Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::1:20 as permitted sender) client-ip=2620:137:e000::1:20;
Date:   Thu, 5 Jan 2023 14:05:09 +0100
From:   =?UTF-8?B?Q2zDqW1lbnQgTMOpZ2Vy?= <clement.leger@bootlin.com>
To:     Yann Sionneau <ysionneau@kalray.eu>
Cc:     linux-kernel@vger.kernel.org, Jules Maselbas <jmaselbas@kalray.eu>,
        Julian Vetter <jvetter@kalray.eu>,
        Marius Gligor <mgligor@kalray.eu>
Subject: Re: [RFC PATCH 16/25] kvx: Add some library functions
Message-ID: <20230105140509.0049ab9a@fixe.home>
In-Reply-To: <20230103164359.24347-17-ysionneau@kalray.eu>
References: <20230103164359.24347-1-ysionneau@kalray.eu>
        <20230103164359.24347-17-ysionneau@kalray.eu>
Organization: Bootlin
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: quoted-printable
Precedence: bulk

Le Tue,  3 Jan 2023 17:43:50 +0100,
Yann Sionneau <ysionneau@kalray.eu> a =C3=A9crit :

> Add some library functions for kvx, including: delay, memset,
> memcpy, strlen, clear_page, copy_page, raw_copy_from/to_user,
> asm_clear_user.
>=20
> CC: linux-kernel@vger.kernel.org
> Co-developed-by: Clement Leger <clement.leger@bootlin.com>
> Signed-off-by: Clement Leger <clement.leger@bootlin.com>
> Co-developed-by: Jules Maselbas <jmaselbas@kalray.eu>
> Signed-off-by: Jules Maselbas <jmaselbas@kalray.eu>
> Co-developed-by: Julian Vetter <jvetter@kalray.eu>
> Signed-off-by: Julian Vetter <jvetter@kalray.eu>
> Co-developed-by: Marius Gligor <mgligor@kalray.eu>
> Signed-off-by: Marius Gligor <mgligor@kalray.eu>
> Co-developed-by: Yann Sionneau <ysionneau@kalray.eu>
> Signed-off-by: Yann Sionneau <ysionneau@kalray.eu>
> ---
>  arch/kvx/include/asm/string.h |  20 ++
>  arch/kvx/kernel/kvx_ksyms.c   |   5 +
>  arch/kvx/lib/clear_page.S     |  40 ++++
>  arch/kvx/lib/copy_page.S      |  90 +++++++++
>  arch/kvx/lib/delay.c          |  39 ++++
>  arch/kvx/lib/memcpy.c         |  70 +++++++
>  arch/kvx/lib/memset.S         | 351 ++++++++++++++++++++++++++++++++++
>  arch/kvx/lib/strlen.S         | 122 ++++++++++++
>  arch/kvx/lib/usercopy.S       |  90 +++++++++
>  9 files changed, 827 insertions(+)
>  create mode 100644 arch/kvx/include/asm/string.h
>  create mode 100644 arch/kvx/lib/clear_page.S
>  create mode 100644 arch/kvx/lib/copy_page.S
>  create mode 100644 arch/kvx/lib/delay.c
>  create mode 100644 arch/kvx/lib/memcpy.c
>  create mode 100644 arch/kvx/lib/memset.S
>  create mode 100644 arch/kvx/lib/strlen.S
>  create mode 100644 arch/kvx/lib/usercopy.S
>=20
> diff --git a/arch/kvx/include/asm/string.h b/arch/kvx/include/asm/string.h
> new file mode 100644
> index 000000000000..677c1393a5cd
> --- /dev/null
> +++ b/arch/kvx/include/asm/string.h
> @@ -0,0 +1,20 @@
> +/* SPDX-License-Identifier: GPL-2.0-only */
> +/*
> + * Copyright (C) 2017-2023 Kalray Inc.
> + * Author(s): Clement Leger
> + *            Jules Maselbas
> + */
> +
> +#ifndef _ASM_KVX_STRING_H
> +#define _ASM_KVX_STRING_H
> +
> +#define __HAVE_ARCH_MEMSET
> +extern void *memset(void *s, int c, size_t n);
> +
> +#define __HAVE_ARCH_MEMCPY
> +extern void *memcpy(void *dest, const void *src, size_t n);
> +
> +#define __HAVE_ARCH_STRLEN
> +extern size_t strlen(const char *s);
> +
> +#endif	/* _ASM_KVX_STRING_H */
> diff --git a/arch/kvx/kernel/kvx_ksyms.c b/arch/kvx/kernel/kvx_ksyms.c
> index 18990aaf259f..678f81716dea 100644
> --- a/arch/kvx/kernel/kvx_ksyms.c
> +++ b/arch/kvx/kernel/kvx_ksyms.c
> @@ -22,3 +22,8 @@ DECLARE_EXPORT(__umoddi3);
>  DECLARE_EXPORT(__divdi3);
>  DECLARE_EXPORT(__udivdi3);
>  DECLARE_EXPORT(__multi3);
> +
> +DECLARE_EXPORT(clear_page);
> +DECLARE_EXPORT(copy_page);
> +DECLARE_EXPORT(memset);
> +DECLARE_EXPORT(asm_clear_user);
> diff --git a/arch/kvx/lib/clear_page.S b/arch/kvx/lib/clear_page.S
> new file mode 100644
> index 000000000000..364fe0663ca2
> --- /dev/null
> +++ b/arch/kvx/lib/clear_page.S
> @@ -0,0 +1,40 @@
> +/* SPDX-License-Identifier: GPL-2.0-only */
> +/*
> + * Copyright (C) 2017-2023 Kalray Inc.
