const u8 *qmul;		/* Q multiplier table */
+
+	p = (u8 *)ptrs[disks-2];
+	q = (u8 *)ptrs[disks-1];
+
+	/* Compute syndrome with zero for the missing data page
+	   Use the dead data page as temporary storage for delta q */
+	dq = (u8 *)ptrs[faila];
+	ptrs[faila] = (void *)raid6_empty_zero_page;
+	ptrs[disks-1] = dq;
+
+	raid6_call.gen_syndrome(disks, bytes, ptrs);
+
+	/* Restore pointer table */
+	ptrs[faila]   = dq;
+	ptrs[disks-1] = q;
+
+	/* Now, pick the proper data tables */
+	qmul  = raid6_gfmul[raid6_gfinv[raid6_gfexp[faila]]];
+
+	/* Now do it... */
+	while ( bytes-- ) {
+		*p++ ^= *dq = qmul[*q ^ *dq];
+		q++; dq++;
+	}
+}
+
+
+const struct raid6_recov_calls raid6_recov_intx1 = {
+	.data2 = raid6_2data_recov_intx1,
+	.datap = raid6_datap_recov_intx1,
+	.valid = NULL,
+	.name = "intx1",
+	.priority = 0,
+};
+
+#ifndef __KERNEL__
+/* Testing only */
+
+/* Recover two failed blocks. */
+void raid6_dual_recov(int disks, size_t bytes, int faila, int failb, void **ptrs)
+{
+	if ( faila > failb ) {
+		int tmp = faila;
+		faila = failb;
+		failb = tmp;
+	}
+
+	if ( failb == disks-1 ) {
+		if ( faila == disks-2 ) {
+			/* P+Q failure.  Just rebuild the syndrome. */
+			raid6_call.gen_syndrome(disks, bytes, ptrs);
+		} else {
+			/* data+Q failure.  Reconstruct data from P,
+			   then rebuild syndrome. */
+			/* NOT IMPLEMENTED - equivalent to RAID-5 */
+		}
+	} else {
+		if ( failb == disks-2 ) {
+			/* data+P failure. */
+			raid6_datap_recov(disks, bytes, faila, ptrs);
+		} else {
+			/* data+data failure. */
+			raid6_2data_recov(disks, bytes, faila, failb, ptrs);
+		}
+	}
+}
+
+#endif
diff --git a/lib/raid6/recov_avx2.c b/lib/raid6/recov_avx2.c
new file mode 100644
index 000000000..4e8095403
--- /dev/null
+++ b/lib/raid6/recov_avx2.c
@@ -0,0 +1,313 @@
+// SPDX-License-Identifier: GPL-2.0-only
+/*
+ * Copyright (C) 2012 Intel Corporation
+ * Author: Jim Kukunas <james.t.kukunas@linux.intel.com>
+ */
+
+#include <linux/raid/pq.h>
+#include "x86.h"
+
+static int raid6_has_avx2(void)
+{
+	return boot_cpu_has(X86_FEATURE_AVX2) &&
+		boot_cpu_has(X86_FEATURE_AVX);
+}
+
+static void raid6_2data_recov_avx2(int disks, size_t bytes, int faila,
+		int failb, void **ptrs)
+{
+	u8 *p, *q, *dp, *dq;
+	const u8 *pbmul;	/* P multiplier table for B data */
+	const u8 *qmul;		/* Q multiplier table (for both) */
+	const u8 x0f = 0x0f;
+
+	p = (u8 *)ptrs[disks-2];
+	q = (u8 *)ptrs[disks-1];
+
+	/* Compute syndrome with zero for the missing data pages
+	   Use the dead data pages as temporary storage for
+	   delta p and delta q */
+	dp = (u8 *)ptrs[faila];
+	ptrs[faila] = (void *)raid6_empty_zero_page;
+	ptrs[disks-2] = dp;
+	dq = (u8 *)ptrs[failb];
+	ptrs[failb] = (void *)raid6_empty_zero_page;
+	ptrs[disks-1] = dq;
+
+	raid6_call.gen_syndrome(disks, bytes, ptrs);
+
+	/* Restore pointer table */
+	ptrs[faila]   = dp;
+	ptrs[failb]   = dq;
+	ptrs[disks-2] = p;
+	ptrs[disks-1] = q;
+
+	/* Now, pick the proper data tables */
+	pbmul = raid6_vgfmul[raid6_gfexi[failb-faila]];
+	qmul  = raid6_vgfmul[raid6_gfinv[raid6_gfexp[faila] ^
+		raid6_gfexp[failb]]];
+
+	kernel_fpu_begin();
+
+	/* ymm0 = x0f[16] */
+	asm volatile("vpbroadcastb %0, %%ymm7" : : "m" (x0f));
+
+	while (bytes) {
+#ifdef CONFIG_X86_64
+		asm volatile("vmovdqa %0, %%ymm1" : : "m" (q[0]));
+		asm volatile("vmovdqa %0, %%ymm9" : : "m" (q[32]));
+		asm volatile("vmovdqa %0, %%ymm0" : : "m" (p[0]));
+		asm volatile("vmovdqa %0, %%ymm8" : : "m" (p[32]));
+		asm volatile("vpxor %0, %%ymm1, %%ymm1" : : "m" (dq[0]));
+		asm volatile("vpxor %0, %%ymm9, %%ymm9" : : "m" (dq[32]));
+		asm volatile("vpxor %0, %%ymm0, %%ymm0" : : "m" (dp[0]));
+		asm volatile("vpxor %0, %%ymm8, %%ymm8" : : "m" (dp[32]));
+
+		/*
+		 * 1 = dq[0]  ^ q[0]
+		 * 9 = dq[32] ^ q[32]
+		 * 0 = dp[0]  ^ p[0]
+		 * 8 = dp[32] ^ p[32]
+		 */
+
+		asm volatile("vbroadcasti128 %0, %%ymm4" : : "m" (qmul[0]));
+		asm volatile("vbroadcasti128 %0, %%ymm5" : : "m" (qmul[16]));
+
+		asm volatile("vpsraw $4, %ymm1, %ymm3");
+		asm volatile("vpsraw $4, %ymm9, %ymm12");
+		asm volatile("vpand %ymm7, %ymm1, %ymm1");
+		asm volatile("vpand %ymm7, %ymm9, %ymm9");
+		asm volatile("vpand %ymm7, %ymm3, %ymm3");
+		asm volatile("vpand %ymm7, %ymm12, %ymm12");
+		asm volatile("vpshufb %ymm9, %ymm4, %ymm14");
+		asm volatile("vpshufb %ymm1, %ymm4, %ymm4");
+		asm volatile("vpshufb %ymm12, %ymm5, %ymm15");
+		asm volatile("vpshufb %ymm3, %ymm5, %ymm5");
+		asm volatile("vpxor %ymm14, %ymm15, %ymm15");
+		asm volatile("vpxor %ymm4, %ymm5, %ymm5");
+
+		/*
+		 * 5 = qx[0]
+		 * 15 = qx[32]
+		 */
+
+		asm volatile("vbroadcasti128 %0, %%ymm4" : : "m" (pbmul[0]));
+		asm volatile("vbroadcasti128 %0, %%ymm1" : : "m" (pbmul[16]));
+		asm volatile("vpsraw $4, %ymm0, %ymm2");
+		asm volatile("vpsraw $4, %ymm8, %ymm6");
+		asm volatile("vpand %ymm7, %ymm0, %ymm3");
+		asm volatile("vpand %ymm7, %ymm8, %ymm14");
+		asm volatile("vpand %ymm7, %ymm2, %ymm2");
+		asm volatile("vpand %ymm7, %ymm6, %ymm6");
+		asm volatile("vpshufb %ymm14, %ymm4, %ymm12");
+		asm volatile("vpshufb %ymm3, %ymm4, %ymm4");
+		asm volatile("vpshufb %ymm6, %ymm1, %ymm13");
+		asm volatile("vpshufb %ymm2, %ymm1, %ymm1");
+		asm volatile("vpxor %ymm4, %ymm1, %ymm1");
+		asm volatile("vpxor %ymm12, %ymm13, %ymm13");
+
+		/*
+		 * 1  = pbmul[px[0]]
+		 * 13 = pbmul[px[32]]
+		 */
+		asm volatile("vpxor %ymm5, %ymm1, %ymm1");
+		asm volatile("vpxor %ymm15, %ymm13, %ymm13");
+
+		/*
+		 * 1 = db = DQ
+		 * 13 = db[32] = DQ[32]
+		 */
+		asm volatile("vmovdqa %%ymm1, %0" : "=m" (dq[0]));
+		asm volatile("vmovdqa %%ymm13,%0" : "=m" (dq[32]));
+		asm volatile("vpxor %ymm1, %ymm0, %ymm0");
+		asm volatile("vpxor %ymm13, %ymm8, %ymm8");
+
+		asm volatile("vmovdqa %%ymm0, %0" : "=m" (dp[0]));
+		asm volatile("vmovdqa %%ymm8, %0" : "=m" (dp[32]));
+
+		bytes -= 64;
+		p += 64;
+		q += 64;
+		dp += 64;
+		dq += 64;
+#else
+		asm volatile("vmovdqa %0, %%ymm1" : : "m" (*q));
+		asm volatile("vmovdqa %0, %%ymm0" : : "m" (*p));
+		asm volatile("vpxor %0, %%ymm1, %%ymm1" : : "m" (*dq));
+		asm volatile("vpxor %0, %%ymm0, %%ymm0" : : "m" (*dp));
+
+		/* 1 = dq ^ q;  0 = dp ^ p */
+
+		asm volatile("vbroadcasti128 %0, %%ymm4" : : "m" (qmul[0]));
+		asm volatile("vbroadcasti128 %0, %%ymm5" : : "m" (qmul[16]));
+
+		/*
+		 * 1 = dq ^ q
+		 * 3 = dq ^ p >> 4
+		 */
+		asm volatile("vpsraw $4, %ymm1, %ymm3");
+		asm volatile("vpand %ymm7, %ymm1, %ymm1");
+		asm volatile("vpand %ymm7, %ymm3, %ymm3");
+		asm volatile("vpshufb %ymm1, %ymm4, %ymm4");
+		asm volatile("vpshufb %ymm3, %ymm5, %ymm5");
+		asm volatile("vpxor %ymm4, %ymm5, %ymm5");
+
+		/* 5 = qx */
+
+		asm volatile("vbroadcasti128 %0, %%ymm4" : : "m" (pbmul[0]));
+		asm volatile("vbroadcasti128 %0, %%ymm1" : : "m" (pbmul[16]));
+
+		asm volatile("vpsraw $4, %ymm0, %ymm2");
+		asm volatile("vpand %ymm7, %ymm0, %ymm3");
+		asm volatile("vpand %ymm7, %ymm2, %ymm2");
+		asm volatile("vpshufb %ymm3, %ymm4, %ymm4");
+		asm volatile("vpshufb %ymm2, %ymm1, %ymm1");
+		asm volatile("vpxor %ymm4, %ymm1, %ymm1");
+
+		/* 1 = pbmul[px] */
+		asm volatile("vpxor %ymm5, %ymm1, %ymm1");
+		/* 1 = db = DQ */
+		asm volatile("vmovdqa %%ymm1, %0" : "=m" (dq[0]));
+
+		asm volatile("vpxor %ymm1, %ymm0, %ymm0");
+		asm volatile("vmovdqa %%ymm0, %0" : "=m" (dp[0]));
+
+		bytes -= 32;
+		p += 32;
+		q += 32;
+		dp += 32;
+		dq += 32;
+#endif
+	}
+
+	kernel_fpu_end();
+}
+
+static void raid6_datap_recov_avx2(int disks, size_t bytes, int faila,
+		void **ptrs)
+{
+	u8 *p, *q, *dq;
+	const u8 *qmul;		/* Q multiplier table */
+	const u8 x0f = 0x0f;
+
+	p = (u8 *)ptrs[disks-2];
+	q = (u8 *)ptrs[disks-1];
+
+	/* Compute syndrome with zero for the missing data page
+	   Use the dead data page as temporary storage for delta q */
+	dq = (u8 *)ptrs[faila];
+	ptrs[faila] = (void *)raid6_empty_zero_page;
+	ptrs[disks-1] = dq;
+
+	raid6_call.gen_syndrome(disks, bytes, ptrs);
+
+	/* Restore pointer table */
+	ptrs[faila]   = dq;
+	ptrs[disks-1] = q;
+
+	/* Now, pick the proper data tables */
+	qmul  = raid6_vgfmul[raid6_gfinv[raid6_gfexp[faila]]];
+
+	kernel_fpu_begin();
+
+	asm volatile("vpbroadcastb %0, %%ymm7" : : "m" (x0f));
+
+	while (bytes) {
+#ifdef CONFIG_X86_64
+		asm volatile("vmovdqa %0, %%ymm3" : : "m" (dq[0]));
+		asm volatile("vmovdqa %0, %%ymm8" : : "m" (dq[32]));
+		asm volatile("vpxor %0, %%ymm3, %%ymm3" : : "m" (q[0]));
+		asm volatile("vpxor %0, %%ymm8, %%ymm8" : : "m" (q[32]));
+
+		/*
+		 * 3 = q[0] ^ dq[0]
+		 * 8 = q[32] ^ dq[32]
+		 */
+		asm volatile("vbroadcasti128 %0, %%ymm0" : : "m" (qmul[0]));
+		asm volatile("vmovapd %ymm0, %ymm13");
+		asm volatile("vbroadcasti128 %0, %%ymm1" : : "m" (qmul[16]));
+		asm volatile("vmovapd %ymm1, %ymm14");
+
+		asm volatile("vpsraw $4, %ymm3, %ymm6");
+		asm volatile("vpsraw $4, %ymm8, %ymm12");
+		asm volatile("vpand %ymm7, %ymm3, %ymm3");
+		asm volatile("vpand %ymm7, %ymm8, %ymm8");
+		asm volatile("vpand %ymm7, %ymm6, %ymm6");
+		asm volatile("vpand %ymm7, %ymm12, %ymm12");
+		asm volatile("vpshufb %ymm3, %ymm0, %ymm0");
