Resguardo inicial

linux-7.2.3/tools/arch/x86/lib/inat.c

@@ -0,0 +1,96 @@
+// SPDX-License-Identifier: GPL-2.0-or-later
+/*
+ * x86 instruction attribute tables
+ *
+ * Written by Masami Hiramatsu <mhiramat@redhat.com>
+ */
+#include "../include/asm/insn.h" /* __ignore_sync_check__ */
+
+/* Attribute tables are generated from opcode map */
+#include "inat-tables.c"
+
+/* Attribute search APIs */
+insn_attr_t inat_get_opcode_attribute(insn_byte_t opcode)
+{
+	return inat_primary_table[opcode];
+}
+
+int inat_get_last_prefix_id(insn_byte_t last_pfx)
+{
+	insn_attr_t lpfx_attr;
+
+	lpfx_attr = inat_get_opcode_attribute(last_pfx);
+	return inat_last_prefix_id(lpfx_attr);
+}
+
+insn_attr_t inat_get_escape_attribute(insn_byte_t opcode, int lpfx_id,
+				      insn_attr_t esc_attr)
+{
+	const insn_attr_t *table;
+	int n;
+
+	n = inat_escape_id(esc_attr);
+
+	table = inat_escape_tables[n][0];
+	if (!table)
+		return 0;
+	if (inat_has_variant(table[opcode]) && lpfx_id) {
+		table = inat_escape_tables[n][lpfx_id];
+		if (!table)
+			return 0;
+	}
+	return table[opcode];
+}
+
+insn_attr_t inat_get_group_attribute(insn_byte_t modrm, int lpfx_id,
+				     insn_attr_t grp_attr)
+{
+	const insn_attr_t *table;
+	int n;
+
+	n = inat_group_id(grp_attr);
+
+	table = inat_group_tables[n][0];
+	if (!table)
+		return inat_group_common_attribute(grp_attr);
+	if (inat_has_variant(table[X86_MODRM_REG(modrm)]) && lpfx_id) {
+		table = inat_group_tables[n][lpfx_id];
+		if (!table)
+			return inat_group_common_attribute(grp_attr);
+	}
+	return table[X86_MODRM_REG(modrm)] |
+	       inat_group_common_attribute(grp_attr);
+}
+
+insn_attr_t inat_get_avx_attribute(insn_byte_t opcode, insn_byte_t vex_m,
+				   insn_byte_t vex_p)
+{
+	const insn_attr_t *table;
+	if (vex_m > X86_VEX_M_MAX || vex_p > INAT_LSTPFX_MAX)
+		return 0;
+	/* At first, this checks the master table */
+	table = inat_avx_tables[vex_m][0];
+	if (!table)
+		return 0;
+	if (!inat_is_group(table[opcode]) && vex_p) {
+		/* If this is not a group, get attribute directly */
+		table = inat_avx_tables[vex_m][vex_p];
+		if (!table)
+			return 0;
+	}
+	return table[opcode];
+}
+
+insn_attr_t inat_get_xop_attribute(insn_byte_t opcode, insn_byte_t map_select)
+{
+	const insn_attr_t *table;
+
+	if (map_select < X86_XOP_M_MIN || map_select > X86_XOP_M_MAX)
+		return 0;
+	map_select -= X86_XOP_M_MIN;
+	/* At first, this checks the master table */
+	table = inat_xop_tables[map_select];
+	if (!table)
+		return 0;
+	return table[opcode];
+}

linux-7.2.3/tools/arch/x86/lib/insn.c

@@ -0,0 +1,792 @@
+// SPDX-License-Identifier: GPL-2.0-or-later
+/*
+ * x86 instruction analysis
+ *
+ * Copyright (C) IBM Corporation, 2002, 2004, 2009
+ */
+
+#include <linux/kernel.h>
+#ifdef __KERNEL__
+#include <linux/string.h>
+#else
+#include <string.h>
+#endif
+#include "../include/asm/inat.h" /* __ignore_sync_check__ */
+#include "../include/asm/insn.h" /* __ignore_sync_check__ */
+#include <linux/unaligned.h> /* __ignore_sync_check__ */
+
+#include <linux/errno.h>
+#include <linux/kconfig.h>
+
+#include "../include/asm/emulate_prefix.h" /* __ignore_sync_check__ */
+
+#define leXX_to_cpu(t, r)						\
+({									\
+	__typeof__(t) v;						\
+	switch (sizeof(t)) {						\
+	case 4: v = le32_to_cpu(r); break;				\
+	case 2: v = le16_to_cpu(r); break;				\
+	case 1:	v = r; break;						\
+	default:							\
+		BUILD_BUG(); break;					\
+	}								\
+	v;								\
+})
+
+/* Verify next sizeof(t) bytes can be on the same instruction */
+#define validate_next(t, insn, n)	\
+	((insn)->next_byte + sizeof(t) + n <= (insn)->end_kaddr)
+
+#define __get_next(t, insn)	\
+	({ t r = get_unaligned((t *)(insn)->next_byte); (insn)->next_byte += sizeof(t); leXX_to_cpu(t, r); })
+
+#define __peek_nbyte_next(t, insn, n)	\
+	({ t r = get_unaligned((t *)(insn)->next_byte + n); leXX_to_cpu(t, r); })
+
+#define get_next(t, insn)	\
+	({ if (unlikely(!validate_next(t, insn, 0))) goto err_out; __get_next(t, insn); })
+
+#define peek_nbyte_next(t, insn, n)	\
+	({ if (unlikely(!validate_next(t, insn, n))) goto err_out; __peek_nbyte_next(t, insn, n); })
+
+#define peek_next(t, insn)	peek_nbyte_next(t, insn, 0)
+
+/**
+ * insn_init() - initialize struct insn
+ * @insn:	&struct insn to be initialized
+ * @kaddr:	address (in kernel memory) of instruction (or copy thereof)
+ * @buf_len:	length of the insn buffer at @kaddr
+ * @x86_64:	!0 for 64-bit kernel or 64-bit app
+ */
+void insn_init(struct insn *insn, const void *kaddr, int buf_len, int x86_64)
+{
+	/*
+	 * Instructions longer than MAX_INSN_SIZE (15 bytes) are invalid
+	 * even if the input buffer is long enough to hold them.