> + * Author(s): Marius Gligor
> + *            Clement Leger
> + */
> +
> +#include <linux/linkage.h>
> +#include <linux/export.h>
> +#include <linux/const.h>
> +
> +#include <asm/cache.h>
> +#include <asm/page.h>
> +
> +#define CLEAR_PAGE_LOOP_COUNT	(PAGE_SIZE / 32)
> +
> +/*
> + * Clear page @dest.
> + *
> + * Parameters:
> + *	r0 - dest page
> + */
> +ENTRY(clear_page)
> +	make $r1 =3D CLEAR_PAGE_LOOP_COUNT
> +	;;
> +	make $r4 =3D 0
> +	make $r5 =3D 0
> +	make $r6 =3D 0
> +	make $r7 =3D 0
> +	;;
> +
> +	loopdo $r1, clear_page_done
> +		;;
> +		so 0[$r0] =3D $r4r5r6r7
> +		addd $r0 =3D $r0, 32
> +		;;
> +	clear_page_done:
> +	ret
> +	;;
> +ENDPROC(clear_page)
> diff --git a/arch/kvx/lib/copy_page.S b/arch/kvx/lib/copy_page.S
> new file mode 100644
> index 000000000000..4bb82d1c964c
> --- /dev/null
> +++ b/arch/kvx/lib/copy_page.S
> @@ -0,0 +1,90 @@
> +/* SPDX-License-Identifier: GPL-2.0-only */
> +/*
> + * Copyright (C) 2017-2023 Kalray Inc.
> + * Author(s): Clement Leger
> + */
> +
> +#include <linux/linkage.h>
> +#include <linux/const.h>
> +
> +#include <asm/page.h>
> +
> +/* We have 8 load/store octuple (32 bytes) per hardware loop */
> +#define COPY_SIZE_PER_LOOP	(32 * 8)
> +#define COPY_PAGE_LOOP_COUNT	(PAGE_SIZE / COPY_SIZE_PER_LOOP)
> +
> +/*
> + * Copy a page from src to dest (both are page aligned)
> + * In order to recover from smem latency, unroll the loop to trigger mul=
tiple
> + * onfly loads and avoid waiting too much for them to return.
> + * We use 8 * 32 load even though we could use more (up to 10 loads) to =
simplify
> + * the handling using a single hardware loop
> + *
> + * Parameters:
> + *	r0 - dest
> + *	r1 - src
> + */
> +ENTRY(copy_page)
> +	make $r2 =3D COPY_PAGE_LOOP_COUNT
> +	make $r3 =3D 0
> +	;;
> +	loopdo $r2, copy_page_done
> +		;;
> +		/*
> +		 * Load 8 * 32 bytes using uncached access to avoid hitting
> +		 * the cache
> +		 */
> +		lo.xs $r32r33r34r35 =3D $r3[$r1]
> +		/* Copy current copy index for store */
> +		copyd $r2 =3D $r3
> +		addd $r3 =3D $r3, 1
> +		;;
> +		lo.xs $r36r37r38r39 =3D $r3[$r1]
> +		addd $r3 =3D $r3, 1
> +		;;
> +		lo.xs $r40r41r42r43 =3D $r3[$r1]
> +		addd $r3 =3D $r3, 1
> +		;;
> +		lo.xs $r44r45r46r47 =3D $r3[$r1]
> +		addd $r3 =3D $r3, 1
> +		;;
> +		lo.xs $r48r49r50r51 =3D $r3[$r1]
> +		addd $r3 =3D $r3, 1
> +		;;
> +		lo.xs $r52r53r54r55 =3D $r3[$r1]
> +		addd $r3 =3D $r3, 1
> +		;;
> +		lo.xs $r56r57r58r59 =3D $r3[$r1]
> +		addd $r3 =3D $r3, 1
> +		;;
> +		lo.xs $r60r61r62r63 =3D $r3[$r1]
> +		addd $r3 =3D $r3, 1
> +		;;
> +		/* And then store all of them */
> +		so.xs $r2[$r0] =3D $r32r33r34r35
> +		addd $r2 =3D $r2, 1
> +		;;
> +		so.xs $r2[$r0] =3D $r36r37r38r39
> +		addd $r2 =3D $r2, 1
> +		;;
> +		so.xs $r2[$r0] =3D $r40r41r42r43
> +		addd $r2 =3D $r2, 1
> +		;;
> +		so.xs $r2[$r0] =3D $r44r45r46r47
> +		addd $r2 =3D $r2, 1
> +		;;
> +		so.xs $r2[$r0] =3D $r48r49r50r51
> +		addd $r2 =3D $r2, 1
> +		;;
> +		so.xs $r2[$r0] =3D $r52r53r54r55
> +		addd $r2 =3D $r2, 1
> +		;;
> +		so.xs $r2[$r0] =3D $r56r57r58r59
> +		addd $r2 =3D $r2, 1
> +		;;
> +		so.xs $r2[$r0] =3D $r60r61r62r63
> +		;;
> +	copy_page_done:
> +	ret
> +	;;
> +ENDPROC(copy_page)
> diff --git a/arch/kvx/lib/delay.c b/arch/kvx/lib/delay.c
> new file mode 100644
> index 000000000000..11295eedc3f5
> --- /dev/null
> +++ b/arch/kvx/lib/delay.c
> @@ -0,0 +1,39 @@
> +// SPDX-License-Identifier: GPL-2.0-only
> +/*
> + * Copyright (C) 2017-2023 Kalray Inc.