+		asm volatile("vpshufb %ymm8, %ymm13, %ymm13");
+		asm volatile("vpshufb %ymm6, %ymm1, %ymm1");
+		asm volatile("vpshufb %ymm12, %ymm14, %ymm14");
+		asm volatile("vpxor %ymm0, %ymm1, %ymm1");
+		asm volatile("vpxor %ymm13, %ymm14, %ymm14");
+
+		/*
+		 * 1  = qmul[q[0]  ^ dq[0]]
+		 * 14 = qmul[q[32] ^ dq[32]]
+		 */
+		asm volatile("vmovdqa %0, %%ymm2" : : "m" (p[0]));
+		asm volatile("vmovdqa %0, %%ymm12" : : "m" (p[32]));
+		asm volatile("vpxor %ymm1, %ymm2, %ymm2");
+		asm volatile("vpxor %ymm14, %ymm12, %ymm12");
+
+		/*
+		 * 2  = p[0]  ^ qmul[q[0]  ^ dq[0]]
+		 * 12 = p[32] ^ qmul[q[32] ^ dq[32]]
+		 */
+
+		asm volatile("vmovdqa %%ymm1, %0" : "=m" (dq[0]));
+		asm volatile("vmovdqa %%ymm14, %0" : "=m" (dq[32]));
+		asm volatile("vmovdqa %%ymm2, %0" : "=m" (p[0]));
+		asm volatile("vmovdqa %%ymm12,%0" : "=m" (p[32]));
+
+		bytes -= 64;
+		p += 64;
+		q += 64;
+		dq += 64;
+#else
+		asm volatile("vmovdqa %0, %%ymm3" : : "m" (dq[0]));
+		asm volatile("vpxor %0, %%ymm3, %%ymm3" : : "m" (q[0]));
+
+		/* 3 = q ^ dq */
+
+		asm volatile("vbroadcasti128 %0, %%ymm0" : : "m" (qmul[0]));
+		asm volatile("vbroadcasti128 %0, %%ymm1" : : "m" (qmul[16]));
+
+		asm volatile("vpsraw $4, %ymm3, %ymm6");
+		asm volatile("vpand %ymm7, %ymm3, %ymm3");
+		asm volatile("vpand %ymm7, %ymm6, %ymm6");
+		asm volatile("vpshufb %ymm3, %ymm0, %ymm0");
+		asm volatile("vpshufb %ymm6, %ymm1, %ymm1");
+		asm volatile("vpxor %ymm0, %ymm1, %ymm1");
+
+		/* 1 = qmul[q ^ dq] */
+
+		asm volatile("vmovdqa %0, %%ymm2" : : "m" (p[0]));
+		asm volatile("vpxor %ymm1, %ymm2, %ymm2");
+
+		/* 2 = p ^ qmul[q ^ dq] */
+
+		asm volatile("vmovdqa %%ymm1, %0" : "=m" (dq[0]));
+		asm volatile("vmovdqa %%ymm2, %0" : "=m" (p[0]));
+
+		bytes -= 32;
+		p += 32;
+		q += 32;
+		dq += 32;
+#endif
+	}
+
+	kernel_fpu_end();
+}
+
+const struct raid6_recov_calls raid6_recov_avx2 = {
+	.data2 = raid6_2data_recov_avx2,
+	.datap = raid6_datap_recov_avx2,
+	.valid = raid6_has_avx2,
+#ifdef CONFIG_X86_64
+	.name = "avx2x2",
+#else
+	.name = "avx2x1",
+#endif
+	.priority = 2,
+};
diff --git a/lib/raid6/recov_avx512.c b/lib/raid6/recov_avx512.c
new file mode 100644
index 000000000..fd9e15bf3
--- /dev/null
+++ b/lib/raid6/recov_avx512.c
@@ -0,0 +1,383 @@
+// SPDX-License-Identifier: GPL-2.0-only
+/*
+ * Copyright (C) 2016 Intel Corporation
+ *
+ * Author: Gayatri Kammela <gayatri.kammela@intel.com>
+ * Author: Megha Dey <megha.dey@linux.intel.com>
+ */
+
+#ifdef CONFIG_AS_AVX512
+
+#include <linux/raid/pq.h>
+#include "x86.h"
+
+static int raid6_has_avx512(void)
+{
+	return boot_cpu_has(X86_FEATURE_AVX2) &&
+		boot_cpu_has(X86_FEATURE_AVX) &&
+		boot_cpu_has(X86_FEATURE_AVX512F) &&
+		boot_cpu_has(X86_FEATURE_AVX512BW) &&
+		boot_cpu_has(X86_FEATURE_AVX512VL) &&
+		boot_cpu_has(X86_FEATURE_AVX512DQ);
+}
+
+static void raid6_2data_recov_avx512(int disks, size_t bytes, int faila,
+				     int failb, void **ptrs)
+{
+	u8 *p, *q, *dp, *dq;
+	const u8 *pbmul;	/* P multiplier table for B data */
+	const u8 *qmul;		/* Q multiplier table (for both) */
+	const u8 x0f = 0x0f;
+
+	p = (u8 *)ptrs[disks-2];
+	q = (u8 *)ptrs[disks-1];
+
+	/*
+	 * Compute syndrome with zero for the missing data pages
+	 * Use the dead data pages as temporary storage for
+	 * delta p and delta q
+	 */
+
+	dp = (u8 *)ptrs[faila];
+	ptrs[faila] = (void *)raid6_empty_zero_page;
+	ptrs[disks-2] = dp;
+	dq = (u8 *)ptrs[failb];
+	ptrs[failb] = (void *)raid6_empty_zero_page;
+	ptrs[disks-1] = dq;
+
+	raid6_call.gen_syndrome(disks, bytes, ptrs);
+
+	/* Restore pointer table */
+	ptrs[faila]   = dp;
+	ptrs[failb]   = dq;
+	ptrs[disks-2] = p;
+	ptrs[disks-1] = q;
+
+	/* Now, pick the proper data tables */
+	pbmul = raid6_vgfmul[raid6_gfexi[failb-faila]];
+	qmul  = raid6_vgfmul[raid6_gfinv[raid6_gfexp[faila] ^
+		raid6_gfexp[failb]]];
+
+	kernel_fpu_begin();
+
+	/* zmm0 = x0f[16] */
+	asm volatile("vpbroadcastb %0, %%zmm7" : : "m" (x0f));
+
+	while (bytes) {
+#ifdef CONFIG_X86_64
+		asm volatile("vmovdqa64 %0, %%zmm1\n\t"
+			     "vmovdqa64 %1, %%zmm9\n\t"
+			     "vmovdqa64 %2, %%zmm0\n\t"
+			     "vmovdqa64 %3, %%zmm8\n\t"
+			     "vpxorq %4, %%zmm1, %%zmm1\n\t"
+			     "vpxorq %5, %%zmm9, %%zmm9\n\t"
+			     "vpxorq %6, %%zmm0, %%zmm0\n\t"
+			     "vpxorq %7, %%zmm8, %%zmm8"
+			     :
+			     : "m" (q[0]), "m" (q[64]), "m" (p[0]),
+			       "m" (p[64]), "m" (dq[0]), "m" (dq[64]),
+			       "m" (dp[0]), "m" (dp[64]));
+
+		/*
+		 * 1 = dq[0]  ^ q[0]
+		 * 9 = dq[64] ^ q[64]
+		 * 0 = dp[0]  ^ p[0]
+		 * 8 = dp[64] ^ p[64]
+		 */
+
+		asm volatile("vbroadcasti64x2 %0, %%zmm4\n\t"
+			     "vbroadcasti64x2 %1, %%zmm5"
+			     :
+			     : "m" (qmul[0]), "m" (qmul[16]));
+
+		asm volatile("vpsraw $4, %%zmm1, %%zmm3\n\t"
+			     "vpsraw $4, %%zmm9, %%zmm12\n\t"
+			     "vpandq %%zmm7, %%zmm1, %%zmm1\n\t"
+			     "vpandq %%zmm7, %%zmm9, %%zmm9\n\t"
+			     "vpandq %%zmm7, %%zmm3, %%zmm3\n\t"
+			     "vpandq %%zmm7, %%zmm12, %%zmm12\n\t"
+			     "vpshufb %%zmm9, %%zmm4, %%zmm14\n\t"
+			     "vpshufb %%zmm1, %%zmm4, %%zmm4\n\t"
+			     "vpshufb %%zmm12, %%zmm5, %%zmm15\n\t"
+			     "vpshufb %%zmm3, %%zmm5, %%zmm5\n\t"
+			     "vpxorq %%zmm14, %%zmm15, %%zmm15\n\t"
+			     "vpxorq %%zmm4, %%zmm5, %%zmm5"
+			     :
+			     : );
+
+		/*
+		 * 5 = qx[0]
+		 * 15 = qx[64]
+		 */
+
+		asm volatile("vbroadcasti64x2 %0, %%zmm4\n\t"
+			     "vbroadcasti64x2 %1, %%zmm1\n\t"
+			     "vpsraw $4, %%zmm0, %%zmm2\n\t"
+			     "vpsraw $4, %%zmm8, %%zmm6\n\t"
+			     "vpandq %%zmm7, %%zmm0, %%zmm3\n\t"
+			     "vpandq %%zmm7, %%zmm8, %%zmm14\n\t"
+			     "vpandq %%zmm7, %%zmm2, %%zmm2\n\t"
+			     "vpandq %%zmm7, %%zmm6, %%zmm6\n\t"
+			     "vpshufb %%zmm14, %%zmm4, %%zmm12\n\t"
+			     "vpshufb %%zmm3, %%zmm4, %%zmm4\n\t"
+			     "vpshufb %%zmm6, %%zmm1, %%zmm13\n\t"
+			     "vpshufb %%zmm2, %%zmm1, %%zmm1\n\t"
+			     "vpxorq %%zmm4, %%zmm1, %%zmm1\n\t"
+			     "vpxorq %%zmm12, %%zmm13, %%zmm13"
+			     :
+			     : "m" (pbmul[0]), "m" (pbmul[16]));
+
+		/*
+		 * 1  = pbmul[px[0]]
+		 * 13 = pbmul[px[64]]
+		 */
+		asm volatile("vpxorq %%zmm5, %%zmm1, %%zmm1\n\t"
+			     "vpxorq %%zmm15, %%zmm13, %%zmm13"
+			     :
+			     : );
+
+		/*
+		 * 1 = db = DQ
+		 * 13 = db[64] = DQ[64]
+		 */
+		asm volatile("vmovdqa64 %%zmm1, %0\n\t"
+			     "vmovdqa64 %%zmm13,%1\n\t"
+			     "vpxorq %%zmm1, %%zmm0, %%zmm0\n\t"
+			     "vpxorq %%zmm13, %%zmm8, %%zmm8"
+			     :
+			     : "m" (dq[0]), "m" (dq[64]));
+
+		asm volatile("vmovdqa64 %%zmm0, %0\n\t"
+			     "vmovdqa64 %%zmm8, %1"
+			     :
+			     : "m" (dp[0]), "m" (dp[64]));
+
+		bytes -= 128;
+		p += 128;
+		q += 128;
+		dp += 128;
+		dq += 128;
+#else
+		asm volatile("vmovdqa64 %0, %%zmm1\n\t"
+			     "vmovdqa64 %1, %%zmm0\n\t"
+			     "vpxorq %2, %%zmm1, %%zmm1\n\t"
+			     "vpxorq %3, %%zmm0, %%zmm0"
+			     :
+			     : "m" (*q), "m" (*p), "m"(*dq), "m" (*dp));
+
+		/* 1 = dq ^ q;  0 = dp ^ p */
+
+		asm volatile("vbroadcasti64x2 %0, %%zmm4\n\t"
+			     "vbroadcasti64x2 %1, %%zmm5"
+			     :
+			     : "m" (qmul[0]), "m" (qmul[16]));
+
+		/*
+		 * 1 = dq ^ q
+		 * 3 = dq ^ p >> 4
+		 */
+		asm volatile("vpsraw $4, %%zmm1, %%zmm3\n\t"
+			     "vpandq %%zmm7, %%zmm1, %%zmm1\n\t"
+			     "vpandq %%zmm7, %%zmm3, %%zmm3\n\t"
+			     "vpshufb %%zmm1, %%zmm4, %%zmm4\n\t"
+			     "vpshufb %%zmm3, %%zmm5, %%zmm5\n\t"
+			     "vpxorq %%zmm4, %%zmm5, %%zmm5"
+			     :
+			     : );
+
+		/* 5 = qx */
+
+		asm volatile("vbroadcasti64x2 %0, %%zmm4\n\t"
+			     "vbroadcasti64x2 %1, %%zmm1"
+			     :
+			     : "m" (pbmul[0]), "m" (pbmul[16]));
+
+		asm volatile("vpsraw $4, %%zmm0, %%zmm2\n\t"
+			     "vpandq %%zmm7, %%zmm0, %%zmm3\n\t"
+			     "vpandq %%zmm7, %%zmm2, %%zmm2\n\t"
+			     "vpshufb %%zmm3, %%zmm4, %%zmm4\n\t"
+			     "vpshufb %%zmm2, %%zmm1, %%zmm1\n\t"
+			     "vpxorq %%zmm4, %%zmm1, %%zmm1"
+			     :
+			     : );
+
+		/* 1 = pbmul[px] */
+		asm volatile("vpxorq %%zmm5, %%zmm1, %%zmm1\n\t"
+			     /* 1 = db = DQ */
+			     "vmovdqa64 %%zmm1, %0\n\t"
+			     :
+			     : "m" (dq[0]));
+
+		asm volatile("vpxorq %%zmm1, %%zmm0, %%zmm0\n\t"
+			     "vmovdqa64 %%zmm0, %0"
+			     :
+			     : "m" (dp[0]));
+
+		bytes -= 64;
+		p += 64;
+		q += 64;
+		dp += 64;
+		dq += 64;
+#endif
+	}
+
+	kernel_fpu_end();
+}
+
+static void raid6_datap_recov_avx512(int disks, size_t bytes, int faila,
+				     void **ptrs)
+{
+	u8 *p, *q, *dq;
+	const u8 *qmul;		/* Q multiplier table */
+	const u8 x0f = 0x0f;
+
+	p = (u8 *)ptrs[disks-2];
+	q = (u8 *)ptrs[disks-1];
+
+	/*
+	 * Compute syndrome with zero for the missing data page
+	 * Use the dead data page as temporary storage for delta q
+	 */
+
+	dq = (u8 *)ptrs[faila];
+	ptrs[faila] = (void *)raid6_empty_zero_page;
+	ptrs[disks-1] = dq;
+
+	raid6_call.gen_syndrome(disks, bytes, ptrs);
+
+	/* Restore pointer table */
+	ptrs[faila]   = dq;
+	ptrs[disks-1] = q;
+
+	/* Now, pick the proper data tables */
+	qmul  = raid6_vgfmul[raid6_gfinv[raid6_gfexp[faila]]];
+
+	kernel_fpu_begin();
+
+	asm volatile("vpbroadcastb %0, %%zmm7" : : "m" (x0f));