+	 */
+	if (buf_len > MAX_INSN_SIZE)
+		buf_len = MAX_INSN_SIZE;
+
+	memset(insn, 0, sizeof(*insn));
+	insn->kaddr = kaddr;
+	insn->end_kaddr = kaddr + buf_len;
+	insn->next_byte = kaddr;
+	insn->x86_64 = x86_64;
+	insn->opnd_bytes = 4;
+	if (x86_64)
+		insn->addr_bytes = 8;
+	else
+		insn->addr_bytes = 4;
+}
+
+static const insn_byte_t xen_prefix[] = { __XEN_EMULATE_PREFIX };
+static const insn_byte_t kvm_prefix[] = { __KVM_EMULATE_PREFIX };
+
+static int __insn_get_emulate_prefix(struct insn *insn,
+				     const insn_byte_t *prefix, size_t len)
+{
+	size_t i;
+
+	for (i = 0; i < len; i++) {
+		if (peek_nbyte_next(insn_byte_t, insn, i) != prefix[i])
+			goto err_out;
+	}
+
+	insn->emulate_prefix_size = len;
+	insn->next_byte += len;
+
+	return 1;
+
+err_out:
+	return 0;
+}
+
+static void insn_get_emulate_prefix(struct insn *insn)
+{
+	if (__insn_get_emulate_prefix(insn, xen_prefix, sizeof(xen_prefix)))
+		return;
+
+	__insn_get_emulate_prefix(insn, kvm_prefix, sizeof(kvm_prefix));
+}
+
+/**
+ * insn_get_prefixes - scan x86 instruction prefix bytes
+ * @insn:	&struct insn containing instruction
+ *
+ * Populates the @insn->prefixes bitmap, and updates @insn->next_byte
+ * to point to the (first) opcode.  No effect if @insn->prefixes.got
+ * is already set.
+ *
+ * * Returns:
+ * 0:  on success
+ * < 0: on error
+ */
+int insn_get_prefixes(struct insn *insn)
+{
+	struct insn_field *prefixes = &insn->prefixes;
+	insn_attr_t attr;
+	insn_byte_t b, lb;
+	int i, nb;
+
+	if (prefixes->got)
+		return 0;
+
+	insn_get_emulate_prefix(insn);
+
+	nb = 0;
+	lb = 0;
+	b = peek_next(insn_byte_t, insn);
+	attr = inat_get_opcode_attribute(b);
+	while (inat_is_legacy_prefix(attr)) {
+		/* Skip if same prefix */
+		for (i = 0; i < nb; i++)
+			if (prefixes->bytes[i] == b)
+				goto found;
+		if (nb == 4)
+			/* Invalid instruction */
+			break;
+		prefixes->bytes[nb++] = b;
+		if (inat_is_address_size_prefix(attr)) {
+			/* address size switches 2/4 or 4/8 */
+			if (insn->x86_64)
+				insn->addr_bytes ^= 12;
+			else
+				insn->addr_bytes ^= 6;
+		} else if (inat_is_operand_size_prefix(attr)) {
+			/* oprand size switches 2/4 */
+			insn->opnd_bytes ^= 6;
+		}
+found:
+		prefixes->nbytes++;
+		insn->next_byte++;
+		lb = b;
+		b = peek_next(insn_byte_t, insn);
+		attr = inat_get_opcode_attribute(b);
+	}
+	/* Set the last prefix */
+	if (lb && lb != insn->prefixes.bytes[3]) {
+		if (unlikely(insn->prefixes.bytes[3])) {
+			/* Swap the last prefix */
+			b = insn->prefixes.bytes[3];
+			for (i = 0; i < nb; i++)
+				if (prefixes->bytes[i] == lb)
+					insn_set_byte(prefixes, i, b);
+		}
+		insn_set_byte(&insn->prefixes, 3, lb);
+	}
+
+	/* Decode REX prefix */
+	if (insn->x86_64) {
+		b = peek_next(insn_byte_t, insn);