> + * Author(s): Clement Leger
> + */
> +
> +#include <linux/export.h>
> +#include <linux/delay.h>
> +
> +#include <asm/param.h>
> +#include <asm/timex.h>
> +
> +void __delay(unsigned long loops)
> +{
> +	cycles_t target_cycle =3D get_cycles() + loops;
> +
> +	while (get_cycles() < target_cycle);
> +}
> +EXPORT_SYMBOL(__delay);
> +
> +inline void __const_udelay(unsigned long xloops)
> +{
> +	u64 loops =3D (u64)xloops * (u64)loops_per_jiffy * HZ;
> +
> +	__delay(loops >> 32);
> +}
> +EXPORT_SYMBOL(__const_udelay);
> +
> +void __udelay(unsigned long usecs)
> +{
> +	__const_udelay(usecs * 0x10C7UL); /* 2**32 / 1000000 (rounded up) */
> +}
> +EXPORT_SYMBOL(__udelay);

This could probably be factorized with  arm64, csky, nios2, etc.

> +
> +void __ndelay(unsigned long nsecs)
> +{
> +	__const_udelay(nsecs * 0x5UL); /* 2**32 / 1000000000 (rounded up) */
> +}
> +EXPORT_SYMBOL(__ndelay);

Same goes for __ndelay I think

> diff --git a/arch/kvx/lib/memcpy.c b/arch/kvx/lib/memcpy.c
> new file mode 100644
> index 000000000000..b81f746a80ee
> --- /dev/null
> +++ b/arch/kvx/lib/memcpy.c
> @@ -0,0 +1,70 @@
> +// SPDX-License-Identifier: GPL-2.0-only
> +/*
> + * Copyright (C) 2017-2023 Kalray Inc.
> + * Author(s): Clement Leger
> + *            Yann Sionneau
> + */
> +
> +#include <linux/export.h>
> +#include <linux/types.h>
> +
> +void *memcpy(void *dest, const void *src, size_t n)
> +{
> +	__uint128_t *tmp128_d =3D dest;
> +	const __uint128_t *tmp128_s =3D src;
> +	uint64_t *tmp64_d;
> +	const uint64_t *tmp64_s;
> +	uint32_t *tmp32_d;
> +	const uint32_t *tmp32_s;
> +	uint16_t *tmp16_d;
> +	const uint16_t *tmp16_s;
> +	uint8_t *tmp8_d;
> +	const uint8_t *tmp8_s;
> +
> +	while (n >=3D 16) {
> +		*tmp128_d =3D *tmp128_s;
> +		tmp128_d++;
> +		tmp128_s++;
> +		n -=3D 16;
> +	}
> +
> +	tmp64_d =3D (uint64_t *) tmp128_d;
> +	tmp64_s =3D (uint64_t *) tmp128_s;
> +	while (n >=3D 8) {
> +		*tmp64_d =3D *tmp64_s;
> +		tmp64_d++;
> +		tmp64_s++;
> +		n -=3D 8;
> +	}
> +
> +	tmp32_d =3D (uint32_t *) tmp64_d;
> +	tmp32_s =3D (uint32_t *) tmp64_s;
> +	while (n >=3D 4) {
> +		*tmp32_d =3D *tmp32_s;
> +		tmp32_d++;
> +		tmp32_s++;
> +		n -=3D 4;
> +	}
> +
> +	tmp16_d =3D (uint16_t *) tmp32_d;
> +	tmp16_s =3D (uint16_t *) tmp32_s;
> +	while (n >=3D 2) {
> +		*tmp16_d =3D *tmp16_s;
> +		tmp16_d++;
> +		tmp16_s++;
> +		n -=3D 2;
> +	}
> +
> +	tmp8_d =3D (uint8_t *) tmp16_d;
> +	tmp8_s =3D (uint8_t *) tmp16_s;
> +	while (n >=3D 1) {
> +		*tmp8_d =3D *tmp8_s;
> +		tmp8_d++;
> +		tmp8_s++;
> +		n--;
> +	}
> +
> +	return dest;
> +}
> +EXPORT_SYMBOL(memcpy);
> +
> diff --git a/arch/kvx/lib/memset.S b/arch/kvx/lib/memset.S
> new file mode 100644
> index 000000000000..9eebc28da2be
> --- /dev/null
> +++ b/arch/kvx/lib/memset.S
> @@ -0,0 +1,351 @@
> +/* SPDX-License-Identifier: GPL-2.0-only */
> +/*
> + * Copyright (C) 2017-2023 Kalray Inc.
