Add endbr64 where needed by inspection. Passes regresson tests.

ok jsing, and kind of tb an earlier version

28 files changed, 113 insertions, 1 deletions

diff --git a/lib/libcrypto/aes/asm/aes-x86_64.pl b/lib/libcrypto/aes/asm/aes-x86_64.pl
index b7399b552ad..78ba20ca59f 100755
--- a/lib/libcrypto/aes/asm/aes-x86_64.pl
+++ b/lib/libcrypto/aes/asm/aes-x86_64.pl
@@ -318,6 +318,7 @@ $code.=<<___;
 .type	_x86_64_AES_encrypt,\@abi-omnipotent
 .align	16
 _x86_64_AES_encrypt:
+	endbr64
 	xor	0($key),$s0			# xor with key
 	xor	4($key),$s1
 	xor	8($key),$s2
@@ -548,6 +549,7 @@ $code.=<<___;
 .type	_x86_64_AES_encrypt_compact,\@abi-omnipotent
 .align	16
 _x86_64_AES_encrypt_compact:
+	endbr64
 	lea	128($sbox),$inp			# size optimization
 	mov	0-128($inp),$acc1		# prefetch Te4
 	mov	32-128($inp),$acc2
@@ -593,6 +595,7 @@ $code.=<<___;
 .hidden	asm_AES_encrypt
 asm_AES_encrypt:
 AES_encrypt:
+	endbr64
 	push	%rbx
 	push	%rbp
 	push	%r12
@@ -884,6 +887,7 @@ $code.=<<___;
 .type	_x86_64_AES_decrypt,\@abi-omnipotent
 .align	16
 _x86_64_AES_decrypt:
+	endbr64
 	xor	0($key),$s0			# xor with key
 	xor	4($key),$s1
 	xor	8($key),$s2
@@ -1138,6 +1142,7 @@ $code.=<<___;
 .type	_x86_64_AES_decrypt_compact,\@abi-omnipotent
 .align	16
 _x86_64_AES_decrypt_compact:
+	endbr64
 	lea	128($sbox),$inp			# size optimization
 	mov	0-128($inp),$acc1		# prefetch Td4
 	mov	32-128($inp),$acc2
@@ -1192,6 +1197,7 @@ $code.=<<___;
 .hidden	asm_AES_decrypt
 asm_AES_decrypt:
 AES_decrypt:
+	endbr64
 	push	%rbx
 	push	%rbp
 	push	%r12
@@ -1291,6 +1297,7 @@ $code.=<<___;
 .type	AES_set_encrypt_key,\@function,3
 .align	16
 AES_set_encrypt_key:
+	endbr64
 	push	%rbx
 	push	%rbp
 	push	%r12			# redundant, but allows to share 
@@ -1316,6 +1323,7 @@ AES_set_encrypt_key:
 .type	_x86_64_AES_set_encrypt_key,\@abi-omnipotent
 .align	16
 _x86_64_AES_set_encrypt_key:
+	endbr64
 	mov	%esi,%ecx			# %ecx=bits
 	mov	%rdi,%rsi			# %rsi=userKey
 	mov	%rdx,%rdi			# %rdi=key
@@ -1561,6 +1569,7 @@ $code.=<<___;
 .type	AES_set_decrypt_key,\@function,3
 .align	16
 AES_set_decrypt_key:
+	endbr64
 	push	%rbx
 	push	%rbp
 	push	%r12
@@ -1660,6 +1669,7 @@ $code.=<<___;
 .hidden	asm_AES_cbc_encrypt
 asm_AES_cbc_encrypt:
 AES_cbc_encrypt:
+	endbr64
 	cmp	\$0,%rdx	# check length
 	je	.Lcbc_epilogue
 	pushfq
@@ -2551,6 +2561,7 @@ $code.=<<___;
 .type	block_se_handler,\@abi-omnipotent
 .align	16
 block_se_handler:
+	endbr64
 	push	%rsi
 	push	%rdi
 	push	%rbx
@@ -2609,6 +2620,7 @@ block_se_handler:
 .type	key_se_handler,\@abi-omnipotent
 .align	16
 key_se_handler:
+	endbr64
 	push	%rsi
 	push	%rdi
 	push	%rbx
@@ -2666,6 +2678,7 @@ key_se_handler:
 .type	cbc_se_handler,\@abi-omnipotent
 .align	16
 cbc_se_handler:
+	endbr64
 	push	%rsi
 	push	%rdi
 	push	%rbx
diff --git a/lib/libcrypto/aes/asm/aesni-sha1-x86_64.pl b/lib/libcrypto/aes/asm/aesni-sha1-x86_64.pl
index bafa906a050..879d16793f5 100644
--- a/lib/libcrypto/aes/asm/aesni-sha1-x86_64.pl
+++ b/lib/libcrypto/aes/asm/aesni-sha1-x86_64.pl
@@ -89,6 +89,7 @@ $code.=<<___;
 .type	aesni_cbc_sha1_enc,\@abi-omnipotent
 .align	16
 aesni_cbc_sha1_enc:
+	endbr64
 	# caller should check for SSSE3 and AES-NI bits
 	mov	OPENSSL_ia32cap_P+0(%rip),%r10d
 	mov	OPENSSL_ia32cap_P+4(%rip),%r11d
@@ -132,6 +133,7 @@ $code.=<<___;
 .type	aesni_cbc_sha1_enc_ssse3,\@function,6
 .align	16
 aesni_cbc_sha1_enc_ssse3:
+	endbr64
 	mov	`($win64?56:8)`(%rsp),$inp	# load 7th argument
 	#shr	\$6,$len			# debugging artefact
 	#jz	.Lepilogue_ssse3		# debugging artefact
@@ -650,6 +652,7 @@ $code.=<<___;
 .type	aesni_cbc_sha1_enc_avx,\@function,6
 .align	16
 aesni_cbc_sha1_enc_avx:
+	endbr64
 	mov	`($win64?56:8)`(%rsp),$inp	# load 7th argument
 	#shr	\$6,$len			# debugging artefact
 	#jz	.Lepilogue_avx			# debugging artefact
@@ -1100,6 +1103,7 @@ $code.=<<___;
 .type	ssse3_handler,\@abi-omnipotent
 .align	16
 ssse3_handler:
+	endbr64
 	push	%rsi
 	push	%rdi
 	push	%rbx
diff --git a/lib/libcrypto/aes/asm/aesni-x86_64.pl b/lib/libcrypto/aes/asm/aesni-x86_64.pl
index d394e6f696b..c261a379580 100644
--- a/lib/libcrypto/aes/asm/aesni-x86_64.pl
+++ b/lib/libcrypto/aes/asm/aesni-x86_64.pl
@@ -242,6 +242,7 @@ $code.=<<___;
 .type	${PREFIX}_encrypt,\@abi-omnipotent
 .align	16
 ${PREFIX}_encrypt:
+	endbr64
 	movups	($inp),$inout0		# load input
 	mov	240($key),$rounds	# key->rounds
 ___
@@ -255,6 +256,7 @@ $code.=<<___;
 .type	${PREFIX}_decrypt,\@abi-omnipotent
 .align	16
 ${PREFIX}_decrypt:
+	endbr64
 	movups	($inp),$inout0		# load input
 	mov	240($key),$rounds	# key->rounds
 ___
@@ -284,6 +286,7 @@ $code.=<<___;
 .type	_aesni_${dir}rypt3,\@abi-omnipotent
 .align	16
 _aesni_${dir}rypt3:
+	endbr64
 	$movkey	($key),$rndkey0
 	shr	\$1,$rounds
 	$movkey	16($key),$rndkey1
@@ -328,6 +331,7 @@ $code.=<<___;
 .type	_aesni_${dir}rypt4,\@abi-omnipotent
 .align	16
 _aesni_${dir}rypt4:
+	endbr64
 	$movkey	($key),$rndkey0
 	shr	\$1,$rounds
 	$movkey	16($key),$rndkey1
@@ -373,6 +377,7 @@ $code.=<<___;
 .type	_aesni_${dir}rypt6,\@abi-omnipotent
 .align	16
 _aesni_${dir}rypt6:
+	endbr64
 	$movkey		($key),$rndkey0
 	shr		\$1,$rounds
 	$movkey		16($key),$rndkey1
@@ -437,6 +442,7 @@ $code.=<<___;
 .type	_aesni_${dir}rypt8,\@abi-omnipotent
 .align	16
 _aesni_${dir}rypt8:
+	endbr64
 	$movkey		($key),$rndkey0
 	shr		\$1,$rounds
 	$movkey		16($key),$rndkey1
@@ -525,6 +531,7 @@ $code.=<<___;
 .type	aesni_ecb_encrypt,\@function,5
 .align	16
 aesni_ecb_encrypt:
+	endbr64
 	and	\$-16,$len
 	jz	.Lecb_ret
 