+		attr = inat_get_opcode_attribute(b);
+		if (inat_is_rex_prefix(attr)) {
+			insn_field_set(&insn->rex_prefix, b, 1);
+			insn->next_byte++;
+			if (X86_REX_W(b))
+				/* REX.W overrides opnd_size */
+				insn->opnd_bytes = 8;
+		} else if (inat_is_rex2_prefix(attr)) {
+			insn_set_byte(&insn->rex_prefix, 0, b);
+			b = peek_nbyte_next(insn_byte_t, insn, 1);
+			insn_set_byte(&insn->rex_prefix, 1, b);
+			insn->rex_prefix.nbytes = 2;
+			insn->next_byte += 2;
+			if (X86_REX_W(b))
+				/* REX.W overrides opnd_size */
+				insn->opnd_bytes = 8;
+			insn->rex_prefix.got = 1;
+			goto vex_end;
+		}
+	}
+	insn->rex_prefix.got = 1;
+
+	/* Decode VEX/XOP prefix */
+	b = peek_next(insn_byte_t, insn);
+	if (inat_is_vex_prefix(attr) || inat_is_xop_prefix(attr)) {
+		insn_byte_t b2 = peek_nbyte_next(insn_byte_t, insn, 1);
+
+		if (inat_is_xop_prefix(attr) && X86_MODRM_REG(b2) == 0) {
+			/* Grp1A.0 is always POP Ev */
+			goto vex_end;
+		} else if (!insn->x86_64) {
+			/*
+			 * In 32-bits mode, if the [7:6] bits (mod bits of
+			 * ModRM) on the second byte are not 11b, it is
+			 * LDS or LES or BOUND.
+			 */
+			if (X86_MODRM_MOD(b2) != 3)
+				goto vex_end;
+		}
+		insn_set_byte(&insn->vex_prefix, 0, b);
+		insn_set_byte(&insn->vex_prefix, 1, b2);
+		if (inat_is_evex_prefix(attr)) {
+			b2 = peek_nbyte_next(insn_byte_t, insn, 2);
+			insn_set_byte(&insn->vex_prefix, 2, b2);
+			b2 = peek_nbyte_next(insn_byte_t, insn, 3);
+			insn_set_byte(&insn->vex_prefix, 3, b2);
+			insn->vex_prefix.nbytes = 4;
+			insn->next_byte += 4;
+			if (insn->x86_64 && X86_VEX_W(b2))
+				/* VEX.W overrides opnd_size */
+				insn->opnd_bytes = 8;
+		} else if (inat_is_vex3_prefix(attr) || inat_is_xop_prefix(attr)) {
+			b2 = peek_nbyte_next(insn_byte_t, insn, 2);
+			insn_set_byte(&insn->vex_prefix, 2, b2);
+			insn->vex_prefix.nbytes = 3;
+			insn->next_byte += 3;
+			if (insn->x86_64 && X86_VEX_W(b2))
+				/* VEX.W/XOP.W overrides opnd_size */
+				insn->opnd_bytes = 8;
+		} else {
+			/*
+			 * For VEX2, fake VEX3-like byte#2.
+			 * Makes it easier to decode vex.W, vex.vvvv,
+			 * vex.L and vex.pp. Masking with 0x7f sets vex.W == 0.
+			 */
+			insn_set_byte(&insn->vex_prefix, 2, b2 & 0x7f);
+			insn->vex_prefix.nbytes = 2;
+			insn->next_byte += 2;
+		}
+	}
+vex_end:
+	insn->vex_prefix.got = 1;
+
+	prefixes->got = 1;
+
+	return 0;
+
+err_out:
+	return -ENODATA;
+}
+
+/**
+ * insn_get_opcode - collect opcode(s)
+ * @insn:	&struct insn containing instruction
+ *
+ * Populates @insn->opcode, updates @insn->next_byte to point past the
+ * opcode byte(s), and set @insn->attr (except for groups).
+ * If necessary, first collects any preceding (prefix) bytes.
+ * Sets @insn->opcode.value = opcode1.  No effect if @insn->opcode.got
+ * is already 1.
