/* SPDX-License-Identifier: GPL-2.0-only */
/*
 * Copyright (C) 2013 ARM Ltd.
 * Copyright (C) 2013 Linaro.
 *
 * This code is based on glibc cortex strings work originally authored by Linaro
 * be found @
 *
 * http://bazaar.launchpad.net/~linaro-toolchain-dev/cortex-strings/trunk/
 * files/head:/src/aarch64/
 */

#include <linux/linkage.h>
#include <asm/assembler.h>

/*
 * determine the length of a fixed-size string
 *
 * Parameters:
 *      x0 - const string pointer
 *      x1 - maximal string length
 * Returns:
 *      x0 - the return length of specific string
 */

/* Arguments and results.  */
srcin           .req    x0
len             .req    x0
limit           .req    x1

/* Locals and temporaries.  */
src             .req    x2
data1           .req    x3
data2           .req    x4
data2a          .req    x5
has_nul1        .req    x6
has_nul2        .req    x7
tmp1            .req    x8
tmp2            .req    x9
tmp3            .req    x10
tmp4            .req    x11
zeroones        .req    x12
pos             .req    x13
limit_wd        .req    x14

#define REP8_01 0x0101010101010101
#define REP8_7f 0x7f7f7f7f7f7f7f7f
#define REP8_80 0x8080808080808080

SYM_FUNC_START(__pi_strnlen)
        cbz     limit, .Lhit_limit
        mov     zeroones, #REP8_01
        bic     src, srcin, #15
        ands    tmp1, srcin, #15
        b.ne    .Lmisaligned
        /* Calculate the number of full and partial words -1.  */
        sub     limit_wd, limit, #1 /* Limit != 0, so no underflow.  */
        lsr     limit_wd, limit_wd, #4  /* Convert to Qwords.  */

        /*
        * NUL detection works on the principle that (X - 1) & (~X) & 0x80
        * (=> (X - 1) & ~(X | 0x7f)) is non-zero iff a byte is zero, and
        * can be done in parallel across the entire word.
        */
        /*
        * The inner loop deals with two Dwords at a time.  This has a
        * slightly higher start-up cost, but we should win quite quickly,
        * especially on cores with a high number of issue slots per
        * cycle, as we get much better parallelism out of the operations.
        */
.Lloop:
        ldp     data1, data2, [src], #16
.Lrealigned:
        sub     tmp1, data1, zeroones
        orr     tmp2, data1, #REP8_7f
        sub     tmp3, data2, zeroones
        orr     tmp4, data2, #REP8_7f
        bic     has_nul1, tmp1, tmp2
        bic     has_nul2, tmp3, tmp4
        subs    limit_wd, limit_wd, #1
        orr     tmp1, has_nul1, has_nul2
        ccmp    tmp1, #0, #0, pl    /* NZCV = 0000  */
        b.eq    .Lloop

        cbz     tmp1, .Lhit_limit   /* No null in final Qword.  */

        /*
        * We know there's a null in the final Qword. The easiest thing
        * to do now is work out the length of the string and return
        * MIN (len, limit).
        */
        sub     len, src, srcin
        cbz     has_nul1, .Lnul_in_data2
CPU_BE( mov     data2, data1 )  /*perpare data to re-calculate the syndrome*/

        sub     len, len, #8
        mov     has_nul2, has_nul1
.Lnul_in_data2:
        /*
        * For big-endian, carry propagation (if the final byte in the
        * string is 0x01) means we cannot use has_nul directly.  The
        * easiest way to get the correct byte is to byte-swap the data
        * and calculate the syndrome a second time.
        */
CPU_BE( rev     data2, data2 )
CPU_BE( sub     tmp1, data2, zeroones )
CPU_BE( orr     tmp2, data2, #REP8_7f )
CPU_BE( bic     has_nul2, tmp1, tmp2 )

        sub     len, len, #8
        rev     has_nul2, has_nul2
        clz     pos, has_nul2
        add     len, len, pos, lsr #3       /* Bits to bytes.  */
        cmp     len, limit
        csel    len, len, limit, ls     /* Return the lower value.  */
        ret

.Lmisaligned:
        /*
        * Deal with a partial first word.
        * We're doing two things in parallel here;
        * 1) Calculate the number of words (but avoiding overflow if
        * limit is near ULONG_MAX) - to do this we need to work out
        * limit + tmp1 - 1 as a 65-bit value before shifting it;
        * 2) Load and mask the initial data words - we force the bytes
        * before the ones we are interested in to 0xff - this ensures
        * early bytes will not hit any zero detection.
        */
        ldp     data1, data2, [src], #16

        sub     limit_wd, limit, #1
        and     tmp3, limit_wd, #15
        lsr     limit_wd, limit_wd, #4

        add     tmp3, tmp3, tmp1
        add     limit_wd, limit_wd, tmp3, lsr #4

        neg     tmp4, tmp1
        lsl     tmp4, tmp4, #3  /* Bytes beyond alignment -> bits.  */

        mov     tmp2, #~0
        /* Big-endian.  Early bytes are at MSB.  */
CPU_BE( lsl     tmp2, tmp2, tmp4 )      /* Shift (tmp1 & 63).  */
        /* Little-endian.  Early bytes are at LSB.  */
CPU_LE( lsr     tmp2, tmp2, tmp4 )      /* Shift (tmp1 & 63).  */

        cmp     tmp1, #8

        orr     data1, data1, tmp2
        orr     data2a, data2, tmp2

        csinv   data1, data1, xzr, le
        csel    data2, data2, data2a, le
        b       .Lrealigned

.Lhit_limit:
        mov     len, limit
        ret
SYM_FUNC_END(__pi_strnlen)

SYM_FUNC_ALIAS_WEAK(strnlen, __pi_strnlen)
EXPORT_SYMBOL_NOKASAN(strnlen)