@@ -830,6 +837,7 @@ $code.=<<___;
 .type	aesni_ccm64_encrypt_blocks,\@function,6
 .align	16
 aesni_ccm64_encrypt_blocks:
+	endbr64
 ___
 $code.=<<___ if ($win64);
 	lea	-0x58(%rsp),%rsp
@@ -2478,6 +2486,7 @@ $code.=<<___;
 .type	${PREFIX}_set_decrypt_key,\@abi-omnipotent
 .align	16
 ${PREFIX}_set_decrypt_key:
+	endbr64
 	sub	\$8,%rsp
 	call	__aesni_set_encrypt_key
 	shl	\$4,$bits		# rounds-1 after _aesni_set_encrypt_key
@@ -2528,6 +2537,7 @@ $code.=<<___;
 .type	${PREFIX}_set_encrypt_key,\@abi-omnipotent
 .align	16
 ${PREFIX}_set_encrypt_key:
+	endbr64
 __aesni_set_encrypt_key:
 	sub	\$8,%rsp
 	mov	\$-1,%rax
@@ -2749,6 +2759,7 @@ $code.=<<___ if ($PREFIX eq "aesni");
 .type	ecb_se_handler,\@abi-omnipotent
 .align	16
 ecb_se_handler:
+	endbr64
 	push	%rsi
 	push	%rdi
 	push	%rbx
@@ -2768,6 +2779,7 @@ ecb_se_handler:
 .type	ccm64_se_handler,\@abi-omnipotent
 .align	16
 ccm64_se_handler:
+	endbr64
 	push	%rsi
 	push	%rdi
 	push	%rbx
@@ -2809,6 +2821,7 @@ ccm64_se_handler:
 .type	ctr32_se_handler,\@abi-omnipotent
 .align	16
 ctr32_se_handler:
+	endbr64
 	push	%rsi
 	push	%rdi
 	push	%rbx
@@ -2844,6 +2857,7 @@ ctr32_se_handler:
 .type	xts_se_handler,\@abi-omnipotent
 .align	16
 xts_se_handler:
+	endbr64
 	push	%rsi
 	push	%rdi
 	push	%rbx
@@ -2885,6 +2899,7 @@ $code.=<<___;
 .type	cbc_se_handler,\@abi-omnipotent
 .align	16
 cbc_se_handler:
+	endbr64
 	push	%rsi
 	push	%rdi
 	push	%rbx
diff --git a/lib/libcrypto/aes/asm/bsaes-x86_64.pl b/lib/libcrypto/aes/asm/bsaes-x86_64.pl
index 20e9e1f71f3..7098ba27f49 100644
--- a/lib/libcrypto/aes/asm/bsaes-x86_64.pl
+++ b/lib/libcrypto/aes/asm/bsaes-x86_64.pl
@@ -813,6 +813,7 @@ $code.=<<___;
 .type	_bsaes_encrypt8,\@abi-omnipotent
 .align	64
 _bsaes_encrypt8:
+	endbr64
 	lea	.LBS0(%rip), $const	# constants table
 