+
+	while (bytes) {
+#ifdef CONFIG_X86_64
+		asm volatile("vmovdqa64 %0, %%zmm3\n\t"
+			     "vmovdqa64 %1, %%zmm8\n\t"
+			     "vpxorq %2, %%zmm3, %%zmm3\n\t"
+			     "vpxorq %3, %%zmm8, %%zmm8"
+			     :
+			     : "m" (dq[0]), "m" (dq[64]), "m" (q[0]),
+			       "m" (q[64]));
+
+		/*
+		 * 3 = q[0] ^ dq[0]
+		 * 8 = q[64] ^ dq[64]
+		 */
+		asm volatile("vbroadcasti64x2 %0, %%zmm0\n\t"
+			     "vmovapd %%zmm0, %%zmm13\n\t"
+			     "vbroadcasti64x2 %1, %%zmm1\n\t"
+			     "vmovapd %%zmm1, %%zmm14"
+			     :
+			     : "m" (qmul[0]), "m" (qmul[16]));
+
+		asm volatile("vpsraw $4, %%zmm3, %%zmm6\n\t"
+			     "vpsraw $4, %%zmm8, %%zmm12\n\t"
+			     "vpandq %%zmm7, %%zmm3, %%zmm3\n\t"
+			     "vpandq %%zmm7, %%zmm8, %%zmm8\n\t"
+			     "vpandq %%zmm7, %%zmm6, %%zmm6\n\t"
+			     "vpandq %%zmm7, %%zmm12, %%zmm12\n\t"
+			     "vpshufb %%zmm3, %%zmm0, %%zmm0\n\t"
+			     "vpshufb %%zmm8, %%zmm13, %%zmm13\n\t"
+			     "vpshufb %%zmm6, %%zmm1, %%zmm1\n\t"
+			     "vpshufb %%zmm12, %%zmm14, %%zmm14\n\t"
+			     "vpxorq %%zmm0, %%zmm1, %%zmm1\n\t"
+			     "vpxorq %%zmm13, %%zmm14, %%zmm14"
+			     :
+			     : );
+
+		/*
+		 * 1  = qmul[q[0]  ^ dq[0]]
+		 * 14 = qmul[q[64] ^ dq[64]]
+		 */
+		asm volatile("vmovdqa64 %0, %%zmm2\n\t"
+			     "vmovdqa64 %1, %%zmm12\n\t"
+			     "vpxorq %%zmm1, %%zmm2, %%zmm2\n\t"
+			     "vpxorq %%zmm14, %%zmm12, %%zmm12"
+			     :
+			     : "m" (p[0]), "m" (p[64]));
+
+		/*
+		 * 2  = p[0]  ^ qmul[q[0]  ^ dq[0]]
+		 * 12 = p[64] ^ qmul[q[64] ^ dq[64]]
+		 */
+
+		asm volatile("vmovdqa64 %%zmm1, %0\n\t"
+			     "vmovdqa64 %%zmm14, %1\n\t"
+			     "vmovdqa64 %%zmm2, %2\n\t"
+			     "vmovdqa64 %%zmm12,%3"
+			     :
+			     : "m" (dq[0]), "m" (dq[64]), "m" (p[0]),
+			       "m" (p[64]));
+
+		bytes -= 128;
+		p += 128;
+		q += 128;
+		dq += 128;
+#else
+		asm volatile("vmovdqa64 %0, %%zmm3\n\t"
+			     "vpxorq %1, %%zmm3, %%zmm3"
+			     :
+			     : "m" (dq[0]), "m" (q[0]));
+
+		/* 3 = q ^ dq */
+
+		asm volatile("vbroadcasti64x2 %0, %%zmm0\n\t"
+			     "vbroadcasti64x2 %1, %%zmm1"
+			     :
+			     : "m" (qmul[0]), "m" (qmul[16]));
+
+		asm volatile("vpsraw $4, %%zmm3, %%zmm6\n\t"
+			     "vpandq %%zmm7, %%zmm3, %%zmm3\n\t"
+			     "vpandq %%zmm7, %%zmm6, %%zmm6\n\t"
+			     "vpshufb %%zmm3, %%zmm0, %%zmm0\n\t"
+			     "vpshufb %%zmm6, %%zmm1, %%zmm1\n\t"
+			     "vpxorq %%zmm0, %%zmm1, %%zmm1"
+			     :
+			     : );
+
+		/* 1 = qmul[q ^ dq] */
+
+		asm volatile("vmovdqa64 %0, %%zmm2\n\t"
+			     "vpxorq %%zmm1, %%zmm2, %%zmm2"
+			     :
+			     : "m" (p[0]));
+
+		/* 2 = p ^ qmul[q ^ dq] */
+
+		asm volatile("vmovdqa64 %%zmm1, %0\n\t"
+			     "vmovdqa64 %%zmm2, %1"
+			     :
+			     : "m" (dq[0]), "m" (p[0]));
+
+		bytes -= 64;
+		p += 64;
+		q += 64;
+		dq += 64;
+#endif
+	}
+
+	kernel_fpu_end();
+}
+
+const struct raid6_recov_calls raid6_recov_avx512 = {
+	.data2 = raid6_2data_recov_avx512,
+	.datap = raid6_datap_recov_avx512,
+	.valid = raid6_has_avx512,
+#ifdef CONFIG_X86_64
+	.name = "avx512x2",
+#else
+	.name = "avx512x1",
+#endif
+	.priority = 3,
+};
+
+#else
+#warning "your version of binutils lacks AVX512 support"
+#endif
diff --git a/lib/raid6/recov_neon.c b/lib/raid6/recov_neon.c
new file mode 100644
index 000000000..d6fba8bf8
--- /dev/null
+++ b/lib/raid6/recov_neon.c
@@ -0,0 +1,106 @@
+// SPDX-License-Identifier: GPL-2.0-only
+/*
+ * Copyright (C) 2012 Intel Corporation
+ * Copyright (C) 2017 Linaro Ltd. <ard.biesheuvel@linaro.org>
+ */
+
+#include <linux/raid/pq.h>
+
+#ifdef __KERNEL__
+#include <asm/neon.h>
+#else
+#define kernel_neon_begin()
+#define kernel_neon_end()
+#define cpu_has_neon()		(1)
+#endif
+
+static int raid6_has_neon(void)
+{
+	return cpu_has_neon();
+}
+
+void __raid6_2data_recov_neon(int bytes, uint8_t *p, uint8_t *q, uint8_t *dp,
+			      uint8_t *dq, const uint8_t *pbmul,
+			      const uint8_t *qmul);
+
+void __raid6_datap_recov_neon(int bytes, uint8_t *p, uint8_t *q, uint8_t *dq,
+			      const uint8_t *qmul);
+
+static void raid6_2data_recov_neon(int disks, size_t bytes, int faila,
+		int failb, void **ptrs)
+{
+	u8 *p, *q, *dp, *dq;
+	const u8 *pbmul;	/* P multiplier table for B data */
+	const u8 *qmul;		/* Q multiplier table (for both) */
+
+	p = (u8 *)ptrs[disks - 2];
+	q = (u8 *)ptrs[disks - 1];
+
+	/*
+	 * Compute syndrome with zero for the missing data pages
+	 * Use the dead data pages as temporary storage for
+	 * delta p and delta q
+	 */
+	dp = (u8 *)ptrs[faila];
+	ptrs[faila] = (void *)raid6_empty_zero_page;
+	ptrs[disks - 2] = dp;
+	dq = (u8 *)ptrs[failb];
+	ptrs[failb] = (void *)raid6_empty_zero_page;
+	ptrs[disks - 1] = dq;
+
+	raid6_call.gen_syndrome(disks, bytes, ptrs);
+
+	/* Restore pointer table */
+	ptrs[faila]     = dp;
+	ptrs[failb]     = dq;
+	ptrs[disks - 2] = p;
+	ptrs[disks - 1] = q;
+
+	/* Now, pick the proper data tables */
+	pbmul = raid6_vgfmul[raid6_gfexi[failb-faila]];
+	qmul  = raid6_vgfmul[raid6_gfinv[raid6_gfexp[faila] ^
+					 raid6_gfexp[failb]]];
+
+	kernel_neon_begin();
+	__raid6_2data_recov_neon(bytes, p, q, dp, dq, pbmul, qmul);
+	kernel_neon_end();
+}
+
+static void raid6_datap_recov_neon(int disks, size_t bytes, int faila,
+		void **ptrs)
+{
+	u8 *p, *q, *dq;
+	const u8 *qmul;		/* Q multiplier table */
+
+	p = (u8 *)ptrs[disks - 2];
+	q = (u8 *)ptrs[disks - 1];
+
+	/*
+	 * Compute syndrome with zero for the missing data page
+	 * Use the dead data page as temporary storage for delta q
+	 */
+	dq = (u8 *)ptrs[faila];
+	ptrs[faila] = (void *)raid6_empty_zero_page;
+	ptrs[disks - 1] = dq;
+
+	raid6_call.gen_syndrome(disks, bytes, ptrs);
+
+	/* Restore pointer table */
+	ptrs[faila]     = dq;
+	ptrs[disks - 1] = q;
+
+	/* Now, pick the proper data tables */
+	qmul = raid6_vgfmul[raid6_gfinv[raid6_gfexp[faila]]];
+
+	kernel_neon_begin();
+	__raid6_datap_recov_neon(bytes, p, q, dq, qmul);
+	kernel_neon_end();
+}
+
+const struct raid6_recov_calls raid6_recov_neon = {
+	.data2		= raid6_2data_recov_neon,
+	.datap		= raid6_datap_recov_neon,
+	.valid		= raid6_has_neon,
+	.name		= "neon",
+	.priority	= 10,
+};
diff --git a/lib/raid6/recov_neon_inner.c b/lib/raid6/recov_neon_inner.c
new file mode 100644
index 000000000..90eb80d43
--- /dev/null
+++ b/lib/raid6/recov_neon_inner.c
@@ -0,0 +1,110 @@
+// SPDX-License-Identifier: GPL-2.0-only
+/*
+ * Copyright (C) 2012 Intel Corporation
+ * Copyright (C) 2017 Linaro Ltd. <ard.biesheuvel@linaro.org>
+ */
+
+#include <arm_neon.h>
+
+#ifdef CONFIG_ARM
+/*
+ * AArch32 does not provide this intrinsic natively because it does not
+ * implement the underlying instruction. AArch32 only provides a 64-bit
+ * wide vtbl.8 instruction, so use that instead.
+ */
+static uint8x16_t vqtbl1q_u8(uint8x16_t a, uint8x16_t b)
+{
+	union {
+		uint8x16_t	val;
+		uint8x8x2_t	pair;
+	} __a = { a };
+
+	return vcombine_u8(vtbl2_u8(__a.pair, vget_low_u8(b)),
+			   vtbl2_u8(__a.pair, vget_high_u8(b)));
+}
+#endif
+
+void __raid6_2data_recov_neon(int bytes, uint8_t *p, uint8_t *q, uint8_t *dp,
+			      uint8_t *dq, const uint8_t *pbmul,
+			      const uint8_t *qmul)
+{
+	uint8x16_t pm0 = vld1q_u8(pbmul);
+	uint8x16_t pm1 = vld1q_u8(pbmul + 16);
+	uint8x16_t qm0 = vld1q_u8(qmul);
+	uint8x16_t qm1 = vld1q_u8(qmul + 16);
+	uint8x16_t x0f = vdupq_n_u8(0x0f);
+
+	/*
+	 * while ( bytes-- ) {
+	 *	uint8_t px, qx, db;
+	 *
+	 *	px    = *p ^ *dp;
+	 *	qx    = qmul[*q ^ *dq];
+	 *	*dq++ = db = pbmul[px] ^ qx;
+	 *	*dp++ = db ^ px;
+	 *	p++; q++;
+	 * }
+	 */
+
+	while (bytes) {
+		uint8x16_t vx, vy, px, qx, db;
+
+		px = veorq_u8(vld1q_u8(p), vld1q_u8(dp));
+		vx = veorq_u8(vld1q_u8(q), vld1q_u8(dq));
+
+		vy = vshrq_n_u8(vx, 4);
+		vx = vqtbl1q_u8(qm0, vandq_u8(vx, x0f));
+		vy = vqtbl1q_u8(qm1, vy);
+		qx = veorq_u8(vx, vy);
+
+		vy = vshrq_n_u8(px, 4);
+		vx = vqtbl1q_u8(pm0, vandq_u8(px, x0f));
+		vy = vqtbl1q_u8(pm1, vy);
+		vx = veorq_u8(vx, vy);
+		db = veorq_u8(vx, qx);
+
+		vst1q_u8(dq, db);
+		vst1q_u8(dp, veorq_u8(db, px));
+
+		bytes -= 16;
+		p += 16;
+		q += 16;
+		dp += 16;
+		dq += 16;
+	}
+}
+
+void __raid6_datap_recov_neon(int bytes, uint8_t *p, uint8_t *q, uint8_t *dq,
+			      const uint8_t *qmul)
+{
+	uint8x16_t qm0 = vld1q_u8(qmul);
+	uint8x16_t qm1 = vld1q_u8(qmul + 16);
+	uint8x16_t x0f = vdupq_n_u8(0x0f);
+
+	/*
+	 * while (bytes--) {
+	 *	*p++ ^= *dq = qmul[*q ^ *dq];
+	 *	q++; dq++;
+	 * }
+	 */
+
+	while (bytes) {
+		uint8x16_t vx, vy;
+
+		vx = veorq_u8(vld1q_u8(q), vld1q_u8(dq));
+
+		vy = vshrq_n_u8(vx, 4);
+		vx = vqtbl1q_u8(qm0, vandq_u8(vx, x0f));
+		vy = vqtbl1q_u8(qm1, vy);
+		vx = veorq_u8(vx, vy);
+		vy = veorq_u8(vx, vld1q_u8(p));
+
+		vst1q_u8(dq, vx);
+		vst1q_u8(p, vy);
+
+		bytes -= 16;
+		p += 16;
+		q += 16;
+		dq += 16;
+	}
+}
diff --git a/lib/raid6/recov_s390xc.c b/lib/raid6/recov_s390xc.c
new file mode 100644
index 000000000..179eec900
--- /dev/null
+++ b/lib/raid6/recov_s390xc.c
@@ -0,0 +1,117 @@
+// SPDX-License-Identifier: GPL-2.0
+/*
+ * RAID-6 data recovery in dual failure mode based on the XC instruction.