+ *
+ * Returns:
+ * 0:  on success
+ * < 0: on error
+ */
+int insn_get_opcode(struct insn *insn)
+{
+	struct insn_field *opcode = &insn->opcode;
+	int pfx_id, ret;
+	insn_byte_t op;
+
+	if (opcode->got)
+		return 0;
+
+	ret = insn_get_prefixes(insn);
+	if (ret)
+		return ret;
+
+	/* Get first opcode */
+	op = get_next(insn_byte_t, insn);
+	insn_set_byte(opcode, 0, op);
+	opcode->nbytes = 1;
+
+	/* Check if there is VEX/XOP prefix or not */
+	if (insn_is_avx_or_xop(insn)) {
+		insn_byte_t m, p;
+
+		/* XOP prefix has different encoding */
+		if (unlikely(avx_insn_is_xop(insn))) {
+			m = insn_xop_map_bits(insn);
+			insn->attr = inat_get_xop_attribute(op, m);
+			if (!inat_accept_xop(insn->attr)) {
+				insn->attr = 0;
+				return -EINVAL;
+			}
+			/* XOP has only 1 byte for opcode */
+			goto end;
+		}
+
+		m = insn_vex_m_bits(insn);
+		p = insn_vex_p_bits(insn);
+		insn->attr = inat_get_avx_attribute(op, m, p);
+		/* SCALABLE EVEX uses p bits to encode operand size */
+		if (inat_evex_scalable(insn->attr) && !insn_vex_w_bit(insn) &&
+		    p == INAT_PFX_OPNDSZ)
+			insn->opnd_bytes = 2;
+		if ((inat_must_evex(insn->attr) && !insn_is_evex(insn)) ||
+		    (!inat_accept_vex(insn->attr) &&
+		     !inat_is_group(insn->attr))) {
+			/* This instruction is bad */
+			insn->attr = 0;
+			return -EINVAL;
+		}
+		/* VEX has only 1 byte for opcode */
+		goto end;
+	}
+
+	/* Check if there is REX2 prefix or not */
+	if (insn_is_rex2(insn)) {
+		if (insn_rex2_m_bit(insn)) {
+			/* map 1 is escape 0x0f */
+			insn_attr_t esc_attr = inat_get_opcode_attribute(0x0f);
+
+			pfx_id = insn_last_prefix_id(insn);
+			insn->attr = inat_get_escape_attribute(op, pfx_id, esc_attr);
+		} else {
+			insn->attr = inat_get_opcode_attribute(op);
+		}
+		goto end;
+	}
+
+	insn->attr = inat_get_opcode_attribute(op);
+	if (insn->x86_64 && inat_is_invalid64(insn->attr)) {
+		/* This instruction is invalid, like UD2. Stop decoding. */
+		insn->attr &= INAT_INV64;
+	}
+
+	while (inat_is_escape(insn->attr)) {
+		/* Get escaped opcode */
+		op = get_next(insn_byte_t, insn);
+		opcode->bytes[opcode->nbytes++] = op;
+		pfx_id = insn_last_prefix_id(insn);
+		insn->attr = inat_get_escape_attribute(op, pfx_id, insn->attr);
+	}
+
+	if (inat_must_vex(insn->attr)) {
+		/* This instruction is bad */
+		insn->attr = 0;
+		return -EINVAL;
+	}
+
+end:
+	opcode->got = 1;
+	return 0;
+
+err_out:
+	return -ENODATA;
+}
+
+/**
+ * insn_get_modrm - collect ModRM byte, if any
+ * @insn:	&struct insn containing instruction
+ *
+ * Populates @insn->modrm and updates @insn->next_byte to point past the
+ * ModRM byte, if any.  If necessary, first collects the preceding bytes
+ * (prefixes and opcode(s)).  No effect if @insn->modrm.got is already 1.
+ *
+ * Returns:
+ * 0:  on success
+ * < 0: on error
+ */
+int insn_get_modrm(struct insn *insn)
+{
+	struct insn_field *modrm = &insn->modrm;
+	insn_byte_t pfx_id, mod;
+	int ret;
+
+	if (modrm->got)
+		return 0;
+
+	ret = insn_get_opcode(insn);
+	if (ret)
+		return ret;
+
+	if (inat_has_modrm(insn->attr)) {
+		mod = get_next(insn_byte_t, insn);
+		insn_field_set(modrm, mod, 1);
+		if (inat_is_group(insn->attr)) {
+			pfx_id = insn_last_prefix_id(insn);
+			insn->attr = inat_get_group_attribute(mod, pfx_id,
+							      insn->attr);
+			if (insn_is_avx_or_xop(insn) && !inat_accept_vex(insn->attr) &&
+			    !inat_accept_xop(insn->attr)) {
+				/* Bad insn */
+				insn->attr = 0;
+				return -EINVAL;
+			}
+		}
+	}
+
+	if (insn->x86_64 && inat_is_force64(insn->attr))
+		insn->opnd_bytes = 8;
+
+	modrm->got = 1;
+	return 0;
+
+err_out:
+	return -ENODATA;
+}
+
+
+/**
+ * insn_rip_relative() - Does instruction use RIP-relative addressing mode?
+ * @insn:	&struct insn containing instruction
+ *
+ * If necessary, first collects the instruction up to and including the
+ * ModRM byte.  No effect if @insn->x86_64 is 0.
+ */
+int insn_rip_relative(struct insn *insn)
+{
+	struct insn_field *modrm = &insn->modrm;
+	int ret;
+
+	if (!insn->x86_64)
+		return 0;
+
+	ret = insn_get_modrm(insn);
+	if (ret)
+		return 0;
+	/*
+	 * For rip-relative instructions, the mod field (top 2 bits)
+	 * is zero and the r/m field (bottom 3 bits) is 0x5.
+	 */
+	return (modrm->nbytes && (modrm->bytes[0] & 0xc7) == 0x5);
+}
+
+/**
+ * insn_get_sib() - Get the SIB byte of instruction
+ * @insn:	&struct insn containing instruction
+ *
+ * If necessary, first collects the instruction up to and including the
+ * ModRM byte.
+ *
+ * Returns:
+ * 0: if decoding succeeded
+ * < 0: otherwise.