 	movdqa	($key), @XMM[9]		# round 0 key
@@ -877,6 +878,7 @@ $code.=<<___;
 .type	_bsaes_decrypt8,\@abi-omnipotent
 .align	64
 _bsaes_decrypt8:
+	endbr64
 	lea	.LBS0(%rip), $const	# constants table
 
 	movdqa	($key), @XMM[9]		# round 0 key
@@ -968,6 +970,7 @@ $code.=<<___;
 .type	_bsaes_key_convert,\@abi-omnipotent
 .align	16
 _bsaes_key_convert:
+	endbr64
 	lea	.Lmasks(%rip), $const
 	movdqu	($inp), %xmm7		# load round 0 key
 	lea	0x10($inp), $inp
@@ -1057,6 +1060,7 @@ $code.=<<___;
 .type	bsaes_enc_key_convert,\@function,2
 .align	16
 bsaes_enc_key_convert:
+	endbr64
 	mov	240($inp),%r10d		# pass rounds
 	mov	$inp,%rcx		# pass key
 	mov	$out,%rax		# pass key schedule
@@ -1071,6 +1075,7 @@ bsaes_enc_key_convert:
 .align	16
 bsaes_encrypt_128:
 .Lenc128_loop:
+	endbr64
 	movdqu	0x00($inp), @XMM[0]	# load input
 	movdqu	0x10($inp), @XMM[1]
 	movdqu	0x20($inp), @XMM[2]
@@ -1103,6 +1108,7 @@ bsaes_encrypt_128:
 .type	bsaes_dec_key_convert,\@function,2
 .align	16
 bsaes_dec_key_convert:
+	endbr64
 	mov	240($inp),%r10d		# pass rounds
 	mov	$inp,%rcx		# pass key
 	mov	$out,%rax		# pass key schedule
@@ -1117,6 +1123,7 @@ bsaes_dec_key_convert:
 .type	bsaes_decrypt_128,\@function,4
 .align	16
 bsaes_decrypt_128:
+	endbr64
 .Ldec128_loop:
 	movdqu	0x00($inp), @XMM[0]	# load input
 	movdqu	0x10($inp), @XMM[1]
@@ -1162,6 +1169,7 @@ $code.=<<___;
 .type	bsaes_ecb_encrypt_blocks,\@abi-omnipotent
 .align	16
 bsaes_ecb_encrypt_blocks:
+	endbr64
 	mov	%rsp, %rax
 .Lecb_enc_prologue:
 	push	%rbp
@@ -1363,6 +1371,7 @@ $code.=<<___;
 .type	bsaes_ecb_decrypt_blocks,\@abi-omnipotent
 .align	16
 bsaes_ecb_decrypt_blocks:
+	endbr64
 	mov	%rsp, %rax
 .Lecb_dec_prologue:
 	push	%rbp
@@ -1568,6 +1577,7 @@ $code.=<<___;
 .type	bsaes_cbc_encrypt,\@abi-omnipotent
 .align	16
 bsaes_cbc_encrypt:
+	endbr64
 ___
 $code.=<<___ if ($win64);
 	mov	48(%rsp),$arg6		# pull direction flag
@@ -1855,6 +1865,7 @@ $code.=<<___;
 .type	bsaes_ctr32_encrypt_blocks,\@abi-omnipotent
 .align	16
 bsaes_ctr32_encrypt_blocks:
+	endbr64
 	mov	%rsp, %rax
 .Lctr_enc_prologue:
 	push	%rbp
@@ -2096,6 +2107,7 @@ $code.=<<___;
 .type	bsaes_xts_encrypt,\@abi-omnipotent
 .align	16
 bsaes_xts_encrypt:
+	endbr64
 	mov	%rsp, %rax
 .Lxts_enc_prologue:
 	push	%rbp
@@ -2477,6 +2489,7 @@ $code.=<<___;
 .type	bsaes_xts_decrypt,\@abi-omnipotent
 .align	16
 bsaes_xts_decrypt:
+	endbr64
 	mov	%rsp, %rax
 .Lxts_dec_prologue:
 	push	%rbp
@@ -2953,6 +2966,7 @@ $code.=<<___;
 .type	se_handler,\@abi-omnipotent
 .align	16
 se_handler:
+	endbr64
 	push	%rsi
 	push	%rdi
 	push	%rbx
diff --git a/lib/libcrypto/aes/asm/vpaes-x86_64.pl b/lib/libcrypto/aes/asm/vpaes-x86_64.pl
index 3ffb1a3038f..8ff8d8602bc 100644
--- a/lib/libcrypto/aes/asm/vpaes-x86_64.pl
+++ b/lib/libcrypto/aes/asm/vpaes-x86_64.pl
@@ -82,6 +82,7 @@ $code.=<<___;
 .type	_vpaes_encrypt_core,\@abi-omnipotent
 .align 16
 _vpaes_encrypt_core:
