diff options
| author | Alex Richardson <arichardson@FreeBSD.org> | 2021-02-22 11:59:12 +0000 |
|---|---|---|
| committer | Alex Richardson <arichardson@FreeBSD.org> | 2021-02-22 11:59:12 +0000 |
| commit | f9f37c002ab5a580accfe26b731eef45e798b435 (patch) | |
| tree | 6c5f3a7746c1d8a0d7a04012d47b6428c7b3057e /string/aarch64/strlen-mte.S | |
Import Arm Optimized Routines v21.02vendor/arm-optimized-routines/v21.02
We already have copies of some of these files in the repository without a
vendor import and we should really be using contrib/ instead.
We should also be able to use some of the math functions to allow the
tests to pass on AArch64 (and other architectures) instead of just x86.
We should also be able to reuse some of the tests for the kyua testsuite.
Imported using
```
curl -L https://github.com/ARM-software/optimized-routines/tarball/e823e3abf5f89ecba58a10fc0fd82c13d9984b6b | tar --strip-components=1 -xvzf -
git add .
```
Diffstat (limited to 'string/aarch64/strlen-mte.S')
| -rw-r--r-- | string/aarch64/strlen-mte.S | 80 |
1 files changed, 80 insertions, 0 deletions
diff --git a/string/aarch64/strlen-mte.S b/string/aarch64/strlen-mte.S new file mode 100644 index 000000000000..7cf41d5c1eac --- /dev/null +++ b/string/aarch64/strlen-mte.S @@ -0,0 +1,80 @@ +/* + * strlen - calculate the length of a string. + * + * Copyright (c) 2020, Arm Limited. + * SPDX-License-Identifier: MIT + */ + +/* Assumptions: + * + * ARMv8-a, AArch64, Advanced SIMD. + * MTE compatible. + */ + +#include "../asmdefs.h" + +#define srcin x0 +#define result x0 + +#define src x1 +#define synd x2 +#define tmp x3 +#define wtmp w3 +#define shift x4 + +#define data q0 +#define vdata v0 +#define vhas_nul v1 +#define vrepmask v2 +#define vend v3 +#define dend d3 + +/* Core algorithm: + + For each 16-byte chunk we calculate a 64-bit syndrome value with four bits + per byte. For even bytes, bits 0-3 are set if the relevant byte matched the + requested character or the byte is NUL. Bits 4-7 must be zero. Bits 4-7 are + set likewise for odd bytes so that adjacent bytes can be merged. Since the + bits in the syndrome reflect the order in which things occur in the original + string, counting trailing zeros identifies exactly which byte matched. */ + +ENTRY (__strlen_aarch64_mte) + PTR_ARG (0) + bic src, srcin, 15 + mov wtmp, 0xf00f + ld1 {vdata.16b}, [src] + dup vrepmask.8h, wtmp + cmeq vhas_nul.16b, vdata.16b, 0 + lsl shift, srcin, 2 + and vhas_nul.16b, vhas_nul.16b, vrepmask.16b + addp vend.16b, vhas_nul.16b, vhas_nul.16b /* 128->64 */ + fmov synd, dend + lsr synd, synd, shift + cbz synd, L(loop) + + rbit synd, synd + clz result, synd + lsr result, result, 2 + ret + + .p2align 5 +L(loop): + ldr data, [src, 16]! + cmeq vhas_nul.16b, vdata.16b, 0 + umaxp vend.16b, vhas_nul.16b, vhas_nul.16b + fmov synd, dend + cbz synd, L(loop) + + and vhas_nul.16b, vhas_nul.16b, vrepmask.16b + addp vend.16b, vhas_nul.16b, vhas_nul.16b /* 128->64 */ + sub result, src, srcin + fmov synd, dend +#ifndef __AARCH64EB__ + rbit synd, synd +#endif + clz tmp, synd + add result, result, tmp, lsr 2 + ret + +END (__strlen_aarch64_mte) + |