+ */
+int insn_get_sib(struct insn *insn)
+{
+	insn_byte_t modrm;
+	int ret;
+
+	if (insn->sib.got)
+		return 0;
+
+	ret = insn_get_modrm(insn);
+	if (ret)
+		return ret;
+
+	if (insn->modrm.nbytes) {
+		modrm = insn->modrm.bytes[0];
+		if (insn->addr_bytes != 2 &&
+		    X86_MODRM_MOD(modrm) != 3 && X86_MODRM_RM(modrm) == 4) {
+			insn_field_set(&insn->sib,
+				       get_next(insn_byte_t, insn), 1);
+		}
+	}
+	insn->sib.got = 1;
+
+	return 0;
+
+err_out:
+	return -ENODATA;
+}
+
+
+/**
+ * insn_get_displacement() - Get the displacement of instruction
+ * @insn:	&struct insn containing instruction
+ *
+ * If necessary, first collects the instruction up to and including the
+ * SIB byte.
+ * Displacement value is sign-expanded.
+ *
+ * * Returns:
+ * 0: if decoding succeeded
+ * < 0: otherwise.
+ */
+int insn_get_displacement(struct insn *insn)
+{
+	insn_byte_t mod, rm, base;
+	int ret;
+
+	if (insn->displacement.got)
+		return 0;
+
+	ret = insn_get_sib(insn);
+	if (ret)
+		return ret;
+
+	if (insn->modrm.nbytes) {
+		/*
+		 * Interpreting the modrm byte:
+		 * mod = 00 - no displacement fields (exceptions below)
+		 * mod = 01 - 1-byte displacement field
+		 * mod = 10 - displacement field is 4 bytes, or 2 bytes if
+		 * 	address size = 2 (0x67 prefix in 32-bit mode)
+		 * mod = 11 - no memory operand
+		 *
+		 * If address size = 2...
+		 * mod = 00, r/m = 110 - displacement field is 2 bytes
+		 *
+		 * If address size != 2...
+		 * mod != 11, r/m = 100 - SIB byte exists
+		 * mod = 00, SIB base = 101 - displacement field is 4 bytes
+		 * mod = 00, r/m = 101 - rip-relative addressing, displacement
+		 * 	field is 4 bytes
+		 */
+		mod = X86_MODRM_MOD(insn->modrm.value);
+		rm = X86_MODRM_RM(insn->modrm.value);
+		base = X86_SIB_BASE(insn->sib.value);
+		if (mod == 3)
+			goto out;
+		if (mod == 1) {
+			insn_field_set(&insn->displacement,
+				       get_next(signed char, insn), 1);
+		} else if (insn->addr_bytes == 2) {
+			if ((mod == 0 && rm == 6) || mod == 2) {
+				insn_field_set(&insn->displacement,
+					       get_next(short, insn), 2);
+			}
+		} else {
+			if ((mod == 0 && rm == 5) || mod == 2 ||
+			    (mod == 0 && base == 5)) {
+				insn_field_set(&insn->displacement,
+					       get_next(int, insn), 4);
+			}
+		}
+	}
+out:
+	insn->displacement.got = 1;
+	return 0;
+
+err_out:
+	return -ENODATA;
+}
+
+/* Decode moffset16/32/64. Return 0 if failed */
+static int __get_moffset(struct insn *insn)
+{
+	switch (insn->addr_bytes) {
+	case 2:
+		insn_field_set(&insn->moffset1, get_next(short, insn), 2);
+		break;
+	case 4:
+		insn_field_set(&insn->moffset1, get_next(int, insn), 4);
+		break;
+	case 8:
+		insn_field_set(&insn->moffset1, get_next(int, insn), 4);
+		insn_field_set(&insn->moffset2, get_next(int, insn), 4);
+		break;
+	default:	/* opnd_bytes must be modified manually */
+		goto err_out;
+	}
+	insn->moffset1.got = insn->moffset2.got = 1;
+
+	return 1;
+
+err_out:
+	return 0;
+}
+
+/* Decode imm v32(Iz). Return 0 if failed */
+static int __get_immv32(struct insn *insn)
+{
+	switch (insn->opnd_bytes) {
+	case 2:
+		insn_field_set(&insn->immediate, get_next(short, insn), 2);
+		break;
+	case 4:
+	case 8:
+		insn_field_set(&insn->immediate, get_next(int, insn), 4);
+		break;
+	default:	/* opnd_bytes must be modified manually */
+		goto err_out;
+	}
+
+	return 1;
+
+err_out:
+	return 0;
+}
+
+/* Decode imm v64(Iv/Ov), Return 0 if failed */
+static int __get_immv(struct insn *insn)
+{
+	switch (insn->opnd_bytes) {
+	case 2:
+		insn_field_set(&insn->immediate1, get_next(short, insn), 2);
+		break;
+	case 4:
+		insn_field_set(&insn->immediate1, get_next(int, insn), 4);
+		insn->immediate1.nbytes = 4;
+		break;
+	case 8:
+		insn_field_set(&insn->immediate1, get_next(int, insn), 4);
+		insn_field_set(&insn->immediate2, get_next(int, insn), 4);
+		break;
+	default:	/* opnd_bytes must be modified manually */
+		goto err_out;
+	}
+	insn->immediate1.got = insn->immediate2.got = 1;
+
+	return 1;
+err_out:
+	return 0;
+}
+
+/* Decode ptr16:16/32(Ap) */
+static int __get_immptr(struct insn *insn)
+{
+	switch (insn->opnd_bytes) {
+	case 2:
+		insn_field_set(&insn->immediate1, get_next(short, insn), 2);
+		break;
+	case 4:
+		insn_field_set(&insn->immediate1, get_next(int, insn), 4);
+		break;
+	case 8:
+		/* ptr16:64 is not exist (no segment) */
+		return 0;
+	default:	/* opnd_bytes must be modified manually */
+		goto err_out;
+	}
+	insn_field_set(&insn->immediate2, get_next(unsigned short, insn), 2);
+	insn->immediate1.got = insn->immediate2.got = 1;
+
+	return 1;
+err_out:
+	return 0;
+}
+
+/**
+ * insn_get_immediate() - Get the immediate in an instruction
+ * @insn:	&struct insn containing instruction
+ *
+ * If necessary, first collects the instruction up to and including the
+ * displacement bytes.