+	endbr64
 	mov	%rdx,	%r9
 	mov	\$16,	%r11
 	mov	240(%rdx),%eax
@@ -172,6 +173,7 @@ _vpaes_encrypt_core:
 .type	_vpaes_decrypt_core,\@abi-omnipotent
 .align	16
 _vpaes_decrypt_core:
+	endbr64
 	mov	%rdx,	%r9		# load key
 	mov	240(%rdx),%eax
 	movdqa	%xmm9,	%xmm1
@@ -279,6 +281,7 @@ _vpaes_decrypt_core:
 .type	_vpaes_schedule_core,\@abi-omnipotent
 .align	16
 _vpaes_schedule_core:
+	endbr64
 	# rdi = key
 	# rsi = size in bits
 	# rdx = buffer
@@ -464,6 +467,7 @@ _vpaes_schedule_core:
 .type	_vpaes_schedule_192_smear,\@abi-omnipotent
 .align	16
 _vpaes_schedule_192_smear:
+	endbr64
 	pshufd	\$0x80,	%xmm6,	%xmm0	# d c 0 0 -> c 0 0 0
 	pxor	%xmm0,	%xmm6		# -> c+d c 0 0
 	pshufd	\$0xFE,	%xmm7,	%xmm0	# b a _ _ -> b b b a
@@ -495,6 +499,7 @@ _vpaes_schedule_192_smear:
 .type	_vpaes_schedule_round,\@abi-omnipotent
 .align	16
 _vpaes_schedule_round:
+	endbr64
 	# extract rcon from xmm8
 	pxor	%xmm1,	%xmm1
 	palignr	\$15,	%xmm8,	%xmm1
@@ -562,6 +567,7 @@ _vpaes_schedule_low_round:
 .type	_vpaes_schedule_transform,\@abi-omnipotent
 .align	16
 _vpaes_schedule_transform:
+	endbr64
 	movdqa	%xmm9,	%xmm1
 	pandn	%xmm0,	%xmm1
 	psrld	\$4,	%xmm1
@@ -600,6 +606,7 @@ _vpaes_schedule_transform:
 .type	_vpaes_schedule_mangle,\@abi-omnipotent
 .align	16
 _vpaes_schedule_mangle:
+	endbr64
 	movdqa	%xmm0,	%xmm4	# save xmm0 for later
 	movdqa	.Lk_mc_forward(%rip),%xmm5
 	test	%rcx, 	%rcx
@@ -673,6 +680,7 @@ _vpaes_schedule_mangle:
 .type	${PREFIX}_set_encrypt_key,\@function,3
 .align	16
 ${PREFIX}_set_encrypt_key:
+	endbr64
 ___
 $code.=<<___ if ($win64);
 	lea	-0xb8(%rsp),%rsp
@@ -721,6 +729,7 @@ $code.=<<___;
 .type	${PREFIX}_set_decrypt_key,\@function,3
 .align	16
 ${PREFIX}_set_decrypt_key:
+	endbr64
 ___
 $code.=<<___ if ($win64);
 	lea	-0xb8(%rsp),%rsp
@@ -774,6 +783,7 @@ $code.=<<___;
 .type	${PREFIX}_encrypt,\@function,3
 .align	16
 ${PREFIX}_encrypt:
+	endbr64
 ___
 $code.=<<___ if ($win64);
 	lea	-0xb8(%rsp),%rsp
@@ -817,6 +827,7 @@ $code.=<<___;
 .type	${PREFIX}_decrypt,\@function,3
 .align	16
 ${PREFIX}_decrypt:
+	endbr64
 ___
 $code.=<<___ if ($win64);
 	lea	-0xb8(%rsp),%rsp
@@ -866,6 +877,7 @@ $code.=<<___;
 .type	${PREFIX}_cbc_encrypt,\@function,6
 .align	16
 ${PREFIX}_cbc_encrypt:
+	endbr64
 	xchg	$key,$len
 ___
 ($len,$key)=($key,$len);
@@ -949,6 +961,7 @@ $code.=<<___;
 .type	_vpaes_preheat,\@abi-omnipotent
 .align	16
 _vpaes_preheat:
+	endbr64
 	lea	.Lk_s0F(%rip), %r10
 	movdqa	-0x20(%r10), %xmm10	# .Lk_inv
 	movdqa	-0x10(%r10), %xmm11	# .Lk_inv+16
@@ -1079,6 +1092,7 @@ $code.=<<___;
 .type	se_handler,\@abi-omnipotent
 .align	16
 se_handler:
+	endbr64
 	push	%rsi
 	push	%rdi
 	push	%rbx
diff --git a/lib/libcrypto/bn/arch/amd64/bignum_add.S b/lib/libcrypto/bn/arch/amd64/bignum_add.S
index d56fa5e3a83..06298ca69e1 100644
--- a/lib/libcrypto/bn/arch/amd64/bignum_add.S
+++ b/lib/libcrypto/bn/arch/amd64/bignum_add.S
@@ -49,6 +49,7 @@
 