> + * Author(s): Clement Leger
> + *            Marius Gligor
> + */
> +
> +#include <linux/linkage.h>
> +
> +#include <asm/cache.h>
> +
> +#define REPLICATE_BYTE_MASK	0x0101010101010101
> +#define MIN_SIZE_FOR_ALIGN	128
> +
> +/*
> + * Optimized memset for kvx architecture
> + *
> + * In order to optimize memset on kvx, we can use various things:
> + * - conditionnal store which avoid branch penalty
> + * - store half/word/double/quad/octuple to store up to 16 bytes at a ti=
me
> + * - dzerol to zero a cacheline when the pattern is '0' (often the case)
> + * - hardware loop for steady cases.
> + *
> + * First, we assume that memset is mainly used for zeroing areas. In ord=
er
> + * to optimize this case, we consider it to be the fast path of the algo=
rithm.
> + * In both cases (0 and non 0 pattern), we start by checking if the size=
 is
> + * below a minimum size. If so, we skip the alignment part. Indeed, the =
kvx
> + * supports misalignment and the penalty for letting it do unaligned acc=
esses is
> + * lower than trying to realigning us. So for small sizes, we don't even=
 bother
> + * to realign. Minor difference is that in the memset with 0, we skip af=
ter the
> + * dzerol loop since dzerol must be cache-line aligned (no misalignment =
of
> + * course).
> + * Regarding the non 0 pattern memset, we use sbmm to replicate the patt=
ern on
> + * all bits on a register in one call.
> + * Once alignment has been reached, we can do the hardware loop for both=
 cases(
> + * store octuple/dzerol) in order to optimize throughput. Care must be t=
aken to
> + * align hardware loops on at least 8 bytes for performances.
> + * Once the main loop has been done, we finish the copy by checking leng=
th to do
> + * the necessary calls to store remaining bytes.
> + *
> + * Pseudo code (applies for non 0 pattern):
> + *
> + * int memset(void *dest, char pattern, long length)
> + * {
> + * 	long dest_align =3D -((long) dest);
> + * 	long copy;
> + * 	long orig_dest =3D dest;
> + *
> + * 	uint64_t pattern =3D sbmm8(pattern, 0x0101010101010101);
> + * 	uint128_t pattern128 =3D pattern << 64 | pattern;
> + * 	uint256_t pattern128 =3D pattern128 << 128 | pattern128;
> + *
> + * 	// Keep only low bits
> + * 	dest_align &=3D 0x1F;
> + * 	length -=3D dest_align;
> + *
> + * 	// Byte align
> + * 	copy =3D align & (1 << 0);
> + * 	if (copy)
> + * 		*((u8 *) dest) =3D pattern;
> + * 	dest +=3D copy;
> + * 	// Half align
> + * 	copy =3D align & (1 << 1);
> + * 	if (copy)
> + * 		*((u16 *) dest) =3D pattern;
> + * 	dest +=3D copy;
> + * 	// Word align
> + * 	copy =3D align & (1 << 2);
> + * 	if (copy)
> + * 		*((u32 *) dest) =3D pattern;
> + * 	dest +=3D copy;
> + * 	// Double align
> + * 	copy =3D align & (1 << 3);
> + * 	if (copy)
> + * 		*((u64 *) dest) =3D pattern;
> + * 	dest +=3D copy;
> + * 	// Quad align
> + * 	copy =3D align & (1 << 4);
> + * 	if (copy)
> + * 		*((u128 *) dest) =3D pattern128;
> + * 	dest +=3D copy;
> + *
> + * 	// We are now aligned on 256 bits
> + * 	loop_octuple_count =3D size >> 5;
> + * 	for (i =3D 0; i < loop_octuple_count; i++) {
> + * 		*((u256 *) dest) =3D pattern256;
> + * 		dest +=3D 32;
> + * 	}
> + *
> + * 	if (length =3D=3D 0)
> + * 		return orig_dest;
> + *
> + * 	// Copy remaining part
> + * 	remain =3D length & (1 << 4);
> + * 	if (copy)
> + * 		*((u128 *) dest) =3D pattern128;
> + * 	dest +=3D remain;
> + * 	remain =3D length & (1 << 3);
> + * 	if (copy)
> + * 		*((u64 *) dest) =3D pattern;
> + * 	dest +=3D remain;
> + * 	remain =3D length & (1 << 2);
> + * 	if (copy)
> + * 		*((u32 *) dest) =3D pattern;
> + * 	dest +=3D remain;
> + * 	remain =3D length & (1 << 1);
> + * 	if (copy)
> + * 		*((u16 *) dest) =3D pattern;