+ * Basically, most of immediates are sign-expanded. Unsigned-value can be
+ * computed by bit masking with ((1 << (nbytes * 8)) - 1)
+ *
+ * Returns:
+ * 0:  on success
+ * < 0: on error
+ */
+int insn_get_immediate(struct insn *insn)
+{
+	int ret;
+
+	if (insn->immediate.got)
+		return 0;
+
+	ret = insn_get_displacement(insn);
+	if (ret)
+		return ret;
+
+	if (inat_has_moffset(insn->attr)) {
+		if (!__get_moffset(insn))
+			goto err_out;
+		goto done;
+	}
+
+	if (!inat_has_immediate(insn->attr))
+		goto done;
+
+	switch (inat_immediate_size(insn->attr)) {
+	case INAT_IMM_BYTE:
+		insn_field_set(&insn->immediate, get_next(signed char, insn), 1);
+		break;
+	case INAT_IMM_WORD:
+		insn_field_set(&insn->immediate, get_next(short, insn), 2);
+		break;
+	case INAT_IMM_DWORD:
+		insn_field_set(&insn->immediate, get_next(int, insn), 4);
+		break;
+	case INAT_IMM_QWORD:
+		insn_field_set(&insn->immediate1, get_next(int, insn), 4);
+		insn_field_set(&insn->immediate2, get_next(int, insn), 4);
+		break;
+	case INAT_IMM_PTR:
+		if (!__get_immptr(insn))
+			goto err_out;
+		break;
+	case INAT_IMM_VWORD32:
+		if (!__get_immv32(insn))
+			goto err_out;
+		break;
+	case INAT_IMM_VWORD:
+		if (!__get_immv(insn))
+			goto err_out;
+		break;
+	default:
+		/* Here, insn must have an immediate, but failed */
+		goto err_out;
+	}
+	if (inat_has_second_immediate(insn->attr)) {
+		insn_field_set(&insn->immediate2, get_next(signed char, insn), 1);
+	}
+done:
+	insn->immediate.got = 1;
+	return 0;
+
+err_out:
+	return -ENODATA;
+}
+
+/**
+ * insn_get_length() - Get the length of instruction
+ * @insn:	&struct insn containing instruction
+ *
+ * If necessary, first collects the instruction up to and including the
+ * immediates bytes.
+ *
+ * Returns:
+ *  - 0 on success
+ *  - < 0 on error
+*/
+int insn_get_length(struct insn *insn)
+{
+	int ret;
+
+	if (insn->length)
+		return 0;
+
+	ret = insn_get_immediate(insn);
+	if (ret)
+		return ret;
+
+	insn->length = (unsigned char)((unsigned long)insn->next_byte
+				     - (unsigned long)insn->kaddr);
+
+	return 0;
+}
+
+/* Ensure this instruction is decoded completely */
+static inline int insn_complete(struct insn *insn)
+{
+	return insn->opcode.got && insn->modrm.got && insn->sib.got &&
+		insn->displacement.got && insn->immediate.got;
+}
+
+/**
+ * insn_decode() - Decode an x86 instruction
+ * @insn:	&struct insn to be initialized
+ * @kaddr:	address (in kernel memory) of instruction (or copy thereof)
+ * @buf_len:	length of the insn buffer at @kaddr
+ * @m:		insn mode, see enum insn_mode
+ *
+ * Returns:
+ * 0: if decoding succeeded
+ * < 0: otherwise.