+ *
+ * Copyright IBM Corp. 2016
+ * Author(s): Martin Schwidefsky <schwidefsky@de.ibm.com>
+ */
+
+#include <linux/export.h>
+#include <linux/raid/pq.h>
+
+static inline void xor_block(u8 *p1, u8 *p2)
+{
+	typedef struct { u8 _[256]; } addrtype;
+
+	asm volatile(
+		"	xc	0(256,%[p1]),0(%[p2])\n"
+		: "+m" (*(addrtype *) p1) : "m" (*(addrtype *) p2),
+		  [p1] "a" (p1), [p2] "a" (p2) : "cc");
+}
+
+/* Recover two failed data blocks. */
+static void raid6_2data_recov_s390xc(int disks, size_t bytes, int faila,
+		int failb, void **ptrs)
+{
+	u8 *p, *q, *dp, *dq;
+	const u8 *pbmul;	/* P multiplier table for B data */
+	const u8 *qmul;		/* Q multiplier table (for both) */
+	int i;
+
+	p = (u8 *)ptrs[disks-2];
+	q = (u8 *)ptrs[disks-1];
+
+	/* Compute syndrome with zero for the missing data pages
+	   Use the dead data pages as temporary storage for
+	   delta p and delta q */
+	dp = (u8 *)ptrs[faila];
+	ptrs[faila] = (void *)raid6_empty_zero_page;
+	ptrs[disks-2] = dp;
+	dq = (u8 *)ptrs[failb];
+	ptrs[failb] = (void *)raid6_empty_zero_page;
+	ptrs[disks-1] = dq;
+
+	raid6_call.gen_syndrome(disks, bytes, ptrs);
+
+	/* Restore pointer table */
+	ptrs[faila]   = dp;
+	ptrs[failb]   = dq;
+	ptrs[disks-2] = p;
+	ptrs[disks-1] = q;
+
+	/* Now, pick the proper data tables */
+	pbmul = raid6_gfmul[raid6_gfexi[failb-faila]];
+	qmul  = raid6_gfmul[raid6_gfinv[raid6_gfexp[faila]^raid6_gfexp[failb]]];
+
+	/* Now do it... */
+	while (bytes) {
+		xor_block(dp, p);
+		xor_block(dq, q);
+		for (i = 0; i < 256; i++)
+			dq[i] = pbmul[dp[i]] ^ qmul[dq[i]];
+		xor_block(dp, dq);
+		p += 256;
+		q += 256;
+		dp += 256;
+		dq += 256;
+		bytes -= 256;
+	}
+}
+
+/* Recover failure of one data block plus the P block */
+static void raid6_datap_recov_s390xc(int disks, size_t bytes, int faila,
+		void **ptrs)
+{
+	u8 *p, *q, *dq;
+	const u8 *qmul;		/* Q multiplier table */
+	int i;
+
+	p = (u8 *)ptrs[disks-2];
+	q = (u8 *)ptrs[disks-1];
+
+	/* Compute syndrome with zero for the missing data page
+	   Use the dead data page as temporary storage for delta q */
+	dq = (u8 *)ptrs[faila];
+	ptrs[faila] = (void *)raid6_empty_zero_page;
+	ptrs[disks-1] = dq;
+
+	raid6_call.gen_syndrome(disks, bytes, ptrs);
+
+	/* Restore pointer table */
+	ptrs[faila]   = dq;
+	ptrs[disks-1] = q;
+
+	/* Now, pick the proper data tables */
+	qmul  = raid6_gfmul[raid6_gfinv[raid6_gfexp[faila]]];
+
+	/* Now do it... */
+	while (bytes) {
+		xor_block(dq, q);
+		for (i = 0; i < 256; i++)
+			dq[i] = qmul[dq[i]];
+		xor_block(p, dq);
+		p += 256;
+		q += 256;
+		dq += 256;
+		bytes -= 256;
+	}
+}
+
+
+const struct raid6_recov_calls raid6_recov_s390xc = {
+	.data2 = raid6_2data_recov_s390xc,
+	.datap = raid6_datap_recov_s390xc,
+	.valid = NULL,
+	.name = "s390xc",
+	.priority = 1,
+};
diff --git a/lib/raid6/recov_ssse3.c b/lib/raid6/recov_ssse3.c
new file mode 100644
index 000000000..4bfa3c6b6
--- /dev/null
+++ b/lib/raid6/recov_ssse3.c
@@ -0,0 +1,328 @@
+// SPDX-License-Identifier: GPL-2.0-only
+/*
+ * Copyright (C) 2012 Intel Corporation
+ */
+
+#include <linux/raid/pq.h>
+#include "x86.h"
+
+static int raid6_has_ssse3(void)
+{
+	return boot_cpu_has(X86_FEATURE_XMM) &&
+		boot_cpu_has(X86_FEATURE_XMM2) &&
+		boot_cpu_has(X86_FEATURE_SSSE3);
+}
+
+static void raid6_2data_recov_ssse3(int disks, size_t bytes, int faila,
+		int failb, void **ptrs)
+{
+	u8 *p, *q, *dp, *dq;
+	const u8 *pbmul;	/* P multiplier table for B data */
+	const u8 *qmul;		/* Q multiplier table (for both) */
+	static const u8 __aligned(16) x0f[16] = {
+		 0x0f, 0x0f, 0x0f, 0x0f, 0x0f, 0x0f, 0x0f, 0x0f,
+		 0x0f, 0x0f, 0x0f, 0x0f, 0x0f, 0x0f, 0x0f, 0x0f};
+
+	p = (u8 *)ptrs[disks-2];
+	q = (u8 *)ptrs[disks-1];
+
+	/* Compute syndrome with zero for the missing data pages
+	   Use the dead data pages as temporary storage for
+	   delta p and delta q */
+	dp = (u8 *)ptrs[faila];
+	ptrs[faila] = (void *)raid6_empty_zero_page;
+	ptrs[disks-2] = dp;
+	dq = (u8 *)ptrs[failb];
+	ptrs[failb] = (void *)raid6_empty_zero_page;
+	ptrs[disks-1] = dq;
+
+	raid6_call.gen_syndrome(disks, bytes, ptrs);
+
+	/* Restore pointer table */
+	ptrs[faila]   = dp;
+	ptrs[failb]   = dq;
+	ptrs[disks-2] = p;
+	ptrs[disks-1] = q;
+
+	/* Now, pick the proper data tables */
+	pbmul = raid6_vgfmul[raid6_gfexi[failb-faila]];
+	qmul  = raid6_vgfmul[raid6_gfinv[raid6_gfexp[faila] ^
+		raid6_gfexp[failb]]];
+
+	kernel_fpu_begin();
+
+	asm volatile("movdqa %0,%%xmm7" : : "m" (x0f[0]));
+
+#ifdef CONFIG_X86_64
+	asm volatile("movdqa %0,%%xmm6" : : "m" (qmul[0]));
+	asm volatile("movdqa %0,%%xmm14" : : "m" (pbmul[0]));
+	asm volatile("movdqa %0,%%xmm15" : : "m" (pbmul[16]));
+#endif
+
+	/* Now do it... */
+	while (bytes) {
+#ifdef CONFIG_X86_64
+		/* xmm6, xmm14, xmm15 */
+
+		asm volatile("movdqa %0,%%xmm1" : : "m" (q[0]));
+		asm volatile("movdqa %0,%%xmm9" : : "m" (q[16]));
+		asm volatile("movdqa %0,%%xmm0" : : "m" (p[0]));
+		asm volatile("movdqa %0,%%xmm8" : : "m" (p[16]));
+		asm volatile("pxor   %0,%%xmm1" : : "m" (dq[0]));
+		asm volatile("pxor   %0,%%xmm9" : : "m" (dq[16]));
+		asm volatile("pxor   %0,%%xmm0" : : "m" (dp[0]));
+		asm volatile("pxor   %0,%%xmm8" : : "m" (dp[16]));
+
+		/* xmm0/8 = px */
+
+		asm volatile("movdqa %xmm6,%xmm4");
+		asm volatile("movdqa %0,%%xmm5" : : "m" (qmul[16]));
+		asm volatile("movdqa %xmm6,%xmm12");
+		asm volatile("movdqa %xmm5,%xmm13");
+		asm volatile("movdqa %xmm1,%xmm3");
+		asm volatile("movdqa %xmm9,%xmm11");
+		asm volatile("movdqa %xmm0,%xmm2"); /* xmm2/10 = px */
+		asm volatile("movdqa %xmm8,%xmm10");
+		asm volatile("psraw  $4,%xmm1");
+		asm volatile("psraw  $4,%xmm9");
+		asm volatile("pand   %xmm7,%xmm3");
+		asm volatile("pand   %xmm7,%xmm11");
+		asm volatile("pand   %xmm7,%xmm1");
+		asm volatile("pand   %xmm7,%xmm9");
+		asm volatile("pshufb %xmm3,%xmm4");
+		asm volatile("pshufb %xmm11,%xmm12");
+		asm volatile("pshufb %xmm1,%xmm5");
+		asm volatile("pshufb %xmm9,%xmm13");
+		asm volatile("pxor   %xmm4,%xmm5");
+		asm volatile("pxor   %xmm12,%xmm13");
+
+		/* xmm5/13 = qx */
+
+		asm volatile("movdqa %xmm14,%xmm4");
+		asm volatile("movdqa %xmm15,%xmm1");
+		asm volatile("movdqa %xmm14,%xmm12");
+		asm volatile("movdqa %xmm15,%xmm9");
+		asm volatile("movdqa %xmm2,%xmm3");
+		asm volatile("movdqa %xmm10,%xmm11");
+		asm volatile("psraw  $4,%xmm2");
+		asm volatile("psraw  $4,%xmm10");
+		asm volatile("pand   %xmm7,%xmm3");
+		asm volatile("pand   %xmm7,%xmm11");
+		asm volatile("pand   %xmm7,%xmm2");
+		asm volatile("pand   %xmm7,%xmm10");
+		asm volatile("pshufb %xmm3,%xmm4");
+		asm volatile("pshufb %xmm11,%xmm12");
+		asm volatile("pshufb %xmm2,%xmm1");
+		asm volatile("pshufb %xmm10,%xmm9");
+		asm volatile("pxor   %xmm4,%xmm1");
+		asm volatile("pxor   %xmm12,%xmm9");
+
+		/* xmm1/9 = pbmul[px] */
+		asm volatile("pxor   %xmm5,%xmm1");
+		asm volatile("pxor   %xmm13,%xmm9");
+		/* xmm1/9 = db = DQ */
+		asm volatile("movdqa %%xmm1,%0" : "=m" (dq[0]));
+		asm volatile("movdqa %%xmm9,%0" : "=m" (dq[16]));
+
+		asm volatile("pxor   %xmm1,%xmm0");
+		asm volatile("pxor   %xmm9,%xmm8");
+		asm volatile("movdqa %%xmm0,%0" : "=m" (dp[0]));
+		asm volatile("movdqa %%xmm8,%0" : "=m" (dp[16]));
+
+		bytes -= 32;
+		p += 32;
+		q += 32;
+		dp += 32;
+		dq += 32;
+#else
+		asm volatile("movdqa %0,%%xmm1" : : "m" (*q));
+		asm volatile("movdqa %0,%%xmm0" : : "m" (*p));
+		asm volatile("pxor   %0,%%xmm1" : : "m" (*dq));
+		asm volatile("pxor   %0,%%xmm0" : : "m" (*dp));
+
+		/* 1 = dq ^ q
+		 * 0 = dp ^ p
+		 */
+		asm volatile("movdqa %0,%%xmm4" : : "m" (qmul[0]));
+		asm volatile("movdqa %0,%%xmm5" : : "m" (qmul[16]));
+
+		asm volatile("movdqa %xmm1,%xmm3");
+		asm volatile("psraw  $4,%xmm1");
+		asm volatile("pand   %xmm7,%xmm3");
+		asm volatile("pand   %xmm7,%xmm1");
+		asm volatile("pshufb %xmm3,%xmm4");
+		asm volatile("pshufb %xmm1,%xmm5");
+		asm volatile("pxor   %xmm4,%xmm5");
+
+		asm volatile("movdqa %xmm0,%xmm2"); /* xmm2 = px */
+
+		/* xmm5 = qx */
+
+		asm volatile("movdqa %0,%%xmm4" : : "m" (pbmul[0]));
+		asm volatile("movdqa %0,%%xmm1" : : "m" (pbmul[16]));
+		asm volatile("movdqa %xmm2,%xmm3");
+		asm volatile("psraw  $4,%xmm2");
+		asm volatile("pand   %xmm7,%xmm3");
+		asm volatile("pand   %xmm7,%xmm2");
+		asm volatile("pshufb %xmm3,%xmm4");
+		asm volatile("pshufb %xmm2,%xmm1");
+		asm volatile("pxor   %xmm4,%xmm1");
+
+		/* xmm1 = pbmul[px] */
+		asm volatile("pxor   %xmm5,%xmm1");
+		/* xmm1 = db = DQ */
+		asm volatile("movdqa %%xmm1,%0" : "=m" (*dq));
+
+		asm volatile("pxor   %xmm1,%xmm0");
+		asm volatile("movdqa %%xmm0,%0" : "=m" (*dp));
+
+		bytes -= 16;
+		p += 16;
+		q += 16;
+		dp += 16;
+		dq += 16;
+#endif
+	}
+
+	kernel_fpu_end();
+}
+
+
+static void raid6_datap_recov_ssse3(int disks, size_t bytes, int faila,
+		void **ptrs)
+{
+	u8 *p, *q, *dq;
+	const u8 *qmul;		/* Q multiplier table */
+	static const u8 __aligned(16) x0f[16] = {
+		 0x0f, 0x0f, 0x0f, 0x0f, 0x0f, 0x0f, 0x0f, 0x0f,
+		 0x0f, 0x0f, 0x0f, 0x0f, 0x0f, 0x0f, 0x0f, 0x0f};
+
+	p = (u8 *)ptrs[disks-2];
+	q = (u8 *)ptrs[disks-1];
+
+	/* Compute syndrome with zero for the missing data page
+	   Use the dead data page as temporary storage for delta q */
+	dq = (u8 *)ptrs[faila];
+	ptrs[faila] = (void *)raid6_empty_zero_page;
+	ptrs[disks-1] = dq;
+
+	raid6_call.gen_syndrome(disks, bytes, ptrs);
+
+	/* Restore pointer table */
+	ptrs[faila]   = dq;
+	ptrs[disks-1] = q;
+
+	/* Now, pick the proper data tables */
+	qmul  = raid6_vgfmul[raid6_gfinv[raid6_gfexp[faila]]];
+
+	kernel_fpu_begin();
+
+	asm volatile("movdqa %0, %%xmm7" : : "m" (x0f[0]));
+
+	while (bytes) {
+#ifdef CONFIG_X86_64
+		asm volatile("movdqa %0, %%xmm3" : : "m" (dq[0]));
+		asm volatile("movdqa %0, %%xmm4" : : "m" (dq[16]));
+		asm volatile("pxor %0, %%xmm3" : : "m" (q[0]));
+		asm volatile("movdqa %0, %%xmm0" : : "m" (qmul[0]));
+
+		/* xmm3 = q[0] ^ dq[0] */
+
+		asm volatile("pxor %0, %%xmm4" : : "m" (q[16]));
+		asm volatile("movdqa %0, %%xmm1" : : "m" (qmul[16]));
+
+		/* xmm4 = q[16] ^ dq[16] */
+
+		asm volatile("movdqa %xmm3, %xmm6");
+		asm volatile("movdqa %xmm4, %xmm8");
+
+		/* xmm4 = xmm8 = q[16] ^ dq[16] */
+
+		asm volatile("psraw $4, %xmm3");
+		asm volatile("pand %xmm7, %xmm6");
+		asm volatile("pand %xmm7, %xmm3");
+		asm volatile("pshufb %xmm6, %xmm0");
+		asm volatile("pshufb %xmm3, %xmm1");
+		asm volatile("movdqa %0, %%xmm10" : : "m" (qmul[0]));
+		asm volatile("pxor %xmm0, %xmm1");
+		asm volatile("movdqa %0, %%xmm11" : : "m" (qmul[16]));
+
+		/* xmm1 = qmul[q[0] ^ dq[0]] */
+
+		asm volatile("psraw $4, %xmm4");
+		asm volatile("pand %xmm7, %xmm8");
+		asm volatile("pand %xmm7, %xmm4");
+		asm volatile("pshufb %xmm8, %xmm10");
+		asm volatile("pshufb %xmm4, %xmm11");
+		asm volatile("movdqa %0, %%xmm2" : : "m" (p[0]));
+		asm volatile("pxor %xmm10, %xmm11");
+		asm volatile("movdqa %0, %%xmm12" : : "m" (p[16]));
+
+		/* xmm11 = qmul[q[16] ^ dq[16]] */
+
+		asm volatile("pxor %xmm1, %xmm2");
+
+		/* xmm2 = p[0] ^ qmul[q[0] ^ dq[0]] */
+
+		asm volatile("pxor %xmm11, %xmm12");
+
+		/* xmm12 = p[16] ^ qmul[q[16] ^ dq[16]] */
+
+		asm volatile("movdqa %%xmm1, %0" : "=m" (dq[0]));
+		asm volatile("movdqa %%xmm11, %0" : "=m" (dq[16]));
+
+		asm volatile("movdqa %%xmm2, %0" : "=m" (p[0]));
+		asm volatile("movdqa %%xmm12, %0" : "=m" (p[16]));
+
+		bytes -= 32;
+		p += 32;
+		q += 32;
+		dq += 32;
+
+#else
+		asm volatile("movdqa %0, %%xmm3" : : "m" (dq[0]));
+		asm volatile("movdqa %0, %%xmm0" : : "m" (qmul[0]));
+		asm volatile("pxor %0, %%xmm3" : : "m" (q[0]));
+		asm volatile("movdqa %0, %%xmm1" : : "m" (qmul[16]));
+
+		/* xmm3 = *q ^ *dq */
+
+		asm volatile("movdqa %xmm3, %xmm6");
+		asm volatile("movdqa %0, %%xmm2" : : "m" (p[0]));
+		asm volatile("psraw $4, %xmm3");
+		asm volatile("pand %xmm7, %xmm6");
+		asm volatile("pand %xmm7, %xmm3");
+		asm volatile("pshufb %xmm6, %xmm0");
+		asm volatile("pshufb %xmm3, %xmm1");
+		asm volatile("pxor %xmm0, %xmm1");
+
+		/* xmm1 = qmul[*q ^ *dq */
+
+		asm volatile("pxor %xmm1, %xmm2");
+
+		/* xmm2 = *p ^ qmul[*q ^ *dq] */
+
+		asm volatile("movdqa %%xmm1, %0" : "=m" (dq[0]));
+		asm volatile("movdqa %%xmm2, %0" : "=m" (p[0]));
+
+		bytes -= 16;
+		p += 16;
+		q += 16;
+		dq += 16;
+#endif
+	}
+
+	kernel_fpu_end();
+}
+
+const struct raid6_recov_calls raid6_recov_ssse3 = {
+	.data2 = raid6_2data_recov_ssse3,
+	.datap = raid6_datap_recov_ssse3,
+	.valid = raid6_has_ssse3,
+#ifdef CONFIG_X86_64
+	.name = "ssse3x2",
+#else
+	.name = "ssse3x1",
+#endif
+	.priority = 1,
+};
diff --git a/lib/raid6/s390vx.uc b/lib/raid6/s390vx.uc
new file mode 100644
index 000000000..9e597e1f9
--- /dev/null
+++ b/lib/raid6/s390vx.uc
@@ -0,0 +1,171 @@
+// SPDX-License-Identifier: GPL-2.0
+/*
+ * raid6_vx$#.c
+ *
+ * $#-way unrolled RAID6 gen/xor functions for s390
+ * based on the vector facility
+ *
+ * Copyright IBM Corp. 2016
+ * Author(s): Martin Schwidefsky <schwidefsky@de.ibm.com>
+ *
+ * This file is postprocessed using unroll.awk.