> + * 	dest +=3D remain;
> + * 	remain =3D length & (1 << 0);
> + * 	if (copy)
> + * 		*((u8 *) dest) =3D pattern;
> + * 	dest +=3D remain;
> + *
> + * 	return orig_dest;
> + * }
> + */
> +
> +.text
> +.align 16
> +ENTRY(memset):
> +	make $r32 =3D 0
> +	make $r33 =3D 0
> +	/* Check if length < KVX_DCACHE_LINE_SIZE */
> +	compd.ltu $r7 =3D $r2, KVX_DCACHE_LINE_SIZE
> +	/* Jump to generic memset if pattern is !=3D 0 */
> +	cb.dnez $r1? memset_non_0_pattern
> +	;;
> +	/* Preserve return value */
> +	copyd $r3 =3D $r0
> +	/* Invert address to compute size to copy to be aligned on 32 bytes */
> +	negd $r5 =3D $r0
> +	/* Remaining bytes for 16 bytes store (for alignment on 64 bytes) */
> +	andd $r8 =3D $r2, (1 << 5)
> +	copyq $r34r35 =3D $r32, $r33
> +	/* Skip loopdo with dzerol if length < KVX_DCACHE_LINE_SIZE */
> +	cb.dnez $r7? .Ldzerol_done
> +	;;
> +	/* Compute the size that will be copied to align on 64 bytes boundary */
> +	andw $r6 =3D $r5, 0x3F
> +	/* Check if address is aligned on 64 bytes */
> +	andw $r9 =3D $r0, 0x3F
> +	/* Alignment */
> +	nop
> +	;;
> +	/* If address already aligned on 64 bytes, jump to dzerol loop */
> +	cb.deqz $r9? .Laligned_64
> +	/* Remove unaligned part from length */
> +	sbfd $r2 =3D $r6, $r2
> +	/* Check if we need to copy 1 byte */
> +	andw $r4 =3D $r5, (1 << 0)
> +	;;
> +	/* If we are not aligned, store byte */
> +	sb.dnez $r4? [$r0] =3D $r32
> +	/* Check if we need to copy 2 bytes */
> +	andw $r4 =3D $r5, (1 << 1)
> +	/* Add potentially copied part for next store offset */
> +	addd $r0 =3D $r0, $r4
> +	;;
> +	sh.dnez $r4? [$r0] =3D $r32
> +	/* Check if we need to copy 4 bytes */
> +	andw $r4 =3D $r5, (1 << 2)
> +	addd $r0 =3D $r0, $r4
> +	;;
> +	sw.dnez $r4? [$r0] =3D $r32
> +	/* Check if we need to copy 8 bytes */
> +	andw $r4 =3D $r5, (1 << 3)
> +	addd $r0 =3D $r0, $r4
> +	;;
> +	sd.dnez $r4? [$r0] =3D $r32
> +	/* Check if we need to copy 16 bytes */
> +	andw $r4 =3D $r5, (1 << 4)
> +	addd $r0 =3D $r0, $r4
> +	;;
> +	sq.dnez $r4? [$r0] =3D $r32r33
> +	/* Check if we need to copy 32 bytes */
> +	andw $r4 =3D $r5, (1 << 5)
> +	addd $r0 =3D $r0, $r4
> +	;;
> +	so.dnez $r4? [$r0] =3D $r32r33r34r35
> +	addd $r0 =3D $r0, $r4
> +	;;
> +.Laligned_64:
> +	/* Prepare amount of data for dzerol */
> +	srld $r10 =3D $r2, 6
> +	/* Size to be handled in loopdo */
> +	andd $r4 =3D $r2, ~0x3F
> +	make $r11 =3D 64
> +	cb.deqz $r2? .Lmemset_done
> +	;;
> +	/* Remaining bytes for 16 bytes store */
> +	andw $r8 =3D $r2, (1 << 5)
> +	/* Skip dzerol if there are not enough data for 64 bytes store */
> +	cb.deqz $r10? .Ldzerol_done
> +	/* Update length to copy */
> +	sbfd $r2 =3D $r4, $r2
> +	;;
> +	loopdo $r10, .Ldzerol_done
> +		;;
> +		so 0[$r0], $r32r33r34r35
> +		;;
> +		so 32[$r0], $r32r33r34r35
> +		addd $r0 =3D $r0, $r11
> +		;;
> +	.Ldzerol_done:
> +	/*
> +	 * Now that we have handled every aligned bytes using 'dzerol', we can
> +	 * handled the remainder of length using store by decrementing size
> +	 * We also exploit the fact we are aligned to simply check remaining
> +	 * size */
> +	so.dnez $r8? [$r0] =3D $r32r33r34r35
> +	addd $r0 =3D $r0, $r8
> +	/* Remaining bytes for 16 bytes store */
> +	andw $r8 =3D $r2, (1 << 4)
> +	cb.deqz $r2? .Lmemset_done
> +	;;
> +	sq.dnez $r8? [$r0] =3D $r32r33
> +	addd $r0 =3D $r0, $r8
> +	/* Remaining bytes for 8 bytes store */
> +	andw $r8 =3D $r2, (1 << 3)
> +	;;
> +	sd.dnez $r8? [$r0] =3D $r32
> +	addd $r0 =3D $r0, $r8
> +	/* Remaining bytes for 4 bytes store */
> +	andw $r8 =3D $r2, (1 << 2)
> +	;;
> +	sw.dnez $r8? [$r0] =3D $r32
> +	addd $r0 =3D $r0, $r8
> +	/* Remaining bytes for 2 bytes store */
> +	andw $r8 =3D $r2, (1 << 1)
> +	;;
> +	sh.dnez $r8? [$r0] =3D $r32
> +	addd $r0 =3D $r0, $r8
> +	;;