+ */
+int insn_decode(struct insn *insn, const void *kaddr, int buf_len, enum insn_mode m)
+{
+	int ret;
+
+#define INSN_MODE_KERN (enum insn_mode)-1 /* __ignore_sync_check__ mode is only valid in the kernel */
+
+	if (m == INSN_MODE_KERN)
+		insn_init(insn, kaddr, buf_len, IS_ENABLED(CONFIG_X86_64));
+	else
+		insn_init(insn, kaddr, buf_len, m == INSN_MODE_64);
+
+	ret = insn_get_length(insn);
+	if (ret)
+		return ret;
+
+	if (insn_complete(insn))
+		return 0;
+
+	return -EINVAL;
+}

linux-7.2.3/tools/arch/x86/lib/memcpy_64.S

@@ -0,0 +1,172 @@
+/* SPDX-License-Identifier: GPL-2.0-only */
+/* Copyright 2002 Andi Kleen */
+
+#include <linux/export.h>
+#include <linux/linkage.h>
+#include <asm/errno.h>
+#include <asm/cpufeatures.h>
+#include <asm/alternative.h>
+
+.section .noinstr.text, "ax"
+
+/*
+ * memcpy - Copy a memory block.
+ *
+ * Input:
+ *  rdi destination
+ *  rsi source
+ *  rdx count
+ *
+ * Output:
+ * rax original destination
+ *
+ * The FSRM alternative should be done inline (avoiding the call and
+ * the disgusting return handling), but that would require some help
+ * from the compiler for better calling conventions.
+ *
+ * The 'rep movsb' itself is small enough to replace the call, but the
+ * two register moves blow up the code. And one of them is "needed"
+ * only for the return value that is the same as the source input,
+ * which the compiler could/should do much better anyway.
+ */
+SYM_TYPED_FUNC_START(__memcpy)
+	ALTERNATIVE "jmp memcpy_orig", "", X86_FEATURE_FSRM
+
+	movq %rdi, %rax
+	movq %rdx, %rcx
+	rep movsb
+	RET
+SYM_FUNC_END(__memcpy)
+EXPORT_SYMBOL(__memcpy)
+
+SYM_FUNC_ALIAS_MEMFUNC(memcpy, __memcpy)
+SYM_PIC_ALIAS(memcpy)
+EXPORT_SYMBOL(memcpy)
+
+SYM_FUNC_START_LOCAL(memcpy_orig)
+	movq %rdi, %rax
+
+	cmpq $0x20, %rdx
+	jb .Lhandle_tail
+
+	/*
+	 * We check whether memory false dependence could occur,
+	 * then jump to corresponding copy mode.
+	 */
+	cmp  %dil, %sil
+	jl .Lcopy_backward
+	subq $0x20, %rdx
+.Lcopy_forward_loop:
+	subq $0x20,	%rdx
+
+	/*
+	 * Move in blocks of 4x8 bytes:
+	 */
+	movq 0*8(%rsi),	%r8
+	movq 1*8(%rsi),	%r9
+	movq 2*8(%rsi),	%r10
+	movq 3*8(%rsi),	%r11
+	leaq 4*8(%rsi),	%rsi
+
+	movq %r8,	0*8(%rdi)
+	movq %r9,	1*8(%rdi)
+	movq %r10,	2*8(%rdi)
+	movq %r11,	3*8(%rdi)
+	leaq 4*8(%rdi),	%rdi
+	jae  .Lcopy_forward_loop
+	addl $0x20,	%edx
+	jmp  .Lhandle_tail
+
+.Lcopy_backward:
+	/*
+	 * Calculate copy position to tail.
+	 */
+	addq %rdx,	%rsi
+	addq %rdx,	%rdi
+	subq $0x20,	%rdx
+	/*
+	 * At most 3 ALU operations in one cycle,
+	 * so append NOPS in the same 16 bytes trunk.
+	 */
+	.p2align 4
+.Lcopy_backward_loop:
+	subq $0x20,	%rdx
+	movq -1*8(%rsi),	%r8
+	movq -2*8(%rsi),	%r9
+	movq -3*8(%rsi),	%r10
+	movq -4*8(%rsi),	%r11
+	leaq -4*8(%rsi),	%rsi
+	movq %r8,		-1*8(%rdi)
+	movq %r9,		-2*8(%rdi)
+	movq %r10,		-3*8(%rdi)
+	movq %r11,		-4*8(%rdi)
+	leaq -4*8(%rdi),	%rdi
+	jae  .Lcopy_backward_loop
+
+	/*
+	 * Calculate copy position to head.
+	 */
+	addl $0x20,	%edx
+	subq %rdx,	%rsi
+	subq %rdx,	%rdi
+.Lhandle_tail:
+	cmpl $16,	%edx
+	jb   .Lless_16bytes
+
+	/*
+	 * Move data from 16 bytes to 31 bytes.
+	 */
+	movq 0*8(%rsi), %r8
+	movq 1*8(%rsi),	%r9
+	movq -2*8(%rsi, %rdx),	%r10
+	movq -1*8(%rsi, %rdx),	%r11
+	movq %r8,	0*8(%rdi)
+	movq %r9,	1*8(%rdi)
+	movq %r10,	-2*8(%rdi, %rdx)
+	movq %r11,	-1*8(%rdi, %rdx)
+	RET
+	.p2align 4
+.Lless_16bytes:
+	cmpl $8,	%edx
+	jb   .Lless_8bytes
+	/*
+	 * Move data from 8 bytes to 15 bytes.