+ */
+
+#include <linux/raid/pq.h>
+#include <asm/fpu/api.h>
+
+asm(".include \"asm/vx-insn.h\"\n");
+
+#define NSIZE 16
+
+static inline void LOAD_CONST(void)
+{
+	asm volatile("VREPIB %v24,7");
+	asm volatile("VREPIB %v25,0x1d");
+}
+
+/*
+ * The SHLBYTE() operation shifts each of the 16 bytes in
+ * vector register y left by 1 bit and stores the result in
+ * vector register x.
+ */
+static inline void SHLBYTE(int x, int y)
+{
+	asm volatile ("VAB %0,%1,%1" : : "i" (x), "i" (y));
+}
+
+/*
+ * For each of the 16 bytes in the vector register y the MASK()
+ * operation returns 0xFF if the high bit of the byte is 1,
+ * or 0x00 if the high bit is 0. The result is stored in vector
+ * register x.
+ */
+static inline void MASK(int x, int y)
+{
+	asm volatile ("VESRAVB	%0,%1,24" : : "i" (x), "i" (y));
+}
+
+static inline void AND(int x, int y, int z)
+{
+	asm volatile ("VN %0,%1,%2" : : "i" (x), "i" (y), "i" (z));
+}
+
+static inline void XOR(int x, int y, int z)
+{
+	asm volatile ("VX %0,%1,%2" : : "i" (x), "i" (y), "i" (z));
+}
+
+static inline void LOAD_DATA(int x, u8 *ptr)
+{
+	typedef struct { u8 _[16 * $#]; } addrtype;
+	register addrtype *__ptr asm("1") = (addrtype *) ptr;
+
+	asm volatile ("VLM %2,%3,0,%1"
+		      : : "m" (*__ptr), "a" (__ptr), "i" (x),
+			  "i" (x + $# - 1));
+}
+
+static inline void STORE_DATA(int x, u8 *ptr)
+{
+	typedef struct { u8 _[16 * $#]; } addrtype;
+	register addrtype *__ptr asm("1") = (addrtype *) ptr;
+
+	asm volatile ("VSTM %2,%3,0,1"
+		      : "=m" (*__ptr) : "a" (__ptr), "i" (x),
+			"i" (x + $# - 1));
+}
+
+static inline void COPY_VEC(int x, int y)
+{
+	asm volatile ("VLR %0,%1" : : "i" (x), "i" (y));
+}
+
+static void raid6_s390vx$#_gen_syndrome(int disks, size_t bytes, void **ptrs)
+{
+	struct kernel_fpu vxstate;
+	u8 **dptr, *p, *q;
+	int d, z, z0;
+
+	kernel_fpu_begin(&vxstate, KERNEL_VXR);
+	LOAD_CONST();
+
+	dptr = (u8 **) ptrs;
+	z0 = disks - 3;		/* Highest data disk */
+	p = dptr[z0 + 1];	/* XOR parity */
+	q = dptr[z0 + 2];	/* RS syndrome */
+
+	for (d = 0; d < bytes; d += $#*NSIZE) {
+		LOAD_DATA(0,&dptr[z0][d]);
+		COPY_VEC(8+$$,0+$$);
+		for (z = z0 - 1; z >= 0; z--) {
+			MASK(16+$$,8+$$);
+			AND(16+$$,16+$$,25);
+			SHLBYTE(8+$$,8+$$);
+			XOR(8+$$,8+$$,16+$$);
+			LOAD_DATA(16,&dptr[z][d]);
+			XOR(0+$$,0+$$,16+$$);
+			XOR(8+$$,8+$$,16+$$);
+		}
+		STORE_DATA(0,&p[d]);
+		STORE_DATA(8,&q[d]);
+	}
+	kernel_fpu_end(&vxstate, KERNEL_VXR);
+}
+
+static void raid6_s390vx$#_xor_syndrome(int disks, int start, int stop,
+					size_t bytes, void **ptrs)
+{
+	struct kernel_fpu vxstate;
+	u8 **dptr, *p, *q;
+	int d, z, z0;
+
+	dptr = (u8 **) ptrs;
+	z0 = stop;		/* P/Q right side optimization */
+	p = dptr[disks - 2];	/* XOR parity */
+	q = dptr[disks - 1];	/* RS syndrome */
+
+	kernel_fpu_begin(&vxstate, KERNEL_VXR);
+	LOAD_CONST();
+
+	for (d = 0; d < bytes; d += $#*NSIZE) {
+		/* P/Q data pages */
+		LOAD_DATA(0,&dptr[z0][d]);
+		COPY_VEC(8+$$,0+$$);
+		for (z = z0 - 1; z >= start; z--) {
+			MASK(16+$$,8+$$);
+			AND(16+$$,16+$$,25);
+			SHLBYTE(8+$$,8+$$);
+			XOR(8+$$,8+$$,16+$$);
+			LOAD_DATA(16,&dptr[z][d]);
+			XOR(0+$$,0+$$,16+$$);
+			XOR(8+$$,8+$$,16+$$);
+		}
+		/* P/Q left side optimization */
+		for (z = start - 1; z >= 0; z--) {
+			MASK(16+$$,8+$$);
+			AND(16+$$,16+$$,25);
+			SHLBYTE(8+$$,8+$$);
+			XOR(8+$$,8+$$,16+$$);
+		}
+		LOAD_DATA(16,&p[d]);
+		XOR(16+$$,16+$$,0+$$);
+		STORE_DATA(16,&p[d]);
+		LOAD_DATA(16,&q[d]);
+		XOR(16+$$,16+$$,8+$$);
+		STORE_DATA(16,&q[d]);
+	}
+	kernel_fpu_end(&vxstate, KERNEL_VXR);
+}
+
+static int raid6_s390vx$#_valid(void)
+{
+	return MACHINE_HAS_VX;
+}
+
+const struct raid6_calls raid6_s390vx$# = {
+	raid6_s390vx$#_gen_syndrome,
+	raid6_s390vx$#_xor_syndrome,
+	raid6_s390vx$#_valid,
+	"vx128x$#",
+	1
+};
diff --git a/lib/raid6/sse1.c b/lib/raid6/sse1.c
new file mode 100644
index 000000000..692fa3a93
--- /dev/null
+++ b/lib/raid6/sse1.c
@@ -0,0 +1,159 @@
+// SPDX-License-Identifier: GPL-2.0-or-later
+/* -*- linux-c -*- ------------------------------------------------------- *
+ *
+ *   Copyright 2002 H. Peter Anvin - All Rights Reserved
+ *
+ * ----------------------------------------------------------------------- */
+
+/*
+ * raid6/sse1.c
+ *
+ * SSE-1/MMXEXT implementation of RAID-6 syndrome functions
+ *
+ * This is really an MMX implementation, but it requires SSE-1 or
+ * AMD MMXEXT for prefetch support and a few other features.  The
+ * support for nontemporal memory accesses is enough to make this
+ * worthwhile as a separate implementation.
+ */
+
+#ifdef CONFIG_X86_32
+
+#include <linux/raid/pq.h>
+#include "x86.h"
+
+/* Defined in raid6/mmx.c */
+extern const struct raid6_mmx_constants {
+	u64 x1d;
+} raid6_mmx_constants;
+
+static int raid6_have_sse1_or_mmxext(void)
+{
+	/* Not really boot_cpu but "all_cpus" */
+	return boot_cpu_has(X86_FEATURE_MMX) &&
+		(boot_cpu_has(X86_FEATURE_XMM) ||
+		 boot_cpu_has(X86_FEATURE_MMXEXT));
+}
+
+/*
+ * Plain SSE1 implementation
+ */
+static void raid6_sse11_gen_syndrome(int disks, size_t bytes, void **ptrs)
+{
+	u8 **dptr = (u8 **)ptrs;
+	u8 *p, *q;
+	int d, z, z0;
+
+	z0 = disks - 3;		/* Highest data disk */
+	p = dptr[z0+1];		/* XOR parity */
+	q = dptr[z0+2];		/* RS syndrome */
+
+	kernel_fpu_begin();
+
+	asm volatile("movq %0,%%mm0" : : "m" (raid6_mmx_constants.x1d));
+	asm volatile("pxor %mm5,%mm5");	/* Zero temp */
+
+	for ( d = 0 ; d < bytes ; d += 8 ) {
+		asm volatile("prefetchnta %0" : : "m" (dptr[z0][d]));
+		asm volatile("movq %0,%%mm2" : : "m" (dptr[z0][d])); /* P[0] */
+		asm volatile("prefetchnta %0" : : "m" (dptr[z0-1][d]));
+		asm volatile("movq %mm2,%mm4");	/* Q[0] */
+		asm volatile("movq %0,%%mm6" : : "m" (dptr[z0-1][d]));
+		for ( z = z0-2 ; z >= 0 ; z-- ) {
+			asm volatile("prefetchnta %0" : : "m" (dptr[z][d]));
+			asm volatile("pcmpgtb %mm4,%mm5");
+			asm volatile("paddb %mm4,%mm4");
+			asm volatile("pand %mm0,%mm5");
+			asm volatile("pxor %mm5,%mm4");
+			asm volatile("pxor %mm5,%mm5");
+			asm volatile("pxor %mm6,%mm2");
+			asm volatile("pxor %mm6,%mm4");
+			asm volatile("movq %0,%%mm6" : : "m" (dptr[z][d]));
+		}
+		asm volatile("pcmpgtb %mm4,%mm5");
+		asm volatile("paddb %mm4,%mm4");
+		asm volatile("pand %mm0,%mm5");
+		asm volatile("pxor %mm5,%mm4");
+		asm volatile("pxor %mm5,%mm5");
+		asm volatile("pxor %mm6,%mm2");
+		asm volatile("pxor %mm6,%mm4");
+
+		asm volatile("movntq %%mm2,%0" : "=m" (p[d]));
+		asm volatile("movntq %%mm4,%0" : "=m" (q[d]));
+	}
+
+	asm volatile("sfence" : : : "memory");
+	kernel_fpu_end();
+}
+
+const struct raid6_calls raid6_sse1x1 = {
+	raid6_sse11_gen_syndrome,
+	NULL,			/* XOR not yet implemented */
+	raid6_have_sse1_or_mmxext,
+	"sse1x1",
+	1			/* Has cache hints */
+};
+
+/*
+ * Unrolled-by-2 SSE1 implementation
+ */
+static void raid6_sse12_gen_syndrome(int disks, size_t bytes, void **ptrs)
+{
+	u8 **dptr = (u8 **)ptrs;
+	u8 *p, *q;
+	int d, z, z0;
+
+	z0 = disks - 3;		/* Highest data disk */
+	p = dptr[z0+1];		/* XOR parity */
+	q = dptr[z0+2];		/* RS syndrome */
+
+	kernel_fpu_begin();
+
+	asm volatile("movq %0,%%mm0" : : "m" (raid6_mmx_constants.x1d));
+	asm volatile("pxor %mm5,%mm5");	/* Zero temp */
+	asm volatile("pxor %mm7,%mm7"); /* Zero temp */
+
+	/* We uniformly assume a single prefetch covers at least 16 bytes */
+	for ( d = 0 ; d < bytes ; d += 16 ) {
+		asm volatile("prefetchnta %0" : : "m" (dptr[z0][d]));
+		asm volatile("movq %0,%%mm2" : : "m" (dptr[z0][d])); /* P[0] */
+		asm volatile("movq %0,%%mm3" : : "m" (dptr[z0][d+8])); /* P[1] */
+		asm volatile("movq %mm2,%mm4");	/* Q[0] */
+		asm volatile("movq %mm3,%mm6"); /* Q[1] */
+		for ( z = z0-1 ; z >= 0 ; z-- ) {
+			asm volatile("prefetchnta %0" : : "m" (dptr[z][d]));
+			asm volatile("pcmpgtb %mm4,%mm5");
+			asm volatile("pcmpgtb %mm6,%mm7");
+			asm volatile("paddb %mm4,%mm4");
+			asm volatile("paddb %mm6,%mm6");
+			asm volatile("pand %mm0,%mm5");
+			asm volatile("pand %mm0,%mm7");
+			asm volatile("pxor %mm5,%mm4");
+			asm volatile("pxor %mm7,%mm6");
+			asm volatile("movq %0,%%mm5" : : "m" (dptr[z][d]));
+			asm volatile("movq %0,%%mm7" : : "m" (dptr[z][d+8]));
+			asm volatile("pxor %mm5,%mm2");
+			asm volatile("pxor %mm7,%mm3");
+			asm volatile("pxor %mm5,%mm4");
+			asm volatile("pxor %mm7,%mm6");
+			asm volatile("pxor %mm5,%mm5");
+			asm volatile("pxor %mm7,%mm7");
+		}
+		asm volatile("movntq %%mm2,%0" : "=m" (p[d]));
+		asm volatile("movntq %%mm3,%0" : "=m" (p[d+8]));
+		asm volatile("movntq %%mm4,%0" : "=m" (q[d]));
+		asm volatile("movntq %%mm6,%0" : "=m" (q[d+8]));
+	}
+
+	asm volatile("sfence" : :: "memory");
+	kernel_fpu_end();
+}
+
+const struct raid6_calls raid6_sse1x2 = {
