diff --git a/arch/x86/crypto/chacha20-avx2-x86_64.S b/arch/x86/crypto/chacha20-avx2-x86_64.S
index 3a2dc3d..f3cd26f 100644
--- a/arch/x86/crypto/chacha20-avx2-x86_64.S
+++ b/arch/x86/crypto/chacha20-avx2-x86_64.S
@@ -45,7 +45,7 @@
 
 	vzeroupper
 	# 4 * 32 byte stack, 32-byte aligned
-	mov		%rsp, %r8
+	lea		8(%rsp),%r10
 	and		$~31, %rsp
 	sub		$0x80, %rsp
 
@@ -443,6 +443,6 @@
 	vmovdqu		%ymm15,0x01e0(%rsi)
 
 	vzeroupper
-	mov		%r8,%rsp
+	lea		-8(%r10),%rsp
 	ret
 ENDPROC(chacha20_8block_xor_avx2)
diff --git a/arch/x86/crypto/chacha20-ssse3-x86_64.S b/arch/x86/crypto/chacha20-ssse3-x86_64.S
index 3f511a7..512a2b5 100644
--- a/arch/x86/crypto/chacha20-ssse3-x86_64.S
+++ b/arch/x86/crypto/chacha20-ssse3-x86_64.S
@@ -160,7 +160,7 @@
 	# done with the slightly better performing SSSE3 byte shuffling,
 	# 7/12-bit word rotation uses traditional shift+OR.
 
-	mov		%rsp,%r11
+	lea		8(%rsp),%r10
 	sub		$0x80,%rsp
 	and		$~63,%rsp
 
@@ -625,6 +625,6 @@
 	pxor		%xmm1,%xmm15
 	movdqu		%xmm15,0xf0(%rsi)
 
-	mov		%r11,%rsp
+	lea		-8(%r10),%rsp
 	ret
 ENDPROC(chacha20_4block_xor_ssse3)