 
 S2N_BN_SYMBOL(bignum_add):
+	endbr64
 
 #if WINDOWS_ABI
         push    rdi
diff --git a/lib/libcrypto/bn/arch/amd64/bignum_cmadd.S b/lib/libcrypto/bn/arch/amd64/bignum_cmadd.S
index 1dc1e58705d..5ad712749f2 100644
--- a/lib/libcrypto/bn/arch/amd64/bignum_cmadd.S
+++ b/lib/libcrypto/bn/arch/amd64/bignum_cmadd.S
@@ -54,6 +54,7 @@
 
 
 S2N_BN_SYMBOL(bignum_cmadd):
+	endbr64
 
 #if WINDOWS_ABI
         push    rdi
diff --git a/lib/libcrypto/bn/arch/amd64/bignum_cmul.S b/lib/libcrypto/bn/arch/amd64/bignum_cmul.S
index c1a23ccea4d..9199c8f48b6 100644
--- a/lib/libcrypto/bn/arch/amd64/bignum_cmul.S
+++ b/lib/libcrypto/bn/arch/amd64/bignum_cmul.S
@@ -51,6 +51,7 @@
 
 
 S2N_BN_SYMBOL(bignum_cmul):
+	endbr64
 
 #if WINDOWS_ABI
         push    rdi
diff --git a/lib/libcrypto/bn/arch/amd64/bignum_mul.S b/lib/libcrypto/bn/arch/amd64/bignum_mul.S
index 42ac988a197..2d7ed190912 100644
--- a/lib/libcrypto/bn/arch/amd64/bignum_mul.S
+++ b/lib/libcrypto/bn/arch/amd64/bignum_mul.S
@@ -59,6 +59,7 @@
 
 
 S2N_BN_SYMBOL(bignum_mul):
+	endbr64
 
 #if WINDOWS_ABI
         push    rdi
diff --git a/lib/libcrypto/bn/arch/amd64/bignum_mul_4_8_alt.S b/lib/libcrypto/bn/arch/amd64/bignum_mul_4_8_alt.S
index 3b7848b285f..f02b09b2887 100644
--- a/lib/libcrypto/bn/arch/amd64/bignum_mul_4_8_alt.S
+++ b/lib/libcrypto/bn/arch/amd64/bignum_mul_4_8_alt.S
@@ -72,6 +72,7 @@
         adc     h, rdx
 
 S2N_BN_SYMBOL(bignum_mul_4_8_alt):
+	endbr64
 
 #if WINDOWS_ABI
         push    rdi
diff --git a/lib/libcrypto/bn/arch/amd64/bignum_mul_8_16_alt.S b/lib/libcrypto/bn/arch/amd64/bignum_mul_8_16_alt.S
index 1be37840df6..97be83e1f72 100644
--- a/lib/libcrypto/bn/arch/amd64/bignum_mul_8_16_alt.S
+++ b/lib/libcrypto/bn/arch/amd64/bignum_mul_8_16_alt.S
@@ -72,6 +72,7 @@
         adc     h, rdx
 
 S2N_BN_SYMBOL(bignum_mul_8_16_alt):
+	endbr64
 
 #if WINDOWS_ABI
         push    rdi
diff --git a/lib/libcrypto/bn/arch/amd64/bignum_sqr.S b/lib/libcrypto/bn/arch/amd64/bignum_sqr.S
index 2e05b9c179a..c4a0cabf352 100644
--- a/lib/libcrypto/bn/arch/amd64/bignum_sqr.S
+++ b/lib/libcrypto/bn/arch/amd64/bignum_sqr.S
@@ -62,6 +62,7 @@
 #define llshort ebp
 
 S2N_BN_SYMBOL(bignum_sqr):
+	endbr64
 
 #if WINDOWS_ABI
         push    rdi
diff --git a/lib/libcrypto/bn/arch/amd64/bignum_sqr_4_8_alt.S b/lib/libcrypto/bn/arch/amd64/bignum_sqr_4_8_alt.S
index a635177c650..b228414dcea 100644
--- a/lib/libcrypto/bn/arch/amd64/bignum_sqr_4_8_alt.S
+++ b/lib/libcrypto/bn/arch/amd64/bignum_sqr_4_8_alt.S
@@ -71,6 +71,7 @@
         adc     c, 0
 
 S2N_BN_SYMBOL(bignum_sqr_4_8_alt):
+	endbr64
 
 #if WINDOWS_ABI
         push    rdi
diff --git a/lib/libcrypto/bn/arch/amd64/bignum_sqr_8_16_alt.S b/lib/libcrypto/bn/arch/amd64/bignum_sqr_8_16_alt.S
index f698202d29c..04efeec7e28 100644
--- a/lib/libcrypto/bn/arch/amd64/bignum_sqr_8_16_alt.S
+++ b/lib/libcrypto/bn/arch/amd64/bignum_sqr_8_16_alt.S
@@ -103,6 +103,7 @@
         adc     c, 0
 
 S2N_BN_SYMBOL(bignum_sqr_8_16_alt):
+	endbr64
 
 #if WINDOWS_ABI
         push    rdi
diff --git a/lib/libcrypto/bn/arch/amd64/bignum_sub.S b/lib/libcrypto/bn/arch/amd64/bignum_sub.S
index f8e1fe35a8f..11a9bd7edd4 100644
--- a/lib/libcrypto/bn/arch/amd64/bignum_sub.S
+++ b/lib/libcrypto/bn/arch/amd64/bignum_sub.S
@@ -49,6 +49,7 @@
 