+	raid6_sse12_gen_syndrome,
+	NULL,			/* XOR not yet implemented */
+	raid6_have_sse1_or_mmxext,
+	"sse1x2",
+	1			/* Has cache hints */
+};
+
+#endif
diff --git a/lib/raid6/sse2.c b/lib/raid6/sse2.c
new file mode 100644
index 000000000..293022024
--- /dev/null
+++ b/lib/raid6/sse2.c
@@ -0,0 +1,480 @@
+// SPDX-License-Identifier: GPL-2.0-or-later
+/* -*- linux-c -*- ------------------------------------------------------- *
+ *
+ *   Copyright 2002 H. Peter Anvin - All Rights Reserved
+ *
+ * ----------------------------------------------------------------------- */
+
+/*
+ * raid6/sse2.c
+ *
+ * SSE-2 implementation of RAID-6 syndrome functions
+ *
+ */
+
+#include <linux/raid/pq.h>
+#include "x86.h"
+
+static const struct raid6_sse_constants {
+	u64 x1d[2];
+} raid6_sse_constants  __attribute__((aligned(16))) = {
+	{ 0x1d1d1d1d1d1d1d1dULL, 0x1d1d1d1d1d1d1d1dULL },
+};
+
+static int raid6_have_sse2(void)
+{
+	/* Not really boot_cpu but "all_cpus" */
+	return boot_cpu_has(X86_FEATURE_MMX) &&
+		boot_cpu_has(X86_FEATURE_FXSR) &&
+		boot_cpu_has(X86_FEATURE_XMM) &&
+		boot_cpu_has(X86_FEATURE_XMM2);
+}
+
+/*
+ * Plain SSE2 implementation
+ */
+static void raid6_sse21_gen_syndrome(int disks, size_t bytes, void **ptrs)
+{
+	u8 **dptr = (u8 **)ptrs;
+	u8 *p, *q;
+	int d, z, z0;
+
+	z0 = disks - 3;		/* Highest data disk */
+	p = dptr[z0+1];		/* XOR parity */
+	q = dptr[z0+2];		/* RS syndrome */
+
+	kernel_fpu_begin();
+
+	asm volatile("movdqa %0,%%xmm0" : : "m" (raid6_sse_constants.x1d[0]));
+	asm volatile("pxor %xmm5,%xmm5");	/* Zero temp */
+
+	for ( d = 0 ; d < bytes ; d += 16 ) {
+		asm volatile("prefetchnta %0" : : "m" (dptr[z0][d]));
+		asm volatile("movdqa %0,%%xmm2" : : "m" (dptr[z0][d])); /* P[0] */
+		asm volatile("prefetchnta %0" : : "m" (dptr[z0-1][d]));
+		asm volatile("movdqa %xmm2,%xmm4"); /* Q[0] */
+		asm volatile("movdqa %0,%%xmm6" : : "m" (dptr[z0-1][d]));
+		for ( z = z0-2 ; z >= 0 ; z-- ) {
+			asm volatile("prefetchnta %0" : : "m" (dptr[z][d]));
+			asm volatile("pcmpgtb %xmm4,%xmm5");
+			asm volatile("paddb %xmm4,%xmm4");
+			asm volatile("pand %xmm0,%xmm5");
+			asm volatile("pxor %xmm5,%xmm4");
+			asm volatile("pxor %xmm5,%xmm5");
+			asm volatile("pxor %xmm6,%xmm2");
+			asm volatile("pxor %xmm6,%xmm4");
+			asm volatile("movdqa %0,%%xmm6" : : "m" (dptr[z][d]));
+		}
+		asm volatile("pcmpgtb %xmm4,%xmm5");
+		asm volatile("paddb %xmm4,%xmm4");
+		asm volatile("pand %xmm0,%xmm5");
+		asm volatile("pxor %xmm5,%xmm4");
+		asm volatile("pxor %xmm5,%xmm5");
+		asm volatile("pxor %xmm6,%xmm2");
+		asm volatile("pxor %xmm6,%xmm4");
+
+		asm volatile("movntdq %%xmm2,%0" : "=m" (p[d]));
+		asm volatile("pxor %xmm2,%xmm2");
+		asm volatile("movntdq %%xmm4,%0" : "=m" (q[d]));
+		asm volatile("pxor %xmm4,%xmm4");
+	}
+
+	asm volatile("sfence" : : : "memory");
+	kernel_fpu_end();
+}
+
+
+static void raid6_sse21_xor_syndrome(int disks, int start, int stop,
+				     size_t bytes, void **ptrs)
+{
+	u8 **dptr = (u8 **)ptrs;
+	u8 *p, *q;
+	int d, z, z0;
+
+	z0 = stop;		/* P/Q right side optimization */
+	p = dptr[disks-2];	/* XOR parity */
+	q = dptr[disks-1];	/* RS syndrome */
+
+	kernel_fpu_begin();
+
+	asm volatile("movdqa %0,%%xmm0" : : "m" (raid6_sse_constants.x1d[0]));
+
+	for ( d = 0 ; d < bytes ; d += 16 ) {
+		asm volatile("movdqa %0,%%xmm4" :: "m" (dptr[z0][d]));
+		asm volatile("movdqa %0,%%xmm2" : : "m" (p[d]));
+		asm volatile("pxor %xmm4,%xmm2");
+		/* P/Q data pages */
+		for ( z = z0-1 ; z >= start ; z-- ) {
+			asm volatile("pxor %xmm5,%xmm5");
+			asm volatile("pcmpgtb %xmm4,%xmm5");
+			asm volatile("paddb %xmm4,%xmm4");
+			asm volatile("pand %xmm0,%xmm5");
+			asm volatile("pxor %xmm5,%xmm4");
+			asm volatile("movdqa %0,%%xmm5" :: "m" (dptr[z][d]));
+			asm volatile("pxor %xmm5,%xmm2");
+			asm volatile("pxor %xmm5,%xmm4");
+		}
+		/* P/Q left side optimization */
+		for ( z = start-1 ; z >= 0 ; z-- ) {
+			asm volatile("pxor %xmm5,%xmm5");
+			asm volatile("pcmpgtb %xmm4,%xmm5");
+			asm volatile("paddb %xmm4,%xmm4");
+			asm volatile("pand %xmm0,%xmm5");
+			asm volatile("pxor %xmm5,%xmm4");
+		}
+		asm volatile("pxor %0,%%xmm4" : : "m" (q[d]));
+		/* Don't use movntdq for r/w memory area < cache line */
+		asm volatile("movdqa %%xmm4,%0" : "=m" (q[d]));
+		asm volatile("movdqa %%xmm2,%0" : "=m" (p[d]));
+	}
+
+	asm volatile("sfence" : : : "memory");
+	kernel_fpu_end();
+}
+
+const struct raid6_calls raid6_sse2x1 = {
+	raid6_sse21_gen_syndrome,
+	raid6_sse21_xor_syndrome,
+	raid6_have_sse2,
+	"sse2x1",
+	1			/* Has cache hints */
+};
+
+/*
+ * Unrolled-by-2 SSE2 implementation
+ */
+static void raid6_sse22_gen_syndrome(int disks, size_t bytes, void **ptrs)
+{
+	u8 **dptr = (u8 **)ptrs;
+	u8 *p, *q;
+	int d, z, z0;
+
+	z0 = disks - 3;		/* Highest data disk */
+	p = dptr[z0+1];		/* XOR parity */
+	q = dptr[z0+2];		/* RS syndrome */
+
+	kernel_fpu_begin();
+
+	asm volatile("movdqa %0,%%xmm0" : : "m" (raid6_sse_constants.x1d[0]));
+	asm volatile("pxor %xmm5,%xmm5"); /* Zero temp */
+	asm volatile("pxor %xmm7,%xmm7"); /* Zero temp */
+
+	/* We uniformly assume a single prefetch covers at least 32 bytes */
+	for ( d = 0 ; d < bytes ; d += 32 ) {
+		asm volatile("prefetchnta %0" : : "m" (dptr[z0][d]));
+		asm volatile("movdqa %0,%%xmm2" : : "m" (dptr[z0][d]));    /* P[0] */
+		asm volatile("movdqa %0,%%xmm3" : : "m" (dptr[z0][d+16])); /* P[1] */
+		asm volatile("movdqa %xmm2,%xmm4"); /* Q[0] */
+		asm volatile("movdqa %xmm3,%xmm6"); /* Q[1] */
+		for ( z = z0-1 ; z >= 0 ; z-- ) {
+			asm volatile("prefetchnta %0" : : "m" (dptr[z][d]));
+			asm volatile("pcmpgtb %xmm4,%xmm5");
+			asm volatile("pcmpgtb %xmm6,%xmm7");
+			asm volatile("paddb %xmm4,%xmm4");
+			asm volatile("paddb %xmm6,%xmm6");
+			asm volatile("pand %xmm0,%xmm5");
+			asm volatile("pand %xmm0,%xmm7");
+			asm volatile("pxor %xmm5,%xmm4");
+			asm volatile("pxor %xmm7,%xmm6");
+			asm volatile("movdqa %0,%%xmm5" : : "m" (dptr[z][d]));
+			asm volatile("movdqa %0,%%xmm7" : : "m" (dptr[z][d+16]));
+			asm volatile("pxor %xmm5,%xmm2");
+			asm volatile("pxor %xmm7,%xmm3");
+			asm volatile("pxor %xmm5,%xmm4");
+			asm volatile("pxor %xmm7,%xmm6");
+			asm volatile("pxor %xmm5,%xmm5");
+			asm volatile("pxor %xmm7,%xmm7");
+		}
+		asm volatile("movntdq %%xmm2,%0" : "=m" (p[d]));
+		asm volatile("movntdq %%xmm3,%0" : "=m" (p[d+16]));
+		asm volatile("movntdq %%xmm4,%0" : "=m" (q[d]));
+		asm volatile("movntdq %%xmm6,%0" : "=m" (q[d+16]));
+	}
+
+	asm volatile("sfence" : : : "memory");
+	kernel_fpu_end();
+}
+
+static void raid6_sse22_xor_syndrome(int disks, int start, int stop,
+				     size_t bytes, void **ptrs)
+{
+	u8 **dptr = (u8 **)ptrs;
+	u8 *p, *q;
+	int d, z, z0;
+
+	z0 = stop;		/* P/Q right side optimization */
+	p = dptr[disks-2];	/* XOR parity */
+	q = dptr[disks-1];	/* RS syndrome */
+
+	kernel_fpu_begin();
+
+	asm volatile("movdqa %0,%%xmm0" : : "m" (raid6_sse_constants.x1d[0]));
+
+	for ( d = 0 ; d < bytes ; d += 32 ) {
+		asm volatile("movdqa %0,%%xmm4" :: "m" (dptr[z0][d]));
+		asm volatile("movdqa %0,%%xmm6" :: "m" (dptr[z0][d+16]));
+		asm volatile("movdqa %0,%%xmm2" : : "m" (p[d]));
+		asm volatile("movdqa %0,%%xmm3" : : "m" (p[d+16]));
+		asm volatile("pxor %xmm4,%xmm2");
+		asm volatile("pxor %xmm6,%xmm3");
+		/* P/Q data pages */
+		for ( z = z0-1 ; z >= start ; z-- ) {
+			asm volatile("pxor %xmm5,%xmm5");
+			asm volatile("pxor %xmm7,%xmm7");
+			asm volatile("pcmpgtb %xmm4,%xmm5");
+			asm volatile("pcmpgtb %xmm6,%xmm7");
+			asm volatile("paddb %xmm4,%xmm4");
+			asm volatile("paddb %xmm6,%xmm6");
+			asm volatile("pand %xmm0,%xmm5");
+			asm volatile("pand %xmm0,%xmm7");
+			asm volatile("pxor %xmm5,%xmm4");
+			asm volatile("pxor %xmm7,%xmm6");
+			asm volatile("movdqa %0,%%xmm5" :: "m" (dptr[z][d]));
+			asm volatile("movdqa %0,%%xmm7" :: "m" (dptr[z][d+16]));
+			asm volatile("pxor %xmm5,%xmm2");
+			asm volatile("pxor %xmm7,%xmm3");
+			asm volatile("pxor %xmm5,%xmm4");
+			asm volatile("pxor %xmm7,%xmm6");
+		}
+		/* P/Q left side optimization */
+		for ( z = start-1 ; z >= 0 ; z-- ) {
+			asm volatile("pxor %xmm5,%xmm5");
+			asm volatile("pxor %xmm7,%xmm7");
+			asm volatile("pcmpgtb %xmm4,%xmm5");
+			asm volatile("pcmpgtb %xmm6,%xmm7");
+			asm volatile("paddb %xmm4,%xmm4");
+			asm volatile("paddb %xmm6,%xmm6");
+			asm volatile("pand %xmm0,%xmm5");
+			asm volatile("pand %xmm0,%xmm7");
+			asm volatile("pxor %xmm5,%xmm4");
+			asm volatile("pxor %xmm7,%xmm6");
+		}
+		asm volatile("pxor %0,%%xmm4" : : "m" (q[d]));
+		asm volatile("pxor %0,%%xmm6" : : "m" (q[d+16]));
+		/* Don't use movntdq for r/w memory area < cache line */
