From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from mail-yw1-x112e.google.com (mail-yw1-x112e.google.com [IPv6:2607:f8b0:4864:20::112e]) by sourceware.org (Postfix) with ESMTPS id 7355D385842B for ; Wed, 1 Feb 2023 19:50:57 +0000 (GMT) DMARC-Filter: OpenDMARC Filter v1.4.2 sourceware.org 7355D385842B Authentication-Results: sourceware.org; dmarc=pass (p=none dis=none) header.from=gmail.com Authentication-Results: sourceware.org; spf=pass smtp.mailfrom=gmail.com Received: by mail-yw1-x112e.google.com with SMTP id 00721157ae682-50aa54cc7c0so233245587b3.8 for ; Wed, 01 Feb 2023 11:50:57 -0800 (PST) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=20210112; h=cc:to:subject:message-id:date:from:in-reply-to:references :mime-version:from:to:cc:subject:date:message-id:reply-to; bh=r5o744Ju3WTWJPKo1z9ssmx34Gcy4iicDu80SxRha8U=; b=S0MuVmhn/V2m8nuK/KvDmcG8dkGtumSvmmH54a6Gl7EezhZ97l6EEs+v2Wt/OdVObq RU0d6a9eqwi+/Q7ZTbYIuagW73/B+YFqqevjHff9DljygwuBS602hku+2ObA0Sz+huyP AdQH8dy6nGk7ObYv4saycctomnCN6EHpmf0sBmnQkOhNKaSkrbTtZIJS8YMnOa9SlSzk fyiJoRgRgQSv23WnnkQZ4uocGFVdElVY+GQZrwTryDuy+/FdT0g2fPZdAxPj3GK/QeJS qsG9cH+aIRzLniK+D9IMK3KFlqOQ+vaREAz1LcpiLo0aeIK9ecNZM1G74g05jMBvNWO+ ohhQ== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20210112; h=cc:to:subject:message-id:date:from:in-reply-to:references :mime-version:x-gm-message-state:from:to:cc:subject:date:message-id :reply-to; bh=r5o744Ju3WTWJPKo1z9ssmx34Gcy4iicDu80SxRha8U=; b=Po769GsvkPNTBLGoiNsFARrI442c2s7cPy9SvglHswksWFg7J2jxkNOGwEvxFFPene lhgXqOtgT1iqoZBqKCTbSo7xr8lKjy5Abe+fcUrDcRu7PMSafjM0i4DWDU3qM6/4fadR 0TNLmnBpvr/a3ccNis/+C1bYBCXOQkwTLv2BRFFQzf87zU7IayJoTMaaQJrh3xyA1AAa 8y/xpARfs4++9eUPmWWAlDmHYkYA2TlmTIrkaXzrTSrMmLFOALg0eyeN7ZdOv/BDS95P OaeDauiFMIkaBoI1ZeDmnUOOStPKI/3O2RXs0/XtAEWuOQHWov5apF7z69vTbExeECQX 9dFw== X-Gm-Message-State: AO0yUKVV1w64QxzmX3eac9dCXGmp4PuMq7IQ6Q0MC3gena7iQkGdOsIy bVY0CPVn1+iUZ/RuE4VmW4drj8nbmVdon3eNf+0= X-Google-Smtp-Source: AK7set/B5i6yvcX8bGQG8HeccMDT6lGq2XFk0DdTPdj2GebeFjAgZlFGQRqdK4n8puBAq1xgLPJ0//Pdr8lToBmFbz8= X-Received: by 2002:a81:46d4:0:b0:506:3d29:a280 with SMTP id t203-20020a8146d4000000b005063d29a280mr405323ywa.421.1675281056659; Wed, 01 Feb 2023 11:50:56 -0800 (PST) MIME-Version: 1.0 References: <20230201170406.303978-1-adhemerval.zanella@linaro.org> <20230201170406.303978-14-adhemerval.zanella@linaro.org> In-Reply-To: <20230201170406.303978-14-adhemerval.zanella@linaro.org> From: Noah Goldstein Date: Wed, 1 Feb 2023 13:50:44 -0600 Message-ID: Subject: Re: [PATCH v11 13/29] string: Improve generic memrchr To: Adhemerval Zanella Cc: libc-alpha@sourceware.org, Richard Henderson , Jeff Law , Xi Ruoyao Content-Type: text/plain; charset="UTF-8" X-Spam-Status: No, score=-8.8 required=5.0 tests=BAYES_00,DKIM_SIGNED,DKIM_VALID,DKIM_VALID_AU,DKIM_VALID_EF,FREEMAIL_FROM,GIT_PATCH_0,KAM_SHORT,RCVD_IN_DNSWL_NONE,SPF_HELO_NONE,SPF_PASS,TXREP autolearn=ham autolearn_force=no version=3.4.6 X-Spam-Checker-Version: SpamAssassin 3.4.6 (2021-04-09) on server2.sourceware.org List-Id: On Wed, Feb 1, 2023 at 11:04 AM Adhemerval Zanella wrote: > > New algorithm read the lastaligned address and mask off the unwanted > bytes. The loop now read word-aligned address and check using the > has_eq macro. > > Checked on x86_64-linux-gnu, i686-linux-gnu, powerpc-linux-gnu, > and powerpc64-linux-gnu by removing the arch-specific assembly > implementation and disabling multi-arch (it covers both LE and BE > for 64 and 32 bits). > > Co-authored-by: Richard Henderson > --- > string/memrchr.c | 196 ++++++++++------------------------------------- > 1 file changed, 39 insertions(+), 157 deletions(-) > > diff --git a/string/memrchr.c b/string/memrchr.c > index 18b20ff76a..b37f2a68c8 100644 > --- a/string/memrchr.c > +++ b/string/memrchr.c > @@ -1,11 +1,6 @@ > /* memrchr -- find the last occurrence of a byte in a memory block > Copyright (C) 1991-2023 Free Software Foundation, Inc. > This file is part of the GNU C Library. > - Based on strlen implementation by Torbjorn Granlund (tege@sics.se), > - with help from Dan Sahlin (dan@sics.se) and > - commentary by Jim Blandy (jimb@ai.mit.edu); > - adaptation to memchr suggested by Dick Karpinski (dick@cca.ucsf.edu), > - and implemented by Roland McGrath (roland@ai.mit.edu). > > The GNU C Library is free software; you can redistribute it and/or > modify it under the terms of the GNU Lesser General Public > @@ -21,177 +16,64 @@ > License along with the GNU C Library; if not, see > . */ > > -#include > - > -#ifdef HAVE_CONFIG_H > -# include > -#endif > - > -#if defined _LIBC > -# include > -# include > -#endif > - > -#if defined HAVE_LIMITS_H || defined _LIBC > -# include > -#endif > - > -#define LONG_MAX_32_BITS 2147483647 > - > -#ifndef LONG_MAX > -# define LONG_MAX LONG_MAX_32_BITS > -#endif > - > -#include > +#include > +#include > +#include > +#include > +#include > +#include > > #undef __memrchr > #undef memrchr > > -#ifndef weak_alias > -# define __memrchr memrchr > +#ifdef MEMRCHR > +# define __memrchr MEMRCHR > #endif > > -/* Search no more than N bytes of S for C. */ > void * > -#ifndef MEMRCHR > -__memrchr > -#else > -MEMRCHR > -#endif > - (const void *s, int c_in, size_t n) > +__memrchr (const void *s, int c_in, size_t n) > { > - const unsigned char *char_ptr; > - const unsigned long int *longword_ptr; > - unsigned long int longword, magic_bits, charmask; > - unsigned char c; > - > - c = (unsigned char) c_in; > - > - /* Handle the last few characters by reading one character at a time. > - Do this until CHAR_PTR is aligned on a longword boundary. */ > - for (char_ptr = (const unsigned char *) s + n; > - n > 0 && ((unsigned long int) char_ptr > - & (sizeof (longword) - 1)) != 0; > - --n) > - if (*--char_ptr == c) > - return (void *) char_ptr; > - > - /* All these elucidatory comments refer to 4-byte longwords, > - but the theory applies equally well to 8-byte longwords. */ > - > - longword_ptr = (const unsigned long int *) char_ptr; > - > - /* Bits 31, 24, 16, and 8 of this number are zero. Call these bits > - the "holes." Note that there is a hole just to the left of > - each byte, with an extra at the end: > + if (__glibc_unlikely (n == 0)) > + return NULL; > > - bits: 01111110 11111110 11111110 11111111 > - bytes: AAAAAAAA BBBBBBBB CCCCCCCC DDDDDDDD > + const op_t *word_ptr = (const op_t *) PTR_ALIGN_UP (s + n, sizeof (op_t)); > + uintptr_t s_int = (uintptr_t) s + n; > > - The 1-bits make sure that carries propagate to the next 0-bit. > - The 0-bits provide holes for carries to fall into. */ > - magic_bits = -1; > - magic_bits = magic_bits / 0xff * 0xfe << 1 >> 1 | 1; > + op_t word = *--word_ptr; > + op_t repeated_c = repeat_bytes (c_in); > > - /* Set up a longword, each of whose bytes is C. */ > - charmask = c | (c << 8); > - charmask |= charmask << 16; > -#if LONG_MAX > LONG_MAX_32_BITS > - charmask |= charmask << 32; > -#endif > + /* Compute the address of the word containing the initial byte. */ > + const op_t *sword = (const op_t *) PTR_ALIGN_DOWN (s, sizeof (op_t)); > > - /* Instead of the traditional loop which tests each character, > - we will test a longword at a time. The tricky part is testing > - if *any of the four* bytes in the longword in question are zero. */ > - while (n >= sizeof (longword)) > + /* If the end of buffer is not op_t aligned, mask off the undesirable bits > + before find the last byte position. */ > + find_t mask = shift_find_last (find_eq_all (word, repeated_c), s_int); > + if (mask != 0) > { > - /* We tentatively exit the loop if adding MAGIC_BITS to > - LONGWORD fails to change any of the hole bits of LONGWORD. > - > - 1) Is this safe? Will it catch all the zero bytes? > - Suppose there is a byte with all zeros. Any carry bits > - propagating from its left will fall into the hole at its > - least significant bit and stop. Since there will be no > - carry from its most significant bit, the LSB of the > - byte to the left will be unchanged, and the zero will be > - detected. > - > - 2) Is this worthwhile? Will it ignore everything except > - zero bytes? Suppose every byte of LONGWORD has a bit set > - somewhere. There will be a carry into bit 8. If bit 8 > - is set, this will carry into bit 16. If bit 8 is clear, > - one of bits 9-15 must be set, so there will be a carry > - into bit 16. Similarly, there will be a carry into bit > - 24. If one of bits 24-30 is set, there will be a carry > - into bit 31, so all of the hole bits will be changed. > - > - The one misfire occurs when bits 24-30 are clear and bit > - 31 is set; in this case, the hole at bit 31 is not > - changed. If we had access to the processor carry flag, > - we could close this loophole by putting the fourth hole > - at bit 32! > - > - So it ignores everything except 128's, when they're aligned > - properly. > - > - 3) But wait! Aren't we looking for C, not zero? > - Good point. So what we do is XOR LONGWORD with a longword, > - each of whose bytes is C. This turns each byte that is C > - into a zero. */ > - > - longword = *--longword_ptr ^ charmask; > - > - /* Add MAGIC_BITS to LONGWORD. */ > - if ((((longword + magic_bits) > - > - /* Set those bits that were unchanged by the addition. */ > - ^ ~longword) > - > - /* Look at only the hole bits. If any of the hole bits > - are unchanged, most likely one of the bytes was a > - zero. */ > - & ~magic_bits) != 0) > - { > - /* Which of the bytes was C? If none of them were, it was > - a misfire; continue the search. */ > - > - const unsigned char *cp = (const unsigned char *) longword_ptr; > - > -#if LONG_MAX > 2147483647 > - if (cp[7] == c) > - return (void *) &cp[7]; > - if (cp[6] == c) > - return (void *) &cp[6]; > - if (cp[5] == c) > - return (void *) &cp[5]; > - if (cp[4] == c) > - return (void *) &cp[4]; > -#endif > - if (cp[3] == c) > - return (void *) &cp[3]; > - if (cp[2] == c) > - return (void *) &cp[2]; > - if (cp[1] == c) > - return (void *) &cp[1]; > - if (cp[0] == c) > - return (void *) cp; > - } > - > - n -= sizeof (longword); > + char *ret = (char *) word_ptr + index_last (mask); > + return ret >= (char *) s ? ret : NULL; > } > + if (word_ptr == sword) > + return NULL; > + word = *--word_ptr; > > - char_ptr = (const unsigned char *) longword_ptr; > - > - while (n-- > 0) > + while (word_ptr != sword) > { > - if (*--char_ptr == c) > - return (void *) char_ptr; > + if (has_eq (word, repeated_c)) > + return (char *) word_ptr + index_last_eq (word, repeated_c); > + word = *--word_ptr; > } > > - return 0; > + if (has_eq (word, repeated_c)) > + { > + /* We found a match, but it might be in a byte past the end of the > + array. */ > + char *ret = (char *) word_ptr + index_last_eq (word, repeated_c); > + if (ret >= (char *) s) > + return ret; > + } > + return NULL; > } > #ifndef MEMRCHR > -# ifdef weak_alias > weak_alias (__memrchr, memrchr) > -# endif > #endif > -- > 2.34.1 > LGTM. Reviewed-by: Noah Goldstein