 
 S2N_BN_SYMBOL(bignum_sub):
+	endbr64
 
 #if WINDOWS_ABI
         push    rdi
diff --git a/lib/libcrypto/bn/arch/amd64/word_clz.S b/lib/libcrypto/bn/arch/amd64/word_clz.S
index 025e98f9cb0..464a9d90fc3 100644
--- a/lib/libcrypto/bn/arch/amd64/word_clz.S
+++ b/lib/libcrypto/bn/arch/amd64/word_clz.S
@@ -30,6 +30,7 @@
         .text
 
 S2N_BN_SYMBOL(word_clz):
+	endbr64
 
 #if WINDOWS_ABI
         push    rdi
diff --git a/lib/libcrypto/bn/asm/modexp512-x86_64.pl b/lib/libcrypto/bn/asm/modexp512-x86_64.pl
index 2e71a7f03dc..af78fff5412 100644
--- a/lib/libcrypto/bn/asm/modexp512-x86_64.pl
+++ b/lib/libcrypto/bn/asm/modexp512-x86_64.pl
@@ -347,6 +347,7 @@ $code.=<<___;
 .type	MULADD_128x512,\@abi-omnipotent
 .align	16
 MULADD_128x512:
+	endbr64
 ___
 	&MULSTEP_512([map("%r$_",(8..15))], "(+8*0)(%rcx)", "%rsi", "%rbp", "%rbx");
 $code.=<<___;
@@ -414,6 +415,7 @@ $code.=<<___;
 .type	mont_reduce,\@abi-omnipotent
 .align	16
 mont_reduce:
+	endbr64
 ___
 
