Revert "x86: bring back rep movsq for user access on CPUs without ERMS"

JIRA: https://issues.redhat.com/browse/RHEL-79780
Upstream Status: RHEL Only

This reverts commit 4c7f1f1.

The change "x86: bring back rep movsq for user access", while aiming at
fixing a performance regression reported at RHEL-74389, unfortunately
brought another regression on lnst mlx5 multistream tcp test. This is like
a short blanket problem where looks like we fix one case in detriment of
other. It appears some AMD Zen processors without advertising ERMS/FSRM
looks to be problematic regarding memory copy and rep movs usage
(and I guess probably they are not advertising/supporting it in some
systems as probably they know it's not behaving well with rep movs).

Lets revert the change for now, until we can find an way to fix the
original issue reported at RHEL-74389 without bringing another regression
with it.

Signed-off-by: Herton R. Krzesinski <herton@redhat.com>

Author: Herton R. Krzesinski

arch/x86/include/asm/uaccess_64.h

@@ -133,7 +133,7 @@ copy_user_generic(void *to, const void *from, unsigned long len)
 		"2:\n"
 		_ASM_EXTABLE_UA(1b, 2b)
 		:"+c" (len), "+D" (to), "+S" (from), ASM_CALL_CONSTRAINT
-		: : "memory", "rax");
+		: : "memory", "rax", "r8", "r9", "r10", "r11");
 	clac();
 	return len;
 }

arch/x86/lib/copy_user_64.S

@@ -27,7 +27,7 @@
  * NOTE! The calling convention is very intentionally the same as
  * for 'rep movs', so that we can rewrite the function call with
  * just a plain 'rep movs' on machines that have FSRM.  But to make
- * it simpler for us, we can clobber rsi/rdi and rax freely.
+ * it simpler for us, we can clobber rsi/rdi and rax/r8-r11 freely.
  */
 SYM_FUNC_START(rep_movs_alternative)
 	cmpq $64,%rcx
@@ -68,24 +68,55 @@ SYM_FUNC_START(rep_movs_alternative)
 	_ASM_EXTABLE_UA( 3b, .Lcopy_user_tail)
 
 .Llarge:
-0:	ALTERNATIVE "jmp .Llarge_movsq", "rep movsb", X86_FEATURE_ERMS
+0:	ALTERNATIVE "jmp .Lunrolled", "rep movsb", X86_FEATURE_ERMS
 1:	RET
 
-	_ASM_EXTABLE_UA( 0b, 1b)
+        _ASM_EXTABLE_UA( 0b, 1b)
 
-.Llarge_movsq:
-	movq %rcx,%rax
-	shrq $3,%rcx
-	andl $7,%eax
-0:	rep movsq
-	movl %eax,%ecx
+	.p2align 4
+.Lunrolled:
+10:	movq (%rsi),%r8
+11:	movq 8(%rsi),%r9
+12:	movq 16(%rsi),%r10
+13:	movq 24(%rsi),%r11
+14:	movq %r8,(%rdi)
+15:	movq %r9,8(%rdi)
+16:	movq %r10,16(%rdi)
+17:	movq %r11,24(%rdi)
+20:	movq 32(%rsi),%r8
+21:	movq 40(%rsi),%r9
+22:	movq 48(%rsi),%r10
+23:	movq 56(%rsi),%r11
+24:	movq %r8,32(%rdi)
+25:	movq %r9,40(%rdi)
+26:	movq %r10,48(%rdi)
+27:	movq %r11,56(%rdi)
+	addq $64,%rsi
+	addq $64,%rdi
+	subq $64,%rcx
+	cmpq $64,%rcx
+	jae .Lunrolled
+	cmpl $8,%ecx
+	jae .Lword
 	testl %ecx,%ecx
 	jne .Lcopy_user_tail
 	RET
 
-1:	leaq (%rax,%rcx,8),%rcx
-	jmp .Lcopy_user_tail
-
-	_ASM_EXTABLE_UA( 0b, 1b)
+	_ASM_EXTABLE_UA(10b, .Lcopy_user_tail)
+	_ASM_EXTABLE_UA(11b, .Lcopy_user_tail)
+	_ASM_EXTABLE_UA(12b, .Lcopy_user_tail)
+	_ASM_EXTABLE_UA(13b, .Lcopy_user_tail)
+	_ASM_EXTABLE_UA(14b, .Lcopy_user_tail)
+	_ASM_EXTABLE_UA(15b, .Lcopy_user_tail)
+	_ASM_EXTABLE_UA(16b, .Lcopy_user_tail)
+	_ASM_EXTABLE_UA(17b, .Lcopy_user_tail)
+	_ASM_EXTABLE_UA(20b, .Lcopy_user_tail)
+	_ASM_EXTABLE_UA(21b, .Lcopy_user_tail)
+	_ASM_EXTABLE_UA(22b, .Lcopy_user_tail)
+	_ASM_EXTABLE_UA(23b, .Lcopy_user_tail)
+	_ASM_EXTABLE_UA(24b, .Lcopy_user_tail)
+	_ASM_EXTABLE_UA(25b, .Lcopy_user_tail)
+	_ASM_EXTABLE_UA(26b, .Lcopy_user_tail)
+	_ASM_EXTABLE_UA(27b, .Lcopy_user_tail)
 SYM_FUNC_END(rep_movs_alternative)
 EXPORT_SYMBOL(rep_movs_alternative)