> +	sb.odd $r2? [$r0] =3D $r32
> +	/* Restore original value */
> +	copyd $r0 =3D $r3
> +	ret
> +	;;
> +
> +.align 16
> +memset_non_0_pattern:
> +	/* Preserve return value */
> +	copyd $r3 =3D $r0
> +	/* Replicate the first pattern byte on all bytes */
> +	sbmm8 $r32 =3D $r1, REPLICATE_BYTE_MASK
> +	/* Check if length < MIN_SIZE_FOR_ALIGN */
> +	compd.geu $r7 =3D $r2, MIN_SIZE_FOR_ALIGN
> +	/* Invert address to compute size to copy to be aligned on 32 bytes */
> +	negd $r5 =3D $r0
> +	;;
> +	/* Check if we are aligned on 32 bytes */
> +	andw $r9 =3D $r0, 0x1F
> +	/* Compute the size that will be copied to align on 32 bytes boundary */
> +	andw $r6 =3D $r5, 0x1F
> +	/*
> +	 * If size < MIN_SIZE_FOR_ALIGN bits, directly go to so, it will be done
> +	 * unaligned but that is still better that what we can do with sb
> +	 */
> +	cb.deqz $r7? .Laligned_32
> +	;;
> +	/* Remove unaligned part from length */
> +	sbfd $r2 =3D $r6, $r2
> +	/* If we are already aligned on 32 bytes, jump to main "so" loop */
> +	cb.deqz $r9? .Laligned_32
> +	/* Check if we need to copy 1 byte */
> +	andw $r4 =3D $r5, (1 << 0)
> +	;;
> +	/* If we are not aligned, store byte */
> +	sb.dnez $r4? [$r0] =3D $r32
> +	/* Check if we need to copy 2 bytes */
> +	andw $r4 =3D $r5, (1 << 1)
> +	/* Add potentially copied part for next store offset */
> +	addd $r0 =3D $r0, $r4
> +	;;
> +	sh.dnez $r4? [$r0] =3D $r32
> +	/* Check if we need to copy 4 bytes */
> +	andw $r4 =3D $r5, (1 << 2)
> +	addd $r0 =3D $r0, $r4
> +	;;
> +	sw.dnez $r4? [$r0] =3D $r32
> +	/* Check if we need to copy 8 bytes */
> +	andw $r4 =3D $r5, (1 << 3)
> +	addd $r0 =3D $r0, $r4
> +	/* Copy second part of pattern for sq */
> +	copyd $r33 =3D $r32
> +	;;
> +	sd.dnez $r4? [$r0] =3D $r32
> +	/* Check if we need to copy 16 bytes */
> +	andw $r4 =3D $r5, (1 << 4)
> +	addd $r0 =3D $r0, $r4
> +	;;
> +	sq.dnez $r4? [$r0] =3D $r32r33
> +	addd $r0 =3D $r0, $r4
> +	;;
> +.Laligned_32:
> +	/* Copy second part of pattern for sq */
> +	copyd $r33 =3D $r32
> +	/* Prepare amount of data for 32 bytes store */
> +	srld $r10 =3D $r2, 5
> +	nop
> +	nop
> +	;;
> +	copyq $r34r35 =3D $r32, $r33
> +	/* Remaining bytes for 16 bytes store */
> +	andw $r8 =3D $r2, (1 << 4)
> +	make $r11 =3D 32
> +	/* Check if there are enough data for 32 bytes store */
> +	cb.deqz $r10? .Laligned_32_done
> +	;;
> +	loopdo $r10, .Laligned_32_done
> +		;;
> +		so 0[$r0] =3D $r32r33r34r35
> +		addd $r0 =3D $r0, $r11
> +		;;
> +	.Laligned_32_done:
> +	/*
> +	 * Now that we have handled every aligned bytes using 'so', we can
> +	 * handled the remainder of length using store by decrementing size
> +	 * We also exploit the fact we are aligned to simply check remaining
> +	 * size */
> +	sq.dnez $r8? [$r0] =3D $r32r33
> +	addd $r0 =3D $r0, $r8
> +	/* Remaining bytes for 8 bytes store */
> +	andw $r8 =3D $r2, (1 << 3)
> +	cb.deqz $r2? .Lmemset_done
> +	;;
> +	sd.dnez $r8? [$r0] =3D $r32
> +	addd $r0 =3D $r0, $r8
> +	/* Remaining bytes for 4 bytes store */
> +	andw $r8 =3D $r2, (1 << 2)
> +	;;
> +	sw.dnez $r8? [$r0] =3D $r32
> +	addd $r0 =3D $r0, $r8
> +	/* Remaining bytes for 2 bytes store */
> +	andw $r8 =3D $r2, (1 << 1)
> +	;;
> +	sh.dnez $r8? [$r0] =3D $r32
> +	addd $r0 =3D $r0, $r8
> +	;;
> +	sb.odd $r2? [$r0] =3D $r32
> +	/* Restore original value */
> +	copyd $r0 =3D $r3
> +	ret
> +	;;
> +.Lmemset_done:
> +	/* Restore original value */
> +	copyd $r0 =3D $r3
> +	ret
> +	;;
> +ENDPROC(memset)
> diff --git a/arch/kvx/lib/strlen.S b/arch/kvx/lib/strlen.S
> new file mode 100644
> index 000000000000..8298402a7898
> --- /dev/null
> +++ b/arch/kvx/lib/strlen.S
> @@ -0,0 +1,122 @@
> +/* SPDX-License-Identifier: GPL-2.0-only */
> +/*
> + * Copyright (C) 2017-2023 Kalray Inc.