 my $STACK_DEPTH         =  8;
@@ -676,6 +678,7 @@ $code.=<<___;
 .type	mont_mul_a3b,\@abi-omnipotent
 .align	16
 mont_mul_a3b:
+	endbr64
 	#
 	# multiply tmp = src1 * src2
 	# For multiply: dst = rcx, src1 = rdi, src2 = rsi
@@ -1077,6 +1080,7 @@ $code.=<<___;
 .type	sqr_reduce,\@abi-omnipotent
 .align	16
 sqr_reduce:
+	endbr64
 	 mov	(+$pResult_offset+8)(%rsp), %rcx
 ___
 	&SQR_512("%rsp+$tmp16_offset+8", "%rcx", [map("%r$_",(10..15,8..9))], "%rbx", "%rbp", "%rsi", "%rdi");
@@ -1106,6 +1110,7 @@ $code.=<<___;
 .globl	mod_exp_512
 .type	mod_exp_512,\@function,4
 mod_exp_512:
+	endbr64
 	 push	%rbp
 	 push	%rbx
 	 push	%r12
diff --git a/lib/libcrypto/bn/asm/x86_64-mont.pl b/lib/libcrypto/bn/asm/x86_64-mont.pl
index cae7309d5ba..6f5ab331e2f 100755
--- a/lib/libcrypto/bn/asm/x86_64-mont.pl
+++ b/lib/libcrypto/bn/asm/x86_64-mont.pl
@@ -63,6 +63,7 @@ $code=<<___;
 .type	bn_mul_mont,\@function,6
 .align	16
 bn_mul_mont:
+	endbr64
 	test	\$3,${num}d
 	jnz	.Lmul_enter
 	cmp	\$8,${num}d
@@ -278,6 +279,7 @@ $code.=<<___;
 .align	16
 bn_mul4x_mont:
 .Lmul4x_enter:
+	endbr64
 	push	%rbx
 	push	%rbp
 	push	%r12
@@ -705,6 +707,7 @@ $code.=<<___;
 .align	16
 bn_sqr4x_mont:
 .Lsqr4x_enter:
+	endbr64
 	push	%rbx
 	push	%rbp
 	push	%r12
diff --git a/lib/libcrypto/bn/asm/x86_64-mont5.pl b/lib/libcrypto/bn/asm/x86_64-mont5.pl
index 7b9c6df2739..3b3325a6ccc 100755
--- a/lib/libcrypto/bn/asm/x86_64-mont5.pl
+++ b/lib/libcrypto/bn/asm/x86_64-mont5.pl
@@ -57,6 +57,7 @@ $code=<<___;
 .type	bn_mul_mont_gather5,\@function,6
 .align	64
 bn_mul_mont_gather5:
+	endbr64
 	test	\$3,${num}d
 	jnz	.Lmul_enter
 	cmp	\$8,${num}d
@@ -387,6 +388,7 @@ $code.=<<___;
 .type	bn_mul4x_mont_gather5,\@function,6
 .align	16
 bn_mul4x_mont_gather5:
+	endbr64
 .Lmul4x_enter:
 	mov	${num}d,${num}d
 	movd	`($win64?56:8)`(%rsp),%xmm5	# load 7th argument
@@ -925,6 +927,7 @@ $code.=<<___;
 .type	bn_scatter5,\@abi-omnipotent
 .align	16
 bn_scatter5:
+	endbr64
 	cmp	\$0, $num
 	jz	.Lscatter_epilogue
 	lea	($tbl,$idx,8),$tbl
@@ -943,6 +946,7 @@ bn_scatter5:
 .type	bn_gather5,\@abi-omnipotent
 .align	16
 bn_gather5:
+	endbr64
 .LSEH_begin_bn_gather5:			# Win64 thing, but harmless in other cases
 	# I can't trust assembler to use specific encoding:-(
 	.byte	0x4c,0x8d,0x14,0x24			# lea    (%rsp),%r10
@@ -1053,6 +1057,7 @@ $code.=<<___;
 .type	mul_handler,\@abi-omnipotent
 .align	16
 mul_handler:
+	endbr64
 	push	%rsi
 	push	%rdi
 	push	%rbx
diff --git a/lib/libcrypto/camellia/asm/cmll-x86_64.pl b/lib/libcrypto/camellia/asm/cmll-x86_64.pl
index 586e5d6e93f..3ceed3e8991 100644
--- a/lib/libcrypto/camellia/asm/cmll-x86_64.pl
+++ b/lib/libcrypto/camellia/asm/cmll-x86_64.pl
@@ -116,6 +116,7 @@ $code=<<___;
 .type	Camellia_EncryptBlock,\@abi-omnipotent
 .align	16
 Camellia_EncryptBlock:
+	endbr64
 	movl	\$128,%eax
 	subl	$arg0d,%eax
 	movl	\$3,$arg0d
@@ -128,6 +129,7 @@ Camellia_EncryptBlock:
 .align	16
 .Lenc_rounds:
 Camellia_EncryptBlock_Rounds:
+	endbr64
 	push	%rbx
 	push	%rbp
 	push	%r13
@@ -176,6 +178,7 @@ Camellia_EncryptBlock_Rounds:
 .type	_x86_64_Camellia_encrypt,\@abi-omnipotent
 .align	16
 _x86_64_Camellia_encrypt:
+	endbr64
 	xor	0($key),@S[1]
 	xor	4($key),@S[0]		# ^=key[0-3]
 	xor	8($key),@S[3]
@@ -226,6 +229,7 @@ $code.=<<___;
 .type	Camellia_DecryptBlock,\@abi-omnipotent
 .align	16
 Camellia_DecryptBlock:
+	endbr64
 	movl	\$128,%eax
 	subl	$arg0d,%eax
 	movl	\$3,$arg0d
@@ -238,6 +242,7 @@ Camellia_DecryptBlock:
 .align	16
 .Ldec_rounds:
 Camellia_DecryptBlock_Rounds:
+	endbr64
 	push	%rbx
 	push	%rbp
 	push	%r13
@@ -286,6 +291,7 @@ Camellia_DecryptBlock_Rounds:
 .type	_x86_64_Camellia_decrypt,\@abi-omnipotent
 .align	16
 _x86_64_Camellia_decrypt:
+	endbr64
 	xor	0($key),@S[1]
 	xor	4($key),@S[0]		# ^=key[0-3]
 	xor	8($key),@S[3]
@@ -400,6 +406,7 @@ $code.=<<___;
 .type	Camellia_Ekeygen,\@function,3
 .align	16
 Camellia_Ekeygen:
+	endbr64
 	push	%rbx
 	push	%rbp
 	push	%r13
@@ -630,6 +637,7 @@ $code.=<<___;
 .type	Camellia_cbc_encrypt,\@function,6
 .align	16
 Camellia_cbc_encrypt:
+	endbr64
 	cmp	\$0,%rdx
 	je	.Lcbc_abort
 	push	%rbx
diff --git a/lib/libcrypto/md5/asm/md5-x86_64.pl b/lib/libcrypto/md5/asm/md5-x86_64.pl
index c902a1b532f..06d69094f44 100755
--- a/lib/libcrypto/md5/asm/md5-x86_64.pl
+++ b/lib/libcrypto/md5/asm/md5-x86_64.pl
@@ -128,6 +128,7 @@ $code .= <<EOF;
 .globl md5_block_asm_data_order
 .type md5_block_asm_data_order,\@function,3
 md5_block_asm_data_order:
+	endbr64
 	push	%rbp
 	push	%rbx
 	push	%r12
diff --git a/lib/libcrypto/modes/asm/ghash-x86_64.pl b/lib/libcrypto/modes/asm/ghash-x86_64.pl
index 71d0822ac91..9ce0c381410 100644
--- a/lib/libcrypto/modes/asm/ghash-x86_64.pl
+++ b/lib/libcrypto/modes/asm/ghash-x86_64.pl
@@ -412,6 +412,7 @@ $code.=<<___;
 .type	gcm_init_clmul,\@abi-omnipotent
 .align	16
 gcm_init_clmul:
+	endbr64
 	movdqu		($Xip),$Hkey
 	pshufd		\$0b01001110,$Hkey,$Hkey	# dword swap
 