+		asm volatile("movdqa %%xmm4,%0" : "=m" (q[d]));
+		asm volatile("movdqa %%xmm6,%0" : "=m" (q[d+16]));
+		asm volatile("movdqa %%xmm2,%0" : "=m" (p[d]));
+		asm volatile("movdqa %%xmm3,%0" : "=m" (p[d+16]));
+	}
+
+	asm volatile("sfence" : : : "memory");
+	kernel_fpu_end();
+}
+
+const struct raid6_calls raid6_sse2x2 = {
+	raid6_sse22_gen_syndrome,
+	raid6_sse22_xor_syndrome,
+	raid6_have_sse2,
+	"sse2x2",
+	1			/* Has cache hints */
+};
+
+#ifdef CONFIG_X86_64
+
+/*
+ * Unrolled-by-4 SSE2 implementation
+ */
+static void raid6_sse24_gen_syndrome(int disks, size_t bytes, void **ptrs)
+{
+	u8 **dptr = (u8 **)ptrs;
+	u8 *p, *q;
+	int d, z, z0;
+
+	z0 = disks - 3;		/* Highest data disk */
+	p = dptr[z0+1];		/* XOR parity */
+	q = dptr[z0+2];		/* RS syndrome */
+
+	kernel_fpu_begin();
+
+	asm volatile("movdqa %0,%%xmm0" :: "m" (raid6_sse_constants.x1d[0]));
+	asm volatile("pxor %xmm2,%xmm2");	/* P[0] */
+	asm volatile("pxor %xmm3,%xmm3");	/* P[1] */
+	asm volatile("pxor %xmm4,%xmm4"); 	/* Q[0] */
+	asm volatile("pxor %xmm5,%xmm5");	/* Zero temp */
+	asm volatile("pxor %xmm6,%xmm6"); 	/* Q[1] */
+	asm volatile("pxor %xmm7,%xmm7"); 	/* Zero temp */
+	asm volatile("pxor %xmm10,%xmm10");	/* P[2] */
+	asm volatile("pxor %xmm11,%xmm11");	/* P[3] */
+	asm volatile("pxor %xmm12,%xmm12"); 	/* Q[2] */
+	asm volatile("pxor %xmm13,%xmm13");	/* Zero temp */
+	asm volatile("pxor %xmm14,%xmm14"); 	/* Q[3] */
+	asm volatile("pxor %xmm15,%xmm15"); 	/* Zero temp */
+
+	for ( d = 0 ; d < bytes ; d += 64 ) {
+		for ( z = z0 ; z >= 0 ; z-- ) {
+			/* The second prefetch seems to improve performance... */
+			asm volatile("prefetchnta %0" :: "m" (dptr[z][d]));
+			asm volatile("prefetchnta %0" :: "m" (dptr[z][d+32]));
+			asm volatile("pcmpgtb %xmm4,%xmm5");
+			asm volatile("pcmpgtb %xmm6,%xmm7");
+			asm volatile("pcmpgtb %xmm12,%xmm13");
+			asm volatile("pcmpgtb %xmm14,%xmm15");
+			asm volatile("paddb %xmm4,%xmm4");
+			asm volatile("paddb %xmm6,%xmm6");
+			asm volatile("paddb %xmm12,%xmm12");
+			asm volatile("paddb %xmm14,%xmm14");
+			asm volatile("pand %xmm0,%xmm5");
+			asm volatile("pand %xmm0,%xmm7");
+			asm volatile("pand %xmm0,%xmm13");
+			asm volatile("pand %xmm0,%xmm15");
+			asm volatile("pxor %xmm5,%xmm4");
+			asm volatile("pxor %xmm7,%xmm6");
+			asm volatile("pxor %xmm13,%xmm12");
+			asm volatile("pxor %xmm15,%xmm14");
+			asm volatile("movdqa %0,%%xmm5" :: "m" (dptr[z][d]));
+			asm volatile("movdqa %0,%%xmm7" :: "m" (dptr[z][d+16]));
+			asm volatile("movdqa %0,%%xmm13" :: "m" (dptr[z][d+32]));
+			asm volatile("movdqa %0,%%xmm15" :: "m" (dptr[z][d+48]));
+			asm volatile("pxor %xmm5,%xmm2");
+			asm volatile("pxor %xmm7,%xmm3");
+			asm volatile("pxor %xmm13,%xmm10");
+			asm volatile("pxor %xmm15,%xmm11");
+			asm volatile("pxor %xmm5,%xmm4");
+			asm volatile("pxor %xmm7,%xmm6");
+			asm volatile("pxor %xmm13,%xmm12");
+			asm volatile("pxor %xmm15,%xmm14");
+			asm volatile("pxor %xmm5,%xmm5");
+			asm volatile("pxor %xmm7,%xmm7");
+			asm volatile("pxor %xmm13,%xmm13");
+			asm volatile("pxor %xmm15,%xmm15");
+		}
+		asm volatile("movntdq %%xmm2,%0" : "=m" (p[d]));
+		asm volatile("pxor %xmm2,%xmm2");
+		asm volatile("movntdq %%xmm3,%0" : "=m" (p[d+16]));
+		asm volatile("pxor %xmm3,%xmm3");
+		asm volatile("movntdq %%xmm10,%0" : "=m" (p[d+32]));
+		asm volatile("pxor %xmm10,%xmm10");
+		asm volatile("movntdq %%xmm11,%0" : "=m" (p[d+48]));
+		asm volatile("pxor %xmm11,%xmm11");
+		asm volatile("movntdq %%xmm4,%0" : "=m" (q[d]));
+		asm volatile("pxor %xmm4,%xmm4");
+		asm volatile("movntdq %%xmm6,%0" : "=m" (q[d+16]));
+		asm volatile("pxor %xmm6,%xmm6");
+		asm volatile("movntdq %%xmm12,%0" : "=m" (q[d+32]));
+		asm volatile("pxor %xmm12,%xmm12");
+		asm volatile("movntdq %%xmm14,%0" : "=m" (q[d+48]));
+		asm volatile("pxor %xmm14,%xmm14");
+	}
+
+	asm volatile("sfence" : : : "memory");
+	kernel_fpu_end();
+}
+
+static void raid6_sse24_xor_syndrome(int disks, int start, int stop,
+				     size_t bytes, void **ptrs)
+{
+	u8 **dptr = (u8 **)ptrs;
+	u8 *p, *q;
+	int d, z, z0;
+
+	z0 = stop;		/* P/Q right side optimization */
+	p = dptr[disks-2];	/* XOR parity */
+	q = dptr[disks-1];	/* RS syndrome */
+
+	kernel_fpu_begin();
+
+	asm volatile("movdqa %0,%%xmm0" :: "m" (raid6_sse_constants.x1d[0]));
+
+	for ( d = 0 ; d < bytes ; d += 64 ) {
+		asm volatile("movdqa %0,%%xmm4" :: "m" (dptr[z0][d]));
+		asm volatile("movdqa %0,%%xmm6" :: "m" (dptr[z0][d+16]));
+		asm volatile("movdqa %0,%%xmm12" :: "m" (dptr[z0][d+32]));
+		asm volatile("movdqa %0,%%xmm14" :: "m" (dptr[z0][d+48]));
+		asm volatile("movdqa %0,%%xmm2" : : "m" (p[d]));
+		asm volatile("movdqa %0,%%xmm3" : : "m" (p[d+16]));
+		asm volatile("movdqa %0,%%xmm10" : : "m" (p[d+32]));
+		asm volatile("movdqa %0,%%xmm11" : : "m" (p[d+48]));
+		asm volatile("pxor %xmm4,%xmm2");
+		asm volatile("pxor %xmm6,%xmm3");
+		asm volatile("pxor %xmm12,%xmm10");
+		asm volatile("pxor %xmm14,%xmm11");
+		/* P/Q data pages */
+		for ( z = z0-1 ; z >= start ; z-- ) {
+			asm volatile("prefetchnta %0" :: "m" (dptr[z][d]));
+			asm volatile("prefetchnta %0" :: "m" (dptr[z][d+32]));
+			asm volatile("pxor %xmm5,%xmm5");
+			asm volatile("pxor %xmm7,%xmm7");
+			asm volatile("pxor %xmm13,%xmm13");
+			asm volatile("pxor %xmm15,%xmm15");
+			asm volatile("pcmpgtb %xmm4,%xmm5");
+			asm volatile("pcmpgtb %xmm6,%xmm7");
+			asm volatile("pcmpgtb %xmm12,%xmm13");
+			asm volatile("pcmpgtb %xmm14,%xmm15");
+			asm volatile("paddb %xmm4,%xmm4");
+			asm volatile("paddb %xmm6,%xmm6");
+			asm volatile("paddb %xmm12,%xmm12");
+			asm volatile("paddb %xmm14,%xmm14");
+			asm volatile("pand %xmm0,%xmm5");
+			asm volatile("pand %xmm0,%xmm7");
+			asm volatile("pand %xmm0,%xmm13");
+			asm volatile("pand %xmm0,%xmm15");
+			asm volatile("pxor %xmm5,%xmm4");
+			asm volatile("pxor %xmm7,%xmm6");
+			asm volatile("pxor %xmm13,%xmm12");
+			asm volatile("pxor %xmm15,%xmm14");
+			asm volatile("movdqa %0,%%xmm5" :: "m" (dptr[z][d]));
+			asm volatile("movdqa %0,%%xmm7" :: "m" (dptr[z][d+16]));
+			asm volatile("movdqa %0,%%xmm13" :: "m" (dptr[z][d+32]));
+			asm volatile("movdqa %0,%%xmm15" :: "m" (dptr[z][d+48]));
+			asm volatile("pxor %xmm5,%xmm2");
+			asm volatile("pxor %xmm7,%xmm3");
+			asm volatile("pxor %xmm13,%xmm10");
+			asm volatile("pxor %xmm15,%xmm11");
+			asm volatile("pxor %xmm5,%xmm4");
+			asm volatile("pxor %xmm7,%xmm6");
+			asm volatile("pxor %xmm13,%xmm12");
+			asm volatile("pxor %xmm15,%xmm14");
+		}
+		asm volatile("prefetchnta %0" :: "m" (q[d]));
+		asm volatile("prefetchnta %0" :: "m" (q[d+32]));
+		/* P/Q left side optimization */
+		for ( z = start-1 ; z >= 0 ; z-- ) {
+			asm volatile("pxor %xmm5,%xmm5");
+			asm volatile("pxor %xmm7,%xmm7");
+			asm volatile("pxor %xmm13,%xmm13");
+			asm volatile("pxor %xmm15,%xmm15");
+			asm volatile("pcmpgtb %xmm4,%xmm5");
+			asm volatile("pcmpgtb %xmm6,%xmm7");
+			asm volatile("pcmpgtb %xmm12,%xmm13");
+			asm volatile("pcmpgtb %xmm14,%xmm15");
+			asm volatile("paddb %xmm4,%xmm4");
+			asm volatile("paddb %xmm6,%xmm6");
+			asm volatile("paddb %xmm12,%xmm12");
+			asm volatile("paddb %xmm14,%xmm14");
+			asm volatile("pand %xmm0,%xmm5");
+			asm volatile("pand %xmm0,%xmm7");
+			asm volatile("pand %xmm0,%xmm13");
+			asm volatile("pand %xmm0,%xmm15");
+			asm volatile("pxor %xmm5,%xmm4");
+			asm volatile("pxor %xmm7,%xmm6");
+			asm volatile("pxor %xmm13,%xmm12");
+			asm volatile("pxor %xmm15,%xmm14");
+		}
+		asm volatile("movntdq %%xmm2,%0" : "=m" (p[d]));
+		asm volatile("movntdq %%xmm3,%0" : "=m" (p[d+16]));
+		asm volatile("movntdq %%xmm10,%0" : "=m" (p[d+32]));
+		asm volatile("movntdq %%xmm11,%0" : "=m" (p[d+48]));
+		asm volatile("pxor %0,%%xmm4" : : "m" (q[d]));
+		asm volatile("pxor %0,%%xmm6" : : "m" (q[d+16]));
+		asm volatile("pxor %0,%%xmm12" : : "m" (q[d+32]));
+		asm volatile("pxor %0,%%xmm14" : : "m" (q[d+48]));
+		asm volatile("movntdq %%xmm4,%0" : "=m" (q[d]));
+		asm volatile("movntdq %%xmm6,%0" : "=m" (q[d+16]));
+		asm volatile("movntdq %%xmm12,%0" : "=m" (q[d+32]));
+		asm volatile("movntdq %%xmm14,%0" : "=m" (q[d+48]));
+	}
+	asm volatile("sfence" : : : "memory");
+	kernel_fpu_end();
+}
+
+
+const struct raid6_calls raid6_sse2x4 = {
+	raid6_sse24_gen_syndrome,
+	raid6_sse24_xor_syndrome,
+	raid6_have_sse2,
+	"sse2x4",
+	1			/* Has cache hints */
+};
+
+#endif /* CONFIG_X86_64 */
diff --git a/lib/raid6/test/Makefile b/lib/raid6/test/Makefile
new file mode 100644
index 000000000..4fb7700a7
--- /dev/null
+++ b/lib/raid6/test/Makefile
@@ -0,0 +1,137 @@
+# SPDX-License-Identifier: GPL-2.0
+#
+# This is a simple Makefile to test some of the RAID-6 code
+# from userspace.