> + * Author(s): Jules Maselbas
> + */
> +#include <linux/linkage.h>
> +#include <asm/export.h>
> +
> +/*
> + *	kvx optimized strlen
> + *
> + *	This implementation of strlen only does aligned memory accesses.
> + *	Since we don't know the total length the idea is to do double word
> + *	load and stop on the first null byte found. As it's always safe to
> + *	read more up to a lower 8-bytes boundary.
> + *
> + *	This implementation of strlen uses a trick to detect if a double
> + *	word contains a null byte [1]:
> + *
> + *	> #define haszero(v) (((v) - 0x01010101UL) & ~(v) & 0x80808080UL)
> + *	> The sub-expression (v - 0x01010101UL), evaluates to a high bit set
> + *	> in any byte whenever the corresponding byte in v is zero or greater
> + *	> than 0x80. The sub-expression ~v & 0x80808080UL evaluates to high
> + *	> bits set in bytes where the byte of v doesn't have its high bit set
> + *	> (so the byte was less than 0x80). Finally, by ANDing these two sub-
> + *	> expressions the result is the high bits set where the bytes in v
> + *	> were zero, since the high bits set due to a value greater than 0x80
> + *	> in the first sub-expression are masked off by the second.
> + *
> + *	[1] http://graphics.stanford.edu/~seander/bithacks.html#ZeroInWord
> + *
> + *	A second trick is used to get the exact number of characters before
> + *	the first null byte in a double word:
> + *
> + *		clz(sbmmt(zero, 0x0102040810204080))
> + *
> + *	This trick uses the haszero result which maps null byte to 0x80 and
> + *	others value to 0x00. The idea is to count the number of consecutive
> + *	null byte in the double word (counting from less significant byte
> + *	to most significant byte). To do so, using the bit matrix transpose
> + *	will "pack" all high bit (0x80) to the most significant byte (MSB).
> + *	It is not possible to count the trailing zeros in this MSB, however
> + *	if a byte swap is done before the bit matrix transpose we still have
> + *	all the information in the MSB but now we can count the leading zeros.
> + *	The instruction sbmmt with the matrix 0x0102040810204080 does exactly
> + *	what we need a byte swap followed by a bit transpose.
> + *
> + *	A last trick is used to handle the first double word misalignment.
> + *	This is done by masking off the N lower bytes (excess read) with N
> + *	between 0 and 7. The mask is applied on haszero results and will
> + *	force the N lower bytes to be considered not null.