+	 */
+	movq 0*8(%rsi),	%r8
+	movq -1*8(%rsi, %rdx),	%r9
+	movq %r8,	0*8(%rdi)
+	movq %r9,	-1*8(%rdi, %rdx)
+	RET
+	.p2align 4
+.Lless_8bytes:
+	cmpl $4,	%edx
+	jb   .Lless_3bytes
+
+	/*
+	 * Move data from 4 bytes to 7 bytes.
+	 */
+	movl (%rsi), %ecx
+	movl -4(%rsi, %rdx), %r8d
+	movl %ecx, (%rdi)
+	movl %r8d, -4(%rdi, %rdx)
+	RET
+	.p2align 4
+.Lless_3bytes:
+	subl $1, %edx
+	jb .Lend
+	/*
+	 * Move data from 1 bytes to 3 bytes.
+	 */
+	movzbl (%rsi), %ecx
+	jz .Lstore_1byte
+	movzbq 1(%rsi), %r8
+	movzbq (%rsi, %rdx), %r9
+	movb %r8b, 1(%rdi)
+	movb %r9b, (%rdi, %rdx)
+.Lstore_1byte:
+	movb %cl, (%rdi)
+
+.Lend:
+	RET
+SYM_FUNC_END(memcpy_orig)
+

linux-7.2.3/tools/arch/x86/lib/memset_64.S

@@ -0,0 +1,119 @@
+/* SPDX-License-Identifier: GPL-2.0 */
+/* Copyright 2002 Andi Kleen, SuSE Labs */
+
+#include <linux/export.h>
+#include <linux/linkage.h>
+#include <linux/cfi_types.h>
+#include <asm/cpufeatures.h>
+#include <asm/alternative.h>
+
+.section .noinstr.text, "ax"
+
+/*
+ * ISO C memset - set a memory block to a byte value. This function uses fast
+ * string to get better performance than the original function. The code is
+ * simpler and shorter than the original function as well.
+ *
+ * rdi   destination
+ * rsi   value (char)
+ * rdx   count (bytes)
+ *
+ * rax   original destination
+ *
+ * The FSRS alternative should be done inline (avoiding the call and
+ * the disgusting return handling), but that would require some help
+ * from the compiler for better calling conventions.
+ *
+ * The 'rep stosb' itself is small enough to replace the call, but all
+ * the register moves blow up the code. And two of them are "needed"
+ * only for the return value that is the same as the source input,
+ * which the compiler could/should do much better anyway.
+ */
+SYM_TYPED_FUNC_START(__memset)
+	ALTERNATIVE "jmp memset_orig", "", X86_FEATURE_FSRS
+
+	movq %rdi,%r9
+	movb %sil,%al
+	movq %rdx,%rcx
+	rep stosb
+	movq %r9,%rax
+	RET
+SYM_FUNC_END(__memset)
+EXPORT_SYMBOL(__memset)
+
+SYM_FUNC_ALIAS_MEMFUNC(memset, __memset)
+SYM_PIC_ALIAS(memset)
+EXPORT_SYMBOL(memset)
+
+SYM_FUNC_START_LOCAL(memset_orig)
+	movq %rdi,%r10
+
+	/* expand byte value  */
+	movzbl %sil,%ecx
+	movabs $0x0101010101010101,%rax
+	imulq  %rcx,%rax
+
+	/* align dst */
+	movl  %edi,%r9d
+	andl  $7,%r9d
+	jnz  .Lbad_alignment
+.Lafter_bad_alignment:
+
+	movq  %rdx,%rcx
+	shrq  $6,%rcx
+	jz	 .Lhandle_tail
+
+	.p2align 4
+.Lloop_64:
+	decq  %rcx
+	movq  %rax,(%rdi)
+	movq  %rax,8(%rdi)
+	movq  %rax,16(%rdi)
+	movq  %rax,24(%rdi)
+	movq  %rax,32(%rdi)
+	movq  %rax,40(%rdi)
+	movq  %rax,48(%rdi)
+	movq  %rax,56(%rdi)
+	leaq  64(%rdi),%rdi
+	jnz    .Lloop_64
+
+	/* Handle tail in loops. The loops should be faster than hard
+	   to predict jump tables. */
+	.p2align 4
+.Lhandle_tail:
+	movl	%edx,%ecx
+	andl    $63&(~7),%ecx
+	jz 		.Lhandle_7
+	shrl	$3,%ecx
+	.p2align 4
+.Lloop_8:
+	decl   %ecx
+	movq  %rax,(%rdi)
+	leaq  8(%rdi),%rdi
+	jnz    .Lloop_8
+
+.Lhandle_7:
+	andl	$7,%edx
+	jz      .Lende
+	.p2align 4
+.Lloop_1:
+	decl    %edx
+	movb 	%al,(%rdi)
+	leaq	1(%rdi),%rdi
+	jnz     .Lloop_1
+
+.Lende:
+	movq	%r10,%rax
+	RET
+
+.Lbad_alignment:
+	cmpq $7,%rdx
+	jbe	.Lhandle_7
+	movq %rax,(%rdi)	/* unaligned store */
+	movq $8,%r8
+	subq %r9,%r8
+	addq %r8,%rdi
+	subq %r8,%rdx
+	jmp .Lafter_bad_alignment
+.Lfinal:
+SYM_FUNC_END(memset_orig)