From: H.J. Lu Date: Tue, 12 Jan 2021 13:15:49 +0000 (-0800) Subject: x86-64: Avoid rep movsb with short distance [BZ #27130] X-Git-Url: http://git.ipfire.org/cgi-bin/gitweb.cgi?a=commitdiff_plain;h=369b47aaf7464d2007448a389647bbb2e255e98b;p=thirdparty%2Fglibc.git x86-64: Avoid rep movsb with short distance [BZ #27130] When copying with "rep movsb", if the distance between source and destination is N*4GB + [1..63] with N >= 0, performance may be very slow. This patch updates memmove-vec-unaligned-erms.S for AVX and AVX512 versions with the distance in RCX: cmpl $63, %ecx // Don't use "rep movsb" if ECX <= 63 jbe L(Don't use rep movsb") Use "rep movsb" Benchtests data with bench-memcpy, bench-memcpy-large, bench-memcpy-random and bench-memcpy-walk on Skylake, Ice Lake and Tiger Lake show that its performance impact is within noise range as "rep movsb" is only used for data size >= 4KB. (cherry picked from commit 3ec5d83d2a237d39e7fd6ef7a0bc8ac4c171a4a5) --- diff --git a/NEWS b/NEWS index 5712eecde81..105fafb44c1 100644 --- a/NEWS +++ b/NEWS @@ -57,6 +57,7 @@ The following bugs are resolved with this release: [25933] Off by one error in __strncmp_avx2 [25966] Incorrect access of __x86_shared_non_temporal_threshold for x32 [25976] nss_compat: internal_end*ent may clobber errno, hiding ERANGE + [27130] "rep movsb" performance issue Version 2.30 diff --git a/sysdeps/x86_64/multiarch/memmove-vec-unaligned-erms.S b/sysdeps/x86_64/multiarch/memmove-vec-unaligned-erms.S index 29cab02f794..95e9bb0e5d3 100644 --- a/sysdeps/x86_64/multiarch/memmove-vec-unaligned-erms.S +++ b/sysdeps/x86_64/multiarch/memmove-vec-unaligned-erms.S @@ -67,6 +67,13 @@ # define REP_MOVSB_THRESHOLD (2048 * (VEC_SIZE / 16)) #endif +/* Avoid short distance rep movsb only with non-SSE vector. */ +#ifndef AVOID_SHORT_DISTANCE_REP_MOVSB +# define AVOID_SHORT_DISTANCE_REP_MOVSB (VEC_SIZE > 16) +#else +# define AVOID_SHORT_DISTANCE_REP_MOVSB 0 +#endif + #ifndef PREFETCH # define PREFETCH(addr) prefetcht0 addr #endif @@ -257,7 +264,21 @@ L(movsb): # error Unsupported REP_MOVSB_THRESHOLD and VEC_SIZE! # endif jb L(more_8x_vec_backward) +# if AVOID_SHORT_DISTANCE_REP_MOVSB + movq %rdi, %rcx + subq %rsi, %rcx + jmp 2f +# endif 1: +# if AVOID_SHORT_DISTANCE_REP_MOVSB + movq %rsi, %rcx + subq %rdi, %rcx +2: +/* Avoid "rep movsb" if RCX, the distance between source and destination, + is N*4GB + [1..63] with N >= 0. */ + cmpl $63, %ecx + jbe L(more_2x_vec) /* Avoid "rep movsb" if ECX <= 63. */ +# endif mov %RDX_LP, %RCX_LP rep movsb L(nop):