> + *
> + *	This is a C implementation of the algorithm described above:
> + *
> + *	size_t strlen(char *s) {
> + *		uint64_t *p    =3D (uint64_t *)((uintptr_t)s) & ~0x7;
> + *		uint64_t rem   =3D ((uintptr_t)s) % 8;
> + *		uint64_t low   =3D -0x0101010101010101;
> + *		uint64_t high  =3D  0x8080808080808080;
> + *		uint64_t dword, zero;
> + *		uint64_t msk, len;
> + *
> + *		dword =3D *p++;
> + *		zero  =3D (dword + low) & ~dword & high;
> + *		msk   =3D 0xffffffffffffffff << (rem * 8);
> + *		zero &=3D msk;
> + *
> + *		while (!zero) {
> + *			dword =3D *p++;
> + *			zero  =3D (dword + low) & ~dword & high;
> + *		}
> + *
> + *		zero =3D __builtin_kvx_sbmmt8(zero, 0x0102040810204080);
> + *		len =3D ((void *)p - (void *)s) - 8;
> + *		len +=3D __builtin_kvx_clzd(zero);
> + *
> + *		return len;
> + *	}
> + */
> +
> +.text
> +.align 16
> +ENTRY(strlen)
> +	andd  $r1 =3D $r0, ~0x7
> +	andd  $r2 =3D $r0,  0x7
> +	make $r10 =3D -0x0101010101010101
> +	make $r11 =3D  0x8080808080808080
> +	;;
> +	ld $r4 =3D 0[$r1]
> +	sllw $r2 =3D $r2, 3
> +	make $r3 =3D 0xffffffffffffffff
> +	;;
> +	slld $r2 =3D $r3, $r2
> +	addd $r5 =3D $r4, $r10
> +	andnd $r6 =3D $r4, $r11
> +	;;
> +	andd $r6 =3D $r6, $r2
> +	make $r3 =3D 0
> +	;;
> +.loop:
> +	andd $r4 =3D $r5, $r6
> +	addd $r1 =3D $r1, 0x8
> +	;;
> +	cb.dnez $r4? .end
> +	ld.deqz $r4? $r4 =3D [$r1]
> +	;;
> +	addd $r5 =3D $r4, $r10
> +	andnd $r6 =3D $r4, $r11
> +	goto .loop
> +	;;
> +.end:
> +	addd $r1 =3D $r1, -0x8
> +	sbmmt8 $r4 =3D $r4, 0x0102040810204080
> +	;;
> +	clzd $r4 =3D $r4
> +	sbfd $r1 =3D $r0, $r1
> +	;;
> +	addd $r0 =3D $r4, $r1
> +	ret
> +	;;
> +ENDPROC(strlen)
> +EXPORT_SYMBOL(strlen)
> diff --git a/arch/kvx/lib/usercopy.S b/arch/kvx/lib/usercopy.S
> new file mode 100644
> index 000000000000..bc7e1a45e1c7
> --- /dev/null
> +++ b/arch/kvx/lib/usercopy.S
> @@ -0,0 +1,90 @@
> +/* SPDX-License-Identifier: GPL-2.0-only */
> +/*
> + * Copyright (C) 2017-2023 Kalray Inc.
> + * Author(s): Clement Leger
> + */
> +#include <linux/linkage.h>
> +
> +/**
> + * Copy from/to a user buffer
> + * r0 =3D to buffer
> + * r1 =3D from buffer
> + * r2 =3D size to copy
> + * This function can trapped when hitting a non-mapped page.
> + * It will trigger a trap NOMAPPING and the trap handler will interpret
> + * it and check if instruction pointer is inside __ex_table.
> + * The next step are described later !
> + */
> +.text
> +ENTRY(raw_copy_from_user)
> +ENTRY(raw_copy_to_user)
> +	/**
> +	 * naive implementation byte per byte
> +	 */
> +	make $r33 =3D 0x0;
> +	/* If size =3D=3D 0, exit directly */
> +	cb.deqz $r2? copy_exit
> +	;;
> +	loopdo $r2, copy_exit
> +		;;
> +0:		lbz $r34 =3D $r33[$r1]
> +		;;
> +1:		sb $r33[$r0] =3D $r34
> +		addd $r33 =3D $r33, 1 /* Ptr increment */
> +		addd $r2 =3D $r2, -1 /* Remaining bytes to copy */
> +		;;
> +	copy_exit:
> +	copyd $r0 =3D $r2
> +	ret
> +	;;
> +ENDPROC(raw_copy_to_user)
> +ENDPROC(raw_copy_from_user)
> +
> +/**
> + * Exception table
> + * each entry correspond to the following:
> + * .dword trapping_addr, restore_addr
> + *
> + * On trap, the handler will try to locate if $spc is matching a
> + * trapping address in the exception table. If so, the restore addr
> + * will  be put in the return address of the trap handler, allowing
> + * to properly finish the copy and return only the bytes copied/cleared
> + */
> +.pushsection __ex_table,"a"
> +.balign 8
> +.dword 0b, copy_exit
> +.dword 1b, copy_exit
> +.popsection
> +
> +/**
> + * Clear a user buffer
> + * r0 =3D buffer to clear
> + * r1 =3D size to clear
> + */
> +.text
> +ENTRY(asm_clear_user)
> +	/**
> +	 * naive implementation byte per byte
> +	 */
> +	make $r33 =3D 0x0;
> +	make $r34 =3D 0x0;
> +	/* If size =3D=3D 0, exit directly */
> +	cb.deqz $r1? clear_exit
> +	;;
> +	loopdo $r1, clear_exit
> +		;;
> +40:		sb $r33[$r0] =3D $r34
> +		addd $r33 =3D $r33, 1 /* Ptr increment */
> +		addd $r1 =3D $r1, -1 /* Remaining bytes to copy */
> +		;;
> +	clear_exit:
> +	copyd $r0 =3D $r1
> +	ret
> +	;;
> +ENDPROC(asm_clear_user)
> +
> +.pushsection __ex_table,"a"
> +.balign 8
> +.dword 40b, clear_exit
> +.popsection
> +


--=20
Cl=C3=A9ment L=C3=A9ger,
Embedded Linux and Kernel engineer at Bootlin
https://bootlin.com