+#
+
+pound := \#
+
+CC	 = gcc
+OPTFLAGS = -O2			# Adjust as desired
+CFLAGS	 = -I.. -I ../../../include -g $(OPTFLAGS)
+LD	 = ld
+AWK	 = awk -f
+AR	 = ar
+RANLIB	 = ranlib
+OBJS	 = int1.o int2.o int4.o int8.o int16.o int32.o recov.o algos.o tables.o
+
+ARCH := $(shell uname -m 2>/dev/null | sed -e /s/i.86/i386/)
+ifeq ($(ARCH),i386)
+        CFLAGS += -DCONFIG_X86_32
+        IS_X86 = yes
+endif
+ifeq ($(ARCH),x86_64)
+        CFLAGS += -DCONFIG_X86_64
+        IS_X86 = yes
+endif
+
+ifeq ($(ARCH),arm)
+        CFLAGS += -I../../../arch/arm/include -mfpu=neon
+        HAS_NEON = yes
+endif
+ifeq ($(ARCH),aarch64)
+        CFLAGS += -I../../../arch/arm64/include
+        HAS_NEON = yes
+endif
+
+ifeq ($(IS_X86),yes)
+        OBJS   += mmx.o sse1.o sse2.o avx2.o recov_ssse3.o recov_avx2.o avx512.o recov_avx512.o
+        CFLAGS += -DCONFIG_X86
+	CFLAGS += $(shell echo "vpmovm2b %k1, %zmm5" |          \
+		    gcc -c -x assembler - >/dev/null 2>&1 &&	\
+		    rm ./-.o && echo -DCONFIG_AS_AVX512=1)
+else ifeq ($(HAS_NEON),yes)
+        OBJS   += neon.o neon1.o neon2.o neon4.o neon8.o recov_neon.o recov_neon_inner.o
+        CFLAGS += -DCONFIG_KERNEL_MODE_NEON=1
+else
+        HAS_ALTIVEC := $(shell printf '$(pound)include <altivec.h>\nvector int a;\n' |\
+                         gcc -c -x c - >/dev/null && rm ./-.o && echo yes)
+        ifeq ($(HAS_ALTIVEC),yes)
+                CFLAGS += -I../../../arch/powerpc/include
+                CFLAGS += -DCONFIG_ALTIVEC
+                OBJS += altivec1.o altivec2.o altivec4.o altivec8.o \
+                        vpermxor1.o vpermxor2.o vpermxor4.o vpermxor8.o
+        endif
+endif
+
+.c.o:
+	$(CC) $(CFLAGS) -c -o $@ $<
+
+%.c: ../%.c
+	cp -f $< $@
+
+%.uc: ../%.uc
+	cp -f $< $@
+
+all:	raid6.a raid6test
+
+raid6.a: $(OBJS)
+	 rm -f $@
+	 $(AR) cq $@ $^
+	 $(RANLIB) $@
+
+raid6test: test.c raid6.a
+	$(CC) $(CFLAGS) -o raid6test $^
+
+neon1.c: neon.uc ../unroll.awk
+	$(AWK) ../unroll.awk -vN=1 < neon.uc > $@
+
+neon2.c: neon.uc ../unroll.awk
+	$(AWK) ../unroll.awk -vN=2 < neon.uc > $@
+
+neon4.c: neon.uc ../unroll.awk
+	$(AWK) ../unroll.awk -vN=4 < neon.uc > $@
+
+neon8.c: neon.uc ../unroll.awk
+	$(AWK) ../unroll.awk -vN=8 < neon.uc > $@
+
+altivec1.c: altivec.uc ../unroll.awk
+	$(AWK) ../unroll.awk -vN=1 < altivec.uc > $@
+
+altivec2.c: altivec.uc ../unroll.awk
+	$(AWK) ../unroll.awk -vN=2 < altivec.uc > $@
+
+altivec4.c: altivec.uc ../unroll.awk
+	$(AWK) ../unroll.awk -vN=4 < altivec.uc > $@
+
+altivec8.c: altivec.uc ../unroll.awk
+	$(AWK) ../unroll.awk -vN=8 < altivec.uc > $@
+
+vpermxor1.c: vpermxor.uc ../unroll.awk
+	$(AWK) ../unroll.awk -vN=1 < vpermxor.uc > $@
+
+vpermxor2.c: vpermxor.uc ../unroll.awk
+	$(AWK) ../unroll.awk -vN=2 < vpermxor.uc > $@
+
+vpermxor4.c: vpermxor.uc ../unroll.awk
+	$(AWK) ../unroll.awk -vN=4 < vpermxor.uc > $@
+
+vpermxor8.c: vpermxor.uc ../unroll.awk
+	$(AWK) ../unroll.awk -vN=8 < vpermxor.uc > $@
+
+int1.c: int.uc ../unroll.awk
+	$(AWK) ../unroll.awk -vN=1 < int.uc > $@
+
+int2.c: int.uc ../unroll.awk
+	$(AWK) ../unroll.awk -vN=2 < int.uc > $@
+
+int4.c: int.uc ../unroll.awk
+	$(AWK) ../unroll.awk -vN=4 < int.uc > $@
+
+int8.c: int.uc ../unroll.awk
+	$(AWK) ../unroll.awk -vN=8 < int.uc > $@
+
+int16.c: int.uc ../unroll.awk
+	$(AWK) ../unroll.awk -vN=16 < int.uc > $@
+
+int32.c: int.uc ../unroll.awk
+	$(AWK) ../unroll.awk -vN=32 < int.uc > $@
+
+tables.c: mktables
+	./mktables > tables.c
+
+clean:
+	rm -f *.o *.a mktables mktables.c *.uc int*.c altivec*.c vpermxor*.c neon*.c tables.c raid6test
+
+spotless: clean
+	rm -f *~
diff --git a/lib/raid6/test/test.c b/lib/raid6/test/test.c
new file mode 100644
index 000000000..841a55242
--- /dev/null
+++ b/lib/raid6/test/test.c
@@ -0,0 +1,152 @@
+// SPDX-License-Identifier: GPL-2.0-or-later
+/* -*- linux-c -*- ------------------------------------------------------- *
+ *
+ *   Copyright 2002-2007 H. Peter Anvin - All Rights Reserved
+ *
+ * ----------------------------------------------------------------------- */
+
+/*
+ * raid6test.c
+ *
+ * Test RAID-6 recovery with various algorithms
+ */
+
+#include <stdlib.h>
+#include <stdio.h>
+#include <string.h>
+#include <linux/raid/pq.h>
+
+#define NDISKS		16	/* Including P and Q */
+
+const char raid6_empty_zero_page[PAGE_SIZE] __attribute__((aligned(PAGE_SIZE)));
+
+char *dataptrs[NDISKS];
+char data[NDISKS][PAGE_SIZE] __attribute__((aligned(PAGE_SIZE)));
+char recovi[PAGE_SIZE] __attribute__((aligned(PAGE_SIZE)));
+char recovj[PAGE_SIZE] __attribute__((aligned(PAGE_SIZE)));
+
+static void makedata(int start, int stop)
+{
+	int i, j;
+
+	for (i = start; i <= stop; i++) {
+		for (j = 0; j < PAGE_SIZE; j++)
+			data[i][j] = rand();
+
+		dataptrs[i] = data[i];
+	}
+}
+
+static char disk_type(int d)
+{
+	switch (d) {
+	case NDISKS-2:
+		return 'P';
+	case NDISKS-1:
+		return 'Q';
+	default:
+		return 'D';
+	}
+}
+
+static int test_disks(int i, int j)
+{
+	int erra, errb;
+
+	memset(recovi, 0xf0, PAGE_SIZE);
+	memset(recovj, 0xba, PAGE_SIZE);
+
+	dataptrs[i] = recovi;
+	dataptrs[j] = recovj;
+
+	raid6_dual_recov(NDISKS, PAGE_SIZE, i, j, (void **)&dataptrs);
+
+	erra = memcmp(data[i], recovi, PAGE_SIZE);
+	errb = memcmp(data[j], recovj, PAGE_SIZE);
+
+	if (i < NDISKS-2 && j == NDISKS-1) {
+		/* We don't implement the DQ failure scenario, since it's
+		   equivalent to a RAID-5 failure (XOR, then recompute Q) */
+		erra = errb = 0;
+	} else {
+		printf("algo=%-8s  faila=%3d(%c)  failb=%3d(%c)  %s\n",
+		       raid6_call.name,
+		       i, disk_type(i),
+		       j, disk_type(j),
+		       (!erra && !errb) ? "OK" :
+		       !erra ? "ERRB" :
+		       !errb ? "ERRA" : "ERRAB");
+	}
+
+	dataptrs[i] = data[i];
+	dataptrs[j] = data[j];
+
+	return erra || errb;
+}
+
+int main(int argc, char *argv[])
+{
+	const struct raid6_calls *const *algo;
+	const struct raid6_recov_calls *const *ra;
+	int i, j, p1, p2;
+	int err = 0;
+
+	makedata(0, NDISKS-1);
+
+	for (ra = raid6_recov_algos; *ra; ra++) {
+		if ((*ra)->valid  && !(*ra)->valid())
+			continue;
+
+		raid6_2data_recov = (*ra)->data2;
+		raid6_datap_recov = (*ra)->datap;
+
+		printf("using recovery %s\n", (*ra)->name);
+
+		for (algo = raid6_algos; *algo; algo++) {
+			if ((*algo)->valid && !(*algo)->valid())
+				continue;
+
+			raid6_call = **algo;
+
+			/* Nuke syndromes */
+			memset(data[NDISKS-2], 0xee, 2*PAGE_SIZE);
+
+			/* Generate assumed good syndrome */
+			raid6_call.gen_syndrome(NDISKS, PAGE_SIZE,
+						(void **)&dataptrs);
+
+			for (i = 0; i < NDISKS-1; i++)
+				for (j = i+1; j < NDISKS; j++)
+					err += test_disks(i, j);
+
+			if (!raid6_call.xor_syndrome)
+				continue;
+
+			for (p1 = 0; p1 < NDISKS-2; p1++)
+				for (p2 = p1; p2 < NDISKS-2; p2++) {
+
+					/* Simulate rmw run */
+					raid6_call.xor_syndrome(NDISKS, p1, p2, PAGE_SIZE,
+								(void **)&dataptrs);
+					makedata(p1, p2);
+					raid6_call.xor_syndrome(NDISKS, p1, p2, PAGE_SIZE,
+                                                                (void **)&dataptrs);
+
+					for (i = 0; i < NDISKS-1; i++)
+						for (j = i+1; j < NDISKS; j++)
+							err += test_disks(i, j);
+				}
+
+		}
+		printf("\n");
+	}
+
+	printf("\n");
+	/* Pick the best algorithm test */
+	raid6_select_algo();
+
+	if (err)
+		printf("\n*** ERRORS FOUND ***\n");
+
+	return err;
+}
diff --git a/lib/raid6/unroll.awk b/lib/raid6/unroll.awk
new file mode 100644
index 000000000..0809805a7
--- /dev/null
+++ b/lib/raid6/unroll.awk
@@ -0,0 +1,20 @@
+
+# This filter requires one command line option of form -vN=n
+# where n must be a decimal number.
+#
+# Repeat each input line containing $$ n times, replacing $$ with 0...n-1.
+# Replace each $# with n, and each $* with a single $.
+
+BEGIN {
+	n = N + 0
+}
+{
+	if (/\$\$/) { rep = n } else { rep = 1 }
+	for (i = 0; i < rep; ++i) {
+		tmp = $0
+		gsub(/\$\$/, i, tmp)
+		gsub(/\$#/, n, tmp)
+		gsub(/\$\*/, "$", tmp)
+		print tmp
+	}
+}
diff --git a/lib/raid6/vpermxor.uc b/lib/raid6/vpermxor.uc
new file mode 100644
index 000000000..1bfb127fb
--- /dev/null
+++ b/lib/raid6/vpermxor.uc
@@ -0,0 +1,105 @@
+/*
+ * Copyright 2017, Matt Brown, IBM Corp.
+ *
+ * This program is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU General Public License
+ * as published by the Free Software Foundation; either version
+ * 2 of the License, or (at your option) any later version.
+ *
+ * vpermxor$#.c
+ *
+ * Based on H. Peter Anvin's paper - The mathematics of RAID-6
+ *
+ * $#-way unrolled portable integer math RAID-6 instruction set
+ * This file is postprocessed using unroll.awk
+ *
+ * vpermxor$#.c makes use of the vpermxor instruction to optimise the RAID6 Q
+ * syndrome calculations.
+ * This can be run on systems which have both Altivec and vpermxor instruction.
+ *
+ * This instruction was introduced in POWER8 - ISA v2.07.
+ */
+
+#include <linux/raid/pq.h>
+#ifdef CONFIG_ALTIVEC
+
+#include <altivec.h>
+#include <asm/ppc-opcode.h>
+#ifdef __KERNEL__
+#include <asm/cputable.h>
+#include <asm/switch_to.h>
+#endif
+
+typedef vector unsigned char unative_t;
+#define NSIZE sizeof(unative_t)
+
+static const vector unsigned char gf_low = {0x1e, 0x1c, 0x1a, 0x18, 0x16, 0x14,
+					    0x12, 0x10, 0x0e, 0x0c, 0x0a, 0x08,
+					    0x06, 0x04, 0x02,0x00};
+static const vector unsigned char gf_high = {0xfd, 0xdd, 0xbd, 0x9d, 0x7d, 0x5d,
+					     0x3d, 0x1d, 0xe0, 0xc0, 0xa0, 0x80,
+					     0x60, 0x40, 0x20, 0x00};
+
+static void noinline raid6_vpermxor$#_gen_syndrome_real(int disks, size_t bytes,
+							void **ptrs)
+{
+	u8 **dptr = (u8 **)ptrs;
+	u8 *p, *q;
+	int d, z, z0;
+	unative_t wp$$, wq$$, wd$$;
+
+	z0 = disks - 3;		/* Highest data disk */
+	p = dptr[z0+1];		/* XOR parity */
+	q = dptr[z0+2];		/* RS syndrome */
+
+	for (d = 0; d < bytes; d += NSIZE*$#) {
+		wp$$ = wq$$ = *(unative_t *)&dptr[z0][d+$$*NSIZE];
+
+		for (z = z0-1; z>=0; z--) {
+			wd$$ = *(unative_t *)&dptr[z][d+$$*NSIZE];
+			/* P syndrome */
+			wp$$ = vec_xor(wp$$, wd$$);
+
+			/* Q syndrome */
+			asm(VPERMXOR(%0,%1,%2,%3):"=v"(wq$$):"v"(gf_high), "v"(gf_low), "v"(wq$$));
+			wq$$ = vec_xor(wq$$, wd$$);
+		}
+		*(unative_t *)&p[d+NSIZE*$$] = wp$$;
+		*(unative_t *)&q[d+NSIZE*$$] = wq$$;
+	}
+}
+
+static void raid6_vpermxor$#_gen_syndrome(int disks, size_t bytes, void **ptrs)
+{
+	preempt_disable();
+	enable_kernel_altivec();
+
+	raid6_vpermxor$#_gen_syndrome_real(disks, bytes, ptrs);
+
+	disable_kernel_altivec();
+	preempt_enable();
+}
+
+int raid6_have_altivec_vpermxor(void);
+#if $# == 1
+int raid6_have_altivec_vpermxor(void)
+{
+	/* Check if arch has both altivec and the vpermxor instructions */
+# ifdef __KERNEL__
+	return (cpu_has_feature(CPU_FTR_ALTIVEC_COMP) &&
+		cpu_has_feature(CPU_FTR_ARCH_207S));
+# else
+	return 1;
+#endif
+
+}
+#endif
+
+const struct raid6_calls raid6_vpermxor$# = {
+	raid6_vpermxor$#_gen_syndrome,
+	NULL,
+	raid6_have_altivec_vpermxor,
+	"vpermxor$#",
+	0
+};
+#endif
diff --git a/lib/raid6/x86.h b/lib/raid6/x86.h
new file mode 100644
index 000000000..9a6ff3711
--- /dev/null
+++ b/lib/raid6/x86.h
@@ -0,0 +1,75 @@
+/* SPDX-License-Identifier: GPL-2.0-or-later */
+/* ----------------------------------------------------------------------- *
+ *
+ *   Copyright 2002-2004 H. Peter Anvin - All Rights Reserved
+ *
+ * ----------------------------------------------------------------------- */
+
+/*
+ * raid6/x86.h
+ *
+ * Definitions common to x86 and x86-64 RAID-6 code only
+ */
+
+#ifndef LINUX_RAID_RAID6X86_H
+#define LINUX_RAID_RAID6X86_H
+
+#if (defined(__i386__) || defined(__x86_64__)) && !defined(__arch_um__)
+
+#ifdef __KERNEL__ /* Real code */
+
+#include <asm/fpu/api.h>
+
+#else /* Dummy code for user space testing */
+
+static inline void kernel_fpu_begin(void)
+{
+}
+
+static inline void kernel_fpu_end(void)
+{
+}
+
+#define __aligned(x) __attribute__((aligned(x)))
+
+#define X86_FEATURE_MMX		(0*32+23) /* Multimedia Extensions */
+#define X86_FEATURE_FXSR	(0*32+24) /* FXSAVE and FXRSTOR instructions
+					   * (fast save and restore) */
+#define X86_FEATURE_XMM		(0*32+25) /* Streaming SIMD Extensions */
+#define X86_FEATURE_XMM2	(0*32+26) /* Streaming SIMD Extensions-2 */
+#define X86_FEATURE_XMM3	(4*32+ 0) /* "pni" SSE-3 */
+#define X86_FEATURE_SSSE3	(4*32+ 9) /* Supplemental SSE-3 */
+#define X86_FEATURE_AVX	(4*32+28) /* Advanced Vector Extensions */
+#define X86_FEATURE_AVX2        (9*32+ 5) /* AVX2 instructions */
+#define X86_FEATURE_AVX512F     (9*32+16) /* AVX-512 Foundation */
+#define X86_FEATURE_AVX512DQ    (9*32+17) /* AVX-512 DQ (Double/Quad granular)
+					   * Instructions
+					   */
+#define X86_FEATURE_AVX512BW    (9*32+30) /* AVX-512 BW (Byte/Word granular)
+					   * Instructions
+					   */
+#define X86_FEATURE_AVX512VL    (9*32+31) /* AVX-512 VL (128/256 Vector Length)
+					   * Extensions
+					   */
+#define X86_FEATURE_MMXEXT	(1*32+22) /* AMD MMX extensions */
+
+/* Should work well enough on modern CPUs for testing */
+static inline int boot_cpu_has(int flag)
+{
+	u32 eax, ebx, ecx, edx;
+
+	eax = (flag & 0x100) ? 7 :
+		(flag & 0x20) ? 0x80000001 : 1;
+	ecx = 0;
+
+	asm volatile("cpuid"
+		     : "+a" (eax), "=b" (ebx), "=d" (edx), "+c" (ecx));
+
+	return ((flag & 0x100 ? ebx :
+		(flag & 0x80) ? ecx : edx) >> (flag & 31)) & 1;
+}
+
+#endif /* ndef __KERNEL__ */
+
+#endif
+#endif