@@ -449,6 +450,7 @@ $code.=<<___;
 .type	gcm_gmult_clmul,\@abi-omnipotent
 .align	16
 gcm_gmult_clmul:
+	endbr64
 	movdqu		($Xip),$Xi
 	movdqa		.Lbswap_mask(%rip),$T3
 	movdqu		($Htbl),$Hkey
@@ -476,6 +478,7 @@ $code.=<<___;
 .type	gcm_ghash_clmul,\@abi-omnipotent
 .align	16
 gcm_ghash_clmul:
+	endbr64
 ___
 $code.=<<___ if ($win64);
 .LSEH_begin_gcm_ghash_clmul:
@@ -686,6 +689,7 @@ $code.=<<___;
 .type	se_handler,\@abi-omnipotent
 .align	16
 se_handler:
+	endbr64
 	push	%rsi
 	push	%rdi
 	push	%rbx
diff --git a/lib/libcrypto/rc4/asm/rc4-md5-x86_64.pl b/lib/libcrypto/rc4/asm/rc4-md5-x86_64.pl
index 501d9e936bb..6d058bd9d4b 100644
--- a/lib/libcrypto/rc4/asm/rc4-md5-x86_64.pl
+++ b/lib/libcrypto/rc4/asm/rc4-md5-x86_64.pl
@@ -109,6 +109,7 @@ $code.=<<___;
 .globl	$func
 .type	$func,\@function,$nargs
 $func:
+	endbr64
 	cmp	\$0,$len
 	je	.Labort
 	push	%rbx
@@ -453,6 +454,7 @@ $code.=<<___;
 .type	RC4_set_key,\@function,3
 .align	16
 RC4_set_key:
+	endbr64
 	lea	8($dat),$dat
 	lea	($inp,$len),$inp
 	neg	$len
@@ -494,6 +496,7 @@ RC4_set_key:
 .type	RC4_options,\@abi-omnipotent
 .align	16
 RC4_options:
+	endbr64
 	lea	.Lopts(%rip),%rax
 	ret
 .align	64
diff --git a/lib/libcrypto/rc4/asm/rc4-x86_64.pl b/lib/libcrypto/rc4/asm/rc4-x86_64.pl
index 8de869489f6..2bac7d744dd 100755
--- a/lib/libcrypto/rc4/asm/rc4-x86_64.pl
+++ b/lib/libcrypto/rc4/asm/rc4-x86_64.pl
@@ -127,7 +127,9 @@ $code=<<___;
 .globl	RC4
 .type	RC4,\@function,4
 .align	16
-RC4:	or	$len,$len
+RC4:
+	endbr64
+	or	$len,$len
 	jne	.Lentry
 	ret
 .Lentry:
@@ -433,6 +435,7 @@ $code.=<<___;
 .type	RC4_set_key,\@function,3
 .align	16
 RC4_set_key:
+	endbr64
 	lea	8($dat),$dat
 	lea	($inp,$len),$inp
 	neg	$len
@@ -505,6 +508,7 @@ RC4_set_key:
 .type	RC4_options,\@abi-omnipotent
 .align	16
 RC4_options:
+	endbr64
 	lea	.Lopts(%rip),%rax
 	mov	OPENSSL_ia32cap_P(%rip),%edx
 	bt	\$IA32CAP_BIT0_INTELP4,%edx
diff --git a/lib/libcrypto/sha/asm/sha1-x86_64.pl b/lib/libcrypto/sha/asm/sha1-x86_64.pl
index 43eee73c4ab..e15ff47f88b 100755
--- a/lib/libcrypto/sha/asm/sha1-x86_64.pl
+++ b/lib/libcrypto/sha/asm/sha1-x86_64.pl
@@ -222,6 +222,7 @@ $code.=<<___;
 .type	sha1_block_data_order,\@function,3
 .align	16
 sha1_block_data_order:
+	endbr64
 	mov	OPENSSL_ia32cap_P+0(%rip),%r9d
 	mov	OPENSSL_ia32cap_P+4(%rip),%r8d
 	test	\$IA32CAP_MASK1_SSSE3,%r8d		# check SSSE3 bit
@@ -309,6 +310,7 @@ $code.=<<___;
 .align	16
 sha1_block_data_order_ssse3:
 _ssse3_shortcut:
+	endbr64
 	push	%rbx
 	push	%rbp
 	push	%r12
@@ -729,6 +731,7 @@ $code.=<<___;
 .align	16
 sha1_block_data_order_avx:
 _avx_shortcut:
+	endbr64
 	push	%rbx
 	push	%rbp
 	push	%r12
@@ -1099,6 +1102,7 @@ $code.=<<___;
 .type	se_handler,\@abi-omnipotent
 .align	16
 se_handler:
+	endbr64
 	push	%rsi
 	push	%rdi
 	push	%rbx
diff --git a/lib/libcrypto/sha/asm/sha512-x86_64.pl b/lib/libcrypto/sha/asm/sha512-x86_64.pl
index 0517eab6602..120693feeb7 100755
--- a/lib/libcrypto/sha/asm/sha512-x86_64.pl
+++ b/lib/libcrypto/sha/asm/sha512-x86_64.pl
@@ -175,6 +175,7 @@ $code=<<___;
 .type	$func,\@function,4
 .align	16
 $func:
+	endbr64
 	push	%rbx
 	push	%rbp
 	push	%r12
diff --git a/lib/libcrypto/whrlpool/asm/wp-x86_64.pl b/lib/libcrypto/whrlpool/asm/wp-x86_64.pl
index de5d3acfb40..7958f6d2897 100644
--- a/lib/libcrypto/whrlpool/asm/wp-x86_64.pl
+++ b/lib/libcrypto/whrlpool/asm/wp-x86_64.pl
@@ -57,6 +57,7 @@ $code=<<___;
 .type	$func,\@function,3
 .align	16
 $func:
+	endbr64
 	push	%rbx
 	push	%rbp
 	push	%r12
diff --git a/lib/libcrypto/x86_64cpuid.pl b/lib/libcrypto/x86_64cpuid.pl
index 1b67d1110f9..dc56732a28b 100644
--- a/lib/libcrypto/x86_64cpuid.pl
+++ b/lib/libcrypto/x86_64cpuid.pl
@@ -18,6 +18,7 @@ print<<___;
 .extern		OPENSSL_cpuid_setup
 .hidden		OPENSSL_cpuid_setup
 .section	.init
+	endbr64
 	call	OPENSSL_cpuid_setup
 
 .extern	OPENSSL_ia32cap_P
@@ -29,6 +30,7 @@ print<<___;
 .type	OPENSSL_ia32_cpuid,\@abi-omnipotent
 .align	16
 OPENSSL_ia32_cpuid:
+	endbr64
 	mov	%rbx,%r8		# save %rbx
 
 	xor	%eax,%eax