Donate to e Foundation | Murena handsets with /e/OS | Own a part of Murena! Learn more

Commit 9d8e2277 authored by Jan Beulich's avatar Jan Beulich Committed by Ingo Molnar
Browse files

x86-64: Handle byte-wise tail copying in memcpy() without a loop 



While hard to measure, reducing the number of possibly/likely
mis-predicted branches can generally be expected to be slightly
better.

Other than apparent at the first glance, this also doesn't grow
the function size (the alignment gap to the next function just
gets smaller).

Signed-off-by: default avatarJan Beulich <jbeulich@suse.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Link: http://lkml.kernel.org/r/4F218584020000780006F422@nat28.tlf.novell.com


Signed-off-by: default avatarIngo Molnar <mingo@elte.hu>
parent 2ab56091

arch/x86/lib/memcpy_64.S

+10 −9
Original line number Original line Diff line number Diff line
@@ -164,18 +164,19 @@ ENTRY(memcpy) 
	retq

	retq

	.p2align 4

	.p2align 4

.Lless_3bytes:

.Lless_3bytes:

	cmpl $0, %edx

	subl $1, %edx

	je .Lend

	jb .Lend

	/*

	/*

	 * Move data from 1 bytes to 3 bytes.

	 * Move data from 1 bytes to 3 bytes.

	 */

	 */

.Lloop_1:

	movzbl (%rsi), %ecx

	movb (%rsi), %r8b

	jz .Lstore_1byte

	movb %r8b, (%rdi)

	movzbq 1(%rsi), %r8

	incq %rdi

	movzbq (%rsi, %rdx), %r9

	incq %rsi

	movb %r8b, 1(%rdi)

	decl %edx

	movb %r9b, (%rdi, %rdx)

	jnz .Lloop_1

.Lstore_1byte:

	movb %cl, (%rdi)





.Lend:

.Lend:

	retq

	retq